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
|
|---|---|---|---|---|---|
As Zika virus continues to spread , decisions regarding resource allocations to control the outbreak underscore the need for a tool to weigh policies according to their cost and the health burden they could avert . For example , to combat the current Zika outbreak the US President requested the allocation of $1 . 8 billion from Congress in February 2016 . Illustrated through an interactive tool , we evaluated how the number of Zika cases averted , the period during pregnancy in which Zika infection poses a risk of microcephaly , and probabilities of microcephaly and Guillain-Barré Syndrome ( GBS ) impact the cost at which an intervention is cost-effective . From Northeast Brazilian microcephaly incidence data , we estimated the probability of microcephaly in infants born to Zika-infected women ( 0 . 49% to 2 . 10% ) . We also estimated the probability of GBS arising from Zika infections in Brazil ( 0 . 02% to 0 . 06% ) and Colombia ( 0 . 08% ) . We calculated that each microcephaly and GBS case incurs the loss of 29 . 95 DALYs and 1 . 25 DALYs per case , as well as direct medical costs for Latin America and the Caribbean of $91 , 102 and $28 , 818 , respectively . We demonstrated the utility of our cost-effectiveness tool with examples evaluating funding commitments by Costa Rica and Brazil , the US presidential proposal , and the novel approach of genetically modified mosquitoes . Our analyses indicate that the commitments and the proposal are likely to be cost-effective , whereas the cost-effectiveness of genetically modified mosquitoes depends on the country of implementation . Current estimates from our tool suggest that the health burden from microcephaly and GBS warrants substantial expenditures focused on Zika virus control . Our results justify the funding committed in Costa Rica and Brazil and many aspects of the budget outlined in the US president’s proposal . As data continue to be collected , new parameter estimates can be customized in real-time within our user-friendly tool to provide updated estimates on cost-effectiveness of interventions and inform policy decisions in country-specific settings .
In April 2015 , the first confirmed case of mosquito-borne Zika virus in the Americas was reported in Brazil [1] . Since then , the virus has spread to 41 countries and territories across the Americas , Oceania , the Pacific Islands , and Africa [2] , with over 1 . 5 million suspected and confirmed cases [1] . In the US , sexually transmitted or travel associated cases have been reported in 40 States and the District of Columbia . Furthermore , transmission has been reported in the Commonwealth of Puerto Rico , the Virgin Islands of the US , and the Territory of American Samoa [3] . There are projections of millions more cases in both the countries Zika has already reached and others within which it has yet to emerge [1] , including predictions of local transmission in the Gulf Coast of the US [4] . There is strong scientific consensus that Zika virus can cause Guillain–Barré syndrome ( GBS ) [5] and that a Zika infection in pregnant women can cause microcephaly in their infants [1 , 6 , 7] , vision-threatening ocular lesions [8] , in utero growth restriction , fetal deaths , stillbirths , and central nervous system lesions [9] . On February 1 , 2016 , the World Health Organization ( WHO ) declared the outbreak an International Health Emergency [10] . Without vaccines or medication to treat Zika infections , vector control remains the only immediate option to combat this outbreak . However , there are hesitations regarding whether extensive efforts likely necessary to contain Zika , such as intensive scale-up of vector control , the application of insecticides , or the implementation of new technologies that could include genetically modified or Wolbachia-infected mosquitoes , would be worth the substantial costs , logistical challenges , and potential environmental repercussions [11] . For example , genetically engineered male Aedes aegypti , the offspring of whom die prior to maturity , are being piloted against the ongoing Zika outbreak . This approach is relatively expensive for middle income countries , requiring $1 . 9 million in the first year and $384 , 000 each year thereafter for an urban population of 50 , 000 [12] . The investment in Zika control should be considered relative to the disease burden that could be averted and the resources available in the country . For example , the Brazil National Development Bank has allocated $136 . 6 million [13] towards combating mosquito-borne diseases . The Costa Rican Department of Social Security has committed $745 , 724 for community-led elimination of breeding sites , through removal of containers of stagnant water in high-risk regions , and the Ministry of Health is planning on allocating $373 , 712 exclusively to the control of Zika . Considering the populations in Brazil and Costa Rica , these investments represent only about $0 . 66 and $0 . 23 per citizen , respectively . Moreover , while Brazil and Costa Rica have similar per capita income , the same investment would be valued differently in countries with different resource constraints . Cost-effectiveness analysis provides a framework under which such investments can be studied for each particular case . As with any global pandemic , international effort to control the outbreak should be led by agencies such as the WHO and by countries with the resources and expertise necessary to confront the threat . To combat the current Zika outbreak , the US President requested from Congress in February the allocation of $1 . 8 billion [14] . Of this total , $250 million has been allocated to the Commonwealth of Puerto Rico for the prevention of Zika infection in pregnant women and for medical costs associated with Zika . To curtail the Zika outbreak internationally , the United States Agency for International Development would receive $335 million and the State Department $41 million to address Zika in Latin America and the Caribbean . The remaining requested funds would be directed towards outbreak management in the US , expanded vector control measures , and vaccine development . Fundamental to these investment decisions is the quantification of the costs , including any environmental risks , of an intervention balanced by its value in terms of the health burden that it would likely avert . We offer quantitative insight into the health burden that an unchecked Zika epidemic might incur and provide an interactive web application ( http://zika . cidma . us/ ) that can be employed by health authorities to evaluate the cost-effectiveness of programs under consideration for Zika control . This study aims to evaluate the cost-effectiveness of expenditures towards Zika control intervention , based on available data . To support these calculations , we also estimated the probability of microcephaly arising from a Zika-infected pregnancy , parameterized with case data from Northeast Brazil , and the probability of GBS following a Zika infection , using Brazilian and Colombian case data . We further estimated the number and burden of cases of Zika-linked microcephaly and GBS expected to occur across the Americas in 2016 , should the Zika epidemic remain unabated . We also provide thresholds of price and effectiveness combinations for interventions that would satisfy WHO criteria for cost-effectiveness . Additionally , our interactive web tool has the flexibility to be updated as more information becomes available on this emerging disease .
We estimated the probability of a microcephaly case given a Zika infection during the first trimester of pregnancy , as the proportion of the Zika-related microcephaly births among all births to mothers infected with Zika during the first trimester of pregnancy [15] . Our web tool also has the flexibility to adjust the duration of risk during pregnancy , as reports emerge suggesting that the risk could extend beyond early pregnancy . Birth counts are totaled for the outbreak in Northeast Brazil , where Zika was first confirmed to have reached Latin America in mid-April 2015 [1 , 16] and where cases have already begun to decline [17] . Given our assumption that only first-trimester Zika infections can cause microcephaly , we expect Zika-liked microcephaly to begin arising around October 15 , 2015 , six months after the first infection . Cumulative case data on suspected and confirmed microcephaly cases were obtained for Northeast Brazil until April 2 , 2016 , with the first report containing cumulative case counts from November 15 , 2015 [17] . To forecast the microcephaly cases yet to arise , we applied a linear regression over the weekly reported cases for 2016 in Northeast Brazil [17] . Excess microcephaly cases , above what would be expected for Northeast Brazil during this time period , were assumed to be linked to Zika infection . This excess is calculated as the difference between the estimated total microcephaly cases during the outbreak in Northeast Brazil and the expected non-Zika related microcephaly births for the same duration and region . To estimate the total microcephaly births for the Northeast Brazil outbreak , we took into account the reporting sensitivity , i . e . , the proportion of confirmed cases from among all investigated microcephaly cases in the region [17] and the expected total reported microcephaly births . The estimated total reported microcephaly births is then the sum of our forecast of newly reported microcephaly cases for the remaining outbreak in Northeast Brazil and the latest available reported microcephaly cases for Northeast Brazil ( 5 , 380 as of April 2 , 2016 ) . The expected microcephaly cases attributable to causes other than Zika were based on prevalence estimates of microcephaly prior to the outbreak from Brazil ( 0 . 5 per 10 , 000 births ) [18] and the highest reported prevalence from the US ( 12 per 10 , 000 births ) [19] to account for possible underreporting before the outbreak . These prevalence values were multiplied by expected births during the Zika-related microcephaly outbreak in Northeast Brazil to estimate expected microcephaly from other causes . The expected births during the outbreak in Northeast Brazil were estimated from the fraction of the Brazilian population in Northeast Brazil ( 28% ) [20] , the Brazilian population [21] , and the Brazilian birth rate ( 14 . 931 per 1000 population ) [22] . Since the attack rate in Northeast Brazil is unknown , to estimate the births by mothers infected in their first trimester , we used attack rates ranging from 23 . 5% ( reported for the latest outbreak in Puerto Rico of chikungunya [23] , a related arbovirus transmitted by the same mosquito species ) to 77% ( upper bound for Zika outbreak in Yap Island , Micronesia in 2007 [24] ) . The 2013–2014 French Polynesia Zika outbreak had an attack rate between those two bounds ( 66% [95%CI: 62–70] [15] ) . The supplement provides a table of the parameters used and their sources , as well as a flow diagram of the equations underlying our calculations of the probability of microcephaly given a Zika infection during the first trimester of pregnancy ( S1 Table and S1 Fig ) . The probability of microcephaly given Zika infection , in the absence of interventions targeted towards pregnant women , is estimated by multiplying the probability of microcephaly given a Zika infection in the first trimester of pregnancy , the geographic-specific birth rate , and the risk period divided by 365 days . We estimated the probability of developing GBS given a Zika infection as the proportion of Zika-related GBS cases from among all Zika infections . Parameterization of our calculations are based on data from a 10-week period in Colombia ( 45 , 314 Zika cases and 231 GBS cases , from January 9th to March 19 , 2016 [1 , 25] ) , adjusting for estimates suggesting that 82% of Zika infections are asymptomatic [24] , and on the average annual GBS incidence for Colombia from 2008 to 2014 ( 242 ) [26] . Similarly , we recalculated the probability of developing GBS using the WHO estimate of 269 Zika-related GBS cases in Brazil [1] using the estimated epidemic size ranging from 443 , 502 to 1 , 301 , 140 for Brazil [27] . The lower and upper limits of this estimate are based on suspected cases of dengue fever for which dengue was ultimately excluded and on the attack rate of Zika virus during the 2013 French Polynesia Zika outbreak , respectively [27] . The health impact of microcephaly and GBS was calculated in disability-adjusted life-years ( DALY ) . We estimated the health impact of a single case of microcephaly in present-value terms ( discounting at 3% annually ) , assuming a 79 . 7% probability of survival through the first year of life [28] , an optimistic life expectancy of 35 years given survival beyond that year , and the 0 . 16 disability weight assigned to living with severe intellectual disability [29] . In the absence of available medical costs for microcephaly , we conservatively used the lifetime direct medical cost associated with a case of mental retardation , estimated as $179 , 760 in lifetime expenses for the US [30 , 31] . In estimating the health impact of a single case of GBS , we conservatively assumed a 5% probability of death [1] , a 9% probability of severe motor impairment ( 0 . 402 disability score ) for 6 months , and an 84% probability of moderate generalized musculoskeletal problems ( 0 . 344 disability score ) for three weeks [29 , 32] . The average age of a case in the Colombian outbreak was 43 years [1] , and we assumed this was also the average age of death ( if death occurred ) . The direct medical cost per case of GBS was $56 , 863 for the US [30 , 33] . The medical costs for both microcephaly and GBS were updated to 2015 USD using the Consumer Price Index Inflation Calculator [30] , and then converted to location-specific costs for each country or region using the World Bank purchasing power indices for medical expenses [34] . Our base case estimates for the costs associated with microcephaly and GBS are highly conservative in that they do not incorporate reduced productivity and quality of life , as well as other indirect costs associated with the conditions , such as educational and support services for microcephaly . To conduct analyses that account for these costs , our interactive tool allows the user to vary the per case cost associated with microcephaly and GBS . We estimated the number of microcephaly and GBS cases averted as the product of Zika infections averted and the probabilities of microcephaly or GBS per Zika infection , respectively . The health burden was estimated as the product of the cases averted and the DALYs lost per case , for both microcephaly and GBS . Similarly , the economic burden was estimated as the product of the cases averted and the direct medical expense for each condition . A net health benefit framework [35] combines health outcomes ( here , in DALYs ) , economic costs , and a willingness-to-pay for DALYs to establish the value of an intervention in a particular setting . The net health benefit of a particular strategy is calculated as the DALYs it averts minus its net cost as a proportion of the willingness to pay threshold . The WHO has established two willingness-to-pay thresholds at the country level: the per-capita GDP or three times the per-capita GDP for interventions to be considered “very cost-effective” or “cost-effective , ” respectively [36] . A positive net health benefit calculation at these willingness-to-pay thresholds indicates that the intervention fulfills the criteria for cost-effectiveness . This analysis is conducted from a government perspective , given our focus on the cost of intervention and direct healthcare costs . We applied the net health benefits framework to both WHO thresholds at which investment would be deemed cost-effective and very cost-effective , respectively , to evaluate: Our interactive web tool was coded in Python ( www . python . org ) using the NumPy package for scientific computing ( http://www . numpy . org/ ) and the Bokeh interactive visualization library ( http://bokeh . pydata . org/ ) .
We estimated that an additional 94 microcephaly cases are likely to occur in Northeast Brazil between the last available report on April 2 , 2016 through the end of April , after which the microcephaly outbreak is projected to dwindle in this region . We estimated the probability of a microcephaly case given a Zika infection during the first trimester to range between 0 . 49% and 2 . 10% . This probability is highly sensitive to the final attack rate for Zika in Northeast Brazil ( Fig 1 ) . Since the reported probability of 0 . 95% for the recent outbreak in French Polynesia [15] falls within our estimated range , we used this value in our base case parameter set . In our most conservative and less conservative scenarios we used 0 . 49% and 2 . 10% , respectively . We also estimated the probability of Zika-related GBS . Specifically , from the Brazilian data , we calculated that the probability of GBS given a Zika infection ranged from 0 . 02% to 0 . 06% , consistent with a recent estimate of 0 . 024% for the French Polynesia outbreak [5] . From the Colombian data our probability estimate was 0 . 08% . We used 0 . 06% in our base case parameter set , and 0 . 02% and 0 . 08% for our most conservative and less conservative scenarios , respectively . The final number of microcephaly and GBS cases predicted to occur throughout Latin America and the Caribbean depends on the attack rate for the region ( Fig 2 ) . For example , an attack rate of 5% would lead to a predicted 665 to 2843 microcephaly cases , 6 , 474 to 25 , 494 GBS cases , and a loss of 43 , 717 to 108 , 951 DALYs , whereas an attack rate of 40% would elevate those predictions to between 5 , 320 and 22 , 743 microcephaly cases , 51 , 790 to 203 , 951 GBS cases , and a loss of 349 , 738 to 871 , 610 DALYs . To put this potential health burden in context , the total dengue fever annual health burden is estimated as 89 , 500 DALYs in Latin America and the Caribbean [37] . We estimated that each microcephaly case conservatively represents the loss of 29 . 95 DALYs and a direct medical cost of $91 , 102 per lifetime . Similarly , for GBS we estimated a health burden of 1 . 25 DALYs per case , as well as a direct medical cost per case of $28 , 818 . The total DALYs likely to be averted by an intervention in a specific setting is influenced by demography ( birth rate and population size ) , and the criteria for cost-effectiveness depend on the per-capita GDP ( Fig 3 ) . For example , our base case analysis suggests that an intervention that prevents 10 , 000 , 000 Zika infections across Latin America and the Caribbean , corresponding to 1 . 6% of the population , would be cost-effective for expenditures below $802 million and very cost-effective below $409 million , using the average regional per capita GDP and birth rate . Preventing infections in only 1 . 6% of the population is likely achievable since previous attack rates of Zika virus in Yap Island and French Polynesia have been estimated to range between 62% and 77% [15 , 24] . Thus , the $376 million for foreign aid in Latin America and the Caribbean proposed by the US President to combat Zika would likely be a very cost-effective investment . These estimations conservatively assumed that the infections averted would be uniformly distributed among the entire population . Interventions that focus on preventing infection among pregnant women would be cost-effective for greater expenditure . We calculated that 13 , 490 pregnant women are at risk of Zika infection if the outbreak is unabated in Puerto Rico , assuming the same attack rate as French Polynesia and the estimated duration of the microcephaly outbreak for Northeast Brazil . If an intervention is able to avert 90% of those infections in pregnant women , it would be cost-effective at $195 . 4 million . This calculation suggests that the expenditure of $250 million for Puerto Rico is justified , given that the funding is allocated not only to the prevention of infection in pregnant women , but also medical costs associated with Zika cases . Conversely , using the same framework we can also calculate the minimal number of Zika infections that an intervention with a fixed cost would have to avert to be deemed cost-effective . For example , the $745 , 724 that the Costa Rican Department of Social Security has allocated for community-led elimination of breeding sites would be cost-effective if it averted 8 , 321 Zika infections . The campaign would be considered very cost-effective if it averted at least 14 , 386 Zika infections . The additional $373 , 712 investment that the Costa Rican Ministry of Health plans to dedicate to the control of Zika needs to avert at least 4 , 170 Zika infections in the country to be considered cost-effective and at least 7 , 210 to be considered very cost-effective . In Brazil , the $136 . 6 million that its National Development Bank has committed , if dedicated exclusively to prevent Zika , would be cost-effective if it averts 1 , 640 , 934 infections and highly cost-effective if it averts 3 , 245 , 553 infections . These analyses are conservative given that the same mosquito vector also transmits dengue , yellow fever and chikungunya . The cost-effectiveness of releasing genetically modified male mosquitoes whose offspring die before reaching adulthood depends on the expected Zika attack rate , the effectiveness of prevention , and the country in which the intervention is implemented . For example , in a Panamanian city with a population of 50 , 000 ( e . g . Santiago de Veraguas ) we estimated that the three-year implementation of this technology was cost-effective if it prevented 27 , 356 infections . The same intervention would not be not cost-effective in similarly populous cities located in countries with lower per-capita GDP ( e . g . , El Salvador or Nicaragua ) or with lower birth rates ( e . g . , Cuba or Thailand ) , because the number of infections that must be prevented within the city would be greater than the entire population of the city . We developed an interactive web tool for the evaluation of interventions beyond those illustrated here ( http://zika . cidma . us/ ) . Our tool allows policy makers to compute the cost-effective expenditure for an intervention that prevents a given number of Zika infections , or the number of infections which must be averted to justify an intervention cost . The user can vary the country or region of interest , which automatically adjusts the GDP , the population size , and the birth rate to generate setting-specific thresholds . Two key parameters that impact the DALY burden are the period of gestation during which a fetus would be at risk of Zika infection if the mother is infected , as well as the probability that a Zika infection of a pregnant woman within that period leads to microcephaly ( Fig 4 ) . The user can interactively modify those two parameters , as well as the probability of a GBS case given a Zika infection . For the analysis of programs with an emphasis on pregnant women , the user can specify a percentage of intervention effort directed specifically to pregnant women . To incorporate indirect medical costs and other expenses , our tool allows the user to specify the per-case cost of microcephaly and GBS . The output of our tool includes the combinations of cost and Zika infections averted for which intervention expenditure would be deemed as very cost-effective , cost-effective , or neither . Hover text displays exact values for the intervention cost , DALYs averted , and net health benefit . As information arises , our web-based tool can be adjusted to provide real-time projections of the health burden of the Zika outbreak in different countries .
Using data-driven analyses , we estimated the health and economic burden of an unchecked Zika epidemic and evaluate the conditions under which proposed interventions would be cost-effective . To parameterize our analyses , we calculated the probability that Zika infection leads to microcephaly or GBS , respectively . We developed a web tool to assist policymakers in their assessment of country-specific control measures . Our results provide conservative estimation both of the burden of Zika , and of the expenditure justified for an intervention . Congenital Zika infection has been associated with a number of conditions beyond microcephaly , including vision-threatening ocular lesions [8] , in utero growth restriction , fetal death , stillbirth , and central nervous system lesions [6 , 9] . Additionally , given the severity of Zika-related cases of microcephaly [38] , the life expectancy of 35 years is likely highly conservative . If the average life expectancy for these microcephaly cases is lower , interventions aimed at curtailing the ongoing Zika outbreak will avert a greater number of DALYs , since each year of life lost has a higher burden than a year lived with disability . As more data become available , the health burden associated with a single Zika infection should be updated . For the results presented here , we did not consider the indirect costs related to GBS or microcephaly . Our tool allows the user to adjust the per-case cost for microcephaly and GBS , giving the user the option to include indirect costs if or when they are available for a specific country . Indirect costs can be important to assessments of cost-effectiveness from the societal perspective , particularly given that the economic losses for caregivers of children with microcephaly may be substantial . For example , in Puerto Rico we estimated total direct medical costs for microcephaly and GBS of $104 million in our base case and from $39 million in our most conservative scenario up to $159 million in our less conservative scenario . However , upon inclusion of both direct non-medical costs and productivity losses [30 , 31 , 33] , the values rise to $736 million in the base case , with $280 million in the most conservative scenario and as high as $1 . 13 billion in the less conservative scenario . Even these estimates do not include other indirect costs such as specialized child-care support , parental productivity losses , or psychological toll of families with children with microcephaly , which are all substantial , yet difficult to quantify . Our analyses are further conservative in their exclusion of the impact on other diseases that could be achieved by interventions that target the Ae . aegypti vector , also responsible for transmitting dengue , chikungunya , and yellow fever . For example , Costa Rica has reported over 20 , 000 dengue fever cases annually , and over 5 , 000 chikungunya cases since the disease appeared in 2014 [39] . If the $376 million proposed by the US President for foreign aid targeted at the Zika outbreak can avert infection among as little as 0 . 7% of the population of Latin America and the Caribbean , our analyses indicate that the intervention would be cost-effective . Averting this number of cases is highly feasible given that prior Zika outbreaks have exhibited attack rates ranging from 62% to 77% [15 , 24] . While vector control is the most immediately available tool for mitigating the Zika burden , the development of an efficacious vaccine would be a more sustainable long-term strategy . The budget requested by the US President to the Congress includes a provision of $200 million for research and development of a Zika virus vaccine . Since a successful vaccine could be used globally to prevent millions of Zika cases over many years , such an investment is more than justified . Our tool identifies the combinations of price and effectiveness for which an intervention would be deemed cost-effective , but it does not predict the number of cases which an intervention will prevent , nor does it predict the net cost of an intervention . As mechanistic models describing Zika transmission and predicting the impact of interventions are developed , they can be integrated within the decision space delineated in this work , and by our web tool . Our interactive application ( http://zika . cidma . us/ ) provides a flexible tool for informing public health policy via a rigorous cost-benefit analysis of available options . While our examples focus on vector-control interventions , our framework would also be applicable to investments in vaccines or therapeutics . Difficult decisions related to next steps confront community members and leaders of areas that are currently facing , or will soon be facing , an epidemic of Zika . Given the potentially high health burden of Zika , the cost of inaction–or even insufficient action–may warrant significant expenditure .
|
Using data on Zika virus , microcephaly , and Guillain-Barré syndrome cases from Brazil and Colombia , we compute ranges for the probability of a microcephaly outcome in infants born to Zika-infected women ( 0 . 49% to 2 . 10% , based on data from Northeast Brazil ) and the probability of Guillain-Barré syndrome in Zika-infected individuals ( 0 . 02% to 0 . 06% in Brazil and 0 . 08% in Colombia ) . These results have allowed us to create a web-based cost-effectiveness tool that quantifies the relationship between the cost of an intervention and the number of Zika virus cases , as well as loss of disability-adjusted life years , that it can avert . Our tool thus identifies the threshold at which a given intervention , such as vector control , may be deemed cost-effective or very cost-effective in a variety of settings according to WHO criteria , in terms of the Zika burden that could be averted and the cost of such an intervention .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"microcephaly",
"pathology",
"and",
"laboratory",
"medicine",
"cost-effectiveness",
"analysis",
"chikungunya",
"infection",
"demography",
"economic",
"analysis",
"pathogens",
"geographical",
"locations",
"microbiology",
"social",
"sciences",
"tropical",
"diseases",
"health",
"care",
"viruses",
"developmental",
"biology",
"rna",
"viruses",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"infectious",
"disease",
"control",
"morphogenesis",
"infectious",
"diseases",
"health",
"economics",
"south",
"america",
"medical",
"microbiology",
"birth",
"defects",
"microbial",
"pathogens",
"congenital",
"disorders",
"economics",
"brazil",
"people",
"and",
"places",
"birth",
"rates",
"population",
"metrics",
"flaviviruses",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms",
"zika",
"virus"
] |
2016
|
A Cost-Effectiveness Tool for Informing Policies on Zika Virus Control
|
Correlations in spike-train ensembles can seriously impair the encoding of information by their spatio-temporal structure . An inevitable source of correlation in finite neural networks is common presynaptic input to pairs of neurons . Recent studies demonstrate that spike correlations in recurrent neural networks are considerably smaller than expected based on the amount of shared presynaptic input . Here , we explain this observation by means of a linear network model and simulations of networks of leaky integrate-and-fire neurons . We show that inhibitory feedback efficiently suppresses pairwise correlations and , hence , population-rate fluctuations , thereby assigning inhibitory neurons the new role of active decorrelation . We quantify this decorrelation by comparing the responses of the intact recurrent network ( feedback system ) and systems where the statistics of the feedback channel is perturbed ( feedforward system ) . Manipulations of the feedback statistics can lead to a significant increase in the power and coherence of the population response . In particular , neglecting correlations within the ensemble of feedback channels or between the external stimulus and the feedback amplifies population-rate fluctuations by orders of magnitude . The fluctuation suppression in homogeneous inhibitory networks is explained by a negative feedback loop in the one-dimensional dynamics of the compound activity . Similarly , a change of coordinates exposes an effective negative feedback loop in the compound dynamics of stable excitatory-inhibitory networks . The suppression of input correlations in finite networks is explained by the population averaged correlations in the linear network model: In purely inhibitory networks , shared-input correlations are canceled by negative spike-train correlations . In excitatory-inhibitory networks , spike-train correlations are typically positive . Here , the suppression of input correlations is not a result of the mere existence of correlations between excitatory ( E ) and inhibitory ( I ) neurons , but a consequence of a particular structure of correlations among the three possible pairings ( EE , EI , II ) .
Neurons generate signals by weighting and combining input spike trains from presynaptic neuron populations . The number of possible signals which can be read out this way from a given spike-train ensemble is maximal if these spike trains span an orthogonal basis , i . e . if they are uncorrelated [1] . If they are correlated , the amount of information which can be encoded in the spatio-temporal structure of these spike trains is limited . In addition , correlations impair the ability of readout neurons to decode information reliably in the presence of noise . This is often discussed in the context of rate coding: for uncorrelated spike trains , the signal-to-noise ratio of the compound spike-count signal can be enhanced by increasing the population size . In the presence of correlations , however , the signal-to-noise ratio is bounded [2] , [3] . The same reasoning holds for any other linear combination of spike trains , also for those where exact spike timing matters ( for example for the coding scheme presented in [4] ) . Thus , the robustness of neuronal responses against noise critically depends on the level of correlated activity within the presynaptic neuron population . Several studies suggested that correlated neural activity could be beneficial for information processing: Spike-train correlations can modulate the gain of postsynaptic neurons and thereby constitute a gating mechanism ( for a review , see [4] ) . Coherent spiking activity might serve as a means to bind elementary representations into more complex objects [5] , [6] . Information represented by correlated firing can be reliably sustained and propagated through feedforward subnetworks ( ‘synfire chains’; [7] , [8] ) . Whether correlated firing has to be considered favorable or not largely depends on the underlying hypothesis , the type of correlation ( e . g . the time scale or the affected frequency band ) or which subpopulations of neurons are involved . Most ideas suggesting a functional benefit of correlated activity rely on the existence of an asynchronous ‘ground state’ . Spontaneously emerging correlations , i . e . correlations which are not triggered by internal or external events , would impose a serious challenge to many of these hypotheses . Functionally relevant synfire activity , for example , cannot be guaranteed in the presence of correlated background input from the embedding network [9] . It is therefore–from several perspectives–important to understand the origin of uncorrelated activity in neural networks . It has recently been shown that spike trains of neighboring cortical neurons can indeed be uncorrelated [10] . Similar results have been obtained in several theoretical studies [11]–[17] . From an anatomical point of view , this observation is puzzling: in general , neurons in finite networks share a certain fraction of their presynaptic sources . In particular for neighboring neurons , the overlap between presynaptic neuron populations is expected to be substantial . This feedforward picture suggests that such presynaptic overlap gives rise to correlated synaptic input and , in turn , to correlated response spike trains . A number of theoretical studies showed that shared-input correlations are only weakly transferred to the output side as a consequence of the nonlinearity of the spike-generation dynamics [15] , [18]–[21] . Unreliable spike transmission due to synaptic failure can further suppress the correlation gain [22] . In [9] , we demonstrated that spike-train correlations in finite-size recurrent networks are even smaller than predicted by the low correlation gain of pairs of neurons with nonlinear spike-generation dynamics . We concluded that this suppression of correlations must be a result of the recurrent network dynamics . In this article , we compare correlations observed in feedforward networks to correlations measured in systems with an intact feedback loop . We refer to the reduction of correlations in the presence of feedback as “decorrelation” . Different mechanisms underlying such a dynamical decorrelation have been suggested in the recent past . Asynchronous states in recurrent neural networks are often attributed to chaotic dynamics [23] , [24] . In fact , networks of nonlinear units with random connectivity and balanced excitation and inhibition typically exhibit chaos [11] , [25] . The high sensitivity to noise may however question the functional relevance of such systems ( [26] , [27]; cf . , however , [28] ) . [29] and [27] demonstrated that asynchronous irregular firing can also emerge in networks with stable dynamics . Employing an analytical framework of correlations in recurrent networks of binary neurons [30] , the balance of excitation and inhibition has recently been proposed as another decorrelation mechanism [17]: In large networks , fluctuations of excitation and inhibition are in phase . Positive correlations between excitatory and inhibitory input spike trains lead to a negative component in the net input correlation which can compensate positive correlations caused by shared input . In the present study , we demonstrate that dynamical decorrelation is a fundamental phenomenon in recurrent systems with negative feedback . We show that negative feedback alone is sufficient to efficiently suppress correlations . Even in purely inhibitory networks , shared-input correlations are compensated by feedback . A balance of excitation and inhibition is thus not required . The underlying mechanism can be understood by means of a simple linear model . This simplifies the theory and helps to gain intuition , but it also confirms that low correlations can emerge in recurrent networks with stable , non-chaotic dynamics . The suppression of pairwise spike-train correlations by inhibitory feedback is reflected in a reduction of population-rate fluctuations . The main effect described in this article can therefore be understood by studying the dynamics of the macroscopic population activity . This approach leads to a simple mathematical description and emphasizes that the described decorrelation mechanism is a general phenomenon which may occur not only in neural networks but also in other ( biological ) systems with inhibitory feedback . In “Results: Suppression of population-rate fluctuations in LIF networks” , we first illustrate the decorrelation effect for random networks of leaky integrate-and-fire ( LIF ) neurons with inhibitory or excitatory-inhibitory coupling . By means of simulations , we show that low-frequency spike-train correlations , and , hence , population-rate fluctuations are substantially smaller than expected given the amount of shared input . As shown in the subsequent section , the “Suppression of population-activity fluctuations by negative feedback” can readily be understood in the framework of a simple one-dimensional linear model with negative feedback . In “Results: Population-activity fluctuations in excitatory-inhibitory networks” , we extend this to a two-population system with excitatory-inhibitory coupling . Here , a simple coordinate transform exposes the inherent negative feedback loop as the underlying cause of the fluctuation suppression in inhibition-dominated networks . The population-rate models of the inhibitory and the excitatory-inhibitory network are sufficient to understand the basic mechanism underlying the decorrelation . They do , however , not describe how feedback in cortical networks affects the detailed structure of pairwise correlations . In “Results: Population averaged correlations in cortical networks” , we therefore compute self-consistent population averaged correlations for a random network of linear excitatory and inhibitory neurons . By determining the parameters of the linear network analytically from the LIF model , we show that the predictions of the linear model are—for a wide and realistic range of parameters—in excellent agreement with the results of the LIF network model . In “Results: Effect of feedback manipulations” , we demonstrate that the active decorrelation in random LIF networks relies on the feedback of the ( sub ) population averaged activity but not on the precise microscopic structure of the feedback signal . In the “Discussion” , we put the consequences of this work into a broader context and point out limitations and possible extensions of the presented theory . The “Methods” contain details on the LIF network model , the derivation of the linear model from the LIF dynamics and the derivation of population-rate spectra and population averaged correlations in the framework of the linear model . This section is meant as a supplement; the basic ideas and the main results can be extracted from the “Results” .
To illustrate the effect of shared input and its suppression by the recurrent dynamics , we compare the spike response of a recurrent random network ( feedback scenario; Fig . 1 A , C , E ) of LIF neurons to the case where the feedback is cut and replaced by a spike-train ensemble , modeled by independent realizations of a stationary Poisson point process ( feedforward scenario; Fig . 1 B , D , F ) . The rate of this Poisson process is identical to the time and population averaged firing rate in the intact recurrent system . In both the feedback and the feedforward case , the ( local ) presynaptic spike trains are fed to the postsynaptic population according to the same connectivity matrix . Therefore , not only the in-degrees and the synaptic weights but also the shared-input statistics are exactly identical . For realistic size and connectivity , asynchronous states of random neural networks [12] , [31] exhibit spike-train correlations which are small but not zero ( compare raster displays in Fig . 1 C and D; see also [15] ) . Although the presynaptic spike trains are , by construction , independent in the feedforward case ( Fig . 1 D ) , the resulting response correlations , and , hence , the population-rate fluctuations , are substantially stronger than those observed in the feedback scenario ( compare Fig . 1 F and E ) . In other words: A theory which is exclusively based on the amount of shared input but neglects the details of the presynaptic spike-train statistics can significantly overestimate correlations and population-rate fluctuations in recurrent neural networks . The same effect can be observed in LIF networks with both purely inhibitory and mixed excitatory-inhibitory coupling ( Fig . 2 ) . To demonstrate this quantitatively , we focus on the fluctuations of the population averaged activity . Its power-spectrum ( or auto-correlation , in the time domain ) is determined both by the power-spectra ( auto-correlations ) of the individual spike trains and the cross-spectra ( cross-correlations ) ( ) of pairs of spike trains ( throughout the article , we use capital letters to represent quantities in frequency [Fourier] space; represents the Fourier transform of the spike train ) . We observe that the spike-train power-spectra ( and auto-correlations ) are barely distinguishable in the feedback and in the feedforward case ( not shown here; the main features of the spike-train auto-correlation are determined by the average single-neuron firing rate and the refractory mechanism; both are identical in the feedback and the feedforward scenario ) . The differences in the population-rate spectra are therefore essentially due to differences in the spike-train cross-spectra . In other words , the fluctuations in the population activity serve as a measure of pairwise spike-train correlations [32]: small ( large ) population averaged spike-train correlations are accompanied by small ( large ) fluctuations in the population rate ( see lower panels in Fig . 1 C–F ) . The power-spectra of the population averaged activity reveal a feedback-induced suppression of the population-rate variance at low frequencies up to several tens of Hertz . For the examples shown in Fig . 2 , this suppression spans more than three orders of magnitude for the inhibitory and more than one order of magnitude for the excitatory-inhibitory network . The suppression of low-frequency fluctuations does not critically depend on the details of the network model . As shown in Fig . 2 , it can , for example , be observed for both networks with zero rise-time synapses ( -shaped synaptic currents ) and short delays and for networks with delayed low-pass filtering synapses ( -shaped synaptic currents ) . In the latter case , the suppression of fluctuations is slightly more restricted to lower frequencies ( ) . Here , the fluctuation suppression is however similarly pronounced as in networks with instantaneous synapses . In Fig . 2 C , D , the power-spectra of the population activity converge to the mean firing rate at high frequencies . This indicates that the spike trains are uncorrelated on short time scales . For instantaneous -synapses , neurons exhibit an immediate response to excitatory input spikes [33] , [34] . This fast response causes spike-train correlations on short time scales . Hence , the compound power at high frequencies is increased . In a recurrent system , this effect is amplified by reverberating simultaneous excitatory spikes . Therefore , the high-frequency power of the compound activity is larger in the feedback case ( Fig . 2 B ) . Note that this high-frequency effect is absent in networks with more realistic low-pass filtering synapses ( Fig . 2 C , D ) and in purely inhibitory networks ( Fig . 2 A ) . Synaptic delays and slow synapses can promote oscillatory modes in certain frequency bands [12] , [31] , thereby leading to peaks in the population-rate spectra in the feedback scenario which exceed the power in the feedforward case ( see peaks at in Fig . 2 C , D ) . Note that , in the feedforward case , the local input was replaced by a stationary Poisson process , whereas in the recurrent network ( feedback case ) the presynaptic spike trains exhibit oscillatory modes . By replacing the feedback by an inhomogeneous Poisson process with a time dependent intensity which is identical to the population rate in the recurrent network , we found that these oscillatory modes are neither suppressed nor amplified by the recurrent dynamics , i . e . the peaks in the resulting power-spectra have the same amplitude in the feedback and in the feedforward case ( data not shown here ) . At low frequencies , however , the results are identical to those obtained by replacing the feedback by a homogeneous Poisson process ( i . e . to those shown in Fig . 2; see “Results: Effect of feedback manipulations” ) . In the present study , we mainly focus on these low-frequency effects . The observation that the suppression of low-frequency fluctuations is particularly pronounced in networks with purely inhibitory coupling indicates that inhibitory feedback may play a key role for the underlying mechanism . In the following subsection , we demonstrate by means of a one-dimensional linear population model that , indeed , negative feedback alone leads to an efficient fluctuation suppression . Average pairwise correlations can be extracted from the spectrum ( 1 ) of the compound activity , provided the single spike-train statistics ( auto-correlations ) is known ( see previous section ) . As the single spike-train statistics is identical in the feedback and in the feedforward scenario , the mechanism underlying the decorrelation in recurrent networks can be understood by studying the dynamics of the population averaged activity . In this and in the next subsection , we consider the linearized dynamics of random networks composed of homogeneous subpopulations of LIF neurons . The high-dimensional dynamics of such systems can be reduced to low-dimensional models describing the dynamics of the compound activity ( for details , see “Methods: Linearized network model” ) . Note that this reduction is exact for networks with homogeneous out-degree ( number of outgoing connections ) . For the networks studied here ( random networks with homogeneous in-degree ) , it serves as a sufficient approximation ( in a network of size where each connection is randomly and independently realized with probability [Erdös-Rényi graph] , the [binomial] in- and out-degree distributions become very narrow for large [relative to the mean in/out-degree]; both in- and out-degree are therefore approximately constant across the population of neurons ) . In this subsection , we first study networks with purely inhibitory coupling . In “Results: Population-activity fluctuations in excitatory-inhibitory networks” , we investigate the effect of mixed excitatory-inhibitory connectivity . Consider a random network of identical neurons with connection probability . Each neuron receives randomly chosen inputs from the local network with synaptic weights . In addition , the neurons are driven by external uncorrelated Gaussian white noise with amplitude , i . e . and . For small input fluctuations , the network dynamics can be linearized . This linearization is based on the averaged response of a single neuron to an incoming spike and describes the activity of an individual neuron by an abstract fluctuating quantity which is defined such that within the linear approximation its auto- and cross-correlations fulfill the same linearized equation as the spiking model in the low-frequency limit . Consequently , also the low-frequency fluctuations of the population spike rate are captured correctly by the reduced model up to linear order . This approach is equivalent to the treatment of finite-size fluctuations in spiking networks ( see , e . g . , [31] ) . For details , see “Methods: Linearized network model” . For large , the population averaged activity can hence be described by a one-dimensional linear system with linear kernel , effective coupling strength and the population averaged noise ( see “Methods: Linearized network model” and Fig . 3 B ) . The coupling strength represents the integrated linear response of the neuron population to a small perturbation in the input rate of a single presynaptic neuron . For a population of LIF neurons , its relation to the synaptic weight ( PSP amplitude ) is derived in “Methods: Linearized network model” and “Methods: Response kernel of the LIF model” . The normalized kernel ( with ) captures the time course of the linear response . It is determined by the single-neuron properties ( e . g . the spike-initiation dynamics [35] , [36] ) , the properties of the synapses ( e . g . synaptic weights and time constants [37] , [38] ) and the properties of the input ( e . g . excitatory vs . inhibitory input [39] ) . For many real and model neurons , the linear population-rate response exhibits low-pass characteristics [13] , [34]–[46] . For illustration ( Fig . 3 ) , we consider a 1st-order low-pass filter , i . e . an exponential impulse response with time constant ( cutoff frequency ; see Fig . 3 A , light gray curve in E ) . The results of our analysis are however independent of the choice of the kernel . The auto-correlation of the external noise is parametrized by the effective noise amplitude . Given the simplified description ( 2 ) , the suppression of response fluctuations by negative feedback can be understood intuitively: Consider first the case where the neurons in the local network are unconnected ( Fig . 3 A; no feedback , ) . Here , the response ( Fig . 3 A ) is simply a low-pass filtered version of the external input ( Fig . 3 A ) , resulting in an exponentially decaying response auto-correlation ( Fig . 3 D; light gray curve ) and a drop in the response power-spectrum at the cutoff frequency ( Fig . 3 E ) . At low frequencies , and are in phase; they are correlated . In the presence of negative feedback ( Fig . 3 B ) , the local input ( Fig . 3 B ) and the low-frequency components of the external input ( Fig . 3 B ) are anticorrelated . They partly cancel out , thereby reducing the response fluctuations ( Fig . 3 B ) . The auto-correlation function and the power-spectrum are suppressed ( Fig . 3 D , E; black curves ) . Due to the low-pass characteristics of the system , mainly the low-frequency components of the external drive are transferred to the output side and , in turn , become available for the feedback signal . Therefore , the canceling of input fluctuations and the resulting suppression of response fluctuations are most efficient at low frequencies . Consequently , the auto-correlation function is sharpened ( see inset in Fig . 3 D ) . The cutoff frequency of the system is increased ( Fig . 3 E; black curve ) . This effect of negative feedback is very general and well known in the engineering literature . It is employed in the design of technical devices , like , e . g . , amplifiers [47] . As the zero-frequency power is identical to the integrated auto-correlation function , the suppression of low-frequency fluctuations is accompanied by a reduction in the auto-correlation area ( Fig . 3 D; black curve ) . Note that the suppression of fluctuations in the feedback case is not merely a result of the additional inhibitory noise source provided by the local input , but follows from the precise temporal alignment of the local and the external input . To illustrate this , let's consider the case where the feedback channel is replaced by a feedforward input ( Fig . 3 C ) which has the same auto-statistics as the response in the feedback case ( Fig . 3 B ) but is uncorrelated to the external drive . In this case , external input fluctuations ( Fig . 3 C ) are not canceled by the local input ( Fig . 3 C ) . Instead , the local feedforward input acts as an additional noise source which leads to an increase in the response fluctuations ( Fig . 3 C ) . The response auto-correlation and power-spectrum ( Fig . 3 D , E; dark gray curves ) are increased . Compared to the unconnected case ( Fig . 3 E; light gray curve ) , the cutoff frequency remains unchanged . The feedback induced suppression of response fluctuations can be quantified by comparing the response power-spectra and in the feedback ( Fig . 3 B ) and the feedforward case ( Fig . 3 C ) , respectively ( see “Methods: Population-activity spectrum of the linear inhibitory network” ) . Here , and denote the Fourier transforms of the response fluctuations in the feedback and the feedforward scenario , respectively , the transfer function ( Fourier transform of the filter kernel ) of the neuron population , and the average across noise realizations . We use the power ratio as a measure of the relative fluctuation suppression caused by feedback . For low frequencies ( ) and strong effective coupling , the power ratio ( 5 ) decays as ( see Fig . 4 A ) : the suppression of population-rate fluctuations is promoted by strong negative feedback . In line with the observations in “Results: Suppression of population-rate fluctuations in LIF networks” , this suppression is restricted to low frequencies; for high frequencies ( , i . e . ) , the power ratio approaches . Note that the power ratio ( 5 ) is independent of the amplitude of the population averaged external input . Therefore , even if we dropped the assumption of the external inputs being uncorrelated , i . e . if for , the power ratio ( 5 ) remained the same . For correlated external input , the power of the population average is different from . The suppression factor , however , is not affected by this . Moreover , it is straightforward to show that the power ratio ( 5 ) is , in fact , independent of the shape of the external-noise spectrum . The same result ( 5 ) is obtained for any type of external input ( e . g . colored noise or oscillating inputs ) . For low frequencies , the transfer function approaches unity ( ) ; the exact shape of the kernel becomes irrelevant . In particular , the cutoff frequency ( or time constant ) of a low-pass kernel has no effect on the zero-frequency power ( integral correlation ) and the zero-frequency power ratio ( Fig . 4 ) . Therefore , the suppression of low-frequency fluctuations does not critically depend on the exact choice of the neuron , synapse or input model . The same reasoning applies to synaptic delays: Replacing the kernel by a delayed kernel leads to an additional phase factor in the transfer function . For sufficiently small frequencies ( long time scales ) , this factor can be neglected ( ) . For networks with purely inhibitory feedback , the absolute power ( 3 ) of the population rate decreases monotonously with increasing coupling strength . As we will demonstrate in “Results: Population-activity fluctuations in excitatory-inhibitory networks” and “Results: Population averaged correlations in cortical networks” , this is qualitatively different in networks with mixed excitatory and inhibitory coupling and , respectively: here , the fluctuations of the compound activity increase with . The power ratio , however , still decreases with . In the foregoing subsection , we have shown that negative feedback alone can efficiently suppress population-rate fluctuations and , hence , spike-train correlations . So far , it is unclear whether the same reasoning applies to networks with mixed excitatory and inhibitory coupling . To clarify this , we now consider a random network composed of a homogeneous excitatory and inhibitory subpopulation and of size and , respectively . Each neuron receives excitatory and inhibitory inputs from and with synaptic weights and , respectively . In addition , the neurons are driven by external Gaussian white noise . As demonstrated in “Methods: Linearized network model” , linearization and averaging across subpopulations leads to a two-dimensional system describing the linearized dynamics of the subpopulation averaged activity . Here , denotes the subpopulation averaged external uncorrelated white-noise input with correlation functions ( , ) , and a normalized linear kernel with . The excitatory and inhibitory subpopulations are coupled through an effective connectivity matrix with effective weight and balance parameter . The two-dimensional system ( 6 ) / ( 7 ) represents a recurrent system with both positive and negative feedback connections ( Fig . 5 A ) . By introducing new coordinates and , , we obtain an equivalent representation of ( 6 ) / ( 7 ) , describing the dynamics of the sum and difference activity and , respectively , i . e . the in- and anti-phase components of the excitatory and inhibitory subpopulations ( see [48]–[50] ) . The new coupling matrix reveals that the sum mode is subject to self-feedback ( ) and receives feedforward input from the difference mode ( ) . All remaining connections are absent ( ) in the new representation ( 8 ) ( see Fig . 5 B ) . The correlation functions of the external noise in the new coordinates are given by with ( ) . The feedforward coupling is positive ( ) : an excitation surplus ( ) will excite all neurons in the network , an excitation deficit ( ) will lead to global inhibition . In inhibition dominated regimes with , the self-feedback of the sum activity is effectively negative ( ) . The dynamics of the sum rate in inhibition-dominated excitatory-inhibitory networks is therefore qualitatively similar to the dynamics in purely inhibitory networks ( “Results: Suppression of population-activity fluctuations by negative feedback” ) . As shown below , the negative feedback loop exposed by the transform ( 8 ) leads to an efficient relative suppression of population-rate fluctuations ( if compared to the feedforward case ) . Mathematically , the coordinate transform ( 8 ) corresponds to a Schur decomposition of the dynamics: Any recurrent system of type ( 6 ) ( with arbitrary coupling matrix ) can be transformed to a system with a triangular coupling matrix ( see , e . g . , [50] ) . The resulting coupling between the different Schur modes can be ordered so that there are only connections from modes with lower index to modes with the same or larger index . In this sense , the resulting system has been termed ‘feedforward’ [50] . The original coupling matrix is typically not normal , i . e . . Its eigenvectors do not form an orthogonal basis . By performing a Gram-Schmidt orthonormalization of the eigenvectors , however , one can obtain a ( normalized ) orthogonal basis , a Schur basis . Our new coordinates ( 8 ) correspond to the amplitudes ( the time evolution ) of two orthogonal Schur modes . The spectra , , and of the subpopulation averaged rates , and the sum mode , respectively , are derived in “Methods: Population-activity spectra of the linear excitatory-inhibitory network” . In contrast to the purely inhibitory network ( see “Results: Suppression of population-activity fluctuations by negative feedback” ) , the population-rate fluctuations of the excitatory-inhibitory network increase monotonously with increasing coupling strength . For strong coupling , approaches from below with . Close to the critical point ( ) , the rate fluctuations become very large; ( 11 ) diverges . Increasing the amount of inhibition by increasing , however , leads to a suppression of these fluctuations . In the limit , and ( 11 ) approach the spectrum of the unconnected network . For strong coupling ( ) , the ratio approaches : the fluctuations of the population averaged excitatory firing rate exceed those of the inhibitory population by a factor ( independently of and ) . Similarly to the strategy we followed in the previous subsections , we will now compare the population-rate fluctuations of the feedback system ( 6 ) , or equivalently ( 9 ) , to the case where the feedback channels are replaced by feedforward input with identical auto-statistics . A straight-forward implementation of this is illustrated in Fig . 5 C: Here , the excitatory and inhibitory feedback channels and are replaced by uncorrelated feedforward inputs and , respectively . The Schur representation of this scenario is depicted in Fig . 5 D . According to ( 6 ) , the Fourier transforms of the response fluctuations of this system read With , and using , , , we can express the spectrum of the sum activity in the feedforward case in terms of the spectra and of the feedback system ( see eq . ( 55 ) ) . For strong coupling ( ) , the zero-frequency component ( ) becomes Thus , for strong coupling , the zero-frequency power ratio reveals a relative suppression of the population-rate fluctuations in the feedback system which is proportional to ( see Fig . 4 B; black dashed line ) . The power ratio for arbitrary weights is depicted in Fig . 4 B ( black dotted curve ) . For a network at the transition point , ( 14 ) equals . Increasing the level of inhibition by increasing leads to a decrease in the power ratio: in the limit , ( 14 ) approaches monotonously . Above , we suggested that the negative self-feedback of the sum mode , weighted by ( Fig . 5 B ) , is responsible for the fluctuation suppression in the recurrent excitatory-inhibitory system . Here , we test this by considering the case where this feedback loop is opened and replaced by uncorrelated feedforward input , weighted by , while the feedforward input from the difference mode , weighted by , is left intact ( see Fig . 5 D′ ) . As before , we assume that the auto-statistics of is identical to the auto-statistics of as obtained in the feedback case , i . e . . According to the Schur representation of the population dynamics ( 9 ) / ( 10 ) , the Fourier transform of the sum mode of this modified system is given by With given in ( 54 ) and , we obtain the power ratio Its zero-frequency component is shown in Fig . 4 B ( gray dotted curve ) . For strong coupling , the power ratio decays as ( gray dashed line in Fig . 4 B ) . Thus , the ( relative ) power in the recurrent system is reduced by strengthening the negative self-feedback loop , i . e . by increasing . So far , we have presented results for the subpopulation averaged firing rates and and the sum mode . The spectrum of the compound rate , i . e . the activity averaged across the entire population , reads In the feedforward scenario depicted in Fig . 5 C , the spectrum of the compound rate ( with ) is given by For strong coupling , the corresponding low-frequency power ratio ( black solid curve in Fig . 4 B ) exhibits qualitatively the same decrease as the sum mode . To summarize the results of this subsection: the population dynamics of a recurrent network with mixed excitatory and inhibitory coupling can be mapped to a two-dimensional system describing the dynamics of the sum and the difference of the excitatory and inhibitory subpopulation activities . This equivalent representation uncovers that , in inhibition dominated networks ( ) , the sum activity is subject to negative self-feedback . Thus , the dynamics of the sum activity in excitatory-inhibitory networks is qualitatively similar to the population dynamics of purely inhibitory networks ( see “Results: Suppression of population-activity fluctuations by negative feedback” ) . Indeed , the comparison of the compound power-spectra of the intact recurrent network and networks where the feedback channels are replaced by feedforward input reveals that the ( effective ) negative feedback in excitatory-inhibitory networks leads to an efficient suppression of population-rate fluctuations . The results presented in the previous subsections describe the fluctuations of the compound activity . Pairwise correlations between the ( centralized ) spike trains are outside the scope of such a description . In this subsection , we consider the same excitatory-inhibitory network as in “Results: Population-activity fluctuations in excitatory-inhibitory networks” and present a theory for the population averaged spike-train cross-correlations . In general , this is a hard problem . To understand the structure of cross-correlations , it is however sufficient to derive a relationship between the cross- and auto-covariances in the network , because the latter can , to good approximation , be understood in mean-field theory . The integral of the auto-covariance function of spiking LIF neurons can be calculated by Fokker-Planck formalism [12] , [31] , [51] . To determine the relation between the cross-covariance and the auto-covariance , we replace the spiking dynamics by a reduced linear model with covariances obeying , to linear order , the same relation . We present the full derivation in “Methods: Linearized network model” . There , we first derive an approximate linear relation between the auto- and cross-covariance functions and , respectively , of the LIF network . A direct solution of this equation is difficult . In the second step , we therefore show that there exists a linear stochastic system with activity and correlations and fulfilling the same equation as the original LIF model . This reduced model can be solved in the frequency domain by standard Fourier methods . Its solution allows us , by construction , to determine the relation between the integral cross-covariances and the integral auto-covariances up to linear order . As we are interested in the covariances averaged over many pairs of neurons , we average the resulting set of linear self-consistency equations ( 56 ) for the covariance matrix in the frequency domain over statistically identical pairs of neurons and many realizations of the random connectivity ( see “Methods: Population averaged correlations in the linear EI network” ) . This yields a four-dimensional linear system ( 76 ) describing the population averaged variances and of the excitatory and inhibitory subpopulations , and the covariances and for unconnected excitatory-excitatory and inhibitory-inhibitory neuron pairs , respectively ( note that we use the terms “variance” and “covariance” to describe the integral of the auto- and cross-correlation function , respectively; in many other studies , they refer to the zero-lag correlation functions instead ) . The dependence of the variances and covariances on the coupling strength , obtained by numerically solving ( 76 ) , is shown in Fig . 6 . We observe that the variances and of excitatory and inhibitory neurons are barely distinguishable ( Fig . 6 A ) . With the approximation , explicit expressions can be obtained for the covariances ( thick dashed curves Fig . 6 E ) : The deviations from the full solutions ( thin solid curves in Fig . 6 E ) , i . e . for , are small . In the reduced model , both the external input and the spiking of individual neurons contribute to an effective noise . As the fluctuations in the reduced model depend linearly on the amplitude of this noise , the variances and covariances ( ) can be expressed in units of the noise variance . Consequently , the correlation coefficients are independent of ( see Fig . 6 ) . The analytical form ( 19 ) of the result shows that the correlations are smaller than expected given the amount of shared input a pair of neurons receives: The quantity in the first line is the contribution of shared input to the covariance . For strong coupling , the prefactor causes a suppression of this contribution . Its structure is typical for a feedback system , similar to the solution ( 3 ) of the one-population or the solution ( 52 ) of the two-population model . The term in the denominator represents the negative feedback of the compound rate . The prefactor in the second line of ( 19 ) is again due to the feedback and suppresses the contribution of the factor , which represents the effect of direct connections between neurons . Our results are consistent with a previous study of the decorrelation mechanism: In [17] , the authors considered how correlations scale with the size of the network where the synaptic weights are chosen as . As a result , the covariance in ( 19 ) caused by shared input is independent of the network size , while the feedback scales—to leading order—as ( see ( 45 ) ) . Consequently , the first line in ( 19 ) scales as . The same scaling holds for the second line in ( 19 ) , explaining the decay of correlations as found in [17] . The first line in ( 19 ) is identical for any pair of neurons . The second line is positive for a pair of excitatory neurons and negative for a pair of inhibitory neurons . In other words , excitatory neurons are more correlated than inhibitory ones . Together with the third line in ( 19 ) , this reveals a characteristic correlation structure: ( Fig . 6 B , E ) . For strong coupling , the difference between the excitatory and inhibitory covariance is . The difference decreases as the level of inhibition is increased , i . e . the further the network is in the inhibition dominated regime , away from the critical point . To understand the suppression of shared-input correlations in recurrent excitatory-inhibitory networks , consider the correlation between the local inputs of a pair of neurons , . The input-correlation coefficient can be expressed in terms of the averaged spike-train covariances: ( see “Methods: Population averaged correlations in the linear EI network”: The input covariance equals the average quantity given in ( 67 ) , the input variance is given by ( 63 ) as ) . The term represents the contribution due to the spike-train variances of the shared presynaptic neurons ( see ( 19 ) ) . This contribution is always positive ( provided the network architecture is consistent with Dale's law; see [15] ) . In a purely feedforward scenario with uncorrelated presynaptic sources , is the only contribution to the input covariance of postsynaptic neurons . The resulting response correlation for this feedforward case is much larger than in the feedback system ( Fig . 6 B , black dotted curve ) . The correlation coefficient between inputs to a pair of neurons in the feedforward case is identical to the network connectivity ( horizontal dotted curve in Fig . 6 D; see [15] ) . In an inhibition dominated recurrent network , spike-train correlations between pairs of different source neurons contribute the additional term , which is negative and of similar absolute value as the shared-input contribution . Thus , the two terms and partly cancel each other ( see Fig . 6 C ) . In consequence , the resulting input correlation coefficient is smaller than ( see Fig . 6 D; here: ) . The correlations in a purely inhibitory network can be obtained from ( 19 ) by replacing , taking into account the negative sign of in and setting and : For finite coupling strength , this expression is negative . The contributions of shared input and spike-train correlations to the input correlation are given by and , respectively ( see ( 19 ) and ( 20 ) ) . Using ( 21 ) , we can directly verify that , because pairwise correlations are negative , leading to a partial cancellation : the right hand side is smaller in magnitude by a factor of compared to each individual contribution . Hence , as in the network with excitation and inhibition , shared-input correlations are partly canceled by the contribution due to presynaptic pairwise spike-train correlations . In the feedforward scenario with zero presynaptic spike-train correlations , in contrast , the response correlations are determined by shared input alone and are therefore increased . The suppression of shared-input correlations in the feedback case is what we call ‘decorrelation’ in the current work . In purely inhibitory networks , this decorrelation is caused by weakly negative pairwise correlations ( 21 ) . For sufficiently strong negative feedback , correlations are smaller in absolute value as compared to the feedforward case . The absolute value of these anti-correlations is bounded by . The similarity in the results obtained for purely inhibitory networks and excitatory-inhibitory networks demonstrates that the suppression of pairwise correlations and population-activity fluctuations is a generic phenomenon in systems with negative feedback . It does not rely on an internal balance between excitation and inhibition . As discussed in “Results: Suppression of population-rate fluctuations in LIF networks” , the suppression of correlations in the recurrent network is accompanied by a reduction of population-activity fluctuations . With the population averaged correlations ( 19 ) , the power ( 1 ) of the population activity reads In “Results: Population-activity fluctuations in excitatory-inhibitory networks” , we showed that the population-activity fluctuations are amplified if the local input in the recurrent system is replaced by feedforward input from independent excitatory and inhibitory populations ( see Fig . 5 C ) . This manipulation corresponds to a neglect of correlations between excitatory and inhibitory neurons . All remaining correlations ( , , , ) are preserved . With the resulting response auto- and cross-correlations and given by ( 84 ) , the power ( 1 ) of the population activity becomes For large effective coupling , the power ratio decays as ( black curve in Fig . 6 F ) . Note that the power ratio derived here is indistinguishable from the one we obtained in the framework of the population model in “Results: Population-activity fluctuations in excitatory-inhibitory networks” ( black solid curve in Fig . 4 B ) . Although the derivation of the macroscopic model in “Results: Population-activity fluctuations in excitatory-inhibitory networks” is different from the one leading to the population averaged correlations described here , the two models are consistent: They describe one and the same system and lead to identical power ratios . The fluctuation suppression is not only observed at the level of the entire network , i . e . for the population activity , but also for each individual subpopulation and , i . e . for the subpopulation averaged activities and . The derivation of the corresponding power ratios and is analog to the one described above . As a result of the correlation structure in the feedback system ( see Fig . 6 B ) , the power of the inhibitory population activity is smaller than the power of the excitatory population activity . In consequence , ( gray curves in Fig . 6 F ) . In ( 22 ) and ( 23 ) , the auto-correlations are scaled by , while the cross-correlations enter with a prefactor of order unity . For large , one may therefore expect that the suppression of population-activity fluctuations is essentially mediated by pairwise correlations . In the recurrent system , however , the cross-correlations ( ) are of order ( see Fig . 6 and ( 19 ) ) . It is therefore a priori not clear whether the fluctuation suppression is indeed dominated by pairwise correlations . In our framework , one can explicitly show that the auto-correlation is irrelevant: Replacing the auto-correlation in ( 23 ) by the average auto-correlation of the intact feedback system has no visible effect on the resulting power ratio ( dashed curves in Fig . 6 F ) . The difference in the spectra of the population activities and is therefore essentially caused by the cross-correlations . The absolute population-activity fluctuations in purely inhibitory and in excitatory-inhibitory networks show a qualitatively different dependence on the synaptic coupling , in agreement with the previous sections . In networks with excitation and inhibition , the correlation coefficient increases with increasing synaptic coupling ( see Fig . 6 E ) . Hence , the population-activity fluctuations grow with increasing coupling strength . In purely inhibitory networks , in contrast , the pairwise spike-train correlation decreases monotonously with increasing magnitude of the coupling strength , see ( 21 ) . In consequence , the population-activity fluctuations decrease . The underlying reason is that , in the inhibitory network , the power of the population activity is directly proportional to the covariance of the input currents , which is actively suppressed , as shown above . For excitatory-inhibitory networks , these two quantities are not proportional ( compare ( 20 ) and ( 1 ) ) due to the different synaptic weights appearing in the input covariance . To compare our theory to simulations of spiking LIF networks , we need to determine the effect of a synaptic input on the response activity of the neuron model . To this end , we employ the Fokker-Planck theory of the LIF model ( see “Methods: Response kernel of the LIF model” ) . In this context , the steady state of the recurrent network is characterized by the mean and the standard deviation of the total synaptic input . Both and depend on the steady-state firing rate in the network . The steady-state firing rate can be determined in a self-consistent manner [12] as the fixed point of the firing rate approximation ( 42 ) . The approximation predicts the firing rate to sufficient accuracy of about ( see Fig . 7 A ) . We then obtain an analytical expression of the low-frequency transfer which relates the fluctuation of a synaptic input to neuron to the fluctuation of neuron 's response firing rate to linear order , so that . This relates the postsynaptic potential in the LIF model to the effective linear coupling in our linear theory . The functional relation can be derived in analytical form by linearization of ( 42 ) about the steady-state working point . Note that depends on and and , hence , on the steady-state firing rate in the network . The derivation outlined in “Methods: Response kernel of the LIF model” constitutes an extension of earlier work [21] , [33] to quadratic order in . The results agree well with those obtained by direct simulation for a large range of synaptic amplitudes ( see Fig . 8 ) . Fig . 7 B compares the population averaged correlation coefficients obtained from the linear reduced model , see ( 19 ) , and simulations of LIF networks . Note that the absolute value of the noise amplitude in the reduced model does not influence the correlation coefficient , as both quantities and depend linearly on . Theory and simulation agree well for synaptic weights up to . For larger synaptic amplitudes , the approximation of the effective linear transfer for a single neuron obtained from the Fokker-Planck theory deviates from its actual value ( see Fig . 8 B ) . Fig . 7 C shows that the cancellation of the input covariance in the LIF network is well explained by the theory . Previous work [17] suggested that positive correlations between excitatory and inhibitory inputs lead to a negative component in the input correlation which , in turn , suppresses shared-input correlations . The mere existence of positive correlations between excitatory and inhibitory inputs is however not sufficient . To explain the effect , it is necessary to take the particular correlation structure into account . To illustrate this , consider the case where the correlation structure is destroyed by replacing all pairwise correlations in the input spike-train ensemble by the overall population average ( homogenization of correlations ) . The resulting response correlations ( upper gray curve in Fig . 7 B ) are derived in “Methods: Population averaged correlations in the linear EI network” , eq . ( 86 ) . In simulations of LIF networks , we study the effect of homogenized spike-train correlations by first recording the activity of the intact recurrent network , randomly reassigning the neuron type ( or ) to each recorded spike train , and feeding this activity into a second population of neurons . Compared to the intact recurrent network , the response correlations are significantly larger ( Fig . 7 B ) . The contribution of homogenized spike-train correlations to the input covariance ( see ( 20 ) ) is given by . For positive spike-train correlations , this contribution is greater or equal zero ( zero for ) . Hence , it cannot compensate the ( positive ) shared-input contribution ( see Fig . 7 C ) . In consequence , input correlations , output correlations and , in turn , population-rate fluctuations ( Fig . 7 D ) cannot be suppressed by homogeneous positive correlations in the input spike-train ensemble . Canceling of shared-input correlations requires either negative spike-train correlations ( as in purely inhibitory networks ) or a heterogeneity in correlations across different pairs of neurons ( e . g . ) . In the previous subsections , we quantified the suppression of population-rate fluctuations in recurrent networks by comparing the activity in the intact recurrent system ( feedback scenario ) to the case where the feedback is replaced by feedforward input with some predefined statistics ( feedforward scenario ) . We particularly studied the effect of neglecting the auto-statistics of the compound feedback , ( the structure of ) correlations within the feedback ensemble and/or correlations between the feedback and the external input . In all cases , we observed a significant amplification of population-activity fluctuations in the feedforward scenario . In this subsection , we further investigate the role of different types of feedback manipulations by means of simulations of LIF networks with excitatory-inhibitory coupling . To this end , we record the spiking activity of the recurrent network ( feedback case ) , apply different types of manipulations to this activity ( described in detail below ) and feed this modified activity into a second population of identical ( unconnected ) neurons ( feedforward case ) . As before , the connectivity structure ( in-degrees , shared-input structure , synaptic weights ) is exactly identical in the feedback and the feedforward case . In “Methods: Linearized network model” , we show that the low-frequency fluctuations of the population rate of the spiking model are captured by the reduced model presented in the previous subsections . To verify that the theory based on excitatory and inhibitory population rates is indeed sufficient to explain the decorrelation mechanism , we first consider the case where the sender identities of the presynaptic spike train are randomly shuffled . Fig . 9 A shows the power-spectrum of the population activity recorded in the original network ( FB ) as well as the spectra obtained after shuffling spike-train identities within the excitatory and inhibitory subpopulations separately ( Shuff2D ) , or across the entire network ( Shuff1D ) . As shuffling of neuron identities does not change the population rates , all three compound spectra are identical . Fig . 9 B shows the response power-spectra of the neuron population receiving the shuffled spike trains . Shuffling within the subpopulations ( Shuff2D ) preserves the population-specific fluctuations and average correlations . The effect on the response fluctuations is negligible ( compare black and light gray curves in Fig . 9 B ) . In particular , the power of low-frequency fluctuations remains unchanged ( Fig . 9 C ) . This result confirms that population models which take excitatory and inhibitory activity separately into account are sufficient to explain the observations . Shuffling of spike-train identities across subpopulations ( Shuff1D ) , in contrast , causes an increase in the population fluctuations by about one order of magnitude ( Fig . 9 B , C; dark gray ) . This outcome is in agreement with the result obtained by homogenizing pairwise correlations ( see Fig . 7 ) and demonstrates that the excitatory and inhibitory subpopulation rates have to be conserved to explain the observed fluctuation suppression . The shuffling experiments and the results of the linear model in the previous subsections suggest that the precise temporal structure of the population averaged activities within homogeneous subpopulations is essential for the suppression of population-rate fluctuations . Preserving the exact structure of individual spike trains is not required . This is confirmed by simulation experiments where new sender identities were randomly reassigned for each individual presynaptic spike ( rather than for each spike train; data not shown ) . This operation destroys the structure of individual spike trains but preserves the compound activities . The results are similar to those reported here . So far , it is unclear how sensitive the fluctuation-suppression mechanism is to perturbations of the temporal structure of the population rates . To address this question , we replaced the excitatory and inhibitory spike trains in the feedback ensemble by independent realizations of inhomogeneous Poisson processes ( PoissI ) with intensities given by the measured excitatory and inhibitory population rates and of the recurrent network , respectively . Note that the compound rates of a single realization of this new spike-train ensemble are similar but not identical to the original population rates , ( in each time window , the resulting spike count is a random number drawn from a Poisson distribution with mean and variance proportional to and , respectively ) . Although the compound spectrum of the resulting local input is barely distinguishable from the compound spectrum of the intact recurrent system ( Fig . 9 D; black and dark gray curves ) , the response spectra are very different: replacing the feedback ensemble by inhomogeneous Poisson processes leads to a substantial amplification of low-frequency fluctuations ( Fig . 9 E; compare black and dark gray curves ) . The effect is as strong as if the temporal structure of the population rates was completely ignored , i . e . if the feedback channels were replaced by realizations of homogeneous Poisson processes with constant rates ( PoissH; light gray curves in Fig . 9 D , E ) . This result indicates that the precise temporal structure of the population rates is essential and that even small deviations can significantly weaken the fluctuation-suppression mechanism . The results of the Poisson experiments can be understood by considering the effect of the additional noise caused by the stochastic realization of individual spikes . Considering the auto-correlation , a Poisson spike-train ensemble with rate profile is equivalent to a sum of the rate profile and a noise term resulting from the stochastic ( Poissonian ) realization of spikes , . Here , denotes a Gaussian white noise with auto-correlation and the mean firing rate . The response fluctuations of the population driven by the rate modulated Poisson activity are , to linear approximation , given by . Inserting , we obtain an additional noise term in the spectrum which explains the increase in power compared to the spectrum of the recurrent network . As a generalization of the Poisson model , one may replace the noise amplitude by some arbitrary prefactor . In simulation experiments , we observed a gradual amplification of the population-rate fluctuations with increasing noise amplitude ( data not shown ) .
We have shown that negative feedback in recurrent neural networks actively suppresses low-frequency fluctuations of the population activity and pairwise correlations . This mechanism allows neurons to fire more independently than expected given the amount of shared presynaptic input . We demonstrated that manipulations of the feedback statistics , e . g . replacing feedback by uncorrelated feedforward input , can lead to a significant amplification of response correlations and population-rate fluctuations . The suppression of correlations and population-rate fluctuations by feedback can be observed in networks with both purely inhibitory and mixed excitatory-inhibitory coupling . In purely inhibitory networks , the effect can be understood by studying the role of the effective negative feedback experienced by the compound activity . In networks of excitatory and inhibitory neurons , a change of coordinates , technically a Schur decomposition , exposes the underlying feedback structure: the sum of the excitatory and inhibitory activity couples negatively to itself if the network is in an inhibition dominated regime ( which is required for its stability; see , e . g . , [12 ) . This negative feedback suppresses fluctuations in a similar way as in purely inhibitory networks . The fluctuation suppression becomes more efficient the further the network is brought into the inhibition dominated regime , away from the critical point of equal recurrent excitation and inhibition ( ) . Having identified negative feedback as the underlying cause of small fluctuations and correlations , we can rule out previous explanations based on a balance between ( correlated ) excitation and inhibition [17] . We presented a self-consistent theory for the average pairwise spike-train correlations which illuminates that the suppression of population-rate fluctuations and the suppression of pairwise correlations are two expressions of the same effect: as the single spike-train auto-covariance is the same in the feedforward and the feedback case , the suppression of population-rate fluctuations implies smaller correlations . Our theory enables us to identify the cancellation of input correlations as a hallmark of small spike-train correlations . In previous studies , shared presynaptic input has often been considered a main source of correlation in recurrent networks ( e . g . [15] , [52] ) . Recently [17] , suspected that correlations between excitatory and inhibitory neurons and the fast tracking of external input by the excitatory and the inhibitory population are responsible for an active decorrelation . We have demonstrated here that the mere fact that excitatory and inhibitory neurons are correlated is not sufficient to suppress shared-input correlations . Rather , we find that the spike-train correlation structure in networks of excitatory and inhibitory networks arranges such that their overall contribution to the covariance between the summed inputs to a pair of neurons becomes negative , canceling partly the effect of shared inputs . This cancellation becomes more precise the stronger the negative compound feedback is . In homogeneous networks where excitatory and inhibitory neurons receive statistically identical input , the particular structure of correlations is . It can further be shown that this structure of correlations is preserved in the limit of large networks ( ) . For non-homogeneous synaptic connectivity , if the synaptic amplitudes depend on the type of the target neuron ( i . e . or ) , the structure of correlations may be different . Still , the correlation structure arranges such that shared input correlation is effectively suppressed . Formally , this can be seen from a self-consistency equation similar to our equation ( 80 ) . The study by [17] has shown that correlations are suppressed in the limit of infinitely large networks of binary neurons receiving randomly drawn inputs from a common external population . Its argument rests on the insight that the population-activity fluctuations in a recurrent balanced network follow the fluctuations of the external common population . An elegant scaling consideration for infinitely large networks with vanishing synaptic efficacy shows that this fast tracking becomes perfect in the limit . This allows to determine the zero-lag pairwise correlations caused by the external input . The analysis methods and the recurrent networks presented here differ in several respects from these previous results: We study networks of a finite number of spiking model neurons . The neurons receive uncorrelated external input , so that correlations are due to the local recurrent connectivity among neurons , not due to tracking of the common external input [17] . Moreover , we consider homogeneous connectivity where synaptic weights depend only on the type of the presynaptic neuron ( as , e . g . , in [12] ) , resulting in a correlation structure . For such connectivity , networks of binary neurons with uncorrelated external input exhibit qualitatively the same correlation structure as reported here ( results not shown ) . In purely inhibitory networks , the decorrelation occurs in an analog manner as in excitatory-inhibitory networks . As only a single population of neurons is available here , population averaged spike-train correlations are negative . This negative contribution compensates the positive contribution of shared input . The structure of integrated spike-train covariances in networks constitutes an experimentally testable prediction . Note , however , that the prediction ( 19 ) obtained in the current work rests on two simplifying assumptions: identical internal dynamics of excitatory and inhibitory neurons and homogeneous connectivity ( i . e . , ; see “Results: Population-activity fluctuations in excitatory-inhibitory networks” ) . For such networks , the structure of correlations is given by . Further , the relation between subthreshold membrane-potential fluctuations and spike responses is the same for both neuron types . Consequently , the above correlation structure can be observed not only at the level of spike trains but also for membrane potentials , provided the assumptions hold true . A recent experimental study [53] reports neuron-type specific cross-correlation functions in the barrel cortex of behaving mice , both for spike trains and membrane potentials . It is however difficult to assess the integral correlations from the published data . A direct test of our predictions requires either a reanalysis of the data or a theory predicting the entire correlation functions . The raw ( unnormalized ) II and EI spike-train correlations in [53] are much more pronounced than the EE correlations ( Fig . 6 in [53] ) . This seems to be in contradiction to our results . Note , however , that the firing rates of excitatory and inhibitory neurons are very different in [53] . In our study , in contrast , the average firing rates of excitatory and inhibitory neurons are identical as a consequence of the assumed network homogeneity . Future theoretical work is needed to generalize our model to networks with heterogeneous firing rates and non-homogeneous connectivity . Recent results on the dependence of the correlation structure on the connectivity may prove useful in this endeavor [54]–[56] . Correlations in spike-train ensembles play a crucial role for the en- and decoding of information . A set of uncorrelated spike trains provides a rich dynamical basis which allows readout neurons to generate a variety of responses by tuning the strength and filter properties of their synapses [1] In the presence of correlations , the number of possible readout signals is limited . Moreover , spike-train correlations impair the precision of such readout signals in the presence of noise . Consider , for example , a linear combination of presynaptic spike trains with arbitrary ( linear ) filter kernels ( e . g . synaptic filters ) . In a realistic scenario , the individual spike trains typically vary across trials [3] , [57] . To understand how robust the resulting readout signal is against this spike-train variability , let's consider the variability of its Fourier transform . Assuming homogeneous spike-train statistics , the ( squared ) signal-to-noise ratio of the readout signal is given by Here , denotes the average across the ensemble of spike-train realizations , the spike-train coherence , and the coefficients and the 1st- and 2nd-order filter statistics . For uncorrelated spike trains , i . e . , and , the signal-to-noise ratio grows unbounded with the population size . Thus , even for noisy spike trains ( ) , the compound signal can be highly reliable if the population size is sufficiently large . In the presence of correlations , , however , converges towards a constant value as grows . Even for large populations , the readout signal remains prone to noise . These findings constitute a generalization of the results reported for population-rate coding , i . e . sums of unweighted spike counts ( see , e . g . , [2] , [3] ) . The above arguments illustrate that the same reasoning applies to coding schemes which are based on the spatio-temporal structure of spike patterns . In a previous study [9] , we demonstrated that active decorrelation in recurrent networks is a necessary prerequisite for a controlled propagation of synchronous volleys of spikes in embedded feedforward subnetworks ( ‘synfire chains’; Fig . 10 ) : A synfire chain receiving background input from a finite population of independent Poisson sources amplifies the resulting shared-input correlations , thereby leading to spontaneous synchronization within the chain ( Fig . 10 B ) . A distinction between these spurious synchronous events and those triggered by an external stimulus is impossible . The synfire chain loses its asynchronous ground state [58] . A synfire chain receiving background inputs from a recurrent network , in contrast , is much more robust . Here , shared-input correlations are actively suppressed by the recurrent-network dynamics . Synchronous events can be triggered by external stimuli in a controlled manner ( Fig . 10 A ) . Apart from the spontaneous synchronization illustrated in Fig . 10 , decorrelation by inhibition might solve another problem arising in embedded synfire structures: In the presence of feedback connections between the synfire chain and the embedding background network , synchronous spike volleys can excite ( high-frequency ) oscillatory modes in the background network which , in turn , interfere with the synfire dynamics and prevent a robust propagation of synchronous activity within the chain ( ‘synfire explosion’; see [59] , [60] ) . The decorrelation mechanism we refer to in our work is efficient only at low frequencies . It cannot prevent the build-up of these oscillations . [61] demonstrated that the ‘synfire explosion’ can be suppressed by adding inhibitory neurons to each synfire layer ( ‘shadow inhibition’ ) which diffusely project to neurons in the embedding network , thereby weakening the impact of synfire activity on the embedding network . In the present work we focus on the integral of the correlation function , nurtured by our interest in the low-frequency fluctuations . An analog treatment can however easily be performed for the zero-lag correlations . In contrast to infinite networks with sparse connectivity ( , ) , in the case of finite networks , pairs of neurons must be distinguished according to whether they are synaptically connected or not in order to arrive at a self-consistent theory for the averaged correlations . Providing explicit expressions for correlations between connected and unconnected neurons , the current work provides the tools to relate experimentally observed spiking correlations to the underlying synaptic connectivity . The quantification of pairwise correlations is a necessary prerequisite to understand how correlation sensitive synaptic plasticity rules , like spike-timing dependent plasticity [62] , interact with the recurrent network dynamics [63] . Existing theories quantifying correlations employ stochastic neuron models and are limited to purely excitatory networks [63]–[65] . Here , we provide an analytical equivalence relation between a reduced linear model and spiking integrate-and-fire neurons describing fluctuations correctly up to linear order . A formally similar approach has been employed earlier to study delayed cumulative inhibition in spiking networks [66] . We show that the correlations observed in recurrent networks in the asynchronous irregular regime are quantitatively captured for realistic synaptic coupling with postsynaptic potentials of up to about . The success of this approach can be explained by the linearization of the neural threshold units by the afferent noise experienced in the asynchronous regime . For linear neural dynamics , the second-order description of fluctuations is closed [67] . We exploit this finding by applying perturbation theory to the Fokker-Planck description of the integrate-and-fire neuron to obtain the linear input-output transfer at low frequencies [33] , thereby determining the effective coupling in our linear model . The scope of the theory presented in the current work is limited mainly by three assumptions . The first is the use of a linear theory which exhibits an instability as soon as a single eigenvalue of the effective connectivity matrix assumes a positive real part . This ultimately happens when increasing the synaptic coupling strength , because the eigenvalues of the random connectivity matrix are located in a circle centered in the left half of the complex plain with a radius given by the square root of the variance of the matrix elements [68] , [69] . Nonlinearities , like those imposed by strictly positive firing rates , prevent such unbounded growth ( or decay ) by saturation . For nonlinear rate models with sigmoidal transfer functions it has been shown that the activity of recurrent random networks of such units makes a transition to chaos at the point where the linearized dynamics would loose stability [70] . However , this point of transition is sharp only in the limit of infinitely large networks . From the population averaged firing rate and the pairwise correlations averaged over pairs of neurons considered in Fig . 7 we cannot conclude whether or not a transition to chaos occurs in the spiking network . In simulations and in the linearized reduced model , we could however observe that the distribution of pairwise correlations broadens when approaching the point of instability . Future work needs to examine this question in detail , e . g . by considering measures related to the Lyapunov exponent . Recently developed semi-analytical theories accounting for nonlinear neural features [71] may be helpful to answer this question . The second limiting factor of the current theory is the use of a perturbative approach to quantify the response of the integrate-and-fire model . Although the steady-state firing rate of the network is found as the fixed point of the nonlinear self-consistency equation , the response to a synaptic fluctuation is determined up to linear order in the amplitude of the afferent rate fluctuation , which is only valid for sufficiently small fluctuations . For larger input fluctuations , nonlinear contributions to the neural response can become more important [33] . Also for strong synaptic coupling , deviations from our theory are to be expected . Thirdly , the employment of Fokker-Planck theory to determine the steady-state firing rate and the response to incoming fluctuations assumes uncorrelated presynaptic firing with Poisson statistics and synaptic amplitudes which are vanishingly small compared to the distance between reset and threshold . For larger synaptic amplitudes , the Fokker-Planck theory becomes approximate and deviations are expected [33] , [34] , [72] , [73] . This can be observed in Fig . 7 A , showing a deviation between the self-consistent firing rate and the analytical prediction at about . In this work , we obtained a sufficiently precise self-consistent approximation of the correlation coefficient by relating the random recurrent network of spiking neurons in the asynchronous irregular state to a reduced linear model which obeys the same relation between and up to linear order . This reduced linear model , however , does not predict the absolute values of the variance and covariance . The variance of the LIF model , for example , is dominated by nonlinear effects , such as the reset mechanism after each action potential . Previous work [12] , [31] has shown that the single spike-train statistics can be approximated in the diffusion approximation if the recurrent firing rate in the network is determined by mean-field theory . One may therefore extend our approach and determine the integral auto-correlation function as with the Fano factor ( see [51] ) . For a renewal process and long observation times , the Fano factor is given by [74] , [75] . The coefficient of variation can be obtained from the diffusion approximation of the membrane-potential dynamics ( App . A . 1 in [12 ) . The covariance can then be determined by ( 19 ) . Another possibility is the use of a refractory-density approach [76] , [77] . The spike-train correlation as a function of the time lag is an experimentally accessible measure . Future theoretical work should therefore also focus on the temporal structure of correlations in recurrent networks , going beyond zero-lag correlations [15] , [17] and the integral measures studied in the current work . This would allow to compare the theoretical predictions to direct experimental observations in a more detailed manner . Moreover , the relative spike timing between pairs of neurons is a decisive property for Hebbian learning [78] in recurrent networks , as implemented by spike timing-dependent plasticity [62] , and suspected to play a role for synapse formation and elimination [79] . The simulation experiments performed in this work revealed that the suppression of correlations is vulnerable to certain types of manipulations of the feedback loop . One particular biological source of additional variability in the feedback loop is probabilistic vesicle release at synapses [80] . In feedforward networks , such unreliable synaptic transmission has been shown to decrease the transmission of correlations by pairs of neurons [22] . Stochastic synaptic release is very similar to the replacement of the population activity in the feedback branch by a rate modulated Poisson processes that conserves the population rate . In these simulations we observed an increase of correlations due to the additional noise caused by the stochastic Poisson realization . Future work should investigate more carefully which of the two opposing effects of probabilistic release on correlations dominates in recurrent networks . The results of our study do not only shed light on the decorrelation of spiking activity in recurrent neural networks . They also demonstrate that a standard modeling approach in theoretical neuroscience is problematic: When studying the dynamics of a local neural network ( e . g . a “cortical column” ) , it is a common strategy to replace external inputs to this neural population by spike-train ensembles with some predefined statistics , e . g . by stationary Poisson processes . Most neural systems , however , exhibit a high degree of recurrence . Nonlocal input to the population , i . e . input from other brain areas , therefore has to be expected to be shaped by the activity within . The omission of these feedback loops can lead to qualitatively wrong predictions of the population statistics . The analytical results for the correlation structure of recurrent networks presented in this study provide the means to a more realistic specification of such external activity .
In the present study , we consider two types of sparsely connected random networks: networks with purely inhibitory coupling ( “I networks” ) and networks with both excitatory and inhibitory interactions ( “EI networks” ) . To illustrate the main findings of this study and to test the predictions of the linear model described in “Methods: Linearized network model” , both architectures were implemented as networks of leaky integrate-and-fire ( LIF ) neurons . The model details and parameters are reported in Table 1 and Table 2 , respectively . All network simulations were carried out with NEST ( www . nest-initiative . org , [81] ) . In this section we show how the dynamics of the spiking network can be reduced to an effective linear model with fluctuations fulfilling , by construction , the same relationship as the original system up to linear order . We first outline the conceptual steps of this reduction , and then provide the formal derivation . We make use of the observation that the effect of a single synaptic impulse on the output activity of a neuron is typically small . Writing the response spike train of a neuron as a functional of the history of all incoming impulses therefore allows us to perform a linearization with respect to each of the afferent spike trains . Formally , this corresponds to a Volterra expansion up to linear order , the generalization of a Taylor series to functionals . In “Methods: Response kernel of the LIF model” , we perform this linearization explicitly for the example of the LIF model . This determines how the linear response kernel depends on the parameters of the LIF model . The linear dependence on the input leads to an approximate convolution equation ( 31 ) linearly connecting the auto- and the cross-correlation functions in the network . As this equation is complicated to solve directly , we introduce a reduced linear model ( 35 ) obeying the same convolution equation . The reduced linear model can be solved by standard Fourier methods and yields an explicit form for the covariance matrix in the frequency domain ( 37 ) . The diagonal and off-diagonal elements of the dimensional covariance matrix in ( 56 ) correspond to the power-spectra of individual neurons and the cross-spectra of individual neuron pairs , respectively . As , in this linear approximation , both the auto- and the cross-covariances are proportional to the variance of the driving noise , the resulting correlation coefficients are independent of the noise amplitude ( see “Methods: Population averaged correlations in the linear EI network” ) . As shown in “Results: Suppression of population-activity fluctuations by negative feedback” and “Results: Population-activity fluctuations in excitatory-inhibitory networks” , the suppression of fluctuations in recurrent networks is most pronounced at low frequencies . It is therefore sufficient to restrict the discussion to the zero-frequency limit . Note that the zero-frequency variances and covariances correspond to the integrals of the auto- and cross-correlation functions in the time domain . In this limit , we may combine the two different sources of fluctuations caused by the spiking of the neurons and by external input to the network into a single source of white noise with variance ( see ( 39 ) ) . In general , the spiking activity of neuron at time is determined by the entire history of the activity of all neurons in the network up to time . Formally , this dependence can be expressed by a functional The subscript in indicates that ( causality ) . In the following , we use the abbreviation . The effect of a single synaptic input on the state of a neuron is typically small . We therefore approximate the influence of an incoming spike train on the activity of the target neuron up to linear order . The sensitivity of neuron 's activity to the input from neuron can be expressed by the functional derivative of with respect to input spike train : It represents the response of the functional to a single -shaped perturbation in input channel at time , normalized by the perturbation amplitude . In ( 27 ) , denotes the unity vector with elements and for all . By introducing the vector of spike trains with the -th component set to zero , can be approximated by Eq . ( 28 ) is a Volterra expansion up to linear order , the formal extension of a Taylor expansion of a function of variables to a functional , truncated after the linear term . With the linearized dynamics ( 28 ) , the pairwise spike-train cross-correlation function between two neurons and is given by Note that ( 29 ) is valid only for positive time lags , because for a possible causal influence of on is not expressed by the functional . Here , denotes the average across the ensemble of realizations of spike trains in the stationary state of the network ( e . g . the ensemble resulting from different initial conditions ) , and the centralized ( zero mean ) spike train . In the last line in ( 29 ) , the average is split into the average across all realizations of spike trains excluding and the average across all realizations of . Note that the latter does not affect the functional derivative because it is , by construction , independent of the actual realization of . A consistent approximation up to linear order is equivalent to the assumption that for all the linear dependence of the functional on is completely contained in the respective derivative with respect to ( 28 ) . Dependencies beyond linear order include higher-order derivatives and are neglected in this approximation . This is equivalent to neglecting the dependence of on for any . Hence , we can average the inner term over and separately . In the stationary state , this correlation can only depend on and equals the auto- or the cross-correlation function: The pairwise spike-train correlation function is therefore given by where we used the fact that for any functional that does not depend on . The average of the functional derivative has the intuitive meaning of a response kernel with respect to a -shaped perturbation of input at time . Averaged over the realizations of the stationary network activity this response can only depend on the relative time . In a homogeneous random network , the input statistics ( number of synaptic inputs and synaptic weights ) and the parameters of the internal dynamics are identical for each cell , so that the temporal shape of the response kernel can be assumed to be the same for all neurons . The synaptic coupling strength from neuron to neuron determines the prefactor : In this notation , the linear equation connecting the auto-correlations and the cross-correlations takes the form Eq . ( 31 ) can be solved numerically or by means of Wiener-Hopf theory taking the symmetry into account [82] . Our aim is to find a simpler model which is equivalent to the LIF dynamics in the sense that it fulfills the same equation ( 31 ) . Let's denote the vector of dynamic variables of this reduced model . Analog to the original model , we define the cross-correlation for and as The simplest functional consistent with equation ( 31 ) is linear in . Since we require equivalence only with respect to the ensemble averaged quantities , i . e . , the reduced activity and therefore can contain a stochastic element which would disappear after averaging . The linear functional with a pairwise uncorrelated , centralized white noise ( ) fulfills the requirement , since for and This equation has the same form as ( 31 ) , so both models , within the linear approximation , exhibit an identical relationship between the auto- and cross-covariances . The physical meaning of the noise is the variance caused by the spiking of the neurons . The auto-correlation function of a spike train of rate has a -peak of weight . The reduced model ( 33 ) exhibits such a -peak if we set . A related approach has been pursued before ( see Sec . 3 . 5 in [31] ) to determine the auto-correlation of the population averaged firing rate . This similarity will be discussed in detail below . So far , we considered a network without external drive , i . e . all spike trains originated from within the network . If the network is driven by external input , each neuron receives , in addition , synaptic input from neurons outside the network . We assume uncorrelated external drive . In the reduced model , this input constitutes a separate source of noise: Here , denotes the convolution and the response kernel with respect to an external input . For simplicity , let's assume that the shape of these kernels is identical for all pairs of pre- and postsynaptic sources , i . e . . If we further absorb the synaptic amplitude of the external drive in the strength of the noise , the linearized dynamics ( 34 ) can be written in matrix notation with . The reduced model ( 35 ) can be solved directly by means of Fourier transform: The full covariance matrix follows by averaging over the sources of noise and as The diagonal elements of represent the auto-covariances , the off-diagonal elements the cross-covariances . Both are proportional to the driving noise . This is consistent with ( 31 ) which is a linear relationship between the cross- and auto-covariances . For networks which can be decomposed into homogeneous subpopulations , the dimensional system ( 35 ) can be further simplified by population averaging . Consider , for example , a homogeneous random network with purely inhibitory coupling . Assume that the neurons are randomly connected with probability and coupling strength . The average number of in/outputs per neuron ( in/out-degree ) is thus given by . By introducing the population averaged external input , the averaged spiking noise , and the effective coupling strength , the dynamics of the population averaged activity becomes Here we assumed that is independent of the presynaptic neuron and can be replaced by . Note that this replacement is exact for networks with homogeneous out-degree , i . e . if the number of outgoing connections is identical for each neuron . For large random networks with binomially distributed out-degrees ( e . g . Erdös-Rényi networks or random networks with constant in-degree ) , ( 38 ) serves as an approximation . To relate our approach to the treatment of finite-size fluctuations in [31] , consider the population-averaged dynamics ( 38 ) of a single population with mean firing rate . We set for all single neuron noises in order for the reduced model's auto-covariances to reproduce the -peak of the spiking dynamics . In the population averaged dynamics , this leads to the variance of the noise given by . This agrees with the variance of the population rate in [31] . Therefore , the dynamics of the population averaged quantity in ( 38 ) agrees with the earlier definition of a population averaged firing rate for the spiking network [31] . In equation ( 38 ) , two distinct sources of noise appear: The noise due to external uncorrelated activity and the noise which is required to obtain the -peak of the auto-correlation functions of the reduced model . The qualitative results of “Results: Suppression of population-activity fluctuations by negative feedback” and “Results: Population-activity fluctuations in excitatory-inhibitory networks” , however can be understood with an even simpler model . As we are mainly concerned with the low-frequency fluctuations , we only need a model with the same limit . As we normalized the kernel so that we can combine both sources of noise and require in ( 36 ) in the zero frequency limit . Hence , in “Results: Suppression of population-activity fluctuations by negative feedback” and “Results: Population-activity fluctuations in excitatory-inhibitory networks” , we consider the model with a pairwise uncorrelated centralized white noise to explain the suppression of fluctuations at low frequencies . As a second example , consider a random network composed of an excitatory and an inhibitory subpopulation and with population sizes and , respectively . Assume that each neuron receives excitatory and inhibitory inputs from and with coupling strengths and , respectively , and probability , such that the average excitatory and inhibitory in/out-degrees are given by and , respectively . The dynamics of the subpopulation averaged activities is given by ( 35 ) with subpopulation averaged noise and and effective coupling Here , denotes the effective coupling strength , the effective balance parameter and and the ( sub ) population averaged external and spiking sources of noise , respectively . Again , the reduction of the -dimensional linear dynamics to the two-dimensional dynamics ( 40 ) is exact if the out-degrees are constant within each subpopulation . As before , both sources of noise can be combined into a single source of noise , if we are only interested in the low-frequency behavior of the model , leading to the dynamics ( 39 ) with the effective coupling ( 40 ) . The linear theory is only valid in the domain of its stability , which is determined by the eigenvalue spectrum of the effective coupling matrix . For random coupling matrices , the eigenvalues are located within a circle with a radius equal to the square root of the variance of the matrix entries [69] . Writing the effective dynamics for the exponential kernel as a differential equation , the eigenvalues of the right hand side matrix are confined to a circle centered at in the complex plain with radius . Given , eigenvalues might exist which have a positive real part , leading to unstable dynamics . This condition is indicated by the vertical dotted lines in Fig . 6 A–F and Fig . 7 B–D near . Beyond this line , the linear model predicts an explosive growth of fluctuations . In the LIF-network model , an unbounded growth is avoided by the nonlinearities of the single-neuron dynamics . We now perform the formal linearization ( 30 ) for a network of LIF neurons . A similar approach has been employed in previous studies to understand the population dynamics in these networks [12] , [31] . We consider the input received by neuron from the local network , where denotes the spike train of the neuron projecting to neuron with synaptic weight . Given the time dependent firing rate of each afferent , and assuming small correlations and small synaptic weights , the total input to neuron can be replaced by a Gaussian white noise with mean and variance , where sums over all synaptic inputs . denotes the amplitude of the postsynaptic potential evoked by synapse . is the membrane time constant of the model . In the stationary state , the firing rate of each afferent is well described by the constant time average . The working point at which we perform the linearization of the neural response ( 30 ) is then given by analog equations as ( 41 ) , resulting in a constant mean and variance . If the amplitude of each postsynaptic potential is small compared to the distance of the membrane potential to threshold , the dynamics of the LIF model can be approximated by a diffusion process , employing Fokker-Planck theory [83] . The stationary firing rate of the neuron is then given by [12] , [31] , [84] with the reset voltage , the threshold voltage and the refractory time . In homogeneous random networks , the stationary rate ( Fig . 7 A ) is the same for all neurons . It is determined in a self-consistent manner [12] as the fixed point of ( 42 ) . The stationary mean and variance are determined by the stationary rate . To determine the kernel ( 30 ) we need to consider how a -shaped deflection in the input to this neuron at time point affects its output up to linear order in the amplitude of the fluctuation . In the stationary state , we may set . It is therefore sufficient to focus on the effect of a single fluctuation We therefore ask how the density of spikes per time of neuron , averaged over different realizations of the remaining inputs to neuron , changes in response to the fluctuation ( 43 ) of the presynaptic neuron in the limit of vanishing amplitude . This kernel ( 30 ) is identical to the impulse response of the neuron and can directly be measured in simulation by trial averaging over many responses to the given -deflection ( 43 ) in the input ( see Fig . 8 A ) . For the theory of low-frequency fluctuations , we only need the integral of the kernel , also known as the DC susceptibility , The second equality follows from the equivalence of the integral of the impulse response and the step response in linear approximation [21] , [33] . Following from [41] , both mean and variance are perturbed as and in response to a step in the afferent rate . Moreover , we used the chain rule . The variation of the afferent firing rate hence co-modulates the mean and the variance and both modulations need to be taken into account to derive the neural response [31] . Although the finite amplitude of postsynaptic potentials has an effect on the response properties [33] , [34] , the integral response is rather insensitive to the granularity of the noise [33] . We therefore employ the diffusion approximation to linearize the dynamics of the LIF neuron around its working point characterized by the mean and the variance of the total synaptic input . In ( 44 ) , we evaluate the partial derivatives of with respect to and using ( 42 ) . First , observe that by chain rule . We then again make use of the chain rule . Analog expressions hold for the derivative with respect to . The first derivative yields , the one with respect to follows analogously , but with a negative sign . We further observe that and with . Taken together , we obtain the explicit result for ( 44 ) Note that the modulation of results in a contribution to that is linear in , whereas the modulation of causes a quadratic dependence on . This expression therefore presents an extension to the integral response presented in [21] , [85] . Fig . 8 B shows the comparison of the analytical expression ( 45 ) and direct simulation . The agreement is good over a large range of synaptic amplitudes in the case of constant background noise caused by small synaptic amplitudes ( here for excitation and for inhibition ) . For background noise caused by stronger impulses , the deviations are expected to grow [33] . The recurrent linear neural dynamics defined in the previous section is conveniently solved in the Fourier domain . The driving external Gaussian white noise is mapped to the response by means of the transfer matrix . According to ( 39 ) , it is given by . The covariance matrix in the frequency domain , the spectral matrix , thus reads where we used and the expectation operator represents an average over noise realizations . To identify the effect of recurrence on the network dynamics , we replace the local feedback input by a feedforward input with spectral matrix . The resulting response firing rate is given by . Assuming that the feedforward input is uncorrelated to the external noise source ( ) yields a response spectrum In the Fourier domain , the solution of the mean-field dynamics ( 38 ) of the inhibitory network is . The power-spectrum hence becomes using the spectrum of the noise . We compare this power-spectrum to the case where the feedback loop is opened , i . e . where the recurrent input is replaced by feedforward input with unchanged auto-statistics , but which is uncorrelated to the external input . The resulting power-spectrum is given by ( 47 ) as . In a homogeneous random network of excitatory and inhibitory neurons , the population averaged activity ( 40 ) can be solved in the Schur basis ( 9 ) introduced in “Results: Population-activity fluctuations in excitatory-inhibitory networks” with and . The power of the population rate therefore is The fluctuations of the excitatory and the inhibitory population follow as So the power-spectra are Replacing the recurrent input of the sum activity by activity with the same auto-statistics , but which is uncorrelated to the remaining input into ( Fig . 5 D′ ) results in the fluctuations The power-spectrum of the sum activity therefore becomes If , alternatively , the excitatory and the inhibitory feedback terms and are replaced by uncorrelated feedforward input and with power-spectra and ( Fig . 5 C , D ) , the spectrum of the sum activity reads The limit ( 14 ) for inhibition dominated networks with can be obtained from this and the former expressions by taking and assuming strong coupling . In this subsection , we derive a self-consistency equation for the covariances in a recurrent network . We start from ( 37 ) ( we drop the superscript of for brevity ) multiply by from left and its transpose from right to obtain We assume a recurrent network of excitatory and inhibitory neurons , in which each neuron receives excitatory inputs of weight and inhibitory inputs of weight drawn randomly from the presynaptic pool of neurons . To obtain a theory for the variances and covariances at zero frequency ( with ) we may abbreviate by . For a population averaged theory , we need to replace in ( 56 ) the variances of an individual neuron by the population average and replace the covariance for a given pair of neurons by the average over pairs that are statistically equivalent to . For a pair of neurons we will show that the set of equivalent pairs depends on the current realization of the connectivity since unconnected pairs are not equivalent to connected ones . Therefore it is necessary to first average the covariance matrix over statistically equivalent neuron pairs given a fixed connectivity and to subsequently average over all possible realizations of the connectivity . The latter will be denoted as . For compactness of the notation , first we perform the averaging for the general case , where neuron belongs to population and neuron to population . We denote by , the sets of neuron indices belonging to populations and , respectively . Subsequently replacing and by all possible combinations , we obtain the averaged self-consistency equations for the network . We denote the number of incoming connections to a neuron of type from the population of neurons of type as and the strength of a synaptic coupling as . Rewriting the self-consistency equation ( 56 ) explicitly with indices yields The last equation shows that for a connected pair of neurons ( or ) either of the first two sums contains a contribution or proportional to the variance of the projecting neuron . We therefore need to perform the averaging separately for connected and for unconnected pairs of neurons . We use the notation for the average covariance over pairs of neurons of types with a connection from neuron to neuron , where is the number neuron pairs connected in this way . An arrow to the right , , denotes a connection from neuron to neuron . Note that we use the same letter for the population averaged covariances and for the covariances of individual pairs . The distinction can be made by the indices: throughout indexes a single neuron , identifies one of the populations . We denote the covariance averaged over unconnected pairs as We further use for the integrated variance averaged over all neurons of type . Connected and the unconnected averaged covariances differ by the term proportional to the variance of the projecting neuron , as mentioned above As a consequence , we can express all quantities in terms of the averaged variance ( 60 ) and the covariance averaged over unconnected pairs ( 59 ) . We now proceed to average the integrated variance over population . Since there are no self-connections in the network , we do not need to distinguish two cases here . Replacing on the right hand side of ( 60 ) , the first term of ( 57 ) contributes From the second to the third step we used that the sum over ( ) yields non-zero contributions only if neuron ( ) connects to neuron . This happens in ( ) cases with the coupling weight ( ) . Therefore the covariance averaged over connected pairs appears on the right hand side . In the last line we used the relation ( 61 ) to express the connected covariance in terms of the variance and the covariance over unconnected pairs . The second term in ( 60 ) is identical because of the symmetry . Up to here , the structure of the network only entered in terms of the in-degree of the neurons . The contribution of the third term follows from a similar calculation From the second to the third step we assumed that among the pairs of neurons projecting to neuron , the fraction has a connection . These pairs contribute with the connected covariance . The connections in opposite direction contribute the other term of similar structure . We ignore multiple and reciprocal connections here , assuming the connection probability is low . We introduce the shorthand for the covariance averaged over all neuron pairs including connected and unconnected pairs This is the covariance which is observed on average when picking a pair of neurons of type and randomly . In this step , beyond the in-degree , the structure of the network entered through the expected number of connections between two populations . Taken all three terms together , we arrive at The averaged covariances follow by similar calculations . Here we only need to calculate the average over unconnected pairs given by ( 59 ) , because the connected covariance follows from ( 61 ) . The first sum in ( 57 ) contributes where due to the absence of a direct connection between and , the term linear in the coupling and proportional to the variance is absent . From the symmetry it follows that the second term corresponds to an exchange of and in the last expression . The third sum in ( 57 ) follows from an analog calculation as before In summary , the contributions from ( 66 ) and ( 67 ) together result in the self-consistency equation for the covariance We now simplify the expressions by assuming that the in-degree of a neuron and the incoming synaptic amplitudes do not depend on the type of the neuron , i . e . that excitatory and inhibitory neurons receive statistically the same input . Formally this means that we need to replace by , the number of incoming connections from population and by , the coupling strength of a projection from a neuron of type . The covariance then has two distinct contributions , that depends on the type of neurons , and that does not . In particular and do not depend on and we omit their subscripts in the following . The variances fulfill the covariances satisfy The disjoint part determines the difference between the covariances for pairs of neurons of different type . Using the parameters , , , , the explicit form is Therefore , also the covariances in the network obey the relation i . e . the mixed covariance can be eliminated and is given by the arithmetic mean of the covariances between neurons of same type . In matrix representation with the vector , the self-consistency equation is The self consistent covariance can then be obtained by solving the system of linear equations The numerical solution shows that the variances for excitatory and inhibitory neurons are approximately the same , as depicted in Fig . 6 A . In the following we therefore assume and then solve ( 76 ) for the covariances . With the abbreviation , the third and fourth line yields the equation for the covariances The structure of the equation suggests to introduce the linear combination which satisfies We solve ( 77 ) for and and insert ( 78 ) for to obtain the covariances as The covariance between unconnected neurons can be related to the covariance between the incoming currents this pair of neurons receives . Expressing the self-consistency ( 68 ) in terms of the covariances averaged over connected and unconnected pairs ( 64 ) uncovers the connection This self-consistency equation yields the argument , why the shared-input correlation ( 19 ) cancels the contribution ( 20 ) due to spike-train correlations in the covariance to the input currents ( see Fig . 6 C , D ) . Rewriting ( 80 ) in terms of these quantities results in If a self-consistent solution with small correlation exists , the right hand side of ( 81 ) must be of the same order of magnitude . The right hand side of this equation has a prefactor which typically is ( for the parameters in Fig . 6 , becomes larger than for ) . The first term in the bracket is proportional to the contribution of shared input , the second term is due to correlations among pairs of different neurons . Each of these terms is of order . Due to the prefactor , however , the sum of the two terms needs to be of order to fulfill the equation . Hence , the terms must have different signs to cause the mutual cancellation . To illustrate how the correlation structure is affected by feedback , let us now consider the case where the feedback activity is perturbed ( “feedforward scenario” ) . We start from ( 47 ) and , again , only consider the fluctuations at zero frequency , First , we consider a manipulation that preserves the single-neuron statistics , and the pairwise correlations , within each subpopulation , but neglects correlations between excitatory and inhibitory neurons . Formally , this corresponds to the block diagonal correlation matrix Here , we have replaced the individual entries of the correlation matrix by the corresponding subpopulation averaged correlations . The calculation of the response auto- and cross-correlation and is similar as for the expressions ( 63 ) and ( 67 ) , with the difference that terms containing are absent: As an alternative type of feedback manipulation , we assume that all correlations are equal , irrespective of the neuron type . To this end , we replace all spike correlations by the population average . Thus , the covariance matrix reads The calculation follows the one leading to the expressions ( 63 ) and ( 67 ) and results in
|
The spatio-temporal activity pattern generated by a recurrent neuronal network can provide a rich dynamical basis which allows readout neurons to generate a variety of responses by tuning the synaptic weights of their inputs . The repertoire of possible responses and the response reliability become maximal if the spike trains of individual neurons are uncorrelated . Spike-train correlations in cortical networks can indeed be very small , even for neighboring neurons . This seems to be at odds with the finding that neighboring neurons receive a considerable fraction of inputs from identical presynaptic sources constituting an inevitable source of correlation . In this article , we show that inhibitory feedback , abundant in biological neuronal networks , actively suppresses correlations . The mechanism is generic: It does not depend on the details of the network nodes and decorrelates networks composed of excitatory and inhibitory neurons as well as purely inhibitory networks . For the case of the leaky integrate-and-fire model , we derive the correlation structure analytically . The new toolbox of formal linearization and a basis transformation exposing the feedback component is applicable to a range of biological systems . We confirm our analytical results by direct simulations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"circuit",
"models",
"computational",
"neuroscience",
"biology",
"computational",
"biology",
"coding",
"mechanisms"
] |
2012
|
Decorrelation of Neural-Network Activity by Inhibitory Feedback
|
Every year about 3 million tourists from around the world visit Brazil , Argentina and Paraguay´s triple border region where the Iguaçu Falls are located . Unfortunately , in recent years an increasing number of autochthonous canine and human visceral leishmaniasis ( VL ) cases have been reported . The parasite is Leishmania ( Leishmania ) infantum and it is transmitted by sand flies ( Phlebotominae ) . To assess the risk factors favorable for the establishment and spread of potential vectors the Centers for Disease Control and Prevention light trap ( CDC-light trap ) collections were made in the Foz do Iguaçu ( FI ) and Santa Terezinha de Itaipu ( STI ) townships and along two transects between them . Our study determined the Phlebotominae fauna , the factors that affect the presence and abundance of Lutzomyia longipalpis and Nyssomyia whitmani , the presence of L . infantum in different sand fly species and which Leishmania species are present in this region . Lutzomyia longipalpis was the prevalent species and its distribution was related to the abundance of dogs . Leishmania infantum was found in Lu . longipalpis , Ny . whitmani , Ny . neivai and a Lutzomyia sp . All the results are discussed within the Stockholm Paradigm and focus on their importance in the elaboration of public health policies in international border areas . This region has all the properties of stable VL endemic foci that can serve as a source of the disease for neighboring municipalities , states and countries . Most of the urban areas of tropical America are propitious for Lu . longipalpis establishment and have large dog populations . Pan American Health Organization´s initiative in supporting the public health policies in the border areas of this study is crucial and laudable . However , if stakeholders do not act quickly in controlling VL in this region , the scenario will inevitable become worse . Moreover , L . ( Viannia ) braziliensis found in this study supports the need to develop public health policies to avoid the spread of cutaneous leishmaniasis . The consequences of socioeconomic attributes , boundaries and frontiers on the spread of diseases cannot be neglected . For an efficient control , it is essential that urban planning is articulated with the neighboring cities .
Brazil is ranked among the countries with the highest number of Visceral Leishmaniasis ( VL ) cases in the world , with 92% of the South American cases ( i . e . , 3 , 289 cases in 2015 ) , and 43% of the people at risk of VL [1 , 2] . Increasing numbers of cases in the country are resulting from the adaptation of rural transmission to urban , a condition that has occurred in many Brazilian regions , especially in the Southeast and Midwest regions of the country [2 , 3] . Cities such as Belo Horizonte ( Minas Gerais state , MG ) , Araguaína ( Tocantins state , TO ) , Campo Grande ( State of Mato Grosso do Sul state ) , Bauru ( Sao Paulo state , SP ) , Palmas ( TO ) , represent 15% of the VL cases [4] . Consequently , there was an unprecedented acceleration in speed of VL dispersion during the 1990’s and the beginning of the 21st century . In the southern region of Brazil , the disease has been notified in Rio Grande do Sul since the 2000’s [5] . Although signed in the neighbor countries in the beginning of this century [6–9] , the VL was not recorded until recently in the extreme-western of Paraná State , region corresponding to the triple border with Argentina and Paraguay . On the Brazilian side , the first record of Lu . longipalpis was in Foz do Iguaçu city ( FI ) in 2012 . Subsequently , canine visceral leishmaniasis ( cVL ) and human ( HVL ) cases were firstly reported in 2013 and 2016 respectively , and the number of VL cases has been increased in the following years [10–12] . For instance , Thomaz-Soccol et al . [13] showed that about 23 . 8% dogs in FI presents cVL . In this way , delineating the sand fly fauna´s distribution is essential to understand adequately VL´s status in the region , and to reduce the risk of future epizootics . If Lu . longipalpis is widely abundant in this region , there is a potential risk that the parasite cycle is established . Moreover , it is critical to assess which environmental conditions provide autochthonous transmission of VL and/or CL , once urban areas of FI and its neighboring cities are expanding in zones with large forest reserves . The study´s objectives are to assess the above described risks by addressing the following questions: 1 ) based on the distribution of phlebotomine sand fly species in the urban , peri-urban ( or ruro-urban ) and rural environment in the western region of Paraná , what are the environmental conditions that allow the installation of Lu . longipalpis populations ? 2 ) are different Leishmania species present sympatrically in this region ? 3 ) are other sand flies hosting L . infantum in areas with recent reports of VL ? 4 ) what is the prevalence of Leishmania spp . in these vectors ? This information is important to help understand the risk factors and scenarios that are propitious for the founding of enzootic foci . This work is in part belonging to the IDRC #107577 research project ( idrc . ca/en/project/addressing-emergence-and-spread-leishmaniasis-borders-argentina-brazil-and-paraguay ) .
The survey was undertaken in the extreme western region of Paraná state in southern Brazil , an area known as the triple border with Argentina and Paraguay ( Fig 1 ) . The three cities in this area , Ciudad del Este ( Paraguay ) , Puerto Iguazú ( Argentina ) , and Foz do Iguaçu ( Brazil ) , represent an urban area in the Atlantic forest with more than 700 , 000 inhabitants . We sampled the sand fly populations of the following Brazilian localities: Foz do Iguaçu ( FI ) , Santa Terezinha de Itaipu ( STI ) , and two transects ( T1 and T2 ) through the rural area between these two cities . The number of dogs in FI is about 57 , 000 ( according to the Zoonosis Control Center ( CCZ ) of Foz do Iguaçu ) , and about 5 , 700 in STI ( data obtained from Sanitaire Vigilance ) . Besides dogs and humans , mice , chickens and feral animals are present in the region , and are potential blood sources . Foz do Iguaçu municipality ( 25° 32' 52" S-54° 35' 17" W ) has an area of 615 . 02 km2 and 256 , 088 inhabitants—a population density of 416 . 38 inhabitants/ km2 [14] . The region is composed of a conservation area ( 22% of its area ) , a rural zone ( 22% ) , the Itaipu Lake ( 24% ) and urban areas ( 32% ) . Santa Terezinha de Itaipu ( 25° 21' 44" S-54° 29' 17" W ) has 22 . 783 inhabitants and a density of 80 . 35 inhabitants/km2 . The economy is based essentially on livestock , soybean and corn cultivation . A considerable part of the area is delimited by Itaipu Lake with areas of remnant forest . The transects were drawn between FI and STI , this area consists of a rural road that connects FI to STI and passes near to the Iguaçu National Park . Sand flies were collected between the 25th and 27th of October 2014 in FI , 21st and 24th November 2014 along the transects , and 19th and 21th October 2015 in STI using automatic CDC-Light traps set at 1 . 5 m above ground level between approximately 6:30 p . m . and 7:30 a . m . , during three consecutive rainless nights . Global Positioning System ( eTrex10 ) registered the coordinates of each trap . For collections in the FI and STI urban areas the cities were divided into a grid of 400 m2 ( patch ) squares ( for methodology see [15] ) . The FI presented 875 patches of 400 m2 along four areas of the city ( A , B , C and D ) . The A area corresponds to the administrative area of the city , with the highest density of buildings and near the Paraná River . It contains two forest remnants that occupy a large part of the locality . The B area is a residential region east of A and west of the rural areas . The density of the dwellings is medium and they were distributed almost evenly . The C area is located in the north of the city , bounded by regions A and B in the south , by the Paraná River in the north and west , and by rural areas in the east . The C area is the largest area and is characterized by discontinuous dwellings interrupted by large "green" spaces ( sports fields , small cultivated areas , etc . ) . The D area is located south of the city and regions A and B , is bordered by the Iguaçu ( south ) and Paraná ( west ) rivers . It has a relatively high density of buildings , with forest remnants that surround the banks of the rivers . Due to the limitation of number of available CDC-LT traps , 123 patches from FI ( 26–28 in the areas A , B , and D , and 42 in the area C , largest area ) were selected for the sampling according to the worst scenario criterion [15] . The “worst scenario” is a functional definition to denote a site within the study patch with the greatest probability of sand fly presence due to habitat conditions . These sites presents dense vegetation which provides shadow , humidity and detritus , soil rich in organic material and access to blood sources without the interference of external light . Minimum and maximum distances between traps settled in different patches were 145 and 475 m , respectively . Sand fly collections were performed in the urban and ruro-urban areas ( Fig 2 ) of STI . The city was also divided into a grid of 400 m2 patches and 33 patches were selected for the sampling according to the worst scenario . Between FI and STI , sand flies were collected along the two transects ( T1 and T2 ) ( Figs 1 and 2 ) , that consisted of two sectors in the ruro-urban areas of FI ( T1 = 17 km ) and the rural area of STI ( T2 = 17 . 5 km ) . Forty CDC-LTs were distributed along the transects , 20 in each sector ( T1 + T2 ) , at intervals of approximately 860 m . Four traps were also installed in the Iguaçu National Park in sites with extensive vegetation cover and in the Itaipu Biological Sanctuary . To collect sand flies , all the owners/residents that collaborated in the study were informed about the practices and signed the informed consent form . In the laboratory , the insects were anesthetized with chloroform and separated by sex . The males were stored at 4o C until morphological identification . The individuals were identified according to the morphological taxonomic keys proposed by Galati [16] . Each female was placed on sterile slides containing one drop of sterile saline solution ( 0 . 9% ) . Firstly , the head was sectioned with the aid of two sterile needles . Then , two needles were inserted in the individual , the first through the thorax and the second through the last two abdominal segments . A first linear and slow traction of the needles was performed to detach the last two segments , the spermathecae and ovarian ducts from the thorax . A second slow traction was performed to remove the digestive tract from the thorax . The cut of the anal duct allowed the complete separation of the digestive tract from the last two fragments . The female’s intestine was carried to a microtube with in ethanol 70% for posterior molecular assessing of the presence and the identification of Leishmania species . The remaining female ( head , thorax , first abdominal segments , spermathecae , ovarian ducts ) and the males ( kept whole ) were clarified using 10% KOH for the visualization of the internal structures as recommended by Rioux et al . [17] . Evandromyia cortelezzii and Evandromyia sallesi females cannot be distinguished by their morphology , thereby they were named as Evandromyia sp . The generic abbreviations of the species proposed by Marcondes [18] were used . Pools of intestines ( two or three females ) from the same trap were extract for assessing the presence of parasites’ DNA . The intestines were macerated in the microtube using a plastic pestle , with 497 μL of lysis buffer ( 100 mM Tris-HCl , 100 mM NaCl , 25 mM EDTA , 0 . 5% SDS , pH 8 . 0 ) . After incubating for 1 h at 55° C with 2% of proteinase K ( 10 mg/mL solution ) , the DNA was precipitated using the phenol-chloroform method [19] . The DNA pellet was resuspended in 20 μL of TE-buffer ( 10 mM Tris-HCl pH 8 . 0 , 1 mM EDTA ) and stored at -20° C . A subset of females ( about 70% , see results for details ) of all positive traps was selected for Leishmania detection according to Schönian et al . ´s [20] protocol . The PCR assays were carried out with 5 μL of DNA template of the female pools in a final reaction volume of 25 μL . The reaction contained 1 X PCR buffer ( 100 Mm Tris-HCl pH 8 . 0 , 0 . 1 mM EDTA , 1 mM DTT , 50% ( v/v ) glycerol ) ( Invitrogen ) , 2 . 5 mM MgCl2 ( Invitrogen ) , DMSO 2 . 5% ( Sigma ) , 200 μM of each dNTP ( Invitrogen ) , 0 . 5 μM LITSR forward primer , 0 . 5 μM L5 . 8S reverse primer and 1 . 4 U of Taq polymerase ( Invitrogen ) . The PCR cycle employed an initial denaturalization step of 94° C for 4 min , followed by 40 cycles of a denaturalization step of 94° C for 30 sec , a hybridization step of 56° C for 30 sec and an extension step of 72° C for 30 sec , and a final extension step of 72° C for 10 min . The cycle procedure was performed in a thermocycler biocycler , MJ96 ( Biosystems ) . DNA of L . infantum reference strain was used as positive control , and water as the negative control in each reaction . In addition , a 220 bp fragment of the constitutive gene IVS6 ( cacophony ) of sand flies was used as internal control [21] . All PCR products were assessed through 1 . 6% agarose gel electrophoresis ( 1 h at 5 V/cm ) , stained with ethidium bromide and visualized under ultraviolet light . The parasite identification was done by performing a RFLP analysis of the amplified ITS1 fragment as proposed by Schonian et al . [20] . The PCR products ( 10–15 μL ) were digested with HaeIII accordingly , using conditions recommended by the supplier ( Hybaid GmbH Heidelberg , Germany ) . The restriction fragments were submitted to electrophoresis in 2% metaphor agarose ( FMC BioProducts Rockland , ME , USA ) at 100 V in 0 . 5× TBE buffer and visualized under ultraviolet light after staining for 15 min in ethidium bromide ( 0 . 5 μg/mL ) . The electrophoresis pattern was compared with the reference strains ( i . e . , the main Leishmania species that cause CL and VL in the Brazil: L . infantum , L . braziliensis , L . amazonensis ) . Positive PCR-RFLP products with electrophoresis pattern different from the reference strains were commercially sequenced by the Macrogen Inc . ( Seoul , South Korea ) to confirm the identification . The electropherograms were manually checked , and the sequences were edited and assembled using BioEdit software and aligned with Guidance 2 . 0 [22] . The sequences were deposited in the GenBank ( accession numbers MG136689 to MG136700 ) . The identification of the sequenced Leishmania samples was obtained from a Bayesian tree constructed using Beast 1 . 8 . 4 [23] . Three runs of 30 million MCMC each ( 3 million of burn-in ) were performed with the substitution model ( Hasegawa-Kishino-Yano , HKY [24] ) defined by Modeltest 2 . 1 . 10 [25] . The blood meal of engorged female sand flies was identified using PCR based analysis on the cytochrome B gene . DNA was extracted from the intestine and its content using the phenol chloroform protocol [19] . The amplifications followed González et al . ‘s [26] protocol . Sequencing was performed commercially by Macrogen Inc . ( Seoul , South Korea ) . The electropherograms were checked manually and the sequences were edited and assembled using BioEdit software . The blood meals were identified using the nucleotide BLAST ( Basic Local Alignment Search ) tool on the GenBank . To assess the environmental variables that affect the presence and abundance of sand fly species , micro and meso scale factors were evaluated . Furthermore , interviews were conducted , by domicile , during the sand fly collection period . Several questions were recorded and environmental variables were measured . Twenty-nine variables were selected for this study ( Table 1 ) , including mean of the minimum and maximum temperatures , and relative humidity of each night . The Normalized Difference Vegetation Index ( NDVI ) was used to highlight the presence of vegetation in each studied area and the plant biomass [27] . Maps were constructed based on bands 6 ( B6 –near infrared band ) and 5 ( B5 –near red band , respectively ) of the LandSat 8 satellite . The months with largest volumes of precipitation are in March and the smallest precipitate volumes in August [28] . To generate the NDVI , three LandSat 8 satellite images were used from February , August , and October 2014 for FI and for transects . For STI the analyses , satellite images was acquired from the same months in 2015 . In the Geographic Information Systems ( GIS ) environment , the following formula was applied using bands 6 and 5: NDVI = [ ( B6 –B5 ) / ( B6 +B5 ) ] . The Normalized Difference Vegetation Index present values ranging from -1 to +1 . Values closer to +1 indicate a high presence of vegetation , and values closer to -1 represent absence of vegetation . The maps of forest fragments ( vegetal remnant ) were generated from the data of the Environment Ministry´s project for managing the Biodiversity of the Paraná River´s corridor [29] . Hypsometry and altimetry maps were generated based on the Shuttle Radar Topography Mission ( SRTM ) and Digital Terrain Model ( DTM ) , with spatial resolution of 30 m in the X band . The DTM was purchased from the United States Geological Survey ( USGS ) website . The maps with the hydrographic network were generated from the official database of the water bodies’ network of the State of Paraná in the scale 1:50 , 000 , in collaboration with several public institutions ( e . g . , SANEPAR , COMEC , ITAIPU , ITCG , Instituto das Águas do Paraná , IAP ) . The average annual and seasonal air temperature data were acquired based on Paula [30] , which generated a serial historical average ( ~30 years ) . The maps were processed using the open software QGIS version 2 . 18 . The relative abundance of the vectors was calculated as: number of Lu . longipalpis by peridomicile/total domicile surveyed [31] . Path Analyses were implemented with the package 'plspm' 0 . 4 . 7 [32] in the R 3 . 3 . 3 [33] for assessing the influence of each environmental variable ( Table 1 ) on sand fly abundance . In this analysis , the variables were grouped at mesoscale ( proportion in a plot of 250 m of land with trees , herbaceous , soil , urban , water and crops: heterogeneity ) , microscale ( proportion in a 25 m plot of land of trees , herbaceous , soil , urban , water and crops , presence of garbage and fallen leaves and fruits: heterogeneity ) , animal supply ( presence of chicks and rodents , and abundance of dogs and chickens ) , public services ( lighting and garbage collection ) , temperature and humidity ( humidity , minimum and maximum temperatures ) , and study areas ( FI , Transect and STI ) . The heterogeneity is the number of different ground cover ( i . e . , trees , herbaceous , soil , urban , water and crops ) in the area ( 25 and 250 m ) of each trap . Altitude , distance from the water and abundances of Lu . longipalpis and Ny . whitmani , were kept isolated . The influence of the environmental variables was calculated only to these two species because only they presented sufficient number of specimens’ ( n > 200 , see results for details ) . The influence of sand fly species abundance on other sand fly species was not tested since their abundances hypothetically do not affect directly the abundances of other species , and their collinearity can be explained due to the sharing of ecological characteristics ( phylogenetic conservatism ) . Additionally , the percentage of contribution of each variable on the presence of each species , in each trap , was assessed with a maximum entropy approach in the Maxent 3 . 4 . 1 [34 , 35] . The validation of the model was assessed using the area under ( AUC ) the receiver operating characteristic ( ROC ) curve . All sequences obtained in the present work , as well as sequence of the reference strain of L . infantum ( MHOM/FR/78/LEM 75 ) , L . braziliensis ( MHOM/BR/75/M2903 ) , and L . amazonensis ( MHOM/BR/73/M2269 ) were used in the Bayesian tree construction . In addition the sequences of L . infantum ( AJ634355 . 1 , FN398341 , GU045592 , KM925006 ) , L . braziliensis ( AJ300483 , FJ753382 , FN398337 ) and L . amazonensis ( AJ000314 , DQ182536 , FJ753373 ) deposed in the GenBank were added for the tree construction . Our isolates regrouped with L . infantum and L . braziliensis ( Fig 3 ) .
A total of 1 , 202 sand fly specimens were collected , 91 . 3% ( 1 , 097 ) of them in FI , 7 . 7% ( 93 ) along the two transects and 1% ( 12 ) in STI city ( Tab 2 ) . Sand flies were captured in 81 of 196 ( 41 . 3% ) patches . The predominant species was Lu . longipalpis ( in 41 traps , 21% ) . Nyssomyia whitmani was present in 34 traps ( 17% ) , and Ny . neivai in 20 ( 10% ) . A total of 1 , 097 sand flies , belonging to seven species , were captured in 44 . 7% of the FI´s 123 traps . Lutzomyia longipalpis was the most abundant species ( 55 . 7% ) in the three habitats ( rural , ruro-urban and urban ) of the municipality’s regions ( A–D ) ( Table 2 ) . Nyssomyia whitmani was present predominantly in sites close to forest remnant . In the two transects ( T1 + T2 ) , the most abundant species were Ny . whitmani ( 50 . 0% ) , Mg . migonei ( 19 . 4% ) and Ny . neivai ( 18 . 3% ) . Nyssomyia whitmani was dominant in both T1 and T2 regions ( 39 . 3 and 67 . 6% , respectively ) , followed by Ny . neivai ( 17 . 9 and 19 . 9% , respectively ) . Migonemyia migonei was present in the T1 ( 30 . 3% ) and T2 ( 2 . 7% ) , Lu longipalpis was not found in the transect captures ( Table 3 ) . In STI , sand flies were found in eight of the 33 traps surveyed ( Table 2 ) . Lutzomyia longipalpis was present in four quadrants ( 7 , 8 , 15 and 18 ) , Ny . whitmani in three quadrants ( 4 , 22 and 26 ) in the ruro-urban region , and Ny . neivai in 14 CDC-LT . The greatest relative abundances of Lu . longipalpis were in zones A ( 9 . 76 ) and C ( 7 . 02 ) in Foz do Iguaçu , and the lowest was in STI ( 0 . 18 ) . The greatest number of Ny . whitmani was recorded in the D zone in FI , followed by transect zones ( 2 . 35 ) ( Table 4 ) . The male to female ratio for Lu . longipalpis was 4 . 64:1 in FI and 0 . 75:1 in STI . For Ny . whitmani , sexual ratio was 1 . 80:1 in FI , 3 . 18:1 in the transect area and 0 . 67:1 in STI urban area ( Fig 2 ) . PCR-RFLP technique was applied on 179 of the 264 female sand flies collected in FI to identify the species of Leishmania that infect them . Among the 37 traps where female sand flies were collected , 12 ( 32 . 4% ) were responsible for the 64 infected females ( 35 . 8% of the total of females assessed ) ( Table 5 ) . A great part of these infections was found in four traps ( 43 . 1% in the trap 321; 21 . 5% in the trap 329; 9 . 8% in the traps 448 and 470 ) . The sequences of the 11 sand flies captured in FI obtained from the PCR-RFLP supported that two Lutzomyia sp . , a Ny . neivai and two Ny . whitmani were positive for L . infantum , and a Ny . neivai and a Ny . whitmani were positive for L . braziliensis ( Table 5 ) . Two Ny . whitmani and a Ny . neivai , from the transects were positive for L . braziliensis . Sixteen engorged Lu . longipalpis females were collected and their blood meal identified . The blood belonged vertebrate species ( Table 6 ) : Canis familiaris ( 6 females– 37 . 5% ) , Mus musculus ( 5 females– 31 . 3% ) , Homo sapiens ( 4 females– 25 . 0% ) and Dasypus novemcinctus ( 1 female– 6 . 2% ) . To understand the spatial distribution of the sand fly species , the most abundant species were chosen and separated by area ( Fig 2 ) . For FI , Lu . longipalpis was present in the four sampled areas , especially in the central ( region A ) and northwest ( region C ) areas . Nyssomyia whitmani was present especially in ruro-urban areas and in forest remnant . In the rural ( T1 + T2 ) areas , Ny . whitmani and Ny . neivai were sympatric . The greatest abundance was in the forest remnant and peridomicile areas in the Iguaçu National Park . In STI , the spatial distribution of the vectors indicated that sand fly fauna was greater near to the federal highway BR 277 , which crosses this city . The study of the environmental variables as hydrography , forest remnant , hypsometry , NDVI , and temperature were performed only for the most abundant species ( i . e . , Lu . longipalpis , Ny . whitmani ) . Lutzomyia longipalpis and Ny . whitmani ( Figs 4 and 5 , respectively ) were present in the areas close to water sources . This variable influences the microclimate of the region , especially humidity , precipitation and wind flows . In STI , sand flies were recorded both close to and distant from water bodies . However , similar to FI , STI is also influenced by the Itaipu reservoir , rivers and other aquatic environs . Although rivers are abundant in the transects , Lu . longipalpis was not found . The distribution of the vectors showed that Ny . whitmani was more abundant near the forest remnants and in the rural zone , including the National Park . On the other hand , Lu . longipalpis was present in urban areas irrespective to the presence of vegetation . In STI , Lu . longipalpis was captured in the downtown zone ( Figs 4 and 5 ) . It was present in areas whose altitude varied from 90 to 270 m above sea level ( masl ) , while in the STI the population was present in higher altitudes , between 250 to 290 masl . Nyssomyia whitmani was present between 90 and 346 masl in both FI and STI . The vegetation index ( NDVI ) indicated that Lu . longipalpis and Ny . whitmani were influenced by the forest corridors within the city of FI ( Figs 4 and 5 ) . Nyssomyia whitmani was positively influenced by the remnant forest in Iguaçu National Park , an Integral Protection Conservation Unit . The presence of forest remnant affected more Ny . whitmani than Lu . longipalpis . The greatest abundance was always near to areas with abundant vegetation . During the sand fly collection period ( three consecutive night ) , in FI , the average minimum temperatures registered were 19 . 0 , 22 . 3 and 21 . 8° C for each night , while maximum temperatures were 27 . 4 , 33 . 2 and 33 . 3° C , respectively . The average temperature for the last 30 years in FI , where Lu . longipalpis was recorded , ranged from 17 . 5 to 18 . 0° C in the winter , and from 25 . 5 to 26 . 2° C in the summer , and the annual temperature ranged from 21 . 8 to 22 . 5° C . In STI , the minimum temperature recorded in our sampling was 25 . 4° C and the maximum was 36 . 4° C . The humidity varied from 71 to 75% . In the winter , the temperature of this region ranged from 17 . 2 to 18 . 0° C , and from 25 . 2 to 26 . 2° C in summer . The Path modeling performed to test the influence of groups of variables on the sand fly abundance ( Lu . longipalpis , Ny . whitmani ) indicated that only the animal food supply group ( loadings: dog abundance = 0 . 68 , presence of rodents = 0 . 02 , presence of chicks = 0 . 52 , abundance of chickens = 0 . 50 ) significantly affects the abundances of Lu . longipalpis ( Path coefficient of 0 . 33 , p < 0 . 01 , R2 of the predictor model = 17% ) . The maximum entropy approach ( AUC = 0 . 834 , Table 1 ) supported that urban 250 m ( 14% ) , minimum temperature ( 13% ) , herbaceous 250 m ( 11% ) , maximum temperature and plenty of dogs ( 8% each ) are the main variables that contributes to the presence of Lu . longipalpis in the region . The abundance of Ny . whitmani was only affected by the temperature and humidity ( loadings: minimum temperature = 0 . 88 , maximum temperature = 0 . 71 , humidity = 0 . 17 ) ( Path coefficient of -0 . 42 , p < 0 . 01 , R2 of the predictor model = 9% ) ( Figs 6 and 7 ) . Similarly , the maximum entropy analysis ( AUC = 0 . 891 , Table 1 ) supported that the humidity ( 27% ) presented the large contribution to the presence of Ny . whitmani , followed by the presence of chicks ( 12% ) and herbaceous at 250 m ( 10% ) . The remaining variables contributed less than 10% to the presence of the species .
Regarding environmental issues , it should be noted that all human activity generates environmental impacts , which compromise the equilibrium and the existing state of an environment . These impacts are man-made according to their needs , which vary in intensity and speed over time [53] . Considering the health concerns , these issues condition and/or intensify environmental vulnerabilities , thus amplifying the possibility of establishing certain types of diseases . The risk generated by man comes from the change in opportunity associated with the modernization process , generating technological and organizational innovations . These factors promote the production of wealth , inequalities and , consequently , unequal environmental risks in the society , if analyzed in micro scales . On the other hand , the risks generated by socioeconomic inequalities can produce a boomerang effect , since people from different social classes can suffer from the consequences of the modernization processes , such as infectious diseases related to elevated demographic concentration . In the last two decades , Brazil has been experiencing the expansion and urbanization of VL not only in large , such as Belo Horizonte and Campo Grande , but also in medium-size cities such as Araçatuba and Governador Valadares . The process of VL´s urbanization has been registered in four of Brazil´s five regions: Northeast ( São Luís , Natal and Aracaju ) , North ( Boa Vista and Santarém ) , Southeast ( Belo Horizonte and Montes Claros ) and Center West ( Cuiabá and Campo Large ) [54] . These scenarios have many common features that may be related to the dispersion of sand fly species . Brazil's agrarian structure underwent deep modifications , associated with environmental and climate changes , reduction of investments in health and education , discontinuity of control actions , adaptation of the vectors to man-made environments , and difficulties in disease control in large urban agglomerates . In the triple frontier , it is important to emphasize that the issues are not only established in the socioeconomic sphere , because the study region presents a great ethnic-cultural discontinuity [55] . The social relationships of immigrants with the country of origin continue even if the individual has moved a long time ago , thus confronting the way of life of the city in which he has established himself [56] . The way of life of the population constitutes a decisive factor for the sanitary campaigns of waste control . Public authorities accept that a population´s way of life can be changed using information from long-term studies . However , this is complicated in cities where there is a high turnover of the inhabitants coming from neighbouring countries . These conditions can make it more difficult to execute successfully preventive health campaigns . This is where articulated urban planning between neighbouring cities , such as the three of the triple border can help to promote conjoint actions in the fight against vector borne diseases , such as the leishmaniases . There are regional and federal policies that are especially linked to MERCOSUL , that advocate intersectoral and intermunicipal articulation . However , there are problems in creating and maintaining programs for the prevention of vector-borne diseases that function in international border areas . It must be emphasized that for border municipalities , health planning needs to project beyond the municipal , state , and federal administrative limits . The emergence of autochthonous VL in Paraná , more specifically in FI and STI , raises the concern of how international borders relate to the appearance of diseases . Farmer [55] , stated that the ineffectiveness of political boundaries helps the spread of pathogenic microorganisms . For this author , political boundaries function as semipermeable membranes , open to the frequency of circulation of diseases , and closed to the free circulation of medicines . This fact demonstrates how countries' governments still do not consider their borders as a gateway to communicable diseases . The dynamics of many emerging diseases are not established within a single national territory and control is insufficient if realized by a single nation-state . However , the author does not ignore the effect of boundaries and frontiers in the spread of diseases and proposes further studies on intercountry borders [57] . The social inequalities of neighbouring countries must be considered beyond a critical sociology which can define the true transnational borders of pandemics . Summary: We have shown that endemic visceral leishmaniases are present on the Brazilian side of triple border . There is therefore a potential risk that this disease could spread to other municipalities of Paraná state , as well as to the neighbouring states or countries , resulting in the installation of new foci . Many urban areas of tropical America are propitious for the establishment of Lu . longipalpis and without exception they have large dog populations . PAHO´s initiative in supporting the public health policies in the border areas of this study is laudable . This action is crucial but if stakeholders do not act quickly in controlling leishmaniases in this region as the scenario will inevitably become worse .
|
Every year about 3 million tourists from around the world visit Brazil , Argentina and Paraguay´s triple border region where the Iguaçu Falls are located . Unfortunately , in recent years an increasing number of autochthonous canine and human visceral leishmaniasis ( VL ) cases have been reported in this area . Our study determined the Phlebotominae fauna of the region , the factors that affect the presence and abundance of Lutzomyia longipalpis and Nyssomyia whitmani; the presence of Leishmania ( Leishmania ) infantum in different sand fly species and what Leishmania species are present . Lutzomyia longipalpis was the prevalent species and its distribution was related to the abundance of dogs . Leishmania ( L . ) infantum was found in Lu . longipalpis , Ny . whitmani , Ny . neivai and Lutzomyia sp . All the results are discussed within the Stockholm Paradigm and focus on their importance in the elaboration of public health policies in international border areas .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"atmospheric",
"science",
"geographical",
"locations",
"tropical",
"diseases",
"vertebrates",
"sand",
"flies",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"animals",
"mammals",
"dogs",
"protozoans",
"leishmania",
"neglected",
"tropical",
"diseases",
"humidity",
"insect",
"vectors",
"infectious",
"diseases",
"geography",
"zoonoses",
"south",
"america",
"protozoan",
"infections",
"brazil",
"disease",
"vectors",
"people",
"and",
"places",
"leishmania",
"infantum",
"eukaryota",
"urban",
"areas",
"meteorology",
"earth",
"sciences",
"leishmaniasis",
"geographic",
"areas",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"amniotes",
"organisms"
] |
2018
|
Hidden danger: Unexpected scenario in the vector-parasite dynamics of leishmaniases in the Brazil side of triple border (Argentina, Brazil and Paraguay)
|
Hyaloperonospora arabidopsidis ( Hpa ) is an obligate biotroph oomycete pathogen of the model plant Arabidopsis thaliana and contains a large set of effector proteins that are translocated to the host to exert virulence functions or trigger immune responses . These effectors are characterized by conserved amino-terminal translocation sequences and highly divergent carboxyl-terminal functional domains . The availability of the Hpa genome sequence allowed the computational prediction of effectors and the development of effector delivery systems enabled validation of the predicted effectors in Arabidopsis . In this study , we identified a novel effector ATR39-1 by computational methods , which was found to trigger a resistance response in the Arabidopsis ecotype Weiningen ( Wei-0 ) . The allelic variant of this effector , ATR39-2 , is not recognized , and two amino acid residues were identified and shown to be critical for this loss of recognition . The resistance protein responsible for recognition of the ATR39-1 effector in Arabidopsis is RPP39 and was identified by map-based cloning . RPP39 is a member of the CC-NBS-LRR family of resistance proteins and requires the signaling gene NDR1 for full activity . Recognition of ATR39-1 in Wei-0 does not inhibit growth of Hpa strains expressing the effector , suggesting complex mechanisms of pathogen evasion of recognition , and is similar to what has been shown in several other cases of plant-oomycete interactions . Identification of this resistance gene/effector pair adds to our knowledge of plant resistance mechanisms and provides the basis for further functional analyses .
Oomycetes comprise a number of agriculturally important plant pathogens , including Phytophthora infestans ( potato and tomato late blight ) , P . ramorum ( sudden oak death ) and P . sojae ( soybean root rot ) . Hyaloperonospora arabidopsidis ( Hpa , downy mildew , formerly known as Peronospora parasitica ) is a naturally occurring oomycete pathogen of the model plant Arabidopsis thaliana . The Hpa/Arabidopsis pathosystem allows the scientific community to take advantage of the genetic tools developed for Arabidopsis in the dissection of plant responses to oomycetes [1] . During their parasitic life stages , oomycetes deliver an arsenal of effector proteins to their plant host , which are hypothesized to target basal defense mechanisms and/or manipulate host metabolism to extract nutrients for the pathogen [2] . The first oomycete effector proteins were , however , identified based on their avirulence functions , i . e . their presence triggered an immune response in the host resulting in resistance to the pathogen . This so-called effector triggered immunity ( ETI ) is characterized by the specific recognition of pathogen avirulence effectors by plant resistance receptors , either directly or indirectly [3] . ETI is often accompanied by localized cell death at the site of infection , the hypersensitive response ( HR ) , which limits the spread of biotrophic pathogens inside the plant host [4] . The Hpa effectors Arabidopsis thaliana recognized 1 ( ATR1 ) , ATR13 and ATR5 as well as Avr1b from P . sojae were identified using classic genetic crosses between virulent and avirulent pathogen strains [5]–[8] , and have been shown to be under strong positive selection . P . infestans effector Avr3a , on the other hand , was isolated using association genetics [9] . Interestingly , despite having no sequence homology , Avr3a and ATR1 reside in conserved syntenic regions within the genome [9] and their three-dimensional structures reveal a similar fold between Avr3a and a sub-domain of ATR1 [10]–[13] . All currently described oomycete effectors were found to have a modular domain structure , containing amino-terminal domains involved in effector translocation and carboxyl-terminal effector domains . The translocation domains typically include a secretion signal sequence followed by the amino acid motif Arg-x-Leu-Arg ( RxLR ) , in which x could be any amino acid . The RxLR motif was shown to be important in translocation of effectors to the host cytoplasm and it is also functionally interchangeable with the translocation motif of Plasmodium falciparum effectors [14] , [15] . The absence of a canonical RxLR motif in the recently cloned effector ATR5 suggests that other sequences may also be involved in translocation of effectors [8] . The carboxyl-terminal effector domains are highly divergent and typically do not have strong sequence similarity to other proteins . The genomes of several Phytophthora species as well as of Hpa strain Emoy2 have recently been sequenced and this has fueled bioinformatics efforts to elucidate the complete arsenal of effectors [16]–[18] . Effector predictions were based on the presence of the amino-terminal translocation domains . While bacterial pathogens such as Pseudomonas syringae contain around 30 to 40 effector genes [19] , the oomycete genomes were found to contain expanded effector repertoires , ranging from around 350 in P . sojae and P . ramorum to more than 700 in P . infestans [20] . Initial effector predictions from the Hpa genome yielded 149 RxLR effector genes [21] , however , the published genome sequence for Hpa strain Emoy2 contains only 134 annotated RxLR effectors [17] . Currently , major efforts are being undertaken in dissecting the effector complement of several oomycetes in order to identify novel avirulence determinants as well as to define effector virulence functions . Several recent large-scale effector screens in different oomycetes focused on the localization of effectors in the plant cell during infection and on their roles in facilitating oomycete infections [22] , [23] . These transgenic approaches identified effectors that localize to the oomycete haustorial feeding structures and may be important in mediating intercellular communication . Another study employed mining of expressed sequence tags ( ESTs ) to identify genes highly expressed during Hpa infection , and investigated their potential functions during compatible interactions [24] . An in planta expression screen of P . infestans effectors was successful in identifying the cognate avirulent effectors ipiO/AVRblb1 and AVRblb2 , recognized by the R proteins Rpi-blb1 and Rpi-blb2 , respectively [25] , [26] . Resistance to different strains of Hpa was mapped to a number of RPP ( Resistance to Peronospora parasitica ) loci in several Arabidopsis ecotypes [27] , [28] . Six of these R genes were subsequently cloned , but the corresponding recognized effectors have only been identified for three of them , RPP1 , RPP13 and recently RPP5 , which recognize ATR1 , ATR13 and ATR5 respectively [5] , [7] , [8] . Similar to the R genes that function against other microbial pathogens , RPP genes belong to the large Nucleotide Binding Site-Leucine Rich Repeat ( NBS-LRR ) gene family in Arabidopsis , which comprises a total of around 150 members , but only a few with an assigned function [29] . Characterized R genes confer resistance to various classes of pathogens including oomycetes , bacteria , fungi and viruses . Additionally , NBS-LRR genes have been implicated in non-self recognition in inter-accession hybrids [30] . Research on RPP1/ATR1 and RPP13/ATR13 has greatly advanced our understanding of effector recognition and resistance signaling . Recognized ATR1 alleles have been shown to associate in planta with the LRR-domain of RPP1 before triggering an immune response [31] . Intracellular recognition of ATR13 by the CC-NBS-LRR protein RPP13 was shown to signal independently of the known signaling genes EDS1 and NDR1 , indicating the presence of additional signaling pathways activated upon effector recognition [32] . In order to gain a better understanding of the interactions between Hpa and its host A . thaliana , we set out to screen 83 Arabidopsis ecotypes for novel recognition specificities with a subset of predicted Hpa effectors . Here , we describe a successful approach to mine the Hpa genome for functional effector proteins based on domain structure similarity to known oomycete RxLR effectors . We identified a novel avirulent RxLR effector , ATR39 , which is recognized by the Arabidopsis ecotype Weiningen . Comparison of ATR39 alleles identified two amino acids that are critical for recognition . We cloned the corresponding R gene , RPP39 , and showed that it is a member of a small cluster of CC-NBS-LRR genes and requires NDR1 for downstream signaling of plant defense responses . Our ability to combine computational predictions with molecular and genetic techniques will facilitate the rapid identification of novel R genes as well as inform our understanding of the evolution of pathogenesis and resistance .
Based on characterized effectors from Phytophthora sp . and Hpa , intracellular oomycete effector proteins are predicted to contain several conserved domains: an N-terminal secretion signal peptide ( SP ) , a central RxLR motif , and a C-terminal variable effector domain . Previously , Win et al . mined the Hpa genome ( version 7 . 0 ) for predicted open reading frames of >70 amino acids , which contained the N-terminal SP and the RxLR motif between amino acids 30 and 60 , and found 149 effectors fulfilling these criteria [21] . In order to refine this search we generated a Hidden Markov Model ( HMM ) from the N-terminal conserved domains ( SP and RxLR ) of previously identified effectors and their homologues ( this set includes 43 proteins , [21] , Figure 1A ) . HMMs are widely used to predict homologies with statistical significance , most prominently in the protein domain database Pfam [33] . This method allowed us to screen the 149 initially predicted effectors with the HMM model and prioritize them for experimental validation , as outlined in Figure 1B . We decided to focus downstream characterization on the 18 highest-scoring predicted effectors , with E-value scores <0 . 001 ( Figure 1C ) . Amino acid and nucleotide sequences for these effectors are available in fasta format online as File S1 and S2 , respectively . Interestingly , some of the predicted effectors scored even higher than two known effectors , ATR1 ( Hp_Contig137 . 3_F55 ) and ATR13 ( Hp_Contig1514 . 4_F2 ) . We tested whether the predicted effectors were expressed during Hpa infection by RT-PCR and confirmed expression for 15 of them at seven days post-inoculation ( Figure 1D ) . Because of Hpa's obligate biotroph lifestyle , studies on Hpa have relied on surrogate systems , delivering oomycete effectors biolistically [34] or using bacterial or viral vectors [35] , [36] . We PCR amplified the highest scoring expressed effectors from the Emoy2 isolate of Hpa , past the predicted signal peptide cleavage site , and cloned them into two Pseudomonas expression vectors . First , the effectors were shuttled into a Gateway-compatible Pseudomonas vector as C-terminal fusions with the AvrRpm1 type three secretion system ( TTSS ) signal peptide ( pPsSP , [35] ) . However , we observed that ATR1 was not functional in this system , but was functional as a C-terminal fusion with the AvrRps4 TTSS signal peptide in the alternative pEDV3 system [36] . We thus decided to also subclone the predicted effectors into pEDV3 and test them in both expression systems . We conjugated these expression plasmids into Pseudomonas fluorescens ( Pf0 ) , a non-pathogenic Pseudomonas strain that lacks an endogenous TTSS and effectors and was engineered to express the Pseudomonas syringae pv . tomato ( Pst ) DC3000 hrp cluster and TTSS [37] . Using this strain allowed us to deliver individual effectors to the plant host , circumventing considerable background we often observed when inoculating a variety of ecotypes with Pst DC3000 ( data not shown ) . We did not observe a reaction to Pf0 carrying an empty vector control in most ecotypes , even after 48 hours post-inoculation ( hpi ) . Effector recognition on the other hand resulted in a visible hypersensitive response within 24 hpi ( Figure 2A ) . We then screened 83 Arabidopsis ecotypes from the Nordborg collection ( 1 , [38] ) with Pf0 expressing each of the predicted effectors . One of the predicted effectors , Hp_Contig399 . 11_F1 , triggered a visible HR in the ecotype Weiningen ( Wei-0 ) , when delivered by Pf0 ( Figure 2A ) . Hp_Contig399 . 11_F1 has an HMM score of 9 . 3 and ranks number 16 of the predicted effectors ( Figure 1C ) . None of the other 82 available ecotypes from the Nordborg collection displayed an HR upon delivery of Hp_Contig399 . 11_F1 . According to nomenclature previously applied to Hpa effectors , we renamed this effector ATR39-1 ( for A . thaliana recognized 39-1 , accession number JQ045572 ) . In order to assess whether ATR39-1 conferred avirulence to pathogenic bacteria , we conducted bacterial growth assays but found that Wei-0 showed natural resistance towards several pathogenic Pseudomonas strains , including Pst DC3000 ( Figure 2B ) and P . syringae pv . maculicola ( Psm ) ES4326 ( Figure S1 ) . We therefore generated an introgression line in which the Wei-0 recognition locus was introduced into the Col-0 background by repeated backcrossing . Using this line ( WC-5BX ) , we showed that ATR39-1 is indeed recognized by the Wei-0 locus and that this recognition results in decreased growth of Pst DC3000 ( Figure 2B ) . Unlike ATR1 , ATR39-1 does not contain the acidic DEER ( Asp-Glu-Glu-Arg ) motif following the RxLR translocation motif . We also tested whether the effector-truncation past the RxLR motif triggers the hypersensitive response and found that the effector domain ( ATR39-1 Δ48 ) is also able to trigger an HR when delivered by Pf0 ( Figure S2 ) . Using the same primers we amplified two alleles of ATR39 from Hpa strain Emoy2 , but only one of them , ATR39-1 , is recognized by Wei-0 ( Figure 2 ) . Both alleles are predicted in our HMM search with scores of 9 . 3 and 8 . 8 for ATR39-1 and ATR39-2 , respectively . The two alleles differ by 10 nucleotides , resulting in 9 amino acid substitutions , and a 2 amino acid insertion in ATR39-2 relative to ATR39-1 ( Figure 3A ) . In the published Hpa Emoy2 genome assembly ATR39-1 is not annotated , however , its allelic variant ATR39-2 is annotated as HaRxL48 [17] . We have generated Illumina paired-end sequencing data for the Emwa1 isolate of Hpa and could identify both alleles in the assembly ( data not shown ) . In this assembly we do not see signatures of duplication events , suggesting that Hpa Emwa1 and possibly also Emoy2 and Noco2 are heterozygous at this locus . Using allele-specific primers we amplified ATR39 alleles from seven isolates of Hpa and found that ATR39-1 is only present and expressed in Emoy2 , Emwa1 and Noco2 isolates ( Figure 3B ) . ATR39-2 , on the other hand , is more prevalent in this set of Hpa isolates . Surprisingly , we could not amplify ATR39-2 from the Emoy2 isolate currently grown in our lab . Since we initially amplified the effectors from a DNA sample obtained from a different source than the Hpa Emoy2 strain , and there is evidence for heterozygosity in the various lab strains , it is possible that our current lab strain is now homozygous for ATR39-1 . In order to determine whether this strain is Emoy2 , we amplified and sequenced the ATR1 effector and verified that our lab strain is indeed Emoy2 ( data not shown ) . We sequenced the ATR39 amplification products and found no nucleotide polymorphisms among ATR39-1 or ATR39-2 alleles from the different Hpa isolates . This conservation is in stark contrast with other characterized Hpa effectors ATR1 , ATR5 and ATR13 , which are highly divergent [5] , [8] , [39] . These findings indicate that the two ATR39 alleles may have an important function in Hpa and may be maintained under strong balancing selection . Sequence comparisons and pattern searches yielded no obvious homologs or putative function for ATR39 . Taken together , we have identified a novel effector from Hpa with unknown function that is able to trigger a resistance response in Arabidopsis . The Wei-0 ecotype was previously shown to be susceptible to multiple Hpa isolates [40] . We confirmed that despite being able to recognize ATR39-1 present in Emoy2 , Emwa1 and Noco2 , Wei-0 still supports growth of these isolates ( Figure 3C ) . This lack of resistance is not due to the lack of ATR39-1 transcript since we detected ATR39-1 expression in infected tissue using RT-PCR ( Figure 3B and Figure S3 ) . A similar suppressed recognition phenotype was observed for one of the ATR1 alleles . The ATR1Emco5 allele is recognized in several Arabidopsis ecotypes , which remain susceptible to infection by Hpa Emco5 [5] , [41] . ATR39-1 and ATR39-2 differ by 9 non-synonimous substitutions , and a two amino acid insertion in ATR39-2 . In order to define the region in ATR39 responsible for differential recognition of the two alleles , we generated ATR39-1in by inserting E168/V169 into ATR39-1 using site-directed mutagenesis . Similarly , we generated ATR39-2del , in which E168/V169 were deleted ( Figure 4A ) . ATR39-2del , but not ATR39-1in triggered an HR in Arabidopsis Wei-0 suggesting that the presence of amino acids E168/V169 blocks recognition of ATR39 ( Figure 4B ) . Additionally , we performed Pst DC3000 growth assays and found that ATR39-2del restricted bacterial growth , whereas Pst DC3000 expressing ATR39-1in grew to similar levels as the non-recognized allele ATR39-2 or empty vector control ( Figure 4C ) . These findings suggest that amino acids E168/V169 in ATR39-2 are critical in evading recognition by the cognate R protein . Resistance to Hpa strains is mediated by a number of RPP ( Resistance to Peronospora parasitica ) loci . In order to identify the RPP39 gene responsible for recognition of ATR39-1 we generated a cross between Wei-0 and Col-0 and found that recognition segregated as a single dominant locus in the F2 progeny . Using 886 F2 plants , we delineated the RPP39 locus to a 150 kilobase region on the bottom of chromosome 1 . In Col-0 this region contains two homologous CC-NB-LRR genes with 91% identity arranged in a tandem repeat ( At1g61180 and At1g61190 ) . Because there is no sequence information available for Wei-0 , we generated a fosmid library and identified six overlapping clones that span this region ( Figure 5B ) . We sequenced the fosmids using Illumina next generation sequencing and found several rearrangements and a transposon insertion relative to the Col-0 sequence ( Figure S4 ) . Wei-0 also contains two CC-NBS-LRR genes at this locus , which were the most promising candidates for RPP39 . We amplified genomic regions containing the two candidate R genes , R_180-Wei-0 ( accession number JQ045574 ) and R_190-Wei-0 ( accession number JQ045573 ) and transformed them into Arabidopsis Col-0 for complementation . Only transgenic plants containing the gene corresponding to the At1g61190 locus ( R_190-Wei-0 ) developed an HR in response to ATR39-1 delivery , indicating that this gene is RPP39 ( Figure 5D ) . Growth assays with Pst DC3000 expressing ATR39-1 performed on plants in the T2 generation confirmed these results as we observed reduction in bacterial growth only in plants containing the R_190-Wei-0/RPP39 transgene ( Figure 5E ) . The predicted RPP39 coding region contains an intron close to the C-terminus , connecting a large N-terminal exon with a short C-terminal exon encoding the last 15 amino acids ( Figure 6A ) . We amplified RPP39 from Wei-0 cDNA and confirmed the presence of this intron . In Agrobacterium-mediated transient expression experiments in Nicotiana benthamiana we showed that the RPP39 cDNA driven by the CaMV35S promoter is able to trigger ATR39-1 dependent HR ( Figure 6B ) . Interestingly , expression of the genomic RPP39 clone in N . benthamiana was not sufficient to trigger HR , yet it was functional in transgenic Arabidopsis . Because a 35S driven clone of the genomic sequence of RPP39 is able to trigger HR ( Figure 6B ) , we believe that the lack of responsiveness of the genomic RPP39 clone is probably due to low expression off the native promoter in the transient assay . RPP39 is very similar to its Col-0 paralogs , At1g61190 ( NM_104800 . 1 ) and At1g61180 ( NM_104799 . 3 ) , with 86% identity at the nucleotide level and 81% to 87% at the amino acid level ( Figure S5 ) . The homologs are most divergent in the C-terminal LRR domain ( Figure S6 ) . The most closely related R proteins outside the RPP39 cluster are RPS5 ( NP_172686 . 1 ) , RFL1 ( AAL65608 . 1 ) and RPS2 ( AAA21874 . 1 ) with 50% , 49% and 27% identity to RPP39 at the amino acid level , respectively ( Figure S5 ) . Taken together , we identified a functional resistance gene as a member of small R gene cluster . Non-specific disease resistance 1 ( NDR1 ) is a common signaling gene required for CC-NBS-LRR mediated resistance functions [42] . Since RPP39 is similar to the resistance genes RPS2 and RPS5 , both of which require NDR1 , we investigated the involvement of NDR1 in RPP39 mediated resistance . We generated stable transgenic plants containing RPP39 or its non-functional paralog R180_Wei-0 in the ndr1 mutant background . RPP39 transgenic plants displayed no difference in their ability to trigger ATR39-1-dependent HR as compared to transgenics in Col-0 wild type background ( Figure 7A ) . However , in Pst DC3000 growth assays we did not see a similar growth reduction in the ndr1 transgenics ( Figure 7B ) . These results indicate that the ability of RPP39 to trigger HR is separable from complete disease resistance , and is reminiscent of RPM1 , which displays similar NDR1 dependency [43] .
We have used a combination of computational prediction methods and phenotypic screening to identify a novel recognized effector from Hpa . Being an obligate biotroph , Hpa is currently not amenable to in depth genetic analysis , and cloning of the previously identified RxLR effectors ATR1 and ATR13 was a lengthy and cumbersome process [5] , [7] . In our approach , we mined the Hpa genome for putative effector sequences and ranked them in a comparison with a Hidden Markov Model ( HMM ) generated based on the N-terminal conserved translocation domains of previously identified oomycete effectors . When we generated the HMM , all characterized effectors contained an RxLR and in our initial ranking we screened the HMM only against predicted RxLR effectors . Recently , Bailey et al . cloned and characterized ATR5 from Hpa strain Emoy2 , the first recognized effector without a canonical RxLR sequence , suggesting that this motif can be modified [8] . In accordance with this , when screening the HMM against all annotated proteins in the final Hpa genome release [17] , we also identified several non-RxLR variants within the top-scoring effector candidates , including ATR5 ( data not shown ) . These findings suggest that our screen is by no means exhaustive and that more recognized effectors await characterization . When screening the HMM against the published annotation of the Hpa protein database , we also noticed that a few of our predicted effectors did not appear in the results because their annotated genes do not include the N-terminal translocation domains due to a difference in annotation of the translational start site . Our choice of using the non-pathogenic P . fluorescens that does not normally trigger a response in Arabidopsis as a surrogate expression and delivery system allowed us to rapidly screen a large number of Arabidopsis ecotypes for novel recognition specificities . Interestingly , despite the fact that Hpa strains are predicted to contain between 10 and 20 avirulent effectors , according to association studies in the 1990s [44] , in our set we only identified one novel effector that was able to trigger resistance in an Arabidopsis ecotype . Under our assay conditions and using our expression vectors we did not detect consistent phenotypes ( either virulence or avirulence ) for most of the tested effectors from the Emoy2 isolate of Hpa . Notably , in this study we compared the two delivery systems pEDV3 and pPsSP , which fuse the TTSS signal peptides from the bacterial effectors AvrRps4 and AvrRpm1 , respectively , upstream of the Hpa effector [35] , [36] . Intriguingly , ATR39-1 was functional when expressed as AvrRpm1 fusion in the pPsSP vector , but not as AvrRps4 fusion in the pEDV3 vector . ATR1 , on the other hand , did not trigger responses when delivered as AvrRpm1 fusion protein in the pPsSP system and is only functional in the pEDV3 system . These results suggest that the choice of expression system can greatly influence the observed phenotypes and should be taken into account . Therefore , we cannot dismiss the possibility that several of the predicted effectors might be functional in different assay conditions or recognized by different Arabidopsis ecotypes not tested in this study . In a set of Arabidopsis ecotypes from the United Kingdom , where all currently available Hpa strains originate from , several ATR13 alleles were tested and were found to be recognized by RPP13 alleles in different ecotypes [45] . The fact that we did not observe prevalence of effector recognition in our subset of Arabidopsis ecotypes suggests that the co-evolution between pathogen and host may play an important role in this obligate biotroph interaction . We also screened various alleles of ATR1 and ATR13 on the Nordborg collection of Arabidopsis ecotypes [38] and found six ecotypes capable of recognizing ATR1Emoy2 , including Ws-0 and Nd-1 , but none that recognized ATR13Emoy2 [41] . Compared with the prevalence of recognition specificities for bacterial effectors such as AvrRpt2 or AvrPphB , which are recognized by RPS2 and RPS5 in more than half of the tested ecotypes [38] , these results indicate a more dynamic effector repertoire in Hpa . We identified two amino acids , E168 and V169 , which abrogate recognition of ATR39-2 . Deletion of these amino acids in ATR39-2 resulted in a gain-of-recognition phenotype while insertion of E168/V169 in ATR39-1 lead to loss-of-recognition by RPP39 . It is possible that this insertion/deletion polymorphism alters the three dimensional structure of ATR39 , thus disrupting the interaction with potential target proteins . Another possibility would be that the polymorphism alters putative enzymatic properties of ATR39 . However , since the primary amino acid sequence of ATR39 is not homologous to any known protein , we can only speculate about the consequences of the insertion on the function of this protein . Experiments to determine the structure and to identify interacting proteins of ATR39 alleles will help elucidate the function of this novel effector . The Arabidopsis ecotype Wei-0 exhibits an interesting resistance pattern: it is resistant against several tested bacterial pathogens but highly susceptible towards Hpa isolates ( [40] , this study ) . Intriguingly , despite being able to recognize ATR39-1 when delivered by a surrogate system , Wei-0 is not able to restrict growth of Hpa expressing ATR39-1 . This finding is reminiscent of ATR1Emco5 , which is recognized in the Arabidopsis ecotype Ws-0 by RPP1-WsB , but this recognition does not abrogate growth of Hpa in a natural infection . Additional Arabidopsis ecotypes recognizing ATR1Emco5 were recently identified , and these were also shown to support growth of Hpa isolate Emco5 , indicating similar mechanisms in virulence may act in these interactions [41] . These data suggest that Hpa may contain a pathogenicity factor , perhaps another secreted effector , which prevents recognition of ATR1 in Emco5 or ATR39-1 in Emwa1 and Emoy2 . Support for this hypothesis comes from a study on different alleles of the P . infestans effector ipiO/Avrblb1 , where expression of one allele , ipiO4 , was found to suppress resistance mediated by the ipiO1/Rblb1 interaction [46] . Alternative explanations for the lack of recognition in these instances could be inhibition of effector translocation to the host , mistimed expression of either effector or R gene or incomplete resistance mediated by the R gene that is not strong enough to contain the pathogen . Experiments aimed at investigating this interesting phenotype should yield important insight into Hpa virulence . RPP39 is a member of a small cluster of CC-NBS-LRR genes , which is rapidly evolving through duplication and inversion events . A dominant mutation in the LRR domain of an RPP39 homolog in Ws-0 , uni-1D ( named after the Japanese word for sea urchin because of its morphological phenotype ) was found to display several defects in growth and hormone signaling which are often seen with gain of function R proteins such as SNC1 [47] , [48] . Interestingly , the uni-1D phenotype was not dependent on NDR1 [47] . We found that RPP39 requires NDR1 to fully suppress P . syringae growth , but activates hypersensitive cell death independently of NDR1 . In future , it will be interesting to more completely dissect the RPP39 signaling pathway . Taken together , our results show the feasibility of employing computational predictions in the identification of functional pathogen effectors . In combination with classical genetic methods it was possible to determine function for one member of the large family of predicted R proteins in Arabidopsis . Preliminary data on several polymorphic effectors from our HMM priority list indicate that Hpa isolates other than Emoy2 contain functional/recognized effector alleles , which will be further pursued and may lead to the identification of additional R genes , the analysis of which will further advance our understanding of R protein signaling .
A set of confirmed oomycete effectors and their close homologs used for the bioinformatic analyses included 43 sequences from Phytophthora species and Hyaloperonospora arabidopsidis and has been previously published [21] . The Signal Peptide and RxLR portion of these sequences ( positions 1 to 90 ) were aligned using the muscle algorithm [49] . The HMM building , calibration , and searches were performed using the HMMER software package with hmmbuild , hmmcalibrate , and hmmsearch algorithms , respectively ( http://hmmer . org/ ) . The HMM search included only one iteration . The predicted effectors were PCR amplified without signal peptide from Hpa strain Emoy2 DNA ( obtained from Jonathan Jones ) using primers listed in Table S2 and cloned into the pENTR/D-Topo vector ( Invitrogen ) . The insert sequences were verified and the effectors recombined with LR clonase ( Invitrogen ) into the binary Pseudomonas expression vector pPsSP [35] . The pEDV3 clones of effectors were generated using restriction digests with SalI/BamHI or compatible restriction enzymes . Site-directed mutants ATR39-1in and ATR39-2del were generated in pENTR using the QuikChange Lightning kit ( Agilent ) and primers listed in Table S2 , and recombined with binary vectors as above . For transient expression in Nicotiana benthamiana , ATR39 alleles were recombined into pEarleygate201 containing a 35S promoter and N-terminal HA-tag [50] . The fasta files containing the amino acid and nucleotide sequences of the predicted effectors are available online as Text S1 and Text S2 . A . thaliana plants were grown on soil in controlled growth chambers at short days ( 8 hrs light/16 hrs dark cycle ) and 24°C . Transgenic Arabidopsis were surface sterilized and selected on MS medium with the appropriate antibiotics . For Hpa growth assays the plants were transferred to a growth chamber with 18°C and 100% humidity . Hpa isolates were obtained from E . Holub ( Maks9 ) , X . Dong ( Emwa1 ) , J . McDowell ( Emoy2 , Emco5 ) , X . Li ( Cala2 ) and J . Jones ( Noco2 ) and maintained on susceptible Arabidopsis plants as previously described [27] . For disease assays , conidiospores were harvested by vortexing in water , adjusted to 5*104 spores/mL and sprayed on 2-week-old Arabidopsis seedlings . The infected plants were kept in growth chambers at 18°C and 100% humidity for 7 days before being stained with lactophenol-trypan blue [40] to assess Hpa growth or resistance . Pseudomonas fluorescens strains were grown on Pseudomonas agar with glycerol ( PAG ) supplemented with the appropriate antibiotics . For HR assays , Pf0 strains were grown on PAG plates for 2 days and resuspended in 10 mM MgCl2 to OD600 nm = 1 ( corresponding to 109 cfu/mL ) . Bacteria were inoculated into halves of pierced Arabidopsis leaves using a blunt syringe . Visible HR symptoms were scored 24 hours post-inoculation . Pseudomonas growth assays were performed as described previously [51] . Markers and probes used in map-based cloning of RPP39 are summarized in Table S3 . A fosmid library of Wei-0 was generated following the instructions in the copy control fosmid kit ( Epicentre ) and screened with Digoxigenin-labeled probes using the DIG DNA labeling and detection kit ( Roche ) . To sequence the RPP39 region , 350 bp-sized fragments of overlapping fosmids were sequenced in a 60 bp paired-end sequencing run on an Illumina G2 . Reads were cleaned for vector sequences and bacterial contamination using MAQ ( http://maq . sourceforge . net/ ) and assembled using CLC genomics workbench ( http://www . clcbio . com/ ) . Gaps and misassembled regions were filled in using Sanger sequencing data . Genomic fragments of about 6 kb length , containing RPP39 ( R190_Wei-0 ) or R180_Wei-0 were amplified using primers specified in Table S2 and introduced into pENTR/D-Topo ( Invitrogen ) . Sequences were confirmed and the fragments recombined into pEarleygate301 [50] for expression in Agrobacterium . The RPP39 cDNA clone was amplified from Wei-0 cDNA and recombined into pEarleygate100 containing a 35 S promoter [50] . The binary vectors were mobilized into Agrobacterium tumefaciens strain GV3101 with tri-parental mating and used to transform Arabidopsis plants following the floral dip protocol [52] . Presence of the transgene in Arabidopsis was confirmed by PCR . Transient expression experiments in N . benthamiana were performed as previously described [35] . Protein expression was determined by Western blotting as described [31] . RNA was extracted from Arabidopsis seedlings infected with Hpa using the RNeasy plant mini kit ( Qiagen ) and reverse transcribed using Superscript III ( Invitrogen ) and oligo-dT primers . RT-PCR was performed with gene specific primers ( Table S2 ) and 25 amplification cycles .
|
Oomycete plant pathogens are among the most devastating agricultural pests and employ arsenals of effector proteins to manipulate their plant hosts . Some of these effectors , however , are recognized in the plant and trigger an immune response . Hyaloperonospora arabidopsidis ( Hpa ) causes downy mildew on the model plant Arabidopsis thaliana and this interaction has been developed as a model system for oomycete pathogenesis . Here , we employ computational predictions to identify a novel effector ATR39-1 , which is highly conserved among different Hpa isolates . A two amino acid-insertion in the alternative allele ATR39-2 correlated with evasion of recognition . We identified the corresponding resistance gene RPP39 and found that the signaling gene NDR1 is required to establish full resistance . Recognition of ATR39-1 by RPP39 in the plant did not inhibit growth of the oomycete , suggesting that complex mechanisms exist to prevent effector recognition . Knowledge of such novel resistance interactions provides the backbone of our understanding of plant resistance mechanisms and will aid in the further dissection of plant immunity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"agriculture",
"biology"
] |
2012
|
Computational Prediction and Molecular Characterization of an Oomycete Effector and the Cognate Arabidopsis Resistance Gene
|
Several strategies have been pursued to increase the extent of exon 7 inclusion during splicing of SMN2 ( survival of motor neuron 2 ) transcripts , for eventual therapeutic use in spinal muscular atrophy ( SMA ) , a genetic neuromuscular disease . Antisense oligonucleotides ( ASOs ) that target an exon or its flanking splice sites usually promote exon skipping . Here we systematically tested a large number of ASOs with a 2′-O-methoxy-ethyl ribose ( MOE ) backbone that hybridize to different positions of SMN2 exon 7 , and identified several that promote greater exon inclusion , others that promote exon skipping , and still others with complex effects on the accumulation of the two alternatively spliced products . This approach provides positional information about presumptive exonic elements or secondary structures with positive or negative effects on exon inclusion . The ASOs are effective not only in cell-free splicing assays , but also when transfected into cultured cells , where they affect splicing of endogenous SMN transcripts . The ASOs that promote exon 7 inclusion increase full-length SMN protein levels , demonstrating that they do not interfere with mRNA export or translation , despite hybridizing to an exon . Some of the ASOs we identified are sufficiently active to proceed with experiments in SMA mouse models .
Spinal muscular atrophy ( SMA ) , the most common genetic cause of infant mortality , is an autosomal recessive neuromuscular disease characterized by progressive loss of α-motor neurons in the anterior horns of the spinal cord , leading to limb and trunk paralysis and atrophy of voluntary muscles . Based on the severity and age of onset , SMA is clinically subdivided into types I , II , and III ( MIMs 253300 , 253550 , and 253400 ) , with type I being the most severe [1] . SMA has an incidence of approximately one in 6 , 000 live births , and a carrier frequency of one in 40 . Loss of function of the survival of motor neuron 1 ( SMN1 ) gene is responsible for SMA [2] . Mice with homozygous SMN1 disruption display massive cell death during early embryogenesis [3] . SMN protein is ubiquitously expressed and is mainly localized in the cytoplasm and in nuclear “gems” [4] . Multiple SMN-interacting partners have been identified , suggesting the involvement of SMN in various cellular processes , such as transcription , mRNA transport , and assembly of ribonucleoprotein particles ( RNPs ) , including small nuclear RNPs ( snRNPs ) , small nucleolar RNPs ( snoRNPs ) , and stress granules [5–8] . Humans have an extra SMN gene copy , designated SMN2 . Both SMN genes reside within a segmental duplication on Chromosome 5q13 as inverted repeats [2] . SMN1 and SMN2 are almost identical , except for 11 nucleotide substitutions: seven in intron 6 , two in intron 7 , one in coding exon 7 ( a translationally silent C to T transition , relative to SMN1 ) , and one in non-coding exon 8 [9] . SMN2 is a prototypical example of alternative splicing caused by a single nucleotide substitution in the affected exon . Exon 7 is efficiently included in spliced mRNA from the SMN1 gene; however , the silent C6T transition in SMN2 exon 7 , which weakens the recognition of the upstream 3′ splice site [10] , results in significant skipping of this exon during pre-mRNA splicing . The C6T transition abrogates a splicing factor 2/alternative splicing factor ( SF2/ASF ) -dependent exonic splicing enhancer ( ESE ) ; this SF2/ASF heptamer motif is present in SMN1 exon 7 ( nucleotides +6 to +13 , CAGACAA ) and supersedes the inhibitory effect of heterogeneous nuclear ribonucleoprotein ( hnRNP ) A1 [11–13] . As a result , SMN2 encodes mostly the exon 7–skipped protein isoform ( SMNΔ7 ) , which is unstable , mislocalized , and at best , only partially functional [14–16] . Although the small amount of full-length SMN protein derived from SMN2 is not sufficient to fully compensate for loss of SMN1 , it is essential for viability in the absence of SMN1 , and is an important disease modifier: in both SMA patients and mouse models , there is an inverse relationship between SMN2 copy number and disease severity [17 , 18] . These properties define SMN2 as an ideal therapeutic target for potential treatment of SMA . Antisense technology was initially employed to down-regulate gene expression by targeting mRNA to induce its degradation or block its translation [19] . Advances in antisense chemistry allowed this technology to be applied as a powerful tool to manipulate pre-mRNA splicing . Modification of the base , sugar , or phosphodiester structure of oligonucleotides generates highly stable molecules with not only high affinity for RNA targets , but also resistance to various nucleases , including RNase H ( which cleaves RNA in RNA/DNA hybrids ) . The Kole laboratory pioneered the use of antisense oligonucleotides ( ASOs ) to correct aberrant splicing by targeting pre-mRNA , demonstrating both in vitro and in vivo that ASOs can restore correct expression of a defective β-globin gene by blocking the cryptic splice sites generated by intronic mutations that cause β-thalassemia [20 , 21] . Recently , ASO applications to redirect pre-mRNA splicing were extended to genes with therapeutic relevance to cancer or other diseases [22–24] . ASOs have been employed to promote skipping of non-essential exons , so as to restore the translational reading frame and suppress the effects of various deletion , nonsense , and frameshifting alleles of the dystrophin gene ( DMD ) [25–27] . Although it is now relatively straightforward to block the use of specific exons by targeting the corresponding 5′ or 3′ splice sites or empirically selected sequences within the exon , it remains much more challenging to find suitable antisense targets for promoting exon inclusion . Several antisense approaches have been documented to promote SMN2 exon 7 splicing . One approach is based on the notion that there is a competition between the 3′ splice sites of exons 7 and 8 for pairing with the 5′ splice site of exon 6 [10] , so impairing the recognition of the 3′ splice site of exon 8 should favor exon 7 inclusion . Increased SMN2 exon 7 inclusion and full-length SMN protein were indeed observed in HeLa cells transfected with a modified U7 small nuclear RNA ( snRNA ) complementary to the intron 7/exon 8 junction [28] . Another approach , called TOES ( targeted oligonucleotide enhancer of splicing ) [29] , relies on bifunctional ASOs composed of one segment complementary to exon 7 and a non-complementary tail consisting of RNA sequences with ESE motifs recognized by a serine/arginine-rich ( SR ) protein . One such bifunctional ASO stimulated SMN2 exon 7 inclusion and led to significant restoration of gem number—an indicator of SMN protein increase—in SMA-patient fibroblasts [30] . Moreover , an in vivo delivery system was recently developed for expression of bifunctional ASOs in an adeno-associated virus vector [31] . Strong intronic splicing silencers ( ISSs ) represent ideal antisense targets; however , if present , they are usually inconspicuous within large introns . An ISS in intron 6 was identified in the context of an SMN2 minigene , and targeting this inhibitory element with an ASO increased exon 7 inclusion in transiently transfected COS-7 cells [32] . An intron 7 ISS , located immediately downstream of the 5′ splice site of exon 7 , was recently reported; an ASO against this silencer efficiently restored exon 7 inclusion of SMN2 and increased SMN protein levels in SMA-patient–derived cells [33] . The ESSENCE ( exon-specific splicing enhancement by small chimeric effectors ) method , which was developed to rescue disease-associated exon skipping or modulate alternative splicing , employs small chimeric effectors designed to emulate SR proteins [34 , 35] . A synthetic ESSENCE molecule comprises two portions: an antisense moiety complementary to a target exon; and a minimal RS domain peptide similar to the splicing activation domain of SR proteins . We described several ESSENCE compounds that significantly promote SMN2 exon 7 inclusion in a cell-free splicing assay [34] . Interestingly , the antisense moiety alone also significantly stimulates exon 7 inclusion , although with approximately 8-fold less potency compared to ESSENCE compounds , suggesting the existence of an exonic splicing silencer ( ESS ) in exon 7 [34] . This observation highlighted the possibility of further optimization of the antisense moiety of ESSENCE compounds to enhance their potency . Here , we used two-step ASO walks along SMN2 exon 7 to define position-dependent effects of ASO-binding sites . With the first coarse whole–exon 7 walk , we identified two ASO target regions separated by a central core sequence , and with the second high-resolution walks , we identified optimal ASOs against each target region . The two best ASOs strongly promoted exon 7 inclusion during SMN2 pre-mRNA splicing in three different splicing assays , and increased SMN protein levels in cultured cells , including SMA type I fibroblasts . Thus , appropriate ASOs targeting a coding exon can increase protein levels by promoting exon inclusion and not interfering with translation . Most importantly , the effective ASOs should be suitable for studies in animals and to further optimize bifunctional or peptide-linked ASOs for better efficacy in rescuing SMN2 splicing and full-length protein expression .
Previously we reported that one PNA ( peptide nucleic acid ) ASO complementary to positions +7 to +18 of SMN2 exon 7 stimulates exon 7 inclusion during splicing in vitro [34] . This unexpected observation suggested that the ASO blocked one or more putative ESSs—either a specific motif or a secondary structure that impairs the recognition of exon 7 . To pinpoint the exact locations of these presumptive elements , we used a systematic ASO walk along the entire length of SMN2 exon 7 . Oligonucleotides with a 2′-O-methoxy-ethyl ribose ( MOE ) phosphodiester backbone were used . This backbone modification imparts a very high affinity for targeted mRNA , resistance to both exo- and endonucleases , and does not support cleavage of hybridized mRNA by RNase H [36 , 37] . We synthesized nine MOE 15-mer ASOs complementary to the 54-nucleotide ( nt ) exon 7; beginning with the first position of the exon , the overlapping ASOs provide coverage in 5-nt increments from 5′ to 3′ , with a last step of 4 nt to the last position of the exon ( Figure 1A , Table 1 ) . An unrelated oligonucleotide , 00–00 , was used as a negative control ( Table 1 ) . We first evaluated these ASOs by in vitro splicing in HeLa cell nuclear extract with a radiolabeled SMN2 minigene transcript [11] . Although the precise extent of exon 7 inclusion varies between extracts or depending on the precise reaction conditions , the difference between SMN1 and SMN2 is highly reproducible . We tested eight different concentrations of each ASO , from 1 to 400 nM , compared to no ASO . As shown in Figure 1B , the nine ASOs had different effects on splicing of the SMN2 pre-mRNA . Four ASOs , 01–15 , 16–30 , 21–35 , and 26–40 , strongly inhibited exon 7 inclusion; ASOs 11–25 and 31–45 slightly inhibited exon 7 inclusion; and ASO 40–54 and the control oligonucleotide 00–00 had no effect on alternative splicing of SMN2 exon 7 . Interestingly , we found two ASOs , 06–20 and 36–50 , that significantly stimulated exon 7 inclusion in the cell-free splicing assay , starting at a concentration of approximately 50 nM . ASO 01–15 targets the exonic portion of the 3′ splice site , whereas ASOs 11–25 , 16–30 , 21–35 , and 26–40 target a central region that comprises a transformer 2 beta 1 protein ( Tra2β1 ) -binding motif in SMN exon 7 [38] . Both elements are required for exon 7 inclusion , which can account for the negative effects of these ASOs . ASO 31–45 has a 10-nt overlap with the stimulatory ASO 36–50 , yet it had a negative effect on splicing of exon 7 , suggesting that there is a motif upstream of the ASO 36–50-binding site that is also important for exon 7 inclusion . ASO 40–54 targets the exonic portion of the exon 7 5′ splice site; however , it is essentially neutral . We offer two explanations for this observation: first , the exonic portion of this 5′ splice site deviates from the consensus sequence , and thus ASO 40–54 might not affect U1 snRNA base-pairing; second , ASO 40–54 has a strong predicted stem-loop structure ( unpublished data ) and might not bind efficiently to its complementary sequence . It is not surprising that ASO 06–20 had a positive effect , because it encompasses the sequence of the antisense PNA we reported previously [34]; however , ASO 36–50 , which is complementary to nucleotides +36 to +50 of exon 7 , defines an additional region in SMN2 exon 7 that can be targeted to increase exon 7 inclusion . To determine whether the ASO-mediated stimulation of exon 7 inclusion can take place in cells , we co-transfected plasmid pCI-SMN2 with each of the nine MOE ASOs into HEK293 cells by electroporation . At the same time , we co-transfected plasmid pBabe Puro and selected the transfected cells by treatment with puromycin . Two days after transfection , we analyzed the transiently expressed RNA by RT-PCR ( Figure 2A ) . Each of the ASOs gave similar effects on splicing of exon 7 in the minigene pre-mRNA as observed by in vitro splicing ( Figure 1B ) , except for two ASOs , 11–25 and 31–45 . These two ASOs , especially the latter one , resulted in robust inhibition of exon 7 inclusion in vivo , whereas only slight effects were observed in vitro . We next tested the effect of the exon 7 ASOs on splicing of transcripts from the endogenous SMN genes in HEK293 cells . This time , only the ASOs and plasmid pBabe Puro were electroporated into the cells . The transfected cells were puromycin-selected as above , and RNA samples were collected 2-d post-transfection . After RT-PCR , we digested the cDNA with DdeI to distinguish SMN2 mRNA from SMN1 mRNA [39 , 40] . As expected , the effects of these MOE ASOs on splicing of transcripts from the endogenous SMN2 gene were very similar to those we observed with the SMN2 minigene ( Figure 2B ) . The only notable difference between the two assays was in the relative efficacy of the two positive ASOs , 06–20 and 36–50 . ASO 06–20 was less efficient in stimulating exon 7 inclusion than ASO 36–50 in the in vitro splicing and minigene splicing assays , but it was reproducibly more efficient than ASO 36–50 in the endogenous SMN2 splicing assay . On the basis of the above results obtained with the three splicing assays , we conclude that exon 7 comprises three distinct regions: a core sequence that is essential for exon 7 inclusion , and two flanking regions that correspond to putative ESSs . These two inhibitory regions correspond to the binding sites for ASOs 06–20 and 36–50 , and are designated region A and region B , respectively . Having roughly delineated regions A and B as effective antisense targets to stimulate exon 7 inclusion , we sought to define their boundaries more precisely by identifying ASOs with optimal sequences and lengths , and thus maximize exon 7 inclusion . To this end , we designed 39 new ASOs: 17 ASOs of length 12 , 15 , or 16 nt , spanning 21 nt within or overlapping region A ( microwalk A ) , and 22 ASOs of length 12 , 15 , or 18 nt , spanning 22 nt within or overlapping region B ( microwalk B ) ( Figure 3A ) . The sequences of all 39 ASOs are shown in Table 1 . We first screened all these MOE ASOs using the in vitro splicing assay and a concentration of 100 nM for each ASO . The 15-mer ASO 07–21 was identified as the most effective one in the case of microwalk A ( Figure 3B ) . Its target site is shifted downstream by one nucleotide , relative to the target site of the original ASO 06–20 . In the presence of this new ASO , the extent of exon 7 inclusion increased to 62% , representing a significant improvement over the original 15-mer ASO 06–20 , which gave 45% exon 7 inclusion in this experiment . Microwalk B generated another improved ASO , the 15-mer ASO 34–48 , though at this concentration , it gave only a slight improvement compared to the original ASO ( Figure 3C and unpublished data ) . The target site for this new ASO is shifted upstream by two nucleotides , relative to the target site of the original 15-mer ASO 36–50 . We further characterized the two improved high-resolution walk ASOs by titration in the in vitro splicing assay . Concentrations ranging from 1 to 400 nM of ASO 07–21 or 34–48 were tested and compared with the original ASO 06–20 or 36–50 identified in the coarse walk . The splicing data and dose-response curves are shown in Figure 4 . The median effective concentration ( EC50 ) values for the two original ASOs 06–20 and 36–50 were 56 nM and 47 nM , respectively , whereas the EC50 values for the improved ASOs 07–21 and 34–48 were reduced to 16 nM and 13 nM , respectively . The effects of the two groups of microwalk ASOs were further examined in vivo , first with the minigene pCI-SMN2 and then with the endogenous SMN1/2 genes , as described above . Each ASO , at a concentration of 10 μM , with or without the SMN2 minigene plasmid , was electroporated into HEK293 cells ( Figure 5A ) . ASO 07–21 stimulated exon 7 inclusion during splicing of the SMN2 minigene pre-mRNA more effectively than ASO 06–20 . As expected , ASO 07–21 also potently stimulated exon 7 inclusion during splicing of the endogenous SMN2 pre-mRNA ( Figure 5B ) . Similarly , ASO 34–48 robustly increased the inclusion of exon 7 for both the SMN2 minigene and the endogenous SMN2 gene transcripts ( Figure 5C and 5D ) . The effects of all the antisense molecules on exon 7 splicing were generally consistent between in vivo minigene and endogenous gene assays . Compared to the in vitro splicing assay , the two in vivo splicing assays appeared more sensitive in terms of the stimulatory or inhibitory effects on exon 7 inclusion caused by the tested ASOs . In addition , some ASOs gave inconsistent effects between the in vitro assay and the two in vivo assays: three ASOs in microwalk A ( 08–22 , 02–16 , and 09–20 ) and three ASOs in microwalk B ( 33–47 , 34–45 , and 33–44 ) all slightly promoted exon 7 inclusion in vitro , but more or less inhibited exon 7 inclusion in vivo; ASOs 12–23 , 35–52 , 34–51 , 33–50 , and 32–49 showed neutral or negligible effects on exon 7 inclusion in vitro , but inhibited exon 7 inclusion in vivo to some extent; and finally , ASO 38–49 showed no effect on exon 7 inclusion in vitro or with the endogenous gene , but promoted exon 7 inclusion in the minigene co-transfections . The microwalk data obtained from the two in vivo assays not only verified the presence of a central core sequence , but also established the precise boundaries between the core sequence and regions A and B on either side of it . In microwalk A , the target site for the improved ASO 07–21 extends to three consecutive As of the proposed Tra2β1-recognition sequence ( AAAGAAGGA ) [38] , suggesting that this portion of the Tra2β1-dependent ESE is not critical for binding of the protein . This interpretation is consistent with the finding that mutating the triple As to triple Us does not abrogate Tra2β1 recognition [38] . The triple As mark the 3′ boundary of region A; extending the walk farther downstream causes strong exon 7 skipping , as in the cases of ASOs 08–22 , 12–23 , and 11–22 ( Figure 5A and 5B ) , presumably because of interference with Tra2β1 binding . In microwalk B , ASOs that bind farther upstream of the improved ASO 34–48 , such as 33–47 , 32–46 , and 33–44 , caused a strong reduction in exon 7 inclusion ( Figure 5C and 5D ) . These three ASOs are complementary to exon 7 with a 5′-most boundary at +33C and +32G , respectively , suggesting that this GC dinucleotide is part of an element or structure important for exon 7 recognition . This observation is also consistent with the in vivo effect of ASO 31–45 in the initial exon 7 walk ( Figure 2 ) . The combined data from the coarse- and high-resolution antisense walks point to the existence of a core sequence in the middle of exon 7 , from +22 to +33 , that is essential for exon 7 inclusion . The microwalk data obtained from the two in vivo assays are also useful for mapping both the 5′ boundary of region A and the 3′ boundary of region B . In microwalk A , ASO 04–18 displayed the least stimulatory effects among the 15-mer ASOs; other ASOs that target further upstream sequences were either neutral or inhibitory . These results define the boundaries of region A as +4 to +21 . In microwalk B , the 15-mer ASOs 39–53 and 38–52 , whose 3′-most complementary-sequence boundaries are +53G and +52G , respectively , were essentially neutral; all other 15-mer ASOs , except 33–47 and 32–46 , which partly target the central core sequence , had stimulatory effects . Therefore , nucleotides +34 to +51 define the boundaries of region B . The results obtained with 12-mer ASOs are generally consistent with those obtained with 15-mer ASOs , except for ASO 04–15 , which targets region A , but was unexpectedly inhibitory with respect to exon 7 inclusion , and ASO 09–20 , which also targets region A , but was inhibitory in the in vivo assays . The two high-resolution walk exon 7 ASOs were further characterized with respect to their effects on splicing of the endogenous SMN1/2 genes in HEK293 cells . We first conducted dose-response experiments with each ASO , using a concentration range from 0 . 2 to 20 μM for electroporation . Although we do not know the intracellular concentration of each ASO after electroporation , inclusion of exon 7 in both SMN1 and SMN2 transcripts showed a clear dependence on the dose of each ASO ( Figure 6A and 6B ) . For both ASOs , a significant increase in exon 7 inclusion was already noticeable at a starting concentration of 0 . 2 μM; at 20 μM ASO , SMN2 exon 7 inclusion reached 87%–89% , compared to 86%–87% SMN1 and 50%–51% SMN2 exon 7 inclusion in the untreated cells . After treatment with ASO 07–21 or ASO 34–48 , the exon 7 inclusion level of the SMN1 gene transcripts rose to 93%–95% , indicating that these two ASOs affect the recognition of sequence elements or secondary structures present in transcripts from both SMN genes . Next , we analyzed the effects of the two ASOs over time , as a measure of their intracellular stability . We transfected HEK293 cells by electroporation with each ASO at 10 μM concentration , and total RNA samples from parallel transfections were collected daily at 1–5 d post-transfection . Robust effects were already apparent on the first day and reached a plateau by the second day ( Figure 6C and 6D ) . The strong effects persisted for at least 5 d , though a slight decline of exon 7 inclusion level was observed , which might partly reflect dilution of the ASOs as the cells divided . Both ASO 07–21 and ASO 34–48 strikingly increased the amount of full-length mRNA expressed from the endogenous SMN1/2 genes . However , because the ASOs might remain associated with exon 7 in the spliced mRNA , they could potentially interfere with mRNA export and/or translation . Therefore , we sought to determine whether the stimulatory effect of these ASOs on SMN1 exon 7 inclusion results in increased full-length SMN protein . We could not accurately determine this in HEK293 cells , because of the high levels of identical SMN protein expressed from the SMN1 gene . Therefore , we generated another SMN2 minigene construct , designated pEGFP-SMN2 . This minigene consists of an N-terminal enhanced green fluorescent protein ( EGFP ) , a short cloning-site sequence , a hemagglutinin epitope tag ( HA tag ) , SMN exon 6 , a shortened intron 6 ( same as in minigene pCI-SMN2 ) , exon 7 , intron 7 , and the 5′ end of exon 8 ( 75 nt ) at the C-terminus ( Figure 7A ) . When the pre-mRNA from this minigene is spliced , it generates two mRNAs via exon inclusion or skipping . Translation of the exon 7–included mRNA gives a larger protein of 308 amino acids ( designated P-Incl ) , comprising EGFP , seven amino acids ( SGLRSRE ) derived from the cloning site , the HA tag , and exons 6 and 7 of SMN . The exon 7–skipped mRNA is translated into a smaller protein of 296 amino acids ( designated P-Excl ) , comprising EGFP , SGLRSRE , the HA tag , exon 6 , and four amino acids ( EMLA ) from exon 8 ( Figure 7A ) . Both proteins can be detected with anti-HA antibody . Before proceeding to the protein-level analysis , we first verified the increase in exon 7 inclusion for the minigene pEGFP-SMN2 after electroporation with ASO 07–21 or 34–48 . Compared with the endogenous SMN2 gene , minigene pEGFP-SMN2 favors the exon 7–skipped mRNA isoform . The extent of exon 7 inclusion was about 28% without any treatment; however , after treatment with 10 μM ASO 07–21 or 34–48 , exon 7 inclusion rose to 74% or 85% , respectively ( Figure 7B ) . Protein samples were generated 3 d after the cells were co-transfected via electroporation with the minigene construct and either ASO at 10 or 30 μM concentration . Surprisingly , using mouse monoclonal anti-HA antibody , we only detected the exon-included protein product , P-Incl , with an apparent mass of approximately 38 kDa ( Figure 7C ) . The effects of both ASOs were dose dependent , and after the treatment with 30 μM ASO , P-Incl reached almost the same level as observed with the SMN1 minigene . We were not able to detect any P-Excl protein produced from the SMN2 Δ7 mRNA transcript; we believe this is due to protein instability . SMNΔ7 , the unstable protein isoform derived from the SMN1/2 genes , and P-Excl share the same C-terminus , corresponding to the peptide coded by exon 6 plus four amino acids derived from exon 8; it is possible that these sequences , when present close to the C-terminus , induce degradation of the protein . To overcome this problem , we constructed another SMN2 minigene , pEGFP-SMN2Δ6; this construct is similar to pEGFP-SMN2 , but lacks most of exon 6 , except for the last 9 nt , so as to maintain an intact natural 5′ splice site ( Figure 7A ) . This minigene encodes two protein isoforms: the exon 7–included one ( P-InclΔ6; 274 amino acids ) and the exon 7–excluded one ( P-ExclΔ6; 262 amino acids ) . Using this minigene , we again tested ASO 07–21 or 34–48 in HEK293 cells; significant increases in both exon 7–included mRNA and protein were observed ( Figure 7B and 7C ) . This time , using anti-HA antibody , we detected two polypeptides from pEGFP-SMN2Δ6 , with sizes of about 32 kDa and 30 kDa , and with the latter giving a stronger signal . After treatment with the ASOs , the P-ExclΔ6 band disappeared , whereas the P-InclΔ6 band became considerably stronger and comparable in intensity to the SMN1 minigene product . To control for sample loading differences , the same blots were reprobed with anti–α-tubulin monoclonal antibody . These results clearly demonstrate that ASOs targeting a coding exon are able to promote exon inclusion at the mRNA level , and the resulting mRNA can be translated into protein . To test whether the in vivo results in human HEK293 cells can be extended to SMA-patient cells , SMA type I 3813 fibroblasts were transfected with either ASO at 100 nM concentration using Lipofectin; this method of delivery was chosen because it was more efficient than electroporation with these primary fibroblast cells , consistently giving approximately 50% transfection efficiency with an EGFP plasmid ( unpublished data ) . At 48 h after transfection , the cells were harvested to prepare mRNA and protein samples . The 3813 cells treated with the control ASO 00–00 or buffer alone gave 31% exon 7 inclusion . After treatment with ASO 07–21 or 34–48 , the apparent level of exon 7 inclusion increased to 58% and 61% , respectively ( Figure 8A , left ) ; based on the standard deviations ( unpublished data; n = 3 ) , this corresponds to a 1 . 9 ± 0 . 04 fold increase for ASO 07–21 , and a 2 . 0 ± 0 . 04 fold increase for ASO 34–48 , relative to the treatment with the control ASO 00–00 . The increase in the percentage of exon 7 inclusion could potentially reflect reduced exon skipping alone , without the desirable concomitant increase in exon inclusion . To rule out this possibility , we used GAPDH mRNA as a loading control , and calculated the ratio of SMN full-length mRNA to GAPDH mRNA , normalized to that observed with the control ASO 00–00 ( Figure 8A , right ) . The ratio increased by approximately 1 . 6-fold in ASO 07–21-treated cells and approximately 1 . 8-fold in ASO 34–48-treated cells . To detect changes at the protein level , we used α-tubulin as a loading control . The normalized ratio significantly increased by approximately 1 . 8-fold and approximately 2 . 1-fold in ASO 07–21- and ASO 34–48-treated samples , respectively , compared to ASO 00–00 ( Figure 8B ) . Assuming a delivery efficiency of approximately 50% , the actual effects should be approximately 2-fold greater in the cells that took up the ASOs . Gems are nuclear bodies in which SMN protein accumulates [4] . A correlation between the number of gems in patient-derived fibroblasts and SMA clinical severity has been demonstrated [41] . To assess the effects of ASOs 07–21 and 34–48 on nuclear gem counts , we used indirect immunofluorescence with an anti-SMN monoclonal antibody , after transfection of 3813 cells with either ASO . The cells treated with the unrelated control oligonucleotide contained an average of 10 gems per 100 cells; however , when the cells were treated with ASO 07–21 or 34–48 , the average gem number increased to 26 or 31 per 100 cells , respectively . The carrier 3814 fibroblast cells had approximately 50 gems per 100 cells ( Figure 8C–8E ) . In addition , after treatment with ASO 07–21 or 34–48 , the number of cells containing multiple gems increased , and the number of cells containing no gems decreased . These results confirm that ASOs 07–21 and 34–48 significantly promote full-length SMN protein expression .
Multiple strategies aimed at correcting exon 7 splicing of SMN2 , a modifying gene for SMA , have been investigated . Mechanistic studies to understand how SMN2 exon 7 is alternatively spliced led to the discovery of several cis-elements and trans-acting factors that can be targeted to stimulate exon 7 inclusion [42] . A number of drugs , including synthetic compounds that can modify SMN2 splicing , have been identified using cell-based high-throughput drug screening [42] . On the basis of the current knowledge of splicing mechanisms , we and others have employed antisense-based technologies to stimulate exon 7 inclusion . In recent years , a growing number of studies have demonstrated that MOE-modified ASOs ( with a phosphodiester or a phosphorothioate backbone ) and antisense PNAs can be valuable tools , not only for dissecting gene function , but also for clinical applications [19] . Our previous work showed that ESSENCE compounds have the potential to treat diseases caused by exon skipping resulting from the loss of ESEs in mutant genes , including SMN2 [34] . A critical parameter in ESSENCE effectiveness is the selection of the optimal binding site along a target exon . In principle , the best target sequences are those that comprise negative exonic splicing signals or secondary structures , such as ESSs . In this study , we systematically analyzed SMN2 exon 7 with a large number of MOE ASOs to identify putative ESSs . We employed a two-step ASO walk method , with an initial coarse ASO walk in 5-nt steps along the entire exon , followed by high-resolution single-nucleotide walks within the regions identified in the first step . With the first-step walk , we identified two ASO targets of potential therapeutic importance . The second-step microwalks optimized the ASOs and defined the apparent boundaries of the two ESS-containing regions . Using three independent splicing assays , we identified two potent ASOs that have significant therapeutic potential for SMA treatment . Our data also suggest that the two-step ASO walk is a powerful general method that can be used for screening inhibitory or stimulatory splicing regions present in the target exon ( s ) and the surrounding intron sequences of any gene . Our data revealed an essential core sequence in the center of the 54-nt exon 7 from +22 to +33 ( GAAGGAAGGTGC ) , which is surrounded by two separate inhibitory regions ( A and B ) containing negative splicing signals , with region A extending close to the upstream 3′ splice site ( +4 to +21 ) and region B extending close to the downstream 5′ splice site ( +34 to +51 ) ( Figure 9 ) . Blocking any part of the central core sequence promoted efficient skipping of exon 7 , whereas blocking either of the two inhibitory regions promoted exon 7 inclusion . The 5′ portion of the central core sequence comprises a previously identified Tra2β1-binding motif ( GAAGGA ) [38]; the 3′ part ( AGGTGC ) may represent another cis-acting element that is also crucial for exon 7 recognition . A relatively long conserved sequence in the middle of exon 7 , surrounded by two short inhibitory sequences , was previously identified by an iterative in vitro selection method [43] . Through an analysis of mutability , three segmental sequences with negative or positive cis-acting elements were inferred: the most conserved residues are located in the middle and form a conserved tract ( +16 to +44 ) , whereas the highly mutable upstream nucleotides form an inhibitory region ( +3 to +15 ) , and seven highly mutable downstream nucleotides ( +45 to +51 ) form another negative element [43] . Our data sharpen the boundaries for these three segments , as verified by three different splicing assays and several ASOs , particularly the two most effective ASOs , 07–21 and 34–48 . The target sequence of the 15-mer ASO 07–21 overlaps by six nucleotides with the previously reported conserved tract , and 11 out of 15 target nucleotides of ASO 34–48 are part of this conserved track . Nucleotides A36 and C37 , which fall within the inhibitory region B , have been suggested to be part of a stimulatory motif , as the double mutation A36U/C37U abolishes exon 7 inclusion in the SMN1 gene [43] . However , the double mutation might create a stronger inhibitory motif , or strengthen a secondary RNA structure that impairs exon 7 recognition . We searched exon 7 for putative ESSs with the Web servers PESX ( http://cubweb . biology . columbia . edu/pesx ) and ACESCAN2 ( http://genes . mit . edu/acescan2/index . html ) . PESX found no ESS motifs , whereas ACESCAN2 found three putative ESS motifs: GGTTTT ( +1 to +6 ) , TTTTAG ( +3 to +8 ) and TTCCTT ( +39 to +44 ) . The first two motifs partially overlap with the inhibitory region A , and the last motif resides within the inhibitory region B , so these motifs might contribute to the inhibitory properties of regions A and B . We note that both optimized ASOs are 15-mers , whereas all five of the 18-mer ASOs ( 36–53 , 35–52 , 34–51 , 33–50 , and 32–49 ) displayed more or less inhibitory effects on SMN2 exon 7 inclusion . The longer 18-mer ASOs are likely to overlap two or more binding sites , resulting in complex effects . The strong inhibitory effect of ASO 32–49 and the relatively weak inhibitory effect of ASO 33–50 might be explained by their interference with the positive regulation conferred by the essential core sequence , because the 5′-most boundary nucleotide of their target sequence is 33G and 34C , respectively . The other three ASOs , whose target sequences lie next to the exon–intron junction , might interfere with recognition of the 5′ splice site . A small fraction of the ASOs had inconsistent effects on exon 7 splicing in the different splicing assays . One type of inconsistency occurred between the minigene splicing assays ( both in vitro and in vivo ) and the endogenous gene assay . The other type of inconsistency occurred between the in vitro splicing assay and the in vivo splicing assays . The first type of inconsistency may reflect the differences in the pre-mRNA substrates , although SMN2 exon 7 was identical in all the assays . In both the in vitro splicing assay and the in vivo minigene splicing assay , a small minigene was used . Higher-order structure differences between the small minigene pre-mRNA and the much larger endogenous pre-mRNA might affect the accessibility of some ASOs to their target sequences in exon 7 , resulting in splicing differences . The second type of inconsistency might reflect differences in the splicing-reaction environments between the cell-free splicing assay and the cell-based splicing assays , even though in general , the in vitro splicing reactions accurately reproduce cellular splicing events . For example , it is possible that the effects of some ASOs are more sensitive to changes in the concentration of certain splicing factors resulting from the nuclear extraction procedure . Another possibility is that the discrepancies result from kinetic differences and the coupling between transcription and mRNA processing that normally occurs in vivo , but not in standard in vitro splicing reactions [44]; moreover , transcription from different promoters—as in the minigene and the endogenous gene—can also affect use of alternative cassette exons in vivo [45] . Why ASOs 07–21 and 34–48 are more effective than the original ones , which bind to their target sequences only one or two nucleotides upstream or downstream , is currently under investigation . Because the ASOs have the same lengths and the same or similar G/C contents , annealing kinetics is probably not the cause , assuming equal accessibility to their target sites in exon 7 . Two possible explanations of the effects of ASOs 07–21 and 34–48 are that they precisely block one or more splicing silencers , or efficiently disrupt an inhibitory RNA secondary structure . It has been reported that hnRNP A1 can inhibit SMN2 and SMN1 exon 7 inclusion [12 , 13] . In region A of SMN2 , SF2/ASF presumably fails to block the cooperative propagation of hnRNP A1 , due to the loss of an SF2/ASF motif caused by the C6T transition [13] . ASO 07–21 binds a 15-nt sequence that spans the mutant SF2/ASF motif; therefore , it is reasonable to assume that one of the roles played by ASO 07–21 is similar to a proposed function of SF2/ASF in SMN1 exon 7 , i . e . , blocking the propagation of hnRNP A1 . ASO 07–21 might also block other inhibitory motifs residing in region A , for which an extended inhibitory context has been proposed [46] . We have unexpectedly observed in microwalk A that the 12-mer ASO 04–15 inhibited exon 7 inclusion; ASO 09–20 was likewise inhibitory , at least in vivo . These results suggest that the inhibitory effect of region A is the result of interplay among multiple trans-acting factors . It is likely that ASOs 04–15 and 09–20 can effectively block an undefined weak stimulatory signal , but cannot efficiently block the propagation of hnRNP A1 and possibly other repressors . A terminal-stem loop has been predicted in the less-studied region B , extending to two intron 7 nucleotides [47] . Though it is possible that the inhibitory property of region B is attributable to the proposed RNA secondary structure itself , other possibilities cannot be ruled out , including the existence of one or more splicing repressors that recognize a specific motif or RNA structure . Additional experimental studies will be required to elucidate the precise splicing-stimulatory mechanism of ASO 34–48 . We cannot exclude the possibility that the stimulatory ASOs act in part by stabilizing the pre-mRNAs and/or mRNAs to which they hybridize . Because only some of the exon 7 ASOs stimulate exon 7 inclusion , such a stabilization effect would presumably involve masking of specific instability elements . However , the effect of the ASOs is primarily , if not exclusively , at the level of splicing , because an effect at the level of mRNA half-life should result in the same proportional increase in SMN1 and SMN2 full-length mRNAs , yet the effects on SMN1 were much smaller , consistent with a switch in alternative splicing from the low basal level of exon 7–skipped mRNA . In addition , we did not observe any ASO-mediated stabilization of the labeled pre-mRNA in the in vitro splicing experiments . We used three different assays to demonstrate that the ASOs that promote exon inclusion are compatible with synthesis of full-length SMN protein , indicating that they do not block mRNA export or translation . First , we showed that ASOs 07–21 and 34–48 promote efficient expression of reporter proteins corresponding to exon 7 inclusion in HEK293 cells . Second , we showed that the same ASOs promote expression of full-length SMN protein in primary fibroblasts from a type I SMA patient; though significant , this effect appears subtle because of the relatively low transfection efficiency with these cells . Finally , we showed that the same ASOs promote an increase in the number of nuclear gems detected with anti-SMN antibody in the patient fibroblasts; this single-cell immunofluorescence assay obviates the problem of transfection efficiency and provides an indirect readout for SMN protein levels [4 , 41] . Our initial goal in this study was to define optimal antisense targets to maximize ESSENCE effectiveness . However , our optimized antisense targets should also be useful with bifunctional ASOs [30 , 31] . The antisense portions of the bifunctional ASOs used in two earlier studies were not optimized . The first study used a 15-mer ASO targeting the exon 7 sequence +2 to +16 [30] , i . e . , the same target sequence as that of our ASO 02–16 . However , we found that this ASO had a slightly negative effect on exon 7 inclusion in vivo with both the SMN2 minigene and the endogenous SMN2 gene , though it had a slightly stimulatory effect on splicing in vitro ( Figures 3B , 5A , and 5B ) . ASO 02–16 was one of few ASOs that gave inconsistent effects among the different splicing assays we used . The second study used a 20-mer ASO targeting the exon 7 sequence +6 to +25 [31] , which partially overlaps the central core sequence that is essential for exon 7 inclusion . On the basis of our results , we predict that more-effective bifunctional ASOs targeting exon 7 can be designed , using the ASO 07–21 or 34–48 sequences as the antisense moiety . Although we expect that even more-potent effectors can be designed by attaching ASOs 07–21 or 34–48 to activation-domain peptides or SR protein–binding sites—as in the ESSENCE [34] and TOES [29] methods—these ASOs appear to be sufficiently effective on their own to be tested in animal models of SMA . ASOs with the same chemistry were recently demonstrated to be active modulators of splicing in mice [48] . More generally , the antisense walk strategy described here should be applicable to identifying exon-targeting ASOs that modulate alternative splicing of normal or mutant genes . At the same time , these ASOs can facilitate the identification of important cis-regulatory sequences that can be further investigated to elucidate splicing mechanisms , and to identify trans-acting factors that may be useful molecular targets for disease therapy .
Synthesis and purification of chimeric 2′-O-methoxyethyl–modified oligonucleotides with a phosphodiester backbone were performed using an Applied Biosystems 380B automated DNA synthesizer ( Applied Biosystems , Foster City , California , United States ) as described previously [49] . The oligonucleotides were dissolved in water . The sequences of all the oligonucleotides are indicated in Table I . SMN minigene constructs used for both the in vitro splicing assay and the in vivo minigene splicing assay in HEK293 cells were pCI-SMN1 and pCI-SMN2 , which are derivatives of pCISMNxΔ6-WT and pCISMNxΔ6-C6T , respectively [11 , 50] . The SMN minigenes comprise the 111-nt–long exon 6 , a 200-nt shortened intron 6 , the 54-nt exon 7 , the 444-nt intron 7 , the first 75 nt of exon 8 , and a consensus 5′ splice site ( CAGGTAAGTACTT ) to promote exon definition in vitro . We amplified the minigenes with primer set T7-F2 ( 5′-TAC TTA ATA CGA CTC ACT ATA GGC TAG CCT CG-3′ ) and SMN8-75-5′R-SalI-NotI ( 5′-GAA GCG GCC GCG TCG ACA AGT ACT TAC CTG TAA CGC TTC-3′ ) , and inserted them into vector pCI-neo ( Promega , Madison , Wisconsin , United States ) using XhoI and NotI . Two restriction sites , SalI and NotI , were included in the reverse primer . Four minigene constructs were used for protein expression: constructs pEGFP-SMN2 and pEGFP-SMN1 have the same minigenes as in pCI-SMN2 and pCI-SMN1 , but without the consensus 5′ splice site . Minigene inserts were amplified with primer set Xho-HA-E6-F ( 5′-GAT CTC GAG AGT ACC CAT ACG ACG TAC CAG ATT ACG CTA TAA TTC CCC CAC CAC CTC CC-3′ ) and Pr-E8–75-BamH-R ( 5′-CGG GAT CCT AAC GCT TCA CAT TCC AGA TC-3′ ) , digested with XhoI and BamHI and subcloned into the XhoI and BamHI sites in pEGFP-C1 ( Clontech , Mountain View , California , United States ) . Constructs pEGFP-SMN2Δ6 and pEGFP-SMN1Δ6 are similar to pEGFP-SMN2 and pEGFP-SMN1 , but most of exon 6 was deleted , with only the last nine nucleotides retained to keep the natural context of the 5′ splice site of exon 6 . Minigene inserts were amplified with Pr-HA-I6-Xho-F ( 5′-ATC TCG AGA GTA CCC ATA CGA CGT ACC AGA TTA CGC TTA TTA TAT GGT AAG TAA TCA CTC-3′ ) and Pr-E8–75-BamH-R , and were subcloned into XhoI and BamHI double-digested pEGFP-C1 . One copy of the HA epitope tag sequence was included in primers Xho-HA-E6-F and Pr-HA-I6-Xho-F , so that all protein products expressed from the four EGFP-fused minigene constructs have an N-terminal HA epitope tag and can be detected with anti-HA antibody . To generate in vitro splicing substrates , we linearized plasmids pCI-SMN1 and pCI-SMN2 with SalI , and transcribed them with T7 RNA polymerase ( Promega ) in the presence of α-32P-UTP and 7MeGpppG cap analog [11 , 51] . Substrates were purified by denaturing polyacrylamide gels and spliced in HeLa cell nuclear extract , as described [52] . We incubated 8 fmol of transcript in 10-μl splicing reactions containing 3 μl of nuclear extract and 1 . 6 mM MgCl2 . After incubation at 30 °C for 4 h , we extracted the RNA and analyzed it on 8% denaturing polyacrylamide gels , followed by autoradiography and phosphorimage analysis with an Image Reader FLA-5100 ( FujiFilm Medical Systems , Stamford , Connecticut , United States ) . We calculated exon 7 inclusion as a percentage of the total amount of spliced mRNA , i . e . , included mRNA × 100/ ( included mRNA + skipped mRNA ) . The signal intensity of each mRNA isoform band was normalized according to its U content . HEK293 cells , SMA type I homozygous and carrier fibroblasts ( 3813 and 3814; Coriell Cell Repositories , Camden , New Jersey , United States ) were cultured in Dulbecco's modified Eagle's medium ( DMEM; Invitrogen , Carlsbad , California , United States ) containing 10% ( v/v ) fetal bovine serum and antibiotics ( 100-U/ml penicillin G and 100-μg/ml streptomycin ) . For transfection of MOE oligonucleotides and plasmids into HEK293 cells , electroporation was used , followed by puromycin selection ( Figures 2 and 5–7 ) . Briefly , 90 μl of freshly-prepared cells ( 1 × 107/ml ) suspended in Opti-MEM medium was mixed with 10 μl of ASO/DNA mixture in a 1-mm cuvette . A total of 2 . 5 μg of the plasmid pBabe Puro [53] was included in the ASO/DNA mixture for selection . For the in vivo minigene splicing assay and the minigene protein assays , 5 μg of each minigene plasmid was used . For electroporation , 80 volts and 500 μF were applied with a Gene Pulser II apparatus ( Bio-Rad , Hercules , California , United States ) . Shocked cells were then cultured in 60-mm dishes in normal medium . Untransfected cells were killed by treatment with 2-μg/ml puromycin for 20 h . For transfection of MOE oligonucleotides into 3813 cells , Lipofectin ( Invitrogen ) was used according to the manufacturer's instructions . Total RNA was isolated with Trizol reagent ( Invitrogen ) , and 2 μg of each RNA sample was used per 20-μl reaction for first-strand cDNA synthesis with Oligo-dT and Super Script II reverse transcriptase ( Invitrogen ) . Splicing products were amplified semi-quantitatively using 22 PCR cycles ( 94 °C for 30 s , 55 °C for 30 s , and 72 °C for 36 s ) . Primer set T7-F2 ( see above ) and E8–75+5′R ( 5′-AAG TAC TTA CCT GTA ACG CTT CAC ATT CCA GAT CTG TC-3′ ) was used to amplify minigene pCI-SMN2 transcripts . Primer set EGFP-C ( 5′-CAT GGT CCT GCT GGA GTT CGT G-3′ ) and Pr-E8–75-BamH-R was used to amplify transcripts from plasmids pEGFP-SMN1 , pEGFP-SMN2 , pEGFP-SMN1Δ6 , and pEGFP-SMN2Δ6 . Primer set E3-F ( 5′-ACT TTC CCC AAT CTG TGA AGT A-3′ ) and E8-R ( 5′-CAT TTA GTG CTG CTC TAT GCC AGC-3 ) was used to amplify endogenous SMN2 transcripts in 3813 cells; the GAPDH mRNA in 3813 cells was used as an internal control , and was amplified with primers GAPDH-F ( 5′-AAG GTG AAG GTC GGA GTC AAC GG-3′ ) and GAPDH-R ( 5′-CCA CTT GAT TTT GGA GGG ATC TC-3′ ) . Primer set E6-F ( 5′-ATA ATT CCC CCA CCA CCT CCC-3′ ) and E8–467 ( 5′-TTG CCA CAT ACG CCT CAC ATA C-3′ ) was used to amplify endogenous SMN1/2 transcripts in HEK293 cells , and the PCR products were then digested with DdeI to distinguish their origin ( SMN1 or SMN2 ) . All PCR products were labeled with α-32P-dCTP and separated on 6% or 8% native polyacrylamide gels . The extent of exon 7 inclusion was calculated as above , and the signal intensity of each cDNA band was normalized according to its G+C content . All curves were plotted with Prism software ( Graphpad Software , http://www . graphpad . com ) . To compare the effects of ASOs 07–21 , 34–48 , 06–20 , and 36–50 on exon 7 inclusion in vitro , sigmoidal curve-fitting was used , with X values plotted on a logarithmic scale . The EC50 value of each ASO was calculated by the software . Sigmoidal curves were also plotted for the effects of ASOs 07–21 and 34–48 , respectively , on endogenous SMN genes in HEK293 cells . The time-course analysis of ASOs 07–21 and 34–48 in HEK293 cells was plotted using exponential-decay curve-fitting . Two or 3 d after transfection , HEK293 , 3813 , or 3814 cells were harvested and lysed with lysis buffer ( 50 mM Tris , [pH 7 . 6] , 150 mM NaCl , 2 mM EDTA , 1 . 0% Nonidet P-40 , 1 mM PMSF ) , supplemented with protease inhibitor cocktail ( Roche Applied Science , Indianapolis , Indiana , United States ) . Protein samples were separated by SDS-PAGE ( sodium dodecyl sulfate -polyacrylamide gel electrophoresis ) and transferred to nylon membranes . The blots were probed with monoclonal mouse anti-SMN ( BD Transduction Laboratories; BD Biosciences , San Jose , California , United States ) , monoclonal mouse anti-HA ( HA . 11; BAbCO , Richmond , California , United States ) , or monoclonal mouse anti–α-tubulin ( Sigma , St . Louis , Missouri , United States ) antibody , followed by secondary goat anti-mouse antibody labeled with yellow-fluorescent Alexa Fluor 532 dye ( Molecular Probes/Invitrogen ) . The signal was detected with an Image Reader FLA-5100 ( FujiFilm Medical Systems ) . The 3813 and 3814 fibroblasts were plated on coverslips and grown in DMEM . After overnight growth , the cells were transfected with 100 nM oligonucleotides . At 48 h post-transfection , the cells were washed with phosphate-buffered saline ( PBS ) and fixed with 3 . 7% paraformaldehyde in PBS for 20 min . After washing with PBS , the cells were permeabilized in 0 . 1% Triton X-100 in PBS for 3 min , blocked for 30 min in blocking buffer ( 5% goat serum , 3% IgG-free bovine serum albumin [BSA] ) , and incubated for 3 h in blocking buffer containing mouse monoclonal anti-SMN antibody . The cells were then washed three times with PBS and incubated for 2 h in blocking buffer containing goat anti-mouse secondary antibody Alexa Fluor 594 ( Molecular Probes/Invitrogen ) . The cells were again washed with PBS three more times , and coverslips were mounted with Prolong Gold mounting solution containing DAPI ( Molecular Probes/Invitrogen ) for nuclear staining . Cells were analyzed using an Axioplan 2i fluorescence microscope ( Carl Zeiss , Thornwood , New York , United States ) equipped with Chroma filters ( Chroma Technology , Rockingham , Vermont , United States ) . OpenLab software ( Improvision , http://www . improvision . com ) was used to collect digital images from a CCD camera ( Hamamatsu , Bridgewater , New Jersey , United States ) . To count gems , first a field was randomly selected using the DAPI channel , the cell number was determined based on the DAPI counterstaining , and then the gems were counted in the red channel . At least three different fields were counted , and each field contained more than 40 cells .
The UniProt ( http://www . pir . uniprot . org ) accession number for SMN is Q16637–1 . The Entrez Gene ( http://www . ncbi . nlm . nih . gov/entrez/query . fcgi ? db=gene ) GeneID numbers for the genes discussed in the paper are SMN1 ( 6606 ) and SMN2 ( 6607 ) .
|
Spinal muscular atrophy ( SMA ) is a severe genetic disease that causes motor-neuron degeneration . SMA patients lack a functional SMN1 ( survival of motor neuron 1 ) gene , but they possess an intact SMN2 gene , which though nearly identical to SMN1 , is only partially functional . The defect in SMN2 gene expression is at the level of pre-mRNA splicing ( skipping of exon 7 ) , and the presence of this gene in all SMA patients makes it an attractive target for potential therapy . Here we have surveyed a large number of antisense oligonucleotides ( ASOs ) that are complementary to different regions of exon 7 in the SMN2 mRNA . A few of these ASOs are able to correct the pre-mRNA splicing defect , presumably because they bind to regions of exon 7 that form RNA structures , or provide protein-binding sites , that normally weaken the recognition of this exon by the splicing machinery in the cell nucleus . We describe optimal ASOs that promote correct expression of SMN2 mRNA and , therefore , normal SMN protein , in cultured cells from SMA patients . These ASOs can now be tested in mouse models of SMA , and may be useful for SMA therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"primates",
"biochemistry",
"cell",
"biology",
"homo",
"(human)",
"animals",
"neuroscience",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2007
|
Enhancement of SMN2 Exon 7 Inclusion by Antisense Oligonucleotides Targeting the Exon
|
Salt fortified with the drug , diethylcarbamazine ( DEC ) , and introduced into a competitive market has the potential to overcome the obstacles associated with tablet-based Lymphatic Filariasis ( LF ) elimination programs . Questions remain , however , regarding the economic viability , production capacity , and effectiveness of this strategy as a sustainable means to bring about LF elimination in resource poor settings . We evaluated the performance and effectiveness of a novel social enterprise-based approach developed and tested in Léogâne , Haiti , as a strategy to sustainably and cost-efficiently distribute DEC-medicated salt into a competitive market at quantities sufficient to bring about the elimination of LF . We undertook a cost-revenue analysis to evaluate the production capability and financial feasibility of the developed DEC salt social enterprise , and a modeling study centered on applying a dynamic mathematical model localized to reflect local LF transmission dynamics to evaluate the cost-effectiveness of using this intervention versus standard annual Mass Drug Administration ( MDA ) for eliminating LF in Léogâne . We show that the salt enterprise because of its mixed product business strategy may have already reached the production capacity for delivering sufficient quantities of edible DEC-medicated salt to bring about LF transmission in the Léogâne study setting . Due to increasing revenues obtained from the sale of DEC salt over time , expansion of its delivery in the population , and greater cumulative impact on the survival of worms leading to shorter timelines to extinction , this strategy could also represent a significantly more cost-effective option than annual DEC tablet-based MDA for accomplishing LF elimination . A social enterprise approach can offer an innovative market-based strategy by which edible salt fortified with DEC could be distributed to communities both on a financially sustainable basis and at sufficient quantity to eliminate LF . Deployment of similarly fashioned intervention strategies would improve current efforts to successfully accomplish the goal of LF elimination , particularly in difficult-to-control settings .
Lymphatic filariasis ( LF ) , a mosquito-borne neglected tropical disease ( NTD ) , commonly known as elephantiasis , is one of only parasitic six diseases currently targeted for potential global eradication by 2020 using preventive mass chemotherapy [1–3] . Despite the impressive expansion of a WHO-led elimination program aimed toward the meeting of this goal in all endemic countries since 2000 , stakeholders committed to global LF elimination have recognized that the current tablet-based mass drug intervention is resource-intensive , can face significant compliance issues with time , and may be difficult to implement in remote or socio-ecologically complex areas , such as urban and socio-politically unstable settings , hampering foreseen elimination goals [4–8] . These difficulties have heightened interest in investigating the impacts of either approaches aimed at scaling-up treatment strategies or inclusion of preventive activities into drug programs ( such as supplemental vector control ) , or evaluation of novel intervention technologies , that can effectively overcome current barriers in order to accelerate parasite elimination [9–11] . Salt fortified with the anti-filarial drug , diethylcarbamazine ( DEC ) , could offer an intervention that avoids many of the above issues connected with tablet-based elimination programs [12–18] . Indeed , DEC-salt has played a major role in the elimination of LF in a number of pilot and region-wide settings in Africa , Central America , and Asia [5 , 12–17] . The low dose of DEC ( 0 . 1–0 . 6% [w/w] ) used in these studies and programs was well tolerated and rarely associated with adverse reactions . It also has the potential to be more effective than tablet-based Mass Drug Administration ( MDA ) programs via reduction of the durations of intervention required to interrupt parasite transmission [18] . Moreover , fortified salt can also be provided to a population without developing a dedicated public distribution system , overcoming the need for developing an effective health infrastructure capable of distributing anti-filarial drugs at the high coverages needed for achieving elimination [5 , 7 , 8] . Haiti is one of only four countries remaining in the Americas where LF is still endemic [8 , 19] . MDA using DEC first started in the country under the National Program to Eliminate LF ( NPELF ) in 2000 , and by 2005 had expanded to mass treatment of some 1 . 6 million people at least once in 24 of the initially endemic 120 communes of the country ( [8] . However , following funding , sociopolitical , and natural disaster-based challenges to further scaling up , the program realized full national coverage of all the endemic communes in the country only by 2012 [5 , 7 , 8 , 20] . This delay together with the technical challenge of interrupting transmission in areas of highest prevalence even with high levels of coverage using the suggested five successive years of annual MDA [5 , 7 , 8 , 20] , indicate that it is unlikely that NPELF will meet the goal of accomplishing LF elimination in the country by the target year of 2020 . In 2006 , partly to overcome the above issues , a project was initiated with the collaboration of Congregation de Sainte Croix , the Notre Dame Haiti Program , and the Ministry of Public Health and the Population ( MSPP ) , focused on the local processing and marketing of DEC-mediated salt co-fortified with potassium iodate as an alternative means to facilitate the elimination of LF ( and prevention of iodine deficiency disorders ) in Haiti [8] . Based on the principle of employing a social enterprise framework for providing goods and services in an entrepreneurial and innovative fashion to solve social problems [21–23] , this project purchases both local and imported raw salt which are then cleaned , sized , fortified and packaged for sale in the local market as food-grade co-fortified salt , raising the potential of using a business-based approach to delivering DEC-medicated salt as a sustainable means to accomplish LF elimination in settings , such as Haiti . Here , our major aim was to examine the economic performance and effectiveness of using the Haitian social enterprise-based framework for producing and marketing DEC-fortified salt as a sustainable , cost-effective , model for achieving the long-term elimination of LF , focusing on Léogâne arrondissement , Haiti . A cost-revenue analysis combined with a mathematical modeling-based evaluation of the cost-effectiveness of the DEC salt social enterprise compared to standard MDA was carried out to undertake this analysis . Specifically , we evaluated the economic performance and social value of the enterprise by assessing: 1 ) the growth in salt production , costs of resources consumed , and revenues from sales gained to determine break-even points , 2 ) the impact of the product-mix used for realizing the socially-relevant sale price of the salt , and 3 ) its cost-effectiveness compared to tablet-based MDA for accomplishing LF elimination in the study setting of Léogâne arrondissement , Haiti . We discuss the results in terms of how using a social enterprise can offer a sustainable and innovative strategy for accomplishing LF elimination in Haiti , and similarly resource-constrained settings , that face both programmatic and social difficulties in delivering long-term tablet-based LF MDAs .
We carried out a cost-revenue analysis to evaluate the production capability and financial feasibility of the developed DEC salt social enterprise via assessment of the relationship between fixed and variable costs versus the revenue received [24–27] and a modeling study centered on applying a dynamic transmission model to evaluate the cost-effectiveness of using this intervention versus standard annual MDA for eliminating LF in Léogâne arrondissement , where the salt enterprise operates [28–31] . The cost-revenue analysis was based on costs and revenue data contained in financial accounts during the production phase of the salt enterprise from 2013 to 2018 , while the break-even analysis was carried out over a time horizon that ranged between 2013 to the year when the break-even point was attained . The predicted timelines , in months or years , to LF elimination along with the costs of annual MDA versus the net cost of supplying DEC-fortified salt until elimination was achieved were used to carry out the cost-effectiveness modeling study .
The total annual costs of salt production ( = Investment + Fixed Costs + Variable Costs ) and revenues attained from the sale of all three types of salt are given in Table 2 . Total annual costs increased as production expanded ( Table 1 ) from US$210 , 206 in 2013 to US$1 , 175 , 000 in 2018 –i . e . , approximately five times–but the figures show that revenue increased even faster , up to 17 times that obtained initially in 2013 ( US$61 , 636 ) to close to cost of production by 2018 ( US$1 , 064 , 000 ) . The net cash flow [24–27] figures in the Table , which represent the difference between total production cost and total revenue , although being negative for all the years from 2013–2018 , capture this increasing revenue returns ( as reflected by the declining negative trend in the net cash flow ) towards the later years , suggesting that the project is close to achieving break-even or profitability in the near future . To estimate the exact time point when the enterprise is likely to break-even ( i . e . , the time point when the total program cost is equal to the total revenue ) , we employed a simple linear model to project forward the total project costs and revenues calculated for 2013–2018 . Fig 1 ( A ) shows that if we use the full data on annual costs and revenues obtained for the whole 2013–2018 period , the project will break-even in 2027 . However , if we use the data from 2016–2018 , when total salt production had reached significant levels ( Table 1 ) , to predict the time point at which the break-even point will be achieved , this will occur earlier by year 2022 ( Fig 1 ( B ) ) . Fig 2 depicts the per ton revenues from sales and costs of producing the three different types of salts for the years 2013–2018 . The results show that for all salt types , while initially there was a large difference between the production cost and revenue per ton ( i . e . , a large negative cash-flow ) , this difference decreased for each salt category with time . This occurred faster , however , in the case of the industrial and single-fortified salt , such that break-even was achieved by 2018 . By contrast , the cash-flow from the production and sale of the double-fortified salt was still negative ( lower revenue compared to production costs ) at the end of the present study period of 2018 . This result shows , first , that the overall break-even estimated in this study ( Fig 1 ) for the project is due to the delay in reaching the break-even year for double-fortified salt . Second , it also highlights how the mixed product strategy of producing different types of salt targeting different market sectors can allow the more profitable products ( industrial , single-fortified salt ) to subsidize the sale of a product ( double-fortified salt ) whose cost ( approximately US$200 per ton ( Fig 2 ) ) needs to be kept competitive with other edible salt sold ( retailed at $US265 per ton in Haiti ) in the local market . This was evaluated by analysis of the difference in the variable costs of producing the single-fortified ( which included only potassium iodate ) versus the double-fortified ( both potassium iodate and DEC included ) salt types , given that the investment and fixed costs going into manufacturing all salt types were shared equally between each type . The variable costs per ton for producing the single-fortified ( coarse and fine ) and double-fortified ( coarse and fine ) salt for two types of bags/bales ( 25 . 0-kg bags and 12 . 5-kg bales ) are provided in Table 3 for the years 2014–2018 when both salt products were produced ( note that production of single-fortified salt began only in 2014 ( Table 1 ) ) . Analysis of the difference in the variable costs for producing the single-fortified versus the double-fortified salt indicate that this was consistently about US$70 per metric ton , irrespective of which type—coarse or fine variety—or types of bags/bales were produced ( Table 3 ) . Given that the price of DEC , as delivered by Syntholab Chemicals to the project was US$21 . 60 per kg ( James Reimer , personal communication ) , and DEC salt in this project was fortified with 0 . 32% DEC by weight or with 3 . 2kg of DEC/ton , it can be seen that the cost of producing one metric ton of DEC salt works to be US$ 69 . 12/ton ( i . e . , 3 . 2 kg DEC x US$21 . 60 per kg ) . This result indicates that the marginal cost , or difference in the variable cost , of adding DEC to single-fortified salt ( US$70; Table 3 ) was simply due to the purchase price of DEC . Table 4 shows the potential increase in demand for salt in Léogâne arrondissement calculated as a function of changes in population size from 2013 to 2018 . It also presents the potential DEC-fortified salt coverage which may be achieved in the setting by increasing sale of the double-fortified salt produced over this period . The annual population size estimates from 2013 onwards were predicted using a growth rate of 1 . 28% [39] , whereas the yearly population demand for edible salt was calculated by assuming that the daily salt consumption per person is 15gm ( average of the reported daily per-capita consumption in Haiti of 10gm and 19gm [5 , 40] ) . Assuming that the sale of double-fortified salt is widespread in the community ( i . e . , not targeted towards one segment of the population ) , coverage of DEC salt in Léogâne can then be roughly estimated simply by dividing the quantity of salt produced over the estimated demand . The results from this calculation , listed in the last column of Table 4 , shows that as production increased rapidly from 2013 to 2018 , this would increase potential population coverage achieved by sales of the DEC-medicated salt from as low as 8 . 45% in 2013 to approximately 65% in 2018 . Fig 3 portrays the predicted timelines to LF elimination in Léogâne under each of the MDA and salt interventions investigated . For MDA , the depicted simulations indicate that it would take up to 84 months at 65% coverage , and 60 months at 80% coverage , respectively , to reach the 1% mf threshold . By contrast , the model predictions show that it will take just 1 year ( 12 months ) at 65% coverage , 5 months at 80% coverage , and 3 years or 36 months if actual population coverage ( Table 4 ) is used to reduce the pre-control prevalence to below this threshold via consumption of the traded DEC salt ( Fig 3 ) . This highlights the dramatic effect that daily consumption of DEC-medicated salt even at low dosages ( 0 . 32% w/w ) would have compared to annual intake of higher dosages of DEC ( and ALB ) as provided by tablet-based MDA for eliminating LF infection in an endemic setting [12 , 15–18] . Note that the actual DEC salt population coverages ( Table 4 ) used in this analysis assumed that salt supply occurred uniformly and sale was restricted to Léogâne arrondissement only . Any changes in these parameters would mean attaining lower annual population coverages than shown in Table 4; however , a sensitivity analysis using coverage values 15–20% lower than those depicted in the Table did not affect the above timelines significantly . The comparative costs of carrying out annual MDA versus supplying DEC-medicated salt are shown in Table 5 . These show that while the total cost of delivering annual MDA ( here fixed at US$0 . 64 per person , inclusive of the cost of the drug [29 , 38] ) , simply scaled with population growth , and will continue to be substantial on a yearly basis until LF elimination is achieved , the net cost of supplying DEC-fortified salt through the social business model will decline dramatically with time as production costs decrease and revenues begin to increase over time ( Fig 2 ) . Indeed , projection forward of the net or revenue—production cost data collected during the years 2013–2018 indicates that the total and per capita net costs of DEC salt provided through the present social enterprise could potentially even become zero at the time point ( 2027 ) when the project breaks even ( Fig 1 ( A ) ) . We used the total costs of implementing each strategy until LF elimination ( crossing below the 1% mf threshold ) is achieved to investigate the cost effectiveness of either strategy . The total costs and effectiveness of using MDA at 65% and 80% coverages and supplying DEC salt at 65% , 80% and the actual coverages given in Table 4 , were evaluated and compared via calculations of the average and incremental cost effectiveness ratios [10 , 41–47] . With respect to DEC salt , we also conducted the analysis using three different net costs of salt production per person per year as recorded in Table 5: ( i ) average of the net cost of salt produced per person calculated for the years 2018–2020 , i . e . , US$0 . 57 per person/year , ( ii ) average of the net cost of salt per person for the years 2021–2023 , i . e . , US$0 . 3 per person/year , and ( iii ) average of the net cost of salt per person for the years 2025–2027 , i . e . , US$0 . 025 per person/year . This was performed to assess the sensitivity of the present results to changes in the steeply declining net cost of salt production over time observed in this study . The results from this exercise are shown in Table 6 , and indicate , principally , that irrespective of coverage , the costs of using MDA are significantly greater than those arising from using the salt strategies , primarily because of its lesser effectiveness as well as higher and stationary unit cost . Among the three salt scenarios , costs for eliminating LF , as expected , declined with decreasing net cost of production over time with scenario three showing the lowest costs . However , for all strategies , while the most effective strategy is to deliver MDA or salt at 80% coverage , the most cost-effective option ( in terms of the incremental cost-effectiveness ratio ( ICER ) ) also occurred at this coverage level for both MDA and the DEC salt strategies with incremental costs of either of these options being negative and incremental effects positive over their corresponding next-effective alternative ( ICER: US$88 , 000 per intervention year saved by the 80% MDA strategy compared to an average CER of US$17 , 333 for carrying out 65% MDA using the cost of implementing annual MDA fixed at US$0 . 64 per person/year in Léogâne [38] , and an additional saving of US$3 , 876 per intervention year saved by the 80% DEC salt option over the strategy delivering DEC salt at actual recorded coverages when the net cost of delivering medicated salt was fixed at US$0 . 57 per person ( based on durations of interventions in years and total costs given in Table 6 ) ) . Note moving from delivering DEC salt at actual recorded coverages to providing salt at 65% coverage results in higher predicted total costs but also a saving of 2 years , and so this strategy is dominated by the strategy that delivers DEC salt at 80% coverage which results in extra reductions in both costs and the time needed to accomplish LF elimination ( Table 6 ) . Similar results were also obtained when the other two net costs of producing DEC salt per person were used in carrying out these calculations . Overall , thus , these results indicate that because of: 1 ) decreasing net cost of DEC salt production over time , and 2 ) expansion of its coverage in the population leading to significantly reduced elimination timelines , the delivery of DEC through a salt enterprise may be significantly more cost-effective than annual DEC tablet-based MDA for accomplishing LF elimination in the Léogâne arrondissement setting . It is also to be appreciated that we used a conservative treatment cost of $0 . 64 per person for modeling the cost-effectiveness of the tablet-based MDA program in this study . This represented a best-case scenario for the MDA program implemented in Léogâne given that the actual drug costs started out higher ( US$1 . 84 over the first 3 MDAs ) before approaching the stabilized value used in our analysis as the program became more efficient [29 , 38] . Indeed , a recent systematic review indicated that MDA program costs can vary substantially between settings , with an average cost that could reach as high as US$1 . 32 [48] . Use of such values or inclusion of the actual change observed in the per person treatment cost over time in Léogâne in the present analysis would clearly further increase the cost of MDA over that presented in Table 5 , which in turn would lead to an even higher cost-effectiveness ratio for MDA compared to those estimated for DEC salt in this study ( Table 6 ) . Nonetheless , depending on the cost of producing and delivering DEC-medicated salt , it is readily apparent that the cost savings to a provider of utilizing a social enterprise framework for DEC delivery can be remarkably high . For example , even at the average cost of US$0 . 907 per person/year ( the worst-case scenario using cost data from 2015–2018 ( Table 5 ) ) , delivery of DEC salt through the current enterprise for eliminating LF in Léogâne is predicted to result in total costs of US$453 , 386 and US$190 , 423 for achieving 65% and 80% coverage of the population respectively , which amounts to only 31% and 15% of the corresponding predicted total costs of using annual MDA at these coverages for achieving the same objective in this study location .
In this study , we have undertaken a performance assessment of a novel social enterprise , developed through a collaboration between Haitian and international partners engaged with LF control in the country , as a means to enhance the delivery of anti-filarial drugs to populations through the trading of salt co-fortified with DEC . Although DEC-fortified salt has been used previously in both pilot and region-wide LF intervention programs in a variety of global regions , ranging from Brazil , Tanzania , India and China to effectively control or eliminate LF [5 , 12–17] , it is to be noted that the developed Haiti salt enterprise is the first attempt anywhere in the world to apply the principles of social entrepreneurship for delivering such an intervention . Recent work has highlighted how such social enterprises–that is , a social mission-driven organization that trades in goods or services for a social purpose–are emerging as a potentially effective supply side solution to the provision of cost-efficient public services in response to government failures , business that seek to extract maximal returns on investment , and unstable non-profit organizations [21–23] . In particular , this work has shown how these business entities can solve social problems via their potential to deliver greater responsiveness , efficiency and cost-effectiveness , through an explicit focus on meeting specific social goals while operating with the financial discipline and innovation of a private-sector business [21–23 , 49 , 50] . Although there is continuing debate as to how best to evaluate the performance of social enterprises , it is clear that at least two basic components related to the bottom line of these entities require assessment [21 , 23] . These primarily include: the economic-financial component for measuring overall organizational efficiency , profitability and hence sustainability , and the social effectiveness of the enterprise [21–23] . Here , we have combined the tools of financial accounting and modeling of cost-effectiveness to measure these components in order to present a first analysis of the utility of using the developed Haitian DEC salt enterprise as a sustainable and economically efficient strategy to bring about LF elimination in programmatically difficult-to-control settings , like the arrondissement of Léogâne , Haiti [10 , 41–47] . Our analysis of the performance of the present salt enterprise for creating social value first focused on the question of capacity to economically produce sufficient amounts of DEC-fortified salt for significantly affecting the elimination of LF in the study setting . The production figures shown in Table 1 indicate , firstly , that while initial production of all types of salt were low during the initial years of operation , by 2018 , and just 5 years after processing began in 2013 , the enterprise had reached high levels of both total ( 5 , 845 metric tons ) and DEC-fortified salt ( 1 , 841 metric tons ) production , respectively ( Table 1 ) . Our analysis of the population coverage that the sale of DEC salt could provide demonstrate that the amounts produced could have potentially resulted in a drug coverage rate of 65% by 2018 ( Table 4 ) , which our previous modeling study [18] and the present cost-effectiveness exercise ( Fig 3 and Table 6 ) indicate is sufficient to accelerate the achievement of LF elimination in the Léogâne setting . These findings suggest that as the result of the expansion phases carried out through new capital investments ( Table 1 ) , the current DEC salt project may have reached the production capacity required to achieve its stated social mission of using a market-based approach for delivering sufficient edible DEC-medicated salt as a means to bring about efficient LF transmission interruption in the present study setting . Assessment of the economic and financial performance of the salt enterprise carried out in this study using cost-revenue analysis and financial forecasting has provided further insights regarding the organization efforts used to reach economic equilibrium and hence trading viability . This is an important consideration for evaluating the performance of social enterprises because first and foremost these entities are enterprises , and therefore their social goals can be pursued only by ensuring economic and financial fidelity [21–23] . Our major result in this area is providing clarity regarding how the enterprise’s achieved outputs in salt production and sale may affect its potential to reach break-even points ( Fig 1 ) . Specifically , we show that given the observed trends in production costs and revenue to 2018 ( Table 2 ) , the present salt enterprise may either break-even by 2027 ( if we forecast linearly using all the data from 2013 to 2018 ) , or as early as by 2022 ( if we use data collected during 2016–2018 after the project had significantly expanded capacity ) . This result is clearly dependent on assuming that capacity to produce the increased amount of salt to meet either break-even points is available within the enterprise without any further expansion , and demand for the produced salt in all sectors ( industrial to edible salt markets ) will also expand commensurately . Nonetheless , the finding that it might be possible to reach the break-even point by 2022 ( i . e . , over the next 4 years ) is encouraging , and suggests that the enterprise is likely to be self-sustaining and could become profitable in the very near future . Indeed , analysis of trends in costs of production and revenues gained per ton of each category of salt ( Fig 2 ) indicate that both the industrial and single-fortified salt categories may already have reached their individual break-even points in 2018 , and that the delay for achieving break-even status by the social enterprise is primarily due to the lag experienced by the production and sale of the double-fortified salt . Although the per ton production cost of the latter salt declined as significantly over time as the other two salt categories ( Fig 2 ) , indicating the achievement of considerable economics of scale , the need to keep the price of the DEC salt below the marginal cost of adding the drug ( $70/metric ton ( see Table 3 ) ) to compete with untreated local edible salt in the market means that either: 1 ) the current market price of the double-fortified DEC salt needs to be revised upwards , or 2 ) further economies of scale need to be found to bring down production costs , or 3 ) cross-subsidy from the more profitable categories of salt produced will be required in order to continue with the processing of DEC salt in this setting . While on the one hand , such a capacity to use a product mix strategy innovatively as a means to subsidize the marketing of a product for meeting a social need is a feature of using an enterprise model , note that this may be a particular effect of developing markets in settings , such as Léogâne and Haiti in general , where a strong market-based economy is only just evolving . For other LF endemic settings with stronger market economies and established salt industries , the need for such subsides may be significantly lower meaning that the sale of DEC salt could occur at nearer the true marginal cost of production , i . e . , at the actual cost of purchasing the drug itself . Note also that our present forecasts do not fully consider the likely impacts of key swing factors that may significantly affect the profitability of the salt social enterprise , such as enforcement of the 2017 law requiring all food salt in Haiti to be fortified , further progress on market segmentation and the resulting product mix , and significant weather events similar to Hurricane Michael in 2016 . Positive changes in the first two factors will clearly enhance the enterprise’s ability to break-even faster and hence attain profitability sooner than predicted in this work . The cost-effectiveness modeling exercise carried out in this study showed that apart from the efficiency of the business model used for achieving economic and financial sustainability , the salt enterprise may also be more cost-effective than the standard tablet-based annual MDA program for accomplishing LF elimination in Léogâne ( Table 6 ) . This is because not only will the population coverage that can be potentially attained by sale of the DEC salt ( Table 4 ) be sufficient to make this strategy more effective than annual MDA in reducing the number of years ( 3 versus 7 years ) required for achieving LF elimination ( as defined by reducing mf prevalence below the WHO threshold of 1% [2] ) , the total overall costs involved—due to both decreasing net cost of production and the need for shorter durations of control—for using the salt approach are also significantly lower than those which will be incurred in running the MDA program . Indeed , this greater social impact of using the present social salt enterprise compared to annual MDA was found to be a general outcome , irrespective of the other intervention coverages investigated ( i . e . , at 65% coverage–the often normal coverage obtained by MDA programs—or at the recommended optimal coverage of 80% [2] ) ( Table 6 ) . These results add to our recent modeling work , which highlighted how the continuous consumption of the drug , even at low daily per capita dosages , by resulting in a cumulative impact on the survival of worms and mf which is significantly higher than that afforded by the higher-dosed annual MDA treatment , make DEC medicated interventions , even when delivered at moderate population coverages , a markedly potent strategy for interrupting LF transmission [18] . Finally , an intriguing possibility highlighted by the break-even analysis and the cost forecasting results shown in Fig 1 and Table 5 is that using a social enterprise strategy for delivering DEC through marketing of medicated salt could in principle also lead to zero disease elimination cost for a provider ( viz . donor or health system ) once the social business attains profitability ( i . e . , return a positive cash-flow ) . This is an important result , and demonstrates how using a social enterprise that pursues a social goal by production of services and goods whilst respecting economic efficiency may offer an effective , financially sustainable , intervention strategy in settings facing major fiscal , infrastructural and logistical barriers to carrying out tablet-based programs aiming to control or eliminate parasitic infections . The present performance evaluation primarily focused on internal ( labor , capital , income and taxes ) and external ( goods and services bought outside the company ) expenses/resources related to the economic viability of the salt enterprise [21] . However , estimation of the full social value of a sustainable health social enterprise must also consider , apart from the social benefits accruing from reducing disease only , the wider consequences for a community [21] . Benefits here could be via the choice and use of resources that further address the community interest , such as choosing local salt suppliers to favor short supply chains , choosing socially certified suppliers , adopting a regime of decent work conditions and even giving employment to workers coming from disadvantaged backgrounds [21] . Such analysis must also include calculation of the larger social benefit associated with the potential for the double-fortified salt to additionally and simultaneously reduce the impacts of iodine deficiency in the population [5] . Note , additionally , that the present salt enterprise represents the first attempt to build industrial-scale capacity on the island for processing large volumes of salt to meet various local needs , which apart from providing a market for local raw salt producers can also act as means to significantly stabilize the price of salt sold in the local markets . These benefits , however , must be contrasted against potential adverse effects , such as domination of the market by the growing enterprise , requiring an analysis of how best to compensate for such loses . Recent developments in applying Social Returns on Investment ( SROI ) approaches for comparing the full monetized social costs of a program with the full monetized social benefits of achieving a health outcome ( or set of outcomes ) may offer a means for undertaking this fuller analysis [22 , 51] . We have also used rough first calculations of the population coverages that could be obtained with the expansion of salt production in the present cost-effectiveness modelling study . Field studies to assess the actual household coverage achieved through the enterprise will be critical for not only more realistically quantifying its effectiveness for accomplishing LF transmission interruption in a community , but also for identifying better marketing strategies to achieve good population coverage . In conclusion , we have presented an economic and financial analysis of the Haitian salt social enterprise , which indicates that it may present a sustainable and socially-responsible strategy for aiding the elimination of LF via the marketing of DEC-medicated salt in settings facing fiscal , infrastructural and logistical challenges for delivering tablet-based elimination programs . Results from the break-even projections carried out in this study indicate that the strategy may even have the potential to achieve zero financial costs to a provider once it attains profitability ( i . e . , results in a positive cash-flow ) . This study further has shown that the Haitian salt enterprise may have already reached production and sales levels that could result in the coverage of the Léogâne study population at proportions sufficient enough to break LF transmission . Finally , our simulation-based cost-effectiveness study has indicated that because of: 1 ) increasing revenue from the sale of the DEC salt obtained over time , 2 ) expansion of its delivery in the population , and 3 ) the effect of continuous consumption of the drug , even at low daily per capita dosages , leading to a cumulative impact on the survival of worms and mf higher than that afforded by the higher-dosed annual MDA treatment [18] , the delivery of DEC through the present Haiti salt enterprise may represent a dramatically more cost-effective option than annual DEC tablet-based MDA for accomplishing LF elimination . While these are encouraging first results and highlight both the economic viability and social effectiveness of using a salt enterprise in the fight against LF , it is clear that efforts to more fully quantify the social value and strategies for developing similar salt social enterprises elsewhere in other endemic settings with different market structures than those of Haiti are now required if the comparative or joint utility of the approach among the current arsenal of LF intervention strategies is to be fully appraised and understood . We note that the means by which the global iodization of edible salt has been accomplished successfully over the past two decades may offer a particularly apt model for building and sustaining the present intervention globally , and suggest that similar tactics used in that program based on introducing DEC medication into prevailing salt production and distribution systems , collaboration with the national and regional salt industries , and engagement with the government sector , civic society and the general public [52] , could also make the universal deployment of DEC-medicated salt eminently possible . With less than three years remaining for meeting the initial 2020 target set by WHO for accomplishing the global elimination of LF , the present results indicate that these appraisals and development of policies and strategies for delivery of DEC-salt , either via deployment of similarly-fashioned salt enterprises , such as the present , or through mobilization of existing salt industries , perhaps along with health system-led MDA and vector-control programs , in socially-challenging environments , like Haiti , would improve our current efforts for meeting this laudable but exacting goal successfully .
|
With less than three years remaining for meeting the initial 2020 target set by WHO for accomplishing the global elimination of Lymphatic Filariasis ( LF ) , concerns are emerging regarding the feasibility of meeting this goal using the current tablet-based Mass Drug Administration strategy . Salt fortified with the antifilarial drug , diethylcarbamazine ( DEC ) , could offer an intervention that avoids many of the barriers connected with tablet-based elimination programs . We analyzed the economic performance and cost-effectiveness of a novel DEC-salt social enterprise developed and tested in Léogâne arrondissement , Haiti , as a particularly significant strategy for accomplishing sustainable LF elimination in such complex settings . We show that because of increasing revenue from the sale of the DEC salt over time , expansion of its delivery in the population , and the adverse effect of continuous consumption of the drug on worms , the delivery of DEC through a salt enterprise can represent a significantly more cost-effective option than annual DEC tablet-based MDA for accomplishing LF elimination in settings , like Léogâne . We indicate that development of policy and research into how to deploy similarly-fashioned interventions , or work with the salt industry to increase population use of medicated salt , would improve present efforts to successfully accomplish the elimination of LF .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"cost-effectiveness",
"analysis",
"chemical",
"compounds",
"economic",
"analysis",
"sodium",
"chloride",
"sociology",
"geographical",
"locations",
"salts",
"social",
"sciences",
"tropical",
"diseases",
"parasitic",
"diseases",
"north",
"america",
"filariasis",
"pharmaceutics",
"neglected",
"tropical",
"diseases",
"caribbean",
"lymphatic",
"filariasis",
"communications",
"chemistry",
"marketing",
"haiti",
"economics",
"drug",
"delivery",
"people",
"and",
"places",
"finance",
"helminth",
"infections",
"physical",
"sciences"
] |
2019
|
Economic performance and cost-effectiveness of using a DEC-salt social enterprise for eliminating the major neglected tropical disease, lymphatic filariasis
|
Quantitative linkages between individual organism movements and the resulting population distributions are fundamental to understanding a wide range of ecological processes , including rates of reproduction , consumption , and mortality , as well as the spread of diseases and invasions . Typically , quantitative data are collected on either movement behaviors or population distributions , rarely both . This study combines empirical observations and model simulations to gain a mechanistic understanding and predictive ability of the linkages between both individual movement behaviors and population distributions of a single-celled planktonic herbivore . In the laboratory , microscopic 3D movements and macroscopic population distributions were simultaneously quantified in a 1L tank , using automated video- and image-analysis routines . The vertical velocity component of cell movements was extracted from the empirical data and used to motivate a series of correlated random walk models that predicted population distributions . Validation of the model predictions with empirical data was essential to distinguish amongst a number of theoretically plausible model formulations . All model predictions captured the essence of the population redistribution ( mean upward drift ) but only models assuming long correlation times ( minute ) , captured the variance in population distribution . Models assuming correlation times of 8 minutes predicted the least deviation from the empirical observations . Autocorrelation analysis of the empirical data failed to identify a de-correlation time in the up to 30-second-long swimming trajectories . These minute-scale estimates are considerably greater than previous estimates of second-scale correlation times . Considerable cell-to-cell variation and behavioral heterogeneity were critical to these results . Strongly correlated random walkers were predicted to have significantly greater dispersal distances and more rapid encounters with remote targets ( e . g . resource patches , predators ) than weakly correlated random walkers . The tendency to disperse rapidly in the absence of aggregative stimuli has important ramifications for the ecology and biogeography of planktonic organisms that perform this kind of random walk .
Movement is fundamental to many ecological processes and often dictates relevant biotic and abiotic encounter rates , particularly for planktonic organisms inhabiting a highly dynamic and heterogeneous habitat . On the individual level , movement impacts encounter rates with favorable ( e . g . mates , resources ) and unfavorable ( e . g . disease , consumers ) targets . On the population level , these microscopic encounters directly affect growth and mortality rates , dispersal rates , population distributions , the spread of disease and invasion , home ranges , reproduction and survival ( e . g . [1] ) . Particularly for micoroganisms , recent methodological advances have enabled the high resolution quantification of organism movements ( e . g . [2] ) , their statistical features ( e . g . [3] ) and changes therein in response to external stimuli ( e . g . [4] ) . Significant efforts have sought to establish mechanistic linkages between these individual movement behaviors and the resulting population distributions ( reviewed in [5] . ) Deciphering these linkages for planktonic organisms , but also others , provides powerful tools to predict rates of organism encounters with environmentally relevant factors and ultimately , their ecological function . Efforts to bridge the gap between individual movement behaviors and large scale population dispersal have been intense , especially in spatial ecology . Random walk theory has been a particularly powerful approach . Founded on observations of the irregular motions of pollen , i . e . Brownian motion [6] , random walk theory relates organism movements in terms of speed , direction or turning rate to probabilities of particle distribution [7]–[9] . Correlated random walk models that assume correlation in successive movement direction , turning angle or velocity , have been particularly successful in linking movements and dispersion in diverse organisms [5] , [10]–[12] . Every formulation of a random walk model rests on a set of assumptions about the underlying movement parameters , their changes over time and dependence on internal or external stimuli [13] . Predicted rates of population distributions are extremely sensitive to the underlying assumptions and to the exact model formulation [14]–[16] . As was recently shown for movement data of single-celled algae , widely used models with differing assumptions may yield significantly different predictions of organism distributions [17] . Thus , it is impossible to determine the most appropriate set of assumptions based on theoretical considerations alone . Concurrent empirical data of both organism movements and their resulting population distributions and the stimuli that modulate these distributions are necessary to inform predictive model formulations . In a recent advancement , [18] have developed empirical methods that allow the simultaneous quantification of individual movement behaviors and population distributions of free swimming , planktonic organisms in stable and spatially structured environments . The approach taken here was to use these methods and empirically motivate a series of individual based , hidden Markov models to predict population distributions and examine the goodness of fit between model predicted and empirically measured distributions . The goals of this study were to ( 1 ) examine the feasibility of reproducing empirically observed population distributions from individual movement behaviors and ( 2 ) to identify the key characteristics necessary to adequately link individual movements with population distributions . Advancement on these goals is necessary to developing analytical solutions to random walks and predicting individual encounter rates , population distributions and ultimately the role of movement in driving organism abundance and distribution patterns . The results of both empirical and numerical analyses strongly suggest that motility patterns of some planktonic protists must have correlation times on the order of minutes .
Organism swimming behaviors and vertical distributions were measured in 3D using vertically moveable , stereo video cameras that recorded in randomized order at 6 vertically separate horizons . Each video segment yielded both individual movement behaviors and abundance of organisms . The footage was processed through a series of automated video-analysis steps that yielded organism positions , which were then used to reconstruct and analyze 3D movement behaviors . The empirical movement data consist of a total 1032 movement trajectories of Oxyrrhis marina swimming freely within 1L , 30 cm high column of filtered seawater , several cm distance from the nearest wall . The minimum trajectory length was 3 seconds , with 124 trajectories exceeding 10 seconds in duration . In total , these observations represent 108 minutes of movement data , with the median trajectories length 5 . 2 seconds and the longest observation 33 seconds . The mean swimming speed was 235 ( 103 ) and the mean swimming direction was 57 ( 34 ) off the vertical axis . The frequency distribution of the over 24000 empirically determined vertical velocities shows that their distribution is non-gaussian , with a significant negative skewness ( Fig . 1 ) . Thus , the population was characterized by few strong down-swimmers and many , relatively slower up-swimmers . The median vertical velocity was 118 with a considerable standard deviation of 110 . There was some indication that the population either underwent behavioral shifts during the time of observation , or that there were multiple behavioral types represented within this clonal lineage of O . marina . Vertical velocity significantly increased over the period of observation ( p = 0 . 01 ) , whereas there were no significant differences among vertical velocities measured at the six depths in the water column ( p = 0 . 13 ) . The frequency distribution of vertical velocities remained positively biased , irrespective of the time elapsed since introduction . Consistent upward swimming bias indicates that this bias was inherent to the organisms and not a function of the point of introduction at the base of the water column . Simultaneously to measuring individual movement behaviors , the population distribution of Oxyrrhis marina was quantified throughout the entirety of the tank over 1 . 5 hours ( Fig . 2 ) . The time course of abundance changes are shown in three successive vertical profiles ( i . e . passes ) that each lasted 20–30 minutes . In the laboratory , the population showed a progressive upward drift , slowly increasing the number of cells at higher horizons . Because cells were introduced at the bottom of the tank , abundances at the upper horizons were initially low . Few individuals were seen rising upward rapidly , arriving at the top of the tank within the first 40 minutes ( pass 1 and 2 ) . The majority of individuals remained in the lower half of the tank for the first 40 minutes . After approximately 1 hr , the population appeared uniformly distributed throughout the tank . An individual-based , biased random walk model was formulated to establish linkages between individual movement behaviors on the microscopic level and the macroscopic population distributions and changes therein . To seed this model , individual movement behaviors needed to be characterized both in terms of vertical velocities as well as their correlation times . The movement paths showed highly periodic movements ( Fig . 3 , left panel ) , with correlation coefficients failing to asymptotically approach zero ( right , bottom panel ) and net distance traveled growing rapidly ( left , top panel ) as would be expected for highly correlated movements . Individual movement paths were characterized by high degrees of auto-correlation , in all three dimensions . De-correlation of velocities was not observable over the measured path durations . The auto-correlation coefficient calculated for the entirety of all trajectories failed to identify a de-correlation time in the up to 33 second long observations . However , sample size for trajectories 15 seconds was low , ( 30 trajectories ) . Thus , autocorrelation analysis suggested that correlation times were 30 seconds but did not identify a distinct correlation time scale . Given this uncertainly , a range of 12 correlation times between = 1 to 1800 seconds were chosen for the model analysis . Predictions of population distribution from empirically measured vertical velocities through a series of individual-based simulation models showed that the empirically observed mean upward drift of the population was captured well by all model predictions irrespective of the assumed correlation time ( Fig . 4 ) . Correlation times of 1 second predicted the population to tightly cluster vertically as cells moved upward through the water column ( Fig . 4 , panels 2 & 3 ) . After 30 minutes of simulation , the mean vertical position of this population was 17 . 5 cm , with a standard deviation of 0 . 5 cm . The empirically observed , greater variance of the population dispersal and the delay in upward flux of the majority of the population were not predicted by model iterations assuming short correlation times . Simulations assuming minute-scale correlation times did capture the increased variance in population distribution ( Fig . 4 , panels 6 & 7 ) . Variance in population distribution increased rapidly with increasing correlation time over the first 30 minutes of model simulation ( Fig . 5 ) . Correlation times of 1 second resulted in low and near constant variance in population distributions . Increased correlation times of 100 seconds lead to more rapid dispersal with standard deviations increasing by approximately 1 mm per minute . Assumed correlation times of 100 seconds predicted cells distributed throughout the water column and standard deviations of the population distributions increased at nearly 5 mm per minute . Increasing correlation time lead to emergence of the behavioral heterogeneity observed in the empirical data , signified by greater cell-to-cell variation in movement and resultant position . As is frequently observed ( e . g . [9] ) , the uncorrelated random walk model predicted a gaussian cohort advancing upward at high cell concentration in close proximity . Longer correlation times resulted in rapid increases in population dispersal and more rapid spreading throughout the water column ( Fig . 6 ) . Although the mean net dispersal distance was identical , irrespective of the correlation time , long correlation times resulted in much higher variance in net dispersal distances because some individuals remained near the point of entry for the entirety of the model simulation , whereas few , fast upward swimming cells reached the surface of the tank within a few minutes . Behavioral heterogeneity was also suggested by the variance in the empirically measured vertical velocities ( Fig . 1 ) . Assumption of longer correlation times reproduced the observed cell-to-cell variation in motility , suggesting behavioral heterogeneity is an important contributor to the observed behaviors and population distributions , even though the source population was clonal . Root mean square error ( RMSE ) of model predictions compared to the empirical distribution data decreased significantly with increasing correlation time ( Fig . 7 ) . Model predictions differed most from empirical observations when assumed correlation was weak . Abundance predictions from highly correlated random walk models with 500 seconds differed least from the empirical data . RMSE was highest and statistically significantly different among models assuming = 1 to 300 seconds . RMSE estimates for 500 were lowest and statistically indistinguishable from one another , suggesting a minimum correlation time of 8 minutes . Further refinement or an upper limit of the correlation time was not identifiable based on this comparison of empirical and predicted population distributions . The time and space scales of the model simulations were expanded to a 15 m water column and run for 24 hrs to explore the consequences of correlation duration on individual dispersal distances as well as population distributions . Total population size , evaluation frequency and all other aspects of the simulation were identical to those used in the simulations evaluated above and stated in the methods . Within patch retention mechanisms have been clearly demonstrated for this species [18] but were not implemented in the simulation . First , expansion of the time and space scales of the model dimensions illustrated how longer correlation times increased population dispersal and thus variance in distribution . Based on the empirically measured , vertical velocity distribution , organisms moving with = 1 second occupied a vertical range of 10 cm after 12 hours . For organisms with correlation times of = 300 and 900 seconds , the predicted vertical ranges were 2 . 5 and 4 m respectively . Thus , correlation times increased population dispersal rates by at least an order of magnitude . Second , individual dispersal distance of the farthest traveling 25th percentile increased rapidly as correlation time increased . An individual with a correlation time of 900 seconds would travel on average twice as far and up to 3 times farther than an individual with a weakly correlated random walk . Therefore , individuals with highly correlated random walk behaviors are expected to encounter remote targets more rapidly than weakly correlated random walkers . Simulation of a 1m thick phytoplankton prey layer within the 15 m water column provides quantitative estimates of the impact of correlation time on the encounter of remote targets . Dimensions of the phytoplankton layer were based on empirical measurements in a shallow , coastal fjord [19] . Higher correlation times lead to considerably earlier arrival of 25% of the population within the prey layer , over 1 hour earlier in the case of = 1800 seconds ( Fig . 8 , top panel ) . Populations with strongly correlated random walks remained within this prey layer over 2 hours longer than populations with weakly correlated random walks ( Fig . 8 , bottom panel ) .
Long correlation time and cell-to-cell variation were identified as key characteristics necessary to reproduce empirically observed population distributions from individual movement behaviors . Simultaneous measurements of both individual movement behaviors and population distributions were essential in linking microscopic movement behaviors with macroscopic population distributions . The results strongly suggest that motility patterns of some planktonic protists must have correlation times on the order of minutes , rather than seconds as is currently thought . Persistent similarity of movement in individual cells resulted in vastly higher dispersal rates for the population and significantly increased predicted rates of encounter with remote targets . The correlation times estimated here far exceed previously measured correlation times . Previous studies suggest that transitions from highly correlated movements to more diffusive motion were observed to occur within 10 seconds for taxonomically diverse planktonic organisms ranging from bacteria to copepods [20] , [21] . Uncorrelated movements were not identifiable in a set of several hundred movement tracks . To identify the correlation time-scales empirically would require minute-scale observations of 100s of individuals . The longer correlation times observed in this study may be due to the much larger than typical observation volume used , which may have resulted in longer free path lengths . The consequences of long correlation times in individual motility patterns of plankton are significant . Planktonic organisms live in a nutritionally dilute environment ( e . g . [ ? ] ) . Recent , high-resolution observations in the ocean have shown that phytoplankton , the principal prey of many heterotrophic protists , are frequently concentrated in discrete layers or patches , rather than uniformly distributed [22] . Early hypotheses identified that planktonic predators must exploit these patches to sustain measured levels of secondary productivity [23] . Asexually reproducing organisms in particular can quickly transfer increased resource availability to increased growth and abundance . Long correlation times of individual movements result in significant increases in predicted dispersal distances of individuals and thus increases in the probable encounter with remote targets , including prey patches . Conversely , the probability of less advantageous encounters , including with predators is also increased [20] unless dispersive escape responses are evoked . For clonal organisms , increased probability of encountering unexploited resource patches may offset the increased risk of mortality due to predator encounters . The exact rate of encounter of remote targets will depend on the distribution , size and persistence of targets . Irrespective of target characteristics , individuals with long correlation time will encounter specific targets faster , given their , on average , almost two and up to three-fold greater dispersal distance . Behavioral modifications in response to prey derived stimuli that lead to consumer aggregations within resource patches are well documented ( e . g . [18] , [24] , [25] ) and are expected to provide further advantages to consumers exploiting dilute environments . At the population level increased rates of population dispersal would erode aggregations and patchiness . It is noteworthy that the population also contained a small fraction of strong downward swimmers , which would further increase population dispersal rates . In the absence of aggregative stimuli the behavioral heterogeneity observed here may serve an important dispersive function and provide adaptive advantages to counteract long term aggregations . Long correlation times may have a homogenizing effect in light of many physical and biological processes that lead to cell aggregations and patchiness . This dispersive behavior could lead to reduced competition among cells [26] , reduced risk from predators attracted to high cell concentrations [27] and reduced risk of the entire population being subjected to a localized risk or condition . Accelerated population dispersion may also counteract the tendency of cluster formation due to rapid asexual reproduction in planktonic organisms [28] . It is unknown how constant the measured rates of dispersal are over time . The observations made here were made shortly after organisms were introduced into the tank , thus population distributions were transient and dispersal rates likely at their maximum . The experimental set up deliberately did not include any stimulus that would either limit ( e . g . aggregation ) or enhance dispersal , so that measured dispersal rates were independent of external stimuli . However , organisms likely modulate their dispersal rates both over time and in response to external cues . Such modulation of correlation time has been suggested as an effective prey search strategy for organisms lacking sensory capacity [15] . Similarly , [29] have identified high variance in the turn rate of freshwater zooplankton ( Daphnia spp . ) and proposed that variation in movement behaviors has adaptive advantages . The observed upward bias was a consistent characteristic of the measured swimming behaviors irrespective of the point of introduction or time of sampling . The same vertical bias was previously observed for the same species and the presence of a prey stimulus significantly reduced but did not eliminate upward bias [18] . In the absence of aggregative stimuli , this upward bias would ultimately lead to surface aggregations of organisms . Although surface aggregations were indeed observed in the laboratory , the stable , convection-suppressing conditions of this laboratory set up are neither realistic nor characteristic of planktonic habitats . The dynamic hydrography , including breaking internal waves , shear instability at boundaries and turbulent mixing , characteristic of the coastal ocean may counteract the observed net upward flux of organisms and prevent aggregations at the surface . Reported eddy diffusivities are an order of magnitude higher than the upward swimming velocities measured here and would counteract surface aggregations [30] . Given these dispersive factors , an inherent up swimming bias may hold adaptive advantages for planktonic organisms in the ocean , which is characterized by weak horizontal but strong and predictable vertical gradients in resource availability . The data presented here strongly suggest that correlation times of motility patterns for some planktonic organisms are significantly longer than currently assumed . Long correlation times suggest that organisms with these motility patterns have higher dispersal rates and higher encounter rates with remote targets than organisms with only weakly correlated random walks . Simultaneous empirical observations of individual movement behaviors and the resulting population distributions were essential in linking statistical properties of cell movements to predictions of population distribution , a connection across disparate time and space scales . Model simulations of organism movements and population distributions were necessary to extrapolate beyond empirically measurable time and space scales . Verification of model predictions against empirical observations helped distinguish among a number of reasonable model formulations and ultimately in estimating the minimum correlation time . Quantifying the magnitude of the correlation time provides a basis for estimating individual encounter rates as well as population distributions . These quantitative tools are indispensable to predicting organism distributions and their function in the environment .
The heterotrophic dinoflagellate Oxyrrhis marina was used to study the effects of swimming behaviors on population dispersal . O . marina is 12–18 m in length and is a globally distributed species [31] . Cells swim and steer with the aid of perpendicular transverse and longitudinal flagellae that each propagate helicoidal waves [32] . O . marina was fed the haptophyte prey alga Isochrysis galbana , grown in nutrient-amended filtered seawater , f/2 [33] . Cultures were maintained on a 16∶8 hr light∶dark cycle , at 18 and 50 mol photons m provided by cool and warm white lights . The cultures were not axenic . The salinity of the medium was 30 . Both predator and prey cultures showed positive growth in all tested media ranging in salinity from 24 to 32 . Cultures were transferred every 4–6 days to maintain exponential growth . Cell concentrations of both predator and prey cultures were determined with a Coulter Multisizer ( Beckman Coulter , Miami , Florida ) just prior to experiments . Predators were starved for 48 hrs prior to the experiment to minimize variation between cells . Organism swimming behaviors and vertical distributions were measured in complete darkness in a 1L , octagonal plexiglas tank of 30 cm height at ambient room temperature of approximately . All organisms were introduced at the bottom of the tank and observations were made without external stimuli . To suppress water movement , the water column was stabilized through a weak , linear salinity gradient , ranging from 28 to 30 . Video images were captured with two infra-red sensitive cameras ( Cohu 4815-3000/000 ) , equipped with Nikon 60 mm Micro Nikkor lenses and illuminated by infra-red light emitting diodes ( Ramsey Electronics , 960 nm ) . The cameras were mounted on a vertically movable stage . Vertical position was controlled through a ruler fixed to the side of the stage . Video was recorded at 15 frames per second . Prior to these experiments , it was verified that some cells reached the top of the experimental tank within 15 minutes and filming was commenced after a waiting period of 15 minutes . Footage was collected in the center of the water column at six equally spaced horizons , approximately 5 cm apart , for 2 minutes every 20 to 30 minutes for a total duration of 1 . 5 hours . This resulted in 3 video segments being collected at each horizon . At the beginning of the experiment the order of sampling horizons was randomized . The position of organisms in the video footage was determined with ImageJ image processing software by removing stationary background objects and thresholding . A three-dimensional calibration grid was used to convert video pixel dimensions to physical units . The stereoscopic field of view was approximately 1 . 8 cm wide , 1 . 3 cm high and 4 . 0 cm deep . Thus , cells within a volume of approximately 9 ml were observed . These movement data were unencumbered by frequently encountered methodological limitations such as low temporal resolution , physical restriction ( e . g . container size ) and the 3D rendition avoids underestimates of swimming velocities and directions . Three-dimensional swimming paths were generated from pixel positions by Tracker3D , a Matlab-based motion analysis package for tracking organism movement written by Danny Grünbaum ( Univ . of Washington ) . Before analysis , swimming paths were smoothed with a cubic spline to remove high-frequency noise . Individual movement statistics were calculated from 3D swimming paths , subsampled at 0 . 25 second intervals , including only trajectories of at least 3 seconds duration . Abundance of O . marina was estimated from the average number of 3D trajectories observed in each video frame . Further details on the water column set-up , filming and data collection are reported in [18] . An individual-based , hidden Markov model was formulated to predict the vertical redistribution of the Oxyrrhis marina population in the water column . The successive positions and movement parameters for each individual were modeled explicitly based on the empirically observed behaviors . The magnitude and frequency distribution of empirically measured vertical velocities provided the basis for modeled velocities ( Fig . 1 ) . Negative velocities indicate downward movement . The -th model organism at time was characterized by a position , vertical velocity and associated with a specific swimming trajectory , randomly drawn from the entirety of observed paths and then assigned the first velocity measured within that path . Triplicate model iterations were evaluated at time increments of 4 Hz with 1000 individuals each . Successive organism positions were calculated as:Model organisms encountering the upper or lower boundaries were assigned movement paths with net downward or upward movements respectively . The model was chosen to be 1-dimensional , since there were no horizontal gradients in external stimuli and the variable of interest was the rate of vertical population redistribution in the water column . The spatial and temporal scales of the model were identical to the laboratory set up . In all model formulations , individuals were randomly assigned new velocities at the model iteration frequency of 4 Hz . In the uncorrelated random walk model , the assigned vertical velocity was drawn from the entirety of all observed vertical velocities . Thus , information on the associated trajectory was meaningless for the uncorrelated random walk model . In the correlated random walk model , subsequent velocities were sampled from the associated swimming trajectory in sequence of observation . New trajectories were assigned at the frequency with probabilityThus , at seconds ( i . e . the iteration frequency of 4 Hz ) , individuals sampled repeatedly and in sequence from the velocities within one empirically determined trajectory . Modeled correlation time ranged from 0 to 1800 seconds . Population size was held constant , since demographic processes were not expected to change abundance over the model duration of 1 . 5 hrs . Comparisons of the vertical velocities over time and at different filming horizons were made using a two-way ANOVA . Sensitivity analyses were conducted to ensure that the path discretization parameters did not significantly change the calculated vertical velocities . Furthermore , artificial data sets were created to test the sensitivity of model predictions to deviations from normality for the frequency distribution of vertical velocities and total sample size . Neither analysis suggested a change in conclusions . The autocorrelation coefficients of vertical velocities were calculated for each path separately , with mean velocity subtracted . To facilitate comparison among paths , correlation coefficients were normalized . The root mean square error ( RMSE ) between empirically observed and model predicted vertical population distributions were calculated to facilitate among model comparisons . To do so , vertical population distribution from model predictions were sampled in the same order and frequency as the empirical data were collected . All RMSE estimates were scaled to the maximum RMSE observed to remove the effect of sample size from estimates . Comparison of RMSE were made with a one-way ANOVA . Significant differences among means were assessed using a Bonferroni corrected post-hoc test . Statistical significance was assigned at . All analyses and simulations were done using the software package Matlab 7 . 9 . 0 . .
|
Organism movement is fundamental to how organisms interact with each other and the environment . Such movements are also important on the population level and determine the spread of disease and invasion , reproduction , consumption , and mortality . Theoretical ecologists have sought to predict population dispersal rates , which are often hard to measure , from individual movement behaviors , which are often easier to measure . This problem has been non-trivial . This manuscript contributes seldom available , simultaneously measured movement behaviors and population distributions of a single celled planktonic organism . The empirical data are used to distinguish amongst a set of plausible theoretical modeling approaches to suggest that organism movements are highly correlated , meaning movement direction and speed is consistent over several minutes . Previous estimates suggested persistence only lasted several seconds . Minute-scale correlations result in much more rapid organism dispersal and greater dispersal distance , indicating that organisms encounter and impact a greater portion of their surrounding habitat than previously suspected .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"ecology/behavioral",
"ecology",
"biophysics/theory",
"and",
"simulation",
"marine",
"and",
"aquatic",
"sciences",
"ecology/environmental",
"microbiology",
"ecology/theoretical",
"ecology"
] |
2010
|
Inherent High Correlation of Individual Motility Enhances Population Dispersal in a Heterotrophic, Planktonic Protist
|
The final stage of bacterial cell division requires the activity of one or more enzymes capable of degrading the layers of peptidoglycan connecting two recently developed daughter cells . Although this is a key step in cell division and is required by all peptidoglycan-containing bacteria , little is known about how these potentially lethal enzymes are regulated . It is likely that regulation is mediated , at least partly , through protein–protein interactions . Two lytic transglycosylases of mycobacteria , known as resuscitation-promoting factor B and E ( RpfB and RpfE ) , have previously been shown to interact with the peptidoglycan-hydrolyzing endopeptidase , Rpf-interacting protein A ( RipA ) . These proteins may form a complex at the septum of dividing bacteria . To investigate the function of this potential complex , we generated depletion strains in M . smegmatis . Here we show that , while depletion of rpfB has no effect on viability or morphology , ripA depletion results in a marked decrease in growth and formation of long , branched chains . These growth and morphological defects could be functionally complemented by the M . tuberculosis ripA orthologue ( rv1477 ) , but not by another ripA-like orthologue ( rv1478 ) . Depletion of ripA also resulted in increased susceptibility to the cell wall–targeting β-lactams . Furthermore , we demonstrate that RipA has hydrolytic activity towards several cell wall substrates and synergizes with RpfB . These data reveal the unusual essentiality of a peptidoglycan hydrolase and suggest a novel protein–protein interaction as one way of regulating its activity .
Though not formally considered virulence factors , genes required for bacterial cell division clearly are necessary for the growth , and thus , pathogenesis , of bacteria . The distinction between homeostatic and virulence genes is blurred when nonessential genes involved in vegetative cell division become essential under specific stressful conditions encountered inside a host . Such an example is seen with the resuscitation-promoting factors ( Rpf ) encoded by many different bacteria , including mycobacteria . These proteins are named for their ability to resuscitate nonreplicating dormant bacteria . The single Rpf-encoding gene in Micrococcus luteus [1] is essential , but as many as three of the five genes encoding RpfA-E can be deleted in Mycobacterium tuberculosis without markedly affecting in vitro growth . However , one single deletion ( rpfB ) and several of the triple combinations yielded strains unable to grow or divide in stressful conditions in vitro and in vivo [2] , [3] . This suggests that certain potential cell division proteins that appear to play nonessential roles in homeostatic processes can become vital in conditions of stress . The vital processes of cell growth and division involve the temporal and spatial coordination of events such as peptidoglycan and cell wall extension , DNA replication , chromosomal partitioning , Z-ring assembly , septum formation , and cytokinesis . Much of the mechanism behind this coordination involves inhibiting and stabilizing proteins that regulate the eventual assembly of a Z-ring at the midcell of bacteria [4] , [5] . This Z-ring consists primarily of a polymerized ring of tubulin-like FtsZ on the cytoplasmic side of the plasma membrane , stabilized by membrane-associated and integral membrane proteins . Assembly occurs in an ordered fashion that is not entirely linear , with some components assembling before joining the Z-ring [6] , [7] , [8] . Some of the last proteins to be recruited to the Z-ring are thought to be the peptidoglycan hydrolyzing enzymes , such as AmiC [9] and EnvC [10] in E . coli . These enzymes digest the peptidoglycan layers connecting two recently developed daughter cells in the final stage of cell division [11] . While crucial to cell division , the regulation of these potentially lethal enzymes is poorly understood . It is thought that protein-protein interactions play a role in regulating activity and localization [12] . Studying cell wall hydrolases has proven difficult due to the large number encoded in most bacterial genomes and the high degree of functional redundancy . For example , seven hydrolases can be deleted from a strain of E . coli without loss of viability [13] . Little is known about the hydrolases involved in mycobacterial cell division . CwlM and Rv2719c were both shown to be mycobacterial cell wall hydrolases [14] , [15] . Two recently identified hydrolases , RpfB and RpfE , were shown to interact with rpf-interacting protein ( RipA ) , a peptidoglycan endopeptidase [16] . Rpf proteins constitute a family of lytic transglycosylase enzymes capable of hydrolyzing the glycosidic bonds in the essential stress-bearing , shape-maintaining peptidoglycan layer [17] . RpfB has a structure similar to the E . coli soluble lytic transglycosylase 70 ( Slt70 ) [18] and is known to hydrolyze the beta-1 , 4-glycosidic bond between N-acetyl muramic acid and N- acetyl glucosamine [19] . RipA has been shown to be a peptidoglycan hydrolase [16] . It is predicted to function as an L , D-endopeptidase , capable of hydrolyzing D-glutamyl-meso-diaminopimelic acid [20] , two amino acids that are part of the crosslinking peptides vital for keeping peptidoglycan rigid and stable [21] . Both RpfB and RipA localize to the septa of dividing bacteria [16] and thus may play a role in the late stages of mycobacterial cell division , possibly during regrowth from a stressed state . Here we show that depletion of ripA in a strain of M . smegmatis results in a significant decrease in growth , formation of long , branched chains , and increased sensitivity to a cell wall–targeting antibiotic . These defects can be functionally complemented with the M . tuberculosis allele of ripA . We demonstrate that the peptidoglycan hydrolytic activity of RipA synergizes with RpfB . Thus , this protein is an unusual example of an essential peptidoglycan hydrolase whose activity may be partially regulated through protein-protein interactions .
RipA interacts with RpfB , a lytic transglycosylase , and colocalizes at the septum [16] . We hypothesized that the RipA-RpfB complex may be involved in degrading peptidoglycan at the septum during cell division . To further investigate the function the individual components of this complex , we attempted to make deletion strains of ripA and rpfB in M . smegmatis . Though disruption of rpfB was successful , we were unable to disrupt the ripA gene in M . smegmatis . We have previously reported that the ripA gene in M . tuberculosis appears to be essential for in vitro growth [22] , suggesting that it might also be essential in M . smegmatis . To test this possibility we constructed depletion strains of M . smegmatis in which ripA ( MSMEG3153 ) or rpfB ( MSMEG5439 ) are transcribed from an inducible tetracycline promoter ( Ptet , Figure 1A ) . We found that the ripA depletion strain had dramatically reduced growth in media lacking inducer ( Figure 1B ) , while the rpfB depletion strain had normal growth in the absence of inducer . Both ripA and rpfB depletion strains grew normally in the presence of inducer . The growth phenotype seen with ripA depletion , as measured by optical density , was dose-dependent . While the optical density of ripA depleted cultures failed to increase in the absence of inducer , we did observe that bacteria formed visible clumps that increased in size during incubation . These clumps ( due to filamentation ) failed to suspend well and , therefore , were poorly measured using spectrophotometry . Remarkably , this phenotype is reversible , as addition of inducer to growth-arrested , ripA depleted cells resulted in the resumption of normal growth ( Figure 1C ) . This indicates that the frequency of septum resolution can be uncoupled from septum formation and cell elongation . High levels of inducer did not result in gross morphological changes or lysis . To further confirm the requirement of ripA for growth , we plated ripA-depletion or rpfB-depletion strains of M . smegmatis on plates containing a gradient of inducer . The ripA-depletion strain grew in a Tet-dependent manner , with the highest growth in the center corresponding to the highest concentration of inducer and no growth near the edges of the plate where inducer was the lowest , confirming the requirement of ripA for growth . Conversely , the rpfB-depletion strain grew similarly throughout the plate in a Tet-independent manner , indicating that rpfB is not required for growth under these conditions . While rpfB depleted cells had normal morphology , ripA depleted cells had markedly abnormal shape ( Figures 1C and 2A ) . They grew in long , branched chains that account for the clumps seen grossly . Staining with the fluorescent membrane dye , TMA-DPH , revealed periodic septa along the chains of bacteria . DNA staining with SYTO 9 revealed nucleoids along the length of the chained bacteria , separated by septa , indicating that DNA segregation and septum formation processes are intact ( Figure 2B ) . Occasional patches where the cell wall appeared pinched or partially degraded were also observed . It is not clear if these represent defective division sites or locations where bacteria had begun to lyse . Branches , not observed in wild-type cells , were seen in almost all ripA-depleted bacteria visualized ( >95% ) . Interestingly , 91% ( 203/223 ) of the branches visualized originated directly adjacent to septa . The M . tuberculosis ripA gene encodes a 472 amino acid protein that has been shown to degrade peptidoglycan [23] . Its C-terminal 105 amino acids contain a putative endopeptidase domain , which has 40% identity with the Listeria monocytogenes p60 protein ( Figure 3A ) . p60 has been shown to be a cell wall endopeptidase by its ability to degrade cell wall [24] , [25] . ripA is the first gene in a bicistronic operon . rv1478 , the downstream gene , encodes a 241-amino acid protein , consisting of a signal sequence followed by a sequence homologous to the C-terminal half of RipA ( 70% putative hydrolase domain identity and 27% overall amino acid identity ) . Both genes in the apparent ripA operon encode predicted endopeptidase domains similar to a known p60 hydrolase [24] ( Figure 3A ) . Because the inserted tetracycline-inducible promoter lies upstream of the operon , presumably transcription of both is dependent on the presence of inducer . Thus , either gene could be responsible for the observed phenotype . There are at least five p60-like genes in many of the mycobacterial species , including M . tuberculosis , M . smegmatis , and M . bovis BCG . These include rv0024 , rv1477 ( ripA ) , rv1478 , rv1566c , and rv2190c . These genes lie in various genomic contexts and there is no function discernable from genomic synteny ( Figure 3B ) . To identify the responsible gene we tested if the M . tuberculosis allele of ripA ( ripA-mtb ) was able to complement the M . smegmatis with diminished native ripA ( ripA-smeg ) production . We expressed ripA-mtb from a zeocin-marked episomal plasmid . A ripA-smeg depletion strain containing the ripA-mtb construct grew similarly to wildtype in the absence of inducer , while a strain carrying an empty plasmid formed chains when ripA-smeg was depleted ( Figure 3C ) . In contrast , the M . tuberculosis allele of the ripA paralogue , rv1478 , was not able to complement an M . smegmatis ripA depletion . These results confirm that ripA is sufficient for complementing the strain depleted of the ripA ( ripA-MSMEG3154 ) operon and show that ripA-mtb is functionally similar to ripA-smeg . Because depletion of ripA may have a marked effect on cell wall structure , we reasoned that strains with diminished expression of ripA might have altered susceptibility to antibiotics that target the cell wall . To test this , we grew the M . Smegmatis regulated ripA strain in the presence of inducer , then washed and spread on plates containing various amounts of inducer ( 0 , 10 , 100 ng/ml Tet ) . A sterilized disc of Whatman paper was placed in the middle of the plate and 10 ul of a single antibiotic was added to the disc . After incubating for 3–4 days , the size of the zone of inhibition was measured ( Figure 4 ) . Complete depletion of ripA resulted in a remarkably high level of susceptibility to β-lactams ( carbenicillin ) . It is unclear why depletion of ripA results in such an increase in susceptibility to β-lactams , which target the transpeptidase reaction required for cross-linking peptidoglycan during cell elongation and division . Susceptibility to cycloserine , an analog of D-alanine that inhibits the formation of the cytoplamsic pentapeptide that is eventually transported across the cell membrane and used to cross-link PG strands , was independent of induction of ripA . While increased permeability is often attributed to an observed increase in susceptibility to an antibiotic , depletion of ripA did not affect susceptibility to cycloserine , suggesting a more specific defect . RipA has been predicted and shown to degrade M . luteus cell wall material [25] , [26] . To test if RipA hydrolyzes peptidoglycan and cell wall material from other species of bacteria , we determined the enzymatic activity of RipA using a variety of FITC-labeled , cell wall–derived substrates . We expressed RipA as a fusion protein with GST in E . coli and purified the fusion protein using affinity chromatography . We found that GST-RipA , but not GST alone , was able to hydrolyze cell wall derived from M . smegmatis as well as peptidoglycan purified from Streptomyces , and had minimal activity against M . luteus cell wall ( Figure 5B–D ) . Therefore , RipA is capable of hydrolyzing cell wall material from several bacterial species . The predicted activity of RipA is to cleave the peptide cross-linkages in peptidoglycan and is distinct from Rpf , which is predicted to cleave glycosidic bonds in peptidoglycan ( Figure 5A ) . Given the close proximity of these predicted cleavage sites , we hypothesized that the interaction of RpfB with RipA may result in enhanced hydrolytic activity . To test this possibility we expressed a portion of RpfB as a fusion protein with GST in E . coli and purified it using affinity chromatography . Using the same assays described above , we found that GST-RpfB alone had minimal ability to degrade cell wall extracts or purified peptidoglycan . However , when GST-RpfB was combined with GST-RipA , activity was more than the sum of individual enzyme activities . The same result was found with all substrates tested ( Figure 5B–D ) . No increase was detected when GST was combined with GST-RpfB or GST-RipA ( data not shown ) . Addition of twice as much Rpf yielded no increase in hydrolysis , while twice as much RipA yielded twice the hydrolysis ( data not shown ) indicating the assay is in the linear range .
In this work , we demonstrate that RipA is essential for normal cell division in M . smegmatis , with its depletion resulting in long , branched filaments and increased susceptibility to a specific cell wall targeting antibiotic . Furthermore , RipA cleaves peptidoglycan and synergizes with RpfB . Taken together , these data support a model where RipA is 1 ) required for the final stage of cell division , where daughter cells are separated and 2 ) has peptidoglycan hydrolytic activity that may be modulated by RpfB under certain conditions . It is unusual that RipA is essential for normal cell division in M . smegmatis and , apparently , M . tuberculosis [22] . Because bacteria encode a number of hydrolytic enzymes that are , at least in part , functionally redundant , strains carrying deletions of single hydrolase genes are generally viable , though combinations of mutations can result in lack of viability [13] . In M . smegmatis and M . tuberculosis , ripA does not appear to be redundant . Conversely , while M . marinum strains carrying mutations in the homologous gene , iipA , do have abnormal morphology , they are still able to divide . Mutations in the hydrolytic domain of IipA abolished complementation of the defect , confirming the importance of the hydrolytic activity of IipA [27] . In M . marinum , different rip paralogues might be able to complement for loss of iipA . None of the rpf genes appears to be essential in M . tuberculosis and combinations of at least three rpf genes can be deleted in M . tuberculosis strains while still maintaining normal in vitro vegetative growth [2] . We demonstrate that RpfB is also not essential in M . smegmatis . It logically follows that the interaction between RpfB and RipA must not be essential for RipA function during vegetative growth . Of course , it is possible that another Rpf protein is able to compensate for the absence of RpfB , resulting in increased RipA-dependent activity . For example , RpfE is able to interact with RipA [16] . It is also possible that the RipA-RpfB interaction , and subsequent enhanced hydrolytic activity , is required only under special circumstances , such as growth under specific conditions of stress . As noted , RpfB is required for resuscitation of M . tuberculosis in a reactivation mouse model [3] . Likewise , deletion of several combinations of three rpf genes results in viable bacteria that are unable to resuscitate from in vitro and in vivo resuscitation assays [2] . Thus , the RipA-RpfB interaction may be necessary under certain conditions . There are several models that might explain the cooperativity seen between RipA and RpfB . One protein might allosterically activate the other , resulting in increased peptidoglycan degradation . Alternatively , both proteins might be fully active , but their association might bring their active sites in close proximity , thus producing cleavage of bonds located near to one another in the peptidoglycan . Since peptidoglycan is a highly cross-linked polymer , nearby cleavages are more likely to effectively degrade peptidoglycan and release fragments . Several of the most effective antibiotics , including many important antimycobacterial agents , target cell wall synthesis . RipA appears to represent a particular vulnerability for M . tuberculosis . In addition to its possible role in reactivation through interaction with Rpf , RipA is essential for normal cell division and is accessible to drugs , given its external localization . Inhibiting the enzymatic activity should block the ability of daughter cells to separate from one another , while blocking protein-protein interactions could result in dysregulation of activity . Thus , RipA is an attractive target for antimycobacterial drug development .
E . coli XL-1 ( Stratagene ) strains were used for cloning and E . coli BL21 ( DE3 ) ( Stratagene ) was used for expression of recombinant proteins from the pET41a ( Novagen ) or pMal ( NEB ) . Mycobacterium smegmatis ( mc2155 ) and Mycobacterium tuberculosis ( H37Rv ) strains were grown at 37°C in Middlebrook 7H9 broth supplemented with ADC and Tween80 and antibiotic when appropriate . The E . coli expression strain , BL21 ( DE3 ) was used to synthesize each protein following the Novagen manual protocol . Protein concentrations were measured using the Bradford assay , normalized , and confirmed by coomassie-stained polyacrylamide gels . Protein samples were combined with 4× Laemmli's SDS PAGE buffer and boiled at 100°C for 10 minutes . Proteins were separated on 10% Tris-tricine polyacrylamide gels by electrophoresis , transferred to nitrocellulose , and probed with specific antibodies using standard techniques . M . smegmatis cell wall was prepared as previously described [14] . Streptomyces peptidoglycan and lyophilized M . luteus cell wall were both obtained from Sigma . The fluorescein isothiocyanate ( FITC ) -labeled bacterial cell wall was prepared by covalently linking FITC to amine groups in the cell wall . 10 mg FITC ( Molecular Probes ) was used to label 10 mg of insoluble peptidoglycan or cell wall material following the protocol from Molecular Probes . Recombinant M . tuberculosis proteins were incubated with several FITC-labeled cell wall substrates and assayed for activity by measuring FITC release . 25 µg of Rpf or RipA alone or 25 µg of Rpf and 25 µg RipA combined , were added to 25 µl of 2 mg/ml substrate and 25 µl 4× reaction buffer ( 50 mM Tris , 10 mM MgCl , 50 mM KCl , 2 mM MnCl , 0 . 01% Chaps , 100 mM KH2PO4 , pH 5 . 75 ) . The final volume was brought to 100 µl with H2O . As a control , 50 µg of lysozyme was added to M . smegmatis cell wall . Similar combinations with GST were also tested . GST alone , as well as buffer alone , were used to determine background release of FITC . After incubating at 30°C with enzyme and buffer for 3–5 days , the insoluble substrate was centrifuged ( 18 , 000×g ) and soluble FITC was measured with filters for excitation 485 nm and emission 538 nm . Depletion strains were generated as previously described [28] . Briefly , M . smegmatis , with the tetracycline repressor gene integrated into the attB site , was transformed with a suicide vector containing the first 600 nucleotides of M . smegmatis ripA gene under control of the tetracycline operator/promoter system ( Ptet ) . Transformants were selected for hygromycin resistance . Appropriate recombination was confirmed using forward primers to Ptet and Prip ( native ripA promoter ) paired with a reverse primer to the 3′ end of ripA . The presence of a product of appropriate size for the former and lacking in the latter , confirmed the desired strain . Attempts to disrupt the ripA gene in M . smegmatis using a nonreplicating suicide vector designed to recombine into the middle of the gene were unsuccessful ( though control knockouts , such as rpfB , were successful ) . The ripA and rpfB depletion strains were initially grown in 7H9 media containing kanamycin ( selecting for TetR ) and hygromycin ( selecting for inserted pTet ) as well as anhydrotetracycline ( Tet ) . Once cultures reached late log-phase or stationary phase , they were centrifuged ( 2500×g for 5 minutes ) , washed once with PBS , and resuspended in media with varying amounts of Tet . To test recovery of ripA depleted cells , Tet was either added directly to cultures grown without Tet or to fresh media inoculated with cells depleted of ripA . To test complementation , the ripA gene and its native promoter from M . tuberculosis was amplified and cloned into an episomal plasmid containing the zeocin gene as a marker . This construct , or the isogenic empty vector , was transformed into the ripA depletion strain of M . smegmatis . Strains were grown in the presence of tetracycline inducer , washed and inoculated into media lacking inducer . Cultures were monitored by OD600 and microscopy . To confirm the essentiality of ripA , depletion strains of M . smegmatis were grown on a plate with a gradient of inducer generated by placing 10 µl of 10 ng/ml Tet on a paper disc in the center of the plate , resulting in a concentration of inducer highest at the middle of the plate and lowest at the edges . The ripA depletion strain of M . smegmatis was spread on LB agar plates containing different amounts of anhydrotetracycline inducer ( ng/ml concentrations ) to regulate the amount of ripA expressed . A filter disc with 10 µl of carbenicillin ( 100 mg/ml ) , isoniazid ( 10 mg/ml ) , or cycloserine ( 100 mg/ml ) was placed in the center of plate and the diameter of inhibition of growth was measured after 4 days of growth . M . smegmatis strains were centrifuged at 2500×g for 2 minutes , washed with 1ml PBS , and resuspended in 20 µl of PBS containing 50 nM TMA-DPH or 5 µM SYTO 9 for staining membranes or DNA , respectively . Samples were imaged with a Nikon TE-200 100× ( NA 1 . 4 ) objective and captured with an Orca-II ER cooled CCD camera ( Hamamatsu ) . Final images were prepared using Adobe Photoshop 7 . 0 .
|
Mycobacteria , like all peptidoglycan-containing bacteria , must extend and cleave the surrounding structurally rigid layer of peptidoglycan to grow and divide . The peptidoglycan hydrolases responsible for this cleavage often have redundant functions , both revealing their importance and making them difficult to study . Furthermore , such hydrolases must be tightly regulated , due to their potentially lytic nature . We recently demonstrated the interaction between a lytic transglycosylase ( Rpf ) and an endopeptidase ( RipA ) at the septum of dividing bacteria . To investigate the role of these two hydrolases , we generated conditional mutants of each and were surprised to find that depletion of ripA resulted in long chains of cells . This phenotype was reversed upon induction of ripA , indicating that cell wall expansion and septum formation can be decoupled from the process of septum resolution . In addition , we present data showing that the combination of Rpf and RipA results in enhanced hydrolysis of peptidoglycan in an in vitro assay , suggesting protein–protein interactions as one potential mechanism of regulation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/bacterial",
"infections",
"microbiology/microbial",
"growth",
"and",
"development",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"microbiology/microbial",
"physiology",
"and",
"metabolism"
] |
2008
|
A Mycobacterial Enzyme Essential for Cell Division Synergizes with Resuscitation-Promoting Factor
|
Lymphatic filariasis ( LF ) is one of the most debilitating neglected tropical diseases ( NTDs ) . It still presents as an important public health problem in many countries in the tropics . In Cameroon , where many NTDs are endemic , only scant data describing the situation regarding LF epidemiology was available . The aim of this study was to describe the current situation regarding LF infection in Cameroon , and to map this infection and accurately delineate areas where mass drug administration ( MDA ) was required . The endemicity status and distribution of LF was assessed in eight of the ten Regions of Cameroon by a rapid-format card test for detection of W . bancrofti antigen ( immunochromatographic test , ICT ) . The baseline data required to monitor the effectiveness of MDA was collected by assessing microfilariaemia in nocturnal calibrated thick blood smears in sentinel sites selected in the health districts where ICT positivity rate was ≥ 1% . Among the 120 health districts visited in the eight Regions during ICT survey , 106 ( 88 . 3% ) were found to be endemic for LF ( i . e . had ICT positivity rate ≥ 1% ) , with infection rate from 1 . 0% ( 95% CI: 0 . 2–5 . 5 ) to 20 . 0% ( 95% CI: 10–30 ) . The overall infection rate during the night blood survey was 0 . 11% ( 95% CI: 0 . 08–0 . 16 ) in 11 health districts out of the 106 surveyed; the arithmetic mean for microfilaria density was 1 . 19 mf/ml ( 95% CI: 0 . 13–2 . 26 ) for the total population examined . ICT card test results showed that LF was endemic in all the Regions and in about 90% of the health districts surveyed . All of these health districts qualified for MDA ( i . e . ICT positivity rate ≥ 1% ) . Microfilariaemia data collected as part of this study provided the national program with baseline data ( sentinel sites ) necessary to measure the impact of MDA on the endemicity level and transmission of LF important for the 2020 deadline for global elimination .
Lymphatic filariasis ( LF ) is a parasitic disease caused by three species of thread-like nematode worms—Wuchereria bancrofti , Brugia malayi and Brugia timori—known as filariae . Of these three parasite species , W . bancrofti accounts for nearly 90 percent of LF infections worldwide . Brugia malayi is prevalent only in some parts of South and Southeast Asia , and B . timori is found only in Indonesia [1] . Currently , nearly 1 . 4 billion people in 73 countries worldwide are threatened by LF , with an estimated number of 120 million people infected , and about 40 million disfigured and incapacitated by the disease [2] . LF is a vector-borne disease; several species of Culex , Anopheles , Aedes , and Mansonia mosquitoes are known to be involved in its transmission . LF is one of the oldest and most debilitating neglected tropical diseases ( NTDs ) [3] . It is one of the important public health problems that face many countries in the developing world and is considered as an indicator of poverty [4] . While the infection is usually acquired in childhood , its visible manifestations occur later in life , causing temporary or permanent disability , such as elephantiasis and hydroceles [2] . Morbidity caused by chronic LF is mostly lifelong , and the disease is considered the world’s second leading cause of permanent long-term disability after mental illness [5 , 6] . LF is also responsible of a huge socio-economic burden . Indeed , patients affected by elephantiasis or hydrocele are often victims of societal discrimination , and the disease impairs their educational and employment opportunities , marriage prospects , and sexual life [1 , 7] . Lymphatic filariasis is among the diseases targeted for elimination . The Global Programme to Eliminate Lymphatic Filariasis ( GPELF ) elaborated a plan to achieve the goal of eliminating LF wherever it is endemic by 2020 . The elimination strategy has two components: ( i ) stop the spread of infection by interrupting transmission , and ( ii ) alleviate the suffering of affected individuals by controlling morbidity . In order to interrupt transmission , community-wide mass treatment program should be implemented to treat the entire at-risk population . The global strategy is a once-yearly single-dose of two-drug regimen utilized by communities at risk for LF , with the goal of reaching at least 65% population coverage yearly , for 5–6 years [8 , 9] . In Africa where onchocerciasis and LF are co-endemic , WHO recommends an annual dose of ivermectin ( 150 μg/kg bodyweight ) , in association with albendazole ( 400 mg ) [2] . As a prerequisite of this strategy , mapping the disease appears as the first programmatic step which should be completed to assess the disease situation in the country and identify areas where MDA is required [3] . Before the current survey , very little was known about LF in Cameroon , especially its distribution . A systematic review of prevalence data revealed that Cameroon was endemic for LF but further investigations were still needed to accurately delineate the endemic areas and/or assess the infection rate [10 , 11] . The present study then aimed at providing a quick-and-easy estimate of endemicity status and intensity of infection in Cameroon in order to draw a countrywide LF map and help the national program to implement mass drug administration ( MDA ) wherever the infection endemicity level is equal or greater than 1% [12] .
The objective of the present study was to map LF in Cameroon , and provide the national elimination program with the baseline data necessary for the evaluation of the impact of treatments on the endemicity and transmission of the disease . The health system in Cameroon has a pyramidal structure . From top to bottom , it is organized into central , intermediate ( Regional ) and peripheral ( District ) levels . Since most of the community interventions ( including treatments ) are implemented at the district level in Cameroon , we chose districts as the implementation units ( IU ) for our surveys . Actually , there are 181 health districts in Cameroon , each one divided into health areas . Health areas are made of communities or villages targeted for interventions . Of the 10 Regions of Cameroon , eight ( Adamawa , Centre , East , Littoral , North West , South , South West and West ) were targeted during the current surveys . The two other Regions ( North and Far North ) were not included in the surveys because treatments had already been implemented and were ongoing at the beginning of the present study , based on the certainty of the presence of the disease according to historical existing data [11] and history of clinical signs . Nevertheless , baseline data were collected in sentinel sites ( see below; data are available upon request ) selected in four health districts ( Kar-Hay , Kousseri , Pette and Yagoua ) of the Far North Region , where the highest LF endemicity level , based on the historical endemicity of the disease , were in this Region [11] . In addition , LF prevalence data concerning these two Regions were retrieved from systematic review documents covering the situation of LF in Cameroon in order to draw a complete map of the disease in the country . The present study was carried out in two phases using two types of surveys: The first phase of the study was performed through antigenaemia survey using the BinaxNow Filariasis immunochromatographic test ( ICT ) ( Alere Scarborough Inc , USA ) to assess the LF endemicity status and distribution in Cameroon . Depending on the ICT survey results , a microfilariaemia survey using nocturnal calibrated thick blood film was carried out in communities with the highest ICT positivity rates to establish a baseline for the evaluation of the program success [13] . Although ICT can also be used for monitoring and evaluation of LF programs in sentinel sites , the classical microscopic approach is still needed for confirmation or follow-up of people positive for ICT [3 , 14] . In fact , antigen rates decrease more slowly than microfilariaemia rates , and can underestimate the effects of MDA , particularly after the first few rounds [15] . This survey was carried out in 2009 following the WHO guidelines for rapid mapping of bancroftian filariasis in Africa [16] . From the very few published literature [10 , 11] and existing data on LF in Cameroon—hospital-based records of the clinical signs ( hydrocele and elephantiasis ) - , two villages or communities where transmission was likely to be ongoing ( previous data reporting LF or clinical signs ) were selected in each IU . In each of these villages or communities , having at least 50 individuals , either male or female , aged 5 years and over , were tested for daytime filarial antigenaemia using ICT—the rapid-format card test for detection of W . bancrofti antigen [17]—adhering to the manufacturer’s instructions . At the time of this testing , a visual inspection of the clinical presentation of each enrollee was performed and the most visible clinical signs ( elephantiasis and hydrocele ) , those most likely to be due to W . bancrofti infection , were recorded . At the end of the antigenaemia survey , a parasitological survey was conducted in 2010 in a limited number of villages selected in health districts with ICT positivity rate equal or greater than 1% . These groups of villages or communities , termed sentinel and spot check sites , are necessary to assess the success of the control program , i . e . to measure the impact of MDA on the endemicity level and transmission of LF . One sentinel site was identified in each health district , and at least 300 individuals ( either male or female ) aged 5 years and over were sampled . The sampling was limited—but not exclusively—to a single community . Regarding the low population density in some health districts , a buffer zone ( not larger than 1km ) was established around the target community in case the number of eligible individuals was less than 300 . Doing so , it was possible to examine the required number of people ( 300–500 ) in a minimum ( 2–3 ) of villages or communities close to each other and with similar epidemiological patterns . From each participant , a calibrated thick blood film was collected at night ( from 10 pm to 2 am the next day ) to take into account the nocturnal periodicity of W . bancrofti [18 , 19] . Qualified and/or trained lab technicians collected a 50μl sample of finger-prick blood from each study participant using non-heparinized capillary tubes . Blood samples were collected in absolute aseptic conditions using sterile and single use materials . Standard procedures were used for the processing and analysis of the blood samples [20] . Slides were examined independently by bright field microscopy ( magnification x100 ) , by two experienced laboratory technicians . W . bancrofti microfilariae were identified and counted and the results expressed as microfilariae per ml of blood ( mf/ml ) [21] . When any discrepancy was found , the preparation was re-examined by both lab technicians . All relevant data for LF were recorded into a purpose-built Microsoft Access database and subsequently exported into PASW Statistics version 18 ( SPSS Inc . , Chicago , IL , USA ) for statistical analysis . Antigenaemia and microfilariaemia endemicity levels were expressed as the percentage of infected individuals among the total number of individuals examined; the 95% confidence interval ( CI ) was calculated using the Wilson method not corrected for continuity [22] . The intensity of infection was computed when the microfilarial count was available as arithmetic means , and the sampling fluctuations estimated using the 95% confidence interval ( CI ) . Chi-square , Mann-Whitney and Kruskal-Wallis tests were used to compare LF endemicity level and mean intensity of infection between Regions , health districts , sexes and age groups , respectively . The geographical coordinates of each sampled village or community were recorded using a high sensitivity global positioning system [GPS eTrex; Garmin ( Europe ) Ltd , Southampton , U . K . ] . A thematic analysis was performed using a geographical information system ( GIS ) software ( ArcGIS , version 10 . 2 , ESRI Inc . ) to draw the LF map in Cameroon . For each IU ( health district ) , maps showing the LF endemicity status were drawn , for antigenaemia and microfilariaemia , and presented separately for each Region for readability requirements ( Figs 1–8 ) . An IU was considered non-endemic when the antigenaemia positivity rate was less than 1% , and endemic when the antigenaemia positivity rate was equal or greater than 1% ) [3] . To provide an estimate of the situation of LF in the entire country , historical data ( based on mf detection ) were retrieved from published literatures [10 , 11] to complement the existing circulating filarial antigen ( CFA ) or mf data . Then , the endemicity rates ( either from antigen or mf ) were interpolated at the IU ( health district ) level using the kriging method; health districts were then delineated and a spatially smoothed contour map drawn [23 , 24] to provide a quick estimate of the disease endemicity in the entire country ( Fig 9 ) . This study was conducted as part of the action plan of the national program to eliminate lymphatic filariasis in Cameroon . The surveys were approved by , and undertaken under the authority of , the Ministry of Public Health of Cameroon . After approval of the local administrative and traditional authorities , the objectives and schedules of the study were first explained to community leaders and to all eligible individuals . Because of low literacy rate in some areas and the logistics constraints related to the substantial number of individuals to be sampled , a written agreement was not requested . Verbal agreements were recommended by the Ministry of Public Health and obtained from those who agree to participate , under the discretion of community leaders . Even after the agreement of minors , the approval of their parents or legal guardians was necessary before any procedure . Each team leader was responsible to record agreements and attribute individual code to each participant for anonymous data analysis .
A total of 10 , 943 individuals aged 5–100 years old ( 5313 males and 5630 females ) were registered and examined in 2009 for the LF antigenaemia survey using ICT cards . The mean age ( standard deviation , sd ) of the enrollee was 37 . 0 ( 19 . 3 ) years old [males: 37 . 2 ( 19 . 6 ) and females: 36 . 8 ( 19 . 0 ) ] ( see S1 Dataset for detailed information ) . Figs 1–8 show the LF endemicity status in each implementation unit , aggregated at the Regional ( not national ) level for readability requirements . Among the 120 health districts ( IU ) visited in the eight Regions surveyed for LF circulating antigen detection , 106 ( 88 . 3% ) were found to be endemic for LF ( ICT positivity rate ≥ 1% ) , with endemicity level increasing from 1 . 0% ( 95%CI: 0 . 2–5 . 5 ) to 20 . 0% ( 95%CI: 10 . 0–30 . 0 ) . Table 1 also summarizes the crude LF endemicity status by Region , sex and age group . These results show that LF is endemic in each of the eight Regions surveyed ( overall infection rate: 3 . 3% ) . The lowest infection rate was found in the West Region [1 . 0% ( 95%CI: 0 . 7–1 . 6 ) ] and the highest one in the East Region [8 . 0% ( 95%CI: 6 . 5–9 . 9 ) ] . Males were significantly more infected than females ( p = 0 . 004 ) and a progressive increase in the LF endemicity level according to age was observed , adults aged 21 years and over being significantly more infected than their younger counterparts ( p< 0 . 0001 ) ( Table 1 ) . In 2010 , night blood samples were collected on 26 , 586 individuals among which 12 , 742 ( 47 . 9% ) were males and 13 , 799 were females ( 51 . 9% ) . The age of the enrollees varied between 5 and 110 years old , with an average of 31 . 2 ( 19 . 7 ) years old [30 . 1 ( 19 . 6 ) in males and 32 . 2 ( 19 . 8 ) in females] ( see S2 Dataset for detailed information ) . Figs 1–8 show endemicity status ( antigenaemia and microfilaraemia ) across the surveyed area , and Table 1 summarizes the crude LF positivity rate by Region , sex and age group . The LF endemicity level , assessed with mf detection in nocturnal blood smears , ranged between 0 . 3% and 4 . 0% in 11 health districts out of the 106 surveyed in the eight Regions . An average mf positivity rate of 0 . 11% ( 95% CI: 0 . 08–0 . 16 ) was recorded across the eight Regions included in this survey . The number of individuals harboring W . bancrofti mf was significantly higher ( p < 0 . 047 ) in the East ( 0 . 4% ) and Centre ( 0 . 2% ) Regions as compared to the other Regions . The infection rate was similar between males and females ( p = 0 . 120 ) as well as between age groups ( p = 0 . 150 ) . In the four health districts of the Far North Region not included in the mapping exercise , 404 individuals were examined for LF and CFA detected in 11 ( 2 . 7%; 95% CI: 1 . 5–4 . 8 ) of them . The variation of W . bancrofti mf densities by Region , sex and age group is given in Table 1 . The overall arithmetic mean mf density was 1 . 2 mf/ml ( 95% CI: 0 . 1–2 . 3 ) in the total population examined . Together with the infection endemicity level , the microfilarial density was significantly higher in the Centre ( 3 . 9 mf/ml ) and East ( 3 . 8 mf/ml ) Regions as compared to the six other Regions surveyed ( p < 0 . 0001 ) . Although the difference was not statistically significant ( p = 0 . 119 ) , the W . bancrofti mf density tended to be higher in males ( 2 . 0 mf/ml ) than in females ( 0 . 4 mf/ml ) . Also , the mf density was similar between age groups ( p = 0 . 150 ) , though an overall progressive increase was observed . In the Far North Region , an overall 1 . 7% ( 95% CI: 0 . 8–3 . 5 ) individuals harboured mf , with an intensity of infection of 7 . 49 mf/ml , all of the infected individuals being found in the Yagoua health district ( 4 . 9%; 95%CI: 2 . 4–9 . 7 ) . Among the 10 , 943 subjects examined during the ICT test survey , the visual inspection of their body revealed that 43 ( 0 . 4% ) and 45 ( 0 . 4% ) of them presented with the elephantiasis of the lower limbs and hydrocele , respectively . The occurrence of elephantiasis of the lower limbs was significantly higher in the Littoral ( 15 cases among the 43 recorded ) and North-West ( 15 cases among the 43 recorded ) Regions ( p< 0 . 0001 ) , and the proportion of individuals with hydrocele was significantly higher in the East ( 11 cases among the 43 recorded ) and North-West ( 23 cases among the 43 recorded ) Regions ( p< 0 . 0001 ) . The proportion of individuals with elephantiasis was even among sexes and age groups ( p> 0 . 06 ) whereas that of hydrocele was significantly higher in individuals aged more than 50 years ( p< 0 . 001 ) . At the community level , the LF endemicity status from the 2009–2010 surveys or retrieved from the literature was below the 1% cut-off for 46 . 4% of them . Among these communities , the disease endemicity level was equal to 0 . 0% for 96 . 1% , and the highest endemicity level was 0 . 7% for the very few communities ( 3 . 9% ) with positive cases . Above the LF endemicity threshold , the median endemicity rate was 4 . 0% ( Minimum = 1 . 1%; Maximum = 23 . 0% ) , with an interquartile range equal to 5 . 5% . From the LF endemicity rate at the community level , the spatial distribution of LF in the entire country was predicted by interpolation ( Kriging technique ) ( Fig 9 ) . Although occurring at very low endemicity level , LF is endemic in almost all the country . The distribution is mostly focal , except for the East , North and South Regions , as well as some parts of the North West and Far North Regions .
The present study was carried out to map the distribution of LF in Cameroon and provide the national program with baseline data for the assessment of the impact of treatments on the level of endemicity and transmission of the infection . LF endemicity status at the community level ( Figs 1–8 ) , together with the interpolation results ( Fig 9 ) , shows that the infection is widely distributed in Cameroon . All the Regions and about 90% of the health districts ( MDA implementation level in Cameroon ) surveyed were endemic with LF . Although the infection rate was quite low in these health districts endemic with LF , they were all qualified for MDA ( CFA rate ≥ 1% ) . Sentinel site identification using night blood smears resulted in negative outcomes in most of the health districts surveyed . Indeed , only 11 sentinel sites were identified in the 120 implementation units included in the survey . Consequently , very few mf baseline data exist for the evaluation of post-MDA implementation . However , although ICT can underestimate the effects of MDA since antigen rates decrease more slowly in the population , particularly after the first few rounds [15 , 25] , it can be used to evaluate the success of the LF elimination program as recommended by WHO [3] . Apart from the northern region of Cameroon ( Adamawa , North and Far North ) , loiasis is also co-endemic in most of the southern parts of the country [24] . It was shown that ICT can be positive in individuals harboring very high Loa mf loads , even when their LF CFA is negative [26] . Then , the LF endemicity level in some communities of the southern Cameroon can be overestimated . This potential for misclassifying LF endemicity because of ICT cross-reactivity with Loa might underestimate the effects of MDA in loiasis endemic areas; therefore , it appears necessary to perform night blood smears and molecular diagnostic ( using dried blood spots ) from people exhibiting positive ICT as confirmation tests . However , the proportion of individuals with W . bancrofti mf in night blood smears as well as higher numbers of hydrocele individuals in these communities presented with the same trend as compare to the CFA rate . Indeed , individuals harboring mf in the nocturnal blood smears were found in four of the eight Regions surveyed , but with very low infection rate ( ≤ 0 . 06% ) . These results support the LF endemicity in these areas , and confirm that the antigenaemia detection with immunochromatographic tests is more sensitive than the mf detection with microcopy [23] . In addition to the poor sensitivity of microscopy especially when the intensity of infection is low , amicrofilariaemic individuals cannot be detected by blood smear examinations for mf . One of the factors which can contribute to lower the rate and intensity of infection is the treatment . Indeed no MDA was ever implemented in the eight Regions surveyed to control LF at the outset of this study . However , as onchocerciasis was shown to be widely distributed in Cameroon—all the Regions are affected with more than 6 million people infected [27 , 28]—ivermectin was already distributed since 1992 ( up to 22 annual rounds in some areas ) as part of community interventions . It was shown that mf rate and intensity for LF infections were quite low in villages regularly treated for many years with ivermectin as compared to untreated villages [29] . It is noteworthy that the Cameroon LF distribution pattern observed in this study confirms the predictions from a multivariate Bayesian generalized linear spatial model develop to map the distribution of LF across Africa [30] . Results of LF antigenaemia evaluation have shown that males were significantly more infected than females , and CFA rate was increasing with age , adults aged 21 years and over being more infected than their younger counterparts . These findings are in accordance with previous literatures and might be explained by the exposition-protection pattern of individuals [23] . In fact , because of the activities they practice , males and adults are the most exposed to mosquito bites . Also , since females and children are the most vulnerable to malaria ( another mosquito transmitting disease ) , malaria control program has developed some control measures ( bed nets distribution to pregnant women for example ) which can finally protect them , as compared to males , both for malaria and lymphatic filariasis . It is important to notice however that such gender and age pattern regarding rate and intensity of LF infection assessed by microfilariaemia was not observed in our survey , as was found elsewhere [23] . This might likely be due to the very low proportion of infected individuals who , in addition , harbor very low mf densities . LF signs observed at the occasion of ICT card tests confirm that the disease is endemic in all the Regions surveyed . Although the migration history of affected individuals was not recorded , these signs were mostly found in North-West and Littoral Region where elephantiasis of non-filarial origin ( podoconiosis ) was also reported [31] . This being said , the immunochromatographic card test targeting specifically adult W . bancrofti antigen circulating in the blood was positive in these Regions , and hydroceles were also noticed . This study , together with historical data , provides the national program for elimination of LF in Cameroon with a countrywide map of the infection and highlight areas where MDA should be prioritized . These data allow the national program to implement MDA only in those health districts where loiasis is not co-endemic or where community directed treatments with ivermectin ( CDTI ) had already been ongoing for onchocerciasis control , although loiasis is co-endemic . Indeed , onchocerciasis maps were realized using the rapid epidemiological mapping strategy [27 , 32–33] , and loiasis map drawn [24] in support to onchocerciasis and LF control or elimination programs . In areas free for loiasis and where LF is endemic ( ICT rate ≥ 1% ) and onchocerciasis meso-endemic ( 20–40% nodules prevalence or 40–60 mf prevalence ) or hyper-endemic ( nodules prevalence > 40% or mf prevalence > 60% ) , annual treatments ( ivermectin in combination with albendazole ) are distributed to all eligible population . In areas where loiasis is endemic , CDTI can be organized in onchocerciasis meso- and hyper-endemic communities following specific recommendations [34] . However , in these loiasis endemic areas , the risk of SAE is considered to be higher as compared to the profit resulting in the treatment , regardless of LF endemicity status and when onchocerciasis is hypo-endemic ( nodules prevalence < 20% or mf prevalence < 40% ) . In the present study , among the health districts currently untreated for LF , some ( in the East and South Regions ) display the highest LF and loiasis endemicity levels . For example , among the 14 health districts of the East Region , treatments are ongoing in only four of them ( three being partially covered ) ; the population at risk in the remaining untreated area being estimated at 971 , 000 inhabitants ( about 90% of the total population of the Region ) . It then appears important to implement treatments in those areas in order to achieve the global target of eliminating LF by 2020 [3 , 35] . Indeed , from untreated health districts , the disease could be reintroduced in the neighboring health districts already freed from the disease by multiple rounds of treatments . An alternative to the currently used bi-therapy ( ivermectin in combination with albendazole ) can be the bi-annual treatment using albendazole alone , with an anti-vector component by the usage of long lasting insecticide nets ( LLINs ) [36–38] . A marked reduction in W . bancrofti infection and infectivity in humans was observed in some areas of northern Uganda [39] and in two states of Nigeria [40] where both MDA and LLINs were used to reduce transmission . This encouraged Nigeria to recently launch the Africa’s first ever nationwide co-implementation plan to defeat LF and malaria [40] . Aside for providing a countrywide LF distribution map , this study also provides the national program with the baseline parasitological data necessary to begin a national elimination program and enable the measurement of any impact of MDA on the endemicity level and transmission of LF in the defined sentinel and spot check sites .
|
Lymphatic filariasis , commonly known as elephantiasis , is a parasitic disease caused by the filarial nematodes Wuchereria bancrofti , Brugia malayi and Brugia timori . It is widely distributed in the tropics where it results in a chronic and debilitating disease . Nearly 1 . 4 billion people in 73 countries worldwide are threatened by lymphatic filariasis , with an estimated 120 million people infected , and more than 40 million disfigured and incapacitated by the disease . Mass drug administration of appropriate chemotherapeutic agents has been successful in eliminating the infection in some endemic areas supporting the contention that global elimination of the infection has become feasible . Before targeting lymphatic filariasis for elimination , it is necessary to map its distribution in order to identify areas where treatment is required . In this present study , two surveys were carried out in each of eight Regions of Cameroon to assess the endemicity status and intensity of the infection . Lymphatic filariasis was found to be endemic in all Regions surveyed and in almost all the constituent health districts . As virtually all of these Regions and health districts were found to be eligible for MDA treatments , baseline data were also acquired that can be used by the national program for the evaluation of the success of mass drug administration on the endemicity and transmission of the disease .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Mapping of Bancroftian Filariasis in Cameroon: Prospects for Elimination
|
Currently , there are no biomarkers that can predict the incidence of dengue shock and/or organ failure , although the early identification of risk factors is important in determining appropriate management to reduce mortality . Therefore , we sought to determine the factors associated with dengue shock and/or organ failure and to evaluate the prognostic value of serum procalcitonin ( PCT ) and peripheral venous lactate ( PVL ) levels as biomarkers of dengue shock and/or organ failure . A prospective observational study was conducted among adults hospitalized for confirmed viral dengue infection at the Hospital for Tropical Diseases in Bangkok , Thailand between October 2013 and July 2015 . Data , including baseline characteristics , clinical parameters , laboratory findings , serum PCT and PVL levels , management , and outcomes , were recorded on pre-defined case report forms . Of 160 patients with dengue , 128 ( 80 . 0% ) patients had dengue without shock or organ failure , whereas 32 ( 20 . 0% ) patients developed dengue with shock and/or organ failure . Using a stepwise multivariate logistic regression analysis , PCT ≥0 . 7 ng/mL ( odds ratio [OR]: 4 . 80; 95% confidence interval [CI]: 1 . 60–14 . 45; p = 0 . 005 ) and PVL ≥2 . 5 mmol/L ( OR: 27 . 99 , 95% CI: 8 . 47–92 . 53; p <0 . 001 ) were independently associated with dengue shock and/or organ failure . A combination of PCT ≥0 . 7 ng/mL and PVL ≥2 . 5 mmol/L provided good prognostic value for predicting dengue shock and/or organ failure , with an area under the receiver operating characteristics curve of 0 . 83 ( 95% CI: 0 . 74–0 . 92 ) , a sensitivity of 81 . 2% ( 95% CI: 63 . 6–92 . 8% ) , and a specificity of 84 . 4% ( 95% CI: 76 . 9–90 . 2% ) . Dengue shock patients with non-clearance of PCT and PVL expired during hospitalization . PCT ≥0 . 7 ng/mL and PVL ≥2 . 5 mmol/L were independently associated with dengue shock and/or organ failure . The combination of PCT and PVL levels could be used as prognostic biomarkers for the prediction of dengue shock and/or organ failure .
Dengue is the most important arthropod-borne viral disease , and it exerts a high burden on populations and public health systems in most tropical countries [1 , 2] . The incidence has dramatically increased during the last 50 years ( by 30-fold ) for all four dengue virus serotypes ( DENV 1–4 ) in more than 100 countries , including those in Southeast Asia , Central and South America , the Western Pacific , Africa , and the Eastern Mediterranean [2 , 3] . A previous report estimated that 390 million people are infected with DENV per year worldwide , of which 96 million show clinical manifestations of dengue [4] . Clinical manifestations range from acute febrile illness to severe dengue , which is a life-threatening condition [1] . In-hospital mortality is observed among 1 . 6–10 . 9% of patients with severe manifestations of dengue , including dengue hemorrhagic fever and/or dengue shock syndrome [5–7] . The World Health Organization ( WHO ) has implemented a goal of reducing dengue mortality by at least 20% and morbidity by 25% by the year 2020 [2] . Early recognition of severe dengue would help clinicians achieve close monitoring and provide proper fluid resuscitation in order to prevent severe disease , which would reduce mortality and morbidity . The revised 2009 WHO case definition was introduced in order to improve early recognition of severe dengue by increasing awareness of warning signs [1] . However , a recent systematic review showed that the definition had a wide range of sensitivity ( 59–98% ) and specificity ( 41–99% ) in the prediction of severe dengue [8] . The pathophysiology of severe dengue is complex , and involves an interplay of host immune and genetic factors with virulent strains of DENV [9 , 10] . The critical phase of severe dengue usually occurs as viremia declines [1] . DENV replication occurs within cells , particularly hepatocytes , monocytes , and macrophages , during systemic infection and the immune-mediated response following DENV infection , which is proportional to the viral load [11 , 12] . Immune-mediated pathogenesis has been considered a major cause of the increased vascular permeability of endothelial cells , leading to plasma leakage [9–11] . Delayed recognition and improper management of patients with plasma leakage can lead to shock and/or organ failure [11] . The prevalence of dengue shock among adults is approximately 18% , and it is the most common cause of death from DENV [13] . A previous systematic review and meta-analysis showed that several clinical factors , including age , female sex , neurological signs , nausea/vomiting , abdominal pain , gastrointestinal bleeding , hemoconcentration , ascites , pleural effusion , hypoalbuminemia , hypoproteinemia , hepatomegaly , high levels of aspartate aminotransferase ( AST ) and alanine aminotransferase ( ALT ) , abnormal coagulators , primary/secondary infection , and DENV-2 , were independently associated with the development of dengue shock [13] . Procalcitonin ( PCT ) is a functional immune modulating protein consisting of 114–116 amino acids , and is currently used as a novel biomarker for diagnostic and prognostic purposes [14] . PCT is produced and released into the bloodstream in response to infection and/or inflammation in various tissues . In particular , hepatocytes and peripheral blood mononuclear cells are potent PCT secretors [15] . A recent systematic review and meta-analysis showed that PCT was a useful biomarker for the early diagnosis of sepsis in critically ill patients , with a sensitivity and specificity of 77% and 79% , respectively [16] . The area under the receiver operating characteristics curve ( AUROC ) was 0 . 85 , indicating moderate diagnostic accuracy [16] . In Southeast Asia , an endemic area for tropical infectious diseases , the AUROC for discrimination between bacterial and viral infections using PCT was 0 . 74 , which was also indicative of moderate diagnostic accuracy [17] . Of the patients with dengue , 72% had a PCT level ≥0 . 1 ng/mL and 25% had a PCT level ≥0 . 5 ng/mL , which was higher than that of patients with influenza ( 34% at a PCT level ≥0 . 1 ng/mL and 16% at a PCT level ≥0 . 5 ng/mL [17] . Previous reports have also shown that PCT levels in patients with sepsis are associated with the severity of organ dysfunction [18] , and that PCT could be used as a prognostic marker for discrimination between patients with and without septic shock , in addition to survival [19] . A previous study showed that PCT levels on admission were significantly higher among patients who died following infection with the 2009 H1N1 strain of influenza , compared with those who survived ( 14 . 5 vs . 1 . 7 ng/mL ) [20] . In addition , arterial or venous lactate may be used as a biomarker for tissue hypoperfusion , regardless of organ failure or shock , particularly among patients with sepsis [21] . Our previous prospective study showed that peripheral venous lactate ( PVL ) concentration was independently associated with severe dengue [22] . In clinical practice , it can be difficult to identify the early stages of dengue shock and/or organ failure using clinical data . PCT and/or PVL may provide a superior prognostic method for predicting dengue severity at the time of hospital admission , particularly in the identification of patients at high risk of developing dengue shock and/or organ failure . At present , there have been no studies assessing the capacity of PCT and/or PVL to predict dengue shock and/or organ failure . Thus , we hypothesized that PCT and/or PVL may discriminate between patients who develop dengue shock and/or organ failure and those who do not . Therefore , we undertook a prospective observational study among hospitalized adults with dengue and determined the factors associated with dengue shock and/or organ failure . The prognostic values of PCT and PVL as biomarkers for predicting dengue shock and/or organ failure were evaluated .
The study design was approved by the Ethics Committee of the Faculty of Tropical Medicine , Mahidol University in Bangkok , Thailand . The Strengthening the Reporting of Observational Studies in Epidemiology ( STROBE ) statement ( S1 Checklist ) and the Standards for the Reporting of Diagnostic ( STARD ) accuracy ( S2 Checklist ) were followed in this study [23 , 24] . Patients aged ≥15 years with clinical dengue , defined as acute fever and ≥2 of the following symptoms were included: 1 ) headache , 2 ) ocular pain , 3 ) myalgia , 4 ) arthralgia , 5 ) rash , 6 ) a positive tourniquet test ( ≥20 petechiae per square inch ) , or 7 ) leukopenia ( white blood cell [WBC] counts <5 . 0 × 103 cells/μL ) . Patients had been admitted to hospital for treatment , and the broad criteria allowed physicians to invite all potential patients to participate in the study at the outpatient and emergency department . Written informed consent was obtained from all patients , or the patient's guardians if the patient was 15–18 years old , before participation in the study . This prospective observational study was performed among patients who were admitted to the Hospital for Tropical Diseases ( Faculty of Tropical Medicine , Mahidol University in Bangkok , Thailand ) between October 2013 and July 2015 . The inclusion criteria were ( i ) age ≥15 years , ( ii ) clinical dengue , and ( iii ) confirmed dengue viral infection by reverse-transcriptase polymerase chain reaction ( RT-PCR ) from a serum sample obtained at admission , and/or positive micro-neutralization test results from serum samples obtained at admission and 2 weeks after admission , and/or dengue-specific immunoglobulin M ( IgM ) and immunoglobulin G ( IgG ) detected using enzyme-linked immunosorbent assays ( ELISAs ) in paired serum samples taken at admission and 2 weeks after admission . Patients with an underlying medical illness , mixed infection , current pregnancy , current use of any non-topical antibiotic , or current fluid therapy were excluded from this study . Laboratory tests were conducted at admission , including a complete blood count and blood chemistry assessment , and samples for the measurement of PCT and PVL were collected . Blood samples for PCT and PVL analysis were collected every 24 h until the patient exhibited a body temperature of <37 . 8°C for 48 h . Treating physicians and investigators were blinded to the PCT and PVL results . All patients received standard management from their treating physicians , according to the 2009 WHO guidelines for dengue [1] . In order to exclude other infections , two blood samples for microbiological cultures were obtained , urinalysis was performed , and plain radiography of the chest was routinely performed at admission . Diagnostic tests for other infectious diseases were also performed when indicated by clinical findings at admission or during hospitalization . Dengue severity and outcomes were summarized on discharge . All patient data , including baseline characteristics , clinical parameters , laboratory findings , management , and outcomes , were recorded on a pre-defined case report form . At a 2-week follow-up appointment , blood samples were collected for complete blood counts and serum creatinine assessment . Subsequent follow-up was required within the following 2 months until the laboratory results reached reference ranges in order to serve as a baseline . The WHO 2009 dengue definition was used to classify dengue shock and organ failure in this study [1] . Dengue shock was defined as plasma leakage with shock . Plasma leakage was defined as ≥20% increase in hematocrit above baseline or clinical fluid accumulation manifested by pleural effusion , ascites , or serum albumin <3 . 5 g/dL . Shock was defined as ( 1 ) a rapidly weak pulse with pulse pressure <20 mmHg , or ( 2 ) a systolic blood pressure of <90 mmHg with tissue hypoperfusion evidenced by one of the following criteria: ( i ) decreased urine output ( <0 . 5 mL/kg/h ) , ( ii ) impaired consciousness , ( iii ) AST >1000 IU/L , ( iv ) ALT >1000 IU/L , ( v ) cold skin , or ( vi ) clammy skin . Organ failure was defined as the presence of one of the following criteria: ( i ) respiratory distress ( a respiratory rate of ≥24 breaths/min with <95% oxygen saturation in room air and/or the need for oxygen therapy ) , ( ii ) serum creatinine increased ≥3-fold from baseline , ( iii ) AST >1000 IU/L , ( iv ) ALT >1000 IU/L , ( v ) myocarditis , ( vi ) encephalitis , or ( vii ) spontaneous gastrointestinal bleeding requiring blood transfusion . The WHO 2009 dengue definitions for warning signs ( WSs ) were also used in this study; WSs included ( 1 ) abdominal pain; ( 2 ) vomiting; ( 3 ) clinical fluid accumulation defined as the presence of pleural effusion determined by plain radiography of the chest or a serum albumin level <3 . 5 g/dL; ( 4 ) lethargy; ( 5 ) a liver span of >15 cm; ( 6 ) bleeding from a mucosal area , including the nose , gums , gastrointestinal tract , or vagina; and ( 7 ) an increase in hematocrit of 2% above the sex-specific reference range for a healthy Thai adult with a platelets of ≤100 × 103/μL . Dengue viral RNA was detected from patient serum at admission using a two-step PCR method , as described by Lanciotti et al . [25] , and modified using the methods of Reynes et al . [26] . Viral RNA was detected from acute serum samples using a PureLink Viral RNA/DNA Mini Kit ( Invitrogen , Grand Island , NY , USA ) , according to the manufacturer’s instructions . Serum samples collected at admission and 2 weeks after admission were assayed for serotype-specific DENV using the micro-neutralization test described by Vorndam et al . [27] , with the slightly modified protocol of Putnak et al . [28] . The micro-neutralization test based on the principle of the plaque reduction neutralization test was used to measure serotype specific anti-DENV neutralizing antibodies against all 4 serotypes . Serum samples were tested in triplicate and sera were serially diluted 2-fold from 1:20 to 1:5120 in a 96-well microplate . Each microplate included media only ( negative control ) , a virus control and sera of known specific DENV serotypes ( positive controls ) . The average number of virus foci were counted , and only assays with a virus control in the range of 50–60 foci per well and a media only control with no foci were included . For control sera of known DENV serotypes , at least 50% inhibition by viral replication was required ( 25–30 foci per well ) , compared with the virus control . The virus neutralization titer was defined as the reciprocal of the serum dilution providing 50% inhibition of viral replication compared with the virus control . A positive serotype specific anti-DENV test was defined as a 4-fold rise in neutralizing antibody titer in paired samples for 1 of the 4 DENV serotypes . All sera collected at admission and 2 weeks after admission were tested using four separate capture ELISA assays for IgM and IgG against dengue virus and Japanese encephalitis virus , as described by Innis et al . [29] . The assay was performed using serum samples diluted 1:100 . Assay results for test samples were expressed as units calculated by the following formula: units = 100 × ( A492test sample–A492NS ) / ( A492PS–A492NS ) , where A492 was an absorbance at 492 nm , NS was a normal human serum negative standard , and PS was pooled sera from flavivirus infected patients . Both acute and convalescent sera were used for the assay . Only sera with either anti-dengue IgM or anti-Japanese encephalitis IgM levels ≥40 units were evaluated . To discriminate between dengue and other flavivirus infections , we determined the ratio of dengue IgM to Japanese encephalitis virus IgM , with a ratio ≥1 . 0 indicating dengue virus infection and a ratio <1 . 0 indicating other flavivirus infection . To discriminate primary from secondary dengue infection , the ratio of anti-dengue IgM to anti-dengue IgG was also calculated , with a ratio ≥1 . 8 indicating primary dengue infection and a ratio <1 . 8 indicating secondary dengue infection . This cut-off value was applied for either acute or convalescent samples as long as either an anti-dengue IgM or anti-dengue IgG response could be detected . Pooling the results for both acute and convalescent sera using the same cut-off value allowed more accurate classification of primary and secondary dengue . PCT was measured using an electrochemiluminescence method ( Elecsys BRAHMS PCT , Roche Diagnostic , Mannheim , Germany ) according to the manufacturer’s instructions using a Cobas e 411 immunoassay analyzer ( Roche Diagnostic , Mannheim , Germany ) . Prior to assessment , frozen serum samples were stored at –80 °C by laboratory personnel blinded to patient status . The detection limit for the PCT assay was 0 . 02 ng/mL . The coefficients of variation for low and high concentrations were 1 . 7% and 1 . 4% , respectively . Blood samples were collected from a vein in an upper extremity without the use of a tourniquet . A 2 mL blood sample was collected into a vacutainer tube containing sodium fluoride , immediately placed on ice , sent to the laboratory , and analyzed for lactate within 10 min . Lactate levels were measured by a colorimetric assay with an enzymatic reaction using an auto-analyzer ( Roche/Hitachi Cobas C Systems , USA ) , according to the manufacturer’s protocol . The laboratory personnel were blinded to the sample sources . The coefficient of variation for the assay in our laboratory was 1 . 1% . A previous prospective study at the Hospital for Tropical Diseases ( Bangkok , Thailand ) indicated that the incidence of dengue shock and/or organ failure was 21 . 0% among hospitalized adults with dengue [30] . Based on this information , we calculated that a sample size of at least 122 patients was needed for this study , using a specificity of 90% with a confidence interval ( CI ) of ±6% . All data were analyzed using SPSS software ( version 18 . 0; SPSS Inc . , Chicago , IL ) . Numerical variables were tested for normality using Kolmogorov-Smirnov tests . Variables with non-normal distribution were summarized as medians and interquartile ranges ( IQRs ) , and were compared using Mann-Whitney U tests for two-group comparisons . Categorical variables were expressed as frequencies and percentages , and were analyzed using chi-squared or Fisher’s exact tests , as appropriate . A univariate logistic regression analysis was performed with each potential factor included as an independent variable , and the presence or absence of dengue shock and/or organ failure as the dependent variable . Any variable with a p-value ≤0 . 2 was considered potentially significant and was further analyzed in a stepwise multivariate logistic regression analysis using a backward selection method for determining significant independent factors . The optimal cut-off values of factors predictive of dengue shock and/or organ failure were determined using ROC curves . Prognostic parameters were evaluated using 2 × 2 tables , and 95% CIs were calculated to determine sensitivity , specificity , negative predictive value ( NPV ) , positive predictive value ( PPV ) , positive likelihood ratio ( LR+ ) , and negative likelihood ratio ( LR– ) . The optimal PCT and PVL cut-off values were then combined in a single “bioscore” , as described by Gibot et al , 2012 [31] . The bioscore attributed one point per biomarker with a value above or equal to the optimal cut-off value . The bioscore was defined as 0 ( both biomarkers below their respective cut-off value ) , 1 ( any one of the two biomarkers above/equal to the cut-off value ) , or 2 ( both biomarkers above/equal to the cut-off value ) . The bioscore was then further tested for prognostic value in predicting dengue shock and/or organ failure by logistic regression analysis . All tests of significance were two-sided , with a p-value <0 . 05 indicating statistical significance .
A total of 189 adults with suspected dengue were admitted to the Hospital for Tropical Diseases ( Bangkok , Thailand ) between October 2013 and July 2015 . Of 189 hospitalized adults with suspected dengue viral infection , 29 patients were excluded due to an underlying illness ( 17 patients , 58 . 6% ) , mixed infection ( 10 patients , 34 . 5% ) , or a negative RT-PCR/micro-neutralization/ELISA for dengue ( 2 patients , 6 . 9% ) . Thus , 160 hospitalized adults with confirmed dengue viral infection were finally recruited for this study . Among the 160 patients , 32 ( 20 . 0% ) patients had dengue shock ( 23 patients [71 . 9%] ) and/or organ failure ( 26 patients [81 . 2%] ) , whereas 128 ( 80 . 0% ) patients had dengue without shock or organ failure ( Fig 1 ) . In the 26 patients with organ failure , respiratory distress ( 11 patients [42 . 3%] ) , AST levels >1000 IU/L and/or ALT >1000 IU/L ( 9 patients [34 . 6%] ) , serum creatinine concentration ≥3-fold greater than baseline ( 6 patients [23 . 1%] ) , spontaneous gastrointestinal bleeding requiring blood transfusion ( 6 patients [23 . 1%] ) , myocarditis ( 4 patients [15 . 4%] ) , and encephalitis ( 3 patients [11 . 5%] ) were observed . At admission , patients with dengue shock and/or organ failure were significantly more likely to have a longer duration of fever ( p = 0 . 031 ) , skin bleeding ( p = 0 . 012 ) , mucosal bleeding ( p <0 . 001 ) , vomiting ( p = 0 . 024 ) , a liver span of >15 cm ( p = 0 . 001 ) , decreased breathing sounds ( p <0 . 001 ) , and increased respiratory rate ( p = 0 . 010 ) . When numerical parameters were categorized , patients aged >40 years ( p = 0 . 023 ) , with a fever duration ≥5 days ( p = 0 . 041 ) , respiratory rate ≥24 breaths/min ( p = 0 . 005 ) , mean arterial pressure <70 mmHg ( p = 0 . 030 ) , or pulse pressure <30 mmHg ( p = 0 . 005 ) were more likely to have dengue shock and/or organ failure ( Table 1 and S1 Table ) . Regarding laboratory parameters ( Table 2 and S2 Table ) , patients with dengue shock and/or organ failure had significantly higher hemoglobin concentrations ( p = 0 . 045 ) , increased hematocrit values above baseline ( p <0 . 001 ) , higher WBC counts ( p = 0 . 044 ) , higher absolute bands ( p = 0 . 022 ) , higher absolute atypical lymphocyte counts ( p = 0 . 007 ) , higher AST levels ( p <0 . 001 ) , higher ALT levels ( p <0 . 001 ) , higher PCT levels ( p = 0 . 001 ) , and higher PVL levels ( p <0 . 001 ) ( Fig 2 ) . However , patients with dengue shock and/or organ failure had significantly lower platelet counts ( p <0 . 001 ) and albumin levels ( p <0 . 001 ) . When laboratory parameters were categorized based on the reference ranges ( Table 2 ) , patients with WBC counts >5 . 0 × 103 cells/μL ( p = 0 . 004 ) , absolute bands >200 cells/μL ( p = 0 . 049 ) , absolute atypical lymphocyte counts >300 cells/μL ( p = 0 . 006 ) , AST >120 IU/L ( p = 0 . 002 ) , ALT >120 IU/L ( p = 0 . 002 ) , PCT ≥0 . 7 ng/mL ( p = 0 . 002 ) , and PVL ≥2 . 5 mmol/L ( p <0 . 001 ) were more likely to have dengue shock and/or organ failure . In addition , patients with platelet counts <50 . 0 × 103 cells/μL ( p = 0 . 012 ) and albumin <3 . 5 g/dL ( p = 0 . 001 ) were also more likely to have dengue shock and/or organ failure . Assessment of patient management and outcomes during hospitalization demonstrated that a significant proportion of patients with dengue shock and/or organ failure received albumin as fluid resuscitation ( p <0 . 001 ) and antibiotics ( p = 0 . 017 ) . Of the 32 patients with dengue shock and/or organ failure , 4 ( 12 . 5% ) received mechanical ventilation , 3 ( 9 . 4% ) received renal replacement therapy , and 2 ( 6 . 2% ) received vasopressors . Patients with dengue shock and/or organ failure had significantly longer durations of hospitalization ( p = 0 . 006 ) . However , only two patients expired during hospitalization , both due to multi-organ failure ( S2 Table ) . A univariate logistic regression analysis was used to determine which of the baseline characteristics , clinical parameters , and laboratory findings were associated with the occurrence of dengue shock and/or organ failure . All clinical factors potentially associated with the occurrence of dengue shock and/or organ failure were included in the univariate logistic regression analysis . The following variables were identified as clinical parameters associated with dengue shock and/or organ failure: ( 1 ) age >40 years , ( 2 ) fever duration ≥5 days , ( 3 ) absolute bands >200 cells/μL , ( 4 ) absolute atypical lymphocyte counts >300 cells/μL , ( 5 ) PCT ≥0 . 7 ng/mL , and ( 6 ) PVL ≥2 . 5 mmol/L ( Table 3 ) . All parameters with a p-value ≤0 . 2 in the univariate logistic regression analysis were then further analyzed by a stepwise multivariate logistic regression analysis using a backward selection method , in order to determine the independent factors significantly associated with the occurrence of dengue shock and/or organ failure . The following clinical and laboratory parameters were found to be independently associated with the occurrence of dengue shock and/or organ failure: ( 1 ) PCT ≥0 . 7 ng/mL ( odds ratio [OR]: 4 . 80; 95% CI: 1 . 60–14 . 45; p = 0 . 005 ) and ( 2 ) PVL ≥2 . 5 mmol/L ( OR: 27 . 99 , 95% CI: 8 . 47–92 . 53; p <0 . 001 ) ( Table 4 ) . The two biomarkers PCT ≥0 . 7 ng/mL and PVL ≥2 . 5 mmol/L were assessed as a combined bioscore using a logistic regression model to evaluate the prognostic capacity in predicting the occurrence of dengue shock and/or organ failure . Higher bioscores were associated with increased occurrence of dengue shock and/or organ failure , with ORs of 22 . 23 ( 95% CI 7 . 85–63 . 00 ) and 30 . 00 ( 95% CI 5 . 76–156 . 31 ) for a bioscore 1 and 2 , respectively ( p <0 . 001 ) ( Table 4 ) . The AUROC for PCT in the prediction of dengue shock and/or organ failure was 0 . 69 ( 95% CI: 0 . 59–0 . 80 ) ( Fig 3A ) . The AUROC for PVL in the prediction of dengue shock and/or organ failure was 0 . 78 ( 95% CI: 0 . 68–0 . 88 ) ( Fig 3B ) . The prognostic values of PCT and PVL at admission for predicting dengue shock and/or organ failure are shown in Table 5 . The sensitivities for nearly all PCT and PVL categories were low , except for PVL ≥1 . 5 mmol/L . The specificities for PCT and PVL categories were high , except for PCT ≥0 . 5 ng/mL and PVL ≥1 . 5 mmol/L . The PPVs and LR+ values for PCT and PVL categories were low , except for PVL ≥2 . 5 mmol/L and ≥3 . 0 mmol/L . The NPVs for all PCT and PVL categories were high . The LR–values for PCT and PVL were high , indicating possible prediction of dengue shock and/or organ failure . In order to accurately predict dengue shock and/or organ failure in a greater number of patients , the optimal levels of PCT ≥0 . 7 ng/mL and PVL ≥2 . 5 mmol/L were combined as a bioscore . The AUROC for a combined bioscore in the prediction of dengue shock and/or organ failure was 0 . 83 ( 95% CI: 0 . 74–0 . 92 ) ( Fig 3C ) . The combined bioscore provided good prognostic value for the prediction of dengue shock and/or organ failure among hospitalized adults with dengue , giving an optimal sensitivity of 81 . 2% ( 95% CI: 63 . 6–92 . 8% ) ; specificity of 84 . 4% ( 95% CI: 76 . 9–90 . 2% ) ; PPV of 56 . 5% ( 95% CI: 41 . 1–71 . 1% ) ; NPV of 94 . 7% ( 95% CI: 88 . 9–98 . 0% ) ; LR+ of 5 . 2 ( 95% CI: 3 . 4–8 . 0 ) ; and LR–of 0 . 2 ( 95% CI: 0 . 1–0 . 5 ) ( Table 5 and S3 Table ) . In order to evaluate the use of WHO 2009 WSs for identifying dengue shock and/or organ failure at admission , the diagnostic values of individual WSs and number of WSs were evaluated ( S4 Table and Table 6 ) . The sensitivities for all individual WSs were low , except for the following: lethargy , 87 . 5% ( 95% CI: 71 . 0–96 . 5% ) ; mucosal bleeding , 78 . 1% ( 95% CI: 60 . 0–90 . 7% ) ; and hematocrit >2% with platelets ≤100 ×103/μL , 79 . 4% ( 95% CI: 40 . 6–76 . 3% ) . Similarly , the specificities for all WSs were low , except that for clinical fluid accumulation ( 89 . 8%; 95% CI: 83 . 3–94 . 5% ) . The PPVs were low , but the NPVs for the WSs were high . The LR+ values for the WSs were low , except that for clinical fluid accumulation ( 5 . 2; 95% CI: 2 . 8–9 . 6 ) . The LR–values for the WSs ranged from 0 . 4 to 0 . 8 ( S4 Table ) . When the number of WSs was used to identify dengue shock and/or organ failure , this resulted in an AUROC of 0 . 77 ( 95% CI: 0 . 68–0 . 87 ) ( Fig 3D ) . WSs ≥4 had an optimal sensitivity of 75 . 0% ( 95% CI: 56 . 6–88 . 5% ) and a specificity of 71 . 1% ( 95% CI: 62 . 4–78 . 8% ) . A low PPV of 39 . 3% ( 95% CI: 27 . 1–52 . 7% ) and an LR+ of 2 . 6 ( 95% CI: 1 . 8–3 . 6 ) , but a high NPV of 91 . 9% ( 95% CI: 84 . 7–96 . 4% ) and an LR–of 0 . 4 ( 95% CI: 0 . 2–0 . 6 ) , were obtained for WSs ≥4 , which indicated a small decrease in the likelihood of developing dengue shock and/or organ failure when the number of WSs was <4 ( Table 6 ) . In order to evaluate the average changes in PCT and PVL during hospitalization , PCT and PVL levels were measured at admission and every 24 h during hospitalization until the patient exhibited a body temperature <37 . 8°C for 48 h . Of the 32 patients who developed dengue shock and/or organ failure , only two ( 6 . 2% ) expired during hospitalization , while 30 ( 93 . 8% ) patients survived with complete recovery of organ function . All 128 patients without dengue shock or organ failure survived . Among the patients who survived , those with dengue shock and/or organ failure had higher PCT and PVL levels at admission and during hospitalization than those without ( Fig 4 ) . However , the patients who expired during hospitalization showed a trend toward increased PCT ( >2 ng/mL ) and PVL ( >10 mmol/L ) levels 24 h after admission ( Fig 4 ) .
The strengths of our study were the prospective observational design and assessment of serial samples for PCT and PVL analysis . In addition , treating physicians and investigators were blinded to results in order to reduce missing data and minimize bias . All participants in this study were enrolled during the febrile phase of dengue; thus , the predictive parameters determined in this study could help physicians with early prediction of dengue shock and organ failure during the critical phase of dengue . However , our study had some limitations , as follows: ( 1 ) this study was conducted in a single center in Thailand , which was the referral center for tropical infectious diseases; ( 2 ) we could not perform cultures from sites requiring invasive investigation , such as peritoneal fluid or pleural fluid , as patients with dengue are at risk of bleeding; ( 3 ) empiric antibiotics were prescribed after hemocultures were taken , and ( 4 ) although all adult patients with clinical dengue were enrolled as described in the inclusion criteria , a number of older patients with dengue do not exhibit the full range of symptoms and may therefore have been inadvertently excluded . Therefore , our study focused on the assessment of younger adults with dengue . The utility of PCT and PVL in older patients with dengue remains unknown . Nonetheless , this study was the first to demonstrate that PCT levels ≥0 . 7 ng/mL and PVL levels ≥2 . 5 mmol/L were independently associated with dengue shock and/or organ failure , and that their combination provided good prognostic value for predicting dengue shock and/or organ failure . Dengue shock patients with non-clearance of PCT or PVL expired during hospitalization . These finding may help clinicians to predict dengue shock and/or organ failure earlier among hospitalized adults with dengue , leading to improved patient management and reduced in-hospital mortality and morbidity among patients with dengue .
|
Dengue is a major global health concern , particularly in tropical countries , and affects all age groups . Mortality rates among patients who have been hospitalized with severe dengue are 1 . 6–10 . 9% , and death in adults is mainly due to the development of dengue shock and organ dysfunction . In states of poor tissue circulation or shock , lactate is produced . Additionally , procalcitonin is a highly specific biomarker of systemic inflammation . Therefore , we assessed whether procalcitonin and peripheral venous lactate could be used to predict the incidence of dengue shock and/or organ failure in patients with dengue . Our study showed that a combination of serum procalcitonin levels ≥0 . 7 ng/mL and peripheral venous lactate levels ≥2 . 5 mmol/L at admission could discriminate between patients who did and did not develop shock and/or organ failure , with high sensitivity and specificity . These parameters may therefore be useful as prognostic biomarkers . Our results suggest that serum procalcitonin is indicative of an extensive early inflammatory response , which may occur during the systemic phase of dengue . Peripheral venous lactate may be produced as a result of the poor tissue circulation that precedes dengue shock . Our findings may help clinicians to predict dengue shock and/or organ failure earlier to reduce in-hospital mortality .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"clinical",
"laboratory",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"biomarkers",
"regression",
"analysis",
"bacterial",
"diseases",
"mathematics",
"signs",
"and",
"symptoms",
"statistics",
"(mathematics)",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"dengue",
"fever",
"mathematical",
"and",
"statistical",
"techniques",
"clinical",
"laboratories",
"blood",
"plasma",
"hematology",
"biochemistry",
"diagnostic",
"medicine",
"blood",
"anatomy",
"fevers",
"physiology",
"hemorrhage",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"physical",
"sciences",
"vascular",
"medicine",
"statistical",
"methods"
] |
2016
|
Serum Procalcitonin and Peripheral Venous Lactate for Predicting Dengue Shock and/or Organ Failure: A Prospective Observational Study
|
How do neuronal populations in the auditory cortex represent acoustic stimuli ? Although sound-evoked neural responses in the anesthetized auditory cortex are mainly transient , recent experiments in the unanesthetized preparation have emphasized subpopulations with other response properties . To quantify the relative contributions of these different subpopulations in the awake preparation , we have estimated the representation of sounds across the neuronal population using a representative ensemble of stimuli . We used cell-attached recording with a glass electrode , a method for which single-unit isolation does not depend on neuronal activity , to quantify the fraction of neurons engaged by acoustic stimuli ( tones , frequency modulated sweeps , white-noise bursts , and natural stimuli ) in the primary auditory cortex of awake head-fixed rats . We find that the population response is sparse , with stimuli typically eliciting high firing rates ( >20 spikes/second ) in less than 5% of neurons at any instant . Some neurons had very low spontaneous firing rates ( <0 . 01 spikes/second ) . At the other extreme , some neurons had driven rates in excess of 50 spikes/second . Interestingly , the overall population response was well described by a lognormal distribution , rather than the exponential distribution that is often reported . Our results represent , to our knowledge , the first quantitative evidence for sparse representations of sounds in the unanesthetized auditory cortex . Our results are compatible with a model in which most neurons are silent much of the time , and in which representations are composed of small dynamic subsets of highly active neurons .
How does a population of cortical neurons encode a sensory stimulus such as a sound ? At one extreme , the neural representation could be dense , engaging a large fraction of neurons , each with a broad receptive field . At the other extreme , the neural representation could be sparse , at any moment of time engaging only a small fraction of neurons , each highly selective with a narrow receptive field . Although a dense code under some conditions makes the most efficient use of the “representational bandwidth” [1] available in a neuronal population—why should a large fraction of neurons remain silent most of the time ? —sparse models have recently gained support on both theoretical [2–4] and experimental [5–11] grounds . However , it is not at present clear which of these is a better model of sensory representations in the auditory cortex . In order to distinguish between these alternatives experimentally , we must know what fraction of neurons responds to a given stimulus . The direct experimental approach to measuring the density of a cortical code would begin by simultaneously recording sound-evoked responses of all the neurons in the auditory cortex to an ensemble of stimuli; one could then simply count the number of spikes elicited by each stimulus . Unfortunately , currently available recording techniques do not permit such a direct approach . An alternative approach is to record the activity of a representative subset of neurons serially , and infer the population response from this sample . In this way , the population code could in principle be inferred by sequentially sampling a large population of single unit responses . We have used cell-attached recording in the primary auditory cortex of unanesthetized rats to sample the population response to brief tones and other stimuli . Because we were interested in the population response , we presented a restricted ensemble of stimuli to each neuron , rather than optimizing the stimulus to drive each neuron to fire maximally [12 , 13] . Thus we could assess the fraction of neurons that responded for each stimulus we presented . The stimuli ranged from simple ( tones , sweeps , white-noise bursts ) to complex ( natural sounds ) . Our data therefore address the question: What is the typical response across the entire neuronal population to a particular stimulus ? Rather than: What is the optimal stimulus for a particular neuron ? ( see also [14] ) . We find that the typical population response in unanesthetized auditory cortex is sparse . Consistent with previous findings in barrel cortex [10 , 15 , 16] , some neurons had very low spontaneous firing rates ( <0 . 01 spikes/second ) ; at the other extreme , some neurons had driven rates in excess of 50 spikes/second . However , a given stimulus typically elicited a high firing rate ( >20 spikes/second ) in less than 5% of the population . Note that sparseness as used here refers only to the fraction of neurons active at a given instant; it is quite possible that each neuron might , under the appropriate conditions ( e . g . , when presented with an optimal stimulus ) , participate in a representation by firing at a high rate . Our results represent , to our knowledge , the first quantitative experimental support for the hypothesis that the representation of sounds in the auditory cortex of unanesthetized animals is sparse .
Consistent with the earliest studies of unanesthetized auditory cortex [19–21] , tones evoked a wide range of response patterns . Tones could elicit either an increase or a decrease in a neuron's firing rate over the background firing rate , or both; the change could be transient , delayed , or sustained; and the response pattern could be different for different tone frequencies in a single neuron . Figure 2 shows some examples of the range of response types we observed . In one neuron ( Figure 2A ) , tones elicited a transient , short latency response of the sort commonly observed in the barbiturate-anesthetized auditory cortex . In a second neuron ( Figure 2B ) , tones elicited a suppression of background activity . In a third neuron ( Figure 2C ) , higher frequency tones ( ∼8–40 kilohertz [kHz] ) elicited vigorous sustained firing; interestingly , lower-frequency tones elicited transient responses in the same neuron , emphasizing that the distinction between “transient” and “sustained” applies to responses , not neurons . Other more complex response patterns were also observed ( Figure 2D–2G . ) Finally , half of the neurons tested ( 50% , see below ) showed no change in firing rate for any stimulus presented ( Figure 2H ) . Because a given neuron could show very different response patterns to stimuli of different frequencies ( e . g . , Figure 2C ) , we could not find a simple and objective scheme for organizing neurons into a small number of distinct classes , such as “transient , ” “sustained , ” “off , ” etc . The neurons shown in Figure 2 are a subset; the complete set of responses from the entire dataset is shown in the Figures S4–S8 . We first analyzed the basic population response elicited by tones , beginning with the response to tones presented at 50 or 60 decibels ( dB SPL ) . We divided the tone-evoked response into four 50-millisecond ( ms ) long “epochs”: spontaneous , early , late , and off ( Figure 3A , also see Materials and Methods ) . To ensure a sufficient number of trials for assessing the statistical significance of putative changes in firing rate over background , we grouped responses across nearby frequencies ( one-octave-wide bins; four- or five-octave bins for each response epoch ) . Control analyses using narrower ( half-octave ) bins gave similar results ( see Materials and Methods ) , as expected from the relatively broad frequency tuning of neurons in the rat primary auditory cortex [22 , 23]; see also [24] . Both spontaneous and evoked firing rates were typically low ( see Figure 3 and Table 1 ) . The median spontaneous firing rate across the population was 2 . 8 spikes/second ( sp/s ) . The mean was somewhat higher ( 4 . 9 sp/s ) because it was dominated by a relatively small set of neurons—possibly interneurons ( see Very Responsive Neurons May Be Narrow-Spiking Interneurons , below ) —with high spontaneous rates . Evoked firing rates showed the same pattern: a low median ( 2 . 0–2 . 7 sp/s ) and a somewhat higher mean ( 5 . 4–7 . 0 sp/s ) . The higher mean rates reflect the fact that in some neurons , some frequencies evoked vigorous firing ( see Figure 2C for an example ) . However , such well-driven responses were the exception rather than the rule; as quantified below , most neurons did not respond vigorously to any of the tones presented . Note that for a neuron to contribute on average at least one spike to the population representation of a sound in a 50-ms window , its evoked firing rate must exceed 20 sp/s . To assess whether the low firing rates resulted from some intrinsic defect of the spike-generating mechanism , perhaps introduced by the cell-attached recording method , we extracted the shortest interspike interval ( ISI ) for each neuron . In most neurons , the shortest ISI was less than 10 ms ( median shortest ISI = 4 ms , n = 145 cells ) . Thus , the low firing rates do not appear to arise from an intrinsic inability of neurons to fire rapidly , but instead presumably arise from differences in the synaptic drive received by different neurons . The distribution of spontaneous firing rates across the population was remarkably well fit with a lognormal distribution—that is , the logarithm of the firing rates was well fit with a Gaussian distribution ( Figure 3B and 3C ) . Because the lognormal distribution has a “heavy tail , ” most spikes were generated by just a few neurons: About 16% of neurons—the subset of 23 neurons firing at higher than 9 . 5 sp/s—accounted for 50% of all spikes . The lognormal distribution fit better than the exponential distribution , particularly at low firing rates ( Figure 4 ) ; because we were using cell-attached recording , we were confident that we were not undersampling the low-firing end of the distribution and that therefore this improved fit was real . Although lognormal distributions have widely been used to describe the ISI distributions from a single neuron , population responses are usually reported to be exponentially distributed [6 , 25 , 26]; this is , to our knowledge , the first report that firing rates across a population of neurons are lognormally distributed . What is the typical response across the entire neuronal population to a particular stimulus ? Figure 5 shows the cumulative distribution of firing rate changes ( with respect to baseline ) for each of the stimuli tested . To simplify the interpretation of these cumulative distributions , we defined an arbitrary threshold of 20 sp/s , beyond which we labeled the response as “well-driven;” Figure 5B shows the fraction of neuronal population exceeding this threshold for each ensemble . The choice of 20 sp/s , which corresponds to only a single extra spike in the 50-ms response bin , we consider was quite conservative; for example , other authors have chosen a higher ( arbitrary ) value of 50 sp/s as the threshold for the “high-firing” regime [27] . The typical stimulus-evoked population response was sparse for all stimulus ensembles tested: tones , sweeps , white-noise bursts , and natural sounds . Only a small fraction—less than 5%—of the population showed a well driven ( >20 sp/s ) response . That is , only a few percent of the neuronal population was likely to fire one or more “extra” action potentials ( above baseline-firing rate ) within any 50-ms window , regardless of the stimulus ( see also Figure S3 ) . Such a sparse response might seem incompatible with evidence from other recording technologies , such as multiunit recordings ( or fMRI ) , indicating that tones and other stimuli can indeed elicit substantial increases in population activity . However , these lines of evidence can be reconciled: even though only a small fraction of neurons was highly active at any instant , the activity in this small fraction could lead to as much as a 50% increase in the mean ( as opposed to the median ) firing rate ( e . g . , 4 . 9 versus 7 . 0 sp/s during the early tone epoch; see Table 1 ) . Thus the presentation of a stimulus caused only a barely discernible change in the activity of most of the population; but an appreciable number of extra spikes were concentrated in a small fraction of neurons . Although the distributions of firing rate changes were similar across stimulus ensembles , we did observe minor differences in the mean driven rate among ensembles . As expected , louder tones elicited a greater population response than quiet tones . Perhaps surprisingly , simple spectrotemporally rich stimuli ( sweeps ) elicited a somewhat greater response than did complex stimuli ( natural sounds ) . Nevertheless , these differences were relatively small , and qualitatively the ensembles elicited similarly sparse population responses . The analysis so far has focused on the fraction of well-driven neurons . What fraction of neurons showed any detectable stimulus-locked change at all , beyond those predicted by chance fluctuations around the baseline ? Note that the definition we use here for responsiveness is quite inclusive: Even if a tone elicited only a 1 sp/s increase over the baseline-firing rate , this response might still be deemed responsive if the spontaneous rate was sufficiently low for us to detect a change . The majority of neurons showed no discernible response to any stimulus during any given response epoch . Figure 6 shows the evoked population response to each stimulus ensemble ( with the exception of natural stimuli and white-noise bursts , for which not enough repetitions were presented; see Materials and Methods ) . During each 50-ms response epoch only about 10% of neurons showed any significant stimulus-locked increase in firing rate ( Figure 6 top , Inc ) , and a smaller fraction showed a significant stimulus-locked decrease ( Figure 6 top , Dec ) . Thus , not only was the fraction of well-driven neurons low , the fraction of neurons driven at all was also low . Half of the cells ( 50% ) did not show any significant change ( increase or decrease ) in firing rate during any response epoch , to any stimulus; an example of such an unresponsive neuron was shown in Figure 2H . At the other extreme , a few broadly tuned cells showed significant changes in firing rate in all ( four or five ) octave bins ( i . e . , across the whole frequency space tested ) for at least one of the response periods . It might appear that the sparseness we report is incompatible with the broad frequency tuning of rat auditory cortical neurons . However , we found that sparseness was not achieved through narrow frequency tuning . Instead , it arose through a combination of factors . First , 50% of the neural population failed to respond to any of the simple stimuli we presented . Second , responses were often brief; in many neurons , the change in firing rate was limited to just one of the three response epochs . Thus , sparseness of the response in time contributed to the overall sparseness of the population response . Finally , even when changes occurred they were typically small; the increase in firing rate exceeded 20 sp/s in only about a quarter of the statistically significant responses . As a result , only a small fraction of neurons responded vigorously to any tone even though frequency tuning was broad . The form of sparseness we report has sometimes been termed “population sparseness , ” to distinguish it from “lifetime sparseness” [2 , 28] . Lifetime sparseness refers to the selectivity of a single neuron probed with different stimuli and can be assessed for a single neuron during a single unit experiment . Population sparseness refers to the response of the population to a given stimulus . Responses in visual cortex have been reported to show population sparseness [29] , but population sparseness has not previously been assessed in auditory cortex . The heterogeneity of response patterns to simple tones led us to wonder whether neurons with similar properties might be clustered into nearby regions of the cortex; for example , neurons with predominantly transient responses might be found in one region , and sustained neurons might be found in another . In some cases , therefore , we recorded from multiple cells in a single electrode penetration . Since the recording electrodes were aligned approximately perpendicular to the cortical surface , the cells recorded in a single electrode penetration likely belonged to the same or neighboring cortical column . We did not detect any clustering of response patterns; highly responsive cells were often very near unresponsive cells . Figure 7A shows an example with five neurons recorded over two penetrations ( three in one penetration , and two more in a penetration approximately 50–100 μm ventro-caudal from the first penetration ) . In the first penetration , one neuron was unresponsive , one showed suppression over a wide range of frequencies , and the third showed enhanced firing over an even wider range of frequencies . In the second penetration , both neurons were unresponsive . The fact that unresponsive neurons were often mixed closely with responsive neurons indicates that unresponsiveness need not indicate gross cortical damage ( see also [21] ) or recording from a region of cortex that was unresponsive to the stimuli we were presenting , but that instead neurons with different selectivity are commingled . We also wondered whether firing rate was correlated with cortical layer . We segregated neurons ( n = 141 ) recorded at different cortical depths ( Figure 7B , depths were estimated using the micromanipulator readings and as such were only approximate; see Materials and Methods ) into six groups corresponding to the cortical layers [30] ( Figure 7C ) . We compared the firing rates using multiple comparisons based on Kruskal-Wallis test and found that the spontaneous and mean evoked firing rates were not significantly different , with the exception of layer II , which displayed firing rates significantly lower ( p < 0 . 01 ) than the other cortical layers ( layer I contained only one neuron and was not included in the comparisons ) . Thus , cortical layer does not seem to account for the diversity of response properties we observed . Because we could record from only a relatively small number of neurons in a single penetration , we cannot rule out the possibility that more thorough sampling of all the nearby neurons in a region might reveal subtler forms of spatial or laminar organization that escaped our detection . Alternatively or additionally , responsiveness might be correlated with single neuron properties such as type , morphology , and molecular expression pattern . Although in this study we did not recover neurons for histological analysis and so could not assess whether there was a correlation with morphology or molecular expression pattern , we did attempt to correlate responsiveness with cell type . Cortical neurons can be grouped into two broad classes: excitatory neurons that release glutamate at their synapses; and inhibitory interneurons , which release gamma-aminobutyric acid ( GABA ) . Most cortical neurons are excitatory . GABAergic neurons can have diverse morphological , physiological , or molecular characteristics [31] . Excitatory and inhibitory neurons can also be distinguished based on a variety of physiological parameters [32 , 33] . In particular , the firing rate of some inhibitory interneurons—the so-called fast-spiking subtype—is higher when stimulated by current injection . Spike width and shape have been used in previous studies to assign spikes recorded extracellularly in vivo to putative excitatory and inhibitory neurons in hippocampus [34] , and cortex [35] . We therefore asked whether spike shape might predict response patterns in our sample . Based on previous studies ( for example , [32 , 33] ) , we expected that fast-spiking interneurons would likely have narrow and symmetric spikes . For each cell we therefore computed the spike width , and also the “spike amplitude index” as a measure of spike symmetry ( see Materials and Methods ) . For our population of cells the spike widths ranged from 0 . 4 ms to 1 . 9 ms , with a median value of 0 . 9 ms . We defined the spike amplitude index as the absolute value of the spike peak-to-valley-ratio . A spike amplitude index of unity indicates a perfectly symmetrical spike , whereas a value greater than unity indicates a tall spike , and a smaller value indicates a spike with a deep valley; a fast-spiking interneuron would be expected to have a low spike amplitude index . Spike amplitude indices ranged from 0 . 8 to 34 . 3 , with a median value of 2 . 0 . Neurons with higher evoked firing rates tended to have narrower spikes ( Figure 8A ) , suggesting that interneurons were overrepresented among the most responsive neurons . Indeed , the seven most responsive neurons—those with a mean evoked firing rate ( computed across all octave bins and response epochs ) higher than 20 sp/s—had narrow spikes with spike widths less than or equal to 0 . 9 ms . Neurons with high firing rates also tended to have symmetrical spikes ( Figure 8B ) . Although spike width and shape are only at best crude surrogates for cell type , the striking correlation between these quantities and tone responsiveness suggest that a substantial fraction of the most responsive neurons may be interneurons .
Cell-attached recording differs from conventional extracellular recording methods—especially from tungsten recordings with single electrode—in its selection bias ( see also [10] ) . In conventional recording , single-unit isolation requires neural activity and neurons with low firing rates—spontaneous or evoked—tend to be undersampled . During patch clamp recording , by contrast , the formation of an electrical seal does not require neuronal activity and the tip of a glass patch pipette is in physical contact with the neuron , so even neurons with very low firing rates are as likely to be included in the sample as those with high firing rates . Although lognormal distributions have been used widely to describe the interspike interval distributions from a single neuron , population responses have usually reported to be exponentially distributed [6 , 12 , 25 , 26] . The exponential and lognormal distributions differ most dramatically at the low end: a lognormally distributed population has fewer nearly silent neurons ( e . g . , neurons with a firing rate lower than 0 . 1 sp/s ) than an exponential population . However , because the cell-attached recording method that we used is not biased away from such nearly silent neurons , we could be confident that their underrepresentation in the population was not due to experimental undersampling . It would be of interest to see whether a lognormal distribution of firing rates is seen in neuronal datasets obtained using different recording techniques with similar recording biases , such as optical [36] , tetrodes [37] , or silicone probes [35] . Interestingly , lognormal distributions have recently been reported in another neurobiological context . The distribution of synaptic weights also follows lognormal distribution [38] . It is , however , unclear how these two observations are related and what mechanisms might give rise to such distributions of synaptic weights and firing rates . The rodent auditory cortex is highly organized , consisting of several auditory fields [39 , 40] . In several areas , including the area in which the present experiments were conducted ( area A1 , the primary auditory cortex ) , responses are organized tonotopically , meaning that neurons in a particular region tend to be tuned to similar frequencies [22] . Tonotopy represents the coarsest level of organization within an area , analogous to retinotopy in the primary visual cortex [41] . However , in the visual cortex of cats and primates , as well as some other cortical areas , neurons are further organized into columns , implying that neurons recorded in a single electrode penetration have similar response properties [42] . In our experiments , however , we failed to detect any organization beyond tonotopy . For example , nearly silent neurons could be situated very nearby responsive neurons . Thus we did not find a columnar organization of response dynamics . Although our failure to find columnar organization is not definitive evidence that no such organization exists—absence of evidence is not evidence of absence—it is consistent with several indirect lines of evidence suggesting that columnar organization in the rodent auditory cortex may be weak . First , recent studies suggest that neurons in the visual cortex of rodents , unlike those in cats and primates , may not be organized into columns [36 , 43]; by analogy , it may be the auditory cortex of the rodent also lacks columnar organization . Second , in vitro experiments with acute rodent cortical slices suggest that local columnar connections may be weaker in auditory cortex than in the barrel cortex [44] . Thus it may be the functional microcircuitry of rodent auditory cortex is organized in a more subtle fashion [45] . We correlated neuronal responsiveness with cell type based on electrophysiological criteria . We computed spike width and spike amplitude index ( as a measure of symmetry of spike waveform amplitude ) , expecting fast-spiking cells ( likely GABAergic interneurons [33 , 46] ) to have narrow and symmetrical spikes due to the fast repolarization [33 , 46 , 47] . Multiple features of spike waveforms seem to be required to classify a given cell as a pyramidal cell or an interneuron [35 , 48] , with narrow-spiking cells usually considered to be interneurons . However , the presence of pyramidal cells with narrow spikes [49] , and the overall complexity of various physiological and morphological features of interneurons [31 , 32 , 50] , further complicate electrophysiological identification of interneurons . Although definitive identification of interneurons requires other techniques such as morphological reconstruction , it is likely that majority of highly responsive cells in our sample were not excitatory pyramidal neurons . We speculate that the high responsiveness of inhibitory interneurons might contribute to population sparseness of stimulus-evoked responses by simply inhibiting responses of pyramidal neurons in the auditory cortex . Such inhibition could then lead to sparse communication between the primary auditory cortex and higher sensory cortical areas in awake animals . The sparse and heterogeneous responses we report are consistent with many previous single-unit studies of auditory cortex in unanesthetized animals , including the classical studies [21]; see also [51–55] . In many anesthetized preparations ( e . g . , barbiturate and ketamine ) , sound-evoked responses are typically transient [17 , 39 , 56 , 57] . With the resurgence of work in the awake preparation in the last decade , many researchers have emphasized the much richer repertoire of responses in the auditory cortex of awake animals , including especially sustained responses to sounds [25 , 58–60] . We propose that response heterogeneity is a hallmark of awake auditory cortex . Our study complements recent work aimed at identifying “optimal” stimuli—stimuli that elicit high sustained firing rate [12 , 13 , 61] . The fact that a stimulus can be optimized to drive a particular neuron well tells us little about how this stimulus is represented across the population . Our data suggest that only a minority of neurons are engaged in the representation of many stimuli; indeed , the fact that most stimuli drive most neurons only weakly explains why finding the optimal stimulus for any given neuron can be such a challenge . Thus , although there may be an optimal stimulus for any given neuron , most stimuli are not optimal for most neurons , and so are represented sparsely across the population . The population sparseness in the awake auditory cortex we described arose through a combination of three factors . First , half of neurons failed to respond to any tone we presented . Second , responses were often brief . Third , the amplitude of responses was usually low . Thus , even though the frequency tuning of single neurons is usually broad [24] , only a small fraction of neurons responded vigorously and most neurons were silent . Experimental evidence for sparse coding has been found in a range of experimental preparations , including the visual [5 , 6] , motor [11] , barrel [10] , and olfactory systems [7 , 62 , 63] , the zebra finch auditory system [8] , and cat lateral geniculate nucleus [9] . However , the sparseness of representations in the auditory cortex has not been explicitly addressed in previous work . Our results constitute the first direct evidence that the representation of sounds in the auditory cortex of unanesthetized animals is sparse . Our data support the “efficient coding hypothesis , ” [64] according to which the goal of sensory processing is to construct an efficient representation of the sensory environment . Sparse codes can provide efficient representations for natural scenes [2 , 65] . Sparse representations may also offer energy efficient coding , where fewer spikes are required compared to dense representations [66–68] . A growing body of theoretical work on sparse representations suggest they may be useful for computation [2–4 , 65 , 69 , 70] . Sparse representations have become increasingly important in statistical machine learning [71] . One intuition underlying this approach is that it can be easier to recognize a sparse pattern in a high-dimensional space than a dense pattern in a low dimensional space . This is illustrated in Text S1 and Figure S1 , where spike trains drawn from a sparse distribution could more easily be discriminated than those drawn from a dense distribution . This discriminability in turn can make the patterns easier to learn rapidly ( see Text S2 and Figure S2 ) . Thus , an advantage of sparse cortical representations may be to facilitate rapid learning of arbitrary auditory patterns .
Sprague Dawley rats ( 21–30 days old ) were anesthetized in strict accordance with the National Institutes of Health guidelines , as approved by the Cold Spring Harbor Laboratory Animal Care and Use Committee . A small craniotomy ( maximum size of 1 . 5 × 1 . 5 mm ) and durotomy were performed over the left ( primary ) auditory cortex . The position of the craniotomy was determined by its distance from bregma ( 4 . 5 mm posterior and 4 mm lateral ) , and its relationship to other bone sutures . The presence of clear auditory single-unit responses and/or local field potentials was further used as physiological criteria to confirm the location of the auditory cortex . Based on the anatomical landmarks and physiological criteria we expect that the neurons recorded in this study were in the primary auditory cortex [39] . The whole area was protected by a plastic well with removable cap . The brain surface was covered with Kwik-Cast ( World Precision Instruments ) between the recording sessions . An aluminum headpost was attached to the skull with Relyx Luting Cement ( 3M ESPE ) . A silver chloride ground wire was implanted subcutaneously on the back of the animal . The animals were allowed at least 24 h of recovery before the first recording session . During the recording session , the head of the animal was fixed in the headpost holder and the animal was positioned inside a plastic tube , which provided a loose restraint for body movements . The plastic cap and Kwik-Cast were removed and the cortex covered with physiological buffer ( in mM: NaCl , 127; Na2CO3 , 25; NaH2PO4 , 1 . 25; KCl , 2 . 5; MgCl2 , 1; and glucose , 25 ) mixed with 1 . 5% agar . The animals sat quietly , occasionally moved their limbs , groomed , whisked , etc . The behavioral state of the animal was monitored by a closed video circuit . Excessive movement , signs of stress , or discomfort of the animal were used to indicate the end of the experiment . We recorded from each animal during several recording sessions ( usually two or three sessions per rat ) . The number of recording sessions was limited by the total number of electrode penetrations . Any appearance of brain edema , or a change in cortex appearance , vasculature , etc . was a sign to discontinue recordings from the animal . Cell-attached recordings were obtained using standard blind patch-clamp recording techniques; for details on this technique see also [17 , 72 , 73] . Electrodes were pulled from filamented , thin-walled , borosilicate glass ( outer diameter , 1 . 5 mm; inner diameter , 1 . 17 mm; World Precision Instruments ) on a vertical two-stage puller ( Narishige ) . Internal solution contained ( in mM ) : KCl , 10; KGluconate , 140; HEPES , 10; MgCl2 , 2; CaCl2 , 0 . 05; Mg-ATP , 4; Na2-GTP , 0 . 4; Na2-Phosphocreatine , 10; BAPTA , 10; and biocytin , 1 % , ( pH 7 . 25 ) ; diluted to 290 mOsm . Resistance to bath was 3 . 5–5 . 0 MΩ before seal formation . Recordings were obtained using Axopatch 200B ( Axon Instruments ) with 2 kHz lowpass filter and a custom data acquisition system written in MATLAB ( Mathworks ) , with a sampling rate of either 4 kHz or 10 kHz . Because cell-attached recording requires a minimum seal of only about 10–20 MΩ ( compared with the >1 GΩ for whole cell recording ) , almost every neuron encountered can be patched . We recorded from 166 neurons ( in 25 animals ) , out of which we identified 145 neurons ( in 24 animals ) with at least eight trials per octave bin ( see also section titled Cell counts , below ) . The search for neurons was conducted solely based on pipette's resistance and not on spiking activity . For inclusion in our sample , each cell had to generate at least one action potential ( to guarantee that it was not , e . g . , a glial cell ) . A great care was taken to exclude neurons that might have been damaged by direct contact between the pipette tip and cellular membrane . Recordings during which we observed gradual increase in spontaneous firing rate were excluded . In the rare cases , in which the spontaneous rate increased suddenly , or the electrode “broke in” after a sudden movement of the animal , we analyzed only the first stationary epoch of the recording . Neurons were recorded from all depths ( Figure 7B ) . The neuron appearing in Figure 2A appeared previously in a review article ( Figure 2B of [1] ) , as did the neuron presented in Figure 2C ( Figure 3 of [1] ) . All experiments were conducted in a double-walled sound booth ( Industrial Acoustics Company ) . Free-field stimuli were presented at 97 . 656 kHz using TDT System 3 ( Tucker-Davis Technologies ) connected to an amplifier ( Stax SRM 313 , STAX Limited ) , which drove a calibrated electrostatic speaker ( taken from the left side of a pair of Stax SR303 headphones ) located 8 cm lateral to , and facing , the contralateral ( right ) ear . The main sets of stimuli consisted of 100-ms long pure-tone pips of 16 , 20 , or 64 different frequencies logarithmically spaced between 1–40 kHz ( 81% of recordings , 134 out of 166 ) presented at either 20 , 50 , 80 dB SPL ( n = 43 ) , or at 0 , 30 , 60 dB SPL ( n = 15 ) , or at 0 , 20 , 40 , 60 dB SPL ( n = 76 ) . For the rest of recordings ( 19% , 32 out of 166 ) the stimulus protocol contained 100-ms long pure-tone pips of 28 frequencies logarithmically spaced between 2–48 kHz presented at 60 dB SPL . All tones were repeatedly presented in a fixed pseudo-random order at a rate of two tones per second . A full tuning curve was obtained for each neuron . In 22 neurons ( 13% of recordings , 22 out of 166 ) we also presented frequency-modulated sweep stimuli . Sweeps covered the frequency range from 1 to 40 kHz , and both upward ( from 1 to 40 kHz ) and downward ( from 40 to 1 kHz ) going sweeps were presented at six different rates ( 25 , 50 , 75 , 100 , 125 , 150 octaves/second ) for each neuron ( see Figure 2I for an example ) . In 43 neurons ( 26% of recordings , 23 out of 166 ) we presented 100-ms long white-noise bursts at 80 dB SPL . Natural sound stimuli were presented for 28 neurons . Of those , 23 neurons were also presented with pure tones ( 14% of recordings , 23 out of 166 ) , and five neurons were presented only with natural sounds . The natural sound stimuli were taken from a commercially available audio compact disc , “The Diversity of Animal Sounds” ( Cornell Laboratory of Ornithology ) , originally sampled at 44 . 1 kHz and resampled at 97 . 656 kHz for stimulus presentation [72] . The sounds chosen had no special relevance to the rats ( unlike , e . g . , rat pup calls ) , and therefore are less likely to engage specialized processing mechanisms; to the extent that these sounds are representative of the acoustic environment of humans , they are also representative for rats , which often share the same habitat as humans . Altogether , four natural sound segments were presented for each neuron , with at least four repeats of each segment per neuron . The segments included Jaguar call ( track 3 , seconds 2 to 11 for total duration of 10 s ) , Bowhead Whale ( track 9 , seconds 1 to 10 , 10-s duration ) , Knudsen's Frog ( track 11 , seconds 1 to 10 , 10-s duration , Figure 2J ) , and Bearded Manakin ( track 19 , seconds 0 . 1 to 5 . 1 , 5-s duration ) . The peak amplitude of each segment was normalized to the ±10 V range of the TDT system , which corresponded to 80 dB SPL . Spikes recorded in cell-attached mode were extracted from raw voltage traces by applying a high-pass filter and thresholding ( Figure 1A ) . Spike times were then assigned to the peaks of suprathreshold segments , and rounded to the nearest millisecond . Individual spikes can assume very different shapes even in a single cell ( Figure 1B ) . In some cases we observed bursts of spikes , during which spike amplitude sometimes decreased several-fold . For the cell shown in Figure 1B , both single spikes and bursts were sometimes evoked approximately 40 ms following tone termination . Such large changes in spike characteristics can result in a failure of spike detection in conventional extracellular tungsten recordings . Spikes were recorded at a sampling rate of 4 kHz for 88 neurons in our sample ( n = 166 neurons ) , and 10 kHz for the remainder of the population . For the analysis of spike shape ( Figure 8 ) , the spike waveforms recorded at 4 kHz were resampled to 10 kHz ( using MATLAB resample function ) . We then computed the mean spike waveform , and defined spike width as the time difference between the peak ( maximum amplitude ) and valley ( minimum amplitude following the peak ) of the waveform . Because the spike waveforms are ( re ) sampled at 10 kHz , the spike widths are rounded to the nearest tenth of a millisecond . For each cell we also computed the amplitude index , the absolute value of peak-to-valley-ratio , of the mean spike waveform . Responses to stimuli were divided into 50-ms-duration time bins . In addition , tone-evoked responses were also binned in frequency space . We use the term response bin to refer to subdivision of a response in general , as defined below for various stimuli . When we explicitly refer to binning in time , or frequency , we use the terms response epoch , or octave bin , respectively . Tone-evoked responses were divided into four 50-ms- long response epochs ( Figure 3A ) . The spontaneous epoch was defined as the 50-ms-long period preceding stimulus onset . The early epoch was defined as the first 50 ms of stimulus duration , the late epoch as the last 50 ms of stimulus duration , and the off epoch as the first 50 ms after stimulus termination . In frequency space the responses were grouped into one-octave-wide bins , which resulted in four or five frequency bins ( octave bins ) per cell ( depending on the stimulus protocol used , see Stimuli above ) . The spontaneous firing rate for each cell was computed as a mean of firing rates across all trials in the spontaneous epoch for the given cell . Evoked firing rates were computed for each combination of response epoch and octave bin as a mean of firing rates of all trials in the specific octave-epoch combination ( Figure 3A ) . The distribution of firing rates across octave bins for each response epoch was fit with a lognormal distribution ( Figure 3 ) . To fit each distribution , the octave bins with zero firing rate were removed , and the mean and variance of the distribution of log-transformed firing rates were computed . The mean and variance obtained directly from data were then used as parameters for the normal distribution fit to log-transformed firing rates . The goodness-of-fit for each distribution was assessed using the Kolmogorov-Smirnov test . The significance of stimulus-evoked changes in firing rates was evaluated with the Wilcoxon signed-rank test , i . e . , a non-parametric paired , two-sided test of the hypothesis that the difference in firing rates between the matched trials in two different epochs comes from a distribution whose median is zero . For each octave and early , late , and off response epochs we tested on a trial-by-trial basis whether the stimulus-evoked firing rate increased or decreased significantly compared to the corresponding spontaneous epoch . For this test we also only considered cells with at least 20 trials per octave bin ( 69% , 100 cells out of 145 ) . For the analysis of responsiveness of single neurons ( in Results see Population Response Is Sparse ) the evaluation of significance involved 15 comparisons for most of the cells , because responses of most cells were binned to 15 response bins ( five octave bins times three response epochs ) . Therefore , we used a significance criterion of either p < 0 . 0033 ( for 15 comparisons , 0 . 05/15 ) , or p < 0 . 0042 ( for 12 comparisons , 0 . 05/12 ) to keep the overall significance criterion for each cell at p < 0 . 05 . To be considered tone-responsive , a cell had to show a significant change in firing rate ( increase or decrease ) in at least one response bin . For the population response analysis ( in Results see Population Response Is Sparse and Figure 6 ) the response bins from all neurons were considered independent and their responsiveness was evaluated with the Wilcoxon signed-rank test using a significance criterion of p < 0 . 01 . To evaluate the population response in the early response epoch ( Figure 6A ) , the fraction of bins showing a significant increase , a significant decrease , or no change in the firing rate was computed for each octave bin in the early response epoch . The fraction of responsive bins in the early response epoch was then defined as the mean of the octave-bin fractions in the epoch . Analogous computations were carried out for the late and off response epochs . To compute the population response across all epochs the fractions of responsive bins were computed from all response bins ( from all neurons ) pooled together . Careful inspection revealed no clear examples of frequency tuning sharper than about one octave , suggesting that it would be appropriate to pool together responses to tones within an octave . To confirm systematically that our results were robust to this choice we repeated this analysis with half-octave wide ( i . e . , narrower ) frequency bins , two , and four octaves wide ( i . e . , wider ) frequency bins , and 50-ms-long response epochs . To control for neurons with more transient or sustained responses we performed the population response analysis with 25 , 75 , and 100 ms duration response epochs and one-octave-wide frequency bins . The results of these analyses , however , were the same as for the basic analysis with one-octave-wide frequency bins and 50-ms-duration response epochs ( unpublished analysis ) . Responses to frequency-modulated ( FM ) sweeps were subdivided to 50-ms-duration response epochs . Slower sweeps , with 25 or 50 octaves/second , contained four or two 50-ms epochs , respectively , during the stimulus presentation . Faster sweeps ( 75 , 100 , 125 , 150 octaves/second ) contained one 50-ms epoch . For all sweep rates we also added an off epoch starting either at the sweep termination ( for 25 , 50 , 75 , 100 octaves/second ) , or immediately after the response epoch ( for 125 , 150 octaves/second ) . Each response was thus divided into 32 response bins ( including upward and downward moving sweeps ) . For the analysis of significance of sweep-evoked responses in individual neurons , we therefore used a significance criterion of p < 0 . 0016 ( 0 . 05/32 ) . To compute the population response to FM-sweeps all response bins were considered statistically independent , and their responsiveness was computed using a significance criterion of p < 0 . 01 . Responses to 80-dB white-noise bursts were divided into four 50-ms-long response epochs ( spontaneous , early , late , and off ) analogous to the tone response epochs described above . Responses to natural sounds were also divided to 50-ms- duration response epochs . 10-s-long segments thus contained 200 response bins each , and 5-s-long segments contained 100 response bins . Natural sound-evoked responses were used only for the analysis of stimulus-evoked changes in firing rate , because none of the recordings met our criterion for the test of evoked response significance ( i . e . , at least 20 trials per response bin ) . For the analysis of stimulus-evoked changes in firing rate ( Table 1 , Figures 4 , 5 , and 7C ) , we identified neurons with at least eight trials per response bin ( five trials for natural sounds , three trials for white-noise bursts ) . For the analysis of significance of stimulus-evoked responses ( Figure 6 ) , we identified neurons with at least 20 trials per response bin . We recorded from 166 neurons ( 100% ) , while presenting pure-tone pips . For further analysis of firing rates evoked by 50- or 60-dB tones we identified 145 neurons ( 87% ) with at least eight trials per response bin . For the analysis of evoked response significance we further identified a subset of 100 neurons ( 60% ) with at least 20 trials per response bin . For 91 neurons ( 55% ) we also presented 30- or 40-dB tones . All of these neurons were used for the firing rate analysis , and 62 neurons ( 37% ) from this subset—those with at least 20 trials per response bin—were used for the analysis of evoked response significance . Accordingly , out of 43 neurons ( 26% ) presented with 80-dB tones we selected 22 ( 13% ) for firing rates analysis , and six ( 4% ) with at least 20 trials per octave bin for the analysis of evoked response significance . FM sweeps were presented for 22 neurons , all of which were used for the firing rates analysis . Seventeen neurons with at least 20 trials for each sweep rate and direction were further selected for the analysis of significance of sweep-evoked responses . White-noise bursts ( 80 dB ) were presented for 43 neurons . For the analysis of evoked firing rates we identified 23 neurons ( 55% ) with at least three trials per response bin . Natural sounds were presented for 28 neurons . Twenty-seven neurons with at least five trials for each natural sound segment were identified for the analysis of stimulus-evoked firing rates . Bootstrap resampling showed that the smaller sample size did not influence our estimates of fraction of well-driven response bins ( see Results . )
|
How do neuronal populations in the auditory cortex represent sounds ? Although sound-evoked neural responses in the anesthetized auditory cortex are mainly transient , recent experiments in the unanesthetized preparation have emphasized subpopulations with other response properties . We quantified the relative contributions of these different subpopulations in the auditory cortex of awake head-fixed rats . We recorded neuronal activity using cell-attached recordings with a glass electrode—a method for which isolation of individual neurons does not depend on neuronal activity—while probing neurons with a representative ensemble of sounds . Our data therefore address the question: What is the typical response to a particular stimulus ? We find that the population response is sparse , with sounds typically eliciting high activity in less than 5% of neurons at any instant . The overall population response was well described by a lognormal distribution , rather than the exponential distribution that is often reported . Our results represent , to our knowledge , the first quantitative evidence for sparse representations of sounds in the unanesthetized auditory cortex . These results are compatible with a model in which most neurons are silent much of the time , and in which representations are composed of small dynamic subsets of highly active neurons .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroscience"
] |
2008
|
Sparse Representation of Sounds in the Unanesthetized Auditory Cortex
|
The developing pancreatic epithelium gives rise to all endocrine and exocrine cells of the mature organ . During organogenesis , the epithelial cells receive essential signals from the overlying mesenchyme . Previous studies , focusing on ex vivo tissue explants or complete knockout mice , have identified an important role for the mesenchyme in regulating the expansion of progenitor cells in the early pancreas epithelium . However , due to the lack of genetic tools directing expression specifically to the mesenchyme , the potential roles of this supporting tissue in vivo , especially in guiding later stages of pancreas organogenesis , have not been elucidated . We employed transgenic tools and fetal surgical techniques to ablate mesenchyme via Cre-mediated mesenchymal expression of Diphtheria Toxin ( DT ) at the onset of pancreas formation , and at later developmental stages via in utero injection of DT into transgenic mice expressing the Diphtheria Toxin receptor ( DTR ) in this tissue . Our results demonstrate that mesenchymal cells regulate pancreatic growth and branching at both early and late developmental stages by supporting proliferation of precursors and differentiated cells , respectively . Interestingly , while cell differentiation was not affected , the expansion of both the endocrine and exocrine compartments was equally impaired . To further elucidate signals required for mesenchymal cell function , we eliminated β-catenin signaling and determined that it is a critical pathway in regulating mesenchyme survival and growth . Our study presents the first in vivo evidence that the embryonic mesenchyme provides critical signals to the epithelium throughout pancreas organogenesis . The findings are novel and relevant as they indicate a critical role for the mesenchyme during late expansion of endocrine and exocrine compartments . In addition , our results provide a molecular mechanism for mesenchymal expansion and survival by identifying β-catenin signaling as an essential mediator of this process . These results have implications for developing strategies to expand pancreas progenitors and β-cells for clinical transplantation .
Organogenesis is a complex and dynamic process that requires tight spatial and temporal regulation of differentiation , proliferation , and morphogenesis . The pancreas serves as an interesting model for the study of these processes as its epithelium gives rise to functionally distinct cells: endocrine cells , including insulin-producing β-cells that release hormones into the blood stream to regulate glucose homeostasis , and exocrine cells that produce , secrete , and transport digestive enzymes . These diverse cell types derive from common progenitors residing in the embryonic pancreatic epithelium through a well-orchestrated multi-step process . While numerous studies have delineated the cascades of transcription factors within the epithelium that guide epithelial cell development ( reviewed in [1] , [2] ) , the role of the surrounding mesenchyme in governing pancreas organogenesis at different stages remains largely unknown . Mesenchymal cells start to coalesce around the nascent gut tube shortly before pancreas epithelial cells evaginate around mouse embryonic day 9 . 5 ( e9 . 5 ) to form the dorsal and ventral buds [1] . At e13 . 5–e14 . 5 Pdx1+ epithelial precursor cells become committed to either the endocrine or the exocrine lineage , and from e15 . 5 until the end of gestation , pancreatic cells undergo final differentiation to give rise to all pancreatic cell types found in the adult organ . The first evidence that mesenchymal cells were required for pancreatic epithelial growth was provided in the 1960s by seminal work by Golosow and Grobstein [3] , in which it was shown that e11 mouse pancreatic epithelium rudiments stripped of their overlying mesenchyme failed to grow in culture . However , further studies addressing the role of the mesenchyme at later stages have been difficult as the expanding pancreas epithelium quickly branches into the surrounding mesenchyme , thus preventing clean physical separation of these two layers after ∼e12 in the mouse . Additionally , while improved culture conditions for organ rudiments mimic embryonic development during early stages quite well [4] , full replication of all in vivo aspects of later pancreas organogenesis have not been achieved ex vivo [5] . As a consequence , studying the role of the mesenchyme at advanced stages of pancreas development using explant systems resulted in controversial findings . A number of studies have shown that while mesenchymal cells have a positive effect on exocrine differentiation and growth in culture , they impair endocrine cell development [6]–[10] . Other studies have observed that close proximity between mesenchyme and epithelium promotes exocrine differentiation , while secreted mesenchymal factors enhance endocrine differentiation over a distance [5] . More recently , a study by Attali and colleagues showed that co-culture of epithelium with mesenchyme promotes the production of insulin-expressing cells , an effect largely due to the expansion of Pdx1+ precursor cells rather than maturation or proliferation of insulin-positive cells [11] . Importantly , endocrine development was highly variable and dependent on the culture conditions such as oxygen levels [11] , further indicating that in vivo manipulation of mesenchymal gene expression is necessary to fully uncover all mesenchymal functions throughout pancreas development . Starting in the 1970s , extensive efforts were made to identify mesenchymal factors responsible for these effects on the epithelial compartment [12] , [13] . A decade ago , Bhushan and colleagues demonstrated that fibroblast growth factor 10 ( Fgf10 ) , expressed by mesenchymal cells from e9 . 5 until e11 . 5 , is essential for pancreas growth and differentiation as it stimulates proliferation of Pdx1-expressing precursor cells [14] . Since then , germ-line knock out mouse lines , genetically manipulated zebra fish , and transfected chick embryos have been used to study a limited number of additional mesenchymal signaling pathways for their role in guiding pancreas formation ( summarized in [1] ) . These studies provided evidence for Retinoic Acid ( RA ) , Wnt , FGF , BMP , TGFβ , and EGF signaling pathways as important regulators of pancreas formation [1] , [10] , [14]–[17] . However , detailed studies of the requirement for individual mesenchymal factors in pancreas development have been hampered by the lack of transgenic tools that permit manipulation of gene expression specifically in the pancreatic mesenchyme . Here , we present experiments that take advantage of Nkx3 . 2 ( Bapx1 ) -Cre transgenic mice in which Cre-expression is directed to the embryonic pancreatic mesenchyme , but not the epithelium . Using this Cre line in conjunction with mouse lines allowing Diphtheria Toxin ( DT ) induced apoptosis , we depleted mesenchymal cell during various stages of in vivo pancreas development . As expected , elimination of mesenchymal cells at the onset of pancreas development completely blocked pancreas organogenesis . Surprisingly , mesenchymal requirement was not restricted to this early stage , as ablation at later developmental stages also led to severe epithelial hypoplasia , reduced branching , and impaired β-cell and exocrine cell expansion . To elucidate the signaling pathways essential for mesenchyme function , we eliminated canonical Wnt signaling from the tissue . Loss of Wnt signaling within the mesenchyme resulted in mesenchymal cell ablation—subsequently leading to reduction in both exocrine and endocrine cell mass . Summarily , our results demonstrate that the pancreatic epithelium depends on mesenchymal signals for proper expansion and morphogenesis throughout development .
In order to manipulate gene expression in pancreatic mesenchyme , but not epithelium , we looked for genes whose expression matches this pattern . Previous studies pointed to the homeobox gene Nkx3 . 2 ( also known as Bapx1 ) , whose expression was found in the forming somites as well as in the mesenchyme of developing pancreas , stomach , and gut [18]–[23] . In contrast , Nkx3 . 2 expression was not detected in endodermally derived cells in these tissues [18] , [20] , [23] . In the pancreatic mesenchyme Nkx3 . 2 is expressed as early as e9 . 5 , and by e12 . 5 its expression becomes restricted to the mesenchymal area , which will give rise to the splenic bud [18]–[20] , [23] . An Nkx3 . 2 ( Bapx1 ) -Cre line , in which one copy of the endogenous Nkx3 . 2 gene was replaced by a transgene encoding the Cre recombinase , had previously been generated [24] , [25] . This transgenic mouse line faithfully replicates the endogenous expression of Nkx3 . 2 and directs Cre activity to the foregut mesenchyme and skeletal somites starting at e9 . 5 [24] . Given that pancreatic expression of the Nkx3 . 2-Cre transgene was not thoroughly analyzed in prior studies , we first crossed the transgenic mice to two reporter strains , the R26-LacZf/+ and the R26-YFPf/+ lines , which express LacZ or YFP , respectively , upon Cre-mediated recombination . YFP expression in Nkx3 . 2-Cre;R26-YFPf/+ embryos ( from here on referred to as Nkx3 . 2/YFP ) was not found in the endodermally-derived pancreatic epithelium marked by E-Cadherin and Pdx1 at e9 . 5 [26] , but was ubiquitously detected in the surrounding mesenchyme ( Figure 1A ) . Similarly , X-gal staining in Nkx3 . 2-Cre;R26-LacZf/+ ( from here on referred to as Nkx3 . 2/LacZ ) indicated LacZ expression was confined to the surrounding mesenchyme at e11 . 5 ( Figure 1B ) . At p0 , Nkx3 . 2/LacZ and Nkx3 . 2/YFP expressing cells with fibroblast-like morphology were observed around islets , ducts , and blood vessels ( Figure 1C , C′ , D ) . Importantly , we could not detect reporter genes' expression in either epithelial ( Figure 1C , C′ , D ) , endothelial , or neuronal cells ( Figure S1 ) , indicating that Nkx3 . 2-Cre activity is excluded from those compartments throughout pancreatic development . Thus , the Nkx3 . 2-Cre line directs Cre-activity exclusively to the mesenchyme during pancreas development and serves as a novel tool to specifically manipulate embryonic gene expression in this tissue . General histological analysis implied that the relative proportion of mesenchyme to epithelium shifts during pancreas organogenesis as epithelial cell numbers expand . We therefore took advantage of Nkx3 . 2-Cre transgenic mice to quantify the mesenchymal area during different developmental stages . By measuring the percentage of the pancreatic area marked by Nkx3 . 2/LacZ and Nkx3 . 2/YFP cells at various developmental stages , we determined that while the relative mesenchymal area is significantly reduced during pancreas organogenesis , it still comprised 11% and 6% of the pancreatic area at e15 . 5 and e18 . 5 , respectively ( Figure 1E ) . Thus , although there is a dramatic reduction in their portion over time , embryonic mesenchymal cells are present throughout pancreas organogenesis . Next , we tested the requirement for mesenchyme during pancreas organogenesis in vivo . Studies using cultured pancreatic rudiments as well as Fgf10 knockout mice demonstrated a crucial role for the mesenchyme in expanding the pool of epithelial pancreatic precursor cells at early developmental stages ( e9 . 5–e11 . 5 ) [3] , [14] , [27] , [28] . In order to determine the role of the mesenchyme during pancreas development in vivo , we decided to ablate this tissue by employing transgenic mice carrying the Diphtheria Toxin ( DT ) active A subunit ( DTA ) flanked by flox sites ( R26-eGFP-DTA mice [29] , from here on referred to as DTA ) . Upon Cre-mediated recombination , the DTA produced by the transgene inhibits protein synthesis , resulting in rapid apoptosis of Cre-positive cells within less than 24 h [29] . Given that Nkx3 . 2-Cre is expressed in mesenchymal cells surrounding the pancreas from the time organ morphogenesis is initiated ( e9 . 5 , Figure 1A ) , Nkx3 . 2-Cre;DTA embryos permit the study of mesenchymal requirement at early stages of pancreas development ( illustrated in Figure 2A ) . We first analyzed potential defects in e10 . 5 embryos . At this stage , Nkx3 . 2-Cre;DTA embryos presented with Pdx1+E-Cadherin+ epithelial pancreatic cells ( Figure 2B , C ) . However , while non-transgenic control pancreatic epithelial cells were completely surrounded by E-Cadherin− mesenchymal cells , Nkx3 . 2-Cre;DTA embryos lacked most of the adjacent mesenchymal cell layer ( Figure 2B , C ) . To assess potential defects in pancreatic bud morphology , we performed whole mount staining with the epithelial marker E-Cadherin . In wild-type embryos this staining revealed the expected organization of stomach , liver , and ventral and dorsal pancreatic buds ( Figure 2D ) . In contrast , pancreatic buds of Nkx3 . 2-Cre;DTA embryos were severely reduced in size and did not evaginate from the foregut epithelium ( Figure 2E ) . Nkx3 . 2-Cre;DTA transgenic mice suffered from embryonic lethality starting at e15 . 5 as well as severe skeletal defects ( Figure 2F , G ) resulting from Nkx3 . 2-Cre activity in the somites [21] , [22] , [24] . Although the few viable Nkx3 . 2-Cre;DTA embryos recovered at e15 . 5 were only slightly smaller than non-transgenic littermates ( Figure 2F , G ) , their gastrointestinal tract was dramatically reduced in size ( Figure 2H , I ) , likely due to Nkx3 . 2-Cre-mediated expression of DTA in the mesenchyme surrounding these tissues [23] , [24] . Notably , while pancreatic tissue was clearly detected in non-transgenic embryos at this stage ( Figure 2H , demarcated by the white line ) , Nkx3 . 2-Cre;DTA embryos had no visible pancreatic tissue ( Figure 2I ) . Histological analysis of gut rudiments confirmed the gross morphology observation and showed only intestine-like tissue in Nkx3 . 2-Cre;DTA embryos , with no discernable stomach , spleen , or pancreatic tissues ( Figure 2J , K ) . Thus , elimination of mesenchyme at the earliest stages of pancreas formation leads to complete agenesis caused by the inability of pancreatic epithelium to evaginate from the forming gut and to expand . Pancreas development is a multistep process during which the epithelium undergoes complex morphological changes while common precursor cells differentiate into the various cells types that form the adult pancreas [30] . To test whether mesenchymal cells play distinct roles during different stages of pancreas development , we depleted the mesenchyme at various time points by injecting DT into developing embryos . Unlike primates , rodent cells lack a high affinity receptor for DT and therefore do not endocytose the toxin [31] . Since DT internalization into the cell cytoplasm is crucial for its ability to trigger the apoptotic machinery , rodent cells are resistant to ectopically administrated DT . However , mouse cells expressing a human DT Receptor ( DTR ) transgene , encoding for the human heparin binding epidermal growth factor ( hbEGF ) , gain sensitivity to DT and are rapidly eliminated upon exposure to the toxin [32] . Prior studies have established that cell specific expression of human DTR in transgenic mice allows the ablation of targeted cells within 6 h following DT administration [33] . By crossing transgenic mice in which DTR expression is activated upon Cre-mediated recombination ( iDTR [34] , from here on referred to as DTR ) with the Nkx3 . 2-Cre mice ( Nkx3 . 2-Cre;DTR ) we were able to specifically ablate the mesenchyme at different embryonic time points during pancreas development upon DT injection . To ensure efficient delivery of DT to the developing pancreas , we injected the agent directly into embryos intraperitoneally ( i . p . ; the experimental procedure is illustrated in Figure 3A and Figure S2A–D ) [35] , [36] . As early as 4 h following DT injection into e13 . 5 Nkx3 . 2-Cre;DTR embryos , we observed an increase in apoptotic mesenchymal cells compared to controls ( Figure S2E , F ) . One day after DT injection we detected only E-Cadherin expressing cells in Nkx3 . 2-Cre;DTR pancreata ( Figure S2G , H ) , strongly indicating that E-Cadherin-negative mesenchymal cells were eliminated . The loss of Nkx3 . 2-Cre;DTR-positive mesenchymal cells was further confirmed by direct staining for human DTR expression ( Figure S2I , J ) . At the end of gestation ( e18 . 5 ) , Nkx3 . 2-Cre;DTR embryos injected with DT at e13 . 5 were viable and appeared grossly normal , with normal body weight ( Figure 3B–D ) . At e18 . 5 the transgenic embryos displayed skeletal dysplasia , gastrointestinal defects , and asplenia ( Figure 3B , C and Figure S3 ) , likely a result of ablation of Nkx3 . 2-Cre expressing cells in these organs , and they died at birth . Therefore , in utero injections of DT into Nkx3 . 2-Cre;DTR do not cause embryonic lethality and permit studying the effects of mesenchyme ablation on epithelial pancreas development during embryogenesis . To elucidate the requirement of mesenchyme at different stages we injected Nkx3 . 2-Cre;DTR embryos and non-transgenic littermates in utero with a single dose of DT at embryonic days 11 . 5 , 12 . 5 , 13 . 5 , 14 . 5 , 15 . 5 , or 16 . 5 ( illustrated in Figure 3E ) . Embryos were then allowed to develop in situ until e18 . 5 when pancreata were dissected and weighed . Surprisingly , both dorsal and ventral pancreatic regions were significantly reduced in size in treated transgenic embryos independent of time of DT administration ( Figure 3F–H ) . The most dramatic reduction in pancreas mass , up to 80% , was observed when transgenic embryos were injected between e11 . 5 and e13 . 5 ( Figure 3H ) . DT injection at later stages , e14 . 5 and e15 . 5 , resulted in an approximately 50% loss of pancreas mass . Notably , mesenchymal elimination as late as e16 . 5 led to a marked reduction in pancreas size to about two-thirds of non-transgenic littermates ( Figure 3H ) . These results demonstrate that the mesenchyme is continuously required for proper pancreas development and organogenesis . Next , we performed an in-depth analysis of pancreas morphogenesis and cell differentiation in transgenic animals in which mesenchyme was depleted mid-way through organogenesis ( DT injections into e13 . 5 Nkx3 . 2-Cre;DTR embryos followed by analysis at e18 . 5; DT e13 . 5→e18 . 5 ) . When compared to normal tissues [37] , mesenchyme-ablated pancreata displayed an abnormal globular morphology . DT-treated Nkx3 . 2-Cre;DTR pancreata were smooth and lacked the typical extension of the left branches ( Figure 4A , B ) as well as the gastric lobe ( Figure 3F , G ) . In addition , DT-treated transgenic pancreata presented with a rounded tail instead of the stereotypical anvil-shaped tail found in non-transgenic controls ( Figure 4A , B ) [37] . Histological analysis further revealed more compacted cellular distribution in mesenchyme-ablated pancreata as compared to control , as shown by severe reduction of typical acellular areas normally found between adjacent lobes ( Figure 4C , D ) . Staining for the endothelial cell marker PECAM1 revealed that pancreatic vasculature , known to be crucial for organ development [38] , [39] , was not overtly disrupted in DT-treated transgenic pancreata ( Figure S4A–D ) . Similarly , Tuj1 ( β-III Tubulin ) -expressing neuronal cells , known to be required for proper endocrine differentiation [40] , [41] , could be found in pancreata of DT-treated Nkx3 . 2-Cre;DTR embryos ( Figure S4E , F ) . Previous in vitro studies have implied that mesenchymal cells may control the differentiation of pancreatic epithelial cells [6] , [9] . In order to study the in vivo effect of the mesenchyme on epithelial cell differentiation , we analyzed pancreatic tissues from DT-treated Nkx3 . 2-Cre;DTR for the expression of exocrine and endocrine markers at the end of gestation ( DT e13 . 5→e18 . 5 ) . Normal expression patterns for both the duct cell marker Mucin1+ and the acinar cell marker Amylase+ in treated transgenic pancreas indicated normal exocrine differentiation ( Figure 4E , F ) . Furthermore , endocrine differentiation was not disturbed by mesenchymal ablation as Insulin , Glucagon , and Somatostatin expressing cells could be detected in transgenic pancreata ( Figure 4G , H ) . Endocrine cells were single-hormone positive , clustered in islet-like structures typical for this developmental stage , and were distributed throughout the pancreas in a normal pattern ( Figure 4C , D , G , H ) . Moreover , β-cells from transgenic embryos expressed the transcription factor MafA ( Figure 4I , J ) , which is critical for full maturation and glucose responsiveness [42] , strongly indicating that mesenchyme ablation does not block their differentiation potential . Thus , while pancreas morphogenesis is impaired upon mesenchyme elimination after the first stages of pancreas formation , differentiation of the major cell types was not blocked . Although each of the specific pancreatic epithelium lineages formed in mesenchyme-depleted pancreata , the dramatic reduction in pancreatic organ size suggests a decrease in the overall number of pancreatic epithelial cells . In order to understand whether mesenchymal ablation affects either endocrine or exocrine mass , Nkx3 . 2-Cre;DTR mice treated with DT at e13 . 5 ( illustrated in Figure 5A ) were analyzed at e18 . 5 for β- and acinar cell masses . Both Insulin+ β-cell and Amylase+ acinar-cell mass were significantly reduced in transgenic mice when compared to non-transgenic littermates ( Figure 5B , C ) , suggesting a requirement for mesenchymal cells during the expansion of both exocrine and endocrine compartments/precursors . The observation that transgenic pancreata maintained a normal acinar to β-cell ratio ( Figure 5D ) indicates that both cell types depend in equal measures on mesenchymal signals for their proliferation . To determine the developmental stage during which mesenchyme ablation affects pancreatic mass , we injected embryos at e13 . 5 and investigated pancreata 2 d later at e15 . 5 ( DT e13 . 5→e15 . 5 ) . At that stage , pancreas mass in transgenic embryos was already reduced by 80% as compared to controls ( Figure 5E ) , similar to the reduction observed in pancreata injected at e13 . 5 and analyzed at e18 . 5 ( Figure 3H ) . Since mesenchymal cells comprise only 11% of pancreatic tissue at e15 . 5 ( Figure 1E ) , the observed reduction in pancreatic weight was likely due to a rapid and significant loss of the epithelial compartment of the organ . Next , we investigated whether cells of either the endocrine or exocrine compartments were already affected in e15 . 5 Nkx3 . 2-Cre;DTR embryos that were DT-treated 2 d before ( i . e . , at e13 . 5 ) . While present in DT-treated transgenic pancreata ( Figure 5F , G ) , the number of cells positive for Neurogenin 3 ( Ngn3 ) , a transcription factor that marks endocrine precursor cells [43] , was significantly reduced compared to littermate control mice ( Figure 5H ) . Similarly , the number of cells expressing Ptf1a , a transcription factor found in exocrine precursor and differentiated acinar cells [44] , was significantly reduced 2-fold in DT e13 . 5→e15 . 5 Nkx3 . 2-Cre;DTR pancreata as compared to non-transgenic controls ( Figure 5I–K ) . Therefore , the reduction in β-cell and acinar cell mass detected at DT e13 . 5→e18 . 5 Nkx3 . 2-Cre;DTR embryos ( Figure 5B , C ) is , at least in part , due to the decreased number of Ngn3+ precursor cells and Ptf1a+ exocrine cells at earlier developmental stages . Since pancreatic growth between e13 . 5 and e15 . 5 relies heavily on proliferation of precursor cells [45] , we next analyzed the effect of mesenchymal depletion on the proliferation rate of these cell populations in e14 . 5 Nkx3 . 2-Cre;DTR embryos treated at e13 . 5 ( DT e13 . 5→e14 . 5 ) . Epithelial tip cells serve as multi-potent progenitors before they become committed to the exocrine lineage around e14 . 5 [44] . Staining these cells , identified as Carboxypeptidase 1 ( Cpa1 ) expressing cells , with an antibody against phosphorylated Histone H3 , a marker of cell proliferation , revealed a 50% reduction in proliferating tip cells in Nkx3 . 2-Cre;DTR embryos compared to controls ( Figure 5L–N ) . In addition , we analyzed proliferation of Sox9 expressing cells , a transcription factor that marks epithelial precursor cells giving rise to exocrine cells as well as to Ngn3+ endocrine precursors [46]–[48] . The percentage of Sox9+ proliferating cells was slightly but significantly smaller in transgenic embryos ( Figure 5O–Q ) . We could not detect apoptotic epithelial cells by TUNEL ( terminal deoxynucleotidyl transferase dUTP biotin nick end labeling ) assays ( unpublished data ) , concluding that depletion of mesenchymal cells affects both endocrine and exocrine mass through reduced proliferative capacity of epithelial progenitor cells rather than their apoptosis . The reduced proliferative potential of progenitor cells at e14 . 5 explained , at least in part , the reduction in pancreas mass in embryos treated with DT at e13 . 5 . However , when mesenchyme was eliminated at e16 . 5 we also observed a significant reduction of about 35% in pancreas mass at e18 . 5 , affecting both the endocrine and exocrine compartments ( DT e16 . 5→e18 . 5; Figures 3H , 6A–D ) . By e16 . 5 , the various pancreatic cell types are committed towards their final differentiation fate and present with many of their mature cell characteristics . Since pancreatic growth at those late stages of development is attributed to proliferation of these differentiated cells [49] , the decrease in pancreatic mass could not be due to reduced proliferation of progenitor cells . While previous studies did not detect effects of mesenchymal cells on β-cell proliferation in culture [11] , in vivo analysis of Nkx3 . 2-Cre;DTR embryos treated with DT at e16 . 5 and analyzed at e17 . 5 revealed decreased proliferative potential of both insulin and amylase expressing cells ( Figure 6E–J ) . In agreement with what we had found at earlier stages , the ratio between Insulin+/Amylase+ areas was not affected in the DT-treated embryos ( Figure 6D ) , suggesting mesenchymal factors have similar effect on cells of these two compartments . Upon determining the requirement for mesenchymal cells to guide epithelial organ formation throughout development , we set out to identify signals and pathways critical for the mesenchymal effects . Canonical Wnt signaling is active in the developing pancreas , and both the mesenchyme and the epithelium express various Wnt ligands and receptors in a dynamic fashion [50] . At e11 . 5 , Wnt signaling is observed in epithelial cells , and its level of activation declines in the following embryonic days [51] , while its activity in the mesenchymal layer has been first reported around e13 . 5 [15] , [52] . In order to directly investigate the role of mesenchymal Wnt signaling in pancreas development , we decided to block this pathway specifically in the mesenchyme by crossing transgenic mice carrying floxed alleles of β-catenin ( βcatf/f ) , an essential mediator of canonical Wnt signaling , with Nkx3 . 2-Cre mice . In addition to its critical role in Wnt signaling , β-catenin has other functions within cells , most notably in maintaining cell-cell interactions as part of a complex with E-Cadherin . However , in pancreatic mesenchymal cells we failed to observe membrane-associated localization of the β-catenin protein ( Figure S5A ) . Therefore , elimination of this gene in Nkx3 . 2-Cre;β-catf/f pancreata is unlikely to perturb cell-cell interactions but should reveal the requirement for β-catenin mediated Wnt signaling in mesenchyme . As expected , elimination of β-catenin did not affect epithelial size at e12 . 5 ( Figure S5B ) prior to the reported onset of mesenchymal Wnt signaling . In contrast , Nkx3 . 2-Cre;β-catf/f pancreata were markedly reduced in size at e15 . 5 and e18 . 5 ( Figure 7A , B ) , indicating that mesenchymal β-catenin signaling is critical for organ formation at later stages . In addition , Nkx3 . 2-Cre;βcatf/f pancreata exhibited aberrant morphology with diminished branching when compared to controls ( Figure 7A , C , D ) . In order to identify the potential effects on pancreatic epithelial development in Nkx3 . 2-Cre;βcatf/f embryos , we stained e18 . 5 knock-out pancreata for various cell markers and assessed acinar- and β-cell mass . All major pancreatic cell types , both of the exocrine ( acinar and duct cells , Figure 7E , F ) and of the endocrine compartments ( α- , β- , and δ-cells , Figure 7G , H ) , were detected in the Nkx3 . 2-Cre;βcatf/f e18 . 5 pancreata . However , both β-cell and acinar-cell mass was significantly reduced in knock-out embryos ( Figure 7I , J ) . Interestingly , the ratio between β- and acinar cells was maintained in Nkx3 . 2-Cre;βcatf/f pancreata ( Figure 7K ) . The Wnt signaling pathway was shown to become activated in the pancreatic mesenchyme around e13 . 5 [15] , [52] . To address whether the reduction in pancreatic mass observed in Nkx3 . 2-Cre;βcatf/f at e15 . 5 and e18 . 5 is due to effects on epithelial growth at earlier stages , we studied epithelial proliferation in these mice at e13 . 5 . At this stage , proliferating Cpa1+ tips cells serve as multipotent pancreatic progenitor cells for both endocrine and exocrine populations [44] . As shown in Figure 7L , the proliferation rate of Cpa1+ cells was significantly lower in Nkx3 . 2-Cre;βcatf/f embryos as compared to controls . Cell death was not apparent as we could not detect apoptotic epithelial cells by TUNEL assays or by staining for cleaved Caspase3 ( unpublished data ) . Therefore , blocking mesenchymal Wnt signaling leads to reduced pancreatic mass by affecting the proliferation capacity of epithelial precursor cells . Wnt signaling is known to regulate cell survival and proliferation [53] . Since pancreata from Nkx3 . 2-Cre;βcatf/f mice phenocopied those from DT-treated Nkx3 . 2-Cre;DTR mice , we wondered whether Wnt signaling is required for mesenchymal cell survival . Indeed , while e13 . 5 pancreatic tissue from wild type embryos contained both E-Cadherin-positive epithelial cells and E-Cadherin-negative mesenchymal cells ( Figure 7M ) , we could detect only E-Cadherin expressing cells in Nkx3 . 2-Cre;βcatf/f tissues ( Figure 7N ) , indicating ablation of the pancreatic mesenchymal layer in transgenic mice . Thus , our results point to mesenchymal Wnt signaling as a critical mediator of mesenchymal cell survival in vivo and therefore of epithelial growth and patterning .
Factors secreted by the pancreas mesenchyme have previously been shown to regulate pancreas organogenesis [6] , including Fgf10 whose function is required for the expansion of common epithelial progenitor cells during early stages of pancreas development [14] . The pancreatic defects we observe in Nkx3 . 2-Cre;DTA embryos are more severe than those previously reported for Fgf10−/− pancreata [14] , a finding likely explained by the absence of mesenchymal cells , and thus reduction of all mesenchymal factors , in transgenic mice . Our results further demonstrate a requirement for mesenchymal cells in promoting proliferation of various epithelial cell types , including precursors and differentiated cells . While it is theoretically possible that these functions are mediated by a limited number of factors throughout all stages of development , the dynamic activation of mesenchymal signaling pathways ( summarized in [1] ) would suggest a more complex interplay of a diverse set of molecules that changes over time . Our findings also suggest that mesenchyme supports proliferation of multiple distinct cell types , even during the same developmental stage . For instance , mesenchyme ablation has similar effects on proliferation of mature acinar and β-cells towards the end of gestation . This observation poses the question as to whether different epithelial cell types rely on the same mesenchymal factor ( s ) for their proliferation , or whether these processes are mediated by distinct signals . Future analysis is required to identify secreted factors expressed by the pancreatic mesenchyme at different developmental stages . The use of the Nkx3 . 2-Cre line will allow specific manipulation of the genes coding for these signals to ascertain their role during pancreas organogenesis . Another important finding concerns the observation that the pancreatic mesenchyme is required for both endocrine and exocrine development in vivo . Previous reports had reached differing conclusions , with some demonstrating a positive role for the mesenchyme on exocrine formation but not endocrine cell development [6] , [7] , [9] , and others indicating that mesenchymal factors promote proliferation of multi-potent pancreas progenitors that subsequently increase the formation of endocrine cells [11] . Some of these conflicting results can be explained by the different culture conditions used in each experiment . In contrast to the cultured studies , in vivo depletion of the mesenchyme investigated here revealed similar requirements for this tissue with regard to the endocrine and exocrine cytodifferentiation . At this point , we cannot exclude that other cells types , including endothelial and neural-crest derived cells [38]–[41] , or cells residing in the adjacent liver , stomach , gut , or kidneys might provide signals that guide epithelial cell differentiation in mesenchyme depleted embryos in vivo . In addition , mesenchymal cells that did not originate from Nkx3 . 2-Cre expressing cells might still be present in our in vivo model and could provide either instructive or permissive signals . Prior organ culture studies proposed another model to explain the various effects of the mesenchyme on the epithelial compartments by demonstrating distinct effects of the mesenchyme on epithelial cells depending on the physical distance and contact between these tissues [5] . In these experiments , close proximity between epithelial and mesenchymal cells promoted exocrine differentiation while at the same time blocked endocrine formation . In contrast , mesenchyme factors supported endocrine differentiation at a distance , indicating that the physical relation between mesenchymal and epithelial cells is critical for endocrine versus exocrine differentiation . Our studies support the notion of mesenchymal signals being important for both endocrine and exocrine development . However , our lineage tracing experiments provide evidence of close physical contact between Nkx3 . 2/LacZ and Nkx3 . 2/YFP expressing cells with endocrine cells , indicating that close proximity between mesenchymal and epithelial cells does not necessarily interfere with endocrine differentiation . However , since mesenchymal cells surround islets , they are likely in close contact only with peripheral endocrine cells , such as α-cells , while direct interactions with centrally located β-cells might not be common . Whether the mesenchyme contributes to β-cell expansion by releasing secreted factors or through cell-cell interactions as well as how the mesenchyme affects other endocrine cells are questions that need to be addressed in future experiments . Furthermore , isolation and characterization of mesenchymal cells throughout development might reveal cell heterogeneity that could explain differential functions with regard to promoting endocrine versus exocrine development . Our results also point to sustained mesenchyme function as a critical regulator of epithelial pancreas development and identify Wnt signaling as an essential mediator of mesenchyme survival . It is not clear as to whether Wnt signaling is activated in an autocrine or paracrine manner , as several Wnt ligands are expressed by both pancreatic epithelial and mesenchymal cells during development [50] . It is noteworthy that the defects we observe in Nkx3 . 2-Cre;βcatf/f only occur after the onset of canonical Wnt signaling in pancreas mesenchyme as measured by expression of transgenic Wnt-reporters ( i . e . , e13 . 5 [15] , [52] ) . The implication of canonical Wnt signaling as the cause for the observed phenotypes is indirectly supported by a previous study using germ-line knock-out mice in which mPygo2 , a critical component of the nuclear β–catenin/Tcf complex required for β-catenin transcriptional activity , has been eliminated [15] . mPygo−/− mice show pancreas hypoplasia and a reduction in endocrine mass [15] , phenotypes that are not observed when this gene is specifically eliminated in pancreas epithelium . Thus , while mesenchyme specific depletion of mPygo2 has not been reported , the absence of pancreas hypoplasia upon epithelial-specific mPygo2 elimination suggests that at least some of the pancreatic defects are caused by reduced mesenchymal Wnt signaling . However , and in contrast to Nkx3 . 2-Cre;βcatf/f pancreata , the exocrine compartment is not affected in mPygo2−/− mutants and mesenchyme depletion was not reported in those mice . Since Wnt signaling is significantly reduced , but not completely blocked in the absence of mPygo2 [15] , it is possible that low level of canonical Wnt signaling is sufficient for mesenchymal cell survival and the production of factors that promote exocrine cell development . Alternatively , β-catenin is known to regulate cell-cell interactions as part of Cadherin complexes and these additional functions might be crucial for the maintenance of the pancreatic mesenchyme . However , we did not observe β-catenin localized to membranes in mesenchymal cells . In order to study whether different levels of mesenchymal Wnt signaling have a different effect on endocrine and exocrine expansion , mice specifically lacking mesenchymal expression of various components of this pathway ( such as mPygo2 ) would need to be examined . In addition to Wnt signaling , other signaling pathways , such as the RA , BMP , and Hedgehog , have been implicated as mesenchymal factors regulating pancreas development [15]–[17] , [54] , [55] . Using Nkx3 . 2-Cre line as a novel tool to manipulate gene expression in the pancreatic mesenchyme will allow direct study of the role of these and potentially other pathways in pancreas organogenesis . In summary , data presented here indicate continuous requirement of mesenchymal cells and/or mesenchyme-derived signals to regulate epithelial pancreas formation from the onset of organ morphogenesis until the end of gestation . Isolation of mesenchymal cells at different stages of pancreas formation might allow identification of candidate factors that regulate expansion of common and endocrine progenitors as well as of differentiated β-cells . Future therapies for both type I and II diabetes rely on renewable sources of functional insulin-producing β-cells [56] . Current protocols allow the formation of pancreas progenitor cells from human embryonic stem cells ( hESC ) in vitro , but not fully differentiated β-cells . Our results demonstrate that mesenchymal factors provide critical signals for the expansion of both precursors and differentiated endocrine and exocrine cells . Thus , mesenchymal signaling factors not yet identified will likely be useful for expansion of hESC derived pancreas progenitor and differentiated β-cells .
Mice used in this study were maintained according to protocols approved by the Committee on Animal Research at the University of California , San Francisco . Nkx3 . 2 ( Bapx1 ) -Cre mice were described previously [24] . R26-YFPflox ( Gt ( ROSA ) 26Sortm1 ( EYFP ) Cos ) , R26-LacZflox ( Gt ( ROSA ) 26Sortm1Sor ) , R26-eGFP-DTA ( Gt ( ROSA ) 26Sortm1 ( DTA ) Jpmb ) , DTR ( iDTR , Gt ( ROSA ) 26Sortm1 ( HBEGF ) Awai ) , and β-cateninflox ( Ctnnb1tm2Kem ) mice were obtained from Jackson Laboratories . Noon on the day a vaginal plug was detected was considered as embryonic day 0 . 5 . Injections were preformed as previously described [35] , [36] . Briefly , pregnant females were anesthetized , a laparotomy was performed , and the uterus was delivered through the incision ( as illustrated in Figure S2A–D ) . Each embryo was micro-injected with 8 ng/gr body weight Diphtheria Toxin ( Sigma ) diluted in 5 µl PBS . The uterus was placed back into the abdominal cavity and the laparotomy was closed . Embryos were allowed to develop in situ until indicated stages . For immunofluorescence , dissected embryos and pancreatic tissues were fixed with Z-fix ( Anatech ) for 2–16 h , embedded in paraffin wax , and sectioned . For Ptf1a staining , tissues were fixed with Z-fix for 2 h , embedded in OCT ( Tissue Tek ) , and cryosectioned . Tissue sections were stained using the following primary antibodies: rabbit anti-Amylase ( 1∶200 , Sigma ) , goat anti-Cpa1 ( 1∶200 , R&D ) , mouse anti-E-Cadherin ( 1∶200 , BD ) , rabbit anti-Glucagon ( 1∶200 , Linco ) , guinea pig anti-Insulin ( 1∶200 , Linco ) , mouse anti-Ki67 ( 1∶200 , BD ) , rabbit anti-MafA ( 1∶200 , Bethyl ) , armenian hamster anti-Mucin1 ( 1∶200 , Neomarker ) , guinea pig anti-Neurogenin 3 ( 1∶400 , Millipore ) , rabbit anti-phosphorylated Histone H3 ( 1∶200 , Millipore ) , rabbit anti-Pdx1 ( 1∶200 , Millipore ) , rabbit anti-Ptf1a ( 1∶600 , a gift from Dr . Helena Edlund ) , rat anti-Somatostatin ( 1∶200 , Chemicon ) , rabbit anti-Sox9 ( 1∶200 , Chemicon ) , and chicken anti-YFP/GFP ( 1∶400 , Abcam ) followed by staining with Alexa Fluor tagged secondary antibodies ( 1∶500 , Invitrogen ) and mounting with DAPI-containing Vectashield media ( Vector ) . For TUNEL analysis , ApopTag Plus Fluorescein In Situ Apoptosis Detection kit ( Millipore ) was used according to the manufacturer's protocol . For embryo wholemount staining , tissues were processed as previously described [57] and stained with rat anti-E-Cadherin ( 1∶1 , 000 , CalBiochem ) , followed by staining with Alexa Fluor 555 anti-rat secondary antibody ( 1∶500 , Invitrogen ) . For x-gal staining , tissues were fixed with 2% PFA and 0 . 25% Glutaraldehyde for 2 h and incubated overnight with 0 . 5 mg/ml x-gal solution ( Roche ) , followed by a second round of fixation in 4% PFA overnight . Tissues were then embedded in paraffin , sectioned , and counter-stained with nuclear Fast Red ( Vector ) . For histological analysis , dissected tissues were fixed with Z-fix ( Anatech ) , for 4 h , and embedded in paraffin wax . Tissue sections were stained with Meyer's Hematoxylin ( Sigma ) followed by staining with Eosin ( Protocol ) . Images were acquired using Zeiss ApoTome , Leica MZ FL3 and SP5 , and Olympus IX70 microscopes . For all quantifications presented in this study , each transgenic tissue was processed and stained in parallel with a littermate control , with each analyzed group comprising at least three pairs of transgenic and control embryos ( i . e . , n≥3 ) as indicated in the figure legends . Throughout each analysis , images were acquired using the same exposure time and magnification . When MetaMorph software was used for image analysis , the same signal-to-noise threshold was applied throughout the experiment . For all measurements presented in this study , with the exception of the measurement of the mesenchymal area at e11 . 5 and Ptf1a+ cell numbers at e15 . 5 , the following regimen was applied: the entire pancreatic tissue , including both dorsal and ventral buds , was embedded in paraffin wax and cut into 5 µm thick sections . Every fifth section ( 20% of total tissue ) was then immuno-stained with indicated antibodies as described above . Images were acquired as detailed below and analyzed blindly . For measurement of mesenchymal areas at e15 . 5 and e18 . 5 , isolated pancreatic tissues from Nkx3 . 2-Cre;R26-YFPflox embryos were stained with an anti-YFP antibody and a fluorescent secondary antibody and entire sections were automatically imaged using Olympus IX70 widefield microscope and MetaMorph software . Over-exposure of the tissue and DAPI staining were used to determine the edges of the section . Images were analyzed using MetaMorph software , which automatically measured the positive area in each channel . To determine the percentage of mesenchymal area , total YFP-positive area was divided by total tissue area of each section . For β- and acinar cell mass , isolated e18 . 5 tissues ( including both dorsal and ventral tissues ) were dissected and weighed . Following fixation , tissues were embedded in paraffin wax , sectioned as described above , and immuno-stained with anti-Insulin and anti-Amylase antibodies . Images were acquired as described above for mesenchymal area measurement , and areas positive for either Amylase or Insulin , as well as the total pancreatic area , were automatically measured using MetaMorph software . To determine the fractions of the β- and acinar cell areas , total Insulin or Amylase positive area was divided by total tissue area . Cell mass was calculated as the fraction of Amylase+ or Insulin+ areas of the total pancreatic area multiplied by gross pancreas weight . To calculate the β-cell/acinar cell ratio , for each embryo Insulin+ and Amylase+ area was determined as described above for cell mass measurement , and Insulin+ area was divided by Amylase+ area . For clarity , the ratio obtained in non-transgenic controls was set to “1 . ” For quantification of Ngn3-expressing cells , whole e15 . 5 pancreatic tissues were isolated and processed as described above . Sections were stained with anti-Ngn3 antibody followed by fluorescent secondary antibody and images were then acquired as described above for mesenchymal area measurement , but positive cells were counted manually . To accommodate for potential differences in the developmental stage of the various litters analyzed , the number of transgenic Ngn3-positive cells was normalized to the number of Ngn3 cells counted in the corresponding non-transgenic littermate controls . For cell proliferation , whole pancreatic tissue ( including both dorsal and ventral tissues ) was isolated from embryos e15 . 5 and older . From embryos at e13 . 5 or e14 . 5 , pancreatic tissue was isolated together with the adjacent stomach and duodenum . Following fixation , tissues were paraffin-embedded and sectioned as described above . Tissue sections were stained with indicated antibodies and imaged using Zeiss ApoTome or Leica SP5 microscopes . For each section , the percentage of proliferating cells was determined via manual counting of either Ki67 or pHH3 positive cells divided by the number of total target cells . To determine mesenchymal area at e11 . 5 , Nkx3 . 2-Cre;R26-LacZf/+ embryos were stained with X-gal as described above . Entire embryos were then cut to obtain 5 µm thick sections , and all sections were counterstained with FastRed dye and imaged using Zeiss ApoTome . The total dorsal pancreatic bud area , identified by its typical localization and morphology , and pancreatic mesenchyme area , identified by blue x-gal staining , were manually selected and measured using MetaMorph software . For Ptf1a+ cell quantification , isolated e15 . 5 pancreatic tissue were fixed , embedded in OCT , frozen , and cryosectioned . 10 µm thick sections were used and every 10th section was stained ( 10% of total tissue ) . Whole sections were imaged using Leica SP5 confocal microscope and the number of positive cells was counted manually . To account for potential differences in developmental stage of each litter , the number of positive cells obtained for each transgenic animal was normalized to the number obtained from the non-transgenic littermate control . P values were determined using unpaired , two-tailed student t test . Error bars in bar diagrams represent standard deviation of the samples .
|
Embryonic development is a highly complex process that requires tight orchestration of cellular proliferation , differentiation , and migration as cells grow within loosely aggregated mesenchyme and more organized epithelial sheets to form organs and tissues . In addition to intrinsic cell-autonomous signals , these events are further regulated by environmental cues provided by neighboring cells . Prior work demonstrated a critical role for the surrounding mesenchyme in guiding epithelial growth during the early stages of pancreas development . However , it remained unclear whether the mesenchyme also guided the later stages of pancreas organogenesis when the functional exocrine and endocrine cells are formed . Here , we show that specific genetic ablation of the mesenchyme at distinct developmental stages in vivo results in the formation of a smaller , misshapen pancreas . Loss of the mesenchyme profoundly impairs the expansion of both endocrine and exocrine pancreatic progenitors , as well as the proliferative capacity of maturing cells , including insulin-producing beta-cells . Thus , our studies reveal unappreciated roles for the mesenchyme in guiding the formation of the epithelial pancreas throughout development . The results suggest that identifying the specific mesenchymal signals might help to optimize cell culture protocols that aim to achieve the differentiation of stem cells into insulin-producing beta cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] |
[
"medicine",
"growth",
"control",
"cell",
"differentiation",
"endocrine",
"physiology",
"developmental",
"biology",
"organism",
"development",
"gastroenterology",
"and",
"hepatology",
"endocrine",
"cells",
"morphogenesis",
"endocrinology",
"organogenesis",
"diabetes",
"and",
"endocrinology",
"biology",
"pancreas"
] |
2011
|
Pancreatic Mesenchyme Regulates Epithelial Organogenesis throughout Development
|
Dendritic cells ( DCs ) and macrophages ( Møs ) internalize and process exogenous HIV-derived antigens for cross-presentation by MHC-I to cytotoxic CD8+ T cells ( CTL ) . However , how degradation patterns of HIV antigens in the cross-presentation pathways affect immunodominance and immune escape is poorly defined . Here , we studied the processing and cross-presentation of dominant and subdominant HIV-1 Gag-derived epitopes and HLA-restricted mutants by monocyte-derived DCs and Møs . The cross-presentation of HIV proteins by both DCs and Møs led to higher CTL responses specific for immunodominant epitopes . The low CTL responses to subdominant epitopes were increased by pretreatment of target cells with peptidase inhibitors , suggestive of higher intracellular degradation of the corresponding peptides . Using DC and Mø cell extracts as a source of cytosolic , endosomal or lysosomal proteases to degrade long HIV peptides , we identified by mass spectrometry cell-specific and compartment-specific degradation patterns , which favored the production of peptides containing immunodominant epitopes in all compartments . The intracellular stability of optimal HIV-1 epitopes prior to loading onto MHC was highly variable and sequence-dependent in all compartments , and followed CTL hierarchy with immunodominant epitopes presenting higher stability rates . Common HLA-associated mutations in a dominant epitope appearing during acute HIV infection modified the degradation patterns of long HIV peptides , reduced intracellular stability and epitope production in cross-presentation-competent cell compartments , showing that impaired epitope production in the cross-presentation pathway contributes to immune escape . These findings highlight the contribution of degradation patterns in the cross-presentation pathway to HIV immunodominance and provide the first demonstration of immune escape affecting epitope cross-presentation .
Cytotoxic CD8+ T cell ( CTL ) responses play an important role in the outcome of viral infections . CTL responses elicited during HIV or HCV infection follow a predictable immunodominance hierarchy , whereby immunodominant T cell responses are defined by a higher frequency in a population sharing a HLA , or by a higher magnitude of interferon-gamma production in an individual [1] . The acute phase of HIV infection is characterized by narrow immunodominance patterns [2 , 3] , and immune pressure leading to frequent escape mutations in immunodominant epitopes changes the T cell response hierarchy during disease progression [4–9] . Since immunodominance established during HIV infection or reproduced by some HIV vaccines does not clear or prevent infection , breaking immunodominance hierarchies to induce the presentation of broader subdominant but protective epitopes provides an interesting alternative for vaccine design . Immunodominance is shaped by multiple factors [10] , including binding affinity to MHC or the TCR [11 , 12] , frequency of CD8+ T cell precursors and the TCR repertoire [13] , kinetics of expression and amount of viral proteins [14] , and efficiency of antigen processing [15–17] . How degradation patterns during cross-presentation of antigens , specifically in the case of highly variable pathogens like HIV , may shape immunodominance and viral evolution is not well understood . Antigen presenting cells ( APC ) such as DCs and Møs cross-present antigens from various sources , such as cell-associated antigens [18–21] , viral particles [22–24] , or viral proteins [25 , 26] for priming or activation of T cell responses . Internalized antigens first undergo proteolytic processing by cathepsins in endocytic compartments [27] where they can be loaded onto MHC I or MHC II molecules for presentation to CD8+ or CD4+ T cells [28] , or eventually escape into the cytosol [29] for additional degradation [30 , 31] , translocation in the ER and cross-presentation by MHC I . The cell type and the trafficking of antigens have a crucial impact on their processing , as different proteases in each compartment can produce or destroy epitopes , thus shaping the surface peptidome [25 , 32 , 33] . Different cell types express individual patterns of proteases , which affect epitope processing as we previously showed for the degradation of several HIV epitopes by cytosolic peptidases [34 , 35] . In a given cell type , the degradation of proteins in the cytosol and in the ER [34–36] contribute to defining the timing and amount of peptides available for presentation , and have been shown to preferentially produce multiple epitopes corresponding to immunodominant responses in HIV [16 , 17] and HCV infection [15] . Moreover , differences in degradation patterns of HIV peptides in cytosolic , endosomal or lysosomal cell extracts isolated from human PBMCs [37 , 38] further highlight the critical role of antigen trafficking on epitope processing . Mutations within and outside epitopes alter degradation patterns by proteasomes and aminopeptidases in the cytosol or in the ER , reduce epitope presentation and lead to immune escape [39–42] . Nothing is known about the impact of these mutations on the degradation patterns during cross-presentation despite its potential impact for T cell priming and activation . The aim of this study was to systematically examine the processing and cross-presentation of dominant and subdominant HIV Gag-derived epitopes and of natural mutants of an immunodominant epitope by monocyte-derived DCs and Møs . We showed a preferential production and a superior intracellular stability of peptides containing immunodominant epitopes in cytosol and endolysosomes . Moreover , we showed that frequent HLA-restricted mutations in an immunodominant peptide associated with shifts in immunodominance patterns , modified the degradation patterns of HIV fragments in endolysosomes and reduced epitope stability and production in the cross-presentation pathway . These results highlight the contribution of degradation patterns in the cross-presentation pathways of APC to immunodominance and immune escape in HIV infection .
Cells were isolated from HLA-typed blood donors or anonymous buffy coats after written informed consent and approval by the Partners Human Research Committee under protocol 2005P001218 ( Boston , USA ) . Human peripheral blood mononuclear cells ( PBMCs ) were isolated by Ficoll-Hypaque ( Sigma-Aldrich ) density centrifugation . Monocytes were enriched using CD14+ magnetic isolation kits ( StemCell ) and differentiated into DCs and Møs during a 6-day culture . DCs were cultured in AIM-V media with 1% human serum AB ( Gemini Bio-Products ) supplemented with 20ng/mL IL-4 and 10ng/mL GM-CSF ( CellGenix ) . On days 2 and 4 , fresh IL-4 and GM-CSF were added . Møs were cultured in ultra low attachment plates ( Corning ) in AIM-V media with 10% human serum AB . Where indicated , maturation of DCs and Møs was induced by TLR ligand stimulation with 2μg/mL LPS , 1μg/mL CL097 , or 1μg/mL R848 ( Invivogen ) for 2 days [35] . Epitope-specific CTL clones were maintained in the presence of 50U/mL IL-2 , using 0 . 1μg/mL CD3-specific mAb 12F6 , and irradiated feeder cells as stimulus for T cell proliferation . Immature DCs and Møs were exposed to recombinant HIV-1 p24-protein , HIV-1 p55-protein or control protein ( Protein Sciences Corporation , USA ) for 1hr at 37°C . Where indicated , cells were pre-incubated for 45 minutes with inhibitors for proteasome ( 10μM MG132 ( Enzo Life Sciences ) ) or cysteine proteases ( 5μM E64 ( Sigma-Aldrich ) ) . Cells pulsed with equivalent molar concentrations of the optimal epitopes were used as controls for antigen presentation and CTL clone specificity . DCs and Møs were thoroughly washed and cultured overnight with epitope-specific CTL clones at a 2:1 effector-to-target ratio in 96-well plates ( Millipore ) coated with anti-IFN-y mAb 1-D1K ( Mabtech ) . ELISPOT plates were washed and developed as described previously [43] . DCs or Møs ( 1x106 cells/mL ) were exposed to the following protease inhibitors for 45 minutes: 10μM MG132 , 5μM E64 , 10μM cathepsin S inhibitor Z-FL-COCHO , 10μM leupeptin ( Enzo Life Sciences ) , 120μM bestatin ( Sigma-Aldrich ) , before incubation with different concentrations of recombinant HIV-1 p24-protein for 1hr at 37°C or 4°C . Samples were thoroughly washed in ice-cold PBS and immediately treated with 3mg/mL pronase E ( Sigma-Aldrich ) in AIM-V media without serum for 10 minutes on ice . Cells were washed , lysed in 0 . 5% Triton X-100 containing lysis buffer and the amount of p24 protein in cell lysates was measured using a standard HIV-1 p24 antigen ELISA ( Perkin Elmer ) . Whole cell extracts from DCs and Møs were prepared by 0 . 125% digitonin permeabilization in ice-cold lysis buffer ( 50mM HEPES , 50mM potassium acetate , 5mM MgCl2 , 1mM DTT , 1mM ATP , 0 . 5mM EDTA , 10% Glycerol , pH 7 . 4 ) , followed by 17 , 762 rcf centrifugation at 4°C for 15 minutes to remove cell debris as previously done [17 , 35 , 44] . The proteolytic activities of cathepsin S ( cell , 50μM; extracts , 10μM Z-VVR-AMC ) , omni cathepsins ( cell , 50μM; extracts , 50μM Z-FR-AMC ) , cathepsin D&E ( extracts , 10μM Mca-GKPILFFRLK-Dnp , Enzo Life Sciences ) , and cathepsin B ( extracts , 50μM Z-RR-AMC , Bachem ) were measured by cleavage of peptide-specific fluorogenic substrates . Incubation with the relevant inhibitor of cathepsin S ( 10μM ZFL-COCHOO , Calbiochem ) , cathepsin B ( 10μM Z-RLVazaglyIV-OMe , Bachem ) , omnicathepsins ( 50μM E64 ) , and cathepsin D&E ( 100μM Pepstatin A , Enzo Life Sciences ) confirmed the specificity of reactions . For cells , 2x104 DCs or Møs in PBS/0 . 0025% digitonin were used to measure the proteolytic activities . For cell extracts , equivalent amounts as determined by total protein concentration were used in reaction buffer ( 50mM sodium chloride , 50mM potassium phosphate , 2mM DTT , 2mM EDTA; pH 7 . 4 , pH5 . 5 or pH4 . 0 , respectively ) . The rate of fluorescence emission , which is proportional to the proteolytic activity , was measured every 5 minutes at 37°C in a Victor-3 Plate Reader ( Perkin Elmer ) [34 , 35] . 2nmol of >98% pure peptides ( Bio-Synthesis , USA ) were digested with 15μg of whole cell extracts , normalized to actin levels , at 37°C in 50μL of degradation buffer ( 50mM Tris-HCl , 137mM potassium acetate , 1mM MgCl2 , and 1mM ATP , pH7 . 4 , pH5 . 5 , or pH4 . 0 ) [37 , 45] . At various time points the reaction was stopped with 2 . 5μL of 100% formic acid ( FA ) and peptide fragments were purified by 5% trichloroacetic acid precipitation . Peptides in the digestion mix were identified by in house mass spectrometry . Equal amounts of peptide degradation samples were injected into a Nano-HPLC ( Eksigent ) and online nanosprayed into an Orbitrap mass spectrometer ( LTQ Orbitrap Discovery , Thermo ) with a flow rate of 400nL/min . A Nano cHiPLC trap column ( 200μm x 0 . 5mm ChromXP c18-CL 5μm 120Å; Eksigent ) was used to remove salts in the sample buffer . Peptides were separated in a Nano cHiPLC column ( 75μm x 15cm ChromXP c18-CL 5μm 300Å; Eksigent ) over a gradient of 2% to 40% buffer B ( buffer A: 0 . 1% FA in water; buffer B: 0 . 1% FA in acetonitrile ) and mass spectra were recorded in the range of 370 to 2000Daltons . In tandem MS/MS mode , the eight most intense peaks were selected with a window of 1Da and fragmented . The collision gas was helium , and the collision voltage was 35V . Masses in the mass spectra were searched against source peptide databases with Proteome Discoverer ( Thermo Scientific ) . The integrated area under a peptide peak is proportional to its abundance . Each sample was run on the mass spectrometer at least twice . One nmol of highly purified peptide was degraded in 15μg of whole cell extracts at 37°C in degradation buffer at pH7 . 4 or pH4 . 0 [37] . Aliquots were taken at 0 , 10 , 30 , and 60 minutes , and the reaction was stopped with 2 . 5μL of 100% TFA . The remaining peptide at each time point was quantified by reversed-phase HPLC ( RP-HPLC; Waters ) . 100% represents the amount of peptide detected at time point 0 calculated as the area under the peptide peak . A stability rate of each peptide was calculated by a nonlinear regression ( one-phase exponential decay ) of the degradation profile obtained over a 60-minute incubation [35 , 44] . Peptides incubated in buffer without cell extracts were used as controls . Spearman’s rank correlation coefficient was used to examine bivariate associations . The Kruskal-Wallis test was used to compare measurements between groups . In figures , p-value criteria are assigned as * p<0 . 05 , ** p<0 . 01 and *** p<0 . 001 . Statistical analyses were conducted using GraphPad Prism ( GraphPad Prism Software , USA ) .
We analyzed the cross-presentation of the three optimally defined HLA-B57 restricted HIV epitopes originating from HIV-1 p24 protein by immature monocyte-derived DCs and Møs: subdominant B57-ISW9 ( ISPRTLNAW , aa 15–23 in Gag p24 ) , dominant B57-KF11 ( KAFSPEVIPMF , aa 30–40 in Gag p24 ) , and dominant B57-TW10 ( TSTLQEQIGW , aa 108–117 in Gag p24 ) [46 , 47] . B57-ISW9-specific CTL responses to cross-presenting DCs were 28-fold and 94-fold lower compared with B57-KF11 and B57-TW10-specific responses , respectively ( Fig . 1A , left panel ) . Similar results were observed with cross-presenting Møs , with 47-fold lower B57-ISW9-specific CTL responses compared with B57-KF11 and B57-TW10-specific CTL responses ( Fig . 1A , right panel ) . DCs and Møs pulsed with increasing amounts of synthetic ISW9 or TW10 peptides similarly activated epitope-specific CTLs ( Fig . 1B ) . Since a previous study showed comparable affinities of ISW9 , KF11 and TW10 peptides for HLA-B57 [48] , our results suggest that differences in CTL responses to cross-presenting DCs and Møs are not due to differences in peptide avidity among the clones , but likely to differential epitope production . To ensure that epitopes cross-presented by DCs and Møs were endogenously processed , we measured the intracellular concentrations of HIV-1 p24 protein after uptake at 37°C or 4°C . In both cell subsets the intracellular concentration of HIV p24 increased with the amount of p24 used for uptake at 37°C whereas the uptake at 4°C was minimal ( Fig . 1C ) . Immature Møs showed at least 5-fold lower intracellular p24 concentrations than DCs , which may indicate a faster degradation of internalized protein by Møs [49] . Moreover , B57-KF11-specific CTL responses increased with the amount of exogenous p24 protein added to cells , in accordance with higher amount of intracellular p24 leading to higher amount of peptide presentation ( Fig . 1D ) . Together , these data show that the higher CTL responses against dominant TW10 and KF11 epitopes after uptake of p24 by DCs and Møs are due to cross-presentation of higher amounts of both peptides compared with subdominant ISW9 . We aimed to identify factors contributing to the production or destruction of the three epitopes in each cell type . Incubation of immature DCs with proteasome inhibitor MG132 resulted in a 43-fold and 4-fold increased presentation of B57-ISW9 and B57-KF11 epitopes , respectively , suggesting that proteasomal degradation of epitope-containing peptides limited the amount of ISW9 and KF11 available for presentation ( Fig . 2A ) . In contrast , inhibition of cysteine proteases by E64 had no effect on the cross-presentation of both epitopes , indicating that fragments escape early into the cytosol before trafficking to compartments with high cysteine protease activity . B57-TW10-specific CTL responses to cross-presenting immature DCs decreased approximately 3-fold upon inhibition of proteasomes , suggesting that proteasomal processing is required for efficient presentation of TW10 . In contrast to DCs , the cross-presentation of B57-ISW9 by immature Møs was not affected upon inhibition of proteasomes , suggesting that the cross-presentation of ISW9 in Møs is proteasome-independent ( Fig . 2B ) . B57-KF11 and B57-TW10 CTL responses decreased 2- and 3-fold respectively upon proteasome inhibition with MG132 or epoxomicin in Møs , suggesting that the processing of both epitopes requires proteasome processing in Møs . Inhibition of cysteine proteases in Møs did not affect the cross-presentation of ISW9 , KF11 and TW10 . Together , these results indicate that cross-presentation of HIV-1 p24 involves distinct proteases in DCs and Møs , which can be essential or detrimental for the processing of epitopes . Exogenous antigens internalized by DCs and Møs first encounter several proteases in endo- and lysosomes [27] , before presentation or additional degradation in the cytosol . In line with previous studies we observed lower omnicathepsin and cathepsin S activities in DCs compared with Møs [49] , which further decreased upon maturation of DCs as shown by an inverse correlation between both activities and the % of mature DCs ( S1A Fig ) . To assess how degradation of HIV peptides along the cross-presentation pathway of immature and TLR ligand-stimulated DCs and Møs may contribute to shaping immunodominance patterns , we used a previously developed degradation assay recapitulating degradation in the cross-presentation compartments [37] . This assay allows the simultaneous analysis of degradation products by cytosolic , endosomal and lysosomal peptidases from the same cells using mass spectrometry . Omnicathepsin and cathepsin S activities measured in live intact cells correlated to their matching cell extracts , as previously demonstrated for cytosolic proteases [35] ( S1B Fig ) and could be activated at different pH values , in accordance with differential cathepsin activation in endosomes and lysosomes ( S1C Fig ) [37] . Degradation of a synthetic 35-mer peptide containing the epitopes B57-ISW9 and B57-KF11 ( MVHQAISPRTLNAWVKVVEEKAFSPEVIPMFAALS , aa 10–44 in Gag p24 ) [34 , 35 , 37] showed the production of peptides of variable lengths at different pH values over time ( S2 Fig ) . To assess and compare the production of peptides in each cell subset and cell compartment , we used the area under each peptide peak identified by mass spectrometry , which we previously showed to be proportional to the amount of the corresponding peptide [35 , 50] . Peptides were grouped according to their lengths or epitope content , and the contribution of each category of peptides to the total degradation products was calculated for each time point . Degradation at pH4 . 0 in cell extracts from immature DCs yielded shorter fragments compared with degradation at pH7 . 4 , with majority of fragments being 8–12 and 13–18 aa long and contributing to 45% and 51% of total peptide intensity at 120 minutes , respectively ( Fig . 3A , upper left panel ) . The degradation of fragments containing both epitopes ( ISW9+/KF11+ ) resulted in the preferential production of B57-KF11 epitope-containing fragments ( ISW9-/KF11+ ) , and only small amounts of B57-ISW9 epitope-containing fragments ( ISW9+/KF11- ) in extracts of immature DCs at all pH values tested ( Fig . 3B , upper left panel ) . KF11- and ISW9-containing fragments were produced more efficiently at pH7 . 4 than at pH5 . 5 and pH4 . 0 , indicating a higher presentation in the direct presentation pathway , or if epitope precursors escape from endolysosomes . Similar results were observed for immature Møs ( Fig . 3A/B , lower left panel ) , in line with comparable cytosolic and endocytic hydrolytic activities in immature DCs and Møs [35] . Degradation of the 35-mer in extracts from DCs and Møs matured with LPS yielded similar degradation patterns with fragments of comparable lengths and higher amounts of fragments containing immunodominant epitope B57-KF11 ( Fig . 3A/B , right panel ) . We further analyzed the cleavage patterns by measuring the relative amount of fragments with a specific N terminus or C terminus ( Fig . 3C , upper or lower graph of each panel ) . After 30 minutes of degradation at pH7 . 4 several minor cleavage sites produced ISW9- and KF11-containing fragments , whereas at pH5 . 5 and pH4 . 0 the generation of fragments with a Tryptophan at the N terminus destroyed ISW9 and fragments with a Methionine and Proline at the C terminus destroyed KF11 . Further trimming resulted in the appearance of new N- and C-terminal cleavage sites , which still preserved KF11-containing peptides at pH7 . 4 , whereas at pH5 . 5 and pH4 . 0 both epitopes were further destroyed . These data indicate that this p24 35-mer is sensitive to degradation in all three cell compartments in DCs and Møs , and favors the production of dominant epitope KF11 over that of subdominant ISW9 , in line with the more efficient cross-presentation of KF11 and the rescue of ISW9 in the presence of protease inhibitors . Moreover , we analyzed the production of 16 well described HIV CD8+ and CD4+ T cell epitopes [46] and epitope precursors , defined as N-terminal extended epitopes , located in this 35-mer ( S3A–S3B Fig ) . Peptides were produced in extracts of both cell subsets at all pH values tested ( B57-KF11 , B15-HL9 , A25-QW11 , B57-FF9 , B57-KP9 ) , preferentially produced at pH7 . 4 ( B57-ISW9 , A02-TV9 ) , or at pH5 . 5 and pH4 . 0 ( B45-VI11 , B15-VF9 , B44-EV9 ) or not produced at any time ( B07-SV9 ) . These results highlight a variable production or degradation of epitopes in different cell compartments , which may affect their capacity to activate CD8+ or CD4+ T cells during infection or cross-presentation . We next extended the analysis to another HIV-1 Gag p24-derived peptide containing the epitope B57-TW10 ( GSDIAGTTSTLQEQIGWMTNNPPIPVGGEIY , aa 101–131 in Gag p24 ) , dominant in HIV acute infection ( Fig . 4 ) . In contrast to p24 35-mer , the majority of degradation products identified after 10 to 120 minutes in extracts from immature DCs and Møs at pH7 . 4 , pH5 . 5 and pH4 . 0 represented the original fragment or long fragments of mostly >26aa ( Fig . 4A ) , containing the TW10 epitope with N- and C-terminal extensions ( Fig . 4B ) . Similarly , degradation in extracts from mature DCs and Møs showed comparable kinetics of degradation and resulted in the production of fragments with similar lengths . Accordingly the few cleavage sites identified after 30 or 120 minutes were mostly located outside B57-TW10 at all three pH values , protecting the antigenic peptide from degradation , in line with the highly efficient cross-presentation of B57-TW10 ( S4 Fig ) . The relative resistance of the B57-TW10-containing fragment to intracellular degradation contrasted with the rapid degradation of the p24 fragment containing B57-ISW9 and B57-KF11 which may contribute to a higher amount of TW10-containing peptide available for presentation by direct or cross-presentation , thus contributing to the dominance of TW10-specific CTL responses during acute HIV infection . We next analyzed the cross-presentation of another immunodominant epitope located in a different HIV-1 protein and restricted by a different HLA allele . The A03-RK9 epitope ( RLRPGGKKK , aa 20–28 in Gag p17 ) is efficiently produced in the endogenous processing pathway for presentation to A03-RK9-specific CTLs [17 , 44] . Immature DCs and Møs , incubated with recombinant HIV-1 p55 protein elicited A03-RK9-specific CTL responses as strong as cells exogenously pulsed with 1 . 2 ug/ml RK9 ( Fig . 5A-B ) . RK9 cross-presentation was not affected by inhibition of proteasome or cysteine proteases , in line with the limited sensitivity of proteasome and cathepsin-mediated degradation resulting in high amounts of peptide for maximum T cell stimulation ( Fig . 5A ) . The incubation of DCs or Møs with different concentrations of p55 protein resulted in concentration-dependent A03-RK9-specific CTL responses ( Fig . 5C ) . In vitro degradation of HIV-1 p17-derived peptide containing the A03-RK9 epitope ( RWEKIRLRPGGKKKYKL , aa 15–31 in p17 ) showed minimal degradation of A03-RK9 at all pH values tested ( Fig . 5D ) . These data indicate that the limited degradation of this peptide may result in more fragments available for cross-presentation compared with B57-ISW9 and B57-KF11 epitopes , thus contributing to immunodominance of RK9-specific CTL responses . The variable stability of epitopes in cytosolic extracts of PBMCs contributes to the amount of peptide available for presentation to T cells [44 , 50 , 51] . We hypothesize that peptide stability in cytosol and endolysosomes of DCs and Møs may contribute to the relative efficiency of cross-presentation of immunodominant epitopes . We first measured the stability of A03-RK9 , a dominant epitope in the acute phase , and B57-KF11 , a dominant epitope in the chronic phase , in whole cell extracts of immature DCs and Møs at pH7 . 4 and pH4 . 0 . In cell extracts of immature DCs , the B57-KF11 epitope was 10-fold faster degraded at pH4 . 0 than at pH7 . 4 ( half-lives of 1 . 9 minutes versus 20 minutes ) whereas the A03-RK9 epitope was 20-fold faster degraded at pH7 . 4 than at pH4 . 0 ( half-lives of 61 minutes versus 1223 minutes ) ( Fig . 6A , left panel ) . Similar results were observed in cell extracts from immature Møs ( Fig . 6A , right panel ) . We next compared the intracellular stability of seven well-defined dominant and subdominant MHC I epitopes located in HIV Gag p24 , Gag p17 and RT , and examined whether their stability corresponded to immunodominance patterns observed in HIV infection . To rank epitopes we calculated a stability rate as done before [44 , 50] . In cell extracts of immature DCs at pH7 . 4 the dominant epitopes A03-RK9 and B57-TW10 showed approximately 5-fold and 4-fold higher stability rates compared with the subdominant epitopes B57-ISW9 and A11-ATK9 ( AIFQSSMTK , aa 158–166 in RT ) , respectively ( Fig . 6B , upper left panel ) . At pH4 . 0 we detected dramatically reduced stability rates for B57-ISW9 , B57-KF11 , B57-FF9 , and A11-ATK9 , indicating a more rapid proteolysis by proteases located in endo- and lysosomes . However , the observed stability rates of all epitopes formed the same hierarchy as seen at pH7 . 4 . Similar results were observed in cell extracts of immature Møs ( Fig . 6B , lower left panel ) . Moreover , the subdominant A03-KK9 epitope overlapping with the dominant A03-RK9 epitope had a 3-fold lower stability rate in both cell subsets at pH4 . 0 , whereas the dominant and overlapping epitopes B57-KF11 ( KAFSPEVIPMF , aa 30–40 in Gag p24 ) and B57-FF9 ( FSPEVIPMF , aa 32–40 in Gag p24 ) [52] had comparable stability rates . In line with our previous study [35] , the cytosolic stability rate of all epitopes was not affected upon maturation of DCs and Møs with LPS ( Fig . 6B , upper and lower right panel ) . Together , these results show that the intracellular stability of optimal HIV epitopes is highly variable in DCs and Møs and in different cell compartments , but follows similar hierarchies and may contribute to differences seen in cross-presentation and immunodominance patterns observed in HIV infection . Immune pressure exerted by T cell immune responses leads to predictable mutations within and outside epitopes altering viral fitness , epitope processing and presentation [44] . In HLA-B57+ patients the TW10 epitope rapidly mutates at residues 3 and 9 during acute infection [53 , 54] , and the dominant TW10 CTL response wanes while KF11-specific CTL responses become dominant [2 , 55] . To assess the impact of escape mutations on degradation patterns in the cross-presentation pathway , peptides containing the TW10 epitope or its naturally occurring variants TW10 T3N or TW10 T3N/G9A were degraded in whole cell extracts from immature DCs and Møs at pH7 . 4 and pH4 . 0 for 10 , 30 and 60 minutes . Degradation of the two variants at pH7 . 4 showed comparable kinetics of disappearance of the original peptides ( 64–67% of original peptides left ) , whereas the WT showed faster degradation with 23% of the original peptide left after 60 minutes ( Fig . 7A ) . However , both mutants generated less N- and C-extended TW10-containing peptides than the WT at pH7 . 4 ( 23% TW10 T3N variant , 19% TW10 T3N/G9A variant vs . 45% TW10 WT at 60 minutes ) . Degradation at pH4 . 0 demonstrated a very fast kinetic of degradation of the T3N/G9A peptide , and generation of a majority of fragments lacking part of the epitope ( antitopes ) ( 39% TW10 WT , 65% T3N variant , and 99% T3N/G9A variant of antitopes produced after 10 minutes ) . The analysis of the N- and C-terminal cleavage sites showed that TW10 WT and TW10 T3N sequences were spared from degradation at pH7 . 4 , whereas a cleavage site between Tryptophane and Alanine partly destroyed the TW10 T3N/G9A variant within 10 minutes of degradation ( Fig . 7B , upper panel ) . At pH4 . 0 , in line with the faster degradation of the long peptides into antitopes , two major cleavage sites produced short antitopes with an Isoleucine at the N-terminus or a Glutamine at the C-terminus that partly destroyed TW10 WT , and more extensively TW10 T3N/G9A . These cleavage sites were 2 . 8-fold and 2 . 6-fold more pronounced in the TW10 T3N/G9A variant compared with the TW10 WT and the TW10 T3N variant ( Fig . 7B , lower panel ) . We compared the cytosolic and lysosomal stability of TW10 epitope and its variants in DCs and Møs ( Fig . 7C ) . Similar intracellular stabilities were observed for B57-TW10 and TW10 T3N in each compartment whereas the stability of TW10 T3N/G9A was reduced by 7- to 8-fold in both compartments . These results demonstrate that the low intracellular stability of TW10 T3N/G9A variant contribute to reducing the epitope presentation in both direct and cross-presentation pathways , in line with our previous findings in cells infected with virus containing TW10 WT or variants [44] . This represents the first demonstration of an escape mutation affecting the cross-presentation of an HIV epitope .
This study shows how degradation patterns in the cross-presentation pathway of APCs favor the production of immunodominant HIV epitopes and provides the first demonstration of antigen processing mutations affecting the cross-presentation of HIV epitopes by APCs to CTLs . The production and presentation of HIV epitopes are affected by both the cell subsets and the cellular compartments in which antigen traffics [34–36 , 38] . Differences in the processing of epitopes between monocyte-derived DCs and Møs include the involvement of distinct sets of peptidases in the production or destruction of epitopes , and in the trafficking of degradation products . While proteasomes were involved in the degradation of B57-ISW9-containing fragments in DCs , its processing in Møs was proteasome-independent , suggesting possible differences in the trafficking of epitope-containing fragments and the involvement of proteases located in different subcellular compartments [33 , 56–58] . The degradation of peptides by specific proteases and the translocation across different cell compartments has been demonstrated to depend on the size of fragments [59 , 60] , which may contribute to the observed differences in cross-presentation of B57-ISW9 and B57-KF11 by DCs and Møs upon proteasome inhibition . The formation of specific peptide pools [61] or DC-specific antigen storage compartments [62] may permit the fast cross-presentation of HIV peptides by DCs and Møs but this remains to be demonstrated . In addition , cell-specific peptidases such as a serine protease uniquely expressed in monocytes [63] , may lead to cell-specific processing of epitopes and may contribute to the differential priming of immune responses by different tissue DC subsets observed in vivo [64–66] . However , further transcriptomics , proteomics [67] and functional analyses are needed to identify additional cell subset-specific peptidases that may shape epitope presentation by various APC subsets . A major difference between DCs and Møs contributing to the ability of DCs for cross-priming is their capacity to tightly control endolysosomal pH at higher values than in Møs [68 , 69] , which leads to lower cathepsin activities and slower degradation rates of proteins [49] . However lower peptidase activities are not always corresponding to better epitope production , as we identified epitopes processed in higher amount in Møs or in monocytes than in DCs [35] . Although technically still challenging it will be essential to determine the relative amount of peptides required for priming or activation of T cell responses by DCs or other infectable cell subsets , and how this will affect the capacity of T cells to recognize infected targets and clear infection . The level of peptidase activities in a given cell type and compartment , and the sensitivity of a given antigenic sequence to degradation in this compartment shapes the amount of epitope available for presentation [70] . In the cytosol the degradation profiles of proteins into epitopes [17 , 42] and the intrinsic stability of HIV peptides before loading onto MHC [35 , 44 , 51] determine the timing and amount of peptide available for presentation and are defined by specific motifs . It is likely that these steps in the endolysosomal pathway will be driven by motifs that still remain to be identified . A direct consequence of this sequence- and compartment-dependent degradation of proteins is the impact of HLA-restricted mutations on HIV epitope processing and cross-presentation . Immune pressure selects variants impairing viral fitness and/or epitope presentation , reducing binding to MHC or to the TCR [71] . Flanking mutations have been shown to prevent the processing of epitopes [16 , 40 , 42 , 72] and intraepitopic mutations can destroy epitopes [44 , 73 , 74] , which demonstrates the adaptation of HIV to antigen processing in the cytosol . This study provides the first demonstration that a frequently detected HLA-restricted mutation during acute infection affects the cross- and direct presentation of an epitope in DCs and Møs . While the WT epitope B57-TW10 , dominant in the acute phase , is efficiently processed and highly stable in endo-lysosomes and the cytosol of DCs and Møs its mutant is degraded faster than WT . The lower amount of epitope available for presentation may lead to less epitope presented to T cells and therefore results most likely in subdominant responses , which coincides with a shift in immunodominance toward B57-KF11 in HLA-B57 persons during chronic HIV infection [2 , 55] . A vaccine eliciting the same immunodominance patterns as natural infection cannot be successful at preventing or clearing HIV infection . Breaking natural immunodominance and targeting immune responses towards protective epitopes [55] is required in the design of a T cell arm of vaccine strategies [75 , 76] . The combination of sequence alterations to modulate epitope production and intracellular peptide stability and to break natural immunodominance , the use of adjuvants modulating epitope processing as well as the targeting of immunogens to specific cell compartments offers ways to modulate epitope presentation to induce protective immunity in HIV infection and beyond .
|
Pathogens such as HIV can enter cells by fusion at the plasma membrane for delivery in the cytosol , or by internalization in endolysosomal vesicles . Pathogens can be degraded in these various compartments into peptides ( epitopes ) displayed at the cell surface by MHC-I . The presentation of pathogen-derived peptides triggers the activation of T cell immune responses and the clearance of infected cells . How the diversity of compartments in which HIV traffics combined with the diversity of HIV sequences affects the degradation of HIV and the recognition of infected cells by immune cells is not understood . We compared the degradation of HIV proteins in subcellular compartments of dendritic cells and macrophages , two cell types targeted by HIV and the subsequent presentation of epitopes to T cells . We show variable degradation patterns of HIV according to compartments , and the preferential production and superior intracellular stability of immunodominant epitopes corresponding to stronger T cell responses . Frequent mutations in immunodominant epitopes during acute infection resulted in decreased production and intracellular stability of these epitopes . Together these results demonstrate the importance of protein degradation patterns in shaping immunodominant epitopes and the contribution of impaired epitope production in all cellular compartments to immune escape during HIV infection .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Variable Processing and Cross-presentation of HIV by Dendritic Cells and Macrophages Shapes CTL Immunodominance and Immune Escape
|
The disease phenotype of bovine spongiform encephalopathy ( BSE ) and the molecular/ biological properties of its prion strain , including the host range and the characteristics of BSE-related disorders , have been extensively studied since its discovery in 1986 . In recent years , systematic testing of the brains of cattle coming to slaughter resulted in the identification of at least two atypical forms of BSE . These emerging disorders are characterized by novel conformers of the bovine pathological prion protein ( PrPTSE ) , named high-type ( BSE-H ) and low-type ( BSE-L ) . We recently reported two Italian atypical cases with a PrPTSE type identical to BSE-L , pathologically characterized by PrP amyloid plaques and known as bovine amyloidotic spongiform encephalopathy ( BASE ) . Several lines of evidence suggest that BASE is highly virulent and easily transmissible to a wide host range . Experimental transmission to transgenic mice overexpressing bovine PrP ( Tgbov XV ) suggested that BASE is caused by a prion strain distinct from the BSE isolate . In the present study , we experimentally infected Friesian and Alpine brown cattle with Italian BSE and BASE isolates via the intracerebral route . BASE-infected cattle developed amyotrophic changes accompanied by mental dullness . The molecular and neuropathological profiles , including PrP deposition pattern , closely matched those observed in the original cases . This study provides clear evidence of BASE as a distinct prion isolate and discloses a novel disease phenotype in cattle .
Prion diseases , or transmissible spongiform encephalopathies ( TSEs ) , are mammalian neurodegenerative disorders of sporadic , genetic , or infectious origin characterized by accumulation and deposition of an abnormal isoform ( PrPTSE ) of the cellular prion protein ( PrPC ) in the brain [1] . TSEs include a wide range of animal and human disorders , such as BSE in cattle , scrapie in sheep and goats , chronic wasting disease in deer and elk , and Creutzfeldt-Jakob disease ( CJD ) in humans [1] . First identified in 1986 in the UK , BSE has been confirmed in over 180 , 000 cases , although more than one million cattle have been estimated to be infected [2] . Evidence of the spread of the BSE agent across certain mammalian species , including humans , indicates that this disease is a major animal and human public health issue [3]–[5] . Common neurological signs in cattle include apprehension , hyperaesthesia , kicking , and pelvic limb ataxia , accompanied by general signs such as reduced milk yield and loss of conditions . In all cases , progression to behavioural , sensory and posture/movement alterations led to death within a few months [6] , [7] . Early transmission studies showed that isolates from field BSE cases and variant CJD ( vCJD ) , its human counterpart , were all caused by a single prion strain [4] . In addition , PrPTSE from BSE and vCJD cases exhibited a distinctive glycotype signature , with high glycosylation site occupancy and similar electrophoretic mobility of the unglycosylated protease-resistant PrPTSE fragment [8] . These PrPTSE traits have been used as biochemical indicators of the BSE prion strain . Until recently , monitoring of BSE in cattle was accomplished by passive surveillance and pathological confirmation of suspected clinical cases . In 2001 , the European Community imposed an active surveillance system based on biochemical tests of brain tissues from all slaughtered cattle over 30 months of age . This strategy led to the recent identification of new PrPTSE types , provisionally termed as “type-H” and “type-L” according to the electrophoretic migration of the unglycosylated proteinase K-resistant PrPTSE , which is higher ( BSE-H ) or lower ( BSE-L ) than classical BSE ( BSE-C ) [9]–[11] . An additional distinctive signature of type-H and type-L is the even representation of di- , mono- , and unglycosylated PrPTSE species . In 2004 , we described two aged asymptomatic Italian cattle of Piemontese and Alpine brown breeds neuropathologically characterized by the presence of PrP-amyloid plaques [12] . This new pathological phenotype , named BASE , was characterized by marked involvement of olfactory areas , hippocampus , and thalamus , with relative sparing of the brainstem . The molecular signature of BASE PrPTSE was similar to that later detected in BSE-L cases [11] . A feature shared between BASE , a condition both well-defined molecularly and pathologically , and “L-type” cases , defined only on a molecular basis , is the older age of the affected animals ( approximately 12 years ) as compared to BSE cases ( 5–6 years ) . Recent studies have shown that BASE and “L-type” isolates exhibit similar biological properties upon transmission to Tgbov XV , and have shorter incubation period and survival time than BSE; these findings are suggestive of a single prion strain for BASE and BSE-L [10] , [13] . In contrast , the H-type phenotype showed an unusually long incubation period in Tgbov XV [10] . To date , all of the available demographic and molecular evidence strongly suggests that H-type BSE and BASE-L represent sporadic forms of bovine spongiform encephalopathies [14] . Human susceptibility to BASE has been suggested by experimental transmission to primates and to PrP humanized transgenic mice [14] . Here we inoculated cattle of different breeds with brain homogenates from Italian BASE and BSE cases , in order to assess the strain attributes and disease phenotype of the above isolates in their natural hosts .
All Friesian cattle were homozygous for six octapeptide repeat copies , and three cattle carried a silent mutation at codon 78 ( CAG/CAA ) . Four out of six Alpine brown cattle were homozygous for six octapeptide repeat copies; one animal carried 6/7 and another 5/7 octapeptide repeat copies . Four different silent mutations were found at codons 78 ( CAG/CAA ) , 23 ( CTC/CTT ) , 95 ( CCA/CCC ) , and 77 ( GGT/GGC ) in four cattle . Homozygosity for 23 bp and 12 bp deletion alleles was present in three Friesian cattle . Results of these genetic studies are summarized in Table 1 . A total of twelve cattle , two groups of three Alpine brown and three Friesian , intracerebrally inoculated with either BSE or BASE , developed neurological signs and were killed at the terminal stage of disease ( Table 2 ) . In contrast , two saline inoculated Friesian cattle are free of clinical signs at the time of writing , i . e . 42 months post-inoculation . In BASE-treated cattle the clinical disease duration was shorter than in BSE-inoculated animals; however , caution in evaluating these differences is dictated by the low number of experimental animals in addition to the undetermined infectivity titre of the inocula . In BSE-inoculated cattle , clinical signs at onset consisted of behavioural changes and hypersensitivity . As the disease progressed , major clinical signs included aggressiveness , frequent bellowing and head shaking , postural abnormalities , exaggerated blink reflex , generalized cutaneous hyperaesthesia , and stimulus-induced myoclonic jerks ( Table 2 and Figure 1A and Video S1 ) . Conversely , early neurological signs in both Friesian and Alpine brown cattle inoculated with BASE consisted of fasciculations of gluteal muscles , a dull coat and postural and behavioural signs of depression , including low head carriage , mild kyphosis , and decreased alertness . With progression , muscle atrophy , beginning in the gluteal region and progressing to the paravertebral region and to other hind limb musculature became apparent ( Figure 1B and 1C ) . Fore-limb muscles were relatively spared ( Video S2 ) . With the exception of the “downer” cattle , neither gait ataxia nor difficulties in rising were observed throughout the disease course . Cattle showed an exaggerated response to facial touch or pinch , but not to light and sound stimuli . Observations via night filming showed that BASE cattle were prone to sudden falls . One Friesian cow ( code # 254 ) showed a “downer” syndrome at onset . Immunoblot analysis of proteinase K-treated ( PK ) brain homogenates from each BSE- and BASE-infected cattle revealed the presence of PrPTSE in all sampled areas . However , all BSE-challenged animals showed a di-glycosylated-dominant PrPTSE type , whereas in all BASE-inoculated cattle a mono-glycosylated-dominant PrPTSE type was detected . In addition , the molecular mass of the PK-resistant unglycosylated fragment was identical to that of the original inoculum in each animal ( Figure 2A–2C ) . In cattle infected with BSE , the highest amounts of PrPTSE were observed in the thalamus , basal ganglia , obex , olfactory areas and hippocampus , whereas very low amounts were seen in cerebral cortices and cerebellum ( Figure 2D and 2E ) . Differently from BSE , in BASE-infected cattle consistently high amounts of PrPTSE were observed in cerebral cortex , hippocampus , and cerebellum ( Figure 2F and 2G ) . In both groups , low amounts of PrPTSE were found in the spinal cord . In all experimentally infected animals , no PrPTSE was detected in peripheral tissues , including cervical and mesenteric lymph nodes , spleen , thymus , liver , lung , peripheral nerves and forelimb and hind limb muscles , either by standard Western blot analysis or following phosphotungstic acid precipitation . Typical neuropathological changes , including spongiosis and gliosis were detected in all cattle ( Figure S1 and S2 ) . The conventional lesion profile , based on vacuolation score , was similar in BSE- and BASE-infected cattle; however , a more severe involvement of central grey matter ( periaqueductal grey ) and rostral colliculus but not the vestibular nuclear complex were observed in BASE-inoculated cattle as compared to BSE-challenged animals , which showed severe involvement of the putamen ( Figure S1 ) . Additional brain areas , including the olfactory areas , amygdalae , hippocampi and dorsal horns of spinal cords , were severely involved in both groups . Ventral and dorsal roots did not show major pathological changes . Friesian and Alpine brown muscle tissue was normal in BSE-infected cattle ( Figure 3A , 3C , 3E and 3G ) , whereas groups of atrophic muscle fibers were observed in the gluteus medius ( Figure 3B , 3D and 3F ) and , to a decreasing extent , in major psoas , longissimus dorsi , and triceps brachii of BASE-infected cattle ( Figure 3H ) . In BSE cattle , a synaptic-punctate and “glial-associated” stellate pattern of PrP deposition was observed in different brain areas , including olfactory areas , cerebral cortex , basal ganglia , thalamus , cerebellum , medulla , and spinal cord ( Figure 4A and inset , 4E , 4G , 4I and inset ) . Conversely , in BASE-inoculated cattle , abundant amyloid PrP plaques were observed in subcortical white matter and in deep grey nuclei , as observed in natural BASE cases ( Figure 4B and inset , 4F; and Figure S2 and S3 ) . No PrP plaques were seen in the olfactory glomeruli , the cerebellum or the spinal cord ( Figure 4I and 4J ) . Neurons from BSE cattle showed intracellular PrP deposition in contrast to the membrane-associated deposits observed in neuronal cells of BASE cattle ( Figure 4C and 4D ) . These patterns of PrP neuronal staining were also seen in ventral horn neurons of the spinal cord ( Figure 4I and J insets ) and in the dorsal root ganglion cells ( data not shown ) . No PrP staining was detected in the peripheral nerves and muscles .
In the present work , we demonstrate that BSE and BASE isolates maintain distinct biological properties and induce different disease phenotypes after transmission in their natural host . The similarity of the molecular typing differences between BASE and BSE PrPTSE in Friesian and Alpine Brown cattle also supports the notion that the two conditions are caused by different prion strains . Cattle inoculated with BASE developed a syndrome characterized by progressive muscle atrophy and behavioural changes . Amyotrophic changes were preceded by fasciculations , findings denoting a lower motor neuron deficit . The absence of anorexia or difficulty in feeding and swallowing suggests that amyotrophy may be caused by motor neuron dysfunction and , therefore , not indicative of a generalized wasting syndrome , such as that observed in chronic wasting disease [15] . Consistent with clinical findings of lower motor neuron involvement , pathological examination of muscle tissues disclosed groups of atrophic fibers more frequently detected in proximal than distal hind limb muscles . However , there was no convincing loss of ventral horn neurons . Pathogenic mechanisms leading to motor neuron dysfunction remain unknown; however , the role of pathological PrP deposition at the plasma membrane of motor neurons or a loss of PrP function , as observed in experimental models of amyotrophic lateral sclerosis , cannot be ruled out [16] . Clinical signs of motor neuron dysfunction , including stiffness , posterior paresis with “clonic spasms of muscle bundles” [17] and generalized weakness , accompanied by severe lethargy and ataxia , were previously reported in cattle experimentally infected with American strains of sheep scrapie ( either at first or at second passage ) . Cattle inoculated with pre-1975 and post-1990 sources of sheep scrapie from the UK presented similarly with ataxia and weakness and most showed dullness with low head carriage and did not over react to external stimuli [17]–[19] . While the clinical characterization described previously in cattle infected with scrapie is suggestive of upper and lower motor neuron involvement , results obtained in BASE cattle point to lower motor neuron dysfunction or to peripheral neuropathy as the cause of amyotrophic changes . In contrast to the amyotrophic changes observed in BASE-inoculated cattle , animals inoculated with BSE presented a disorder characterized by apprehension and hypersensitivity to external stimuli , overlapping clinical features described in early accounts of UK BSE [20] . Molecular features of PrPTSE from BASE and BSE donor cattle were preserved with high fidelity in recipient animals . In particular , the conformation of PrPTSE , as assessed by the electrophoretic motility of the core fragment , and the glycosylation status were indistinguishable in recipient animals of different breeds compared to the original inocula . These PrPTSE traits were maintained in all cortical and subcortical investigated brain regions . Taken together , PrPTSE molecular traits and PrPTSE regional distribution showed distinct patterns in the two groups of animals , supporting the notion of two different prion strains , as also suggested by results from experimental transmission to transgenic mice expressing bovine PrP [10] , [13] . Differences between BASE- and BSE-inoculated animals were also observed at the neuropathological and immunohistochemical levels . At variance with original BASE cases , where no spongiform changes were observed , as a likely effect of early disease stage , marked vacuolation was seen in BASE-inoculated cattle with a lesion profile divergent from that seen in BSE-treated animals in at least four regions . Indeed , extensive vacuolar pathology was seen in the hindbrain of BASE-inoculated cattle , whereas severe involvement of the putamen was a distinguishing feature in the BSE group . The divergent biological properties of the two strains were further confirmed by the different patterns of PrP deposition , indistinguishable from those patterns seen in naturally occurring BSE and BASE . Moreover , the distinct neural and microglial cells involved in the two groups and the subcellular sites of PrP accumulation denote a different trafficking and propagation of PrP . Taken together , intraspecies transmission of BASE and BSE recapitulated the key neuropathological hallmarks observed in these naturally occurring cattle TSEs . This is at variance with the alternate patterns of PrPTSE depositions seen in inoculated Tgbov XV mice , i . e . , uni- and multicentric plaques in BSE-challenged animals and diffuse/focal PrP deposition , but not amyloid plaques , in BASE-inoculated mice [10] , [13] . We recently showed that in TgBov XV mice challenged with the same inocula used in the present study , BASE-inoculated mice had significantly shorter incubation periods and survival times than BSE-inoculated mice , consistent with results from another laboratory [10] , [13] . This effect was not influenced by any species barrier phenomenon and is therefore likely to be strain-dependent . An exception to diverging phenotypic characteristics observed in BSE- and BASE-inoculated cattle was the incubation time observed for BSE-inoculated Alpine brown ( but not BSE-treated Friesian animals ) which did not significantly differ from times assessed for BASE-inoculated Friesian and Alpine brown cattle . However , the small number of investigated animals and the individual variability in incubation times dictate caution in the interpretation of these data . In contrast , the breed-associated effect in BASE-inoculated cattle , with significantly shorter incubation periods and survival times in Friesian than in Alpine brown , suggests that disease-modifier genetic loci other than known PRNP polymorphisms could be relevant to both of these parameters . While it is now clear that vCJD originated from human exposure to BSE , it is still uncertain whether emerging cattle TSEs , including BASE , or L-BSE , and H-BSE have infected humans or to which extent they can be potentially dangerous for human and animal health . Recent experimental data show that the BASE strain is efficiently transmitted to Tgbov XV mice and to TgOv mice; in the latter , BASE transmits at first passage with a 100% attack rate , as opposed to cattle BSE that transmits with a low attack rate [21] . Moreover , transmitted BASE shows shorter incubation periods than BSE in Cynomolgus monkeys [14] . Paradoxically , while BASE is efficiently transmitted at first passage and with a high attack rate to 129 Met/Met Tg humanized mice [22] , human transgenic lines of all genotypes at codon 129 are resistant to BSE transmission [22] , [23] . Taken together , these data might suggest that the BASE agent could transmit to humans more efficiently than the BSE agent .
All procedures involving animals and their care were conducted in conformity with national and international laws and policies ( EEC Council Directive 86/609 , OJL358 , 1 , 12 December 1987; Italian Legislative Decree 116/92 , Gazzetta Ufficiale della Repubblica Italiana 10 , 18 February 1992; and Guide for the Care and Use of Laboratory Animals , U . S . National Research Council , 1996 ) , and the study was approved by the authors' Institutional Review board . PRNP ORF amplification , sequencing and determination of the octapeptide repeat copy number was performed as previously described [12] . Polymorphisms of the 12-bp indel , located within intron 1 , and 23-bp indel , located in the promoter region , were determined as previously reported [24] . Primer pairs 5′-CCTGTTGAGCGTGCTCGT/5′-ACCTGCGGCTCCTCTACC-3′ and 5′-GAAGTCACGTGAAGGCACT-3′/5′-CAAAGAGTTGGACAGGCACA-3′ were used to amplify the 12-bp indel ( 202 bp/214 bp ) and 23-bp indel ( 167 bp/190 bp ) , respectively , as described above . PCR was performed as 30 cycles of 30 sec at 94°C , 30 sec at 55°C and 45 sec at 72°C . High resolution agarose ( 3 . 5% ) gel electrophoresis was used to visualise the allelic PCR products whose specificity and length was also confirmed by direct sequencing with the same primer used for the PCR described above . 10% brain homogenates from the thalamus of a BSE-affected Friesian ( code #128204 ) and a BASE-affected Piemontese ( code #1088 ) were prepared in phosphate-buffered saline . These cattle carried the same PrP genotype with six octapeptide repeats , and were extensively studied in our previous work [12] . BSE and BASE inocula were prepared to obtain a comparable amount of PrPTSE as assessed by Western blot analysis with the 6H4 anti-PrP monoclonal antibody ( Prionics ) . Eight Friesian and six Alpine brown cattle ( 4 months old ) were purchased from Italian herds in which no cases of BSE had ever been recorded . All calves were free of neurological signs . Prior to inoculation , animals remained in the new environment for one month for adaptation . Inoculation was carried out using a semi-stereotaxic technique in surgical aseptic conditions . Calves were anesthetized with xylazine ( 50 μg/kg ) , a midline incision was made at the junction of the parietal and frontal bones , and a 1-mm hole was drilled through the calvarium . The inoculum was injected into the frontal lobe via a 22-gauge 9-cm-long disposable needle while the needle was withdrawn . Two groups of animals , each comprising three Friesians and three Alpine brown cattle , were inoculated with 1 milliliter of 10% brain homogenate from BSE-and BASE-affected animals , respectively . Conversely , two Friesians cattle were challenged with 1 milliliter of phosphate buffered saline and used as controls . To avoid potential cross-contamination , BSE and BASE transmission experiments were performed on different days and the facility was decontaminated with 10% sodium hypochlorite solution after each inoculation . Clinical evaluations were comprised of a bi-weekly observation by the veterinarian and two daily observations by animal husbandry staff who reported any observed motor and/or behavioural changes . For assessment of the gait cows were walked along the corridor outside the pen . The cattle were filmed nightly with closed circuit television monitoring to record signs of disease that may not have been observed during the day . Once a month , a veterinarian trained in neurology examined each cattle by conventional neurological scale evaluations [25] , [26] . Animals were considered symptomatic when they showed two of the following neurological signs observed in two separate consecutive examinations: abnormal behaviour , abnormal posture , aberrant reactions , or hyperreactivity to sensitive stimuli , light and sound . Cattle at the terminal stage were euthanized with pentobarbital administered intravenously . Peripheral organs , including cervical and mesenteric lymph nodes , spleen , thymus , liver , lung , peripheral nerves and forelimb and hind limb muscles ( m . triceps brachii , m . longissimus dorsi , m . gluteus medius and m . major psoas ) , were sampled and each sample was divided equally; one portion was fixed in 4% buffered formaldehyde for H&E stain and PrP immunohistochemistry , and the other was frozen . Serial 10-μm-thick muscle cryosections were stained with H&E and adenosine triphosphatase ( ATPase ) , after pre-incubation at pH 4 . 3 , 4 . 6 and 10 . 4 . Nervous tissue was removed in a separate area to avoid cross-contamination . The fixed half of the brain sample was used for neuropathological examination , while the frozen brain sample was stored at −80°C for biochemical analyses . The spinal cord was sampled at cervical , thoracic , lumbar and sacral levels and sections were fixed in 10% buffered formaldehyde . The remaining tissue was frozen at −80°C for further studies . From each neural tissue sample , including optic nerve , olfactory bulb , frontal cortex , occipital cortex , hippocampus , nucleus caudatus , putamen , globus pallidus , thalamus , cerebellum , obex , and cervical , thoracic , lumbar and sacral spinal cord , 100 mg of wet tissue was homogenized in 9 volumes of lysis buffer ( 100 mM sodium chloride/10 mM EDTA/0 . 5% Nonidet P-40/0 . 5% sodium deoxycholate/10 mM Tris , pH 7 . 4 ) and digested with 50 μg/ml of proteinase K ( Boehringer Mannheim ) for 1 h at 37 °C . Digestion was blocked by the addition of phenylmethylsulfonyl fluoride at 2 mM . For deglycosylation , proteinase K-digested samples were deglycosylated with recombinant peptide N-glycosidase F ( PNGase F ) according to manufacturer's instructions ( Boehringer Mannheim ) . Samples , equivalent to 400 μg of wet tissue , were resolved on 13% polyacrylamide gels and then transferred onto PVDF membrane ( Immobilon P; Millipore , Bedford MA ) for 2 hours at 60V . Membranes were blocked with 1% non-fat dry milk in TBST ( 10 mM Tris/150 mM sodium chloride/0 . 1% Tween-20 , pH 7 . 5 ) for 1 hour at 37°C and incubated overnight at 4°C with anti-PrP monoclonal antibody 6H4 ( Prionics ) diluted to 1/5 , 000 . Blots were developed using the Amersham enhanced chemiluminescence ( ECL ) system , as described by the supplier and visualized on an autoradiographic film . Films were scanned by using a densitometer ( GS-710 , Biorad ) , calculating the relative amounts of PrPSc in a semiquantitative manner . To enhance PrPTSE detection , extraneural tissues , including cervical and mesenteric lymph nodes , spleen , thymus , liver , lung , peripheral nerves and forelimb and hind limb muscles , that were negative on a standard immunoblot were subjected to phosphotungstic acid ( PTA ) precipitation and analyzed by Western blot , as previously described [27] . Briefly , 100 mg of wet tissue were homogenized in nine volumes of 2% sarkosyl in phosphate-buffered saline , pH 7 . 4 . Cellular debris were removed by centrifugation at 1 , 000 rpm for 2 minutes and samples were incubated for 30 minutes at 37°C with constant agitation in phosphate-buffered saline containing 50 units/mL Benzonase and 1 mmol/L magnesium chloride . Subsequently , samples were adjusted to 0 . 3% sodium phosphotungstic acid , incubated at 37°C for 30 minutes and centrifuged at 14 , 000 rpm for 30 minutes . The supernatant was saved , and the pellet dissolved in 20 μl phosphate-buffered saline , pH 7 . 4 , containing 0 . 1% sarkosyl . The supernatant and the pellet were adjusted to a final concentration of 20 μg of proteinase K per milliliter and incubated at 37 °C for 30 minutes . Three-mm thick samples were embedded in paraffin after decontamination with 98% formic acid for 1 hour . The paraffin-embedded blocks selected for the study included coronal sections at the level of the olfactory bulb , the frontal , parietal and occipital cortices , the pyriform lobus , hippocampus , striatum , thalamus , brainstem , sagittal sections through the cerebellum and spinal cord at cervical , thoracic , lumbar and sacral levels . Histological sections were deparaffinized , rehydrated , and stained with hematoxylin and eosin . Additional sections were stained with thioflavin S . The distribution of spongiosis , was determined by using a conventional lesion profile , which allows to characterize strain tropism and to compare the present results with those of previous studies on field and experimental BSE [28] . The definition of each score was performed by three independent observers blinded to the animal identification , as follows: 0 no vacuolation , 1 a few vacuoles ( minimum 3 per field×10 objective ) , 2 several vacuoles evenly distributed , 3 moderate numbers or many vacuoles evenly distributed , and 4 numerous vacuoles some of which coalescing , as previously described [28] . For immunohistochemical study , sections obtained form nervous and extraneural tissues , were rehydrated and treated with 98% formic acid for 20 min at room temperature , followed by hydrated autoclaving in distilled water at 121°C for 30 min . After rinsing , sections were incubated overnight at 4°C with anti-PrP monoclonal antibody F99/97 . 6 . 1 ( VMRD , inc . ; diluted to 1/1 , 000 ) , recognizing a conserved epitope ( QYQRES ) on the cattle PrP . Subsequent antibody detection was carried out using a biotinylated goat anti-mouse secondary antibody diluted to 1/200 for 20 min ( Vector Laboratories , Burlingame , CA ) at room temperature , followed by the avidin-biotin-peroxidase complex ( Vectastain ABC kit; Vector Laboratories ) according to manufacturer's protocol . Immunoreactivity was visualized using 3 , 3′-diaminobenzidine as chromogen .
|
For approximately two decades , bovine spongiform encephalopathy ( BSE ) , now termed classical BSE ( BSE-C ) , has been regarded as the only and exclusive prion disorder affecting cattle . However , over the last 4 years , two additional bovine prion strains , bovine amyloidotic spongiform encephalopathy ( BASE , also named BSE-L ) and BSE-H , have been discovered and characterized in Canada , the United States , Japan , and nine European countries , which applied an active surveillance program on slaughtered cattle . Although a total of 20 BSE-L and 16 BSE-H have been reported to date , the disease phenotype of these conditions remains largely unknown . Intriguingly , recent evidence has been provided that the BSE-C and BASE strains disclose converging properties after transmission to inbred mice . Here , we show that intraspecies transmission of BASE induces a disease phenotype characterized by dullness and progressive amyotrophy , the latter highly suggestive of a motor neuron disorder . This is at variance with the over-reactivity and hypersensitivity , but not muscle changes , observed in BSE-transmitted cattle . This study confirms that BASE and BSE represent two distinct prion disorders in cattle with diverging molecular features and disease phenotypes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neurological",
"disorders/prion",
"diseases"
] |
2008
|
Intraspecies Transmission of BASE Induces Clinical Dullness and Amyotrophic Changes
|
Ornaments used in courtship often vary wildly among species , reflecting the evolutionary interplay between mate preference functions and the constraints imposed by natural selection . Consequently , understanding the evolutionary dynamics responsible for ornament diversification has been a longstanding challenge in evolutionary biology . However , comparing radically different ornaments across species , as well as different classes of ornaments within species , is a profound challenge to understanding diversification of sexual signals . Using novel methods and a unique natural history dataset , we explore evolutionary patterns of ornament evolution in a group—the birds-of-paradise—exhibiting dramatic phenotypic diversification widely assumed to be driven by sexual selection . Rather than the tradeoff between ornament types originally envisioned by Darwin and Wallace , we found positive correlations among cross-modal ( visual/acoustic ) signals indicating functional integration of ornamental traits into a composite unit—the “courtship phenotype . ” Furthermore , given the broad theoretical and empirical support for the idea that systemic robustness—functional overlap and interdependency—promotes evolutionary innovation , we posit that birds-of-paradise have radiated extensively through ornamental phenotype space as a consequence of the robustness in the courtship phenotype that we document at a phylogenetic scale . We suggest that the degree of robustness in courtship phenotypes among taxa can provide new insights into the relative influence of sexual and natural selection on phenotypic radiations .
Adaptive radiations are driven by ecological differences that promote processes of diversification and speciation [1] . In contrast , phenotypic radiations , which occur in the absence of clear ecological differentiation , are less well understood . One commonly investigated mechanism for phenotypic diversification among ecologically similar taxa is variation in social and sexual selection pressures promoting signal or ornament diversification . Ornamental radiations may come about as a consequence of variation in signaling environment [2 , 3] , sensory capabilities [4 , 5] , or pseudorandomly via mutation-order selection [6 , 7] or Fisher-Lande-Kirkpatrick processes [8–11] . Most studies investigating patterns of ornamental diversification have focused on individual trait classes and simplified axes of variation; however , sexual selection does not act on single traits in isolation . A more complete understanding of the processes driving ornamental diversification is possible only by investigating evolutionary relationships between the full suites of ornamental traits under selection . Many animals rely on multiple ornamental traits to attract mates . Advantages of multiple ornaments may include increased information transfer ( multiple messages ) , increased reliability ( redundancy ) , increased flexibility ( ensuring information transfer across contexts and environments ) , and increased memorability/discriminability [12–16] . Multiple ornaments may be more common when costs associated with the display or evaluation of those ornaments are low [17] , as is likely the case in lekking species [12] . Though we now have broad empirical support for many of the proposed adaptive benefits of multiple signals at the level of individual species , how these specific hypotheses map onto our understanding of phylogenetic patterns of ornament evolution is less clear . Gaining insights into the macroevolutionary patterns of multiple ornament evolution is challenging , in part , owing to the difficulties of comparing highly divergent phenotypic traits across species . For instance , even focusing on evolutionary patterns of a single trait ( e . g . , plumage color in birds ) across species can be difficult when traits possess different axes of variation ( e . g . , red versus blue ) . Though ingenious new methods have been devised to compare highly divergent ornaments of a single signal type ( e . g . , plumage color [18] , electrical signals [19] , or song [20] ) , comparing ornamental complexity across signal types presents yet an additional layer of complication . However , understanding the interrelationships of different classes of ornaments across phylogenetic scales can potentially provide valuable information about the evolutionary processes of communication , phenotypic radiation , and speciation that cannot be gathered from single-trait or single-species studies . Following the evolution of multiple ornaments , selective pressures may favor different interrelationships among signal types . If ornamental investment is governed by evolutionary tradeoffs , investment in one class of ornaments will come only at the expense of investment in another . Evidence suggests that signal tradeoffs manifest as a negative correlation among ornament types across evolutionary time [21–27] , reflecting strong , consistent constraints imposed by ecology , physiology , and natural selection [28 , 29] . Alternatively , instances in which ornamental traits show no evolutionary relationships [30–35] suggest long-term patterns of independent evolutionary trajectories . In such cases , signals are functionally independent and may even have evolved for use in different contexts ( e . g . , territorial defense versus mate attraction ) . When might we expect positive correlations among ornament classes across species ? Theoretical [12 , 36] and empirical [37 , 38] work suggests that positive correlations among signals across species may reflect consistent selection acting similarly on separate axes of ornamental evolution . Strong , consistent intersexual selection could generate these positive correlations ( sensu [37] ) , especially if the signals convey separate information [36] , resulting in functional integration among ornament elements [39 , 40] . In such cases , positive correlations among signals across species would arise when selection favors an “integrated whole” of ornamental traits [41 , 42] , which we call the “courtship phenotype . ” The courtship phenotype is the composite expression of all ornamental classes evaluated during courtship and may represent the composite target of selection . Evolution may favor integrated , holistic mate evaluation strategies because of advantages that sensory overlap and redundancy offer ( e . g . , increased accuracy ) [12–16] . Here , we examine broad evolutionary patterns of ornamental signal investment and complexity across the wildly diverse [43 , 44] , monophyletic [45] birds-of-paradise ( Paradisaeidae ) “…in which the process of sexual selection has gone to fantastic extremes” [46] ( Fig 1A ) . We focus on the birds-of-paradise because this family exhibits extreme variation across species in multiple ornamental axes [43] ( e . g . , color [47–49] and behavior [50 , 51] ) while possessing broadly similar life histories and mating systems [43 , 44] . Consequently , insights about the strength , direction , and diversification of ornamental phenotypes in this group may shed light on key processes of sexual selection and its power to generate phenotypic radiation when natural-selection–imposed constraints are minimized . In this study , we use a unique natural history dataset to quantitatively evaluate behavioral , acoustic , and colorimetric ornamentation across 40 species of birds-of-paradise , as well as relationships between signals and display environment .
Comparisons across signal types are inherently challenging for evolutionary biologists given that such signals are necessarily measured in different ways . Additionally , comparisons within color , acoustic , and behavioral repertoires across taxa that vary widely ( e . g . , the birds-of-paradise ) present an additional methodological challenge: how does one compare phenotypes that may share no obvious overlapping characters ? We addressed this obstacle with a two-pronged approach to quantify ornamental complexity for behavior , color , and sounds in the birds-of-paradise . First , we broke down each ornament into a taxonomically unbounded character space that allowed classification of subunits across all species . Second , we used the specific attributes of a given ornament for each individual for each species to categorize the ornament components before quantifying two conceptually aligned measures of complexity for each signal type . Specifically , we evaluated richness ( the number of unique elements ) and diversity ( using an index dependent on the number and relative contribution of each element type ) using phylogenetic comparative approaches ( see Methods for additional details ) . For behavioral analyses , we first broke down the courtship behaviors of all species into distinct subunits shared across species ( e . g . , S1 Video ) . We then analyzed composite behavioral sequences across time using sliding-window analyses to compare maximally diverse behavioral repertoires for a set duration across species ( Fig 1B ) . For colorimetric analyses , we relied on visual modeling of multispectral images to quantify the number and relative abundances of perceptually distinct color types across individuals and species . Though different colors may have different underlying production mechanisms , our analyses simply focused on the number and distribution of distinguishable colors ( Fig 1C ) . Similar to our behavioral analysis pipeline , we used acoustic properties and agglomerative clustering to classify distinct sound types used by birds-of-paradise in courtship contexts before employing a similar sliding-window analysis to identify maximally diverse acoustic sequences , facilitating comparisons across species ( Fig 1D ) . In total , we analyzed 961 video clips , 176 audio clips , and 393 museum specimens . From these analyses , we obtained quantitative diversity and richness metrics of ornamental complexity across the birds-of-paradise ( Fig 2 ) , which allowed us to rigorously evaluate patterns of correlated character evolution , as well as facilitating our investigation of the influence of breeding system and display environment on ornamental complexity . Using multiple phylogenetic generalized least squares ( mPGLS ) analyses , which allowed us to control for the nonindependence of species due to their shared evolutionary history , as well as the potentially confounding influences of display environment ( both display height and proximity to courting conspecific males ) , we uncovered positive correlations between color and acoustic diversity ( Fig 3A ) , as well as between behavioral and acoustic diversity ( Fig 3B ) , consistent with the hypothesis that selection has acted similarly on these axes of ornamental complexity . Interestingly , however , there was no significant relationship between color diversity and behavioral diversity , indicating independent evolutionary trajectories for these visually encoded aspects of courtship ornamentation ( S1 Table and S1 Fig ) . Analyses of ornamental richness revealed the same pattern to those uncovered for ornamental diversity ( S2 Table ) . Specifically , behavioral richness and acoustic richness were correlated , as were color richness and acoustic richness ( as was the case for both relationships involving ornamental diversity ) . Behavior richness and acoustic richness , but not color richness , were influenced by stratum of the forest in which species display ( Fig 4A–4C ) . Specifically , we found that behavioral richness exhibited a negative relationship with display height among birds-of-paradise , such that species that display on the forest floor had the largest behavioral repertoires ( S2 Table and Fig 4B ) . Species that display on the forest floor are typically operating with lower-light environments , and consequently , these species appear to rely more heavily on complex dance sequences to attract mates . Additionally , birds-of-paradise show increased acoustic ( Fig 4C ) richness as their display locations increase in height ( S2 Table ) , a result that partially corresponds to the predictions of sensory drive [52 , 53] whereby the openness of the upper canopy favors increasingly complex acoustic displays . Similar to the patterns we uncovered for signal richness , we also found that behavioral diversity and acoustic diversity were influenced by display height ( S1 Table ) . Species displaying in the forest understory exhibiting a marginally significant ( p = 0 . 051 ) trend for greater acoustic diversity relative to ground displaying species , and the behavioral diversity for ground-displaying species was higher than for both understory and canopy species ( S1 Table ) . However , color diversity was not significantly influenced by display height . Birds-of-paradise that display in classic leks have greater color richness ( Fig 4D and S2 Table ) , corresponding to the increased strength of sexual selection on males to “stand out” visually when being evaluated simultaneously in lekking contexts . However , neither behavioral nor acoustic richness were significantly associated with the spatial distribution of displaying males . Furthermore , none of the diversity metrics ( color , behavior , sound ) were significantly associated with the breeding system structure ( S1 Table ) .
Our study provides evidence that selection has favored correlated levels of ornamental diversity across multiple signals among the birds-of-paradise . This pattern of positive correlation among distinct ornament classes across evolutionary timescales and species suggests strong sexual selection on functionally integrated courtship phenotypes . The degree to which phenotypic traits are coexpressed and functionally dependent upon one another can be referred to as functional integration [54] or interdependence [40] . Courtship phenotypes with greater functional integration are therefore composed of ornaments that are typically expressed at similar levels and that are mutually interdependent in order to influence mate choice [39 , 41 , 42] . Correlations among the signals that comprise the courtship phenotype also suggest a previously undescribed robustness in bird-of-paradise courtship phenotypes that may have played a key role in the extreme ornamental radiation exhibited by this taxon ( Fig 1 ) . Evolutionary biologists dating back to Mayr [55] and even Darwin [56] have recognized the potential evolutionary implications of functional redundancy ( two or more structures performing the same function ) . Functional redundancy , including “true” redundancy ( i . e . , structurally identical components with identical functions ) and degeneracy ( i . e . , structurally distinct components with similar functions ) [57] , facilitates evolutionary innovation ( i . e . , increases “evolvability” ) by increasing robustness . Robust systems are those in which the overall structure and interconnectedness of parts provide protection from environmental or mutational instability [58] such that a given function is not lost if a single component fails . Robustness increases evolvability by enabling elements to react to selection independently and diverge while maintaining original functions [57 , 59] . All redundancy ( both “true” redundancy and degeneracy ) provides a measure of robustness , but robust systems are not necessarily redundant [60] . Given the broad theoretical [59 , 61 , 62] and empirical [63–65] support for the idea that robustness can promote evolvability across a wide array of biological domains , we posit that the correlations among signal types within birds-of-paradise courtship phenotypes are at least partially responsible for the dramatic diversification and radiation of courtship signals displayed by birds-of-paradise . If female birds-of-paradise make mate choice decisions based on sensory input from the multiple signals that comprise a composite courtship phenotype and information from those channels is correlated , then novel mutations changing the structure or form of a given ornament may occur without “necessary” information being lost [66] . Consequently , over evolutionary time , we suggest it is the inherent functional overlap ( redundancy/degeneracy ) and structural interdependency ( robustness ) of courtship phenotypes that leads to increased phenotypic diversification ( evolvability ) in birds-of-paradise . Phenotypic radiations in the absence of clear ecological differentiation may arise stochastically [67 , 68] and be heavily influenced by the specific intricacies of female choice [7 , 10 , 69 , 70] . Birds-of-paradise clearly exhibit some ecological differentiation [43] , but broadly speaking , they tend to be heavily frugivorous and predominantly polygynous [71] . They do not , however , all display to potential mates in the same contexts or microenvironments . Some species display high in the canopy , some down on the forest floor , and others in the understory in between . Likewise , some species display in large , cacophonous leks , some species in exploded leks ( exp leks ) in which males can hear but not see one another , and other species display solitarily . Our results suggest that these differences have shaped the specific courtship and signaling strategies of each species ( Fig 4 and S1 and S2 Tables ) . Birds with richer acoustic repertoires display high in the canopy , where there is less environmental interference ( e . g . , from cluttered branches ) , increasing the likelihood that females will be able to detect and discern numerous , elaborate sounds [53] . Likewise , more behaviorally complex birds tend to display near the forest floor where there is less light ( and ability to perceive subtle variation in color ) but more area available for a courtship stage or “dance floor . ” Birds that display in true leks have more colorful plumage , perhaps because females need to identify attractive individuals based on relatively unchanging traits , allowing them to compare among multiple displaying males simultaneously . Display site and display context thus influence the specific forms of ornamentation possessed by individual species [72] , and taking them into account from an analytical perspective allows us to better understand patterns of signal coevolution and the potential importance of a functionally integrated courtship phenotype . Signal efficacy and information content can exert strong influence on receiver preferences , and understanding both elements is integral when examining the evolution of complex , multicomponent courtship phenotypes [14 , 70 , 73 , 74] . The influence of receiver preference is difficult to overstate , particularly in birds-of-paradise , for which recent work indicates that selection acting on female preferences controls the rate , extent , and phenotypic space available for ornamental radiations [70] . Importantly , receiver preferences are influenced by the perceptual abilities [75 , 76] and psychology of signal receivers [77 , 78] , as well as the environments through which signals are transmitted [52]—all of which can markedly influence signal efficacy . Additionally , the information content of multiple signals may increase the net amount of information transferred ( e . g . , multiple messages [16] ) or increase accuracy and reliability if multiple signals communicate the same message ( e . g . , redundant signals [12 , 16] ) . The perceptual channels by which birds-of-paradise attract mates and those channels that are correlated at a phylogenetic scale provide tantalizing , though tentative , insights into the processes of efficient information transfer and receiver stimulation regulating mate choice in this group . Specifically , the fact that significant positive correlations exist between acoustic and color signals ( auditory , visual ) , and between acoustic and behavioral signals ( auditory , visual ) , but not between color and behavioral signals ( visual , visual ) aligns with psychometric literature on information and sensory input . When multiple sources of information are provided , information may be maximized if that information comes from separate channels ( e . g . , acoustic , visual ) and lost when arriving through a single sensory channel [79] ( but see [80] ) . What exactly this “information” might be in birds-of-paradise ( quality [81] , attractiveness [69] , motivation [82] , etc . ) is not clear , but this result provides an interesting starting point for future investigations . Phylogenetic comparative investigations of animal signals hold the potential to answer important questions about the evolutionary trajectories of communication over time [83 , 84] . However , the data used to tackle key questions of signal evolution necessarily place upper and lower bounds on the confidence and interpretations one can make from such comparative studies . It is our hope that the novel approaches we have developed to quantify color , sound , and behavior will be useful to other researchers interested in understanding signal variation at different scales . Though our primary aim was to generate methodological pipelines that facilitated comparisons among the highly divergent birds-of-paradise , the basic framework we describe here may also be useful for comparisons of more similar taxa—including studies of intraspecific variation in signaling effort ( e . g . , through sliding-window analyses focused on bouts of maximal complexity ) or investment ( e . g . , by using receiver visual models to identify the number and perceptual similarity of color patches across individuals ) . Consequently , we feel that our approaches complement recent suggestions for incorporating a systems biology approach to the study of animal communication [57] , wherein more comprehensive , higher-resolution data will only improve the validity and interpretability of analyses incorporating fitness surfaces and communication networks . Evolutionary tradeoffs—increases in trait expression linked to reductions in another—are ubiquitous: “If there were no tradeoffs , then selection would drive all traits correlated with fitness to limits imposed by history and design” [85] . Tradeoff thinking can inform our interpretations of both the marked interspecific variation in overall signal complexity ( Fig 2 ) and the finding that the ornaments of birds-of-paradise are positively correlated at phylogenetic scale ( Fig 3 ) . Firstly , interspecific variation in overall signal complexity suggests tradeoffs between investment in courtship and some other , unmeasured , variable that differs across species ( e . g . , microenvironment , paternal care , resource competition , etc . ) . Secondly , the absence of tradeoffs among signal types indicates an absence of differential costs on acoustic , behavioral , and chromatic signals . Further , the correlation among ornamental classes suggests that selection is acting on functionally integrated courtship phenotypes for birds-of-paradise , a finding that indicates female birds-of-paradise make mate choice decisions incorporating holistic , multicomponent information sets comprised of the various ornaments possessed by males of their species . Rather than being unique to birds-of-paradise , however , we suggest that this phenomenon is widespread among animals—though it is by varying degrees constrained , impeded , or obfuscated by conflicting and constraining processes and limitations imposed by ecology and natural selection . The degree to which selection has facilitated the evolution of integrated , robust courtship phenotypes may in fact serve as a proxy for the overall strength and consistency of female-driven sexual selection in any taxa , for which the integration and correlation among ornaments comprising the courtship phenotype may shed important light on the history and strength of sexual selection in that particular group .
The study was focused on vertebrates ( birds-of-paradise ) but used museum specimens ( physical and media ) , so no IACUC protocol was required . We quantified the behavioral complexity of courtship display behaviors for the birds-of-paradise by scoring field-recorded video clips of 32 ( 80% ) paradisaeid species , primarily from the Macaulay Library at the Cornell Lab of Ornithology ( macaulaylibrary . org , S3 Table ) . In total , we watched 961 clips from 122 individuals totaling 47 , 707 . 2 s ( approximately 795 . 12 min; mean clip duration = 49 . 64 s ) . Courtship display behavior is highly variable among bird-of-paradise species , necessitating broad behavioral categories to facilitate investigations of behavioral evolution . Specifically , one of us ( CDD ) blindly evaluated video clips of male birds-of-paradise displaying species-typical courtship behaviors [43] using a customized ethogram of behavioral units that enabled us to quantify all state and event behaviors exhibited by all species of Paradisaeidae ( S4 Table ) . As with display behaviors , we quantified the acoustic complexity of courtship sounds produced by analyzing field-recorded audio/video clips of 32 ( 80% ) bird-of-paradise species . In total , we analyzed sound from 176 clips from 59 individuals totaling 24 , 670 . 9 s ( approximately 411 min; mean clip duration = 140 . 18 s; S8 Table ) . Though birds can generate sounds ( both vocally and mechanically ) in numerous contexts , we focused our analysis on recordings from known display sites or those matching written descriptions of courtship sound production [43] .
|
Animals frequently vary widely in ornamentation , even among closely related species . Understanding the patterns that underlie this variation is a significant challenge , requiring comparisons among drastically different traits—like comparing apples to oranges . Here , we use novel analytical approaches to quantify variation in ornamental diversity and richness across the wildly divergent birds-of-paradise , a textbook example of how sexual selection can profoundly shape organismal phenotypes . We find that color and acoustic complexity , along with behavior and acoustic complexity , are positively correlated across evolutionary timescales . Positive links among ornament classes suggests that selection is acting on correlated suites of traits—a composite courtship phenotype—and this integration may be partially responsible for the extreme variation in signal form that we see in birds-of-paradise .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"taxonomy",
"organismal",
"evolution",
"acoustics",
"ecology",
"and",
"environmental",
"sciences",
"social",
"sciences",
"vertebrates",
"animals",
"animal",
"phylogenetics",
"phylogenetics",
"data",
"management",
"animal",
"behavior",
"zoology",
"ecological",
"metrics",
"computer",
"and",
"information",
"sciences",
"animal",
"communication",
"birds",
"behavior",
"bioacoustics",
"evolutionary",
"systematics",
"species",
"diversity",
"physics",
"psychology",
"eukaryota",
"ecology",
"animal",
"evolution",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"evolutionary",
"biology",
"amniotes",
"organisms"
] |
2018
|
Evolution of correlated complexity in the radically different courtship signals of birds-of-paradise
|
Our ability to recreate complex biochemical mechanisms in designed , artificial systems provides a stringent test of our understanding of these mechanisms and opens the door to their exploitation in artificial biotechnologies . Motivated by this philosophy , here we have recapitulated in vitro the “target sequestration” mechanism used by nature to improve the sensitivity ( the steepness of the input/output curve ) of many regulatory cascades . Specifically , we have employed molecular beacons , a commonly employed optical DNA sensor , to recreate the sequestration mechanism and performed an exhaustive , quantitative study of its key determinants ( e . g . , the relative concentrations and affinities of probe and depletant ) . We show that , using sequestration , we can narrow the pseudo-linear range of a traditional molecular beacon from 81-fold ( i . e . , the transition from 10% to 90% target occupancy spans an 81-fold change in target concentration ) to just 1 . 5-fold . This narrowing of the dynamic range improves the sensitivity of molecular beacons to that equivalent of an oligomeric , allosteric receptor with a Hill coefficient greater than 9 . Following this we have adapted the sequestration mechanism to steepen the binding-site occupancy curve of a common transcription factor by an order of magnitude over the sensitivity observed in the absence of sequestration . Given the success with which the sequestration mechanism has been employed by nature , we believe that this strategy could dramatically improve the performance of synthetic biological systems and artificial biosensors .
In order to test the extent to which we understand complex biochemical systems -and to improve our ability to exploit them in man-made technologies ( e . g . , synthetic biology; biosensors ) - it is important to reconstruct these processes in the laboratory . Illustrative examples of this include recent demonstrations of synthetic genetic networks in which genetic elements are “mixed and matched” in order to create artificial bistable “toggle switches , ” genetic oscillators and other complex , non-linear input/output behaviors ( e . g . , [1]-[5] ) . Other examples include the recent de novo design of proteins , including enzymes , unrelated to any naturally occurring sequences ( e . g . , [6]-[9] ) and the artificial selection of new proteins [10]-[11] . And while these studies do not ( and cannot ) prove that our knowledge of , for example , genetic regulatory networks and protein folding and evolution is exhaustively complete , they nevertheless suggest that our understanding of these naturally occurring systems is sufficient to enable the design of similarly complex , artificial systems [12] . Motivated by the above philosophy , here we recreate in vitro the “sequestration” mechanism thought to underlie the extraordinary sensitivity ( the steepness of the input/output function ) of a number of genetic networks ( e . g . , [5] , [13]-[18] ) . In the sequestration mechanism , low concentrations of a given target molecule are sequestered by binding to a high affinity ( low dissociation constant ) receptor that acts as a “depletant , ” which serves as a “sink” that prevents the accumulation of free target without generating an output signal ( Figure 1a ) . When the total target concentration surpasses the concentration of the depletant ( saturating the sink ) , a threshold response is achieved in which the addition of any further target produces a large rise in the relative concentration of free target ( Figure 1b , top ) . The rapidly rising concentration of free target then binds to –and thus activates– a second , lower affinity ( higher dissociation constant ) receptor ( or “probe” ) that , unlike the depletant , generates an output signal . This threshold effect generates a “pseudo-cooperative” dose-response curve , which is much more sensitive ( much steeper ) than the hyperbolic “Langmuir isotherm” produced by simple , single site binding ( Figure 1b , Bottom ) [17]-[19] . ( At this juncture we must note an important semantic distinction . The sensitivity of biological systems , such as metabolic networks or signal transduction pathways , is defined as the ratio of the relative change in output to the relative change in input [15] , [19] . The term ultrasensitivity thus describes systems for which the upstroke of the input/ouput function is steeper than the simple , hyperbolic curve obtained for single site binding , such as is observed for “classic” Michaelis-Menten enzymes [15] , [18] . This definition of sensitivity is distinct from “analytical sensitivity , ” which represents the smallest input ( rather than the smallest change in input ) that the method is capable of resolving above the noise floor . Indeed , as we show here , ultrasensitive behaviour is often produced at the cost of a reduced analytical sensitivity as a steeper input/output function is often achieved at the cost of increasing the smallest input signal that can be robustly detected . In this paper we use only the former , steep-input/output-function definition of the terms “sensitivity” and “ultrasensitivity . ” ) . Sequestration is thought to underlie the ultrasensitive responses of many cellular processes . An example is the depletion of specific messenger RNA by the binding of small regulatory RNA , which generates ultrasensitive thresholds leading , in turn , to the highly sensitive regulation of gene expression [20]-[24] . The binding of many transcription factors is likewise thought to be rendered ultrasensitive via a sequestration mechanism in which high affinity “decoy” binding sites scattered across the genome ( or inhibitory proteins that compete for the transcription factor [25]-[26] act as depletants leading to a steep , effectively “all-or-none” transcriptional response [13] , [27]-[28] . As a test of this hypothesis , Buchler and co-workers have recently recreated the sequestration mechanism in vivo in a synthetic genetic circuit that converts a graded transcriptional response into an ultrasensitive response via the addition of a depletant [16] . Here we build on their study by recreating the sequestration mechanism in vitro using molecular beacons , a well-established biosensor architecture , as our model system . The experimental parameters that define this in vitro model can be controlled with great precision , providing a means to dissect and test a quantitative model of sequestration in unprecedented detail . Following this we have adapted the sequestration mechanism to steepen the concentration-occupancy relationship for the binding of a common transcription factor by an order of magnitude over the sensitivity observed in the absence of sequestration .
Molecular beacons , synthetic biomolecular switches developed by Kramer and coworkers for the detection of specific DNA or RNA sequences [29] , are now widely used in the diagnosis of genetic and infectious diseases [30]-[32] . Consisting of a stem-loop DNA modified with a fluorophore/quencher pair , molecular beacons are quantitatively described by a simple three-state population-shift model in which the equilibrium between a non-binding , non-signaling state and the binding-competent , signaling state is pushed towards the latter upon target binding [33] . This linkage between a conformational equilibrium and target binding allows us to rationally tune the affinity of molecular beacons –without affecting their specificity– by altering the stability of the stem . Indeed , using this approach we have previously shown that it is possible to tune the affinity of molecular beacons across more than 4 orders of magnitude [33] . For the studies reported here we have used a set of six molecular beacons sharing a common recognition element but spanning this same 10 , 000-fold range of target affinities ( Table 1 ) . The input/output function of each of these six molecular beacons is well described by the hyperbolic curve expected for single site binding , ( 1 ) in which F[T] is the fluorescence output as a function of target concentration , [T] , F0 and FB are the fluorescence of the unbound and bound states respectively , and Kdprobe is the dissociation constant of the probe/target duplex . We introduce sequestration into molecular beacons by combining a relatively low affinity signaling probe ( i . e . , fluorophore/quencher labeled ) with an excess of a higher affinity ( but unlabeled and thus non-signaling ) stem-loop that serves as the depletant ( dep ) ( Figure 1c ) . Doing so , we convert the hyperbolic binding curve associated with a traditional molecular beacon ( Eq . 1 ) into a much steeper , ultrasensitive response ( Figure 2 ) . A physically reasonable description for the proposed sequestration mechanism is easily derived from the above hyperbolic binding curve by replacing [T] with the effective concentration of free ( unbound , un-sequestered ) target . As per Buchler and Louis [15] , this concentration goes with: ( 2 ) where [T]t is the total amount of target added and Kddep is the dissociation constant of the depletant/target complex . By combining Eq . 1 and Eq . 2 we see that the ratio of the depletant concentration to the probe affinity ( [dep]/Kdprobe ) is a crucial determinant of ultrasensitivity . To validate this we employed the relatively low affinity molecular beacon 2GCprobe ( Kdprobe = 310 nM; table 1 ) as our probe and the higher affinity unlabeled molecular beacon 0GCdep ( Kddep = 5 . 2 nM ) as our depletant . When we do so we observe ultrasensitivity as soon as the depletant concentration rises above the probe dissociation constant ( i . e . , as [dep]/Kdprobe increases above unity; Figure 2 , left ) . With further increases in depletant concentration the steepness of the dose-response curve increases monotonically to the highest [dep]/Kdprobe ratios we have investigated . Moreover , by applying equations 1 and 2 to these data we see that the sequestration model fits these ultrasensitive responses quantitatively ( R2 ≥ 0 . 998 ) without the use of any fitted parameters ( solid lines , Figure 2 ) . That is , we can quantitatively fit our observations with this model using parameters values –the affinities and fluorescence signals of the two molecular beacons- determined independently in previous studies [33] . The Hill coefficient is commonly employed to describe ultrasensitive systems in biochemistry [34] . And while it is not a physically correct description of sequestration ( as it was originally derived to describe allosteric cooperativity ) , we find that the pseudo-Hill coefficient we obtain by fitting the Hill equation to our data provides a convenient way of comparing the ultrasensitive behavior generated by the sequestration mechanism . As expected , we observe a pseudo-Hill coefficient near unity ( 0 . 90±0 . 02 ) for a binding curve obtained in the absence of depletant ( dotted lines in Figure 2 , left ) . Upon the addition of depletant this value climbs , reaching 1 . 3 at a [dep]/Kdprobe ratio of 0 . 3 before ultimately reaching a value of 9 . 4 at a ratio of 80 , the highest [dep]/Kdprobe ratio we have investigated ( Figure 2 , right ) . That is , sequestration ultimately compresses the normally 81-fold dynamic range of a molecular beacon ( i . e . , the transition from 10% to 90% target occupancy spans an 81-fold change in target concentration; see ref . 17 ) into a 1 . 5-fold dynamic range , significantly increasing the steepness of the input/output function of the molecular beacon and , in turn , improving its ability to detect small changes in relative target concentration . Our in vitro model also provides an opportunity to explore , for the first time , the extent to which sequestration-derived ultrasensitivity depends on other parameters , including , for example , the relative affinities of the depletant and the probe ( i . e . , Kdprobe/Kddep ) . To do so , we varied the depletants affinity ( using Kd ranging from 5 . 2 nM to 3 µM ) at a constant [dep]/Kdprobe ratio of 3 . 2 ( Kdprobe = 310 nM –molecular beacon 2GC- with and a [dep] of 1 µM ) . As expected we find that , while high affinity depletants ( i . e . , 0GCdep , 1GCdep ) produce clear ultrasensitive responses ( pseudo-Hill coefficients of 3 . 9 and 3 . 6 respectively ) , depletants with affinities similar to ( i . e . , 2GCdep , 3GCdep ) or poorer than ( i . e . , 4GCdep and 5GCdep ) those of the probe produce only minor improvements in sensitivity ( pseudo-Hill coefficients <1 . 3 ) ( Figure 3 , left ) . Again , all of the data so obtained are well modeled by Eq . 1 and 2 without the use of any fitted parameters ( i . e . , by fixing all parameters to the values obtained from independent experimental conditions ) , providing another high precision test of the quantitative sequestration model ( Figure 3 , right ) . While the two ratios described above , [dep]/Kdprobe and Kdprobe/Kddep , play crucial roles in generating ultrasensitivity they do not work independently of one another . For example , if [dep]/Kdprobe falls well below unity we will not obtain ultrasensitivity no matter how high the Kdprobe/Kddep ratio climbs . This occurs because , when the probe dissociation constant is significantly higher than the concentration of the depletant , the free target concentration at the “threshold” is too low to saturate the probe , leading to a more-or-less hyperbolic response approximating that seen in the absence of depletant ( Figure S1 ) . To better understand the interplay between these two ratios ( i . e . , to illustrate the parameter space over which ultraensitive behavior is obtained ) we can plot the pseudo-Hill coefficient as a function of [dep]/Kdprobe and Kdprobe/Kddep ( Figure 4 ) . Doing so we find that , when [dep]/Kdprobe falls below 0 . 91 it is not possible to achieve ultrasensitivity with any reasonable value of Kdprobe/Kddep . Similarly , if Kdprobe/Kddep falls below 0 . 94 we do not generate a pseudo-Hill coefficient above 2 even at the highest depletant concentration we have employed . Finally , the ease with which we can manipulate our in vitro system renders it possible to also characterize the effects of varying Kdprobe at a constant depletant concentration . To do so , we increased the length of our target sequence , which lowers the dissociation constants of the probe and depletant for the target by the same extent . Using targets ranging from 13 to 17 nucleotides ( and thus increasing the [dep]/Kdprobe ratio from 0 . 3 to more than 100 ) , we observe the expected monotonic increase in sensitivity ( Figure S2 ) . Moreover , these data too , fit equations 1 and 2 quantitatively ( R2>0 . 995 ) without the use of any adjustable parameters . In the above studies molecular beacons served as a convenient , synthetic toolkit to quantitatively and precisely test the sequestration model . Obviously , however , molecular beacons are not themselves of any specific biological relevance . We have thus also developed an in vitro system to test the extent to which sequestration can cause pseudo-cooperative , highly sensitive changes in the concentration of free transcription factor , thus increasing the sensitivity with which a transcription factor-binding site is occupied . A difficulty in demonstrating this with precision is that traditional methods for quantifying the occupancy of a transcription-factor binding site , including gene activation , gel shift assays and ELISAs , provide only relatively “low resolution” measurements of site occupancy . As our read-out we have thus instead employed a recently developed , highly precise optical reporter for transcription factor binding activity termed “transcription factor beacon” ( Figure 5 , left ) [35] . Specifically , in order to detect the binding of our transcription factor , TATA binding protein , we have used a transcription factor switch that exhibits a 45 nM dissociation constant for this target , reporting its binding via a large change in fluorescence output . As our depletant we have employed a hairpin DNA that contains TATA binding protein’s double stranded recognition sequence but that lacks a fluorophore/quencher reporting pair . Unlike the transcription factor switch , the hairpin does not undergo any binding-induced conformational change and thus its affinity for TATA binding protein is , as required by the sequestration mechanism , significantly greater than that of the reporting probe . Using a 1∶10 mixture of this probe/depletant pair we achieve a pseudo-Hill coefficient of 4 . 3 , compressing the normally 81-fold psuedolinear range of the occupancy of this transcription factor binding site to a mere 4-fold and thus significantly increasing the sensitivity with which it is occupied ( Figure 5 , right ) .
Using molecular beacons and transcription-factor binding as model systems we have recreated , in vitro , the sequestration mechanism that Nature employs to generate ultrasensitive behavior in many genetic networks . Doing so we have demonstrated that the simple , quantitative model proposed by Buchler and Louis accurately and precisely predicts the relationships between ultrasensitivity and the concentrations and affinities of the depletant and probe . We have also demonstrated , more generally , the utility of employing DNA-based in vitro models in the high precision testing and dissection of biologically important regulatory mechanisms . As noted above , Buchler and co-workers [15] , [16] were among the first to test the sequestration mechanism using a synthetic genetic circuit in vivo that they converted from a graded transcriptional response into an ultrasensitive output via the addition of a depletant [16] . Their work confirmed earlier suggestions that sequestration could underlie bistable or oscillatory circuits in natural regulatory systems . It also highlighted the important determinants of the sequestration mechanism . Due to the intrinsic complexity of in vivo systems , however , it proved difficult to use this model system to test all the determinants of sequestration with high precision . Buchler and co-worker , for example , were unable to evaluate the range of Kdprobe/Kddep over which various degree of ultrasensitive behavior could be observed . Here , in contrast , we have employed molecular beacons , a well-defined , easily controllable , in vitro system , as a tool to dissect the sequestration mechanism [15] , [16] in more detail and with greater precision than has hitherto proven possible . While the impressive specificity , affinity and versatility of biomolecular recognition have motivated decades of research in the development of sensors and other biotechnologies based on this effect [36] , the hyperbolic –and thus not particularly sensitive– concentration/occupancy curves characteristic of single site binding limits their precision . This , in turn , limits the utility of these biotechnologies in many applications . Given this we believe that the use of sequestration in vitro may be of use in increasing the sensitivity of synthetic biosystems , such as biosensors , in vitro . Specifically , we have shown that it possible to narrow the 81-fold pseudo-linear dynamic range of the well-known molecular beacon platform by almost 2 orders of magnitude and of a transcription factor switch sensor by a factor of 20 . The modified sensors so produced exhibit ultrasensitive responses equivalent to Hill coefficients of greater than 9 and greater than 4 respectively , converting them into high precision analytical approaches . Given that sequestration requires only the availability of depletants that bind the target in question with greater affinity than that of the signal-generating probe , we anticipate that the mechanism can be adapted to many other biotechnologies , an argument bolstered by the frequency with which this mechanism is employed in the cell [15] , [27] , [28] . Moreover , the sequestration mechanism appears more straightforward to implement than the other mechanisms Nature has employed to achieve improved sensitivity . It appears far easier to implement , for example , than positive allosteric cooperativity as the later would involve the design of probes containing two or more precisely interacting sites for target binding [37]-[38] . Despite these advantages , the sequestration strategy is not without a limitation: the generation of ultrasensitive response is achieved at the cost of a reduced affinity , which shifts the minimum target concentration producing a detectable signal ( the detection limit ) towards higher concentrations . The extremely steep input/output functions demonstrated here would appear to open the door to new applications across biosensors and synthetic biology . Perhaps the most obvious application would be in the monitoring of , for example , drugs with such narrow therapeutic windows that only high precision measurements of their concentration achieve clinical relevance . More speculatively , the availability of sensors that , in contrast to the graded ( analog ) outputs of most biosensors , produce an effectively “all-or-none” ( digital ) response may be useful in the development of molecular logic gates [39]-[43] . These , in turn , may enable the development of molecular-scale computers and “autonomously regulated” chemical systems , ideas that have attracted significant recent interest [44] , [45] .
The following HPLC-purified constructs were purchased from Sigma-Genosys and used as received ( the bases in italic are those constituting the stem ) : 1GCprobe: 5’- ( FAM ) -A-CTATT-GATCGGCGTTTTA-AATAG-G - ( BHQ ) -3’ 2GCprobe: 5’- ( FAM ) -A-CTCTT-GATCGGCGTTTTA-AAGAG-G - ( BHQ ) -3’ 3GCprobe: 5’- ( FAM ) -A-CTCTC-GATCGGCGTTTTA-GAGAG-G - ( BHQ ) -3’ where FAM and BHQ represent 6-carboxyfluorescein and Black Hole Quencher respectively . The following HPLC-purified , un-modified depletants and targets were purchased from Sigma-Genosys and were used as received ( the bases in italic are those constituting the stem ) : 0GCdep: 5’-A-TTATT -GATCGGCGTTTTA-AATAA-G -3’ 1GCdep: 5’-A-CTATT -GATCGGCGTTTTA-AATAG-G -3’ 2GCdep: 5’-A-CTCTT -GATCGGCGTTTTA-AAGAG-G -3’ 3GCdep: 5’-A-CTCTC-GATCGGCGTTTTA-GAGAG-G -3’ 4GCdep: 5’-A-CTCGC-GATCGGCGTTTTA-GCGAG-G -3’ 5GCdep: 5’-A-CGCGC-GATCGGCGTTTTA-GCGCG-G -3’ 13-base target: 5’-TAAAACGCCGATC-3’ 15-base target: 5’-TTAAAACGCCGATCA-3’ 17-base target: 5’-TTTAAAACGCCGATCAA-3’ HPLC purified DNA modified with FAM and internal BHQ-1 inserted on a thymine residue was purchased from Biosearch Technologies ( Novato , CA ) and has the following sequence: TFprobe: 5’- ( FAM ) -TACTT TTATATAAAT AAGT T ( BHQ ) GTGA TTTTTATATATT TCAC -3’ The following HPLC-purified , un-modified depletant was purchased from Sigma-Genosys and was used as received: TFDep: 5’-CGTATATAAAGG TTTTTTT CCTTTATATACG -3’ This protein was expressed recombinantly , purified , and characterized as previously described [46] . All fluorescent experiments were conducted at pH 7 . 0 in 50 mM sodium phosphate buffer , 150 mM NaCl , at 45 °C . For all experiments with TATA binding protein , the buffer was supplemented with 5 mM MgCl2 , as it is essential for efficient DNA binding and the measurement were conducted at 25 °C . Equilibrium fluorescence measurements were obtained using a Cary Eclipse Fluorimeter with excitation at 480 ( ± 5 ) nm and acquisition at 517 ( ± 5 ) nm . Binding curves were obtained using solutions of 10 nM of labeled molecular beacons ( or TF switch ) and varying concentrations of unlabeled beacons ( or DNA binding protein recognition sequence ) as depletant and by sequentially increasing the target concentrations via the addition of small volumes of solution with increasing concentrations of target . Dissociation constants of the labeled beacons were obtained from the literature [33] .
|
Here we recreate in vitro the sequestration mechanism thought to underlie the extraordinary sensitivity ( the steepness of the input/output function ) of a number of genetic networks . We do so first using fluorescent molecular beacons , a well-established , DNA-based biosensor architecture , as our model system . The experimental parameters that define this in vitro model can be controlled with great precision , allowing us to dissect and test a quantitative model of sequestration in unprecedented detail . Following on this we employ the sequestration mechanism to steepen the binding-site occupancy curve of a common transcription factor by an order of magnitude over the sensitivity observed in the absence of sequestration . Our study thus highlights the versatility with which this approach can be used to improve the performance of both synthetic biological systems and artificial biosensors .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"systems",
"biology",
"analytical",
"chemistry",
"synthetic",
"biology",
"chemistry",
"biology",
"computational",
"biology",
"chemical",
"biology",
"biophysics",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2011
|
High-Precision, In Vitro Validation of the Sequestration Mechanism for Generating Ultrasensitive Dose-Response Curves in Regulatory Networks
|
Adipocyte progenitors reside in the stromal vascular fraction ( SVF ) of adipose tissues that are composed of fibroblasts , immune cells , and endothelial cells . It remains to be elucidated how the SVF regulates adipocyte progenitor fate determination and adipose homeostasis . Here , we report that fibroblast-specific protein-1 ( FSP1 ) + fibroblasts in the SVF are essential to adipose homeostasis . FSP1+ fibroblasts , devoid of adipogenic potential , are adjacent to the preadipocytes in the SVF . Ablation of FSP1+ fibroblasts in mice severely diminishes fat content of adipose depots . Activation of canonical Wnt signaling in the FSP1+ fibroblasts results in gradual loss of adipose tissues and resistance to diet-induced obesity . Alterations in the FSP1+ fibroblasts reduce platelet-derived growth factor ( PDGF ) -BB signaling and result in the loss of preadipocytes . Reduced PDGF-BB signaling , meanwhile , impairs the adipogenic differentiation capability of preadipocytes by regulating matrix metalloproteinase ( MMP ) expression , extracellular matrix remodeling , and the activation of Yes-associated protein ( YAP ) signaling . Thus , FSP1+ fibroblasts are an important niche essential to the maintenance of the preadipocyte pool and its adipogenic potential in adipose homeostasis .
Adult adipose tissue contains adipocyte progenitors that are critical to adipose homeostatic turnover as well as adaptive hyperplastic expansion and regeneration [1–4] . Peroxisome proliferator-activated receptor-γ ( PPARγ ) + adipocyte progenitors and preadipocytes—characteristic of cell surface markers , e . g . , cluster of differentiation 34 ( CD34 ) and stem cell antigen-1 ( Sca1 ) —reside in the adipose vasculature expressing mural cell markers α-smooth muscle actin ( αSMA ) , platelet-derived growth factor receptor-β ( PDGFR-β ) , and neural glial antigen 2 ( NG2 ) [1 , 2 , 5 , 6] . Stromal vascular fraction ( SVF ) -resident preadipocytes are highly committed and are capable of proliferating and differentiating into mature adipocytes in vitro and in vivo [1–3] . Cell fate and differentiation capability of the preadipocytes are regulated by signaling pathways and transcriptional and epigenetic programs . Wnt signaling plays crucial roles during development [7] . Activation of β-catenin-dependent canonical Wnt signaling in the preadipocytes inhibits adipogenesis [5 , 8 , 9] . Adipogenic potential of preadipocytes is determined not only by their intrinsic properties but also by the surrounding microenvironment . Diet-induced obesity is regulated by the adipose microenvironment , but not by cell-intrinsic mechanisms , highlighting the importance of microenvironmental regulation in adipose homeostasis [10] . Adipose tissues contain multiple types of stromal cells , including fibroblasts , endothelial cells , and immune cells . While macrophage-related chronic inflammation in the adipose tissues was reported to play a crucial role in the development of obesity and obesity-related insulin resistance [11 , 12] , roles of other cell types in preadipocyte fate determination and adipose homeostasis are largely unknown . Fibroblasts are resident cell types in the adipose tissues with a mesenchymal lineage origin . Fibroblast-specific protein 1 ( FSP1; also known as S100A4 ) is a reliable marker for detecting tissue-resident fibroblasts , in addition to other mesenchymal markers αSMA , PDGFR-β , and NG2 [13–15] . While the roles of fibroblasts in adipose homeostasis remain elusive , fibroblasts have multifaceted regulatory roles in tissue morphogenesis , wound healing , fibrosis , and cancer by modulating the behavior and functions of epithelial cells and stem cells [13 , 15–17] . Here , we report that FSP1+ fibroblasts in the SVF are a niche for adipogenesis . Activation of canonical Wnt signaling in the FSP1+ fibroblasts or ablation of FSP1+ fibroblasts disturbs adipose homeostasis and results in loss of adiposity . Alterations in the FSP1+ fibroblasts resulted in decreased platelet-derived growth factor ( PDGF ) expression . On the one hand , PDGF maintains the preadipocyte number . On the other hand , PDGF regulates the adipogenic potential of preadipocytes by regulating extracellular matrix remodeling in the microenvironment and Yes-associated protein ( YAP ) activation . Thus , FSP1+ fibroblasts are a niche for preadipocytes orchestrating the functions of preadipocytes and their microenvironment .
Preadipocytes express mesenchymal markers αSMA , PDGFR-β , and NG2 . To investigate whether FSP1+ fibroblasts are part of the adipocyte lineage , Fsp1-Cre mice [15] were crossed with Rosa26-floxed stop tdTomato reporter mice . Only very weak tdTomato signal was detected in the inguinal and epididymal white adipose tissues ( I-WATs and E-WATs ) from the Fsp1-Cre;tdTomato compound mice on a normal-chow diet ( ND ) ( S1A Fig ) , in contrast to strong tdTomato signal in the tail tip fibroblasts ( S1A and S1B Fig ) . Fluorescence-activated cell sorting ( FACS ) analysis indicated that a small percentage of the SVF cells , which expressed fibroblast markers αSMA and vimentin ( S1C and S1D Fig ) , expressed tdTomato ( Fig 1A ) . Upon high-fat diet ( HFD ) -induced obesity , percentage of the FSP1+/tdTomato+ SVF cells increased ( I-WAT: 9 . 26% ± 0 . 28% [ND] versus 18 . 17% ± 1 . 36% [HFD]; E-WAT: 7 . 68% ± 0 . 96% [ND] versus 12 . 33 ± 0 . 67% [HFD] ) . SVF preadipocytes differentiated to mature adipocytes upon adipogenic induction ( Fig 1B and 1C ) . tdTomato+ SVF cells , however , did not differentiate to adipocytes ( Fig 1B and 1C ) . Consistently , CD34+Sca1+ preadipocytes were enriched in the tdTomato− , but not the tdTomato+ , SVF populations ( Fig 1D and S1E Fig ) . To gain a better idea of the spatial localization of the FSP1+ cells , immunostaining of PPARγ ( preadipocytes ) and green fluorescent protein ( GFP; FSP1+ fibroblasts ) was performed on the white adipose tissue ( WAT ) sections of the Fsp1-Cre;Rosa26-flox-membrane-targeted Tomato-flox-membrane-targeted GFP ( F-mTmG ) mice . PPARγ+ and GFP+ SVF cells were adjacent to each other but mutually exclusive ( Fig 1E ) . These data collectively suggested that the FSP1+ fibroblasts were not in the adipocyte lineage but were localized in the microenvironment of preadipocytes in the SVF . Wnt regulates adipose homeostasis by activating β-catenin in preadipocytes to inhibit their differentiation [8 , 9] . In addition to its direct inhibitory effect on preadipocyte differentiation [8] , Wnt may regulate adipose homeostasis through cell types in the microenvironment , e . g . , fibroblasts . Expression of β-catenin target genes were altered in the FSP1+/tdTomato+ fibroblasts upon HFD-induced obesity ( S2A Fig ) . To investigate whether activation of Wnt signaling in FSP1+ fibroblasts would affect adipose homeostasis , Ctnnb1exon 3 fl/+ mice were crossed with Fsp1-Cre mice to generate the Fsp1-Cre;Ctnnb1exon 3 fl/+ ( F-BCA ) compound mice . Fsp1-Cre-driven deletion of exon 3 in β-catenin produced a smaller–molecular-weight β-catenin protein in tail tip fibroblasts ( Fig 2A and S2B Fig ) . Despite the fact that the activated form of β-catenin was present in the SVF ( Fig 2A and S2B–S2D Fig ) , there was no detectable activated form of β-catenin in the WAT or adipocytes from the F-BCA mice ( Fig 2A and S2B Fig ) . These data further suggested that FSP1+ fibroblasts were not directly involved in the adipocyte lineage ( Fig 1 ) . In line with this notion , tdTomato+/FSP1+ SVF cells were negative for preadipocyte markers in F-BCA mice ( S2E Fig ) . Fsp1-Cre-driven activation of canonical Wnt signaling in fibroblasts did not affect embryonic development . F-BCA mice were born at expected Mendelian ratio . Mice with an Fsp1-Cre or Ctnnb1exon 3 fl/+ genotype were phenotypically similar to the wild-type ( WT ) littermates throughout the lifespan . Mice with a WT , Fsp1-Cre , or Ctnnb1exon 3 fl/+ genotype were therefore classified as controls in comparison to their F-BCA littermates . At weaning , F-BCA male mice were phenotypically normal , with similar amounts of adipose depots to the control mice ( S3A–S3D Fig ) . Dissected I-WATs and E-WATs were similar in size to those from the control mice ( S3C Fig ) . However , F-BCA male mice gain less weight after puberty compared with the control mice ( Fig 2B ) . At 4 months of age , F-BCA male mice had less fat compared with the control mice ( Fig 2C–2F ) . Dissected I-WATs and E-WATs were significantly smaller in size and weighed less in the F-BCA mice ( Fig 2E and 2F ) . Histology inspection on adipose tissue sections showed that the F-BCA adipocyte diameter was significantly smaller ( Fig 2G ) . F-BCA male mice at 8 months of age were virtually free of subcutaneous or visceral adipose depots ( S3E–S3H Fig ) . The adipose phenotype was further confirmed by dissecting the WATs from the control and F-BCA male mice ( S3G Fig ) and by histology inspection on the adipose tissue sections ( S3H Fig ) . Metabolic cage analyses revealed that the F-BCA male mice had similar food uptake , respiratory exchange ratio , and physical activity , but slightly reduced energy expenditure compared with the control mice ( S4A–S4D Fig ) . Such defects in adipose homeostasis were also observed in the female F-BCA mice ( S5 Fig ) . WAT is an important metabolic regulator . Adipose defects observed in the F-BCA mice mimic lipodystrophy , a disorder accompanied by metabolic disturbances , including hyperglycemia , insulin resistance , and ectopic lipid deposition in secondary organs [8] . F-BCA male mice did not show the metabolic abnormality observed in the classical lipodystrophy mouse models . Rather activation of β-catenin in the FSP1+ fibroblasts offered metabolic benefits in the mice ( Fig 2H–2K ) . F-BCA male mice were able to more efficiently clear glucose ( Fig 2H and 2I ) but retained the same insulin responsiveness as the control mice ( Fig 2J and 2K ) . Livers weighed less , and no ectopic lipid deposition was observed in the livers in the F-BCA male mice ( S4I–S4K Fig ) . Activation of preadipocytes drives adipocyte hyperplasia in diet-induced obesity [10] . Such a process is regulated by the adipose microenvironment but not by cell-intrinsic mechanisms [10] . To investigate whether activation of canonical Wnt signaling in the FSP1+ fibroblasts would affect diet-induced obesity , F-BCA male mice were fed with an HFD for 12 weeks . On HFD , F-BCA male mice consumed more food and had significantly enhanced respiratory exchange ratio and physical activity compared with the control mice ( S4E–S4H Fig ) . Unlike the control mice , which accumulated a significant amount of fat upon HFD feeding ( Fig 3A–3F ) , F-BCA male mice were resistant to HFD-induced obesity ( Fig 3A–3F ) . Control mice fed with an HFD were glucose intolerant and insulin resistant ( Fig 3G–3J ) . F-BCA mice on an HFD cleared glucose as efficiently as those on an ND ( Fig 3G and 3H ) . F-BCA mice on an HFD were more sensitive to insulin ( Fig 3I and 3J ) . Although control mice accumulated a large amount of fat in the liver upon HFD feeding , F-BCA mice were protected from HFD-induced steatosis ( S4J and S4K Fig ) . Activation of canonical Wnt signaling in the FSP1+ fibroblasts disturbed adipose homeostasis . Gene set enrichment analysis ( GSEA ) on gene expression of SVF cells indicated impaired adipogenic potential of the SVF cells isolated from the F-BCA mice ( Fig 4A ) . Indeed , F-BCA SVF cells less efficiently differentiated into adipocytes upon adipogenic induction compared with the control SVF cells ( Fig 4B–4E ) . Inefficient differentiation of F-BCA SVF cells may result from a reduced number of preadipocytes or from impaired differentiation potential of the preadipocytes . F-BCA SVF cells had lower expression levels of preadipocyte markers than the control SVF cells ( Fig 4F ) . FACS analyses revealed reduced numbers of CD34+Sca1+ preadipocytes in the F-BCA SVF ( Fig 4G and 4H ) . Conditioned medium from control SVF cells promoted F-BCA SVF cell differentiation ( Fig 5A and 5B ) , indicating that secreted factors regulated preadipocyte differentiation . Expression of PDGFB was drastically decreased in the F-BCA SVF cells compared with that in the control SVF cells ( Fig 5C ) . Immunohistochemical staining on the control and F-BCA WAT sections indicated reduced PDGFR-β phosphorylation in the F-BCA WATs ( Fig 5D ) . Preadipocytes express PDGF receptors [2 , 5 , 6] . We next studied whether reduced PDGF expression is responsible for the loss of preadipocytes in the F-BCA SVFs . Treatment of the F-BCA SVF cells with PDGF-BB significantly increased the number of preadipocytes and restored the adipogenic potential ( Fig 5E–5H ) . To study whether PDGF-BB is responsible for preadipocyte pool maintenance in vivo , growth factor-reduced Matrigel supplemented with PDGF-BB was implanted into the I-WATs of the F-BCA mice . Matrigel alone minimally affected the number and the adipogenic potential of the F-BCA SVF cells ( Fig 5I–5K ) , whereas PDGF-BB containing Matrigel in the contralateral I-WAT significantly increased the percentage of preadipocytes and restored the adipogenic potential of the F-BCA SVF cells ( Fig 5I–5K ) . Activation of Wnt signaling in fibroblasts promotes tissue fibrosis [18–20] . Sirius red staining revealed significantly more collagen deposition in the F-BCA adipose tissues , particularly around the SVF zones ( Fig 6A ) . More abundant type I collagen ( Col I ) expression was detected in the F-BCA SVF cells ( Fig 6B ) , although the mRNA levels were unchanged upon activation of canonical Wnt signaling in the FSP1+ fibroblasts ( Fig 6C ) . The matrix metalloproteinase ( MMP ) family is primarily responsible for the degradation and remodeling of extracellular matrix [6 , 21] . Proteolytic activity of MMPs is regulated by the tissue inhibitors of metalloproteinase ( TIMPs ) [6 , 21] . MMPs and TIMPs are differentially expressed in the adipose tissues during obesity [22] and modulate adipocyte differentiation [22–26] . Inhibition of MMP activity diminished the adipogenesis capability of preadipocytes [22] . F-BCA SVF cells had much reduced MMP expression and up-regulated TIMP-3 expression ( Fig 6D and 6E ) . Gelatin zymography revealed significantly lower MMP expression in the F-BCA cells ( Fig 6F ) . PDGF-BB treatment increased MMP expression and decreased TIMP3 expression ( Fig 6G , S6A and S6B Fig ) . Despite the fact that the mRNA levels of collagen were not changed , Col I protein levels decreased upon PDGF-BB treatment ( Fig 6H and S6C Fig ) . Excess collagen deposition and a stiff microenvironment inhibit preadipocyte differentiation [25–28] . Hippo pathway transcription factors YAP and TAZ are the major effectors sensing the mechanical signals exerted by extracellular matrix ( ECM ) physical property [28–32] that inhibits adipogenesis of mesenchymal stem cells [26 , 28 , 33] . Indeed , YAP protein accumulated in the F-BCA SVF cells ( Fig 6B ) . The YAP signature was significantly enriched , and expression of YAP target genes CTGF and ANKRD1 was significantly up-regulated in the F-BCA SVF cells ( Fig 6I and 6J ) . PDGF-BB treatment reduced the expression of YAP and its target genes ( Fig 6H and 6K ) . To study whether YAP activation is responsible for the impaired adipogenic potential of F-BCA SVF cells , F-BCA SVF cells were pretreated with YAP pharmacological inhibitor verteporfin ( VP ) [34] . Inhibition of YAP signaling with VP restored the adipogenic potential of the F-BCA SVF cells , without affected Col I and MMP expression ( Fig 6L–6N and S6D–S6G Fig ) . VP treatment did not increase the percentage of preadipocytes , suggesting that YAP mainly regulated differentiation capability ( S6H Fig ) . To investigate the physiological requirement of the FSP1+ fibroblasts in the adipose development , FSP1+ fibroblasts were ablated by generating the Fsp1-Cre;Rosa26-DTA ( F-DTA ) compound mice ( Fig 7A ) . At 4 months of age , F-DTA mice were significantly lean , with markedly reduced fat compared with the control mice ( WT , Fsp1-Cre , or Rosa26-DTA ) ( Fig 7B and 7C , S7A and S7B Fig ) . Dissected WATs were smaller in size and weighed less in the F-DTA mice ( Fig 7D and 7E ) . F-DTA male mice had significantly more food uptake , similar respiratory exchange ratio and physical activity , and slightly reduced energy expenditure compared with the control mice ( S8A–S8D Fig ) , whereas F-DTA female mice had comparable food uptake , respiratory exchange ratio and physical activity , and slightly increased energy expenditure ( S7C–S7F Fig ) . F-DTA mice more efficiently cleared glucose , but retained the same responsiveness to insulin , similar to that of the F-BCA mice ( S8E–S8H Fig ) . No ectopic lipid deposition was observed in the livers of the F-DTA mice ( S8I–S8K Fig ) . F-DTA SVF cells were defective in adipogenesis ( Fig 7F–7H ) . F-DTA SVF cells contained much fewer preadipocytes ( Fig 7I ) . PDGFB expression decreased in the F-DTA SVF cells ( Fig 7J ) . PDGF-BB treatment increased the number of preadipocytes and restored the adipogenic potential of F-DTA SVF cells ( Fig 7K–7M ) . Similar to the F-BCA SVF cells , F-DTA SVF cells had much reduced MMP expression , up-regulated TIMP-1 and TIMP-3 expression , and YAP activation ( S9 Fig , Fig 7N and 7O ) . Inhibition of YAP signaling restored the adipogenic potential of the F-DTA SVF cells ( Fig 7P and 7Q ) .
Adipocytes continuously turn over in adults . Like other adult stem cells and progenitor cells , preadipocytes are resident in a highly specialized niche . In this report , we identified FSP1+ fibroblasts as the niche for preadipocytes . FSP1+ fibroblasts are crucial to the maintenance of adipose homeostasis . Alteration in FSP1+ fibroblasts disturbs adipose homeostasis and results in loss of adiposity . Preadipocytes rapidly expand from the preexisting pool during the first postnatal month [3] . During adulthood , preadipocytes and adipocytes are constantly renewed [35 , 36] . Turnover of preadipocytes and adipocytes is observed in both humans and mice [35 , 36] . Obese mice have increased adipocyte formation [35 , 36] . Adipogenic niches , including macrophage-related chronic inflammation [11 , 12] , may have predominant roles in regulating the turnover of preadipocytes and adipocytes during adulthood and obesity . Numbers of FSP1+ fibroblasts increase during obesity . FSP1+ fibroblasts regulate adipose homeostasis by maintaining the preadipocyte pool and its differentiation potential . Alterations in FSP1+ fibroblasts result in diminished adipose depots at adulthood but not at puberty . Therefore , FSP1+ fibroblasts may represent a class of microenvironmental cues in regulating the turnover of preadipocytes and adipocytes and adipose homeostasis during adulthood and obesity . FSP1 is broadly expressed in mesenchymal cells . It should be noted that fibroblasts are highly heterogeneous . The heterogeneity of fibroblasts are reflected not only by the marker expression but also their biological functions . Subtypes of fibroblasts , similar to those of macrophages , were proposed [13] . Obesity induces macrophage polarization in adipose tissues [12] . It warrants further investigation into whether FSP1+ fibroblasts undergo similar polarization in obesity and whether subpopulations of FSP1+ fibroblasts regulate preadipocyte renewal and maintenance of adipose homeostasis . Conditioned medium from the control SVF cells can restore the differentiation potential of the F-BCA SVF cells , suggesting that FSP1+ fibroblasts regulate preadipocyte renewal and adipose homeostasis via soluble factors . Alterations in FSP1+ fibroblasts greatly reduce PDGF-BB expression . PDGF-BB can restore preadipocyte numbers both in vitro and in vivo . In addition to the maintenance of the preadipocyte population , PDGF-BB also regulates adipogenic differentiation capability of the preadipocytes . Preadipocytes express PDGF receptors [2 , 5 , 6] . PDGF receptor expression in preadipocytes not only reflects the mesenchymal origin of the preadipocytes but may also play a central role in adipose homeostasis by maintaining both the preadipocyte pool and its adipogenic potential . The regulation of maintenance of the preadipocyte population and adipogenic potential by PDGF-BB may be exerted through distinct signaling pathways . Alterations in FSP1+ fibroblasts disturb extracellular matrix remodeling and YAP signaling in the adipose tissue microenvironment . PDGF-BB treatment restores extracellular matrix remodeling and alleviates YAP activation . However , inhibition of YAP signaling restores the adipogenic capability of the SVF cells but does not affect the number of preadipocytes . It remains to be elucidated whether and how FSP1+ fibroblasts are regulated during adult adipose homeostasis . Upon high-fat–induced obesity , β-catenin target gene expression was altered in the FSP1+ fibroblasts . β-catenin-dependent canonical Wnt signaling is inhibitory to adipogenesis [5 , 8] . Wnt ligands produced by the preadipocytes , in particular Wnt10B , negatively regulate adipose homeostasis . Transgenic mice expressing Wnt10B showed a decrease in adipose mass [9] . Such a phenotype is largely attributed to activation of β-catenin in the preadipocytes [8 , 9] . Mice with β-catenin activation in mature adipocytes had normal adipose depots and metabolism , whereas activation of canonical Wnt signaling in PPARγ+ preadipocytes resulted in severe loss of adiposity [8 , 9] , suggesting autocrine regulation of preadipocyte-origin Wnt ligands . However , Wnt ligands produced by the preadipocytes may also regulate adipose homeostasis via paracrine mechanisms . Activation of canonical Wnt signaling in the FSP1+ fibroblasts results in loss of adiposity and beneficial metabolism , similar to that observed in the mice with canonical Wnt signaling activation in the preadipocytes . Thus , Wnt ligands secreted by the preadipocytes may regulate adipose homeostasis via both autocrine and paracrine mechanisms orchestrating the functions of preadipocytes and their microenvironment . In summary , FSP1+ fibroblasts are essential to the maintenance of the pool and differentiation potential of preadipocytes via PDGF signaling , extracellular matrix remodeling , and YAP activation . FSP1+ fibroblasts are a niche for preadipocytes orchestrating the functions of preadipocytes and their microenvironment .
All mice were housed in a specific pathogen-free environment at the Shanghai Institute of Biochemistry and Cell Biology and treated in strict accordance with protocols approved by the Institutional Animal Care and Use Committee of Shanghai Institute of Biochemistry and Cell Biology ( approval number: NAF-15-003-S325-006 ) . Fsp1-Cre mice [15] were from the Jackson Laboratory . β-cat exon3flox/+ mice were a generous gift from Professor Lijian Hui ( Shanghai Institute of Biochemistry and Cell Biology ) . Fsp1-Cre and β-cat exon3flox/+ mice were backcrossed to the FVB background for >9 generations . Rosa26-tdTomato , Rosa26-mTmG , and Rosa26-DTA mice at C57Bl/6 and 129 mixed background were generous gifts from Professor Yi Zeng ( Shanghai Institute of Biochemistry and Cell Biology ) . For diet-induced obesity studies , starting at 4 weeks of age , mice were fed with HFDs containing 60% kcal from fat for 12 weeks ( Research Diets , New Brunswick , New Jersey ) . Glucose tolerance tests ( GTTs ) and insulin tolerance tests ( ITTs ) were performed as described [37] . For GTTs , mice were fasted overnight and received an intraperitoneal injection of 2 g/kg body weight glucose ( Sigma-Aldrich , St . Louis , Missouri ) . For ITTs , mice were injected intraperitoneally with 0 . 5 U/kg body weight insulin ( Biosynthetic Human Insulin , 100 U/mL; Novo Nordisk , Bagsvaerd , Denmark ) after a 4-hour fast . Tail blood glucose levels were measured at 0 , 15 , 30 , 60 , 90 , and 120 minutes after challenge using the Onetouch Ultra blood glucose monitoring system ( LifeScan , Shanghai , China ) . AUC was calculated using GraphPad Prism . Metabolic measurements with indirect calorimetry were performed on mice as described [38] . Animals were maintained on a ND or an HFD at ambient temperature under 12-hour light and dark cycles . Mice were acclimated in a comprehensive lab animal monitoring system ( Columbus Instruments , Columbus , Ohio ) for 1 day before recording over 2 days following the manufacturer's instruction . Mouse WATs were isolated and fixed in 4% PFA followed by embedding in paraffin . Paraffin-embedded tissues were sectioned and stained with hematoxylin–eosin ( HE ) . The immunostaining was performed as previously described [39] . Sections were incubated at 4°C overnight with primary antibodies . Histology or immunostained samples were viewed under microscope ( IX71; OLYMPUS , Tokyo , Japan ) with a UPlan-FLN 4× objective/0 . 13 PhL , a UPlan-FLN 10× objective/0 . 30 PhL , or a LUCPlan-FLN 20× objective/0 . 45 PhL . Images were captured with a digital camera ( IX-SPT; OLYMPUS , Tokyo , Japan ) and Digital Acquire software ( DPController; OLYMPUS , Tokyo , Japan ) . WATs were viewed under microscope ( SZX16; OLYMPUS , Tokyo , Japan ) with an SDF-PLAPO 1× PF . Images were captured with a digital camera ( U-LH100HGAPO; OLYMPUS , Tokyo , Japan ) and Digital Acquire software ( DPController; OLYMPUS , Tokyo , Japan ) . I-WAT and E-WAT were washed with PBS , cut into small pieces , and digested with 1 mg/mL Collagenase ( Worthington , Lakewood , New Jersey ) at 37°C for 60 minutes . SVF cell culture , FACS analyses , and adipogenic induction were performed as described [40 , 41] . For PDGF and VP treatment experiments , SVF cells were treated with PDGF or VP for 4 days before FACS analyses and adipogenic induction . To isolate floated adipocytes , collagenase-treated adipose depot mixture was centrifuged at 200 g for 1 minute . Floating cell layer was collected as adipocytes . Western blot analysis was performed as previously described [39] . Total RNA was prepared from SVF cells using Trizol reagents ( Invitrogen ) as previously described [39] . Equal amounts of RNA were subjected to quantitative RT-PCR using SYBR green with the BIO-RAD Q-PCR Systems according to the manufacturer's protocol . Relative expression levels were calculated using the comparative CT method . Gene expression levels were normalized to Actin . The primers used are listed in S1 Table . SVF cells were isolated from pooled WATs from control , F-BCA , and F-DTA mice . Total RNA was extracted and purified with TRIzol . Two biological replicates were subjected to complementary DNA library preparation and sequencing according to the Illumina standard protocol . RNA-seq reads were mapped to mm9 reference genome using TopHat2 . Mapped reads of the 2 replicates in each group were merged together for further analysis . Expression for each known gene from RefSeq was determined by covered reads and normalized with RPKM . Normalized gene expression of the SVF cells are listed in S2 Table . Differentially expressed genes ( DEGs ) were identified by gene expression comparison between control and F-BCA or F-DTA SVF cells and were defined with parameters including p < 0 . 01 ( Wald test ) , fold change ≥ 1 . 5 or ≤ 0 . 667 , and expression level ≥ 5 RPKM in at least one sample . Gene ontology ( GO ) analysis was performed with the DAVID online tool . Top GO categories were selected according to the p-value after Benjamini-Hochberg correction . GSEA was performed on gene signatures obtained from the MSigDB database version 5 . 0 ( March 2015 release ) [42] . Statistical significance was assessed by comparing the enrichment score to enrichment results generated from 1 , 000 random permutations of the gene set to obtain p-values ( nominal p-value ) . Sample sizes for each figure are denoted in the figure legends . Data are representative of at least 3 biologically independent experiments . For animal experiments , the sample size is determined on the basis of our prior knowledge of the variability of experimental output . Age of animals was matched . The experiments were not randomized , and the investigators were not blinded to allocation during experiments and outcome assessment . Statistical significance between conditions was assessed by unpaired or paired two-tailed student t tests or ANOVAs followed by Bonferroni's multiple comparison test . All error bars represent SEM , and significance between conditions is denoted *p < 0 . 05; **p < 0 . 01; and *p < 0 . 001 . “NS” denotes not significant . The numerical data used in all figures are included in S1 Data .
|
White adipose tissue ( WAT ) , which consists mostly of adipocytes , is not only a passive energy storage but also an active metabolic and endocrine organ in the body . The importance of maintaining proper adipose mass is emphasized by the fact that both adipose tissue excess—in obese individuals—and deficiency have adverse metabolic consequences . In order to maintain the number of adipocytes , there is a continuous turnover from preadipocytes in adults . Like any other adult stem cells and progenitor cells , cell fate and differentiation capability of preadipocytes are tightly regulated by a highly specialized niche . However , what constitutes the preadipocyte niche , and how the niche regulates preadipocyte function and adipose homeostasis , remain poorly known . In this study , we have identified fibroblast-specific protein-1 ( FSP1 ) + fibroblasts in the WAT stromal vascular fraction ( SVF ) of mice as the niche for preadipocytes . We show that FSP1+ fibroblasts with aberrant Wnt signaling fail to maintain the preadipocyte pool and its differentiation potential , resulting in loss of adipose tissue . We conclude that FSP1+ fibroblasts are a niche for preadipocytes and regulate adipose tissue homeostasis in adult mice .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"body",
"weight",
"medicine",
"and",
"health",
"sciences",
"fibroblasts",
"animal",
"models",
"fluorescence-activated",
"cell",
"sorting",
"physiological",
"processes",
"model",
"organisms",
"adipocytes",
"physiological",
"parameters",
"homeostasis",
"connective",
"tissue",
"cells",
"experimental",
"organism",
"systems",
"obesity",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"connective",
"tissue",
"biological",
"tissue",
"mouse",
"models",
"spectrophotometry",
"cytophotometry",
"signal",
"transduction",
"cell",
"biology",
"anatomy",
"adipose",
"tissue",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"wnt",
"signaling",
"cascade",
"cell",
"signaling",
"spectrum",
"analysis",
"techniques",
"signaling",
"cascades"
] |
2018
|
FSP1-positive fibroblasts are adipogenic niche and regulate adipose homeostasis
|
Leprosy , caused by Mycobacterium leprae , affects over 200 , 000 people annually worldwide and remains endemic in the ethnically diverse , mountainous and underdeveloped southwestern provinces of China . Delayed diagnosis of leprosy persists in China , thus , additional knowledge to support early diagnosis , especially early diagnosis of paucibacillary ( PB ) patients , based on the host immune responses induced by specific M . leprae antigens is needed . The current study aimed to investigate leprosy patients and controls in Southwest China by comparing supernatants after stimulation with specific M . leprae antigens in an overnight whole-blood assay ( WBA ) to determine whether host markers induced by specific M . leprae antigens improve the diagnosis or discrimination of PB patients with leprosy . Leprosy patients [13 multibacillary ( MB ) patients and 7 PB patients] and nonleprosy controls [21 healthy household contacts ( HHCs ) , 20 endemic controls ( ECs ) and 19 tuberculosis ( TB ) patients] were enrolled in this study . The supernatant levels of ten host markers stimulated by specific M . leprae antigens were evaluated by overnight WBA and multiplex Luminex assays . The diagnostic value in PB patients and ECs and the discriminatory value between PB patients and HHCs or TB patients were evaluated by receiver operator characteristics ( ROC ) analysis . ML2044-stimulated CXCL8/IL-8 achieved the highest sensitivity of 100% , with a specificity of 73 . 68% , for PB diagnosis . Compared to single markers , a 3-marker combination model that included ML2044-induced CXCL8/IL-8 , CCL4/MIP-1 beta , and IL-6 improved the diagnostic specificity to 94 . 7% for PB patients . ML2044-stimulated IL-4 and CXCL8/IL-8 achieved the highest sensitivity ( 85 . 71% and 100% ) and the highest specificity ( 95 . 24% and 84 . 21% ) for discriminating PB patients from HHCs and TB patients , respectively . Our findings suggest that the host markers induced by specific M . leprae antigens in an overnight WBA increase diagnostic and discriminatory value in PB patients with leprosy , with a particularly strong association with interleukin 8 .
Leprosy is a treatable infection that is caused by Mycobacterium leprae ( M . leprae ) and can result in skin lesions , nerve degeneration , and deformities . The current World Health Organization ( WHO ) directives for leprosy control programs encourage widespread administration of multidrug therapy ( MDT ) to treat patients and early diagnosis [1] . The implementation of WHO MDT treatment has drastically reduced the number of registered leprosy cases from the approximately 12 million reported cases in 1985 to fewer than 250 , 000 reported cases in 2006 [2] . In 2017 , 210 , 671 new cases of leprosy were detected , and the registered prevalence was 192 , 713 cases [3] . Indeed , assays for detecting M . leprae-specific IgM antibodies against PGL-I have been developed successfully [4 , 5] , and these assays are able to identify multibacillary ( MB ) leprosy patients ( with strong humoral immunity against M . leprae ) [6] . However , as only 1–5 skin lesions , 1–2 damaged nerves , and few bacteria are present , the diagnosis and discrimination of paucibacillary ( PB ) patients with leprosy remain challenging . Therefore , antigen-specific immune diagnostic tools , especially antigen-specific secretion of host markers in whole-blood assays ( WBAs ) , have been an important topic in leprosy research . Commercially available IFN-gamma ( IFN-γ ) release assays ( IGRAs ) such as Quanti-FERON-TB Gold have been developed successfully for specific detection of M . tuberculosis infection and discrimination from all ( nonvirulent ) BCG strains and most other nontuberculous mycobacteria ( NTMs ) [7] , which has inspired research into the feasibility of developing similar peptide-based assays for the identification of asymptomatic leprosy [6] . Multiple M . leprae-specific antigens and host markers , including IFN-γ as a candidate host marker , have been studied widely [8–17] . Several studies have explored the immune response in M . leprae-stimulated WBAs [8–17] . Some of these studies have examined the supernatant levels of IFN-γ stimulated with multiple M . leprae antigens in infected patients [8–14] , and others focused on a panel of multiple M . leprae antigen-induced host markers by WBA [15–17] . Sampaio et al . [13] reported previously that 9 of 33 M . leprae recombinant proteins could induce IFN-γ secretion in tuberculoid ( TT ) /borderline tuberculoid ( BT ) patients and HHCs by a WBA in a Brazilian population . However , our laboratory recently reported that IFN-γ secretion induced by stimulation with M . leprae antigens ( LID-1 , ML89 , ML2044 , and ML2028 ) achieved higher positive response rates in PB patients than in MB patients in Southwest China [14 , 15] . This marker could distinguish PB patients from tuberculosis ( TB ) patients after stimulation with ML2044 and ML2028 , but it could not distinguish PB patients from healthy household contacts ( HHC ) or endemic controls ( ECs ) [15] . This result was consistent with those of a previous study in an Ethiopian population , in which M . leprae proteins did not distinguish patients from ECs in one leprosy endemic area based on IFN-γ [12] . Geluk et al . [12] reported that M . leprae and ML2478 induced significantly higher concentrations of MCP-1 , MIP-1β and IL-1β in PB patients than in ECs in Ethiopia . These studies suggested that M . leprae antigen-specific IFN-γ secretion in a WBA had a limited ability to discriminate PB patients from nonleprosy controls; indeed , additional M . leprae-specific antigens and additional host biomarkers should be investigated for their ability to assist in the diagnosis and discrimination of PB patients . The purpose of this study was to explore a new panel of host markers stimulated by specific M . leprae antigens in an overnight WBA to improve the diagnosis ( PB patients vs . ECs ) and discrimination ( PB patients vs . HHCs or TB patients ) of PB patients with leprosy in a hyperendemic area in China . The M . leprae antigens used for the WBA in this study were Leprosy IDRI diagnostic-1 ( LID-1 ) and ML2044 . Additional host markers , including tumor necrosis factor alpha ( TNF-α ) , interleukin ( IL ) -4 , interleukin 6 , interleukin 10 , CC chemokine ligand ( CCL ) 2 , CCL4 , CXC chemokine ligand ( CXCL ) 8 , CXCL10 , granulocyte colony-stimulating factor ( G-CSF ) and granulocyte-macrophage colony-stimulating factor ( GM-CSF ) , were also studied .
This study was approved by the Medical Ethics Committee of Beijing Friendship Hospital , Capital Medical University , Beijing , P . R . China . Written informed consent was obtained from all adult participants . All of the procedures in this study that involved human participants were performed in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards . We recruited 13 MB patients , 7 PB patients , 21 HHCs , 20 ECs , and 19 TB patients from the Honghe Autonomous Prefecture in the Yunnan Province in the southwestern region of China during May 2016 . The leprosy incidence in most provinces and regions of China was lower than 0 . 1 per 100 , 000 in 2010 , and the case notification rate was up to 0 . 85 per 100 , 000 ( 190/4 , 470 , 000 ) from 2010–2014 in the communities in the Honghe Autonomous Prefecture [18] . The most common route of identification for the leprosy patients was out-patient ( 61/190 ) , followed by clue investigation ( 53/190 ) , family members ( 40/190 ) , disease-reporting ( 13/190 ) , epidemic area ( 12/190 ) , self-reporting ( 10/190 ) , and group survey ( 1/190 ) . Leprosy diagnosis was established based on clinical signs and symptoms , skin smears , skin biopsy , and neurophysiologic examinations . The leprosy patients were classified into groups based on Ridley and Jopling [19] . The leprosy patients were also classified into two groups , PB and MB , according to the WHO operational classification [20] . HHCs had been living in the same house as an adult leprosy patient . ECs within the normal controls lived in the same community as leprosy patients . TB patients were referred to the Honghe Autonomous Prefecture Disease Prevention and Control Center ( CDC ) . The specific M . leprae antigens used for the WBA in this study were LID-1 , a fusion protein developed by fusing the ML0405 and ML2331 genes [21 , 22] , and ML2044 , which were provided by Dr . M . S . Duthie from the Infectious Disease Research Institute ( IDRI ) in Seattle , USA . The list of accession numbers/ID numbers for genes and proteins that were mentioned in the text and included in the NCBI search is shown in S1 Table . An undiluted WBA with overnight incubation was performed as previously described [23] . Briefly , undiluted venous whole blood was collected into a heparinized tube or green top BD vacutainer . A 48-well ( flat-bottom ) plate was set up with the antigens and controls in a final volume of 0 . 5 ml . The specific M . leprae antigens ML2044 and LID-1 ( 100 μg/ml ) were used as stimuli . Phytohemagglutinin ( PHA-M , Cat No: L2646 , 750 μg/ml , Sigma-aldrich , Fluke , USA ) was included as a positive control , and 0 . 01 mol/l PBS was included as a negative control . The antigens were added in a volume of 50 μl per well , followed by the addition of 450 μl of blood . The plate was sealed with micropore tape to avoid evaporation during incubation at 37°C with 5% CO2 . After 24 hours of incubation , the supernatants were harvested and stored at −20°C until they were assayed for cytokines and chemokines by Luminex multiplex assays . The potential diagnostic value of host markers was analyzed with Luminex multiplex assays of supernatant samples collected from an overnight WBA; these markers included TNF-α , IL-4 , IL-6 , IL-10 , CCL 2 , CCL4 , CXCL 8 , CXCL10 , G-CSF and GM-CSF . The concentrations of these cytokines were measured by a customized Human Premixed Multi-Analyte Kit ( Cat . No . LXSAHM-10 ) on the Luminex-200™ system and the Xmap Platform ( Luminex Corporation , Austin , TX ) . The list of accession numbers/ID numbers for genes and proteins of the host markers mentioned in the text and included in the HGNC and NCBI search is shown in S2 Table . Statistical analysis was performed primarily with GraphPad Prism software version 5 . 0 ( GraphPad Software Inc . , San Diego , CA , USA ) . The nonparametric Mann-Whitney U test was used to analyze differences between two groups ( PB patients vs . MB patients , HHCs , ECs , or TB patients ) . Probability ( p ) values less than 0 . 05 were considered significant . The diagnostic utility of individual M . leprae antigen-specific responses for leprosy , including sensitivity , specificity , p value , 95% confidence intervals ( CI ) , cutoff value , area under the receiver operator characteristic curve ( AUC ) , and receiver operator characteristics ( ROC ) , was ascertained by ROC curve analysis based on the highest likelihood ratio .
A total of eighty participants , including leprosy patients ( 13 MB patients and 7 PB patients ) and controls ( 21 HHCs , 20 ECs , and 19 TB patients ) were included in the study . The basic characteristics of the participants are presented in Table 1 . Both specific M . leprae antigens ( ML2044 and LID-1 ) were evaluated in all 13 MB cases , 7 PB cases , 21 HHCs , 19 TB cases , and in different numbers of ECs ( 19 ECs for ML2044 and 20 ECs for LID-1 ) . Newly diagnosed MB and PB leprosy patients undergoing MDT treatment . The median and interquartile range ( IQR ) of the treatment duration were 9 months ( 5–16 months ) and 10 months ( 2–12 months ) for MB and PB patients , respectively . After stimulation with ML2044 ( S1 Fig ) and LID-1 ( S2 Fig ) in an overnight WBA , the concentrations of selected host markers were determined in each participant and compared between PB patients and MB patients , HCCs , ECs , and TB patients ( Table 2 ) . After stimulation with ML2044 , the supernatants of PB patients had a higher level of IL-4 ( median 46 pg/ml ) than those of HHCs ( median 5 . 24 pg/ml , p = 0 . 0004 ) , ECs ( median 5 . 24 pg/ml , p = 0 . 0026 ) , or TB patients ( median 5 . 24 pg/ml , p = 0 . 0005 ) . The same trend was also found for IL-6 ( PB patients vs . HHCs , ECs , and TB patients: median = 30 . 66 vs . 2 . 086 , 3 . 661 , and 3 . 271 pg/ml , p = 0 . 0225 , 0 . 0109 , and 0 . 0192 , respectively ) ; CCL4/MIP-1 beta ( PB patients vs . HHCs , ECs , and TB patients: median = 2414 vs . 470 . 9 , 470 . 9 , and 193 . 9 pg/ml , p = 0 . 0101 , 0 . 0066 , and 0 . 0022 , respectively ) ; CXCL8/IL-8 ( PB patients vs . HHCs , ECs , and TB patients: median = 1060 vs . 708 . 5 , 766 . 1 , and 222 . 7 pg/ml , p = 0 . 0139 , 0 . 0185 , and 0 . 0019 , respectively ) ; G-CSF ( PB patients vs . HHCs , ECs , and TB patients: median = 77 . 56 vs . 17 . 33 , 17 . 33 , and 17 . 33 pg/ml , p = 0 . 0013 , 0 . 0006 , 0 . 0046 , respectively ) ; and TNF-α ( PB patients vs . TB patients: median = 8 . 438 vs . 2 . 963 pg/ml , p = 0 . 0423 ) . For ML2044-stimulated CXCL10/IP-10 , the concentration detected in PB patients ( median = 76 . 06 pg/ml ) was higher than that detected in HHCs ( median = 59 . 5 pg/ml , p = 0 . 0496 ) but lower than that detected in ECs ( median = 179 . 4 , p = 0 . 0093 ) . Although the median of the ML2044-induced IL-10 concentration differed significantly between the PB and EC groups ( median 1 . 207 vs . 1 . 207 , IQR 0 . 2437–1 . 207 vs . 1 . 207–1 . 207 , p = 0 . 0058 ) , the ROC analysis precluded the determination of a discriminatory value for PB patients and ECs . The concentration of CCL4/MIP-1 beta in whole blood cells during the response to LID-1 was elevated in PB patients compared with that in ECs or TB patients ( median 1379 vs . 338 , and 362 . 2 pg/ml , p = 0 . 0217 , and 0 . 0242 , respectively ) , as was the concentration of CXCL8/IL-8 between PB patients and ECs ( median 1325 vs . 349 . 8 pg/ml , p = 0 . 0355 ) and the concentration of G-CSF between PB patients and HHCs ( median 70 . 75 vs . 51 . 43 pg/ml , p = 0 . 0214 ) . In contrast , the concentration of LID-1-induced CXCL10/IP-10 was lower in PB patients than in TB patients ( median 62 . 32 vs . 124 . 3 pg/ml , respectively , p = 0 . 0129 ) . No differences in host immune markers induced by ML2044 or LID-1 stimulation were found between PB patients and MB patients . The levels of six out of the 10 markers evaluated in this study ( IL-4 , IL-6 , CCL4/MIP-1 beta , CXCL8/IL-8 , CXCL10/IP-10 , and G-CSF ) were significantly higher in ML2044-stimulated supernatants from PB patients than in those from ECs ( S3 Table ) . The AUC values ranged from 0 . 81 to 0 . 95 ( S3 Table , Fig 1A–1F ) . For LID-1 , the levels of two markers ( CCL4/MIP-1 beta and CXCL8/IL-8 ) were significantly higher in PB patients than in ECs . These two analytes discriminated between PB patients and ECs with AUC = 0 . 80 , and 0 . 77 , respectively , with a sensitivity of 57 . 14% and a specificity of 95 . 00% for both markers ( S3 Table , Fig 1G and 1H ) . Combining the results of the host markers stimulated by the two M . leprae antigens , ML2044-stimulated CXCL8/IL-8 achieved the highest sensitivity of 100% , with a specificity of 73 . 68% , for PB diagnosis . When ML2044-stimulated analyte levels were compared between the PB patients and HHCs , similar to the results obtained for PB patients vs . ECs , significant differences were obtained for IL-4 , IL-6 , CCL4/MIP-1 beta , CXCL8 , CXCL10/IP-10 , and G-CSF . The AUC values for all six of these markers were ≥0 . 73 by ROC analysis . The sensitivities of the six analytes for PB patients ranged from 28 . 75% to 85 . 71% , with specificities from 90 . 48% to 95 . 24% ( Fig 2 , S4 Table ) . For LID-1-stimulated supernatants , significant differences between the PB patients and HHCs were obtained for one host marker , G-CSF . After ROC analysis , the AUC for LID-1 stimulated G-CSF was 0 . 80 . Among the host markers stimulated by the two M . leprae antigens , ML2044-stimulated IL-4 achieved the highest sensitivity ( 85 . 71% ) and specificity ( 95 . 24% ) for discriminating PB patients from HHCs . ML2044-induced IL-4 , IL-6 , CCL4/MIP-1 beta , CXCL8/IL-8 , G-CSF and TNF-α levels were significantly higher in the PB patients than in the TB patients . After ROC analysis , the AUC values for all six analytes ranged from 0 . 7669 to 0 . 9549 ( Fig 3 , S5 Table ) . For LID-1-stimulated supernatants , significant differences between the PB patients and TB patients were obtained for two host markers , CCL4/MIP-1 beta and CXCL10/IP-10 . After ROC analysis , the AUC values for these markers were 0 . 80 and 0 . 83 , respectively . Among the host markers stimulated by the two M . leprae antigens , ML2044-stimulated CXCL8/IL-8 achieved the highest sensitivity of 100% , with a specificity of 84 . 21% , for discriminating PB patients from TB patients . Although the sensitivity of ML2044-induced CXCL8/IL-8 ( up to 100% for the discrimination of PB patients from ECs ) predominated over any other single host marker induced by ML2044 or LID-1 , the specificity of this marker reached only 73 . 68% , far lower than those of IL-6 , CXCL10/IP-10 , CCL4/MIP-1 beta , IL-4 , and G-CSF induced by ML2044 ( specificity = 94 . 74% for each marker ) or CXCL8/IL-8 and CCL4/MIP-1 beta induced by LID-1 ( specificity = 95% for each marker ) , as shown in S3 Table . To increase the specificity of diagnosis between PB patients and ECs , different combination models of 3–6 M . leprae-specific antigen-induced host markers were used to classify the participants . The cutoff value obtained using the ROC analysis described above was used to classify individuals into two groups . A 3-marker model ( ML2044-induced CXCL8/IL-8 , CCL4/MIP-1 beta , and IL-6 ) was the best model and was superior to the use of two M . leprae-specific antigens ( ML2044 and LID-1 ) or 4 , 5 or 6 host markers in combination . Participants were allocated to clinical groups according to the results of the majority of the individual marker tests ( 2/3 or 3/3 ) . Accordingly , the ML2044-induced 3-host marker combination model classified 24 of the 26 participants ( 92 . 31% ) in the correct clinical groups , with 85 . 71% sensitivity , which was lower than the sensitivity of ML2044-induced IL-8 as a single marker ( 100% ) , but the specificity of diagnosis for PB leprosy patients increased from 73 . 68% to 94 . 74% . Two ( 7 . 7% ) participants were misclassified , with 1 false-negative and 1 false-positive classification . The classification of individual participants , sensitivity values and specificity values are shown in Fig 4 . Despite the high cost and low efficiency , combined utilization of ML2044-induced CXCL8/IL-8 as a single marker and the 3-marker model ( ML2044-induced CXCL8/IL-8 , CCL4/MIP-1 beta , and IL-6 ) would help to both maintain high sensitivity and enhance specificity . Seven PB patients , as defined by Jopling , and 19 HCCs were analyzed . The concentrations and the best cutoff were determined for each cytokine or chemokine , as described in S3 Table . The cutoff was used to define whether the concentrations of cytokines and chemokines indicated that the participant was a PB patient or an EC . Each cytokine or chemokine was used as an independent marker . Prediction of PB patients is shown in dark gray , and prediction of ECs is shown in light gray , for 3 different phenotypic markers . When ≥ 2 phenotypes supported one of the diagnoses , a final diagnosis of either PB ( black ) or EC ( white ) was made . Subsequently , we explored a model with 3 or more markers for distinguishing PB patients from HHCs or TB patients in the same way . The sensitivity and the specificity of a single marker ( ML2044-induced IL-4 ) reached 85 . 71% and 100% for distinguishing PB patients from HHCs . Only the 3-marker model ( ML2044-induced IL-4 , LID-1-induced G-CSF , and ML2044-induced CCL4/MIP-1 beta ) achieved the same sensitivity and specificity as the single marker ( ML2044-induced IL-4 ) , as shown in S3 Fig . In addition , the sensitivity and the specificity of a single marker ( ML2044-induced CXCL8 ) reached 100% and 94 . 74% . Only the 3-marker model ( ML2044-induced CXCL8 , CCL4/MIP-1 beta , and IL-4 ) was found to have the same sensitivity and specificity as ML2044-induced CXCL8 , as shown in S4 Fig .
In conclusion , we identified a biosignature of a single M . leprae-specific host marker in antigen-stimulated overnight WBAs ( ML2044-induced CXCL8/IL-8 ) that showed potential for the diagnosis of PB disease , with accurate prediction of 100% of PB cases and 73 . 68% of EC cases . The sensitivity of this analyte model was better than that of any other single host marker , but the specificity was relatively low . A 3-marker model of ML2044-induced CXCL8/IL-8 , CCL4/MIP-1 beta , and IL-6 improved the specificity of diagnosis between PB patients and ECs to 94 . 74% . Moreover , ML2044-induced CXCL8/IL-8 and ML2044-induced IL-4 dominated in discriminating PB patients from TB patients and HHCs , respectively . However , both diagnostic performance at the time of screening and assessment of immunological changes in host markers during MDT therapy need to be further evaluated before these diagnostic approaches can be recommended for routine clinical practice .
|
Leprosy , caused by Mycobacterium leprae , affects over 200 , 000 people annually worldwide and remains endemic in the ethnically diverse , mountainous and underdeveloped southwestern regions of China . Although it is curable , delayed diagnosis of leprosy persists in China , with a disability rate as high as 20% nationwide . To identify the source of infection and block transmission more effectively , further knowledge about the diagnostic value of antigen-specific induced host immune responses in patients is needed . The current study aimed to evaluate the diagnostic value of an overnight whole-blood assay for the diagnosis of paucibacillary ( PB ) leprosy patients , and differences in the ability of specific M . leprae antigens to stimulate a panel of host markers was tested by an overnight whole-blood assay . Our findings suggest that host markers induced by specific M . leprae antigens in an overnight WBA have diagnostic value in leprosy patients and discriminatory value between leprosy patients and healthy household contacts ( HHCs ) or tuberculosis ( TB ) patients .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2019
|
Host immune responses induced by specific Mycobacterium leprae antigens in an overnight whole-blood assay correlate with the diagnosis of paucibacillary leprosy patients in China
|
Bordetella adenylate cyclase toxin ( CyaA ) binds the αMβ2 integrin ( CD11b/CD18 , Mac-1 , or CR3 ) of myeloid phagocytes and delivers into their cytosol an adenylate cyclase ( AC ) enzyme that converts ATP into the key signaling molecule cAMP . We show that penetration of the AC domain across cell membrane proceeds in two steps . It starts by membrane insertion of a toxin ‘translocation intermediate’ , which can be ‘locked’ in the membrane by the 3D1 antibody blocking AC domain translocation . Insertion of the ‘intermediate’ permeabilizes cells for influx of extracellular calcium ions and thus activates calpain-mediated cleavage of the talin tether . Recruitment of the integrin-CyaA complex into lipid rafts follows and the cholesterol-rich lipid environment promotes translocation of the AC domain across cell membrane . AC translocation into cells was inhibited upon raft disruption by cholesterol depletion , or when CyaA mobilization into rafts was blocked by inhibition of talin processing . Furthermore , CyaA mutants unable to mobilize calcium into cells failed to relocate into lipid rafts , and failed to translocate the AC domain across cell membrane , unless rescued by Ca2+ influx promoted in trans by ionomycin or another CyaA protein . Hence , by mobilizing calcium ions into phagocytes , the ‘translocation intermediate’ promotes toxin piggybacking on integrin into lipid rafts and enables AC enzyme delivery into host cytosol .
The secreted adenylate cyclase toxin-hemolysin ( CyaA , ACT , or AC-Hly ) plays a key role in virulence of Bordetellae . This multifuctional protein binds the αMβ2 integrin ( CD11b/CD18 , CR3 or Mac-1 ) of myeloid phagocytic cells and delivers into their cytosol a calmodulin-activated adenylate cyclase enzyme that ablates bactericidal capacities of phagocytes by uncontrolled conversion of cytosolic ATP to the key signaling molecule cAMP [1]–[5] . In parallel , the hemolysin moiety of CyaA forms oligomeric pores that permeabilize cell membrane for monovalent cations and contribute to overall cytoxicity of CyaA towards phagocytes [6]–[10] . The toxin is a 1706 residues-long protein , in which a calmodulin-activated adenylate cyclase ( AC ) enzyme domain of ∼400 N-terminal residues is fused to a ∼1300 residue-long RTX ( Repeats in ToXin ) cytolysin moiety [11] . The latter consists itself of three functional domains typical for RTX hemolysins . It harbors , respectively , ( i ) a hydrophobic pore-forming domain , ( ii ) a segment recognized by the protein acyltransferase CyaC , activating proCyaA by covalent post-translational palmitoylation at ε-amino groups of Lys860 and Lys983 [12] , [13] , and ( iii ) an assembly of five blocks of the characteristic glycine and aspartate-rich nonapeptide RTX repeats that form numerous ( ∼40 ) calcium-binding sites [14] . Since no structural information on the RTX cytolysin moiety is available , the mechanistic details of toxin translocation across the lipid bilayer of cell membrane remain poorly understood . Delivery of the AC domain into cells occurs directly across the cytoplasmic membrane , without the need for toxin endocytosis [15] and requires structural integrity of the CyaA molecule [6] , unfolding of the AC domain [16] and a negative membrane potential [17] . Recently , we described that CyaA forms a calcium-conductive path in cell membrane and mediates influx of extracellular Ca2+ ions into cell cytosol concomitantly with translocation of the AC domain polypeptide into cells [18] . The current working model predicts that both Ca2+ influx and AC translocation depend on a different membrane-inserted CyaA conformer than the pore-forming activity [19] , [20] . The two membrane activities of CyaA , however , appear to use the same essential amphipatic transmembrane segments within the pore-forming domain ( α-helix502–522 and α-helix565–591 ) , employing them in an alternative and mutually exclusive way . These segments harbor two pairs of negatively charged glutamate residues ( Glu509/Glu516 and Glu570/Glu581 ) that were found to play a central role in toxin action on cell membrane . These control , respectively , the translocation of the positively charged AC domain , the formation of oligomeric CyaA pores and the cation-selectivity of the CyaA pore . Charge-reversing , neutral or helix-breaking substitutions of these glutamates were found to shift the balance between AC translocating and pore-forming activities of CyaA on cell membrane [8] , [10] , [19] , [20] . The very high specific AC enzyme activity of CyaA allowed previously to detect its capacity to promiscuously bind and penetrate at reduced levels also numerous cell types lacking the CD11b/CD18 receptor [21] , [22] . This is likely due to a weak lectin activity of CyaA , which would enable interaction of the toxin with cell surface gangliosides [23] and glycoproteins [24] . Indeed , binding of CyaA to CD11b/CD18 was recently found to depend on initial interaction with the N-linked glycan antenna of the receptor [24] , where the specificity of CyaA for CD11b/CD18 appears to be determined by a segment of the stalk domain of the CD11b subunit ( Osicka et al . , manuscript in preparation ) . CD11b/CD18 belongs to the β2 subfamily of polyfunctional integrins playing a major role in leukocyte function . The same β2 subunit ( CD18 ) can , indeed , pair with four distinct α subunits to yield the αLβ2 ( CD11a/CD18 , LFA-1 ) , αMβ2 ( CD11b/CD18 , CR3 , Mac1 ) , αXβ2 ( CD11c/CD18 , p150/195 ) and αDβ2 ( CD11d/CD18 ) receptors , respectively [25] . Among key features of these integrins is their capacity of bi-directional signaling , where the avidity and conformation of the integrins is regulated by intracellular signals in the ‘inside-out’ signaling mode . In turn , binding of ligands or counter-receptors results in ‘outside-in’ signaling [26] . Among other effects , the latter yields actin cytoskeletal rearrangements and can result in lateral segregation of the β2 integrins from the bulk phase of the plasma membrane into distinct lipid assemblies known as lipid rafts [27] , [28] . These were first detected as detergent-resistant membrane ( DRM ) , characterized by insolubility in some detergents under certain conditions and enriched in cholesterol , sphingolipids , and glycosylphosphatidylinositol-anchored proteins [29]–[32] . Besides playing an important role in signal transduction , receptor internalization , vesicular sorting or cholesterol transport [33] , the components of lipid rafts are often exploited as specific receptors mediating cell entry of toxins , pathogenic bacteria , or viruses [34]–[36] . Here , we show that CyaA-mediated influx of Ca2+ ions into cells induces mobilization of the toxin-receptor complex into lipid rafts , where translocation of the AC domain across cytoplasmic membrane is accomplished .
To examine whether CyaA localizes to lipid rafts , murine J774A . 1 monocytes exposed to 1 nM CyaA ( 176 ng/ml , 37°C , 10 min ) were lyzed with ice-cold Triton X-100 and detergent-resistant membrane ( DRM ) was separated from soluble cell extracts by flotation through sucrose density gradients . As shown in Fig . 1A , while the CD71 marker of bulk membrane phase was exclusively detected in the soluble extract at the bottom of the gradient , up to 30% of total loaded CyaA was found to float in fraction 3 at a lower buoyant density towards the top of the gradient , together with the DRM marker protein NTAL ( see Fig . 1E for quantification ) . Notably , while over 50% of the full-length CyaA molecules ( ∼200 kDa ) remained in the soluble phase at the bottom of the gradient , the floating DRM fractions were selectively enriched in a processed CyaA form of ∼160 kDa , representing up to 60% of total CyaA in the fraction 3 of the gradient ( see Fig . 1F for quantification ) . This appeared to have the entire AC domain-cleaved off , as it could only be detected by the 9D4 antibody recognizing the C-terminal RTX repeats and not by the 3D1 antibody binding between residues 373 and 399 of the C-terminal end of the AC domain of CyaA [37] . Most CyaA molecules accumulating in DRM appeared , hence , to have the AC domain translocated across cellular membrane and accessible to processing by intracellular proteases [10] , [38] . As further documented in Fig . 1B , no CD11b was floating with DRM from mock-treated J774A . 1 cells , or when toxin binding occurred at 4°C . In turn , exposure of monocytes to 1 nM CyaA at 37°C , resulted in mobilization of over 30% of total cellular CD11b into the floating DRM , showing that CyaA relocated from the bulk of the membrane to DRM together with its αMβ2 integrin receptor . This was , however , not due to any generalized clustering and mobilization of β2 integrins into rafts resulting from toxin action , as the highly homologous CD11a subunit of the other β2 integrin expressed by J774A . 1 cells ( LFA-1 ) , was not mobilized into DRM ( Fig . 1B ) . Hence , the relocation of CD11b/CD18 into rafts was specifically due to interaction with CyaA . To assess whether CyaA association with DRM depended on toxin interaction with CD11b/CD18 , we used Chinese hamster ovary ( CHO ) cells that do not express any β2 integrins unless transfected by genes encoding CD11b and CD18 subunits ( CHO-CD11b/CD18 ) . As shown in Fig . 1C , even when the CyaA concentration was raised to 113 nM , to obtain detectable amounts of the toxin associated with mock-transfected CHO cells , CyaA was detected exclusively in the soluble extract at the bottom of the gradient . In contrast , association of CyaA and CD11b/CD18 with DRM was detected already upon treatment of CHO-CD11b/CD18 transfectants with 1 nM CyaA ( Fig . 1D ) . This showed that CyaA depended on binding to CD11b/CD18 for association with DRM and it was able to mobilize CD11b/CD18 into DRM independently of the myeloid cell background . Since CyaA exerts several activities on cells in parallel , we analyzed which of them enabled mobilization of the CyaA-CD11b/CD18 complex into DRM . Towards this aim , we used a specific set of CyaA variants that retain the capacity to bind CD11b/CD18 , while lacking one or more of the other CyaA activities ( Table 1 ) . As documented in Fig . 2A , the capacity of CyaA to elevate cellular cAMP concentrations was not required for mobilization of CyaA into DRM . The enzymatically-inactive CyaA-AC− construct , unable to catalyze conversion of ATP into cAMP , was indeed accumulating in DRM with the same efficacy as the intact CyaA ( fractions 3–4 ) . Fatty-acylation of CyaA as such was also not essential for association of CyaA with DRM . As further shown in Fig . 2A , the non-acylated proCyaA was detected in DRM despite an importantly reduced capacity to associate with cells . Moreover , the pore-forming activity of CyaA was both insufficient and dispensable for mobilization of CyaA into DRM . The acylated CyaAΔAC construct , lacking the entire AC domain but retaining an intact pore-forming ( hemolytic ) capacity , was unable to mobilize into DRM ( Fig . 2A ) . In contrast , the CyaA-E570Q+K860R-AC− construct unable to permeabilize cells to any significant extent , but exhibiting an intact capacity to translocate the AC domain across cell membrane was , indeed , recruited into DRM together with CD11b/CD18 as efficiently as intact CyaA . In turn , the CyaA-E570K+E581P double mutant that was unable to form CyaA pores , or to translocate the AC domain across membrane , and retained only the CD11b-binding capacity ( Table 1 ) , was also unable to associate with DRM . To corroborate these observations , we used fluorescence microscopy to examine the distribution of individual CyaA proteins in cell membrane . As documented in Fig . 2B , the intact CyaA , CyaA-AC− and CyaA-E570Q+K860R-AC− proteins were found to induce formation of , and to localize within , patches on cell membrane . Moreover , the same patches were labeled to high extent also with B subunit of cholera toxin ( CtxB ) , which specifically binds the GM1 ganglioside accumulating in lipid rafts . Hence , the three CyaA variants capable of associating with DRM ( cf . Fig . 2A ) were also found to co-localize with CtxB within membrane patches . In turn , no formation of membrane patches , a diffuse distribution on cell surface , and low if any co-localization with CtxB , were observed for the CyaAΔAC and CyaA-E570K+E581P constructs that were unable to associate with DRM , too . The pattern of DRM association , processing to the 160 kDa form and co-localization of the different CyaA variants with CtxB , respectively , resembled strongly the pattern of structure-function relationships observed recently for the capacity of CyaA to promote influx of extracellular calcium ions into J774A . 1 cells [18] . Indeed , as documented in Fig . 2C by measurements of intracellular calcium concentrations ( [Ca2+]i ) , the CyaA , CyaA-AC− and CyaA-E570Q+K860R-AC−proteins ( 17 nM ) exhibited an expected capacity to promote Ca2+ influx into J774A . 1 cells ( see [18] for details on different kinetics of Ca2+ entry for AC− and AC+ constructs ) . In contrast , the CyaA-E570K+E581P and CyaAΔAC constructs , failed to mediate any increase of [Ca2+]i even when used at a 113 nM concentration ( Fig . 2D ) . Collectively , hence , these results show that the capacity of different CyaA variants to associate with DRM and co-localize with CtxB within coalesced lipid rafts was mirrored by the capacity to promote Ca2+ influx into cells . We showed recently that CyaA-mediated influx of Ca2+ into cells is independent of the AC enzyme or pore-forming ( hemolytic ) activities of CyaA and occurs concomitantly to translocation of the AC domain across target cell membrane [18] . It remained , however , to assess whether it was the mere insertion of a CyaA translocation precursor into cell membrane , or whether the accomplishment of translocation of the AC domain across cell membrane was required for formation of a calcium conductive path in cell membrane . Towards this aim , we used the 3D1 monoclonal antibody ( MAb ) that binds to the distal end of the AC domain ( residues 373 to 400 ) and was previously shown to block membrane translocation of the AC domain of cell-associated CyaA [39] . As expected and documented in Fig . 3A , preincubation of CyaA with the 3D1 MAb did not affect the capacity of CyaA to bind J774A . 1 cells , while strongly inhibiting AC domain delivery and cAMP concentration elevation in cells . However , as revealed by detection of both CyaA and 3D1 in the fraction 3 of the sucrose gradient shown in Fig . 3B , the membrane-inserted CyaA with bound 3D1 MAb still associated with DRM at the same levels as CyaA alone . Moreover , as also shown in Fig . 3B , due to 3D1-mediated inhibition of AC domain translocation , the relative amount of processed CyaA ( ∼160 kDa ) in the DRM fractions decreased to 10 to 15% of total present CyaA , while over 50% of CyaA in DRM was processed in the presence of isotype control . As shown in Fig . 3C , however , despite arrested AC domain translocation , the CyaA-3D1 complex promoted elevation of [Ca2+]i in cells with kinetics resembling the Ca2+ influx produced by CyaA-AC− ( cf . Fig . 2C ) . Hence , 3D1 binding uncoupled translocation of the AC domain from membrane insertion of CyaA and ‘locked’ the toxin in the conformation of a ‘translocation intermediate’ that permeabilized cells for Ca2+ ions and associated with DRM . Next , we aimed to determine whether elevation of [Ca2+]i as such would mobilize into DRM also the CyaA-E570K+E581P protein unable to associate with DRM on its own ( cf . Fig . 2 ) . As demonstrated in Fig . 4 , upon permeabilization of cells for extracellular Ca2+ ions with the Ca2+ ionophore ionomycin ( 500 nM ) , up to 15% of the added CyaA-E570K+E581P was found associated with DRM . In contrast , no association of CyaA-E570K+E581P with DRM was observed upon treatment of cells with 1 µM thapsigargin that increases [Ca2+]i by triggering Ca2+ release from intracellular stores . This showed that entry of extracellular Ca2+ across the cytoplasmic membrane was required for mobilization of CyaA into DRM . Influx of extracellular Ca2+ during leukocyte activation was reported to induce mobilization of integrins in cell membrane by calpain-mediated cleavage of talin that tethers β2 integrins to actin cytoskeleton [40] , [41] . Therefore , we examined whether CyaA-promoted recruitment of the toxin receptor into rafts depended on talin processing . As shown in Fig . 5A , intact talin ( ∼270 kDa ) was largely predominating in lyzates of cells treated with the CyaA-ΔAC or CyaA-E570K-E581P proteins that are unable to promote Ca2+ entry into cells . In contrast exposure of cells to the CyaA , CyaA-AC− , or CyaA-E570Q+K860R-AC− proteins , promoting influx of Ca2+ into cells , increased about seven-fold the detected amounts of the ∼220 kDa C-terminal fragment of processed talin ( Fig . 5A , left panel ) . Concomitantly , increased amounts of the 47-kDa N-terminal fragment of talin ( talin head ) were detected in cell lyzates . Moreover , tightly associated talin head was found to float together with the CD11b/CD18 heterodimer in DRM ( Fig . 5B ) and could be co-immunoprecipitated with the integrin on beads coated with anti-CD11b antibody ( Fig . 5C ) . This CyaA-induced processing of talin was clearly due to activation of calpain , as preincubation of cells with 100 µM calpain inhibitor , calpeptin , blocked talin cleavage in CyaA-treated cells ( Fig . 5A , right panel ) . Remarkably , pretreatment of cells with calpeptin strongly inhibited also the association of CyaA with DRM ( Fig . 5D ) and decreased by at least a factor of two the capacity of cell-associated CyaA to translocate the AC enzyme into target cells ( Fig . 5E ) . In turn , no effect of calpain inhibition was observed for CyaA-mediated Ca2+ influx ( Fig . 5F ) . In line with these results , pretreatment of cells with 100 µM calpeptin blocked effectively also the formation of CyaA-AC− patches in cell membrane and ablated co-localization of CyaA-AC− with CtxB , as documented in Fig . 5G . It can , hence , be concluded that CyaA-mediated influx of Ca2+ into cells activated cleavage of talin by calpain and this was required for mobilization of CyaA-CD11b/CD18 complexes into lipid rafts . To determine what role does association of CyaA with lipid rafts play in the mechanism of toxin action on cellular membrane , we analyzed the activities of CyaA on cells having the rafts disrupted by depletion of cholesterol . As shown in Table 2 , the total cholesterol content of J774A . 1 cells could be decreased about two-fold by cholesterol extraction with 10 mM MβCD for 30 min . While the disruption of raft structures did not impact on association of CyaA with cells ( Fig . 6A and Fig . S1 ) , the modest decrease of cellular cholesterol content yielded an about five-fold decrease of the capacity of CyaA to translocate the AC domain across cell membrane . This defect was further mirrored by decreased DRM association of CyaA in MβCD-extracted cells , as shown in Fig . 6B . In parallel , the specific capacity of CyaA to promote Ca2+ influx into cholesterol-depleted cells was reduced and the [Ca2+]i increase ensuing toxin addition was delayed by several minutes , reaching a plateau at about a half-maximal [Ca2+]i concentration , as compared to non-depleted cells ( Fig . 6C ) . In line with this , the two-fold decrease of cellular cholesterol level moderately decreased also the co-localization of CyaA with CtxB in lipid rafts ( Fig . 6D ) . Therefore , the above described experiments were replicated on monocytic U937 histiocytic lymphoma cells ( CD11b+ ) that are defective in endogenous cholesterol synthesis . These cells can be efficiently depleted of cholesterol without losing viability , by them growing for 48 hours in media containing cholesterol-free ( delipidated ) serum . As shown in Table 2 , such treatment reduced the cholesterol content of U937 cells almost 10-times . As shown in Fig . 6E , a pronounced , over ten-fold decrease of specific AC translocation capacity of CyaA was observed on U937 cells grown in media with delipidated serum , as compared to CyaA activity on cells grown with standard serum . At the same time , however , the total amounts of cell-associated CyaA remained equal , irrespective of cell treatment . However , by difference to well-detectable DRM association of CyaA on cholesterol-replete U937 cells , grown with standard serum , no association of CyaA with DRM was observed in lyzates of cholesterol-depleted U937 cells grown in delipidated serum , respectively ( Fig . 6F ) . Intriguingly , compared to the Ca2+ influx elicited by equal concentrations of CyaA in J774A . 1 cells , about an order of magnitude lower amplitude and delayed kinetics of CyaA-mediated Ca2+ influx was observed for U937 cells grown in media with standard serum ( cf . Fig . 6C and Fig . 6G ) . These cells exhibited a 3-fold lower cholesterol content than the J774A . 1 cells ( see Table 2 ) , suggesting that the low cholesterol content of U937 cells might have accounted for the poor capacity of CyaA to elicit Ca2+ influx in these cells . Indeed , when cholesterol content of J774A . 1 cells was reduced about two-fold by cholesterol extraction with 10 mM MβCD , a delayed kinetics of CyaA-induced influx of Ca2+ into J774A . 1 cells and a two-fold lower final [Ca2+]i reached in 20 minutes , were also observed ( cf . Fig . 6C ) . Similarly , a delayed influx of Ca2+ and a lower final level of [Ca2+]i was observed also upon addition of equal CyaA concentrations to U937 cells depleted of cholesterol by growth in delipidated media , as compared to U937 cells grown in standard media , as shown in Fig . 6G . At the same time , however , the respective amounts of CyaA associated per 106 J774A . 1 or U937 cells remained the same ( ∼5 ng of CyaA bound per 106 cells ) , irrespective of whether the cholesterol content of cells was decreased by the treatments ( cf . Fig . 6A and 6E ) . These results , hence , strongly point towards a close relation between the overall content of cholesterol in cellular membrane and the propensity of the membrane-inserted CyaA to adopt the ‘translocation intermediate’ conformation , which would account for the Ca2+ conducting path across cell membrane ( cf . Fig . 3 and [18] ) . Finally , a correspondingly reduced CtxB binding and little if any co-localization of CtxB with CyaA were observed on cholesterol-depleted U937 cells , grown in delipidated serum , as compared to binding and some observable co-localization of CyaA with CtxB on cholesterol-replete U937 cells ( Fig . 6 ) . In the light of the above results , we aimed to test the hypothesis that AC translocation across membrane was supported and accomplished upon recruitment of the membrane-associated toxin into the cholesterol-rich environment of lipid rafts . Therefore , we examined whether the inactive CyaA-E570K+E581P construct would gain any capacity to translocate its enzymatically active AC domain across cellular membrane upon mobilization into lipid rafts . Since this mutant is intact for receptor binding but fails to promote Ca2+ influx into cells , we reasoned that mobilizing Ca2+ ions into cells in trans , by co-incubation with a translocating CyaA-AC− toxoid , might promote recruitment of CyaA-E570K+E581P mutant into rafts to some extent . As shown in Fig . 7A , when biotinylated CyaA-E570K+E581P was added to cells alone , or when it was co-incubated with equal amounts of the enzymatically inactive CyaA-E570K+E581P-AC− toxoid , unable to cause calcium influx , the CyaA-E570K+E581P-biotin failed to associate with DRM . In contrast , upon co-incubation with equal amounts of the translocating CyaA-AC− toxoid ( 1∶1 ) , a significant fraction of CyaA-E570K+E581P-biotin associated with DRM . Moreover , as shown in Fig . 7B , this mobilization into DRM was paralleled by a doubling of the residual capacity of the CyaA-E570K+E581P variant to deliver the AC domain across cell membrane and elevate cytosolic cAMP concentrations ( Fig . 7B ) . Thus , recruitment into cholesterol-rich lipid rafts enhanced the residual AC translocating activity of this defective CyaA variant .
We show here that membrane translocation of the adenylate cyclase domain of CyaA occurs by a two step mechanism and involves toxin piggybacking on the αMβ2 integrin for relocation into lipid rafts . The present results allow us to propose a new model of CyaA mechanism of action , as summarized in Fig . 8 . Upon initial binding of CyaA to the CD11b/CD18 receptor distributed in the bulk phase of cell membrane , a ‘translocation intermediate’ of CyaA would insert into the cytoplasmic membrane . It is assumed that in this ‘translocation intermediate’ a part of the AC domain is already inserted within the membrane and is shielded form the lipids by association with the amphipathic α-helical transmembrane segments of the hydrophobic domain of CyaA ( residues 502–522 , 529–549 , 571–591 , 607–627 and 678–698 [19] , [20] ) . This ‘translocation intermediate’ then forms a path conducting external Ca2+ ions across cellular membrane into the submembrane compartment of cells . Incoming calcium ions activate the Ca2+-dependent protease calpain , located in the submembrane compartment , which produces cleavage of the talin tether . This liberates the toxin-receptor complex from association with actin cytoskeleton and mobilizes it for recruitment into lipid rafts . Within the specific liquid-ordered environment of cholesterol-rich lipid rafts , translocation of the positively charged AC domain across the cellular membrane is completed , driven by the negative gradient of membrane potential . Deciphering this fine-tuned mechanism of toxin action on cell membrane fosters our understanding of the key role played by CyaA in virulence of Bordetellae during the early phases of bacterial colonization of host respiratory mucosa . It allows to propose the following scenario . The produced CyaA targets the CD11b/CD18 receptor of incoming myeloid phagocytic cells , such as neutrophils , macrophages and dendritic cells [42] . As CyaA action does not depend on receptor-mediated endocytosis , the toxin recruited into lipid rafts can rapidly translocate its highly active AC enzyme domain across the cytoplasmic membrane of cells , in a process exhibiting a half-time of only about ∼30 seconds [38] . Mobilization of toxin-receptor complexes into lipid rafts than promotes their clustering and potentially induces recruitment of cellular cAMP-responding elements , such as the protein kinase A anchored to AKAPs , the specific A-kinase anchoring scaffolds [43]–[45] . This would allow maximization of toxin action through subversive cAMP production in close vicinity of components of the cAMP-regulated PKA signaling pathway . This capacity to hijack the spatio-temporal regulation of cellular cAMP/PKA signaling would then endow CyaA with the high potency in paralyzing the central bactericidal mechanisms employed by myeloid phagocytic cells . Indeed , few picomoles of CyaA ( 1 ng/ml or less ) were previously reported to instantaneously suppress the oxidative burst capacity of neutrophils [46] , or the phagocytosis of complement-opsonized particles by macrophages [2] . Several other bacterial protein toxins appear , indeed , to utilize lipid rafts as a portal of cell entry , exploiting as specific receptors directly certain raft components , such as cholesterol , sphingolipids or GPI-anchored proteins [47]–[50] . In contrast , we found here that CyaA associates with rafts only upon binding and mobilization ( hijacking ) of its receptor CD11b/CD18 . Unless activated in the process of leukocyte activation , this β2 integrin is distributed diffusely over the entire cellular membrane . As outlined above , we show here that upon binding of CyaA the integrin relocates into lipid rafts , due to toxin-induced and calcium-activated cleavage of talin by calpain . Moreover , the recently discovered capacity of CyaA to bind N-linked oligosaccharides of CD11b/CD18 [24] might also play a role in this process . It is , indeed , plausible to propose that CyaA interaction with terminal sialic acid residues of glycan chains of raft sphingolipids might also be contributing to accumulation of the CyaA-CD11b/CD18 complex in lipid rafts , as well as it may contribute to clustering of lipid rafts containing CyaA later-on . An evidence for CyaA interactions with gangliosides can , indeed , be deduced from the previously observed inhibition of CyaA activity on macrophages by the presence of micromolar concentrations of free gangliosides , such as GT1b [23] . It remains to be addressed in future studies if CyaA can form oligomeric pores also once engaged in interaction with the target cell membrane through binding of the CD11b/CD18 receptor and whether CyaA can form pore-forming oligomers also in phagocyte membrane . We have recently succeeded in demonstrating the presence of the long-predicted CyaA oligomers within the membrane of cells lacking the receptor CD11b/CD18 , such as erythrocytes [10] . Vojtova-Vodolanova with co-authors ( 2009 ) , indeed , showed that formation of CyaA oligomers underlies the pore-forming activity of CyaA towards erythrocytes . However , despite significantly higher amounts of CyaA binding per single phagocyte cell through the CD11b/CD18 receptor , fairly high concentrations ( >1 µg/ml ) of the recombinant enzymatically inactive but fully pore-forming CyaA-AC− variants are needed to provoke lysis of cells like J774A . 1 monocytes in several hours [8] . While this resistance to colloid-osmotic lysis is likely to be to large extent due to membrane recycling mechanisms and pore removal form phagocyte cytoplasmic membrane , it remains to be shown that CyaA can form oligomeric pores in leukocyte membrane as well . The results presented here do not indicate any role of CyaA oligomers in promoting calcium influx , toxin mobilization into rafts , or AC enzyme translocation into CD11b+ phagocytes . Early dose-dependence studies indicated that the AC domain was delivered across target cell membrane by CyaA monomers . Indeed , toxin molecules with the AC domain cleaved-off by cytosolic proteases , upon AC translocation into cells , were detected exclusively in form of CyaA monomers within erythrocyte membranes and were excluded from the detected CyaA oligomers [10] . Moreover , we used here the CyaA-E570Q+K860R-AC− protein , which essentially lacks any pore-forming activity and fails to permeabilize the membrane of J774A . 1 cells , thus being unlikely to form any CyaA oligomers ( Table 1 and [51] ) . On the other hand , this construct is fully capable to translocate the AC domain into cytosol of CD11b-expressing J774A . 1 cells , to promote calcium influx and to associate with DRM , or to co-localize with CtxB in coalesced rafts , respectively ( cf . Fig . 2 ) . It appears , therefore , unlikely that oligomerization plays a role in DRM association of CyaA . We also observed here that the levels of binding of CyaA to CD11b-expressing cells were not affected upon cholesterol depletion of cell membrane , while the translocation of the AC domain across the membrane depended strongly on the cholesterol content . This suggests that by modulating the physical properties lipid bilayers , cholesterol was specifically supporting the translocation of the AC domain across cell membrane . Indeed , cholesterol removal was previously found to impair the residual penetration capacity of CyaA on artificial membranes and erythrocytes [52] , [53] . This goes well with the impact of cholesterol concentration on membrane fluidity , lateral phase separation , formation of liquid-ordered structures and the propensity of lipids to adopt the inverted hexagonal phase [54] , [55] . The same membrane properties would also be expected to support AC domain translocation into cells by lowering the energy barrier for polypeptide penetration into and across the lipid bilayer [56] . It is plausible to speculate that membrane translocation of the AC domain requires the presence of cholesterol-dependent liquid-ordered ( lo ) phase , in which the acyl chains of lipids are tightly packed , while the individual lipid molecules have a high degree of lateral mobility . The relative mobility of lipids in lo domains represents , indeed , a likely prerequisite for passage of the AC domain across lipid bilayer . A high condensation and immobility of lipids in liquid-disordered ( ld ) -phase domains would , in turn , be expected to interfere with AC polypeptide translocation . The requirement for sufficient membrane fluidity for AC translocation to occur is also indicated by the block of AC translocation at 4°C [38] . Recently , we demonstrated that AC domain translocation across target cell membrane is accompanied by entry of Ca2+ ions into cells . Moreover , the AC domain polypeptide as such was found to participate in formation of the transiently opened calcium influx path in cell membrane [18] . Here , we used the 3D1 MAb recognizing a distal segment of the AC domain and show that blocking of AC domain translocation across cell membrane can lock CyaA in a ‘translocation intermediate’ conformation that forms a path for Ca2+ influx across cell membrane ( cf . Fig . 3 ) . Moreover , this ‘translocation intermediate’ was found to be recruited into lipid rafts ( Fig . 3 ) . The sum of the data hence allows us to answer the question what happens first , whether calcium influx precedes toxin mobilization into rafts , or whether recruitment of CyaA into rafts precedes calcium influx and AC translocation . We showed here that calpeptin-mediated inhibition of calcium-activated processing of talin by calpain yields ( i ) inhibition of CyaA recruitment into rafts and ( ii ) it inhibits AC translocation across membrane . Collectively , hence , these results strongly suggest that the transient influx of Ca2+ into cells accompanies the earliest step of membrane insertion of the toxin ‘translocation intermediate’ . This would precede and be essential for subsequent recruitment of CyaA into lipid rafts , whereupon AC translocation is accomplished . It remains , however , to be determined what is the threshold of the calcium signal required for initiation of talin cleavage and mobilization of CyaA into lipid rafts . Two major isoforms of calpain have , indeed , been so far identified in eukaryotic cells . The calpain I ( μ-calpain ) is activated at µM Ca2+ concentrations , while calpain II ( m-calpain ) only responds to mM concentrations of Ca2+ [57] . Here we observed that CyaA relocalization into DRM occurred at 1 nM toxin concentration , which is about two-times less than the lowest CyaA concentrations still allowing to elicit a [Ca2+]i increase detectable in cells by the Fura-2/AM probe [18] . Moreover , only influx of extracellular Ca2+ ions into cells , and not the elevation of cytosolic [Ca2+]i due to Ca2+ release from intracellular stores , enabled the accumulation of CyaA in DRM ( cf . Fig . 4 ) . This differs importantly from the mechanism reported for localization of the leukotoxin ( LtxA ) of Actinobacillus actinomycetemcomitans into rafts . LtxA binds yet another β2 integrin of human leukocytes , the LFA-1 or CD11a/CD18 heterodimer . Horeover , LtxA appears to first adsorb on cell membrane of T lymphocytes in a receptor-independent manner , to trigger , somehow the store-operated elevation of cytosolic [Ca2+]i , to induce talin cleavage , and upon relocation into rafts , the Ltx clusters with LFA-1 within rafts to promote cell lysis [58] . With CyaA , all the Ca2+ ions entering macrophage cytoplasm due to toxin action appear to come from extracellular medium [18] . It is generally accepted that there exists a gradient of about four orders of magnitude in Ca2+ concentrations between the external medium ( ∼2 mM ) and cell cytosol ( ∼100 nM ) . Therefore , numerous Ca2+-buffering proteins accumulate beneath the inner face of cell membrane , accounting for formation of local Ca2+ gradients and controlling signaling induced by alterations of Ca2+ concentrations in the submembrane compartment . These concentrations can , indeed , be still much higher , and rise more rapidly , than the bulk Ca2+ levels in cell cytosol [59] . Therefore , it is likely that even an importantly lower CyaA concentration than used here ( 1 nM = 176 ng/ml ) , may still be generating sufficiently high local Ca2+ signal beneath cell membrane in order to promote activation of μ-calpain at the inner face of cell membrane . It appears , thus , plausible to assume that mobilization of CyaA into rafts in phagocyte membrane , and translocation of the AC domain from rafts directly into the cytosolic compartment of phagocytes , are indeed taking place also during natural Bordetella infections in vivo . This would account for the remarkable efficacy of CyaA in disarming the sentinel cells of the host innate defense .
Intact recombinant CyaA and its mutant variants were expressed and purified as previously described [19] . Except of pro-CyaA , the CyaA proteins were produced in E . coli XL1-Blue in the presence of the co-expressed toxin-activating acyltransferase CyaC , as previously described [20] . Lipopolysaccharide was eliminated by repeated 60% isoporopanol washes of CyaA bound to the Phenyl Sepharose resin [60] . This reduced the final endotoxin content below 50 EU/mg of purified protein , as determined by the Limulus amebocyte lyzate assay ( QCL-1000 , Cambrex , NJ , USA ) . For fluorescence microscopy , the CyaA proteins were labeled while bound to Phenyl-Sepharose resin during the final purification step . Briefly , the CyaA eluates from a DEAE-Sepharose columns ( GE Healthcare ) in 50 mM Tris-HCl ( pH 8 ) , 8 M urea , 0 . 2 mM CaCl2 , 200 mM NaCl , were diluted 1∶4 with a buffer containing 50 mM Tris-HCl ( pH 8 ) , 1 M NaCl and 1 mg of CyaA was loaded on an 0 . 5 ml Phenyl-Sepharose column . The columns were extensively washed with 0 . 1 M sodium bicarbonate ( pH 9 ) , 1 M NaCl . Next 10 µg/ml Alexa Fluor 488 succinimidylester solution ( Molecular Probes ) was loaded and labeling proceeded at 25°C for 1 hour . The columns were washed with 50 mM Tris-HCl ( pH 8 ) , 1 M NaCl , and the CyaA-Alexa Fluor 488 conjugates were eluted in a buffer containing 50 mM Tris-HCl ( pH 8 ) , 8 M urea and 2 mM EDTA . Unreacted dye was separated from labeled CyaA on Sephadex G-25 columns ( GE Healthcare ) . Efficiency of protein labeling was assessed spectrophotometrically and a molar ratio of about 1∶4 ( protein∶dye ) was found for all CyaA preparations . It was verified that this extent of labeling did not affect the biological activities of CyaA . See Protocol S1 for full description . Detergent-resistant membranes ( DRM ) were separated by flotation in discontinuous sucrose density gradients . Briefly , J774A . 1 cells ( 2 . 107 ) were washed with prewarmed DMEM and incubated with 1 nM CyaA proteins at 37°C for 10 min . Cells were washed with ice-cold phosphate-buffered saline ( PBS ) , scraped from the Petri dish and extracted at 4°C for 60 min using 200 µl of TBS buffer ( 20 mM Tris-HCl , pH 7 . 5 , 150 mM NaCl ) containing 1% Triton X-100 , 1 mM EDTA , 10 mM NaF and a Complete Mini proteinase inhibitor cocktail ( Roche , Basel , Switzerland ) . The lyzates were clarified by centrifugation at 250×g for 5 min and the post-nuclear supernatants were mixed with equal volumes of 90% sucrose in TBS . The suspensions were placed at the bottom of centrifuge tubes and overlaid with 2 . 5 ml of 30% sucrose and 1 . 5 ml of 5% sucrose in TBS . Membrane flotation according buoyant density was achieved by centrifugation at 150 , 000×g in a Beckman SW60Ti rotor for 16 h at 4°C . Fractions of 0 . 5 ml were removed from the top of the gradient . Calcium influx into J774A . 1 and U937 cells was measured as previously described [18] . Briefly , cells were loaded with 3 µM Fura-2/AM ( Molecular Probes ) at 25°C for 30 min and the time course of calcium entry into cells induced by addition 3 µg/ml of CyaA proteins was determined as ratio of fluorescence intensities ( excitation at 340/380 nm , emmision 505 nm ) , using a FluoroMax-3 spectrofluorometer equipped with DataMax software ( Jobin Yvon Horriba , France ) . J774A . 1 cells were incubated in DMEM supplemented with 10 mM methyl-β-cyclodextrin ( MβCD ) at 37°C for 30 min . Cholesterol-depleted U937 cells were obtained upon growth in RPMI medium supplemented with 10% of delipidated serum ( lipoprotein-deficient serum from fetal calf , Sigma ) for 48 h . Cholesterol content was determined using an Amplex Red Cholesterol Assay Kit ( Molecular Probes , Invitrogen ) according to manufacturer's instructions . Viability of cells was tested by trypan blue staining and no significant cell death occurred upon cholesterol extraction . U937 cells were grown in media with 10% standard , or delipidated serum , and were mounted on polylysin-coated coverslips prior to incubation with labeled proteins . J774 . A1 cells ( 5 . 104 ) were grown directly on coverslips ( ∅ φ12 mm ) and incubated with Alexa Fluor 488-labeled CyaA proteins ( 6 nM ) at 37°C for 10 min , before cells were washed and 5 µg/ml of Alexa Fluor 594-labeled cholera toxin subunit B ( CtxB ) was added for additional 5 min . The unbound proteins were washed-off with ice-cold PBS , cells were fixed with 4% paraformaldehyde in PBS at 25°C for 20 min , and mounted in Mowiol solution ( Sigma ) . Fluorescence images were taken using a CellR Imaging Station ( Olympus , Hamburg , Germany ) based on Olympus IX 81 fluorescence microscope , using a 100× oil immersion objective ( N . A . 1 . 3 ) . Digital images were processed using ImageJ software . J774 . A1 cells ( 106 ) were incubated with 17 nM CyaA in DMEM for 30 min at 37°C , washed with Hank's Buffered Salt Solution buffer ( HBSS ) , and lyzed at 4°C during 30 min in 500 µl of Tris-buffered saline ( pH 7 . 4 ) supplemented with 1% Triton X-100 and EDTA-free Complete Mini proteinase inhibitor cocktail ( Roche , Basel , Switzerland ) . The lyzate was centrifuged for 15 min at 10 , 000×g at 4°C , and the supernatant was incubated with MEM-174 MAb covalently linked to CNBr-activated Sepharose beads ( GE Healthcare ) at 4°C for 1 h . The beads were washed five times with 1 ml of the lysis buffer and the bound proteins were eluted with SDS-PAGE loading buffer and analyzed by SDS-PAGE followed by Western blotting . J774A . 1 or U937 cells ( 106 ) were incubated with 6 nM CyaA proteins at 37°C for 10 min , washed repeatedly in buffer , and the amount of cell-associated adenylate cyclase ( AC ) activity was determined in cell lyzates as previously described by [61] . J774A . 1 or U937 cells ( 106 ) were incubated with CyaA proteins at indicated concentrations for 10 min at 37°C and intracelular cAMP concentrations were determined in cell lyzates using a competitive ELISA as previously described [62] .
|
The adenylate cyclase toxin ( CyaA ) of pathogenic Bordetellae eliminates the first line of host innate immune defense . It penetrates myeloid phagocytes , such as neutrophils , macrophage or dendritic cells , and subverts their signaling by catalyzing an extremely rapid conversion of intracellular ATP to the key signaling molecule cAMP . This efficiently inhibits the oxidative burst and complement-mediated opsonophagocytic killing of bacteria , thus enabling the pathogen to colonize host airways . We show that translocation of CyaA into phagocyte cytosol occurs in two steps . The toxin first binds the integrin CD11b/CD18 and inserts into phagocyte membrane to mediate influx of calcium ions into cells . This promotes relocation of the toxin-receptor complex into specific lipid microdomains within cell membrane called rafts . The increased concentrations of cholesterol within rafts and their particular lipid organization then support translocation of the adenylate cyclase enzyme directly into the cytoplasmic compartment of cells . The mechanism of CyaA penetration into cells sets a new paradigm for membrane translocation of toxins of the RTX family .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"biochemistry/membrane",
"proteins",
"and",
"energy",
"transduction",
"microbiology/innate",
"immunity"
] |
2010
|
Bordetella Adenylate Cyclase Toxin Mobilizes Its β2 Integrin Receptor into Lipid Rafts to Accomplish Translocation across Target Cell Membrane in Two Steps
|
The ability of cells to accurately control gene expression levels in response to extracellular cues is limited by the inherently stochastic nature of transcriptional regulation . A change in transcription factor ( TF ) activity results in changes in the expression of its targets , but the way in which cell-to-cell variability in expression ( noise ) changes as a function of TF activity , and whether targets of the same TF behave similarly , is not known . Here , we measure expression and noise as a function of TF activity for 16 native targets of the transcription factor Zap1 that are regulated by it through diverse mechanisms . For most activated and repressed Zap1 targets , noise decreases as expression increases . Kinetic modeling suggests that this is due to two distinct Zap1-mediated mechanisms that both change the frequency of transcriptional bursts . Notably , we found that another mechanism of repression by Zap1 , which is encoded in the promoter DNA , likely decreases the size of transcriptional bursts , producing a unique transcriptional state characterized by low expression and low noise . In addition , we find that further reduction in noise is achieved when a single TF both activates and represses a single target gene . Our results suggest a global principle whereby at low TF concentrations , the dominant source of differences in expression between promoters stems from differences in burst frequency , whereas at high TF concentrations differences in burst size dominate . Taken together , we show that the precise amount by which noise changes with expression is specific to the regulatory mechanism of transcription and translation that acts at each gene .
The cellular response to environmental changes is mediated through activation of TFs and subsequent coordinated activation and repression of dozens of target genes . However , gene expression is noisy [1] , and this limits the precision with which cells can regulate protein levels . Genome-wide , noise ( σ2/μ2 , variance/mean2 ) decreases as expression increases [2]–[4] . Along this global trend , individual genes with the same average expression in the population differ in their amount of noise . The level of noise for each gene is related to its function and is determined by the mechanisms of regulation [5] . However , the precise mechanisms by which control of noise is accomplished for native genes are not known . Two quantities that describe the dynamics of gene expression , and have been related to the distribution of protein abundances , are burst size and burst frequency . Burst frequency is determined by the rate at which the promoter switches from an inactive to an active transcriptional state due to transcription factor ( TF ) binding and subsequent PolII recruitment ( promoter on-switching ) . Burst size is the number of proteins produced during each promoter on-event [6]–[8] . Native genes differ in the relative contribution of burst frequency and size to expression [4] , [9] , [10] , suggesting that evolution can tune both parameters in order to reach an optimal level of expression and noise for each gene [11] . When an increase in gene expression is caused by an increase in the rate of promoter on-switching ( burst frequency ) , noise ( σ2/μ2 ) decreases monotonically with expression [12] . In contrast , an increase in burst size ( due to a decrease in promoter off-switching rate or an increase in the transcription or translation rate ) results in an increase in expression and in noise strength ( σ2/μ ) , and no change in noise [12] . Mutations in the TATA box in yeast [13] and the ribosome binding site in B . subtilis [14] both affect noise strength , but not noise . The former is thought to be involved in transcription re-initiation [15] , thus extending the time of each active state of the promoter , while the latter affects the number of proteins produced from each mRNA molecule . These observations strengthen the claim that changes in mean expression but not noise stem from molecular mechanisms that affect the number of proteins produced during each transcriptional event , but not the frequency of such events . Taken together , these data support a model of gene expression in which changes in promoter dynamics , such as changes in on-switching rates and transcription and translation rates , can be deduced by measuring how noise changes with expression [4] , [7] , [8] . Since most genes are regulated through multiple mechanisms , each of which can affect burst size and burst frequently differently , different genes should exhibit different relationships between mean expression and noise . However , measurements of a set of seven different promoters in E . coli all showed similar changes in expression and noise throughout induction [16] . Gene regulation in eukaryotes is more complex , and we hypothesized that burst frequency and burst size would be differentially regulated for each gene and , as a consequence , that the relationship between noise and expression would be different for different genes . To characterize the relationship between mean expression and noise for native promoters in response to environmentally stimulated changes in TF activity , we generated a set of 16 strains in which distinct promoters are fused upstream of a yellow fluorescent protein reporter ( YFP ) . In each strain , we extracted a different Zap1 binding-site containing promoter from its native locus , integrated it into the his3 locus , and measured its expression and noise at 12 different zinc concentrations ( induction levels ) . Decreasing zinc concentration increases the activity and expression of Zap1 and changes the expression of Zap1 target promoters [17] . The resulting Zap1 dose-response curves of these targets show activation , repression , and a combination of activation and repression , consistent with previous observations [17] . We found that for Zap1-activated targets , an increase in Zap1 causes an increase in expression and a decrease in noise . Similarly , Zap1-repressed targets exhibit the same relationship between expression and noise , whereby an increase in Zap1 causes a decrease in expression and an increase in noise . Despite this general trend that has previously been reported [2]–[4] , we found that the slope of noise versus expression is unique for each promoter , showing that noise is not determined by expression level alone . The most notable exception to expression determined noise is the ZRT2 promoter , which is both activated and repressed by Zap1 [18] , in which we found a different and novel relationship between mean expression level and the distribution of expression . Repression of ZRT2 by Zap1 results in a decrease in both expression and noise , leading to a transcriptional state of low expression and low noise that is unique among the 16 tested promoters . This behavior is predicted by a kinetic model in which repression is due to a secondary binding event near the TATA that causes a decrease in transcription rate ( burst size ) , thereby preventing the typical increase in noise that accompanies repression due to a reduction in burst frequency . These results suggest that the relationship between noise and expression is unique to each promoter and is determined by the regulatory mechanism encoded in the promoter DNA sequence and not by mean expression level alone . We hypothesized that further noise reduction will occur when activation and repression are performed by the same TF . Using a model of noise that takes into account the sensitivity to TF level fluctuations and an experiment in which we decouple activator from repressor , we find strong evidence supporting our hypothesis that coupling between activator and repressor is a mechanism for noise reduction . Finally , analysis of the data from all measured Zap1 targets brings forward a global principle of regulation in which the major source of differences in expression between promoters changes with induction . Our results strongly support a model in which at low Zap1 activity , differences in expression between Zap1 targets are due to variability in the frequency of transcriptional bursts , while at high Zap1 activity , differences are due to variability in the number of proteins produced during each transcriptional burst . This model suggests that such behavior is a general property of transcriptional regulation .
To study how expression of different native promoters is regulated by environmental-induced changes in TF activity , we measured promoter-driven expression in single cells for 16 targets of the TF Zap1 in response to changes in extracellular zinc . To do this we used an experimental system that we previously developed in which a promoter of interest drives YFP expression from the genomic his3 locus ( Figure 1A ) [19] . We generated a set of 16 promoter-YFP fusion strains and used flow-cytometry to perform quantitative single-cell measurements of promoter-driven expression at 12 induction points ( Figure 1C ) . These promoters ( Figure 1B , Table S1 ) have diverse activation curves ( Figure 1D , Figure S1 ) and , while the response of each promoter correlates with the predicted Zap1 occupancy along the promoter ( Figure S2 ) , the diversity of responses suggests that the way in which Zap1 alters expression is different for different promoters . In addition , we examined the changes in noise and noise strength along the induction curves ( Figure 1E , G ) . For most activated ( 11/13 ) and repressed ( 2/3 ) promoters , noise decreases as expression increases ( Figure 1E , average Pearson correlation for all promoter of −0 . 73 , Figure S3A ) , consistent with observed genome-wide trends [2]–[4] . In contrast , noise strength changes less consistently across Zap1 targets ( Figure 1G , average Pearson of −0 . 09 , Figure S3B ) . Surprisingly , not only do different promoters exhibit different amounts of noise at the same level of expression ( Figure 1C ) , but also the way in which noise and noise strength change with expression is unique to each promoter ( Figure 1E , G , Figure S4 ) . Interestingly , a single promoter ( ZRT2 ) that is both activated and repressed by Zap1 ( Figure 1D , lower right ) [18] shows very different amounts of noise at the same mean expression ( Figure 1F ) . Because different molecular mechanisms of gene regulation can lead to the same change in mean expression but different changes in noise [20] , these results suggest that the precise molecular mechanism by which a change in Zap1 activity causes a change in expression may be different at each promoter . To better understand what determines the relationship between expression and noise we used an analytical model of gene regulation ( Figure 2A ) [21] to predict changes in expression and noise in response to changes in TF activity ( see Materials and Methods ) . We fit this model to measurements of ZRT1 expression and noise and find that the model replicates our experimental results when an increase Zap1 activity causes an increase in the promoter on-switching rate ( Kon ) ( Figure 2B , C ) . To further challenge the model we created a set of seven start codon context mutants of the ZRT1 promoter ( NNNNATG ) and measured the expression distribution of these variants at 12 different levels of TF activity ( Figure 2D–F ) ( only three mutants are shown for clarity ) . These mutations change translational efficiency and therefore the number of proteins produced per mRNA ( b ) , without affecting promoter dynamics ( Figure S5 ) . We find that ATG context variants at a single induction point differ in expression but not in noise , consistent with similar experiments in B . subtilis [14] . In support of the above hypothesis , we obtain the best fit of the model to our data when TF induction is modeled as changing Kon while ATG context variants change b ( Figures S6 and S7 ) . Furthermore , when fitting our model to data , we find that the optimal rate constants are on the order of experimentally measured promoter switching rates [9] , [22] , [23] , and not TF binding/unbinding rates [24] . This suggests that promoter switching rates probably correlate with , and are partially determined by , TF concentration and binding kinetics . However , each TF binding event does not necessarily lead to transcription initiation . These results suggest that increases in TF activity increase the frequency of transcriptional bursts , while increases in translational efficiency cause an increase in the size ( number of proteins produced ) of each burst . We note that this is in contrast to observations in E coli [16] and in yeast at the GAL1 promoter [25] , in which TF induction appears to change the promoter off-switching rate ( Koff ) but consistent with measurements of the PHO5 promoter [20] . In addition to increasing expression of target genes , Zap1 can also act as a repressor . Zap1 represses two targets ( ADH1 and ADH3 ) by binding upstream of the core promoter and inducing intergenic transcription through the core promoter , probably promoting dissociation of the activating TF Rap1 ( Figure 3A , B ) [26] . Two mechanisms have been proposed for repression by transcriptional interference: dislodgement of TFs and the Pol II pre-initiation complex by RNA Polymerase [27] , and competitive binding , one form of which is deposition of nucleosomes in the otherwise nucleosome-free region where the activating TFs and Pol II bind [28] . We hypothesized that deposition of nucleosomes would result in occlusion of the activating binding site , the TATA box , and Pol II binding , thus reducing the effective TF concentration and lowering the frequency of transcriptional activation . We model this mechanism as a reduction in Kon . Alternatively , passage of RNA polymerase may dislodge already bound Rap1 , TBP , and/or the RNA polymerase pre-initiation Complex . This would shift the promoter from the “on” into the “off” state , thus reducing the length of each transcriptional on state and therefore the number of mRNA molecules produced during each transcriptional burst ( Figure 3C ) . We model this mechanism as an increase in Koff . To determine the ability of dislodgement by Pol II ( TD ) or occlusion of TF binding by nucleosomes ( NO ) to explain our experiments , we fit each model to the data . We find that the NO model fits our data better than the TD model ( Figure 3D ) ( see Materials and Methods ) . Furthermore , the NO model consistently fits the data better in the case in which we vary each parameter by up to 2-fold . The increased robustness ( Figure 3E ) and decreased sensitivity ( Figure S8 ) of the NO model gives us further reason [29] to favor a model in which repression by Zap1 at the ADH1 and ADH3 promoters occurs by inducing intergenic transcription and nucleosome deposition over the core promoter and/or Rap1 binding site . Uniquely among Zap1 target promoters , ZRT2 responds nonmonotonically to an increase in Zap1 activity , whereby its expression first increases then decreases in response to increasing Zap1 activity [18] . In the activating regime of ZRT2 , noise decreases as expression increases , suggesting a Kon ( burst frequency ) dominated change that is similar to the purely activated targets . However , in contrast to the repressed targets ADH1 and ADH3 , where noise increases with the decrease in expression , in the regime where ZRT2 expression decreases noise remains constant . These results suggest that the decrease in ZRT2 expression is a result of a decrease in burst size ( see below ) , with the consequence of having induction points that have the same mean expression level but different expression distributions ( Figure 1F ) . At high induction , the distribution is less noisy ( Figure 1F , blue ) than at low induction ( Figure 1F , red ) . Thus , the ZRT2 promoter reaches a state that is unique amongst Zap1 targets that is characterized by both low expression and low noise . Taken together , these findings suggest that although ADH1 , ADH3 , and ZRT2 are all repressed by Zap1 , the mechanism by which ZRT2 is repressed is unique . In response to increasing Zap1 , ZRT2 expression first increases and then decreases . The activation by Zap1 is a result of Zap1 binding at activating binding sites 250–300 bp upstream of the start codon , while the repression is due to the presence of repressive Zap1 binding sites near the TATA box ( between −90 and −112 ) [18] . We made a variant of the ZRT2 promoter ( ZRT2-zre ) that lacks the repressive binding sites ( Figure 4B ) . We hypothesized that a model of the ZRT2 promoter should include promoter states in which Zap1 is bound as an activator , as a repressor , and both as activator and repressor ( Figure 4A ) . Based on experimental evidence [18] , we model the binding site affinity for the repressive site as weaker than that of the activating site . We assume that binding of Zap1 to the repressive site turns off the promoter but does not affect the transition probabilities between states . When the model was simultaneously fit to both the ZRT2-WT and ZRT2-zre experimental data , we find that the model obtains a good fit to data when , like with ZRT1 , an increase in Zap1 activity increases Kon and does not affect any other parameters . Interestingly , we find that the repressed state ( state 4 , Figure 4A ) is not fully off , but has a small , but not insignificant , transcription rate relative to the transcription rate of the active state ( state 2 , Figure 4A ) . Notably the only parameter change required to change from ZRT2-WT to ZRT2-zre is setting Koffrep to be very high , mimicking the mutation of the repressive binding sites ( Figure 4C , D ) . These experimental and modeling results suggest that binding of the transcriptional activator Zap1 to a binding site between the TATA box and TSS is necessary to generate a promoter state with low transcriptional activity . Notably , a very simple promoter model is able to replicate a nonmonotonic response to changes in TF activity . Furthermore , it suggests that ZRT2 is able to reach a state of low expression and low noise purely through transcriptional regulation due to a promoter state with high burst frequency ( due to binding of activating Zap1 ) and low burst size ( due to binding of repressive Zap1 ) . These results suggest that in the ZRT2 promoter , an increase in Zap1 both increases the frequency and decreases the size of transcriptional bursts . Therefore , our simple kinetic model shows that adding a repressive binding site for the activating TF is sufficient for explaining both ZRT2 expression and noise as a function of induction . Repression of ZRT2 is accompanied by a decrease in noise strength , suggesting that repression occurs via a decrease in burst size . We therefore hypothesized that addition of a repressive Zap1 binding site to a native Zap1 target that lacks repression would cause a decrease in expression and burst size . To test this hypothesis , we added a consensus Zap1 binding site ( ACCTTAAGGT ) upstream of the transcription start site of ZRT1 ( Figure 4E , ZRT1pr+ZRE ) . Consistent with our hypothesis that this repressive site reduces expression through a decrease in burst size , this additional site results in a constant ∼2-fold decrease in expression , a decrease in noise strength , and no change in noise ( Figure 4F ) . A model identical to the ZRT2 model ( Figure 4A ) , except that the repressive site has a higher affinity to Zap1 than the activating site , replicates the experimental data ( Figure 4F ) . Interestingly , we find that while both models require the repressed state to be partially active , the repressed state of the ZRT1 promoter has higher activity ( in model and data ) than for the ZRT2 promoter . This may be because ZRT2 has at least two repressive Zap1 binding sites , while we only introduced a single repressive binding site into ZRT1 . Nevertheless , these results show that the presence of a Zap1 binding site between the TATA box and transcription start site is both necessary and sufficient for repression mediated by a decrease in burst size . A computational search for Zap1 binding sites between the TATA box and the transcription start site identified three weak Zap1 binding sites in the ZRT3 promoter ( Figure 5A ) . A closer look at the ZRT3 induction curve at very low zinc concentrations showed that expression of ZRT3 decreases slightly at high Zap1 induction ( Figure 5B , inset ) . To determine if these weak Zap1 sites were functional , we mutated them and measured expression of the wild-type and mutant ZRT3 promoters . Consistent with our hypothesis that Zap1 binding sites around the TSS are repressive , removal of the presumptive Zap1 binding sites increased expression ( Figure 5B ) , in particular at higher induction , consistent with our model in which repression is a function of repressor activity . This suggests that low-affinity binding sites may be functional at high TF concentration , perhaps mostly at promoters that have additional high-affinity binding sites . The ZRT2 promoter presents a case in which the activator and repressor are the same TF . We were intrigued by this mechanism and wondered whether this affects the noise properties . Many promoters in yeast are regulated by the binding of both activators and repressors to different binding sites in the promoter [30] . The activator and repressor can be different proteins ( e . g . , ADH1 is activated by Gcr1 and Rap1 and repressed by Zap1 ) or the same protein ( such as ZRT2 that is both activated and repressed by Zap1 ) ( Figure 6A ) . We hypothesized that the sensitivity to TF fluctuations for a promoter that is both activated and repressed depends on the coupling between activator and repressor . For example , we expect that when activation and repression are done by the same TF , in a regime where a change in activator binding has the exact opposite result on expression as the same change in repressor binding , the promoter is insensitive to any fluctuations in TF levels . This is because any random fluctuation in the concentration or activity of the TF will have an equal activating and repressive effect and therefore result in no net change in target activity . To study this hypothesized phenomenon , we used our kinetic model of ZRT2 and simulated the case where activator and repressor are different ( decoupled ) and where they are the same TF ( coupled ) . We then calculated the contribution of TF fluctuations to expression noise throughout the induction ( Figure 6B ) ( see Materials and Methods for a detailed description of the model ) . Coupling of the activator and repressor reduces the sensitivity to TF fluctuations throughout induction and places the point of minimal sensitivity to TF fluctuations at the point of maximum target gene expression ( Figure 6B , blue line ) . Our model predicts that the total sensitivity to TF fluctuations is reduced throughout the induction curve , and that this reduction is greatest at maximal promoter expression ( Figure 6B , point 1 ) . To test this we measured extrinsic noise ( the contribution of variance in all factors; e . g . , ribosomes , Zap1 , Pol II ) for the native ZRT2 promoter using a dual-reporter . We find that extrinsic noise is constant across the induction ( Figure 6D , purple , see also Text S1 and Figure S11 ) . However , when we remove as much global extrinsic noise [4] as possible using a very narrow forward and side scatter gate ( Figure S9 ) [25] we hypothesize that we are left with mostly pathway-specific noise ( e . g . , noise due to TF level fluctuations ) . In support of this hypothesis , we find that pathway-specific noise is not constant , but rather varies greatly ( around 10 fold ) with induction . We find that this signal , which we expect to be dominated by changes in TF sensitivity , does indeed drop around the point of maximal expression ( Figure 6D , blue ) , consistent with our model . In fact , the extrinsic noise replicates quite well the general predicted change in TF sensitivity with induction . Finally , our model predicts that decoupling of activator and repressor will increase total noise as the sensitivity to TF fluctuations is increased . To test this we replaced the two activating Zap1 binding sites of ZRT2 with two Gal4 binding sites ( Figure 6A ) and measured expression and noise throughout the repressive regime at high Gal4 induction ( 0 . 5% galactose ) ( Figure S10 ) . Consistent with our model , the Gal4-Zap1 regulated ZRT2 variant has higher noise than the wild-type promoter ( Figure 7C ) . These results show that , while repression is able to reduce expression and keep noise constant ( Figure 4 ) , a transcriptional regulatory motif , in which the activator and repressor are the same protein , is capable of reducing noise even further . This suggests that the coupling of activator and repressor can be a mechanism to regulate gene expression with less variability . We hypothesized that the source of differences in expression between genes might change with TF concentration . At low TF concentrations , promoters will be inactive most of the time , and differences in expression may depend mostly on differential recruitment of the TF . In this case , the major source of differences in expression between promoters should stem from the frequency with which transcriptional bursts occur . Alternatively , at saturating concentrations of activating TF , the promoter should be “on” most of the time and the major difference in expression between promoters should arise from the transcription and translation rates of each promoter . Thus , as the concentration of TF changes from negligible to saturating , we expect the transcription and translation rates of each promoter to become more important in determining expression differences between genes . To determine whether burst frequency or burst size dominate the differences in expression between promoters , for each induction level , we measured the correlation between expression and noise or noise strength across promoters . Consistent with the above hypothesis , across all promoters , noise is highly correlated with expression at low levels of Zap1 activity ( R = −0 . 66 , p<0 . 01 ) , while noise strength is uncorrelated ( R = −0 . 02 , p<0 . 94 ) ( Figure 7A ) . This suggests that at low TF concentration , burst frequency determines the differences in expression across promoters . Conversely , at high levels of Zap1 activity , noise strength is correlated with expression ( R = 0 . 63 , p = 0 . 01 ) , and noise is slightly less correlated ( R = −0 . 55 , p = 0 . 04 ) ( Figure 7A ) . Overall , we found a continual increase in the correlation between noise strength and expression with increasing TF activity ( unpublished data ) . To test the hypothesis that these differences are due to a change in the dominant source of expression difference between promoters , we generated 50 random genes in-silico that differ only in their rates of promoter on-switching ( KON ) and translation ( KTL ) . We then performed an induction by increasing KON for each promoter to 20 times its original value . This results in a mean to noise and mean to noise strength scaling that is strikingly similar to what we observed for the native Zap1 targets ( Figure 7B ) . Taken together , our results suggest that as a set of targets of the same TF are induced , the major source of expression differences between them changes from being dominated by burst frequency to a combination of burst frequency and burst size .
Similar to the global trend [3] , our data suggest that changes in expression of individual promoters are dominated by differences in burst frequency . This is consistent with Zap1 binding to promoters being limiting for transcriptional activation , especially at low Zap1 concentrations , and with the proposal that the rate-limiting step in transcription for yeast is promoter firing rate , which is determined by TF search times [31] . However , the observation that different activated targets have different scaling between noise and expression suggests that while activation by Zap1 acts only through burst frequency at most activated promoters , it may act partially or even completely through burst size at other activated promoters . This is entirely reasonable; Zap1 is not the only TF acting at these promoters , and the promoters differ in both nucleosome organization and the presence and location of TATA boxes . Experiments that placed a tetO sequence at different locations within the FLO11 promoter suggest that the same TF can have different effects on promoter dynamics , depending on the location of binding sites within the promoter [23] . Unfortunately , there are not enough strongly induced Zap1 targets in S . cerevisiae to identify the promoter architecture features that determine the source of the promoter-specific slope . It will be interesting to perform dose-response curves for a larger set of promoters from other yeasts , or on synthetic promoters , in order to identify promoter architectures that determine the promoter-specific slope . Our observation that repression by production of an upstream interfering transcript causes an increase in noise , while repression when the TF binds near the TATA box causes a decrease in noise , suggests that different dynamics occur at each promoter during repression . This , along with previous observations [18] , [23] , [25] , [28] , [32] , suggests that the mechanism of regulation by any TF is determined in cis by the promoter architecture . Binding sites between the TATA box and TSS decrease burst size , binding sites within a few hundred bases upstream of the TATA box increase burst frequency , and binding sites further upstream , with a nearby downstream TATA box , repress through a reduction in burst frequency . These data show , to our knowledge , for the first time that different promoter architectures can cause a similar change in expression in response to changes in TF activity , but exhibit different changes in noise . If the genome-wide scaling of expression and noise extends to proteins with very low expression , then a large fraction of cells will have zero molecules of protein [33] . Single-molecule studies have confirmed this: many cells have zero molecules of proteins with low levels of expression [4] . However , many proteins expressed at low levels are essential . This raises the question: How does the cell maintain a low level of both expression and noise for essential proteins , so that all cells have the minimum number of proteins ? Our results showing that burst size regulation can reduce expression without increasing noise suggest a way out of this trap . Lowly expressed genes tend to be bound by many transcriptional regulators , both activators and repressors [30] . Low levels of an activating TF result in low expression and high noise . Notably , a motif in which weak transcription but efficient translation generates high noise may exist at the comK gene in B . subtilis [34] . In contrast , combinatorial regulation that results in high burst frequency and low burst size ( approaching the Poisson limit [9] ) provides a regulatory motif through which cells can produce low levels of protein with low cell-to-cell variability . Our identification of this same regulatory motif in the ZRT3 promoter suggests that this motif may be common . This regulatory strategy may be used to prevent some cells from having zero molecules of protein when expression is low . The concentrations of TFs , like those of all other proteins , vary greatly from cell to cell . We expect that these variations have a significant impact on the cell-to-cell variability of target gene expression [4] , and therefore wondered how cells deal with this source of noise . Interestingly , we find that ZRT2 is able to reduce noise through its reduced sensitivity to fluctuations in TF levels , as a result of activator and repressor being the same molecule . Mechanisms for extrinsic noise reduction have been previously reported [21] . However , to the best of our knowledge , we are the first to propose theoretically and confirm experimentally a mechanism for desensitizing promoters to TF noise . Noise as a result of TF fluctuations has been proposed theoretically in several studies [4] , [35] . In fact , Bai et al . propose a dual-reporter experiment to investigate extrinsic noise resulting from TF fluctuations , which we have performed in this work ( Figure 6D ) . We note that noise from TF fluctuations is a special case of noise propagation in a gene network , where the noise of a downstream gene is a function of its intrinsic noise and the noise from any upstream genes [36] . An alternative mechanism for a similar reduction in sensitivity would be the regulation by multiple different decoupled TFs . We hypothesize that as the number of different TFs increases , target sensitivity ( and therefore noise ) decreases , if the TFs are sufficiently de-correlated . This potential mechanism , as well as the general characterization of the effect of TF noise on target noise , would make the subject of a meaningful follow-up study . The observed change in the scaling between noise and expression throughout the increase in TF concentration suggests that variability between promoters in burst size ( transcription efficiency , translation efficiency , and promoter off-switching rate ) becomes more important as TF concentration is increased . This suggests that differences in promoter architecture play different roles at low and high TF concentrations . In the presence of limiting TF , promoter architecture may determine expression by determining TF search time , through the number of accessible TF binding sites . However , at high TF concentration , promoters are mostly bound by TFs , and the transcription and translation efficiency of each gene may play a greater role in determining expression . This idea is supported by the positive correlation between noise strength and expression at high TF concentration , as would be expected from theory [20] . In addition , differences in burst frequency cannot account for the measured single-cell expression distributions at high TF concentration . These data suggest that the dominant sources of gene-to-gene variability in expression change with TF concentration: at low TF concentration burst frequency ( the ability of the promoter to recruit TF ) differences dominate , whereas at high TF concentration burst size ( transcriptional and translational efficiency ) differences dominate . Overall , our results show that the relationship between expression and noise is highly dependent on the promoter architecture . One implication of this finding is that using only a single TF , evolution can implement diverse expression profiles with unique noise properties . The fact that repression of ZRT2 by Zap1 is evolutionarily conserved suggests that there is an advantage to this ability .
Construction of promoter-YFP strains was performed as described previously [19] . In brief , a master strain , his3::TEF2pr-mCherry-YFP-NatMX4 , was created in the background strain Y8205 [37] by homologous recombination . Each promoter-YFP strain was created by integration of a PCR product containing the native promoter along with URA3 as a selection marker . Integration by homologous recombination upstream of YFP was confirmed by DNA sequencing and by identical expression and growth of multiple independent transformants . To introduce , alter , and remove elements within ∼150 bp of the ATG , we developed a method in which an existing URA3-promoter-YFP cassette is amplified over multiple rounds of PCR . In each subsequent round , a new primer is used that further extends the product towards the YFP and optionally introduces designed mutations . Thus multiple site-directed mutations can be tiled onto the 3′ end of the promoter . All promoter variants were confirmed by DNA sequencing . Yeast strains were grown overnight to saturation in YPD , resuspended in low zinc medium [18] , and grown overnight to saturation in media lacking zinc . Cultures were then diluted 1∶40 in water and 6 µl of this dilution was inoculated in 130 µl of low zinc media supplemented with various concentrations of zinc . Cells were grown in round-bottom 96-well plates shaking at 30°C a minimum of 12 h , to approximately 5*106 cells/ml prior to expression measurements . For galactose inductions , cells were pregrown overnight to saturation in low zinc media with 0 . 5% galactose as the sole carbon source to induce expression , then resuspended in media with varying concentrations of zinc with 0 . 5% galactose similar to the above experiments . Flow cytometry was performed on a BD LSRII . YFP and mCherry were excited using 488 nm and 561 nm lasers and emitted light was collected with 525/50 nm and 610/20 nm band-pass filters , respectively . There is no detectable spillover of YPF or mCherry into the other channel using these filters and lasers . Expression and noise measurements were collected and calculated using the ratio of YFP over mCherry for each cell . To obtain expression and noise measurements from each well , a relatively homogenous subpopulation of mostly G1 cells was chosen by gating on forward and side scattering . Wells containing fewer than 500 cells after gating or with obvious contamination were excluded from further analysis . Noise was quantified as the variance over the mean squared and noise strength as the variance over the mean . We model stochastic promoter state switching , transcription , and translation using the master equation following the approach described by Sanchez et al . [38] , which in turn is an adaptation of previous derivations of the master equation for gene regulation [38]–[40] . In this description of promoter regulation , TF binding and unbinding events determine the transitions between promoter states . A change in transcriptional activity occurs when a transition is made to a state with differing transcription rate . Each promoter state is modeled to have a low ( including zero ) or relatively high transcription rate to describe in-active ( “off” ) or active ( “on” ) states , respectively . Translation occurs in bursts with the probability of a burst described by a geometric distribution . The master equation ( in matrix notation ) takes the form: ( 1 ) where is the vector of probabilities of having n proteins in the cell for each promoter state . describes the time evolution of these probabilities . is the matrix of promoter state transition rates , where is the rate of transitioning from state j to state i and is −1 times the sum over all outgoing rates from i . is the diagonal matrix of transcription rates with on the diagonal ( ) , where is the transcription rate of state i . is the identity matrix . b is the average burst size ( proteins produced per mRNA ) . δ is the protein degradation rate . h ( β ) describes a geometric distribution and is the probability of producing a burst of size β . To derive the mean protein abundance and variance , we solve this system at steady state , thus for , we get mean protein abundance: ( 2 ) where is the zeroth partial moment of the distribution of mRNA abundance and is the solution to: ( 3 ) We can get noise ( σ2/μ2 ) and noise strength ( σ2/μ ) by deriving: ( 4 ) where is the first partial moment of the distribution of protein abundance and is the solution to: ( 5 ) Variance ( σ2 ) is: ( 6 ) Therefore , noise ( σ2/μ2 ) becomes: ( 7 ) and noise strength ( σ2/μ ) : ( 8 ) We solve the master equation for a number of different promoter architectures , where we define and for each system to describe the specific promoter states , the transitions between them , and the transcriptional activity of each state . We fit the kinetic scheme's analytical solutions of mean and noise of protein abundance to the measured mean and noise of fluorescence intensity ( see Text S1 for a detailed description of the fitting procedure and parameter constraints ) . The goodness of fit is measured by the root mean squared error ( distance , Δ ) of both mean and noise . To investigate the hypothesized effect of promoter mutations , we simultaneously fit the model to wild-type and mutant promoters while only one parameter is allowed to change between the fits . We distinguish between two hypothesized ADH1 regulatory mechanisms by fitting two models to the measured data . While ADH1 nucleosome occlusion gives a better fit than TF dislodgement , both models have a good fit to the data . To investigate if nucleosome occlusion gives a significantly better fit to the data , for each fit found by optimization , we perform 2-fold perturbations on each parameter . By looking at the distribution of fits after perturbation , we get an idea of which model is more robust and as a result is more likely to be the correct model [29] , [41] . We find that the NO model is significantly more robust than the TD model . To measure the sensitivity of the kinetic model to variations in each parameter , we performed a rigorous sensitivity analysis procedure described by Marino et al . [42] that uses the Latin Hypercube Sampling–based Partial Rank Correlation Coefficient ( LHS-PRCC ) . First , we uniformly sampled 10 , 000 instances of the model ( without fitting ) , each with a unique parameter setting , sampled from the entire allowed parameter space using Latin Hypercube sampling , and evaluated each of these models by measuring the distanced to the experimental data . Next , we calculate the Partial Rank Correlation Coefficient of the parameter value to the model score ( goodness of fit ) to measure the sensitivity of that parameter . We find that the NO model is significantly less sensitive to parameter variation than the TD model ( see Figures S7 and S8 for sensitivity analyses of the ZRT1 and ADH1/3 models , respectively ) . To investigate the effect on gene expression noise of activation and repression by the same TF ( coupled ) versus activation and repression by two different TFs ( decoupled ) , we extended the ZRT2 kinetic model to incorporate fluctuations in the concentration of TF . We use a kinetic scheme in which on-switching rates for activator and repressor can be changed independently ( Konact and Konrep , see Eq . 16 ) . These rates are determined by the distributions of activator TF and repressor TF , respectively . We therefore assume that the on-switching rates have Gamma distributions with a constant shape parameter ( burst size ) and varying scale parameter ( burst frequency ) as the activator and repressor are induced . The means of the on-switching rates were chosen to be in the range of our model fits ( 10−3 to 101 ) , which are in accordance with previously determined promoter switching rates [9] , [22] , [31] . Next , we calculate the shape parameter of the distribution of Kon using the ratio between the mean of the measured protein distribution and the chosen mean of the on-switching rates ( ratio of ∼104 ) , which we apply to the measured noise strength ( ∼103 ) . This gives a shape parameter value of around 10−1 . Because the product of shape and scale is equal to the mean , we can compute the values of the scale parameter ( 10−2 to 102 ) . We note that the qualitative result of predicted noise reduction ( Figure 7B ) is robust to 10-fold changes ( up and down ) of both shape and scale parameter of the distributions of Konact and Konrep . Finally , to simulate coupled and decoupled activation and repression , we sampled the on-switching rates of the activator and repressor from a bivariate gamma distribution with a normalized covariance of zero ( decoupled ) or one ( coupled ) . Each sample represents a single cell with some amount of activator and repressor , and therefore some Konact and Konrep . We then computed the mean expression for each “cell” using the analytical solution of the ZRT2 model and calculated the predicted noise that results from fluctuations of activator and repressor as the squared coefficient of variation ( sensitivity to TF fluctuations , η2TF ) . ( 16 )
|
In response to environmental changes , cells regulate the activity of transcription factors ( TFs ) , which in turn change the expression of dozens of downstream target genes by binding to their promoters . The response of each target gene is determined by the interplay between TF concentration and the context in which TF binding sites occur in each target promoter . To examine the relationship between promoter sequence , mechanism of regulation , and response to TF activity , we measured expression of 16 target genes of a single TF in response to changes in TF concentration in single cells . We found that different native promoters that are all targets of the same TF exhibit diverse responses to changing TF levels in terms of both gene expression level and cell-to-cell variability ( noise ) in expression . Using computational modeling and mutations of specific promoter elements , we show that the molecular mechanisms of regulation can be inferred by measuring how noise changes with expression . These results show that a single TF can regulate transcription through multiple mechanisms , resulting in similar changes in mean expression but vastly different changes in cell-to-cell variability .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"flow",
"cytometry",
"gene",
"regulation",
"microbiology",
"dna",
"transcription",
"model",
"organisms",
"molecular",
"genetics",
"cytometry",
"chromatin",
"gene",
"expression",
"biology",
"molecular",
"biology",
"systems",
"biology",
"biochemical",
"simulations",
"cell",
"biology",
"protein",
"translation",
"genetics",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"computational",
"biology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] |
2013
|
Promoter Sequence Determines the Relationship between Expression Level and Noise
|
RNA sequencing provides a new perspective on the genome of Mycobacterium tuberculosis by revealing an extensive presence of non-coding RNA , including long 5’ and 3’ untranslated regions , antisense transcripts , and intergenic small RNA ( sRNA ) molecules . More than a quarter of all sequence reads mapping outside of ribosomal RNA genes represent non-coding RNA , and the density of reads mapping to intergenic regions was more than two-fold higher than that mapping to annotated coding sequences . Selected sRNAs were found at increased abundance in stationary phase cultures and accumulated to remarkably high levels in the lungs of chronically infected mice , indicating a potential contribution to pathogenesis . The ability of tubercle bacilli to adapt to changing environments within the host is critical to their ability to cause disease and to persist during drug treatment; it is likely that novel post-transcriptional regulatory networks will play an important role in these adaptive responses .
Mycobacterium tuberculosis presents a major threat to global health , causing around 10 million new cases of tuberculosis and 2 million deaths every year [1] . The bacteria typically establish a prolonged asymptomatic infection that progresses to active disease in only a minority of individuals . During the course of infection , M . tuberculosis has to adapt to survival in a range of different microenvironments , combining persistence in a non-replicating state with periods of active cell division [2] . Transcriptional regulation in response to environmental change has been extensively analyzed in M . tuberculosis by RT-PCR and hybridization to microarrays representing the complete set of predicted coding sequences ( CDSs ) [3] , and several regulons induced under physiologically relevant growth conditions have been characterized [4]–[9] . Thirteen sigma factors , eleven two-component regulators , eleven serine-threonine protein kinases and more than one hundred predicted transcription factors have been identified in the genome of M . tuberculosis [10] . Less attention has been given to the potential role in M . tuberculosis of regulatory processes that occur subsequent to the initiation of transcription . These are often mediated by RNA and involve alterations in the efficiency of transcription , translation and stability of messenger RNAs ( reviewed in [11]–[13] ) . There are two broad categories of non-coding regulatory RNA [11]–[13] . The first is based on cis-acting regulatory elements that are present on the mRNA transcript , generally as a 5’ untranslated region ( UTR ) . 5’ UTRs can regulate gene expression by forming secondary structural features that enhance or inhibit transcription or translation of their cognate mRNA . Their structural conformation is often determined by interaction with particular protein , tRNA or small molecule ligands . 5’ UTRs that respond to small molecules are referred to as riboswitches [14] . The second category of RNA regulators is small RNA molecules ( sRNAs ) that bind in trans to their mRNA target . These include antisense RNAs that are transcribed from the DNA strand opposite to a CDS , inhibiting translation and promoting degradation by base-pairing with the corresponding mRNA transcript [13] . A second class of sRNAs are transcribed from a separate location on the genome – typically an intergenic region ( IGR ) , though attenuated 5’ UTR transcripts can also act as sRNAs [15] – and again influence translation and degradation of mRNAs by a more limited degree of base pairing . These “trans-encoded” sRNAs are functionally analogous to microRNAs in eukaryotic cells [16] . RNA regulators are frequently implicated in adaptive responses of bacteria to environmental change and may therefore be expected to play a role in the pathogenesis of M . tuberculosis . sRNAs have been shown to regulate expression of virulence determinants in a number of bacterial pathogens , located in some cases within defined pathogenicity islands and carried specifically by pathogenic strains [17]–[20] ( reviewed in [21] ) . Initial studies have described the occurrence of sRNAs in M . tuberculosis though the extent of the regulatory RNA network in these bacteria , and its relevance to gene expression during infection , are unknown [22] , [23] . Knowledge of regulatory networks controlled by non-coding RNA in bacteria has undergone a rapid expansion over the last decade , as a result of transcriptional profiling approaches based on high-density tiling arrays and deep sequencing of whole genome cDNAs ( RNA-seq ) . Sequence-based transcriptomes have been described for a range of bacteria , including several human pathogens [17] , [24]–[29] . The aim of the present study was to apply an analogous approach to characterize the transcriptome of M . tuberculosis , with a particular focus on identification of the nature and extent of non-coding RNA .
RNA extracted from three independent exponential phase cultures of M . tuberculosis was used to generate cDNA preparations that were then analyzed by Illumina-based sequencing . Two of the samples were analyzed as technical replicates in separate sequencing runs , generating a set of five total transcriptome profiles ( Table 1 ) . After removal of reads mapping to rRNA molecules an average of just over 3 million reads , corresponding to 8% of total , mapped to annotated protein CDSs in the sense orientation . A further 0 . 5 million reads mapped to CDSs in the antisense orientation , and 0 . 7 million reads mapped to IGRs . IGRs make up less than 10% of the genome , and thus the density of reads mapping outside of CDSs was more than two-fold higher than that seen for predicted coding transcripts . Calculation of pairwise correlation coefficients demonstrated a high degree of reproducibility between samples ( r≥0 . 93 Table S1 ) , and the total transcriptome data are recorded as an average of the five samples in Table S2 ( sense and antisense reads for each CDS ) and Table S3 ( reads mapping to IGRs of at least 50 nucleotides ) . To facilitate comparison of expression levels of different genes , data for CDSs are also presented in the form of reads per kilobase per million reads ( RPKM ) . cDNA was also sequenced from two stationary phase cultures of M . tuberculosis . The correlation coefficient between the samples ( r = 0 . 82 ) indicated a lower level of reproducibility than that seen in exponential phase . This was driven mainly by a lower amount of mRNA , with only 2% of stationary phase reads mapping to CDSs as compared to 8% in exponential phase as well as a dramatic increase in the level of intergenic reads ( Table 2 ) . However , the overall pattern of relative gene expression was very similar between the replicate samples , and data are expressed as average values in Tables S2 and S3 . To explore the functional role of the dominant stationary phase sRNA , MTS2823 we engineered its over-expression in exponential phase cultures of M . tuberculosis H37Rv under the control of a strong rRNA promoter . Over-expression of MTS2823 had a slight but clear effect on the growth rate ( Figure 8A ) . However , microarray analysis revealed a striking overall down-regulation of multiple genes with 301 genes showing ≥2 . 5-fold change ( P−value ≤0 . 05 ) ( Figure S6 , Table S5 ) . Apart from MTS2823 itself ( represented by microarray probe MtCDC1551-3762 ) , only two genes were up-regulated; Rv2035 encoding a potential activator of HspG ( up 3 . 2-fold ) and Rv3229c , encoding a fatty acyl desaturase DesA3 ( up 3 . 1-fold ) . The down-regulated subset showed a significant over-representation of genes involved in energy metabolism ( class I . B; P−value<0 . 01 after FDR correction ) as well as a trend towards over-representation of genes involved in macromolecular synthesis ( class II . A ) and insertion elements ( Figure S5 ) . Rv1131 ( prpC/gltA1 ) encoding methyl citrate synthase , showed the greatest change of expression , with a 15-fold down-regulation . Related genes involved in the citrate synthase cycle , encoding methyl citrate dehydratase ( Rv1130 , prpD; down 7-fold ) and an associated transcriptional regulator ( Rv1129c , lrpG; down 5 . 6-fold ) , were similarly repressed . Analysis of associated partners identified in the STRING database [47] revealed down-regulation of an extended methyl citrate network ( Figure 8B ) . Down-regulation of prpC in exponential phase cultures has previously been observed as a consequence of deletion of several sigma factors [48] , [49] . sigE , sigB , sigG and sigA transcripts were all down-regulated in the MTS2823 over-expressing strain ( Table S5 ) . Five of the ten vapC toxin homologues were down-regulated ≥2 . 5-fold while none of their antitoxin partners were affected ( Table S5 ) . Quantitative RT-PCR ( qRT-PCR ) analysis of selected genes confirmed the pattern seen by microarray , over a more extended dynamic range ( Table 11 ) . The abundance of selected sRNAs in stationary phase cultures encouraged us to explore their expression in non-replicating M . tuberculosis during infection . Following aerosol infection in mice , M . tuberculosis undergoes a period of active replication before engagement of the adaptive immune response . Infection then persists for months with little or no change in bacterial load before mice die from cumulative lung damage . There is uncertainty as to whether the chronic phase of infection is associated with persistence of non-replicating bacteria , or a balance between bacterial replication and killing [50]–[52] . To assess expression of non-coding RNAs during the chronic stage of infection , we prepared mycobacterial RNA from lungs of mice and used qRT-PCR to measure selected sRNA and mRNA transcripts . Results were normalized to 16S rRNA . Consistent with RNA-seq and Northern blotting , qRT-PCR analysis confirmed the abundance of MTS2823 in exponential cultures with further increase in stationary phase , and the stationary phase induction of MTS0997 and MTS1338 ( Figure 9 ) . All three of the sRNAs were present at very high levels in chronically infected lung tissue; each accumulating to a level relative to 16S rRNA that was increased over and above that observed in stationary phase ( Figure 9 ) . By comparison , groES , a highly abundant mRNA in exponential cultures was markedly reduced in stationary phase , with an intermediate level in infected tissues suggesting the presence of replicating as well as non-replicating populations in line with the model suggested by Chao and Rubin [53] ( Figure 9 ) .
Application of deep sequencing technologies to study the transcriptome of M . tuberculosis has uncovered an abundance of non-coding RNAs including cis-encoded regulatory elements , antisense transcripts and intergenic sRNAs . After removal of signals from ribosomal RNAs , the percentage of reads mapping in antisense orientation and to IGRs represents 28% of the total transcriptome in exponential phase M . tuberculosis . This is broadly similar to the 25% reported for Salmonella typhi and the 27% reported for Helicobacter pylori [24] , [25] , and contributes to a growing appreciation of the prominence of non-coding RNA in bacteria . Non-coding RNAs can regulate gene expression at a post-transcriptional level , providing a level of control intermediate between conventional transcriptional control and post-translational protein turnover , that is particularly useful in the rapid response to stress stimuli and may play an important role in generation of population heterogeneity [54] . Characterisation of non-coding RNA is likely to be important in understanding the complex biology underlying tuberculosis infection [2] . Sequence-based transcriptional profiling has advantages over hybridization-based microarray analysis in displaying a greater dynamic range with single-nucleotide resolution . The number of reads mapping to individual sequences provides a realistic assessment of relative transcript abundance , although the potential for variability in the efficiency of reverse transcription precludes precise quantitation . M . tuberculosis has only a single ribosomal RNA operon , but rRNA represents more than 80% of total RNA content , and we compared strategies to reduce this signal by physical removal of rRNA prior to sequencing or by computational removal of rRNA prior to sequencing or computational removal of rRNA reads after sequencing . The latter approach has the advantage of limiting potential RNA degradation or otherwise skewing of the RNA composition during processing [55] . This method still provided an average level of coverage of approximately one read per nucleotide ( outside rrn ) in exponential phase samples , and was used to generate the datasets for the present analysis . To enhance representation of shorter transcripts , we included an RNA ligation step . In spite of this , RPKM values for 5S rRNA ( 115 nucleotides ) were between 5- and 10-fold lower than those for 16S and 23S rRNA ( 1 , 537 and 3 , 138 nucleotides , respectively ) , and only a low number of reads mapped to tRNAs ( <100 nucleotides ) , Therefore it is likely that sRNAs are under-represented , quantitatively and perhaps qualitatively , in our final transcriptome . Our classification of RNA as “non-coding” is contingent on current genome annotations , and we cannot exclude the possibility that some of these newly identified transcripts encode short peptides . The coding transcriptome of M . tuberculosis reveals a profile consistent with expectations for exponentially growing bacteria , with genes involved in energy production and macromolecule biosynthesis significantly over-represented in the transcriptome as compared to the overall genome . M . tuberculosis-specific features include abundant representation of transcripts encoding cargo proteins exported by Type VII secretion systems [30] with genes encoding ESX-1-exported proteins ESAT6 , CFP10 , EspC and EspD amongst the top fifty transcripts . Transcript abundance is a feature of many , though not all of the ESX export proteins and , while bearing in mind potential changes in transcriptional regulation during infection , this may contribute to their differential immunogenicity [56] . Transcripts encoding several TA pairs are also highly abundant . The genome of M . tuberculosis is remarkably rich in Type II TA systems believed to modulate gene expression through ribonuclease activity [33] , but their role in mycobacterial physiology and pathogenesis remains unclear . The stationary phase transcriptome is dominated by an abundance of genes induced as part of the DosR regulon that has been extensively characterized in response to a hypoxic environment [5] , [8] . We had not set out to induce hypoxic conditions and we were surprised by the strength of the DosR signal . Enhanced regular aeration of cultures resulted in increased bacterial growth in our culture system , however , and we assume that the stationary phase arrest seen in this study reflects some level of oxygen depletion . The distribution of abundant transcripts across the antisense transcriptome is largely the inverse of the coding transcriptome , with energy metabolism and macromolecule synthesis under-represented at the expense of conserved hypotheticals and proteins of unknown function . Reflecting a pattern that is emerging from a range of sequence-based bacterial transcriptome studies [57] the antisense transcriptome of M . tuberculosis includes a diverse size-range of transcripts arising both within genes and as a result of 3’ UTR overlaps between convergent gene pairs . The abundance of long 3’ UTRs observed in M . tuberculosis may be due to inefficient termination resulting from replacement of the characteristic bacterial L-shaped intrinsic transcriptional terminator consisting of a hairpin loop followed by a poly-U stretch [58] by an I-shaped terminator lacking a poly-U stretch [59] , [60] . 3’ UTR antisense transcripts have the potential to provide a regulatory connection between neighbouring genes [57] and enrichment of cell envelope and toxin genes in 3’-3’ convergent pairs may reflect an organizational motif that has functional consequences for gene expression . Several antisense transcripts are associated with foreign genetic elements , including an abundant transcript that maps to CRISPR-association genes . The CRISPR locus is a non-coding RNA defence system that bacteria use to combat invasion by foreign genetic sequences [36] , [37] . In E . coli the CRISPR-associated ( cas ) proteins involved in this process are silenced by the H-NS DNA-binding protein during exponential growth [61] and the antisense transcript extending across cas-encoding Rv2816c and Rv2817c may perform an analogous regulatory function in M . tuberculosis . Although we had not specifically enriched our libraries for primary 5’ ends [25] we were able to estimate the minimum length of most 5’ UTRs by examining Artemis profiles , and identify multiple very long ( ≥150 nt ) 5’ UTRs . By analogy with other bacteria , we anticipate that these will play a role in coordination of gene expression and in regulation by small molecule ligands . Many of the genes with long , highly expressed 5’ UTRs encode ribosomal proteins or other factors involved in translation . In Escherichia coli coordinated expression of ribosomal RNA , ribosomal proteins and ribosome-associated factors involves the interaction of ribosomal proteins with the 5’ UTR of mRNAs causing attenuation of expression [41] . Similarly , in B . subtilis 5’ UTR binding by ribosomal protein L20 controls the expression of the infC operon [62] In M . tuberculosis sequence alignments of 5’ UTRs of rpoA with rpsM and of infA with infC suggest that the upstream sequences may represent elements involved in coordinate regulation of core transcriptional and translation processes , analogous to the systems described in E . coli and B . subtilis [41] , [62] . The 5’ UTRs of groE genes overlap the CIRCE elements that are recognized by the HrcA transcriptional repressor [39] but also play a role in post-transcriptional control of mRNA stability in B . subtilis [63] and in Rhodobacter capsulatus [64] . It is likely that the transcriptional control of the heat shock response in M . tuberculosis is complemented by an additional layer of post-transcriptional regulation via this element . A further subset of 5’ UTRs correspond to predicted riboswitches that regulate expression of their cognate gene in response to the presence or absence of small molecules . In the absence of activation signal , riboswitch-mediated attenuation generates truncated transcripts mapping to the 5’ end of genes; a profile that is associated with many of abundant 5’ UTRs identified in the non-coding transcriptome of M . tuberculosis . It has been suggested that 2% of B . subtilis genes may be regulated by riboswitches [13] and this form of regulation may have similar importance for M . tuberculosis . Small regulatory RNAs , generally encoded in IGRs , have provided a focus for recent interest with respect to their role in bacterial responses to environmental change and pathogenesis ( reviewed in [12] ) . Trans-encoded sRNAs typically function by base-pairing with a panel of mRNA targets in a manner that prevents translation and accelerates their degradation . In the case of the best-characterised examples , the interaction of sRNA with mRNA is mediated by the Hfq chaperone . M . tuberculosis belongs to the subset of bacteria which lack an identified Hfq homologue , but RNA-seq analysis is consistent with previous descriptions of the presence of multiple intergenic sRNAs . To explore the functional role of sRNAs in M . tuberculosis , we manipulated the expression of MTS2823 . This sRNA is present in all growth phases , but is expressed at very high levels during stationary phase; we reasoned that we might be able to recapitulate its stationary phase function by over-expression in exponential phase . Over-expression of MTS2823 caused widespread down-regulation of gene expression , with the most profound effect on the gene classes preferentially associated with exponential growth . Genes linked to the methyl citrate network , in particular prpC , and to a lesser extent prpD were most strongly down-regulated . This could represent a preferential targeting by MTS2823 , either to reduce the utilization of substrates ( i . e . propionyl-CoA and/or oxaloacetate ) or to reduce the accumulation of potentially toxic intermediates such as methyl citrate . The participation of these genes within a feed-forward regulatory loop provides an alternative explanation for their amplified response [65] , [66] . Methyl citrate genes are also down-regulated in response to hypoxia and DNA damage [8] , [67] . During transition to stationary phase , the down-regulation of genes required for active bacterial replication is commonly mediated by expression of a 6S RNA molecule that interferes with transcription by RNA polymerase associated with the principal sigma factor [68] . An M . smegmatis homologue of MTS2823 was recently identified in a bioinformatics screen based on suboptimal structural criteria for 6S RNA , but detailed analysis , including lack of binding to RNA polymerase , led the authors to conclude that it is not a genuine 6S RNA homologue [69] . Our results suggest that MTS2823 may have functional homology with 6S RNA in its ability to mediate down-regulation of exponential phase genes , although the molecular mechanism remains to be clarified . In vitro models of mycobacterial persistence typically involve growth arrest , associated with down-regulation of a broad panel of genes and up-regulation of a more limited set of condition-specific genes [8] , [70] . MTS2823 may be an important mediator of the common down-regulatory component . We anticipate that the majority of M . tuberculosis sRNAs will function in post-transcriptional regulation of more restricted sets of mRNAs involved in adaptive responses to specific environmental stimuli . MTS1338 provides a characteristic example . Expression of MTS1338 is strongly induced under the conditions used to generate stationary phase cultures , by a mechanism that is at least partly dependent on the hypoxia-responsive DosRS two-component transcriptional regulator . sRNAs involved in Hfq-mediated interactions are generally degraded along with their mRNA target [13] . In contrast , we observed accumulation rather than degradation of M . tuberculosis sRNAs in the stationary phase induction model . Even higher levels of sRNA accumulation were observed in the lungs of mice during chronic tuberculosis infection , with MTS2823 present at 16% of the level of 16S rRNA , MTS0997 at 5% and MTS1338 at 0 . 5% . The abundance of sRNAs in infected tissues suggests that they may provide useful biomarkers , and potentially important functional mediators , during the course of disease . In summary , sequence-based analysis of the transcriptome of M . tuberculosis uncovers a wide range of novel non-coding RNAs with the potential to influence patterns of gene expression in vitro and during infection . Targeted analysis by deletion and over-expression may provide insights into the molecular pathogenesis of this important human disease .
The NIAID DIR Animal Care and Use Program adheres to the U . S . Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing , Research , and Training , the PHS Policy on Humane Care and Use of Laboratory Animals , the Guide for the Care and Use of Laboratory Animals , and the U . S . Animal Welfare Regulations . We operate in accord with NIH Policy Manual 3040-2 ( Animal Care and Use in the Intramural Program ) and in compliance with the provisions of the NIH's intramural Institutional Assurance ( NIH IRP , PHS Assurance A4149-01 ) on file with the NIH Office of Laboratory Animal Welfare ( OLAW ) . The NIH Intramural Research Program is fully accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care , International . The Animal Care and Use Committee ( ACUC ) of the National Institute of Allergy and Infectious Diseases , Division of Intramural Research , with permit number NIH IRP , PHS Assurance A4149-01 , approved the animal study protocol LCID-3E under which all animal experiments were performed . Table S6 lists all oligos used in this study . M . tuberculosis H37Rv and M . tuberculosis H37Rv:: δdosR::kan [5] were grown in Middlebrook 7H9 supplemented with 0 . 2% glycerol and 10% ADC in roller bottle culture . Exponential phase cultures were harvested at an OD600 0 . 6 to 0 . 8; stationary phase cultures were harvested one week after OD600 had reached 1 . 0 . M . tuberculosis H37Rv:: δRv3676 [45] had a significantly slower growth rate and was grown two weeks after OD600 1 . 0 to generate stationary phase cultures . To monitor β-galactosidease activity , M . tuberculosis and M . smegmatis were grown on solid 7H9 supplemented with 25 µg/ml kanamycin and 50 µg/ml X-gal . Eight-week-old C57Bl/6 mice were infected with 100 colony forming units of M . tuberculosis H37Rv by aerosol using a BioAerosol nebulizing generator ( CH Technologies Inc . , Westwood , NJ ) for 10 minutes and the infection was allowed to develop for 9 months by which time the CFUs had reached 7 . 5±2 . 5×105 . Mice were sacrificed , each lung homogenized in 5 ml Trizol reagent ( Invitrogen Corporation , Carlsbad , CA ) and bacteria were separated from the lysed lung tissue by centrifugation at 3500 rpm for 10 minutes at 4°C . The cell pellet was washed once in Trizol , followed by resuspension in 1 ml Trizol containing 5 µg/ml glycogen and stored at −80°C until bacterial RNA extraction . lacZ reporter fusions were made by PCR amplification of the MTS0997 promoter region with primers P0997 . f+r . The resulting fragment was inserted into pEJ414 [71] . Mutation of the −10 box was done with site-directed mutagenesis using primers SDM0997 ( −10 ) . f+r and Pfu Ultra ( Stratagene ) according to manufacturer's instructions . Reporter plasmids were electroporated into M . smegmatis mc2155 [72] as well as M . tuberculosis H37Rv . Over-expression construct of MTS2823 . The over-expression plasmid was made by replacing the rrnB promoter fragment in a previously described over-expression plasmid [22] with the region spanning −80 to −8 from the same promoter to generate pKA303 . The −10 box is immediately followed by a HindIII site where the sRNAs can be inserted , which means that the sRNAs can be expressed from their own transcription start site . A 316 bp HindIII fragment containing MTS2823 ( +1 to +304 ) was PCR amplified using primers ox2823 . f and ox2823 . r . The fragment was cloned into pCR-Blunt II-TOPO ( Invitrogen ) , sequenced and subsequently sub-cloned into pKA303 . The resulting plasmid was electroporated into M . tuberculosis H37Rv . Cultures were cooled rapidly by the addition of ice directly to the culture before centrifugation . RNA was isolated by means of the FastRNA Pro blue kit from QBiogene/MP Bio according to manufacturer's instructions . RNA was treated Turbo DNase ( Ambion ) until DNA free . The quality of RNA was assessed using a Nanodrop ( ND-1000 , Labtech ) and Agilent bioanalyser . In order to enrich for small transcripts in the RNA-seq reactions , total RNA was treated with tobacco acid pyrophosphatase ( TAP , Epicentre technologies ) and subsequently ligated with T4 RNA ligase ( Ambion ) before reverse transcription . Bacteria from mouse lungs were disrupted by bead beating in Trizol in a Fastprep instrument , chloroform extracted and ethanol precipitated . Due to the low bacterial count RNA from three mice was pooled . The RNA was subsequently purified with Turbo DNase and phenol extraction before reverse transcription . Northern blots , RT-PCR for investigating co-transcription and 5’ RACE were carried out as described previously [22] . Construction of cDNA libraries and sequencing was carried out essentially as described previously [73] . Sequence reads were aligned to the reference sequence of M . tuberculosis H37Rv ( EMBL accession code AL123456 ) as paired end data using BWA [74] . The orientation of the second read in correctly mapped pairs was reversed using Samtools [74] before producing coverage plots in order to maintain the directional fidelity of the data . Novel intergenic features were annotated by visual inspection of the transcriptome using the Artemis genome browser . RPKM values were calculated using only sequence reads that mapped to annotated features unambiguously ( the ‘XA’ note in the alignment file was used to identify alternative mapping locations ) and on the correct strand . To ascertain the proportion of a gene to which reads could be unambiguously mapped , 54 nucleotide paired end data were simulated from the genome sequence and aligned to the sequence under the settings used for mapping the RNA-seq reads . cDNA for quantitative RT-PCR and RNA-seq was made with random primers and Superscript III according to manufacturer's instructions ( Invitrogen ) with an additional incubation step for one hour at 55°C in order to optimize read-through of highly structured sequences . qRT-PCR was carried out on a 7500 Fast Real-Time PCR System ( Applied Biosystems ) using Fast SYBR Green Master Mix ( Applied Biosystems ) . RNA without RT ( RT- ) was analyzed alongside cDNA ( RT+ ) . Standard curves were performed for each gene analyzed and the quantities of cDNA within the samples were calculated from cycle threshold values . Data were averaged , adjusted for chromosomal DNA contamination ( RT+ minus RT- ) and normalized to corresponding 16S RNA values . Microarrays were performed on three biological replicates of over-expression strain against the control strain containing empty vector . Reverse transcription , hybridization and subsequent analysis were carried out as described [75] . All analyses where carried out using STATA s . e . m version 10 . To compare the frequencies of different functional categories we used Fisher's exact test . For the comparison between the annotated genome and the transcriptome , and for the comparison of the transcriptome with the antisense transcripts we used all the categories and therefore the False Discovery Rate correction for multiple testing [76] was applied . For the rest of the analyses only selected categories were tested and no test for multiple correction was required .
|
Tuberculosis bacteria are able to hide quietly inside the body for years or decades before reawakening to cause disease . If we knew more about how the bacteria change from a harmless persistent form to an aggressive disease-causing form , we could develop drugs that would be more effective in treating active tuberculosis and may also allow us to eliminate the infection before it erupts into disease . The key to this is in knowing how the bacteria determine which of their genes to express at different times . By applying modern sequencing technologies we have discovered a new putative network of gene regulation in Mycobacterium tuberculosis that is based on RNA molecules rather than protein molecules . We anticipate that this finding will open the way for new research that will allow us to understand the fundamental mechanisms underlying this deadly human disease , and that will help us to design better tools for prevention and treatment of TB .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"microbial",
"metabolism",
"microbiology",
"rna",
"synthesis",
"bacterial",
"pathogens",
"microbial",
"physiology",
"microbial",
"pathogens",
"biology",
"gram",
"positive",
"biochemistry",
"rna",
"bacterial",
"physiology",
"nucleic",
"acids"
] |
2011
|
Sequence-Based Analysis Uncovers an Abundance of Non-Coding RNA in the Total Transcriptome of Mycobacterium tuberculosis
|
The role of APOBEC3 ( A3 ) protein family members in inhibiting retrovirus infection and mobile element retrotransposition is well established . However , the evolutionary effects these restriction factors may have had on active retroviruses such as HIV-1 are less well understood . An HIV-1 variant that has been highly G-to-A mutated is unlikely to be transmitted due to accumulation of deleterious mutations . However , G-to-A mutated hA3G target sequences within which the mutations are the least deleterious are more likely to survive selection pressure . Thus , among hA3G targets in HIV-1 , the ratio of nonsynonymous to synonymous changes will increase with virus generations , leaving a footprint of past activity . To study such footprints in HIV-1 evolution , we developed an in silico model based on calculated hA3G target probabilities derived from G-to-A mutation sequence contexts in the literature . We simulated G-to-A changes iteratively in independent sequential HIV-1 infections until a stop codon was introduced into any gene . In addition to our simulation results , we observed higher ratios of nonsynonymous to synonymous mutation at hA3G targets in extant HIV-1 genomes than in their putative ancestral genomes , compared to random controls , implying that moderate levels of A3G-mediated G-to-A mutation have been a factor in HIV-1 evolution . Results from in vitro passaging experiments of HIV-1 modified to be highly susceptible to hA3G mutagenesis verified our simulation accuracy . We also used our simulation to examine the possible role of A3G-induced mutations in the origin of drug resistance . We found that hA3G activity could have been responsible for only a small increase in mutations at known drug resistance sites and propose that concerns for increased resistance to other antiviral drugs should not prevent Vif from being considered a suitable target for development of new drugs .
The human and mouse APOBEC3 ( A3 ) protein family members , hA3F and G and mA3 , respectively , are well-studied host factors that restrict retrovirus replication . [1]–[3] . These components of the innate cellular defense system inhibit retroviral propagation by inducing deamination of C-to-U residues in the negative strand of retroviral DNA during reverse transcription [4] , resulting in a mutated provirus with biased G-to-A changes on the plus strand [5]–[8] . Due to the specificity of A3 for single-stranded DNA , the frequency of induced mutations forms an increasing gradient in the genome from the primer binding site ( PBS ) to the polypurine tract ( PPT ) [9] . In the case of HIV-1 , whose genome contains a central PPT ( cPPT ) , this effect results in dual gradients of G-to-A mutations from the PBS to the cPPT and the cPPT to the PPT [10] . A role of A3-mediated defense in more distant retroviral evolution became apparent when we recently showed that mA3 likely contributed to inactivation of the infectivity of some types of endogenous MLV at the time of their integration into the host germline around a million years ago [11] . Taken together , these observation show that A3 was active as an antiviral factor in the distant evolutionary past , and that it is so today . However , the effects these APOBEC3 restriction factors have had on the evolution of actively replicating retroviruses such as HIV-1 are less clear . The present study was undertaken to look for a signal of past A3-induced G-to-A changes in the sequence of modern day HIV-1 genomes . We hypothesized that the observed variation in G∶A ratios in different viruses and in their sensitivities to A3 [5] , [12] , [13] is partly an effect of earlier A3-induced mutation and selection pressure on the virus genomes . Highly G-to-A mutated sequences are unlikely to be transmitted due to deleterious mutations . However , less efficient A3 activity , and thus fewer G-to-A mutations , could allow virus transmission , although with reduced efficiency . In this scenario , mutated A3 target sequences within which G-to-A mutations are the least deleterious are more likely to survive purifying selection pressure . This model predicts that as mutations accumulate with time , the ratio of nonsynonymous ( NS ) to synonymous ( S ) sites in probable A3 target sequences will increase with increasing virus generations more rapidly than at sites that are not A3 targets . To study the potential of human APOBEC3G ( hA3G ) to affect HIV-1 evolution , we developed an in silico model that takes into account the observed mutation gradient and the probabilities of hA3G-mediated mutation at each site , derived from the sequence contexts of mutation sites ( hereafter referred to as “contexts” ) found in the literature [9] , [14] . After adjustment for location in the provirus in our simulations , we found that the frequency of G-to-A mutations introduced into HIV-1 could be partly explained by the virus's genetic profile as well as its distribution of probable target sequence contexts . Our simulation approach is supported by independent experimental results obtained by passaging HIV-1 under conditions that permit a high level of A3G-induced mutation . Further analyses showed that the ratio of NS to S-mutations at hA3G target sites has increased during HIV-1 evolution compared to random controls , implying an evolution where A3 sites have had higher mutation rates than non-A3 sites and therefore been subjected to greater influence from purifying selection . We also analyzed the potential of low-level hA3G-mediated G-to-A mutations to give rise to naturally occurring drug resistance mutations .
To calculate the likelihood of hA3G-mediated G-to-A mutation at different sites in HIV-1 , we analyzed mutation site contexts surrounding 1324 observed G-to-A mutations attributed to hA3G from the literature [9] , [14] . Given the variability in these data and to minimize false predictions , we used our previous experience with mutation site context analyses [11] and included as many significant sites as possible ( Figure S1 ) . Thus , we collected sequences up to 3 bases 5′ and 5 bases 3′ of the G-to-A mutations . Using nucleotide frequencies at each offset position from the mutation target G-nucleotide , we constructed a position-specific scoring matrix ( PSSM , Figure S1 ) using odds ratios ( i . e . , observed relative to expected nucleotide frequencies ) and similar to the one derived for our previous study on endogenous nonecotropic MLVs [11] . Due to limited representation of observed non-A3G mediated G-to-A mutation sites in the literature , we included an artificial nucleotide frequency background to compensate for sampling artifacts using a previously described pseudo count method ( Figure S1 ) [15] . Using automated PERL scripts we simulated a collection of 106 random mutation site contexts based on nucleotide frequencies observed in HIV-1 . These data were used to derive a score distribution and a cumulative frequency distribution for probability calculation of context scores . Scores for the previously described G-to-A contexts were calculated and compared to the newly calculated score and cumulative frequency distributions . An analysis of G-to-A mutation site contexts collected from endogenous MLVs [11] supported this putative cutoff for likely A3 targets ( data not shown ) . In addition to our simulations using all hA3G targets , we also included separate analyses using the 80% probability cutoff for introduction of G-to-A mutations ( see below ) . Mutations were simulated in HIV-1 ( AF033819 ) using automated PERL scripts as outlined in Figure 1 . Briefly , a mutation position was randomly selected using the previously observed dual gradients of increasing likelihood for mutation in the PBS-cPPT and cPPT-3′PPT regions [9] , [10] . A score for the sequence context surrounding each randomly selected G-nucleotide was compared against cumulative frequency distributions of simulated random viral hA3G target contexts , and used to test whether a G-to-A mutation should be introduced at that site . If the context likelihood exceeded a random number between 0 and 1 a G-to-A mutation was introduced at that site . Failure to introduce a mutation initiated a new attempt at another random position . Information from successful mutations , including their location and whether they were synonymous ( S ) or nonsynonymous ( NS ) ( i . e . unchanged and changed amino acid respectively ) – was collected and followed by another round of simulated G-to-A mutation . Simulations were repeated until a stop codon was introduced into any gene or a normal ATG initiation codon was altered , after which the HIV-1 sequence with accumulated G-to-A mutations was collected into a database . In this way , we collected 4000 independent genomes with accumulated G-to-A mutations . All simulation rounds were performed using either contexts with probabilities over 80% or all G-nucleotides independent of context . To analyze the potential evolutionary footprints of hA3G-mediated G-to-A mutations in HIV-1 we used automated PERL scripts to test the ratios between the summarized probabilities of nonsynonymous to synonymous G-nucleotide contexts . To look for mutations that might have occurred during the evolutionary history of the virus , the same analysis was performed for putative ancestral G-nucleotide targets observed as A-nucleotides in the virus . Ratios were compared using a χ2 test ( 1 degree of freedom ) for HIV-1 ( AF033819 ) and HIVcon ( group M consensus sequence derived from subgroup gene alignment consensus sequences ) downloaded from www . hiv . lanl . gov . Locations of known HIV-1 protease inhibitor ( PI ) , nucleoside RT inhibitors ( NRTI ) , non-nucleoside RT inhibitor ( NNRTI ) , integrase inhibitor ( INI ) , and fusion inhibitor ( FI ) resistance mutation sites were collected from the literature [16] and the HIV drug resistance database ( http://hivdb . stanford . edu ) . We then analyzed all G-nucleotide contexts and used their calculated probabilities in our in silico model to estimate the potential likelihood for hA3G-mediated G-to-A mutations leading to drug resistance . Highly mutated HIV-1 samples were obtained from a passaging experiment . A detailed description of these studies [17] is available upon request . In brief , the YRHHY>A5 mutation at amino acids 40 to 44 , which renders HIV-1 Vif unable to efficiently bind to hA3G but still be effective against hA3F [18] , was inserted into the replication-competent HIV-1 plasmid pNL4-3 ( AF324493 ) using overlapping PCR to generate pNL4-3-YRHHY>A5 . For virus production , 4×106 293T cells were transfected with 20 µg of either pNL4-3 or pNL4-3-YRHHY>A5 and 1 . 2 µg pGL , which expresses the green fluorescent protein from a cytomegalovirus immediate early promoter ( Invitrogen ) to determine transfection efficiency . The virus-containing supernatant was harvested 48-hours after transfection , and assayed for RT activity . On day one of infection , 1×106 CEM cells were infected with 200 µl of virus supernatant ( 1000 RT units of virus ) for five hours . The volume of medium was then increased to 5 ml . A 4 ml aliquot of cells and virus-containing supernatant was removed at two day intervals post infection ( days 3 , 5 , 7 etc . ) , and stored at −70°C for subsequent analysis . A 4 ml aliquot of fresh CEM-CM ( complete medium , RPMI+10% fetal calf serum+1% penicillin-streptomycin , in which the CEM cells were maintained ) . was then added to the remaining 2 ml cell and virus suspension and the sample incubated for another 2 days . PCR reactions using 2 µl of extracted DNA from the infected cells , 1 µl High Fidelity Platinum Taq polymerase ( Invitrogen ) and 20 pmoles each of Vif F ( 5′CAGGGAGATTCTAAAAG3′ ) and Vif R ( 5′GGATAAACAGCAGTTGTTGC3′ ) primers were performed at 98 C for 2 minutes , followed by 30 cycles at 98 C for 30 seconds , 55 C for 30 seconds and 72 C for one minute . Final PCR product extension was performed at 72 C for 10 minutes . The PCR product was then cloned into the TOPO TA vector ( Invitrogen ) and the individual clones were sequenced using the primer NL43 seq 4921F ( 5′GAGATCCAGTTTGGAAAGGAC3′ ) . Mutation site contexts from the in vitro experiments described above were collected using automated PERL scripts as described earlier [11] . Scores and likelihoods were calculated in our current model . In these experiments , G-to-A mutations were considered either natural RT errors or hA3G-mediated changes since the YRHHY>A5 vif mutant is still capable of inhibiting hA3F [18] . To evaluate our in silico model , we simulated the number of mutations we observed in vitro on the same vif region of pNL4-3 ( AF324493; positions 5041–5770 ) . We then isolated all contexts surrounding simulated G-to-A mutations using our automated PERL scripts and calculated scores and likelihoods for direct comparison to the in vitro results .
Using automated PERL scripts and previously identified hA3G-induced mutations [9] , [14] , we extracted sequence contexts extending 3 bases upstream and 5 bases downstream of each potential G-to-A mutation position in the HIV-1 genome and constructed a Log position-specific scoring matrix ( PSSM , Figure S1 ) . We next created a random set of 106 mutation site contexts ( Figure 2A , circles ) based on the HIV-1 base composition and plotted the histogram and cumulative frequencies of that distribution ( Figure 2A , diamonds ) . This analysis yielded an approximately normal distribution with a mean score of −0 . 7 and a standard deviation of 1 . 5 ( black boxplot ) . To determine the pattern of actual mutations , we then tested context scores of previously observed G-to-A mutation sites [14] against this distribution . The distribution of calculated probabilities of these mutation “hotspots” [14] was significantly skewed relative to random , with a mean of 2 . 3 and a standard deviation of 1 . 1 ( Figure 2A; red boxplot ) , with 96% of mutations at sites with a score of 80% or greater , suggesting that a score threshold of 80% would capture the large majority of actual hA3G-mutation targets . We also found support for an 80% cutoff when G-to-A mutation site contexts collected from endogenous MLVs [11] were analyzed in the same way using simulated score distributions . To estimate the potential of hA3G-mediated mutations to contribute to HIV-1 evolution , we used our in silico model ( Figure 1 ) to simulate stochastic accumulation of G-to-A mutations in the genome before an obviously lethal mutation ( introduction of a stop codon or loss of an initiation codon ) appeared in any gene ( Figure 2B and C ) . The simulation took into account both the dual hA3G mutation gradients ( PBS-cPPT and cPPT-3′PPT ) and the target probability at each G-nucleotide calculated from the surrounding sequence context . A collection of 4000 HIV-1 sequences – with nonsynonymous ( NS , Figure 2B ) and synonymous ( S , Figure 2C ) mutations accumulated until one lethal mutation was introduced – was analyzed . Most of the simulated mutations displayed the expected dual gradient ( black circles ) . Very infrequent mutations at the bottom of the graphs result from contexts that are at or very close to the probability cutoff . To ensure that this pattern did not reflect a nonuniform distribution of high probability hA3G targets , we also plotted the distribution of context probability scores for all G-residues . As can be seen ( Figure 2B and C , gray squares ) , there is no indication of an uneven distribution of context probabilities across the HIV-1 provirus . The distribution of the number of accumulated G-to-A changes per provirus resulting from our simulations is presented in Figure 3 . For comparison , simulations were also repeated in which all G-nucleotide sites were considered for successful mutations and compared to those derived using the 80% probability cutoff ( Figure 2 and above ) . As expected from our model , without selection pressure on the virus sequence , NS ( red ) to S ( blue ) mutation ratios were consistent ( approximately 3∶1 ) independent of the number of mutations that were introduced into the viruses . Further , the numbers of random G-to-A mutations introduced before a stop codon appeared averaged about 1 . 6 S and 4 . 0 NS mutations , but some individual genomes accumulated as many as 40 substitutions when an 80% hA3G target probability cutoff was used ( Figure 3 ) . This value increased to 72 when all G-nucleotides were considered . Thus , when all sites were used in simulations , mutations were more widespread and high scoring hA3G target contexts accumulated slightly fewer mutations than in simulations where a probability cutoff was used ( Figure S2 ) . This pattern reflects the fact that higher scoring sites have a larger proportion of potential stop codons resulting from G-to-A mutations . Together , these simulation results suggest that a moderate level of G-to-A mutations is tolerable in the virus sequence during its evolution without causing significant harm ( i . e . premature stops ) , although the significance of other NS mutations was not assessed in our simplified in silico model . Given the number of G-to-A mutations that could be introduced in our in silico model , it is conceivable that a low-level of hA3G-mediated mutations has been introduced into the HIV-1 sequence during its recent history and that these mutations might still be distinguishable from the more random RT-induced G-to-A mutations . To investigate this possibility , we used our in silico model ( Figure 2A ) to calculate the likelihood of each G-nucleotide context fit to be a hA3G target as outlined in Figure 4A . Briefly , we summarized the likelihood scores for each G-nucleotide context and tested for mutation type when changed into an A-nucleotide . We then calculated a ratio of NS to S sites ( Figure 4A ) . In the same way , we summarized likelihood scores for currently residing A-nucleotides , which we treated as putative ancestral G-nucleotides . Evolutionary theory predicts that the least costly mutations will have a higher chance to become fixed in the viral sequence and that deleterious mutations , i . e . mostly NS , will naturally be sorted out from the virus population due to purifying selection pressure . Conversely , among target sequences in which the target G-nucleotide has been changed to an A ( referred to here as “putative” targets ) , the NS/S ratio should increase with time . Thus , the action of A3 on ancestral virus should have left a footprint on the viral genome in the form of an increased NS/S ratio at A3G targets relative to putative ancestral targets ( Figure 4A ) . In search of such a footprint , we tested a modern HIV-1 , subtype B sequence for the score and likelihood ratios at potential hA3G target sites and found an increased NS/S ratio ( p = 0 . 07 , χ2 test ) when compared to the cognate ratios at putative ancestral sites ( Figure 4B ) . Indeed , analyses using the group M consensus sequence reconstructed from subgroup gene consensus sequences ( HIVcon ) , which most likely has eliminated some background variation , supported our finding with a higher level of significance ( p = 0 . 04 ) . Increasing the target score cutoff led to slightly higher ratios for HIV-1 , whereas a decreased cutoff had the opposite effect on both sequences ( data not shown ) . To ensure that differences in NS/S ratios between currently observed G-nucleotides and putative ancestral hA3G mutation site contexts were not artifacts from random HIV-1 RT errors , we repeated our analyses by random sampling of codons from the virus sequences while not considering context scores or likelihoods . In these controls we found no significant differences in nonsynonymous to synonymous ratios between the current and putative ancestral G-nucleotides ( Figure 4B ) . Thus , we propose that hA3G-induced mutation has had a significant influence on the HIV-1 genome in its evolutionary past . Our analysis implies that , over its evolutionary history , HIV-1 has been subjected to sublethal G-to-A mutations , which have not had sufficient effect to stop virus proliferation and transmission . An imperfectly functioning Vif protein might explain this result . Vif has become an interesting target for new anti-viral drugs [19]–[21] since inhibition of its activity could increase the chance for hA3G to more efficiently restrict virus spread . However , suboptimal blocking of Vif might also lead to an increase in sublethal hA3G-induced G-to-A mutations , and potentially more rapid appearance of drug resistance . To test this possibility , we analyzed the G-nucleotide contexts in the major HIV-1 genes gag , pro , pol and env ( Figure S2 ) . We hypothesized that combinations of HIV-1 inhibitory drugs , including a potential Vif-targeting drug , could lead to an increased selection pressure for HIV-1 , which might then benefit from low-level hA3G-mediated G-to-A mutations , allowing it to escape other drugs like protease- , RT- , integrase- and fusion-inhibitors . To test this possibility , we screened a total of about 52 , 000 mutations resulting from our in silico simulations using either 80% probability cutoffs or all G-nucleotide context probabilities ( Figure 3 ) , our analyses showed a total of 695 mutations introduced at known drug resistance sites ( Figure 5 and Figure S2 ) . However , in no case was the frequency of mutations at sites with high probability scores greatly different from the probability for all G-to-A mutations regardless of score . Thus , it seems unlikely that increased drug resistance would arise as a significant side effect of an increased , but still low-level , hA3G-induced mutation rate resulting from using Vif as an anti HIV-1 drug target . However , caution is indicated where , due to an altered selection pressure from anti-viral regimens used in combination , multiple G-to-A mutations in adjoining contexts could potentially lead to new drug resistance ( Figure S2 ) . Our current in silico model is based on previously published experimental data [9] , [10] , [14] . To test the role of hA3G more directly , we performed in vitro passaging experiments on CEM cells using a previously described HIV-1 vif mutant [18] that was previously shown to have lost the ability to inhibit hA3G , but could still inhibit hA3F; therefore G-to-A mutations were considered to result from either RT-induced errors or hA3G-mediated changes [17] , [18] . Western blotting analysis indicated that the T cell line used did not express detectable levels of hA3F ( data not shown ) , further supporting the view that hA3F did not significantly contribute to the pool of G-to-A mutations analyzed . We also analyzed another vif mutant which is unable to block the activity of hA3F but can block the activity of hA3G ( DRMR>A4 ) . This mutant replicated with kinetics that were indistinguishable from those of the wild type virus ( unpublished observations ) . Thus , we are confident that these cells express little or no hA3F . To assess the distribution of accumulated mutations , we passaged the virus sequentially for three rounds over three months , and then sequenced the vif region ( corresponding to positions 5041–5770 of pNL4-3 ) , amounting to about 166 , 000 nucleotides from 227 separate genomes . We identified a total of 585 G-to-A mutations and calculated scores and probabilities for their respective hA3G mutation site contexts . In the raw sequence data [17] , [18] , we noted that approximately 87% of the mutations were in GG dinucleotide contexts , agreeing with frequencies observed by us and others for vif-defective viruses produced from cells that exclusively express hA3G ( transfected 293T cells ) . To evaluate our in silico model predictions we simulated the same number of mutations observed in vitro into the same pNL4-3 sequence . A comparison of the observed and predicted frequencies of G-to-A mutations as a function of context probability is shown in Figure 6A , and the corresponding cumulative distributions are shown in Figure 6B . We could not detect any significant differences in the mean values ( P = 0 . 41 , χ2 test , 30 degrees of freedom ) between our in silico predictions and observations from in vitro experiments . The larger variation observed in vitro ( red confidence intervals and red outliers ) compared to the in silico results ( blue confidence intervals ) indicates that other factors such as nucleic acid structure might also influence the efficiency of hA3G-mediated mutation . However , given the similarity of the cumulative distribution , this contribution is likely to be small compared to the primary sequence likelihood used in our predictions , implying that our model is capable of accurately capturing hA3G-mediated mutagenesis during HIV replication .
Human A3F and A3G are potent antiviral factors whose activities , if not blocked by Vif , are capable of causing high levels of G-to-A mutations in newly made viral DNA . However , their role in the genesis of naturally occurring mutations is unclear . In this report , we have built an in silico model to simulate G-to-A mutations and to analyze their potential role in HIV-1 evolution and drug resistance . We have found that hA3G activity acting on prior generations of virus has left detectable footprints in the HIV-1 genome . Although there are at least ten human APOBEC members ( reviewed in [2] , [12] , [22] ) , including at least one more potential HIV-1 restriction factor , hA3H [23] , [24]; we have chosen to focus on the effects of hA3G on HIV-1 to limit our predictions and simplify our model in order to reach realistic conclusions . We also elected to include a wider mutation site context than the more commonly used GG-dinucleotide , consistent with our previous studies of nonrandom mA3 targets in endogenous nonecotropic MLV [11] , to distinguish hA3G-induced G-to-A changes from randomly introduced HIV-1 RT errors . G-to-A hypermutation in HIV-1 is clearly capable of limiting virus proliferation due to the introduction of both missense and nonsense mutations , which can be directly lethal to the virus [7] , [8] . The origin of G-to-A changes and their contribution to the bias and A-richness of the HIV-1 genome have been studied , as has the role of hA3G as a potential driver of HIV-1 evolution , although mutations introduced by RT are likely to be of greater general importance in the evolution of drug-resistant variants [25] . The hypothesis that hA3G has played an important role in HIV-1 evolution was also discussed in another recent study of HIV-1 nucleotide and codon usage bias [26] . To address the potential role of hA3G-mediated G-to-A mutations , as distinguished from random HIV-1 RT-introduced errors , in HIV-1 evolution , we analyzed individual mutation sites in hA3G target contexts , calculated from independent experimental data . To minimize bias due to purifying selection , the impaired vif gene was chosen for analysis of G-to-A mutation frequency . Another practical reason was that we needed to verify with each sequence that the Vif inactivating mutation was not reverted to regain Vif function . Thus , analyzing sequences from the Vif/Vpr region allowed us to verify the Vif-deficient phenotype . When the results obtained from this in silico analysis were compared with the accumulation of G-to-A mutations in vif ( - ) HIV-1 , passaged repeatedly in CEM cells , we saw no significant difference in mean number of mutations as a function of their context score , although the variance of this value was much greater in the latter analysis , perhaps reflecting effects of secondary structure or some other feature on the A3G mediated mutation frequency . It should be emphasized that , although we used the same total number of mutations observed in vitro for our simulation , the in silico model was otherwise completely independent of the in vitro study , yet the two did not give significantly different results . Thus , the similarity between experiment and simulation supports the accuracy of our model , as well as its suitability for estimating mutational events . Further support for our hA3G footprint observations in HIV-1 and our estimates of moderately high tolerance of G-to-A changes before introduction of deleterious stop mutations in our simulations could also be observed in a recent study of genetic variation in early founder HIV-1 virus populations from primary infections [27] . Our results imply that although hA3G-mediated G-to-A mutations likely have had a detectable effect on HIV-1 evolution that is traceable through mutational footprints Indeed , several studies of endogenous retroviruses from different genera and host species have recently shown evolutionary effects caused by A3-introduced G-to-A mutations , some of which likely contributed to provirus inactivation [11] , [28] , [29] . However , it is likely that other mutagens , including other A3 proteins , as well as HIV-1 RT are also important contributors to the nucleotide bias in the virus genome [25] , [26] . The HIV-1 Vif protein has drawn interest as a drug target since interfering with its action could theoretically allow hA3G to escape degradation and thus restrict virus proliferation [2] , [19] , [21] . However , an increased rate of G-to-A mutations compared to that from the error prone RT enzyme , as a result of an incomplete block of Vif , could potentially be used by HIV-1 to accelerate evolution of drug resistance and immune escape [30] . Variation at G-nucleotide sites is not uncommon within a HIV-1 population and polymorphism at HIV-1 drug resistance sites has been observed in single genome sequencing studies [31] , [32] . Here , we used our model to estimate the chance for HIV-1 to benefit from such increased genetic variation due to hA3G-introduced mutations . G-to-A-induced drug resistance has been described , but until now its discussion has focused on GG-dinucleotide contexts [30] , [33] . Our in silico model instead uses the extended information of the sequence context surrounding the mutation site and enabled us to calculate a probability score for each possible site . This approach enabled us to screen the entire HIV-1 genome and estimate the likelihood of mutations in the context of the dual gradients arising during reverse transcription [9] , [10] . The moderate number of hA3G-mediated G-to-A mutations at known drug resistance sites implies that the hA3G-Vif interaction can be a potential target for new antiviral drugs , with little or no concern for enhancement or appearance of resistance to other antiviral drugs . The increased rate of G-to-A mutations from hA3G does not seem overly problematic for the evolution of drug resistance in HIV-1 . However , as pointed out by Mulder et al , [34] it cannot be entirely excluded that other A3-induced changes in high scoring hA3G target sites in the vicinity of current drug targets ( Figure S2 ) not associated with drug resistance to date , might give rise to compensatory or even new resistance mutations . In that study , the authors describe a mutation , M184I in RT , which emerged in a partially impaired HIV-1 vif mutant . The conclusion that hA3G was causing this mutation was based on the dinucleotide context and the results are somewhat contradictory to our findings ( Figure 5 ) . Given that G-to-A mutations are the most common RT errors and our previous experience and support from larger mutation site contexts derived from the literature , we emphasize that such a short dinucleotide context is unlikely to be sufficient to readily distinguish bona fide hA3G induced G-to-A mutations from RT error . In conclusion , we have found evidence that low levels of APOBEC3G-induced mutagenesis have affected HIV-1 evolution , leading to enhanced rates of variation at sites with a high probability of match to the optimum context for hA3G-mediated deamination . The low levels of G-to-A mutation could allow the viruses to achieve enhanced genetic variation . However , since the predicted effect on resistance to standard antiviral drugs is likely to be small , we propose that concerns over increased resistance mutations should not impede development of HIV-1 Vif as a candidate drug target .
|
The search for new drugs to battle HIV-1 infections is a continuing struggle . APOBEC3G proteins have been shown to deaminate C-residues in HIV-1 minus strand DNA during its synthesis , resulting in G-to-A mutations in the RNA genome . The HIV-1 Vif protein has evolved to counteract APOBEC3G and thereby escape these frequently deleterious mutations , making Vif an attractive target for new drugs . However , a partial block of Vif could result in an increased although low-level HIV-1 G-to-A mutation rate . Here we investigated APOBEC3G mutation footprints in HIV-1 evolution and the potential risk for known drug resistance from sublethal G-to-A mutations . Using computer simulations , the accuracies of which were verified by infection experiments , we detected evolutionary APOBEC3G mutation footprints in the HIV-1 genome . We predict that the risk that APOBEC3G-induced G-to-A mutations will cause drug resistance is very low . We therefore propose that concerns for increased resistance to other antiviral drugs should not prevent Vif from being considered a suitable target for development of new drugs .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"virology/virus",
"evolution",
"and",
"symbiosis",
"virology/immunodeficiency",
"viruses",
"virology",
"infectious",
"diseases/hiv",
"infection",
"and",
"aids",
"virology/host",
"antiviral",
"responses"
] |
2009
|
Likely Role of APOBEC3G-Mediated G-to-A Mutations in HIV-1 Evolution and Drug Resistance
|
Transcriptional activity has been shown to relate to the organization of chromosomes in the eukaryotic nucleus and in the bacterial nucleoid . In particular , highly transcribed genes , RNA polymerases and transcription factors gather into discrete spatial foci called transcription factories . However , the mechanisms underlying the formation of these foci and the resulting topological order of the chromosome remain to be elucidated . Here we consider a thermodynamic framework based on a worm-like chain model of chromosomes where sparse designated sites along the DNA are able to interact whenever they are spatially close by . This is motivated by recurrent evidence that there exist physical interactions between genes that operate together . Three important results come out of this simple framework . First , the resulting formation of transcription foci can be viewed as a micro-phase separation of the interacting sites from the rest of the DNA . In this respect , a thermodynamic analysis suggests transcription factors to be appropriate candidates for mediating the physical interactions between genes . Next , numerical simulations of the polymer reveal a rich variety of phases that are associated with different topological orderings , each providing a way to increase the local concentrations of the interacting sites . Finally , the numerical results show that both one-dimensional clustering and periodic location of the binding sites along the DNA , which have been observed in several organisms , make the spatial co-localization of multiple families of genes particularly efficient .
Statistical properties of long DNA chains in good solvents are accurately described by worm-like chain ( WLC ) models [28] . These types of models provide a coarse-grained description of protein-coated DNA ( e . g . the eukaryotic chromatin ) . They are simple enough to allow some analytical treatment and to be investigated numerically . They include the typical elastic behavior of DNA , which has been measured in vitro and in vitro . More precisely , the WLC model is defined by a bending energy where is the variation of the tangent vector along the curvilinear abscissa of the polymer , is the bending modulus , and is the total length of the polymer . In our study , we further take into account the short range electrostatic repulsion of DNA ( DNA is negatively charged ) . Due to the screening of the charges in vivo , it is commonplace to model this repulsion as a hard-core potential . Our framework therefore consists of a self-avoiding WLC with a hard-core radius . The persistence length along the polymer is defined as the distance beyond which the WLC loses most of its orientational order – see Fig . 1 . For an infinitely thin chain ( is then the only energy ) , one has where is the temperature in Kelvin and is the Boltzmann constant . For naked DNA , nm; moreover typical in vivo ionic conditions lead to nm [29] , though may appear larger or smaller due to the presence of DNA bound proteins such as histone like proteins . In the case of eukaryotes , one can model the 30 nm chromatin fiber by taking ; can vary between 50 and 250 nm , depending on the compaction level of the chromatin . Within this framework , genes along the DNA are associated with specific sites on the polymer ( Fig . 1 ) . Part of these genes will participate to the co-localization process . In this regard , several possible scenarios have been proposed ( see the introduction ) . Here , we investigate the effect of thermodynamic interactions ( e . g . van der Walls or ionic ) between proteins and DNA and discuss whether these interactions can lead to a well coordinated self-organization of the chromosome . We therefore do not consider proteinic complex assemblings that require energy consumption nor possible active forces – e . g . induced by molecular motors – that would drive chromosome loci to the transcription factories . Finally , the binding of proteins on DNA is treated implicitly , that is , two chromosome loci that can be bridged by a proteinic complex interact according to a short-range attractive potential . Here is the step function that is if and otherwise , is the interaction range , and is the strength of the potential . This interaction mimics a free energy term resulting from the bridging of the chromosome , either by bivalent TFs such as the Lac repressor [8] , or by TF multimerization such as in the phage [26]; similarly , tethering may be mediated by the transcription factories , or more generally by RNA polymerase/TF complexes , which occur for instance during the transcriptional activation of some bacterial genes [30]; values of therefore lie between several nanometers and several tens of nanometers . The free energy gain comprises that due to protein-DNA binding and , when multimerization comes into play , that due to protein complex formation . In any case , free energies ( i . e . , ) are expected to be a few kcal/mol ( and thus a few ) [25] , [26] . Our coarse-graining procedure allows to tackle , within the same formalism , different mechanisms that may lead a gene to be an interacting site . For instance , our model can mimic the effect of the chromatin condensation ( heterochromatin ) , which can prevent a site from participating to the interaction just by hiding it or making it unaccessible . A more realistic modeling of chromosome structuration would include heterogeneities in the interaction between the sites ( different for different pairs of sites ) and also the explicit presence of solvent molecules . However , our goal here is to provide a plausible general picture for the formation of transcription factories that can be cast within a formalism as simple as possible . Overall , our framework consists of a self-avoiding WLC along which specific sites are distributed sparsely and are able to interact ( Fig . 1 ) . In this context , we define as the mean distance between two successive interacting sites along the DNA . We also define the capacity of a site as the number of other sites it can interact with simultaneously . For the results shown here , we take for simplicity no limit on the capacity . The maximum number of partners of a site , hereafter referred to as , will be limited only by the steric constraint: one cannot pack more than some maximum number of sites within a given distance of a point . Notice that the possibility of multiple interactions is compatible with the fact that gene regulatory regions frequently have several TF binding sites . Moreover , large protein complexes , which are likely to appear around transcription factories , should favor the simultaneous interaction of several binding sites . Our results can be divided into three parts . First , we show that transcription factories can be viewed as the result of a micro-structuration mechanism , which is an archetype of a self-organizing process . In particular , our calculation highlights the range of parameters for which the micro-structuration is expected . Next , we use numerical simulations to address the topological ordering of DNA around the transcription factories . Finally , we tackle the problem of forming transcription factories in the presence of different families of interacting sites , i . e . families corresponding to different regulatory properties .
Before dealing with the mechanisms that are responsible for the formation of discrete foci , we quickly recall the basic phenomenology of a self-attracting and self-avoiding WLC . Within the framework of our model , this corresponds to considering a dense distribution of interacting sites along the DNA , so the interacting sites are close-by along the whole DNA . Such WLCs have been extensively studied for more than forty years [31] , [32] . Depending on the values of the parameters , they mainly lead to three typical conformations , which are also known to arise for chromosomes in vitro and in vivo [33] , [34] . First , in the absence of the self-attracting interaction , the WLC behaves as a self-avoiding random walk , at least on length scales larger than the persistence length; this leads to the so-called “swollen” state – in the physics of an isolated polymer chain , some phases may arise only when the polymer is short; as suggested in [35] , we refer to these pseudo-phases as “states” . Second , introduce an attractive interaction . For a sufficiently strong attraction , the polymer goes to one of two possible compact conformations . The densest conformation is obtained by having the polymer wind many times around a circle , forming in effect a kind of torus; accordingly , this has been called the “toroidal” state [32] . For a weaker interaction , less dense and less ordered conformations arise , the so-called “globules” where the polymer forms a ball but otherwise seems rather random . Which of these macroscopic states – swollen , toroidal or globule – describes the equilibrium state depends on the parameters , the two most important ones being the attractive force and the polymer stiffness [35] . Coming back to our system consisting of a single chain with sparse interacting sites , its peculiarity is that only a few designated sites of the chain are subject to the attraction; this means that a further organization of the chain on smaller length scales can arise as now explained . Suppose that the interacting sites are sparsely distributed along the polymer . Starting in one of the compact states , the energy can be enhanced by local rearrangements , keeping the polymer compactness roughly unchanged . At a coarse-grained level , one can focus on the local density of the interacting sites in three-dimensional space . In a random conformation , the density will be uniform . By contrast , after local rearrangements , the density will vary , leading to clumping in some areas , and voids in others . In essence , a uniform density is energetically unstable , and so the system will spontaneously structure so as to form regions of high and low densities of interacting sites ( Fig . 2 ) . This leads to a micro-phase separation between interacting and non-interacting sites , which is reminiscent of what is observed in block co-polymers [32] . In the following , we investigate in detail this micro-structuration , tackling the problem in two ways . First , we use a mean-field theory of polymer physics . This allows us to qualitatively capture the transition between the homogeneous states with a uniform distribution of interacting sites and the micro-structured states with a spatially modulated distribution of interacting sites . Next , we use Monte-Carlo simulations to both validate our analytical results and to further study the DNA conformations around the foci . Within the scope of chromosome structuration via the bridging of co-regulated genes , the macroscopic state diagram for not too strong attractive forces is limited to two states: the swollen state and the micro-structured globule – see Fig . 3 . Generally speaking , the micro-structured globule tends to be favored thermodynamically over the homogeneous globule for interacting sites that are sparsely distributed along the WLC; the homogeneous density of binding sites is unstable to a modulation , at least if the capacity of sites is not too small . In this situation , the number of interacting sites lying within the foci of the micro-structured globule ( ) , which is kept fixed in our calculation for the sake of simplicity , determines the position of the transition between the two states . We now present the principles of the underlying calculation , emphasizing the crucial parameters that determine the balance between the states . This allows us to discuss the mechanisms responsible for the shift between the states . The best way to determine the thermodynamically favored state is to compute the free energy of each state as a function of the model parameters , which is explicitly done in section 3 in Text S1; the state with the lowest free energy is the favored one . In the following , for the sake of simplicity , we do not tackle the issue of the toroidal phase , considering only the micro-structured globule , the swollen state and the homogeneous globule . Ignoring their internal structure , these isotropic states look like balls . As a consequence , they can be characterized by a radius and a free energy . In the most general case , the free energy can be decomposed into four terms: ( 1 ) is the free energy due to the attractive potential between the interacting sites . is the contribution from the bending energy . is the free energy cost due to the excluded volume of the polymer within an area of extension , which stems from the repulsion of the hard-core monomers constituting the polymer . is the entropy related to the number of polymer configurations that are compatible with a radius [32] . For a given type of organization ( e . g . a micro-structured globule ) , the free energy calculation consists in first determining the that minimizes the free energy . Then , is plugged into the free energy relation Eq . ( 1 ) , which gives the corresponding free energy of the state , that is . One must compare the free energies of each state . The explicit dependence on the radius of each term is calculated using a standard mean-field single chain polymer theory , also known as Flory theory [32] , focusing on the bulk contribution to the free energies ( large ) . In the following , we skip the technical details and give the final results of the calculation , as well as its interpretation . For more details on the derivation , we refer the reader to section 3 in Text S1 . In the case of a sparse distribution of interacting sites , the position of the sites may be crucial , as we shall see in the last section . To simplify our discussion , we therefore consider , in a first stage , sites that are regularly spaced by along the DNA . Our WLC offers a single framework to discuss the formation of transcription factories both in bacteria and in eukaryotes . Within the context of transcriptional regulation , genes participating to the same transcription factories are believed to participate to specific cellular functions . can therefore be evaluated as the typical distance between two consecutive genes that are co-regulated by the same TF or that are known to participate to the same function . As a consequence , is expected to be larger than the distance separating two genes , e . g . 1 kbps in bacteria ( i . e . , 300 nm ) , and 100 kbps in mammals ( 600 nm ) – we have used 150 bps/nm for the chromatin fiber [36] . This leads to factors in relation ( 2 ) . Hence , within the scope of our model , for biologically relevant values of and , the homogeneous globule state ( with a uniform distribution of actively transcribed genes ) is thermodynamically unlikely both in bacteria and eukaryotes . As far as the micro-structured globule with discrete foci is concerned , in eukaryotes one can approximate as the typical number of active RNA polymerases within one transcription factory , i . e . , [37] . By considering nm and Mbps ( 6 mm ) , which corresponds to the mean distance between two consecutive genes regulated by a TF in the human genome [38] , one finds . The regulatory regions of eukaryotic genes often have several TF binding sites of the same type , which can be interpreted as . Hence , a bridging induced by TFs ( with binding energies of several 's per TF ) , or induced by a proteinic complex involving TFs , is sufficient to induce the formation of transcription factories according to the above micro-phase separation ( ) . Moreover , given the parameters of chromatin , the values of and lie in a range that allows to switch between a state with discrete foci and the swollen state – see Fig . 3 . This suggests that the micro-phase structuration is also a possible mechanism for fine tuning the global genetic regulation of a cell . In bacteria , an interesting case concerns the formation of the putative transcription factories during the transcription of rRNA operons [5] . In this situation , 7 operons scattered along 2 Mbps have to be co-localized . nm then leads to . In the same way , both co-regulated genes [15] and genes that are thought to be functionally related [17] have been shown to be periodically spaced according to a 100 kbps period . Supposing these genes are co-localized by groups of at least ten , this leads again to . Hence , in bacteria , if one considers one single binding site per gene , large binding energies are required for the formation of transcription factories . However , this should be balanced by the overall negative supercoiling of bacterial DNA which is beyond the scope of our model . Together with the action of nucleoid-associated proteins ( e . g . histone like proteins such as Fis , H-NS or HU ) , this effect would tend to condense the chromosome and hence to dampen consequences of thermal fluctuations . Numerical simulations of polymer models are useful to investigate the principles of chromosome organization within space [19] , [39] , [40] . In this respect , simulations of our self-avoiding WLC ( see Methods for details ) confirm that gene foci arise for persistence lengths , binding free energies and inter-gene distances that are typical of bacteria and eukaryotes ( see Fig . 4 for two such examples ) . Simulations are also useful to see how the foci organize in 3-dimensional space . Indeed , a priori , foci may form regular lattices , random lattices , or they may wander with time . In this respect , our results suggest a rich variety of equilibrium conformations that depend on the parameters of the system . However , from a computational point of view , we are not able to investigate the thermodynamic state diagram when the DNA chain becomes relatively large because the different metastable states last the whole time window of the simulation once they are formed; in particular , we do not see switches between the states as would arise in a situation of co-existence . Thus we are limited to considering the most likely structures that form when starting from a random coil ( swollen ) configuration as we progressively increase the value from an initial zero value . Note that from a biological point of view , such metastable states may be just as relevant as the true equilibrium states . The resulting structures can be divided into three main groups , as we now describe . Recent experiments in monkey Cos7 cells have shown that different transcription factories recruit different genes depending on their promoter type [41] . In the same spirit , one may hypothesize that genes regulated by the same TFs preferentially co-localize in space [6] , [7] , [14] . This would explain for instance , in yeast , the tendency of co-regulated genes to be clustered along the chromosomes [14] , [42] , [43] . A somewhat analogous issue arises in bacteria: one often finds that a TF coding gene , the binding site of that TF , and the corresponding regulated gene ( s ) are all close-by along the DNA ( see [16] and references therein ) . This is thought to optimize the three-dimensional targeting process of the TF toward its binding site because , in bacteria , protein translation occurs close to the coding gene . Accordingly , space co-localization of distant binding sites for each TF type may very well occur since it is a natural way to make three-dimensional targeting and assembly of complexes more efficient . The investigation of gene positions in E . coli and yeast suggests that in these organisms a near periodic arrangement on the DNA of co-regulated genes may be at the base of a good 3-dimensional spatial co-localization [15] , [16] . However , there are hundreds of TF types both in bacteria and yeast , and thousands in higher eukaryotes so that the satisfaction of all the separate co-localization constraints may be a hard problem for the organism to solve . We have used our framework to numerically model the spatial co-localization process when different types of TFs regulate a large number of genes . Specifically , we have types of binding sites , where two binding sites interact only if they are of the same type . The way these sites ( and their types ) are positioned along the chain can affect the way the different foci form . We have therefore compared the co-localization process using four kinds of positioning of these binding sites , namely: i ) sites ordered – and thus clustered – according to their types ii ) randomly distributed sites and types; iii ) periodically distributed sites and types; iv ) sites that are spaced according to random multiples of , hereafter referred to as random periodic: there is approximate periodicity in the site positions while the site types are taken to be completely random ( see Figure S4 for an illustrative explanation ) . Situation corresponds to the one-dimensional clustering of nearby binding sites whereas situations to correspond to the interaction of binding sites that can be distant from each other . In particular , situation is useful to determine whether regularity in the site types is necessary for co-localization , even if there is some regularity in the site positions along the DNA . In this context , the mean distance ( measured along the chain ) between two consecutive sites regardless of their type is a useful additional parameter to characterize the one-dimensional site properties along the DNA . To simplify our study , we take a number of interacting sites that is roughly the same for each site type so that , being the mean distance between two sites of the same type .
Within a fairly general framework , we investigated the topological organization of a model chromosome . Using an effective attractive potential between selected genes on a DNA chain , we found that these could organize into discrete foci , with the DNA visiting the foci in several topologically distinguishable ways . The foci are composed of genes that can be far away from each other along the DNA , which is supported by the recent observation of numerous Mbps-range DNA loops [6] , [7] . Of course , in vivo , numerous obstacles might prevent chromosomes from achieving the conformations we predict: supercoiling , chromatin remodeling and confinement introduce other interactions that may dominate for some parameter values . Another point is that we have focused on equilibrium conformations whereas in reality cellular processes operate away from equilibrium . However , a pure equilibrium approach is useful because it shows the natural organizational trend of the system . Several conclusions transpire from our framework . First , in bacteria and eukaryotes , the formation of transcription factories may be related to a self-organizing process akin to the folding transition of single polymer chains . The underlying thermodynamic mechanism is a spatial micro-phase separation driven by regions of DNA where genes are subject to similar transcriptional regulation . In effect , due to the very nature of the self-avoiding DNA chain , all genes cannot cluster together to enhance transcription rates; instead , discrete foci must form in space . Our results therefore confirm that self-organization may play a crucial role in the structuring of chromosomes [10] , [44] . Interestingly , the interaction strength needed between distant sites along the DNA in order to induce the micro-structuration is compatible with the binding of TFs to DNA . The bridging can be achieved via a bivalent TF , or more generally through the formation of large protein complexes , e . g . by tethering the DNA-bound TF to ongoing transcription factories . This corroborates TFs as possible entities for mediating the effective attractive potential; our model therefore predicts a 3D co-localization of co-regulated genes . In eukaryotes , this can be tested by a combination of 3D fluorescence in situ hybridizations ( FISH ) and chromosome conformation capture techniques [45] as exemplified in [6] , [7] , [45] , [46] . In bacteria , this can be tested by using the site-specific recombination system of the bacteriophage [13] . Furthermore , as illustrated by Eq . ( 3 ) , the number of co-regulated genes that can be co-localized within the same focus depends both on the number of TF binding sites per gene and on the binding energies . This leads to the prediction that the presence of aptamers which can compete with TFs for binding to cognate DNA sites will lead to smaller transcription factories or even none at all . Second , using numerical simulations of our model , we have shown that the topology of the DNA conformations fall into several classes according to the way the foci are visited , and that two of these classes had been previously hypothesized on the basis of biological evidence . For instance , starting from a toroidal organization of DNA which has been observed in some organisms [34] , if the interacting sites that stabilize this structure become less dense , there should be a micro-phase separation whereby distinct foci appear along the ring , which fits the solenoidal model proposed in [14] . As interacting site density decreases further , rosettes may form as proposed in [18] . Or the DNA may successively visit the different foci in a random fashion , corresponding to our “traveling chain” topology . Third , which topological ordering arises generally depends on the way the binding sites are positioned along the one-dimensional DNA . We find that some periodic regularities and some clustering in the positioning of co-regulated genes , as observed respectively in [15] , [16] and in e . g . [42] , [43] , strongly favor the formation of well-separated foci with a homogeneous size and content , and disfavor the presence of genes outside of the foci . To this end , we considered the possibility of having multiple types of protein binding sites , thought to be associated with different transcription factor families or gene functions . We found that having periodically-positioned targets of multiple TFs favored the solenoidal topology whereas the necklace of rosettes topology was favored if groups of genes were one-dimensionally clustered along the DNA ( Fig . 8 ) .
Numerical simulations of the continuous self-avoiding WLC model are based on an off-lattice semi-flexible polymer composed of jointed cylinders of radius and length ( Figure S1 ) . The cylinders are impenetrable ( hard-core interactions ) and two consecutive cylinders , that form a bending angle contribute a bending energy to the total energy . The solvent is implicit , it is not treated explicitly . Interacting sites are taken to be located at the joints between two consecutive cylinders; a joint can contain or not an interacting site . They interact via a uniform short range square potential of depth and interaction range ( Figure S1 ) . Thus , if two non-consecutive interacting sites and can interact , they contribute an energy if the distance between them is less than . As a result , the total energy of the system reads: ( 4 ) where means that the non-consecutive interacting sites and are able to interact . is the step function that is equal to if and is otherwise . is the bending modulus of the polymer; it depends on the type of polymer ( DNA or chromatin ) that is described . To have results that are insensitive to the discrete nature of the polymer representation , one should use a segment length that is a small fraction of the persistence length; in all our simulations , we take this factor to be one fifth . Note also that it is best to work off-lattice as lattice anisotropy is known to induce geometrical artifacts that could spoil the interpretation of the results . The persistence length and the radius of our polymer representation of DNA depend on the type of organism to be modeled . In the limit of an infinitely thin polymer ( ) , the persistence length is related to the bending modulus via . In the case of the self-avoiding polymers presented in this work , this relation holds well ( data not shown ) so that the bending energy is enough to define . To sample the state space of our polymer model , we use standard Monte Carlo procedures with the Metropolis accept/rejection rule , which guarantees reaching thermodynamic equilibrium if ergodicity is not broken . The Monte Carlo method consists in 1 ) picking at random two joints ( a joint being the point where two consecutive cylinders coincide ) , and 2 ) applying a 3-dimensional rotation around the axis that passes through the two joints according to a random angle in . Here we take a relatively small value , ( at larger values the acceptance rate goes down ) . The largest timescale for the conformational relaxation of a single coiled polymer scales as [31] . From a numerical point of view , this results in a relaxation times that scales as for local microscopic evolution rules , i . e . , where only a finite number of cylinders are updated at each time step . In our Monte Carlo simulation , time is counted in number of sweeps , one sweep consisting of attempts to rotate part of the chain; also monomers are updated during one single rotation . This leads to a relaxation time that scales as . Nevertheless , we still need roughly computer operations to thermalize the system in the regime of interest where interaction effects are dominant . This prevents the current method from scaling up to very long chromosomes , although we can deal with interesting systems . We present simulations with up to cylinders; this corresponds to 50 kbps in the case of naked DNA and 5 Mbps in the case of the chromatin fiber . In this case , we are not able to sample the equilibrium space of the condensed polymer because the different metastable states are very stable . In situations of slow temporal evolution , defining a steady state may be a tricky operation . For the parameters we used , our results suggest to consider as steady a state that lasts for more than sweeps . For random positions of the interacting sites , the folding time can exceed sweeps . Notice then that sweeps correspond to Monte-Carlo steps for ( the largest size we report here ) .
|
The good operation of cells relies on a coordination between chromosome structure and genetic regulation which is yet to be understood . This can be seen in particular from the transcription machinery: in some eukaryotes and bacteria , transcription of highly active genes occurs within discrete foci called transcription factories , where RNA polymerases , transcription factors and their target genes co-localize . The mechanisms underlying the formation of these foci and the resulting topological structure of the chromosome remain to be elucidated . Here , we propose a thermodynamic framework based on a polymer description of DNA in which genes effectively interact through attractive forces in physical space . The formation of transcription foci then corresponds to a self-organizing process whereby the interacting genes and the non-interacting DNA form two phases that tend to separate . Numerical simulations of the model unveil a rich zoology of the topological ordering of DNA around the foci and show that regularities in the positions of the interacting genes make the spatial co-localization of multiple families of genes particularly efficient . Experimental testing of the predictions of our model should shed new light on the relation between transcriptional regulation and cellular conformations of chromosomes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biophysics/macromolecular",
"assemblies",
"and",
"machines",
"physics/interdisciplinary",
"physics",
"cell",
"biology/nuclear",
"structure",
"and",
"function",
"physics/condensed",
"matter",
"biophysics/theory",
"and",
"simulation",
"computational",
"biology/molecular",
"dynamics",
"computer",
"science/numerical",
"analysis",
"and",
"theoretical",
"computing",
"biophysics",
"biophysics/transcription",
"and",
"translation",
"computational",
"biology/systems",
"biology"
] |
2010
|
Spatial and Topological Organization of DNA Chains Induced by Gene Co-localization
|
Characterizing interactions between drugs is important to avoid potentially harmful combinations , to reduce off-target effects of treatments and to fight antibiotic resistant pathogens , among others . Here we present a network inference algorithm to predict uncharacterized drug-drug interactions . Our algorithm takes , as its only input , sets of previously reported interactions , and does not require any pharmacological or biochemical information about the drugs , their targets or their mechanisms of action . Because the models we use are abstract , our approach can deal with adverse interactions , synergistic/antagonistic/suppressing interactions , or any other type of drug interaction . We show that our method is able to accurately predict interactions , both in exhaustive pairwise interaction data between small sets of drugs , and in large-scale databases . We also demonstrate that our algorithm can be used efficiently to discover interactions of new drugs as part of the drug discovery process .
Understanding interactions between drugs is becoming increasingly important . A recent large-scale study of older adults ( ages 57–85 ) in the U . S . found that 29% of them use five or more prescription medications concurrently , and that as many as 4% may be at risk of having a major adverse drug-drug interaction [1] . For this reason , the evaluation of drug interactions is “an integral part of drug development and regulatory review prior to its market approval” [2] , and institutions like the FDA put much effort in developing guidelines for in vitro and in vivo studies , as well as for developing in silico models and methods . Potentially beneficial effects of drug interactions , on the other hand , are equally important . Indeed , some drugs show synergistic effects against their targets , which not only increases the efficacy of treatments but may also improve the selectivity and reduce off-target effects [3] . Antagonistic interactions can be used to study the mechanisms of action of drugs [4] , and even suppressing interactions between drugs , in which one drug inhibits the action of the other , have been found to be potentially very relevant in the fight against antibiotic-resistant pathogens [5] . More broadly , it is becoming increasingly clear that drug interactions leading to network effects at a systems level are the norm in pharmacology , rather than the exception [6]–[11] . According to some , these network effects may even be at the root of the dismal results of attempts to develop single-target drugs , and of the simultaneous decline of drug development productivity [7] . Therefore , network pharmacology is emerging as a new paradigm in drug discovery . However , despite the conceptual appeal of abstract network approaches to drug development , one may argue that the contributions of network analysis have so far been relatively modest . Indeed , most of these contributions have been related to pointing out network properties that make certain proteins more likely to be good targets [8] , for example connector versus non-connector enzymes [12] , [13] , or central versus peripheral proteins [11] . These contributions notwithstanding , there is little in the form of actual , concrete , examples where network analysis has resulted in a clear application to the discovery of new drugs or to the study of the effects of existing drugs . Here we present one such application . In particular , we use the information that is encoded in networks of reported drug interactions to predict uncharacterized interactions . Because the models we use are abstract , our approach can deal with adverse interactions as well as synergistic/antagonistic/suppressing interactions or any other type of drug interaction . We show that our method is able to accurately predict drug interactions , and that it can be used efficiently to discover interactions of new drugs as part of the drug discovery process .
For specific drug pairs , interactions can be predicted in silico from mechanistic or flux balance models of the pathways and processes in which their targets are involved [6] , [14] . However , this approach is difficult to generalize and is , therefore , inappropriate for large-scale identification of interactions and for the identification of interactions between drugs whose mechanisms are not fully understood . Another approach is to use statistical models based on molecular and pharmacological data [15] but , again , such data is not always available . Finally , there are mechanism-independent methods to predict multidrug interactions based on maximum entropy approaches , but these require knowledge of pair interactions [16] , which is what we aim to uncover here . As in other biological problems , network theory [11] , [17] , [18] provides a useful , although abstract , alternative to mechanistic and molecular modeling . In a network representation of drug interactions , each node represents a drug and each link represents an interaction between the corresponding pair of drugs . Interactions of different types ( for example , synergistic versus antagonistic ) are represented by links of different types ( Fig . 1A ) . Drug interaction networks contain explicit information about the interactions that are known , but also about implicit information about interactions that have never been tested; the question we are concerned with is how to extract this information from the network . Here , we present a network-based approach to predict an interaction between drugs and from a network of known drug interactions ( which includes and but no explicit information about their interaction ) . Our approach deals rigorously with the information contained in the network by means of Bayesian model averaging [19] ( Methods ) . The approach is completely unsupervised and parameter-free . Within our Bayesian model averaging approach , the only relevant modeling question is what family of models can accurately describe the network of drug interactions . In this regard , it is well established that pairwise drug interactions are largely determined by the cellular functions targeted by the drugs [20]–[22] . In network terms , this means that the interaction is determined by the cellular functions and of and , respectively; in other words , nodes can be partitioned into groups ( by cellular function ) such that the interaction between any pair of nodes depends only on the groups to which they belong ( Fig . 1B–C ) . Stochastic block models are a family of network models that mathematically formalize the idea of group-dependent interactions 23–25 . Although originally proposed in the context of social interactions , stochastic block models are increasingly used to describe the structure of complex networks in general [19] , [26] and for network inference [19] ( Methods ) . Again , after this choice of plausible models the resulting algorithm is completely unsupervised and parameter-free ( Methods ) . To benchmark the performance of our algorithm , we consider two alternative heuristic approaches . The first benchmark is based on the idea that similar drugs have similar interactions . In this spirit , we set , where ( respectively , ) is a drug whose known interactions are as similar as possible to those of ( ) , and is a known interaction ( Methods ) . Second , we consider an approach based on the Prism algorithm , which was developed to identify groups of drugs ( or genes ) with similar interactions to other drugs [20] , [27] . Instead of averaging over all possible partitions of drugs into groups as done in our Bayesian model averaging approach , we take the partition proposed by Prism and use that partition to make the prediction ( Methods ) . Additionally , we consider as a baseline the simplest possible algorithm for predicting , which is to use the overall rate of each interaction type in the network . For example , if 60% of known interactions in a network are synergistic ( ) and 40% are antagonistic ( ) , then we set with 60% probability and with 40% probability . This baseline captures the fact that it is harder to make a prediction when the ratio of interactions is 60/40 than when the ratio is , for example , 95/5 . We start by testing the algorithms described above against two experiments in which all pairwise interactions between a small set of drugs were exhaustively tested [20] , [28] . In the first experiment , Yeh and coworkers tested the effect of all pairwise combinations of 21 antibiotics on E . coli's growth [20] . They classified each interaction as synergistic , additive , antagonistic or suppressing . In the second experiment , Cokol and coworkers studied the effect of all pairwise combinations of 13 anti-fungal drugs on the growth of S . cerevisiae [28] . They classified interactions as synergistic , additive or antagonistic ( except for some interactions that were unresolved ) . To study the performance of the algorithms , we simulate situations in which not all pairwise interactions are known . In particular , we simulate a situation in which only a fraction of all interactions are observed , and then try to predict the unobserved interactions ( repeated random sub-sampling validation ) . In each case , we measure the fraction of predictions that are exactly correct ( exact classification ) , as well as the fraction of predictions that deviate from the experimental observation by at most one level ( classification ) . For example , miss-predicting a synergistic interaction as additive is considered correct by the classification metric , but miss-predicting a synergistic as antagonistic or suppressing ( or vice versa ) , or an additive as suppressing ( or vice versa ) is considered incorrect . In Fig . 2 we show the results of the validation . As expected , the stochastic block model , the neighbor-based and the Prism-baed predictions have accuracies well above the baseline , even when as many as 80% of the interactions are unobserved . In the majority of cases , the stochastic block model is significantly and consistently more accurate than the neighbor-based and the Prism-based predictions with one exception: when the fraction of observed interactions is high ( % ) in the Cokol dataset , in which the neighbor-based prediction is best . Note that as soon as the number of interaction types grows ( from 3 in Cokol to 4 in Yeh ) or the fraction of observed interactions decreases , the stochastic block model becomes more accurate . Moreover , even when the neighbor-based exact predictions are more accurate , ±1 predictions are always more accurate for the stochastic block model . Although the absolute differences of prediction accuracy between the stochastic block model and the neighbor-based approach may seem modest ( typically , between 5 and 10 percent points ) , it is important to note that relative to the baseline the improvements are quite major ( Fig . 2E–F ) . Indeed , when the fraction of observed interactions is 50% , the stochastic block model represents a 29% and a 63% improvement ( for the Cokol and Yeh datasets , respectively ) in exact classifications over the neighbor-based approach , and a 55% and 66% over the Prism-based approach ( always , with respect to the baseline ) . When we only observe 20% of the interactions , the relative improvements are 126% and 133% over neighbor-based predictions , and 61% and 154% over Prism-based predictions . Next , we test our algorithm against the existence of adverse drug interactions in two drug interaction databases: the database available through the web site Drugs . com and the DrugBank database [29] , [30] . For the Drugs . com database , we restrict our analysis to major adverse interactions between generic drugs; for the DrugBank , we consider all interactions . We consider two snapshots of each of the databases . For the Drugs . com database , we collected the first snapshot in May 10 , 2010 , and the second one in February 22 , 2012 . A total of 1 , 518 drugs are listed in both snapshots . There are 32 , 074 drug interactions present in both instances of the network; novel interactions present in the 2012 dataset but not in the 2010 dataset , and spurious interactions present in the 2010 dataset but not present in the 2012 dataset . For the DrugBank dataset , the first snapshot corresponds to January 2009 , and the second to April 2012 . A total of 1 , 012 drugs are listed in both snapshots; there are 9 , 113 drug interactions present in both instances of the network , with and . We evaluate to what extent could our network algorithms have predicted which interactions needed to be added to each of the first snapshots ( that is , to what extent can the algorithms uncover novel interactions ) , and which ones needed to be removed ( that is , to what extent can they detect spurious interactions ) . As we show in Fig . 3 , the algorithm based on stochastic block models is able to accurately uncover spurious and , especially , novel interactions . In contrast , neighbor-based and Prism-based predictions perform only marginally better than the baseline . First , we measure the area under the receiver operating characteristic ( AUROC ) curve ( Fig . 3A–B ) [31] . In the case of uncovering novel interactions , the AUROC gives the probability that an interaction randomly chosen from those that were added to the first snapshot has a higher score than one randomly chosen from the set of interactions that were never added to the network . For the Drugs . com database , we find this probability to be 0 . 87 for the stochastic block model , 0 . 53 for neighbor-based predictions , and 0 . 52 for Prism-based predictions . For the DrugBank dataset , these probabilities are 0 . 71 , 0 . 52 and 0 . 53 , respectively . Similarly , when dealing with spurious interactions , the AUROC gives the probability that an interaction randomly chosen from those that were removed from the 2010 snapshot has a lower score than one randomly chosen from the set of interactions that were not removed from the network . For the Drugs . com database , we find this probability to be 0 . 73 for the stochastic block model , 0 . 51 for neighbor-based predictions , and 0 . 45 for Prism-based predictions . For the DrugBank dataset , these probabilities are 0 . 61 , 0 . 50 and 0 . 50 , respectively . It is also interesting to analyze the sensitivity-specificity curves ( Fig . 3C–F ) . Consider first the results for the Drugs . com database . For the most pressing case of uncovering previously unreported major drug interactions ( Fig . 3C ) , we find that at 95% sensitivity , the stochastic block model has a specificity of 62% , that is , that we could have built , in 2010 , a list of potential interactions containing 95% of the interactions that were actually added to the database , and excluding 62% of those that were never added . Conversely , at 95% specificity we obtain a sensitivity of 45% , that is , a list containing only 5% of the interactions that were never added to the network would have included close to half of all the interactions that were actually added to the database . While results for spurious interactions and for the DrugBank dataset are more modest , our method , unlike the neighbor-based or the Prism-based algorithms , has significant predictive power in all the cases we study . Finally , we demonstrate that our algorithm can be used to discover interactions of novel drugs as part of the drug discovery process . In particular , consider a lab that has developed a new drug which is known to have a harmful interaction with another drug . Ideally , the lab wants to identify all other drugs that also have harmful interactions with . Since in principle there are as many potential interactions as drugs in the market ( more than 1 , 000 , according to the Drugbank and Drugs . com databases ) , it would be extremely costly to test all possible interactions experimentally . Considering that the typical drug interacts with approximately 20–40 other drugs ( in DrugBank and Drugs . com , respectively ) , random testing for interactions would require 35–55 experiments to uncover a single harmful interaction . Lacking any knowledge about ( other than its interaction with ) , our algorithm can guide experiments by identifying those drugs that are most likely to interact with . In particular , we could use the stochastic block model inference approach to predict the most likely interaction , test it in the lab , and iterate the process adding , at each iteration , whatever interaction information the lab assay gave . To test whether such an approach would work in practice , we have simulated the discovery of two drugs whose interactions are in fact known and reported in the 2012 snapshot of DrugBank—acetophenazine and cinacalcet ( these drugs were selected randomly among those with 10 to 20 interactions ) . For each of these drugs , we proceed exactly as if no data were available in the database except for one seed interaction , which we also choose at random . From the seed interaction and interaction data for all drugs other than , we use the stochastic block model approach to infer the next most likely interaction of , check if the interaction truly exist , add this information to the network , and iterate . As we show in Fig . 4 , the results are very promising . For acetophenazine , the 16 iterations we carry out are enough to discover 11 of the 15 interactions that are reported in DrugBank . For cinacalcet , we are able to uncover 8 of the 12 reported interactions . As mentioned above , these numbers need to be compared with the approximately 55 experiments that would be necessary to uncover a single interaction without any guidance .
There is a pressing need to elucidate and understand interactions between drugs . With thousands of drugs in the market , and hundreds or thousands being tested and developed , it is clear that we cannot rely only on experimental assays to uncover interactions . Therefore , we need to develop computational data-mining methods to guide experimental analysis . There are many possible approaches to predict drug interactions computationally . One is to mine patient data that are collected as part of post-marketing surveillance . However , this approach is problematic because of confounding factors that may not be properly accounted for in existing reporting systems [32] . Another approach is to use models based on molecular and pharmacological data [15] . Our approach is complementary to these efforts , and exploits the information that is encoded in the network of known drug interactions—since known interactions are the result of certain ( known or unknown ) “pharmacological rules” , we can infer “rules” from known interactions and then use the inferred “rules” to , in turn , predict unreported interactions ( as we show in the Supporting Text S1 and Fig . S2 , the inferred “rules” correlate strongly with drug structure , category and target ) . Although the network approach has been frequently invoked as a new paradigm in pharmacology [7] , [8] and there are large-scale databases that compile and report drug interactions [10] , [30] , this is , to the best of our knowledge , the first attempt to use network inference to predict drug interactions . The network inference algorithm we have presented is very abstract and does not take into consideration any information other than reported interactions . It may be necessary in the future to complement the method with chemical , biological and/or pharmacological information . However , one advantage of our abstract approach is that , precisely because it is abstract , it can be easily extended to other kinds of pharmacological interaction data that can be represented as networks . For example , it is straightforward to extend our approach to predict associations between drugs and adverse side effects from pharmacosafety networks [33] , protein- and target-drug interactions [34] , [35] , or associations between drugs and therapies [15] and drugs and diseases [36] , which may help to guide drug repositioning . Our approach can even be used to predict gene-disease associations [37] and , therefore , to uncover novel targets . Another interesting extension of our approach is to predict multidrug interactions ( that is , interactions between groups of three or more drugs ) , which are relevant to cancer treatment among others . Although it seems that knowledge of pair interactions may be enough to describe higher-order interactions [16] , within our framework tertiary interactions could also be modeled using three-dimensional stochastic block models in which the probability that three drugs , and interact depends only on the groups , and to which they belong . The generalization to interactions between any number of drugs is straightforward . All in all , we think that our approach opens the door to new ways of looking at and making predictions from pharmacological networks .
For the Yeh et al . dataset , we collected the data on pairwise combinations of 21 antibiotics from Figs . 3 and 4a of [20] . For the Cokol et al . dataset , we collected the data on pairwise combinations of 13 anti-fungal drugs from Fig . 3 of [28] . For the Drugs . com dataset , we collected all drug interactions that were listed in the website , starting from a small set of highly connected seed drugs . Drugs that are not connected , directly or indirectly , to the seed drugs are not included in our analysis . We limited our searches to “generic drugs” ( which include common combinations of generic drugs such as acetaminophen/hydrocodone ) and to “major interactions . ” We consider two snapshots of the database from May 10 , 2010 , and February 22 , 2012 . Finally , for the DrugBank , we downloaded two snapshots of the complete database , corresponding to January 2009 and April 2012 , from http://www . drugbank . ca/downloads [29] , [30] . The fundamental assumption of our approach is that the structure of the drug interaction network can be satisfactorily accounted for by a model , which is unknown but belongs to a family of models , that is , a group of models that can be parametrized in some consistent way . Then , the probability that given the observed network is [19] ( 1 ) To estimate this integral we rewrite it , using Bayes theorem , as [19] , [38] ( 2 ) Here , is the probability of the observed interactions given a model and is the a priori probability of a model , which we assume to be model-independent . For the family of stochastic block models , each model is completely determined by a partition of drugs into groups and the group-to-group interaction probability matrices . Here , is the total number of interaction types ( for example , if interactions can be synergistic , additive or antagonistic , then ) and , for a given partition , the matrix element is the probability that a drug in group and a drug in group interact with each other ( these matrices verify that for all pairs of groups ) . Thus , if belongs to group and to group we have that [38] ( 3 ) and ( 4 ) where is the number of interactions of type between drug groups and . The integral over all models in can be separated into a sum over all possible partitions of the drugs into groups , and an integral over all possible values of each . Using this together with Eqs . ( 2 ) , ( 3 ) and ( 4 ) , and under the assumption of no prior knowledge about the models ( ) , we have ( 5 ) where the integral is over all within the subspace that satisfies the normalization constraints , and is the normalizing constant ( or partition function ) . These integrals factorize into terms corresponding to all pairs [38] , each with the general formUsing these expressions in Eq . ( 5 ) , one obtains ( 6 ) where the sum is over all partitions of the drugs , is the total number of known interactions between groups and , and is a function that depends on the partition only ( 7 ) This sum can be estimated using the Metropolis algorithm [19] , [39] as detailed next . The sum in Eq . ( 6 ) cannot be computed exactly because the number of possible partitions is combinatorially large , but can be estimated using the Metropolis algorithm [19] , [39] . This amounts to generating a sequence of partitions in the following way . From the current partition , select a random drug and move it to a random new group giving a new partition . If , always accept the move; otherwise , accept the move only with probability . By doing this , one gets a sequence of partitions such that [39] ( 8 ) where is the number of samples in . In practice , it is useful to “thin” the sample , that is , to consider only a small fraction of evenly spaced partitions so as to avoid the computational cost of sampling very similar partitions which provide very little additional information . Moreover , one needs to make sure that sampling starts only when the sampler is “thermalized” , that is , when sampled partitions are drawn from the desired probability distribution ( which in our case is given by ) . Our implementation automatically determines a reasonable thinning of the sample , and only starts sampling when certain thermalization conditions are met . Therefore , the whole process is completely unsupervised . The source code of our implementation of the algorithm is publicly available from http://seeslab . info/downloads/drugraph/ and http://github . com/seeslab/drugraph . As often happens in Metropolis sampling , in general it is better to run many short independent sampling processes that a single very long sampler . Results reported here are obtained using 50 independent sampling processes of 200 ( conveniently thinned ) partitions each . These sampling processes can be run in parallel , taking on the order of 1–2 days to complete on high-end CPUs for the largest network considered here ( with over 1 , 500 drugs ) . Sampling an equivalent 10 , 000 partitions with a single run can take 2–3 weeks . The Prism algorithm [27] was originally developed to identify groups of genes that interact monochromatically , that is , that have the same type of interactions with genes in other groups . Yeh and coworkers then introduced Prism II [20] to identify groups of drugs relaxing the requirement for perfect monochromaticity . Our implementation of Prism II is as follows . Each drug is initially placed in a group by itself . Then , groups are sequentially merged until all drugs belong to a single group . At each step , we merge the two groups with the smallest “distance” to each other . The distance between groups and is ( 9 ) Here , the normalized drug-drug distance between drugs and is ( 10 ) with the number of interactions reported for both and . The change of monochromaticity entropy is ( 11 ) where is a vector with the number of synergistic ( − ) and antagonistic ( + ) interactions between groups and , and ( 12 ) with and By itself , the Prism II algorithm returns a tree of nested drug groupings . To make interaction predictions , we need to: ( i ) set the free parameter ; ( ii ) cut the tree at a certain level to get a single partition of the drugs into groups ( a process that needs to be unsupervised ) ; and ( iii ) given those groups , determine the probability of each type of interaction . To cut the tree , we choose the partition with the smallest number of groups among those with total monochromaticity entropy that satisfies , where is the partition that corresponds to putting all drugs in a single group . Additionally , we set to get results consistent with those reported in Ref . [20] ( we also checked that these parameters yield good results for the Cokol dataset , and that the results do not improve using other values of ; see Supporting Text S1 and Fig . S1 ) . Finally , once the groups are defined , we estimate the probability as ( 13 ) where and are defined as above . With our implementation , the Prism-based algorithm takes 1–2 days on high-end CPUs to generate interaction predictions for the large networks considered here ( with over 1 , 000 drugs ) . Given a network of drug interactions , we define the interaction similarity between drugs and as the fraction of interactions with other drugs that are equal for and , over the total number of interactions that are reported for both drugs . In particular , and if two drugs do not have any equal interaction with others . To predict the interaction between drugs and , we order all possible drug pairs by decreasing value of the product of similarities to the query drugs . We then select the pair with the highest product for which the interaction is known , and use that value as our prediction of . Note that we may have , that is , we may use the known interaction between and a drug that is very similar to to predict .
|
Over one in four adults older than 57 in the US take five or more prescriptions at the same time; as many as 4% are at risk of a major adverse drug-drug interaction . Potentially beneficial effects of drug combinations , on the other hand , are also important . For example , combinations of drugs with synergistic effects increase the efficacy of treatments and reduce side effects; and suppressing interactions between drugs , in which one drug inhibits the action of the other , have been found to be effective in the fight against antibiotic-resistant pathogens . With thousands of drugs in the market , and hundreds or thousands being tested and developed , it is clear that we cannot rely only on experimental assays , or even mechanistic pharmacological models , to uncover new interactions . Here we present an algorithm that is able to predict such interactions . Our algorithm is parameter-free , unsupervised , and takes , as its only input , sets of previously reported interactions . We show that our method is able to accurately predict interactions , even in large-scale databases containing thousands of drugs , and that it can be used efficiently to discover interactions of new drugs as part of the drug discovery process .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
A Network Inference Method for Large-Scale Unsupervised Identification of Novel Drug-Drug Interactions
|
We recently reported that centrosomal protein 164 ( CEP164 ) regulates both cilia and the DNA damage response in the autosomal recessive polycystic kidney disease nephronophthisis . Here we examine the functional role of CEP164 in nephronophthisis-related ciliopathies and concomitant fibrosis . Live cell imaging of RPE-FUCCI ( fluorescent , ubiquitination-based cell cycle indicator ) cells after siRNA knockdown of CEP164 revealed an overall quicker cell cycle than control cells , although early S-phase was significantly longer . Follow-up FACS experiments with renal IMCD3 cells confirm that Cep164 siRNA knockdown promotes cells to accumulate in S-phase . We demonstrate that this effect can be rescued by human wild-type CEP164 , but not disease-associated mutants . siRNA of CEP164 revealed a proliferation defect over time , as measured by CyQuant assays . The discrepancy between accelerated cell cycle and inhibited overall proliferation could be explained by induction of apoptosis and epithelial-to-mesenchymal transition . Reduction of CEP164 levels induces apoptosis in immunofluorescence , FACS and RT-QPCR experiments . Furthermore , knockdown of Cep164 or overexpression of dominant negative mutant allele CEP164 Q525X induces epithelial-to-mesenchymal transition , and concomitant upregulation of genes associated with fibrosis . Zebrafish injected with cep164 morpholinos likewise manifest developmental abnormalities , impaired DNA damage signaling , apoptosis and a pro-fibrotic response in vivo . This study reveals a novel role for CEP164 in the pathogenesis of nephronophthisis , in which mutations cause ciliary defects coupled with DNA damage induced replicative stress , cell death , and epithelial-to-mesenchymal transition , and suggests that these events drive the characteristic fibrosis observed in nephronophthisis kidneys .
Nephronophthisis ( NPHP ) is an autosomal recessive polycystic kidney disease ( PKD ) attributed to dysfunction of the primary cilia [1] , antennae-like structures projecting from the cell surface which have sensory or mechanical functions [2] . To date , mutations in seventeen genes have been identified as causing NPHP , yet fewer than half of all NPHP cases segregate with these disease loci [3] . Although ciliary dysfunction with consequent defective planar cell polarity among the epithelial cells in the kidney is believed to be the fundamental etiology of cystogenesis in both NPHP and other types of PKD [4] , the overall size of kidneys in NPHP is considerably smaller than in autosomal dominant PKD [5] . This discrepancy is partly due to tubulointerstitial renal fibrosis in NPHP , which is far more evident than in autosomal dominant PKD-affected kidneys . Epithelial-to-mesenchymal transition ( EMT ) is a hallmark of tubulointerstitial renal fibrosis [6] . Recent studies associating NPHP proteins with defective DNA damage response ( DDR ) signaling [7] , [8] support the notion that accumulation of DNA damage and cilia loss result in cell cycle arrest or cell death with associated renal function loss and fibrosis [9] , but exactly how these processes are linked remains unknown . One of the proteins linking these cellular processes in NPHP is centrosomal protein 164 ( CEP164 ) ( NM_014956 , NP_055771 ) . CEP164 regulates primary cilium formation [10] by promoting vesicular trafficking to the mother centriole during initiation of ciliogenesis [11] . Germline mutations in CEP164 have been reported in families with NPHP15 ( MIM:614845 ) [7] . Furthermore , CEP164 has a role in DDR signaling [7] , [12] , [13] . Cep164 interacts with checkpoint kinases ATR and ATRIP in vivo [12] and localizes with DNA damage proteins TIP60 , SC-35 and phosphorylated Chk1 [7] . CEP164 expression is cell cycle stage-dependent; most protein is present at the end of S phase and the beginning of the G2/M phase when cilia are not typically present . Reduction of endogenous levels of CEP164 by siRNA knockdown in HeLa cells abrogates the G2/M checkpoint [12] , suggesting a critical role in cell cycle regulation . Because disturbance of the cell cycle contributes to the cystic and fibrotic renal phenotype of NPHP [14] , we interrogated whether these non-ciliary functions of CEP164 might contribute to the particular phenotype observed in NPHP kidneys . Here we investigate the role of CEP164 in the cell cycle , particularly in S-phase progression and proliferation . We test wild-type and mutant alleles of CEP164 and verify that disease alleles of CEP164 affect cilia as well as cell cycle progression . Live cell imaging studies suggest that CEP164 protects cells from apoptosis . Furthermore we observe upregulation of EMT and fibrosis markers as a result of reduced cellular levels of CEP164 in vitro and in vivo that could partially explain the cystic and fibrotic renal phenotype of CEP164 mutant patients .
To establish the cell cycle progression of cells after knockdown of endogenous CEP164 , we generated RPE-FUCCI cells [15] stably expressing mKO2-hCdt1 ( 30/120 ) ( red ) and mAG-hGem ( 1/110 ) ( green ) [16] and confirmed that knockdown due to either a pool of four siRNAs ( siCEP164-p ) or an individual siRNA ( siCEP164-i ) causes down-regulation of CEP164 mRNA levels ( Figure S1A ) , resulting in a 5-fold reduction of cilia in these cells capable of forming cilia after serum starvation ( Figure S1B–D ) . Live cell imaging of unsynchronized RPE-FUCCI cells for 72 hours ( Figure S1E and Movies S1–2 ) reveals a significantly shorter cell cycle in siCEP164 transfected cells than control non-targeting siRNA ( siControl ) transfected RPE-FUCCI cells ( ∼35 hours versus ∼48 hours ) ( Figure 1A ) . However , these same videos show that siCEP164 transfected cells remain significantly longer in early S-phase ( 8 . 6 hours ) compared to siControl transfected cells ( 5 . 7 hours ) . Accordingly , G1 , G2 and M phases in siCEP164 transfected RPE-FUCCI cells are shorter ( Figure 1B , S1E and Movies S1–2 ) . RPE-FUCCI cells expressing both mKO2-hCdt1 ( 30/120 ) and mAG-hGem ( 1/110 ) , always demonstrated EdU incorporation , supporting the accuracy of the early S-phase values scored ( Figure 1C ) . To determine how defective ciliogenesis in the siCEP164 cells is affecting cell cycle progression , we performed a time series experiment with RPE-FUCCI cells synchronized at G0/G1 . The reduced ciliation frequency observed in cells with reduced levels of CEP164 ( Figure S1B–C ) is consistent with the increased tendency to enter the cell cycle and with the speed with which they proceed through the cell cycle . Both decreased ciliary frequency and increased cell cycle entry were observed . It takes siControl transfected cells about 10 hours to enter S-phase , whereas siCEP164 treated cells require only 6 hours ( Figure 1D ) . We next wanted to validate whether the accelerated cell cycling of siCEP164 knockdown cells conferred a growth disadvantage as had been suggested by Chaki et al . [7] . We performed CyQUANT NF Cell proliferation assays and then measured fluorescence 72 hours after siRNA knockdown of CEP164 or siControl in RPE and IMCD3 cells . Mouse Cep164 siRNA reduces endogenous mouse Cep164 expression significantly ( Figure S2A–B ) . DNA staining by the CyQuant assay revealed a decreased cell number after knockdown in both cell lines ( **p<0 . 01; Figure 1E ) . Standard growth curves in IMCD3 cells reveal a significant growth advantage for cells treated with siCtrl over cells treated with siCep164 . Previously , these results were seen to be rescued by WT-CEP164 in IMCD3 cells [7] . These results are supporting the conclusion that increased cell cycle progression does not result in decreased population doubling time in the context of CEP164 loss . We stably transfected murine renal inner medullary collecting duct ( IMCD3 ) cells and RPE cells with doxycycline ( Dox ) -inducible constructs expressing GFP-tagged CEP164 wild-type ( Figure 2D , S2D , S5A ) or human disease-associated cDNA N-GFP-CEP164- , -R93W and -Q525X . Upon induction with doxycycline , all constructs showed GFP expression at the base of the cilium , in the cytoplasm and occasionally in the nucleus ( Figure S2D , S5A ) . These cells were then transfected with either siControl or siCep164-p/-i or siCEP164-p/-i to reduce endogenous expression ( Figure S2A , B ) . We observed a reduction in the percent of ciliated cells in all cell lines after knockdown of endogenous Cep164/CEP164 ( Figure S2E , S5B ) . Rescue of cilia numbers in cells was observed upon induction of wild-type allele CEP164-WT ( Figure S2E , S5B ) by the addition of doxycyline , a finding which is consistent with our previously published results and extends upon them , although not quantified in a 3D experimental setting in this study [7] . To induce cell cycle arrest and attain synchronization , we used a double thymidine block prior to release ( Figure S2C ) and then followed the IMCD3 inducible stable cell lines using FACS . Upon siRNA knockdown of Cep164 , cell cycle histograms of IMCD3-N-GFP-CEP164-WT cells revealed increased accumulation of DNA in S-phase at the expense of G2/M phase ( 40%±2 ) when compared to the control siRNA treated cells ( 30%±3 ) ( Figure 2A ) indicating cell cycle delay or arrest in transition from S to G2/M phase . Upon doxycycline induction of wild-type human CEP164 construct N-GFP-CEP164-WT cells were rescued from S-phase arrest ( 32 . 7%±1 . 2 ) ( Figure 2A ) . The observed S-phase block cannot be rescued by overexpression of the human nonsense mutant CEP164-Q525X , a mutation from NPHP family F59 [7] ( Figure 2B ) ; however , it should be noted that simply expressing the CEP164-Q525X allele alone had a nearly identical statistically significant effect as siCep164 treatment , suggesting a dominant negative interference of N-GFP-CEP164-Q525X with murine endogenous Cep164 function ( Figure 2B ) . To rule out the possibility of clonal drift , we repeated these experiments in IMCD3 polyclonal lines expressing N-GFP-CEP164-WT and N-GFP-CEP164-Q525X ( Figures S2F ) and again observed a rescue with the wild-type allele and a dominant negative effect upon expression of the Q525X allele . Cells expressing NPHP missense mutation N-GFP-CEP164-R93W [7] exhibited S-phase accumulation upon siRNA knockdown of endogenous mouse Cep164 ( increase from 30% to 36 . 7%±3 . 2 ) indicating that this disease-causing variant affects this function of CEP164 ( Figure 2C ) . We conclude that CEP164 plays a role in early S-phase progression and that mutations in CEP164 associated with NPHP are defective in this function . Because S-phase arrest may reflect increased DNA damage response signaling , we examined Cep164 levels in the presence or absence of DNA damage and observed increased levels of Cep164 protein ( Figure 2E ) . Importantly , IMCD3 and RPE cells depleted of Cep164 accumulate the DNA damage marker phosphorylated H2AX ( γH2AX ) ( Figure 2F–G ) , which is accompanied by stabilization of PCNA , after transfection with a species-specific pooled ( siCEP164-p/siCep164-p ) or individual siRNA ( siCEP164-i/siCep164-i ) or exposure to replication stress agent aphidicolin ( APH ) ( Figure 2I ) . To examine the pathophysiological relevance of these data , we obtained a urine sample from a newly diagnosed and untransplanted NPHP patient and isolated urine-derived renal epithelial cells . Compared to a healthy age- and gender-matched control , localization of CEP164 was observed to be more nuclear and γH2AX was quantitatively more evident ( ***p<0 . 001 ) ( Figure 2H ) . Although this is an isolated patient , these data would indicate that DDR processes are relevant to advent of NPHP . Despite the fact that reduction of cellular levels of CEP164 by siCEP164 knockdown results in a quicker cell cycle than controls ( Figure 1A ) , we consistently observed a decreased cell number during CyQUANT assays using different cell lines ( Figure 1E ) . This paradox led us to investigate whether apoptosis might explain the discrepancy between the accelerated cell cycle in RPE-FUCCI cells after knockdown of CEP164 and decreased net proliferation which was determined by several assays . The time-lapse data from RPE-FUCCI cells transfected with siCEP164 or control were analyzed for the number of cells characteristically appearing apoptotic ( passive or Brownian movement , blebbing , detaching , and/or lacking fluorescence ) within 72 hours of filming . These events were significantly higher after CEP164 knockdown , normalizing for the total number of cells per field ( *p<0 . 05 ) ( Figure 3A ) . For molecular analysis of apoptotic markers , RNA from RPE-FUCCI cells and IMCD3 cells was isolated after transfection with control or siCEP164-p or siCep164-p oligos respectively and we performed RT-QPCR to measure Caspase-3 mRNA expression ( Figure 3E ) . Twenty-four to 48 hours after transfection there was increased expression of Caspase-3 mRNA ( ***p<0 . 0001 ) which we visualized by live cell imaging of dual immunofluorescent staining of Annexin V and Caspase-3 substrate ( Figure S5C–D ) . As further validation , we transfected CEP164 siRNA into RPE cells stably expressing 53BP1-GFP , a protein which accumulates at double strand breaks [17] , and stained those cells for Caspase-3 after fixation . RPE nuclei showing 53BP1-GFP foci were counted as well as Caspase-3 positive nuclei and were normalized against the total number of nuclei analyzed . After 32 hours more 53BP1-GFP positive cells were counted in siCEP164 cells compared to controls ( *p<0 . 05 ) ( Figure 3B–C ) . During all time points significantly ( *p<0 . 05 ) more apoptosis events were scored ( Figure 3B , D ) . We performed a FACS assay for measuring apoptosis of Cep164 siRNA transfected IMCD3 cells incubated with 50 nM aphidicolin ( APH ) for 16 hours , a treatment which causes replicative stress and synchronizes cells in S-phase [8] . Cep164 knockdown caused apoptosis of IMCD3 cells which was further enhanced by APH treatment ( Figure 3F ) , suggesting that S-phase prolongation and/or replicative stress may predispose renal cells to apoptosis ( Figure 2 ) . Epithelial-to-mesenchymal transition ( EMT ) slows down cell proliferation and provides cells an alternative to apoptosis [18] . A pro-fibrotic mesenchymal transition is characterized by expression of Snail [19] . E-cadherin expression decreases as cells lose their epithelial characteristics and become more mesenchymal [20] . We investigated the role of siCep164 ( Figure S3F ) in the induction of EMT using TGFβ1 incubation ( 5 ng/mL ) as a positive control for EMT in IMCD3 cells [21] , [22] . Six days after knockdown of Cep164 in IMCD3 cells we measured decreased gene expression levels of E-cadherin ( Fig . 4A ) and increased levels of Snail ( Fig . 4B ) . Tieg1 , TGF-β1 , αSMA , Fibronectin1 and CTGF ( Figure S3A–E ) are concomitantly upregulated after Cep164 depletion as measured by RT-QPCR ( *p<0 . 05 ) . Because mesenchymal cells migrate faster than epithelial cell populations [21] , we performed a scratch wound migration assay [23] to investigate the cell migration capacity of IMCD3 cells after knockdown of Cep164 . Cells with reduced levels of Cep164 migrated significantly faster than siControl cells ( *p<0 . 05 ) . TGFβ1 incubation had a comparable effect on migration in this experimental set-up ( Figure 4C–D ) . Accordingly , expressing the dominant negative allele N-GFP-CEP164-Q525X resulted in increased Snail and γH2AX protein levels ( *p<0 . 05 ) ( Fig . 4E–F ) . Finally , we investigated the role of siCep164 in the induction of fibrosis in mouse embryonic fibroblasts ( MEFs ) . Six days after knockdown of Cep164 in MEFs ( Figure S4F ) we measured increased levels of TGF-β1 , Fibronectin1 and CTGF , but not of Tieg1 and αSMA ( Figure S4A–E ) , by RT-QPCR . We conclude that Cep164 has a role in inducing EMT and fibrosis in renal epithelial and mesenchymal cells . Furthermore , our data suggest this effect to be possibly specific to the kidney; RPE cells in similar experimental settings do not undergo EMT and do not migrate ( Figure S5E–G ) . To evaluate cep164 loss of function in vivo we performed morpholino oligonucleotide ( MO ) knockdown in zebrafish using a different MO targeting the splice donor site of exon 3 ( Figure 5O ) than previously published [7] . This more effective morpholino induced consistent and robust developmental abnormalities . cep164 knockdown caused microcephaly , shortened body axis , axis curvature and edema ( Figure 5B , L ) compared to wildtype zebrafish after control injection ( Figure 5A ) and in general recapitulates the results of the other published MO [7] . This phenotype was associated with massive cell death as demonstrated by widespread acridine orange ( Ac . Or . green ) staining in morphant embryos ( Figure 5D ) compared to control injected siblings ( Figure 5C ) . Homozygous p53 mutant zebrafish embryos injected with cep164 morpholino displayed cell death as well compared to control injections ( Figure 5E–F ) . DDR signaling was also activated in morphant embryos as indicated by an increased signal of γH2AX immunofluorescence compared to control ( Figure 5G–H ) . 72 ( hours post fertilization , hpf ) morphant embryos displayed increased apoptosis ( Figure 5I–L , 5P ) and increased γH2AX levels as well ( Figure 5M–N–Q ) . No specific pronephros DNA damage or apoptosis accumulation was observed; however , the pronephros at this embryonic stage is exquisitely regenerative . Similarly , we examined cep164 morphant zebrafish for induction of EMT and a profibrotic response after depletion of cep164 . Indeed , RT-QPCR revealed significant induction of snail and fibronectin1 in cep164 MO injected embryos at 32 ( Figure 5R ) , 72 and 96 ( Figure S6 ) hpf .
NPHP is a common cause of renal end-stage disease in children and young adults . Although NPHP-associated ciliary defects and impaired DNA damage response have been associated with CEP164 dysfunction ( NPHP15 ) [7] , [10] , [12] , [13] , the exact mechanism linking these processes to NPHP is unclear . Here we identify novel functions for CEP164 relevant to NPHP pathogenesis , namely in cell cycle progression , apoptosis , EMT and fibrosis regulation . We show that despite accelerated cell cycle progression , total cell number is decreased after CEP164 knockdown . Our data further indicate a role for CEP164 in S-phase progression . Accumulation of cells in S-phase could be rescued by wild-type CEP164 , but not by its disease-associated variant alleles . We observed that Cep164-loss promotes apoptosis in vitro characterized by increased levels of 53BP1 and γH2AX . Zebrafish with reduced levels of cep164 show developmental abnormalities , increased apoptosis , enhanced DDR signaling and a profibrotic response , demonstrating the in vivo relevance of our findings . These novel functions are highly relevant to the etiology of NPHP which features increased apoptosis and fibrosis . Our data suggests functional similarity between CEP164 and at least one other NPHP protein , GLIS2 ( NPHP7 ) , which also protects renal cells from apoptosis and fibrosis [5] . With two of the seventeen known nephronophthisis-associated proteins clearly associated with these processes , as well as the ongoing large-scale proteomics efforts to understand the nephronphthisis interactome ( www . syscilia . org ) [24] , we anticipate that additional NPHP-genes will be implicated in these processes . In short , we propose that these non-ciliary functions of NPHP genes help to explain differences in disease progression between NPHP and other types of PKD ( Figure 6 ) . Studies from Graser et . al . show that CEP164 expression is cell cycle stage-dependent [10] . Most protein is present at the end of the S-phase and the beginning of G2/M in HeLa cells . Knockdown of CEP164 in HeLa cells showed abrogation of the G2/M checkpoint [12] , suggesting a role of CEP164 in the G2/M checkpoint . In line with these published findings , we find that CEP164 knockdown in RPE-FUCCI cells accelerates G1 , G2 and M cell cycle phase durations but delays S-phase progression . Human wild-type CEP164 , but not its disease-associated mutants , rescue IMCD3 cells from accumulation in S-phase . Activation of the intra-S checkpoint occurs when replication forks cannot stall at damaged DNA [25] , ensuring that the cell cycle progression only reinitiates after damage repair . Because CEP164 mutant alleles were not able to rescue S-phase accumulation , we suggest that replicative stress might contribute to NPHP development in general . In a similar manner , Nek8 ( NPHP9 ) -mediated replication stress contributes to the NPHP etiology [8] . EMT is a hallmark of tubulointerstitial renal fibrosis [6] . Upon CEP164 mutation , induction of apoptosis might compete with EMT as has previously been described [18] , [26] , [27] . The reduced total cell number after siCep164 can partially be explained by induced apoptosis but probably by transdifferentiation of the epithelial kidney cells as well . Accordingly , mesenchymal marker Snail is increased upon reduction of cellular levels of Cep164 in conjunction with a decrease of epithelial marker E-cadherin , indicating that cells lose their epithelial characteristics . The pathological significance of the tubular EMT in renal fibrosis is becoming increasingly accepted [28] . Snail activation associates with patients' renal fibrosis , and disrupts renal homeostasis [29] . The principal effector cells of renal fibrosis are myofibroblasts; evidence suggests that both cells of epithelial origin and cells of mesenchymal origin as progeny for myofibroblasts [6] . We observe that epithelial cells in vitro and in vivo undergo the process of EMT upon Cep164 knockdown , a proposed mechanism contributing to the rise of myofibroblasts during fibrosis . Furthermore , we saw that Cep164 depletion was capable of inducing fibrotic genes in MEF cells , presumably because MEFs are already mesenchymal and thus no longer require EMT . These results obtained from cell types of both epithelial and mesenchymal origin indicate that Cep164 loss can induce fibrosis regardless of the origin of the myofibroblast . In NPHP patients , excessive deposition of extracellular matrix ( fibrosis ) by mesenchymal cells replaces functional tissue [30] . Briefly , this manuscript shows in vitro S-phase arrest , quicker cell cycle progression and EMT and fibrosis induction upon loss of Cep164 . Accumulation of DNA damage signaling during replicative stress could be the cause of the observed apoptosis . Apoptosis is known to contribute to initiation of renal cyst formation [31] , [32] . Our data support the hypothesis proposed by Choi et al . which states that , in addition to cilia loss-of-function , replicative stress contributes to the disease mechanism of NPHP as well [8] . We show that loss of Cep164 results in EMT and fibrosis in different cell types as well as in zebrafish . The induced overall pro-apoptotic and pro-fibrotic response of different cell types may explain the non-cystic features of nephronophthisis such as reduced kidney size ( Figure 6 ) . Since the fibrotic nature of NPHP kidneys is progressive with a time window of several years for therapeutic intervention , understanding and curing this aspect of juvenile kidney disease will potentially delay the need for renal replacement therapy . Our data support the hypothesis that the NPHP-interactome encoded by the 17 NPHP genes coordinates cilia loss-of-function with concomitant DNA damage response , apoptosis , and the creation of a pro-fibrotic environment , all of which directly contribute to the renal phenotype in these patients .
Renal epithelial cells were obtained from a nephronophthisis patient that had been included in the AGORA ( Aetiologic research into Genetic and Occupational/environmental Risk factors for Anomalies in children ) biobank project . The regional Committee on Research involving Human Subjects ( CMO Arnhem/Nijmegen ) approved the study protocol . Written informed consent was obtained from the patient and the parents . All zebrafish experiments were approved by the Animal Care Committee of the University Medical Center Utrecht in the Netherlands . Renal epithelial cells were obtained from a nephronophthisis patient and a healthy gender- and age-matched control . The patient was determined to have isolated clinical diagnosis of NPHP . Urine-derived renal epithelial cells were derived as we have previously described [33] . Mouse Inner Medullar Collecting Duct ( IMCD3 ) cells were cultured in Dulbecco's Modified Eagle's Medium ( DMEM ) :F12 ( 1∶1 ) ( GlutaMAX , Gibco ) , supplemented with 10% Fetal Calf Serum ( FCS ) and penicillin and streptomycin ( 1% P/S ) . Cells were incubated at 37°C in 5% carbon dioxide ( CO2 ) to approximately 90% confluence . IMCD3 cells were stably transfected with CEP164 constructs in a retroviral vector ( pRetroX-Tight-Pur ) for doxycyclin-inducible expression . Inducible overexpression was obtained of N terminally GFP-tagged human full-length CEP164 isoform 1 ( NGFP-CEP164-WT ) , or truncated CEP164 , corresponding with the p . Q525X mutation , or non-functional CEP164 , corresponding with the p . R93W mutation by addition of 2 ng/ml doxycycline [7] . Human retinal pigment epithelial ( RPE ) cells were cultured in DMEM∶F12 ( 1∶1 ) ( GlutaMAX , Gibco ) , supplemented with 10% FCS and 1% P/S . Cells were incubated at 37°C in 5% CO2 to approximately 90% confluence . RPE cells were transfected with lentiviral vectors containing mKO2-hCdt1 ( 30/120 ) and mAG-hGem ( 1/110 ) . Fluorescent , ubiquitination-based cell cycle indicator ( FUCCI ) [16] expressing stable transformants were generated [15] . RPE 53BP1-GFP cells are described in Janssen et . al . [17] RPE cells were stably transfected with CEP164 constructs in a retroviral vector ( pRetroX-Tight-Pur ) for doxycyclin-inducible expression . Inducible overexpression was obtained of N terminally GFP-tagged human full-length CEP164 isoform 1 ( NGFP-CEP164-WT ) . RPE cells were serum starved>24 hour in experiments for cilia quantification . Mouse embryonic fibroblasts ( MEFs ) were cultured in DMEM ( Gibco ) , supplemented with 10% FCS and 1% P/S . Cells were incubated at 37°C in 5% CO2 to approximately 90% confluence . At least 6 hours after plating , cells were transfected with Lipofectamine RNAimax ( Invitrogen , 13778-075 ) , according to the supplier's protocol . Opti-MEM ( Invitrogen , 31985-062 ) was used to dilute the ON-TARGETplus siRNA SMARTpools ( Thermo Scientific Dharmacon ) for Non-targeting pool UGGUUUACAUGUCGACUAA/UGGUUUACAUGUUGUGUGA/UGGUUUACAUGUUUUCUGA/UGGUUUACAUGUUUUCCUA ( D-001810-10 ) , human CEP164 GAGUGAAGGUGUAUCGCUU/GAGAAGUGGCGCAAGUAUU/GGACCAUCCAUGUGACGAA/GAAGAGUGAACCUAAGAUU ( L-020351-02 ) , human CEP164 GAGUGAAGGUGUAUCGCUU ( J-020351-17 ) , or mouse Cep164 GGAGAGUGCAGGAGGGAGA/ACCCAGUGCAGGCAGGAAA/AGUCAGAGAUCCACGGACA/CCACAGAAAGAAAACGAGA ( L-057068-01 ) , mouse Cep164 CCACAGAAAGAAAACGAGA ( J-057068-09 ) to 20 nM . RPE-FUCCI cells were seeded in 8 well Lab-Tek Chamber Slides ( Thermo Scientific ) without addition of serum . After 16 hours , the cells were transfected with Lipofectamine RNAimax . Seven hours after transfection , medium in the Lab-Tek Chamber Slides was replaced with Leibovitz's medium without phenol red ( Gibco ) , supplemented with 6% FCS , 1% P/S and 1% Ultraglutamine . Real time imaging was performed using a Zeiss microscope using a 10× lens . Every 15 minutes images were made of the RPE-FUCCI cells in LED , GFP and dsRED channels for 72 hours . Three positions were imaged per experimental condition . Images were processed with the MetaMorph software . GraphPad Prism 5 . 0 was used to perform one-way ANOVA with Dunnett's post test . For immunostaining , IMCD3 , RPE-FUCCI , urine derived renal epithelial cells or RPE 53BP1-GFP cells were grown on coverslips and fixed for 30 minutes in 4%PFA at the indicated time points , followed by a 15 minutes permeabilization step in 0 . 5%Triton-X100/1%BSA/PBS . Primary antibody incubations ( mouse anti-acetylated tubulin ( Sigma , T7451 , dilution 1∶20000 ) , rabbit anti-CEP164 , Novus 45330002 , 1∶500 , mouse anti-phospho-Histone H2A . X ( Ser139 ) , clone JBW301 , Millipore 05-636 , 1∶500 or rabbit anti-active Caspase-3 ( BD Pharmingen , 559565 , dilution 1∶250 ) were performed overnight in 1% BSA/PBS . Goat anti-mouse/rabbit Alexa 647 secondary antibody ( Invitrogen , dilution 1∶500 ) incubations were performed for 2 . 5 hours at RT . DAPI incubations were performed for 10 minutes at RT . Coverslips were mounted in Fluormount G ( Cell Lab , Beckman Coulter ) . Confocal imaging was performed using Zeiss Confocal laser microscope and images were processed with the ZEN 2011 software . Approximately 250 events per condition were scored . GraphPad Prism 5 . 0 was used to perform statistical analysis . To observe centrosomal localization of N-GFP-CEP164-WT , clonally inducible IMCD3 cell lines doxycylcin ( Dox ) -inducibly expressing human N-GFP-CEP164-WT were treated by double thymidine block ( 2 mM ) . Cells were also induced with doxycycline ( 10 ng/mL ) during the thymidine block to express N-GFP-CEP164-WT . Cells were stained with CEP164-SR antibody followed by anti-rabbit-alexa fluor 594 antibody for confocal imaging to observe colocalization with the induced N-GFP-CEP164-WT-expressing cells . For live cell imaging , RPE cells were seeded in Lab-Tek Chamber Slides with cells at 30% confluency . RPE cells were transfected and after 16 hours , wells were washed once with PBS and once with 1× Binding Buffer ( NucView Dual Apoptosis Kit for Live Cells , Biotium , 30067 ) . Cells were incubated 40 minutes at RT with NucView 488 Caspase-3 substrate ( 5 µM ) and CF 594-Annexin V ( 1∶40 ) in 1× Binding Buffer . Cells were washed once with 1× Binding buffer and confocal imaging was performed using a Zeiss Confocal laser microscope and images were processed with the LSM500 software . GraphPad Prism 5 . 0 was used to perform two-tailed student t-tests . To examine cells in S phase , RPE-FUCCI cells were seeded on coverslips . After 48 hour EdU ( Invitrogen , A10044 ) incorporation took place for 30 minutes using 10 µM in culture medium . The cells were fixed in 3% PFA and washed with PBS . Cells were shortly incubated with EdU staining buffer ( 100 mM Tris pH 8 , 5; 1 mM CuSO4 ) . Then the cells were incubated with EdU staining buffer containing Alexa Fluor 647 azide ( 1∶1000 ) ( Invitrogen , A10277 ) and ascorbic acid ( 0 . 1 M ) ( Merck ) for 30 minutes at RT in the dark . The coverslips were washed twice with PBS and incubated with DAPI for 30 minutes at RT in the dark . The coverslips were mounted with Fluormount G ( Cell Lab , Beckman Coulter ) after washing them once with PBS . Confocal imaging was performed using Zeiss Confocal laser microscope and images were processed with the ZEN 2011 software . RPE and IMCD3 cells were transfected in 96 well plates seeded with cells at 30% confluency . CyQUANT NF reagent ( Invitrogen , C35006 ) was prepared according to the manufacturer's protocol . After 72 hour of incubation , 50 µl of CyQUANT NF Cell Proliferation Assay reagent was added to each well after aspiration of medium . After incubation for 30 minutes at 37°C , fluorescence was measured ( excitation 485 nm , emission 538 nm ) on a Fluoroskan Ascent FL apparatus ( Thermo Scientific , 374-90441C ) using Ascent Software version 2 . 6 . Blanc measurement subtraction was performed and GraphPad Prism 5 . 0 was used to perform two-tailed student t-tests . Cells were lysed and total RNA was isolated ( RNeasy Mini Kit , Qiagen , 74106 ) and measured ( NanoDrop spectrophotometer ND-1000 , Thermo Fischer Scientific Inc . ) . cDNA was synthesized from 500 ng RNA template using the iScript cDNA Synthesis Kit ( Bio-Rad , 170-8891 ) according to the supplier's protocol . Dilutions were made for RT-QPCR analysis to determine mRNA expression levels which were normalized against a reference gene . The iQ SYBR Green Supermix ( Bio-Rad , 170-8880 ) was used to multiply and measure the cDNA with a CFX96 Touch Real-Time PCR Detection System ( Bio-Rad ) . All samples were run in triplicate in 20 µl reactions . The following PCR program was used: 95°C for 3 min , followed by 40 cycles of 10 s at 95°C , 30 s at the indicated annealing temperature and 30 s at 72°C , then 10 s at 95°C followed by a melt of the product from 65°C–95°C . The primer sequences ( Sigma ) used and concomitant annealing temperatures are: hCEP164 forward 5′-GGCAAAGCTGTCAACTTCTGG , hCEP164 reverse 5′-GAACTGGGGCTAATGAGGAAC , 61°C , mCep164 forward 5′-AGAGTGACAACCAGAGTGTCC , mCep164 reverse 5′-GGAGACTCCTCGTACTCAAAGTT , 61°C , hRPLP0 forward 5′-TGCACAATGGCAGCATCTAC , hRPLP0 reverse 5′-ATCCGTCTCCACAGACAAGG , 58°C , hCaspase-3 forward 5′-ACATGGCGTGTCATAAAATACC , hCaspase-3 reverse 5′-CACAAAGCGACTGGATGAAC , 60°C , mE-cadherin forward 5′-CAGTTCCGAGGTCTACACCTT , mE-cadherin reverse 5′-TGAATCGGGAGTCTTCCGAAAA , 66°C , mCaspase-3 forward 5′-GGCTTGCCAGAAGATACCGGT , mCaspase-3 reverse 5′-GCATAAATTCTAGCTTGTGCGCGT , 67°C , mSnail forward 5′- CACACGCTGCCTTGTGTCT , mSnail reverse 5′- GGTCAGCAAAAGCACGGTT , 66°C , hSnail forward 5′-TCGGAAGCCTAACTACAGCGA , hSnail reverse 5′-AGATGAGCATTGGCAGCGAG , 64°C , mRPL27 forward 5′-CGCCCTCCTTTCCTTTCTGC , mRPL27 reverse 5′-GGTGCCATCGTCAATGTTCTTC , 53°C , hVimentin forward 5′-GACAATGCGTCTCTGGCACGTCTT , hVimentin reverse 5′-TCCTCCGCCTCCTGCAGGTTCTT , 67°C , ZfαSMA forward 5′- CATGTACCCGGGCATTGCAGA , , ZfαSMA reverse 5′- GGAAGGTGGAGAGAGAGGCCA , zfSnail forward 5′-CTCCTGCCCACACTGTAACCG , zfSnail reverse 5′-CATGCGACTGAAGGTGCGAGA , zfFibronectin1 forward 5′-TCCCAGACATCACGGGCTACA , zfFibronectin1 reverse 5′-GCATGAGTTCTGTCCGGCCTT , zfβ-actin forward 5′-TCTGGATCTGGCTGGTCGTGA , zfβ-actin reverse 5′- CTCCTGCTCAAAGTCCAGGGC 63°C . Taqman assays were performed to measure mouse CTGF ( Applied Biosystems , probe number Mm01546133_m1 ) , Tieg1 ( Mm00449812_m1 ) , TGFβ1 ( Mm01178820_m1 ) , ACTA2 ( αSMA , Mm00725412_s1 ) , and Fn1 ( Mm01256744_m1 ) gene expression levels . The following PCR program was used: 40 cycles of 15 s at 95°C , 60 s at 60°C . The ΔΔCT method was used for statistical analysis to determine gene expression levels . Protein lysates were prepared using RIPA lysis buffer . To correct for protein content BCA protein assay ( Pierce ) was performed . Western blots were performed for Cep164 . β-actin was used as loading control in combination with Coomassie Blue staining . After blotting , the PVDF membranes were blocked in 5% dried skim milk in TBS with 0 . 5% Tween . And western blots were performed for γH2AX , PCNA and Snail . H2AX and β-actin were used as loading control in combination with Coomassie Blue staining . After dry blotting ( iBlot Dry Blotting System , Invitrogen , IB3010-01 ) , the nitrocellulose membranes were blocked in 5% BSA in TBS with 0 . 5% Tween . The primary antibodies ( rabbit anti-Cep164 , Novus 45330002 , 1∶2000 , rabbit anti-H2AX ( pSer139 ) , Calbiochem DR1017 , 1∶1000 , mouse anti-phospho-Histone H2A . X ( Ser139 ) , clone JBW301 , Millipore 05-636 , 1∶1000 ( Specificity of the gamma H2AX antibodies was determined by pre-treatment with phosphatases ) , rabbit anti-Histone H2A . X , Millipore 070627 , 1∶1000 , rat anti-PCNA , Antibodies Online ABIN334654 , 1∶1000 , rabbit anti-Snai1 , Santa Cruz sc-28199 , 1∶400 , and mouse anti-β-actin AC-15 , Sigma A5441 , 1∶15000 , rabbit anti-GFP Abcam , 1∶1000 ) were incubated overnight at 4°C . The secondary swine anti rabbit , goat anti rat and rabbit anti mouse antibodies which are HRP conjugated ( DAKO , dilution 1∶2000 ) were incubated for 1 hour at RT . The ECL Chemiluminescent Peroxidase Substrate kit ( Sigma , CPS1120-1KT ) was used for development . Scans of the blots were made with the BioRad ChemiDoc XRS+ device with Image Lab software 4 . 0 . GraphPad Prism 5 . 0 was used to perform two-tailed student t-tests . To investigate S-phase progression , dox-inducible non-clonally and clonally selected mouse IMCD3 cells expressing wild type human CEP164 cDNA construct N-GFP-CEP164-WT or mutant human CEP164 construct N-GFP-CEP164-Q525X or N-GFP-CEP164-R93W were transfected with either negative control siRNA ( 50 nM ) or anti-mouse Cep164 siRNA ( 50 nM ) using Polyplus transfection reagents . Cells were treated for double thymidine block ( 2 mM ) from time point 24–42 to 50–68 hrs post transfection . Cells were then also induced with doxycycline ( 10 ng/ml ) at 24 hrs post siRNA transfection for expression of human wild type construct N-GFP-CEP164-WT or human mutant constructs . Cells were released from second thymidine block for 6 hrs and fixed with 2% PFA and stained with PI/RNAse staining solution . Events were acquired in a FACSCalibur flow cytometer ( BD Biosciences ) for the cell cycle histogram Mean and SD of percent of DNA amount for different phases ( triplicate samples ) were calculated and plotted as histograms . To quantify apoptosis , IMCD3 cells were plated and transfected with siControl or siCep164 . After 24 hours cells were exposed to 0 and 50 nM aphidicolin for 16 hours . Cells were harvested and washed once with 1% BSA-PBS . Cells were collected in FACS tubes in 200 µl 1% BSA-PBS containing Vybrant DyeCycle Violet Stain ( Invitrogen , V35003 , 1∶1000 ) to stain living and early apoptotic cells ( 7 minutes at 37°C ) and 7-AAD viability stain ( eBioscience , 00-6993 , 1∶60 ) to stain late apoptotic cells ( 10 minutes on ice ) . Cells were measured ( 10 . 000 events ) with a BD FACSCanto II flowcytometer and analyzed using BD FACSDiva Software [34] . GraphPad Prism 5 . 0 was used to perform two-way ANOVA with Bonferroni post hoc test . IMCD3 cells were transfected overnight with non-targeting siControl or siCep164 oligonucleotides in 24 well plates seeded with cells at 40% confluency . 48 hour later , when the cells were>85% confluent , a plastic disposable pipette tip was used to create a scratch wound in the cell monolayer . After washing the wells once with PBS , the cells were incubated with serum-free medium for 18 hours containing no or 5 ng/mL TGFβ ( Peprotech , 100-21 ) . Images of the same positions of the scratch were made with a light microscope ( 4× objective ) after 0 and 18 hours . Migration of cells was measured with Image-Pro . GraphPad Prism 5 . 0 was used to perform two-way ANOVA with Bonferroni post hoc test . Wild-type and p53-/- embryos ( tp53 M214K ) [35] at the 1–2 cell stage were injected with 1 or 2 nL of a 0 . 1 mM antisense morpholino oligonucleotide targeting Cep164 exon 3 in pure water with 0 . 1% Phenol Red using a nanoject2000 microinjector ( World Precision Instruments ) . The sequence of the exon 3 morpholino was: TGTGTTGTGGAGTGTGTGTTACCAT . The sequence of the standard control morpholino was: 5′-CCT CTTACCTCAGTTACAATTTATA-3′ . Primers amplifying exon 3 ( Cep164 ex 2–4 product length = 187 ) forward primer GGTGCTGGAGGAGGATTATG and reverse primer GTAGTAGACCTCGCCCGTCA were used . For western blot 15 embryos were pooled in 60 µL Triton X-100 lysis buffer . For RT-QPCR 5 embryos were pooled in 100 µL TRIzol reagent ( Invitrogen , 15596-026 ) and RNA was isolated following standard procedures . 24 and 72 ( PTU treated ) hpf live dechorionated embryos are incubated in a 2 mg/mL solution of acridine orange ( Sigma ) in PBS for 30 min at room temperature . Embryos are washed quickly in E3 , then 5×5 minutes in E3 and visualized on a Zeiss LSM5 Pascal confocal microscope . No autofluorescence was detected in the regions analyzed . GraphPad Prism 5 . 0 was used to perform two-tailed student t-tests . Anti-phospho H2AX antibody was a kind gift from James Amatruda ( University of Texas Southwestern Medical Center , Dallas , Texas 75390 ) . Embryos were fixed in 4% PFA overnight at 4°C and stained with Anti-phospho H2AX ( 1∶1500 ) , Alexa 488 goat-anti-rabbit ( Invitrogen A11008 ) secondary antibody and visualized on a Zeiss LSM5 confocal microscope .
|
Nephronophthisis is a leading inherited cause of renal failure in children and young adults . This work contributes to understanding of the disease mechanism of nephronophthisis , which is characterized by multi-cystic and fibrotic kidneys . The genes mutated in patients with nephronophthisis all seem to encode proteins involved in cilia function , and some of them are recently reported to also function in DNA damage signaling . We investigated how loss of cilia and impaired DNA damage signaling could cause the excessive fibrosis seen in nephronophthisis . Studies during the past decade have focused on treating the cysts of this early-onset renal disease . However , we think that understanding and curing the fibrosis seen in these patients will provide new treatment opportunities . Our work gives insight into the orchestration of downstream effects on the cellular level after loss of nephronophthisis gene CEP164 as a result of loss of cilia and accumulating DNA damage signaling .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"chronic",
"kidney",
"disease",
"genetics",
"pediatric",
"nephrology",
"biology",
"and",
"life",
"sciences",
"nephrology",
"gene",
"function"
] |
2014
|
Nephronophthisis-Associated CEP164 Regulates Cell Cycle Progression, Apoptosis and Epithelial-to-Mesenchymal Transition
|
Dengue virus is transmitted by Aedes mosquitoes and infects at least 100 million people every year . Progressive urbanization in Asia and South-Central America and the geographic expansion of Aedes mosquito habitats have accelerated the global spread of dengue , resulting in a continuously increasing number of cases . A cost-effective , safe vaccine conferring protection with ideally a single injection could stop dengue transmission . Current vaccine candidates require several booster injections or do not provide protection against all four serotypes . Here we demonstrate that dengue virus mutants lacking 2′-O-methyltransferase activity are highly sensitive to type I IFN inhibition . The mutant viruses are attenuated in mice and rhesus monkeys and elicit a strong adaptive immune response . Monkeys immunized with a single dose of 2′-O-methyltransferase mutant virus showed 100% sero-conversion even when a dose as low as 1 , 000 plaque forming units was administrated . Animals were fully protected against a homologous challenge . Furthermore , mosquitoes feeding on blood containing the mutant virus were not infected , whereas those feeding on blood containing wild-type virus were infected and thus able to transmit it . These results show the potential of 2′-O-methyltransferase mutant virus as a safe , rationally designed dengue vaccine that restrains itself due to the increased susceptibility to the host's innate immune response .
Dengue virus ( DENV ) is a member of the Flaviviridae family . DENV infection causes dengue fever ( DF ) and the more severe forms of the disease , dengue hemorrhagic fever ( DHF ) and dengue shock syndrome ( DSS ) . DENV has four serotypes ( DENV-1 to -4 ) , each of which is capable of causing severe disease . The frequency , severity , and geographical spread of cases have increased over the past decades [1] , [2] . Every year , one hundred million new cases of DF and 250 , 000 DHF/DSS are estimated by the WHO . At present , despite intensive global research efforts , no vaccine or antiviral treatment for dengue infection is available . Vaccine development is complex due to multiple factors . ( i ) An effective vaccine must consist of a tetravalent formulation protecting against each of the four serotypes because more than one serotype typically circulates in a region . ( ii ) A sub-protective vaccine potentially increases the risk of vaccinees to develop the more severe forms of dengue during repeated infection because of a known association of pre-existing immunity with severity [3] , [4] . ( iii ) Since most infections occur in developing countries , an ideal vaccine should be affordable and fully protective [5] . Taken together , a vaccine inducing a robust level of immunity ideally with only one inoculation is required . Live-attenuated vaccines are replication-competent viruses , which can induce an immune response and an immune memory that are comparable to those induced by the wild-type virus . Live-attenuated viruses do not cause disease because of the low level of replication and hence low levels of inflammation . Prominent examples of successful live-attenuated vaccines providing long-term immunity are those against vaccinia virus , poliovirus ( Sabin ) , and two members of the Flaviviridae family , yellow fever virus ( YF-17D ) and Japanese encephalitis virus ( JEV SA14-14-2 ) [6] . Live-attenuated DENV vaccines have been shown to induce protective neutralizing antibody titers in mice , monkeys , and humans [7]–[9] . In addition , evidence that a balanced T cell response contributes to protection is accumulating , emphasizing the importance of T cell epitopes in a vaccine [8] . Flaviviruses are positive-sense , single-stranded RNA viruses . The flavivirus genome encodes for 3 structural ( C , prM , and E ) and 7 non-structural proteins ( NS1 , NS2A , NS2B , NS3 , NS4A , NS4B , and NS5 ) . NS5 is a multifunctional protein , consisting of the RNA-dependent RNA polymerase [10] and methyltransferase ( MTase ) activities responsible for 5′ RNA cap formation [11] , [12] as well as internal RNA methylation [13] . While N-7-methylation is essential for RNA translation and stability , the function of 2′-O-methylation has remained elusive until recently . We and others demonstrated that while 2′-O-MTase is not essential for viral replication in vitro , viruses bearing mutations in the highly conserved MTase catalytic K-D-K-E tetrad are severely attenuated in the host due to the inability of the virus to shield viral RNA from recognition by host innate immune factors [14] , [15] . DENV RNA binds to RIG-I and MDA5 [16] , [17] , which activates interferon ( IFN ) -β production via a cascade involving IFN-β promoter stimulator 1 ( IPS-1 ) [17] . IFNs in turn activate IFN stimulated genes ( ISGs ) , which induce antiviral responses in infected and neighbouring cells . IFN-induced proteins with tetratricopeptide repeats ( IFITs ) are critical for the inhibition of viral infections , although their functions are only partially understood [14] , [18] . The human IFIT gene family comprises four members: IFIT1 , IFIT2 , IFIT3 ( = IFIT4 ) , and IFIT5; whereas mice only express IFIT1 , 2 and 3 [19] . Interestingly , IFIT homologs are conserved from amphibians to mammals , suggesting that they play a central role in the innate immune response [19] . IFIT1 and 2 bind to eukaryotic Initiation Factor 3 ( eIF3 ) and inhibit translation [20] , whereas IFIT3 amplifies the antiviral signal by connecting IPS-1 and TBK1 , resulting in more IFN production [21] . The role of human IFIT5 is less well understood . Here we demonstrate that DENV strains bearing a mutation in the catalytic site of the 2′-O-MTase replicated to high titres in cell culture whereas these mutant viruses were highly attenuated in mice and rhesus monkeys . The mutation was stable over several passages and reversion to wild-type ( WT ) was not observed . For further safety improvement , a second mutation in the 2′-O-MTase catalytic tetrad was introduced without affecting the viability of the virus in vitro . A single dose administration to rhesus macaques ( RM ) conferred protection to homologous DENV challenge . Mice immunized with a single dose of a divalent ( DENV-1/2 ) formulation of the mutant viruses and mice immunized with the monovalent formulation showed comparable antibody responses , demonstrating that there was no interference between two serotypes of the DENV MTase mutants . Moreover , no enhanced infection and increased TNF-α levels were observed in immunized mice upon challenge with heterologous virus . Overexpression of IFITs in HEK-DC-SIGN cells suggested a role for IFIT1 in the attenuation of MTase mutant in human cells . Taken together , these results demonstrate the potential of 2′-O-MTase mutants as a DENV vaccine . To our knowledge , this is the first live-attenuated rational vaccine approach , tailored to optimally activate the innate and adaptive immune response while being severely attenuated due to its susceptibility to the IFN response .
Flaviviruses replicate in the cytoplasm . The cytoplasm-replicating viruses have evolved N7- and 2′-O-methyltransferases ( MTase ) to methylate their viral mRNA 5′ cap structures [22] . We have previously shown for West Nile virus ( WNV ) and DENV-1 that mutation of the Asp of the tetrad K-D-K-E completely abolished both N7- and 2′-O-MTase activities and was lethal for viral replication; mutations of the other three residues of the tetrad abolished 2′-O-methylation ( with a decrease in N7-methylation ) , and led to attenuated viruses [14] , [23] . Since there are four serotypes of DENV , we introduced the same MTase mutations into DENV-2 to examine whether the same approach was feasible with more than one serotype . A WT recombinant MTase , representing the N-terminal 296 amino acids of the DENV-2 NS5 ( strain TSV01 ) , was cloned and expressed . Two mutant MTases containing Ala-substitutions at the K-D-K-E tetrad ( Fig . 1A ) were prepared: one with a single E217A mutation and another with double K61A+E217A mutations . The mutant enzymes retained 95% and 77% of the WT N7-methylation activity , respectively; neither mutant exhibited any 2′-O-methylation activity ( Fig . 1B ) . BHK-21 cells transfected with equal amounts of WT and mutant ( E217A and K61A+E217A ) genome-length RNAs of DENV-2 generated equivalent numbers of viral E protein-expressing cells ( Fig . 1C ) . Both WT and mutant RNAs produced infectious viruses ( passage 0 ) with similar plaque morphologies ( Fig . 1D ) . The replication of mutant viruses was attenuated in mammalian Vero and mosquito C3/36 cells ( Fig . 1E ) . Continuous culturing of the mutant viruses on Vero cells or HEK-293 cells expressing DC-SIGN ( HEK-DC-SIGN ) for ten rounds ( 3–4 days per round ) did not change their plaque morphologies ( Fig . 1D and data not shown ) . The expression of DC-SIGN facilitates DENV infection [24] . Sequencing of the passage 0 and 10 viruses from both Vero and HEK-DC-SIGN cells showed that the engineered mutations were retained ( Supplementary Fig . S1a and S1b ) . Similar results were obtained for DENV-1 containing the E216A ( E216 in DENV-1 MTase is equivalent to E217 in DENV-2 MTase ) or K61A+E216A mutation in MTase ( Supplementary Fig . S2 ) . Collectively , the results demonstrate that the 2′-O-MTase mutant DENV-1 and -2 are slightly attenuated , but stable in cell culture . We infected AG129 mice with the WT and 2′-O-MTase mutants ( called “E216A” for DENV-1 and “E217A” for DENV-2 from this point ) to assess viral replication and immunogenicity in vivo . AG129 mice lack the receptors for type I and type II IFNs , and have been used widely for antiviral and vaccine testing [25]–[28] . Mice were intraperitoneally ( i . p . ) infected with 2 . 75×105 plaque-forming unit ( PFU ) of WT or mutant viruses . The viremia result showed that mutating K61A or E216A in DENV-1 and mutating E217A in DENV-2 attenuated the virus compared to the WT virus ( Fig . 2A and B ) . Next , we examined a combination of two MTase mutants ( E216A and E217A ) representing DENV-1 and DENV-2 to address a potential competition effect that has been described previously with attenuated strains in humans [29] and in mice [25] . To this end , mice were injected i . p . with 2 . 75×105 PFU of E216A or E217A or a combination of both ( a total of 5 . 5×105 PFU viruses ) . At 30 days post immunization , mice were challenged i . p . with 1×106 PFU of WT DENV-1 or 5×106 WT DENV-2 . DENV specific IgG titers and viremia were observed . All mice immunized with E216A and/or E217A were protected against homologous challenge ( Fig . 2C ) , demonstrating that the immune response was protective even though the IgG titers in E216A and/or E217A-infected mice were 2 to10 times lower than those in the WT virus-infected mice ( Fig . 2D and E ) . A general concern for live attenuated vaccines is their theoretical potential to mutate back to WT under pressure of the immune system . To address this in our system , virus from mice infected with mutant DENV1 or DENV2 was isolated at day 3 after infection and the mutations were found to be stable ( Supplementary Fig . S1c ) . To rule out that compensatory mutations were introduced into the viral genome the input and output ( day 3 after infection ) virus was sequenced using Illumina deep sequencing technology . As summarized in Supplementary Table S1 , only the single nucleotide polymorphisms ( SNPs ) responsible for the E216A or E217A mutation were found when comparing the sequences to wild-type DENV-1 or -2 , respectively . We next compared the neutralization and infection enhancing capacity of serum collected 30 days post immunization ( Table 1 and Supplementary Fig . S3 ) [30] . Mutant viruses cause the same or less antibody-dependent enhancement ( ADE ) than the respective wild-type viruses in the heterologous setting ( 0 . 51±0 . 16 vs . 0 . 74±0 . 2 for DENV-1 immunization and ADE tested against DENV-2 and 0 . 64±0 . 22 vs . 0 . 62±0 . 14 for DENV-2 immunization and ADE tested against DENV-1 ) ( Table 1 ) . More importantly , we did not observe enhanced infection in vivo ( Fig . 2C and see later challenge experiments with a virulent DENV-2 strain ) . These data suggest that vaccination with the E216A/E217A mutants does not cause ADE during heterologous challenge even though lower neutralizing Ab titers are generated by the mutant strains compared to the wild-type virus . While antibodies are crucial to reduce the viral load by binding and neutralizing virus particles , T cells are necessary for efficient viral clearance [31] , [32] . AG129 mice are not suitable to study T cell responses because of their lack of IFN-γ signaling , which is critical to activate T cells . We therefore used IFNAR mice lacking the receptor for IFN-α/β [33] . IFNAR mice were immunized with 2 . 75×105 Pfu DENV-2 E217A or DENV-2 WT and spleens were harvested at day 7 for restimulation in vitro and detection of IFN-γ production ( Fig . 3A ) . Mutant and WT virus elicited a strong CD4 and CD8 T cell response after re-stimulation with DENV-2 . The CD4 response was weaker in E217A-immunized mice , likely due to the lower total viral load in E217A-immunized mice compared to mice immunized with the WT virus ( Fig . 3B ) . To test for targeted DENV T cell response splenocytes were re-stimulated with a pool of NS4B and NS5 CD8 peptides described by Yauch et al [32] . No significant difference in the NS4B and NS5-specific T cell response was seen between mice immunized with E217A or WT DENV-2 ( Fig . 3B ) . Taken together , DENV 2′-O-MTase mutants induce a T cell response and epitope presentation that is similar to WT infection . Nevertheless , additional studies in mice and monkeys are necessary to assess the T cell response in greater detail and to test its functional contribution to protection . DENV-1 strain 05K3126 and DENV-2 strain TSV01 do not cause pathology in mice . To test for protection against a more virulent strain we immunized mice with DENV-1 E216A , DENV-2 E217A , a mixture of E216A and E217A , WT DENV-1 ( Westpac ) or WT DENV-2 ( TSV01 ) or PBS and challenged them with the virulent DENV-2 strain D2Y98P [34] 30 days later ( Fig . 4 ) . DENV-2 E217A protected against the homologous challenge ( Fig . 4A ) . Immunization with DENV-1 E216A protected 70% of the mice , showing limited cross-protection after infection with D2Y98P ( Fig . 4A and 4B ) . No enhanced disease was detected after heterologous challenge . Increased TNF-α levels were associated with pathology in the AG129 mouse model in the context of ADE [35] . To further assess the possibility of ADE-associated pathology , we measured TNF-α levels in plasma three days after challenge . High levels of TNF-α were only detected in unimmunized ( PBS ) mice , showing that TNF-α as a marker of pathology was independent of ADE , and that immunization with E216A did not cause ADE after heterologous challenge . These data demonstrate that immunization with E217A protects mice against challenge with an aggressive , virulent DENV-2 strain that causes 100% mortality in unimmunized mice . To assess the safety ( viremia profile ) and efficacy ( neutralizing antibody response and protection against challenge ) of the 2′-O-MTase mutant DENV vaccine approach in an immunologically competent host , three groups of Rhesus monkeys ( RM ) were immunized with different doses of E217A . One group received a low dose ( 1×103 PFU ) , one group a medium dose ( 1×104 PFU ) , and one group a high dose ( 1×105 PFU ) of E217A virus . Viremia was monitored during 10 days after inoculation . The E217A virus was severely attenuated , and no viremia was detected except for one animal ( R0105 ) that had received a high dose ( 1×105 PFU ) and developed a low viremia ( Table 2 ) . Virus was extracted for sequencing , and it was confirmed that the E217A mutation was retained in the virus extracted at days 3 , 4 and 7 from this animal . Importantly , full virus genome sequencing of the viral RNA recovered at day 7 showed that no compensatory mutations were introduced ( data not shown ) . All immunized monkeys developed neutralizing antibodies to DENV-2 on day 15 after immunization ( Table 3 ) . ADE was analyzed in a K562 assay and a similar enhancement pattern was observed for both heterologous and homologous infection in vitro: ADE correlated with the neutralizing titer , ie the higher the NT50 the higher the enhancement ( Supplementary Fig . S4 ) . This argues against a physiologically relevant infection enhancement , which would only be expected after heterologous infection . By day 30 after immunization , all monkeys including the ones with low dose immunization developed high titers ( GMT≥92 ) of neutralizing antibodies ( Table 3 ) . The monkeys were then challenged with 1×105 PFU of WT DENV-2 on day 64 post-immunization . No viremia was detected in all immunized monkey , whereas all four PBS-immunized controls had a mean peak virus titer of 2 . 5 log10 PFU/ml and mean viremia duration of 4 . 8 days ( Table 4 ) . In all animals except one ( R0055 ) , anamnestic antibody responses were observed after challenge ( Table 3 ) . These data demonstrate that live , attenuated DENV MTase mutant virus , even when administrated at low dose ( 1×103 PFU ) , can induce protective immunity in non-human primates . The 2′-O-methylation of the 5′ cap of WNV and coronavirus RNA functions to subvert innate host antiviral response through escape of IFIT-mediated suppression [14] , [15] . To assess whether this is true for DENV as well , we pretreated HEK-DC-SIGN cells with an increasing dose of IFN-β for 24 h . While HEK-DC-SIGN cells are susceptible to type I IFN , they do not produce detectable levels of IFN-β after infection with mutant or WT virus ( data not shown ) . The IFN-β-treated cells were infected with WT or mutant E217A DENV-2 . The E217A virus was significantly more sensitive to IFN-β pretreatment than the WT virus , as demonstrated by the percentage of infected cells ( Fig . 5A ) as well as the viral titers in culture supernatants ( Fig . 5B ) . To test the stability of the mutation under IFN pressure and in different cell types we passaged the virus in the presence of 0 , 20 and 200 U/ml IFN-β in HEK-DC-SIGN and U937-DC-SIGN . As illustrated in Supplementary Fig . S5 , E217A was lost in the presence of IFN , whereas wild-type virus resisted the IFN pressure in both cell lines . E217A isolated from passage three in HEK-DC-SIGN and from passage one in U937-DC-SIGN was isolated for sequencing . The E217A mutation was retained and no compensatory mutations were introduced ( data not shown ) . To elucidate the molecular mechanism of attenuation , we over-expressed human IFIT1 , 2 , 3 , or 5 in HEK-DC-SIGN cells . The cells were infected with WT or mutant DENV-2 and assessed for the number of infected cells by flow cytometry ( Fig . 5C ) . The WT virus infection was not affected , whereas E217A mutants were significantly inhibited by IFIT1 , but not IFIT2 , 3 , or 5 . However , IFIT1 over-expression did not completely block E217A infection nor did it affect virus output from the infected cells ( Fig . 5D ) , suggesting that other IFN-mediated signals are involved in the response against DENV . Both mutant and WT virus show similar growth kinetics in untreated cells ( Fig . 5E ) . We currently don't know why the mutant virus is attenuated in Vero cells but not in HEK-DC-SIGN since both lines are deficient in IFN production . It should be noted that the maximum antiviral effect of IFITs could be underestimated due to the low transfection efficiency ( 30–50% ) of the IFIT-expressing plasmids . We compared the effect of 2′-O-MTase mutation on viral fitness in mosquito Ae . aegypti , the natural transmission vector for DENV . The mosquitoes were fed with blood containing WT or E217A . After the mosquitoes were fed at a titer of 1×105 PFU/ml , significant differences in oral infection and dissemination between the WT and mutant viruses were observed 15 days post-infection ( Table 5 ) . The WT virus infected 29% of mosquitoes at the highest titer ( 1×105 PFU/ml ) , but only 1–6% of mosquitoes at lower titers ( 1×103 and 1×104 PFU/ml ) . When orally fed with 1×105 PFU/ml WT virus , approximately 10% of mosquitoes were infected after 9; the WT virus disseminated in 24% of the mosquitoes ( Table 5 ) . When fed with 1×103 and 1×104 PFU/ml WT virus , the dissemination rates reached 1–4% . In contrast , the mutant virus was unable to infect the Ae . aegypti and , subsequently , no dissemination was observed for all titers ( Table 5 ) . To examine whether the E217A mutant could replicate in vivo , we intra-thoracically inoculated the WT and mutant viruses into Ae . aegypti mosquitoes . Intra-thoracic inoculation bypasses the mosquito midgut , which is the key barrier to establish infection during natural feeding route . Both WT and mutant viruses reached 100% infection rate upon intra-thoracic inoculation . The mean genome copy number reached 4 . 6×109 and 6 . 2×109 , respectively ( Supplementary Fig . S6 ) . The genome copy number of the WT virus was approximately 35% higher than that of the mutant virus ( p = 0 . 1054 ) . Overall , the results demonstrate that the 2′-O-MTase mutant virus is compromised in vector fitness .
Various dengue vaccine strategies are currently under development , including live attenuated virus , subunit vaccines , chimeric viruses , and DNA vaccines [36] , [37] . The YFV 17D-based chimeric dengue vaccine developed by Sanofi-Pasteur is the most advanced in clinical testing [38] , [39] . The establishment of reverse genetic manipulation of DENV has greatly facilitated the generation of promising vaccine candidates [36] , [38] . The recent progress in understanding the mechanism of attenuation of 2′-O-MTase mutant flaviviruses has provided a novel approach for vaccine and antiviral development [40] . Here we show in a proof-of-concept study that MTase mutant E216A DENV-1 and E217A DENV-2 strains are stable in vitro , and safe and immunogenic in vivo . Importantly , enhancement of infection was not observed after heterologous infection of immunized mice . The fear in a clinical setting is that sub-neutralizing titers of antibodies could enhance infections , even though this has so far not happened in the context of vaccine trials in humans [41] . A commonly used approach to address ADE in vitro is to infect K562 cells in the presence of antibodies . Virus alone is not able to infect K562 cells efficiently , whereas virus-antibody immune complexes bind to K562 cells via Fc-γ receptors ( FcγR's ) , assisting the internalization of the virus and infection of the cells . We found that K562 cells could be infected in the presence of serum from immunized mice and monkeys at dilutions that were approximately 50% neutralizing in the U937-DC-SIGN system ( Supplementary Fig . S3 and S4 ) . This is in line with a previous report , which found that even strongly neutralizing antibodies are enhancing at concentrations that are close to the 50% neutralizing titer [42] . Clinically relevant ADE would be expected at sub-neutralizing titers and only after heterologous infection , and this was not observed in our experiments . A caveat of the K562 system is that the cells do not express inhibitory FcγRIIb , which is present on human target cells ( dendritic cells and macrophages ) and which negatively regulates ADE [43] , [44] . Physiological amounts of complement , another negative regulator of ADE , are also not taken into account [45] . In summary , while the K562 assays done here did not show more ADE for heterologous infections , we cannot exclude ADE because of the limitations of the assay . Potential ADE will have to be addressed in further monkey studies . Live attenuated dengue vaccine candidates have several advantages . Importantly , they can induce long lasting humoral and cellular immune responses to both structural and non-structural viral proteins . In this study we show a CD8 response to NS4B and NS5 peptides that is similar in mice immunized with mutant or WT virus , suggesting that the response is qualitatively equivalent . Chimeric viruses using the same backbone for all four DENV serotype glycoproteins would induce a type-specific response restricted to the structural proteins of one DENV serotype [36] . The interdependence of the T and B cell response for the efficient generation of immune memory has been demonstrated in a number of human studies [46] , [47] . We speculate that the advantage of an attenuated non-chimeric DENV that includes all naturally occurring T and B cell epitopes could be that only one vaccination is required to confer long-term immunity to re-infection , as seen for natural DENV infections [48] , [49] . A single-dose vaccine would facilitate the logistics of a vaccination program and would significantly reduce its cost compared to candidates requiring several booster immunizations . The 2′-O-MTase mutant DENV vaccine approach , with a known mechanism of attenuation , can be readily generated using a reverse genetic system . This is in contrast to the method to develop live attenuated vaccines by passaging of WT viruses in cell lines , leading to the introduction of random mutations . The reverse genetic system-based rational vaccine ensures that the vaccine maintains the attenuated genotype . Additionally , a tetravalent formulation would contain the same attenuating mutation in all four serotype recombinant vaccine strains , making the generation of a more pathogenic virus by intra vaccine-strain recombination impossible . Moreover , recombination in cell culture is hardly observed in flaviviruses , suggesting that flaviviruses are not prone to evolution by recombination [50] . By introducing additional mutations in the K-D-K-E tetrad of 2′-O-MTase , further safety and attenuation can be achieved . Only a virus that has at least two mutations will be acceptable in the clinical setting . Our data demonstrate that the 2′-O-MTase E217A virus is attenuated in mice and monkeys . We cannot explain fully why the 2′-O-MTase mutant virus was attenuated ( 10-fold lower virus titer compared to WT virus ) in AG129 mice , which are unable to respond to IFN-signals . It is likely that pattern recognition receptors and downstream pathways activated by the mutant virus trigger antiviral defense mechanisms in an IFN-dependent and IFN-independent manner . Whether the balance between low virulence and high immunogenicity is achieved in humans by 2′-O-MTase mutants remains to be elucidated . Our studies in human HEK293 cells show increased susceptibility of DENV2 E217A mutant to IFN-β in vitro , suggesting that DENV E217A mutants will be attenuated in humans as well . In the monkey immunization experiments , one monkey out of four in the high dose group experienced peak viremia of about 100 PFU , which is comparable to other live attenuated vaccine candidates [51] . Indeed , weak replication of the vaccine approach is desirable in order to induce a strong protective cellular immune response . Replication should be restricted enough to preclude onset of illness , whereas sub-clinical symptoms such as mild rash , transient leukopenia , and mildly elevated liver enzyme values are generally accepted [52]–[54] . Furthermore , studies with murine hepatitis virus have shown that MTase mutants are highly attenuated in its natural host , induce IFN , which could further induce the immunogenicity of a vaccine , and are genetically stable in vivo [15] . Moreover , the replication level of WNV 2′-O-MTase mutant in mice was largely decreased in the spleen , serum , or brain in comparison with the WT WNV infection . Intracranial inoculation of 1×105 PFU of 2′-O-MTase mutant WNV did not cause any mortality and morbidity in mice , demonstrating the safety of this vaccine approach [14] . Taken together , these evidences demonstrate the safety and immunogenicity of the MTase-mutant vaccine approach . We are currently working on the tetravalent formulation to develop the strategy towards a clinical application .
All experimental procedures involving Rhesus Monkeys were approved by and carried out in strict accordance with the guidelines of the Animal Experiment Committee of State Key Laboratory of Pathogen and Biosecurity , Beijing , China . All procedures were performed under sodium pentobarbital anesthesia by trained technicians and all efforts were made to ameliorate the welfare and to minimize animal suffering in accordance with the “Weatherall report for the use of non-human primates” recommendations . The mouse experiments were conducted according to the rules and guidelines of the Agri-Food and Veterinary Authority ( AVA ) and the National Advisory Committee for Laboratory Animal Research ( NACLAR ) , Singapore . The experiments were reviewed and approved by the Institutional review board of Biological Resource Center , Singapore ( IACUC protocols 90474 and 100536 ) . BHK-21 , C6/36 , and HEK-293 were purchased from the American type culture collection ( http://www . atcc . org ) . HEK-293 and U937 cells expressing DC-SIGN were obtained by lentiviral transfection and subsequent cell sorting . All cells were maintained in minimal essential medium supplemented with fetal bovine serum ( 5%–10% ) . WT MTases representing the N-terminal 262 and 296 amino acids of DENV-1 and -2 NS5 , respectively , were cloned , expressed , and purified as reported previously [11] . Mutagenesis of MTase was performed using QuikChange II XL site-directed mutagenesis kit ( Stratagene ) . The complete sequence of each mutant MTase was verified by DNA sequencing . N7- and 2′-O-methylation assays were performed as described before [11] . Full-length infectious cDNA clones of DENV-1 ( Western Pacific 74 strain ) and DENV-2 ( TSV01 strain ) [55] , [56] were used to generate WT and mutant viruses . A standard mutagenesis protocol was used to engineer mutations into the MTase region as reported previously [11] . The protocols for in vitro transcription , RNA transfection , IFA , plaque assay , and growth kinetics were reported previously [23] . Strain D2Y98P was described previously [34] . Female or male 6–8 week old IFN α/β/γ receptor deficient mice ( AG129 ) were purchased from B&K Universal Limited with permission from Dr . M . Aguet ( ISREC , School of Life Sciences Ecole Polytechnique Fédérale ( EPFL ) ) . IFN α/β receptor deficient mice ( IFNAR ) on a C57BL/6 background were provided by Prof . Ulrich Kalinke [33] . All mice were bred and kept under specific pathogen-free conditions at the Biomedical Resource Centre , Singapore . For immunization , BHK-21 derived mutant and WT viruses were used . For challenge experiments DENV produced in C6/36 cells was used . Fourteen RMs , weighing from 3 . 4 to 5 . 0 kg , were pre-screened negative for IgG antibodies against DENV and JEV by indirect immunofluorescence assay . Animals were randomly divided into four groups and inoculated subcutaneously ( s . c . ) in the deltoid region of left arm with 0 . 5 ml of DENV-2 E217A dilutions containing 5 . 0 , 4 . 0 , 3 . 0 log10 PFU , respectively . Animals in the control group received PBS . Blood was collected from each RM daily post immunization for 10 days to detect viremia . For neutralizing antibody tests , blood was taken before immunization ( day −1 ) and on days 15 and 30 post-immunization . On day 64 post-immunization , all immunized animals including the PBS-treated control animals were challenged by s . c . inoculation of 0 . 5 ml containing 5 log10 PFU of DENV-2 ( TSV-01 ) . For the following 9 days , blood was collected for determination of viremia . Neutralizing antibody levels in serum were measured by the standard 50% plaque reduction neutralization test ( PRNT50 ) on days 15 and 30 post-challenge , respectively . Viremia in serum samples was determined by plaque assay in BHK-21 cell monolayers in 12-well plates . Undiluted serum or serial 10-fold dilutions of serum were inoculated onto BHK cells . After 1 h of adsorption at 37°C , wells were overlaid with 1 ml of DMEM supplemented with 2% FBS and 1% agarose . Plates were incubated for 4 days at 37°C in 5% CO2 . Monolayers were fixed by addition of 1 ml of 4% formalin solution to the overlay medium . After 1 h of fixation at room temperature , the fixative was removed , wells were washed with water , and monolayers were stained with 1% crystal violet in 70% methanol . Plaques were counted , and titers were expressed as PFU/ml . For determination of dengue virus-neutralizing antibody titers , serial twofold dilutions of serum ( starting at a dilution of 1∶10 ) were mixed with equal volumes of a suspension of ∼500 PFU of DENV-2 TSV01/ml . The serum-virus mixtures were incubated at 37°C for 1 h and tested ( 0 . 2 ml/well ) for concentration of infectious virus using the plaque assay described above . The neutralization titer was defined as the lowest serum dilution at which the infectious virus concentration was reduced by 50% from the concentration found when virus was incubated with culture medium . Cells were seeded at 1×105 per well in a 24-well plate and treated 24 hours prior to infection with medium or varying concentrations of human recombinant IFN-β ( Immunotools ) . Cells were then infected at an MOI of 1 with WT or MTase mutant virus ( TSV01 ) , respectively , incubated for 72 hours and harvested and processed for flow cytometry as described . Supernatants were collected for plaque assay . IFIT expression plasmids were a kind gift from A . Pichlmair ( 14 ) . For IFIT overexpression cells were seeded at 1×106 per well in a 6-well plate . 24 hours later cells were transfected using 293fectin according to manufacturer's protocol . One day post transfection cells were trypsinized and seeded in a 24-well plate at 1×105 per well . After 24 hours of incubation cells were infected and analysed as described previously . Transfection rate was 30–50% judged by parallel experiments with GFP expression plasmid . To determine the percentage of infected cells , cells were harvested , washed in PBS and fixed and permeabilized with Cytofix/Cytoperm ( BD ) . Intracellular dengue E protein was stained with antibody 4G2 conjugated to Alexa 647 and fluorescent cells were measured by flow cytometry . For the assessment of ADE , 4G2 or serum/plasma was serially diluted and a constant amount of virus was added . The antibody-virus mixture was incubated at 37°C for 30 min and then 50 µl of the mixture was added to 25'000 K562 cells per 96-plate well ( MOI 0 . 5–1 ) . After 2 h of infection 150 µl RPMI medium containing FCS was added . After 2 . 5 days of incubation the infected cells were fixed and stained intracellularly with 4G2-Alexa 647 . The percentage of infected cells was quantified by flow cytometry . For the measurement of neutralization , 4G2 or heat-inactivated serum/plasma was serially diluted and a constant amount of virus was added . The antibody-virus mixture was incubated at 37°C for 30 min and then 50 µl of the mixture was added to 200 , 000 U937 cells ( ATCC ) stably transfected with human DC-SIGN ( MOI 0 . 1–1 ) . After 2 h of infection 150 µl RPMI medium containing FCS was added . After incubation over night the infected cells were fixed and stained intracellularly with 4G2-Alexa 647 . The percentage of infected cells was quantified by flow cytometry and data were analyzed with GraphPad Prism software for the calculation of the NT50 . Spleens were harvested at day 7 after infection and single cell suspensions were incubated with live virus or a pool of the following peptides: NS4B59-66 , NS4B99-107 and NS5237-245 [32] overnight . Brefeldin ( Biolegend ) was added for 5 h before cells were washed and stained with antibodies CD4-APC , CD8-PercPCy5 . 5 and IFN-γ-PE-Cy7 ( Biolegend ) . Cells were acquired on a FACSCantoII ( BectonDickinson ) and data were analyzed with FlowJo ( Treestar ltd . ) 96-well polystyrene plates were coated with concentrated , heat inactivated dengue virus . Plates were incubated overnight at 4° . Before use , plates were washed three times in PBS ( pH 7 . 2 ) containing 0 . 05% Tween-20 ( PBS-T ) . Non-specific binding was blocked with 2% non-fat dry milk diluted in PBS ( PBS-M ) for 2 h at room temperature ( RT ) . After washing , sera were diluted 1∶50 in PBS-M , heat inactivated for 1 hour at 55°C and threefold serial dilutions were added to the wells . Plates were incubated for 1 h at RT , followed by three washes with PBS-T . Peroxidase-conjugated rabbit anti-mouse IgG , in PBS-M was added , followed by 1 h of incubation at RT and three additional washes with PBS-T . TMB was used as the enzyme substrate . The reaction was stopped with 1 M HCl and the optical densities were read at 450 nm using an automatic ELISA plate reader . Endpoint titers were defined as the lowest dilution of plasma in which binding was twofold greater than the mean binding observed with the negative controls . Vector competence experiments were performed using a colony of Ae . aegypti mosquitoes in which 10% of the population is derived from field obtained eggs each month . Batches of 50–75 female mosquitoes , aged 5–7 days were fed with pig blood containing WT or MTase mutant DENV-2 at titers of 5 , 4 , and 3 log 10 PFU/ml . Fully engorged mosquitoes were held at 27°C , 80% relative humidity , and 12 h photoperiod for 15 days , after which the abdomen was separated from the thorax and homogenized . Homogenates were inoculated into Vero cell culture . After culturing the inoculated cells for 5 days , viral infection was assayed using an indirect fluorescent antibody test ( IFA ) . Antibody 6B6C-1 against flavivirus group E protein ( at 1∶10 dilution; provided by the USA CDC as a mouse hybridoma ) and an anti-mouse antibody conjugated with FITC were used as a primary and secondary antibody , respectively . Positive fluorescence determinations were performed manually using an inverted fluorescent microscope ( Olympus IX71 ) . Chi-square and contingency table statistical tests were performed to detail heterogeneity in vector competence within/between WT and mutant viruses . Intra-thoracic inoculation of 0 . 17 µl of WT DENV-2 and E217A at a titer of 105 PFU/ml was performed using 10 female mosquitoes each . Following inoculation , mosquitoes were held for seven days under the same conditions as described above . Mosquitoes were then killed by freezing and homogenized . Viral RNA was quantified by real-time qRT-PCR using primers and methods reported previously [57] . Briefly , whole mosquito homogenate viral RNA was extracted using QIAamp Viral RNA Mini Kit ( Qiagen ) . qRT-PCR was completed using Invitrogen SuperScript III Platinum One-Step qRT-PCR mix ( without ROX ) and CFX96 Real-Time PCR Detection System ( BioRad ) . Cycling parameters performed were 50°C for 30 min , 95°C for 2 min , followed by 45 cycles of 95°C for 10 sec , 60°C for 30 sec . A two-tailed unpaired t-test was performed to determine the statistic difference between the mean genomic equivalents calculated for WT and mutant viruses . Virus was isolated from mouse serum with Qiagen Viral RNA extraction Kit . Fifty ng of viral RNA were used to prepare cDNA libraries using the Illumina TruSeq RNA sample preparation kit according to manufacturer's protocol . The only protocol modification was the removal of the mRNA enrichment step . The cDNA libraries were sequenced as a multiplex in a single lane of an Illumina HiSeq2000 ( Next Generation Sequencing Core facility , Genomic Institute of Singapore ) . One to 2 million 50 bp paired-end reads were generated for each virus . Wild-type and mutant virus samples were mapped to their respective reference genomes using Bowtie 2 [58] . Mapping statistics and genotype calls were made with SAMtools [59] . Data analysis was performed in Pipeline Pilot ( http://www . accelrys . com ) . At least two reads with an alternate base at a given position were defined as a SNP . Statisitical tests were performed with GraphPad Prism software , using students t test , two-way ANOVA or Chi-square and contingency table statistical tests as indicated in the figure legends .
|
The four serotypes of dengue virus cause severe outbreaks globally in tropical countries with thousands of patients requiring hospitalization . The health care and indirect economic cost of dengue in endemic countries is huge . Despite this , no clinically approved vaccine or antiviral treatment is currently available . Dengue transmission could be stopped with a vaccine that provides full protection to all serotypes . Dengue afflicts many developing countries and a vaccine should therefore be cost-effective and should provide protection with ideally a single injection . Here we present a novel dengue vaccine approach that harbours mutation ( s ) in the 2′-O-methyltransferase ( MTase ) , a viral enzyme that methylates viral RNA as a strategy to escape the host immune response . Non-methylated RNA is recognized as “foreign” and triggers an interferon response in the cell . The MTase mutant virus is immediately recognized by the host's immune response and hardly has a chance to spread in the organism while an immune response is efficiently triggered by the initially infected cells . Mice and monkeys infected with the mutant virus developed an immune response that fully protected them from a challenge with wild-type virus . Furthermore , we show that MTase mutant dengue virus cannot infect Aedes mosquitoes . Collectively , the results suggest 2′-O-MTase mutant dengue virus as a safe , highly immunogenic vaccine approach .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"infectious",
"diseases",
"clinical",
"immunology",
"immunology",
"biology"
] |
2013
|
Rational Design of a Live Attenuated Dengue Vaccine: 2′-O-Methyltransferase Mutants Are Highly Attenuated and Immunogenic in Mice and Macaques
|
The immune response to influenza virus infection comprises both innate and adaptive defenses . NK cells play an early role in the destruction of tumors and virally-infected cells . NK cells express a variety of inhibitory receptors , including those of the Ly49 family , which are functional homologs of human killer-cell immunoglobulin-like receptors ( KIR ) . Like human KIR , Ly49 receptors inhibit NK cell-mediated lysis by binding to major histocompatibility complex class I ( MHC-I ) molecules that are expressed on normal cells . During NK cell maturation , the interaction of NK cell inhibitory Ly49 receptors with their MHC-I ligands results in two types of NK cells: licensed ( “functional” ) , or unlicensed ( “hypofunctional” ) . Despite being completely dysfunctional with regard to rejecting MHC-I-deficient cells , unlicensed NK cells represent up to half of the mature NK cell pool in rodents and humans , suggesting an alternative role for these cells in host defense . Here , we demonstrate that after influenza infection , MHC-I expression on lung epithelial cells is upregulated , and mice bearing unlicensed NK cells ( Ly49-deficient NKCKD and MHC-I-deficient B2m-/- mice ) survive the infection better than WT mice . Importantly , transgenic expression of an inhibitory self-MHC-I-specific Ly49 receptor in NKCKD mice restores WT influenza susceptibility , confirming a direct role for Ly49 . Conversely , F ( ab’ ) 2-mediated blockade of self-MHC-I-specific Ly49 inhibitory receptors protects WT mice from influenza virus infection . Mechanistically , perforin-deficient NKCKD mice succumb to influenza infection rapidly , indicating that direct cytotoxicity is necessary for unlicensed NK cell-mediated protection . Our findings demonstrate that Ly49:MHC-I interactions play a critical role in influenza virus pathogenesis . We suggest a similar role may be conserved in human KIR , and their blockade may be protective in humans .
Influenza viruses are classified as members of the Orthomyxoviridae family , which are enveloped viruses with a segmented , negative , single-stranded RNA ( ssRNA ) genome that contains 7–8 gene segments . Structurally , influenza A virus expresses two surface glycoproteins , hemagglutinin and neuraminidase [1 , 2] . Influenza A virus can cause severe human illness , including upper and lower respiratory tract infections and pneumonia , and is associated with major human pandemics . Seasonal influenza epidemics result in 250 , 000–500 , 000 deaths worldwide annually [3] . NK cells are innate lymphocytes that play a critical role in host defense against tumors and virus infection , both by directly eliminating them and by enhancing the rapid development of adaptive responses [4–6] . NK cells are important for protection against influenza virus infection in various animal models [5 , 7 , 8] . In response to NK cell cytolytic function , influenza virus has developed several evasion strategies to escape NK cell recognition [9 , 10] . Importantly , influenza virus infection was shown to induce accumulation of MHC-I molecules in the lipid raft microdomains of infected cells , leading to increased binding of the NK cell inhibitory receptor KIR2DL1 and inhibition of human NK cell cytotoxicity in vitro [11 , 12] . NK cell effector functions are tightly controlled by the combination of signals received through germline-encoded activating and inhibitory receptors [6 , 13] . Mouse NK receptors include the Ly49 , NKG2 , and NKR-P1 families of receptors encoded in the Natural Killer gene Complex ( NKC ) on chromosome 6 [13 , 14] . Inhibitory receptors engage molecular indicators of health , while activating receptors engage indicators of disease . By integrating these signals , the NK cell can appropriately spare or destroy a potential target [14] . Ly49 family members are type II transmembrane glycoproteins , part of the C-type lectin superfamily that forms disulphide-linked homodimers [15] . The mouse Ly49 are functionally equivalent to human killer-cell immunoglobulin-like receptors ( KIR ) . The ligands for KIR and Ly49 receptors are self MHC-I molecules or MHC-I related molecules that are expressed by pathogens upon infection [5 , 16 , 17] . Beyond regulating NK cell killing , interactions between MHC-I and Ly49 receptors are required for NK cell education . The licensing hypothesis states that , to be fully functional , a developing NK cell must successfully engage a self-ligand with an inhibitory receptor [18 , 19] . In a C57BL/6 mouse , this is canonically achieved by engagement of MHC-I by Ly49C and/or Ly49I . Accordingly , NK cells that do not express Ly49C/I , or cells from MHC-I-deficient or Ly49-deficient ( NKCKD ) mice , are ‘unlicensed’ , displaying attenuated responses to MHC-I-deficient tumors in vitro and in vivo [19–22] . NKCKD mice also develop lymphomas earlier than WT mice , again suggesting a degree of dysfunction in unlicensed NK cells [22] . Despite being unlicensed , however , these Ly49C/I- cells represent up to half of the population of mature NK cells in a healthy , WT mouse [9 , 18 , 23–25] , suggesting a role for these cells in host defense . Since these unlicensed cells are dysfunctional with regard to rejecting MHC-I-deficient tumors , their role in host defense may be in NK-mediated anti-pathogen activity . MHC-I-deficient ( B2m-/- ) mice , which bear only unlicensed NK cells , exhibit robust NK cell responses and can control mouse cytomegalovirus ( MCMV ) infection as efficiently as WT mice [26 , 27] . A publication by Orr et al . found that adoptively transferred unlicensed Ly49C/I- Ly49G2+ NK cells into MCMV-infected neonates enhanced their survival better than the licensed , Ly49C/I+ Ly49G2+ cells [28] . For a more in-depth analysis of the role of unlicensed NK cells in viral infection , we have used Ly49-deficient ( NKCKD ) mice generated in our laboratory [20] , in which approximately 80% of NK cells are unlicensed . Thus , NKCKD mice serve as a model to study the role of unlicensed NK cells during viral infections . In this study , we explore the interactions of influenza virus with licensed and unlicensed NK cells . We present evidence that influenza effectively evades licensed NK cells , but is eliminated by unlicensed NK cells in a perforin-dependent manner . Importantly , genetic and physical disruption of Ly49 binding to its MHC-I ligands results in enhanced NK cell-mediated control of influenza virus infection in vivo , implicating virus-induced MHC-I expression as an immunoevasion strategy .
C57BL/6 ( B6 ) and B6 . 129P2-B2mtm1Unc/J ( B2m-/- ) mice were purchased from The Jackson Laboratory ( Bar Harbor , ME ) , Ifnar1-/- were obtained from Dr . Subash Sad ( University of Ottawa , Ottawa , ON ) . B6 . NKCKD , B6 . NKCKD-Ly49Itg , B6 . Ly49QKO , and their congenic control , B6 . Ly49129 mice , have been previously described [20 , 29] . Mice deficient in both Ly49 ( NKCKD ) and perforin ( Prf-/- ) were produced by mating B6 . NKCKD with B6 . Prf-/- mice . B6 . NKCKD , B6 . Ly49QKO and WT controls were bred as homozygous pairs . B6 . NKCKD-Ly49Itg and B6 . Prf-/- mice were bred as heterozygous mating pairs , and littermates were used for experimentation . Assessment of genotypes was performed by PCR . All mice were maintained in a specific-pathogen-free environment . All breeding and manipulations performed on animals were approved by the University of Ottawa animal care committee ( protocol BMI-2049 ) in accordance with the principles published in the Canadian Council on Animal Care’s “Guide to the Care and Use of Experimental Animals” and with the Animals for Research Act , R . S . O . 1990 , c . 22 , s . 17 ( 1–3 ) . Groups of age and sex-matched mice ( 6–8 weeks old ) were anesthetized with isoflurane and inoculated intranasally with 600 or 1050 PFU of mouse-adapted A/FM/1/47-MA ( FM-MA ) strain influenza virus [30] . Influenza-infected mice were housed in a level 2 confinement area for the duration of the experiment . Body weight was measured daily . Animals were considered to be at endpoint if weight loss exceeded 25% of the body weight prior to infection , or if the animal was moribund . Viral loads of infected mice were determined by plaque assay , as described previously [30] . Virus titer is expressed as the number of plaque forming units per gram of lung ( PFU/g ) . Anti-NK1 . 1 mAb ( clone PK136 ) , anti-IFN-γ mAb ( clone XMG1 . 2 ) , and anti-Ly49C/I F ( ab' ) 2 mAb ( clone 5E6 ) were injected i . p . into groups of age and sex-matched WT mice . 200 μg of mAb per mouse were injected i . p . two days prior to influenza virus infection , on the day of infection , and every two days post-infection until day 10 p . i . Anti-AsialoGM1 antibody ( Wako Pure Chemical Industries , Osaka , Japan ) was injected i . p . two days prior to influenza virus infection ( 25 μl ) , on the day of infection ( 25 μl ) , and every three days post-infection ( 10 μl ) until day 10 p . i . Lungs were removed and minced in 5 ml RPMI with 0 . 5 mg/ml collagenase D ( Roche ) , followed by incubation for 1 h at 37°C with agitation . The minced pieces were crushed on a 70 μm cell strainer to prepare single cell suspensions for flow cytometry as previously described . Anti-mouse CD18 ( LFA-1 ) , CD326 ( EpCAM ) , MHC-I ( H-2Kb ) , 5E6 ( anti-Ly49C/I ) , 4D11 ( anti-Ly49G ) , CD8 ( CD8β ) , CD4 , CD3 , TCRβ , NKp46 ( CD335 ) , NKG2D ( CD314 ) , NKG2A ( 16a11 ) , NKG2A/C/E ( 20d5 ) , CD27 , CD11b , CD107a ( 1D4B ) , IFN-γ ( XMG1 . 2 ) , and Live/Dead stain were purchased from eBioscience ( eBioscience , San Diego , CA , USA ) . Anti-NKG2D ( CD314 ) was purchased from BioLegend ( BioLegend , San Diego , CA , USA ) . Anti-mouse TCRβ chain was purchased from BD Biosciences ( BD Biosciences , Mississauga , Ontario , Canada ) . PK136 ( anti-NK1 . 1 ) , 5E6 ( anti-Ly49C/IB6 ) , and XMG1 . 2 ( anti-IFN-γ ) hybridomas were kind gifts from Drs . James Carlyle ( Sunnybrook Research Institute , Toronto , ON ) , Charles Sentman ( Dartmouth Hitchcock Medical Center , Lebanon , New Hampshire ) , and Subash Sad ( University of Ottawa , Ottawa , ON ) , respectively . Cell fluorescence data was acquired with a CyAN-ADP flow cytometer ( Beckman Coulter ) and analyzed with Kaluza software ( Beckman Coulter , New Jersey , USA ) . The levels of cytokines and chemokines in lung tissue homogenates were measured by bead array flow cytometry using the mouse Th1/Th2/Th17/Th22-13plex FlowCytomix multiplex kit and mouse chemokine 6plex kit ( eBioscience , San Diego , CA , USA ) . Streptavidin-PE conjugated influenza A non-structural protein ( NS2 ) 114-121 ( RTFSFQLI ) and nucleocapsid protein ( NP ) 311-325 ( QVYSLIRPNENPAHK ) tetramers were kindly provided by the NIH Tetramer Core Facility at Emory University ( Emory University Vaccine Center , Atlanta , GA ) . 5x105 lung cells were stained with 1 μg of tetramer in 20 μL of cRPMI and incubated for 1 h at 37°C . Influenza virus-specific CD8+ T cells were stained with H-2KbNS2114-121 ( RTFSFQLI ) tetramer , while influenza virus-specific CD4+ T cells were stained with the I-Ab ( NP ) 311-325 ( QVYSLIRPNENPAHK ) tetramer . Total lymphocytes isolated from the lungs were incubated with YAC-1 cells at 1:1 ratio or with phorbol 12-myristate 13-acetate ( PMA , 10 μg/ml ) and ionomycin ( 1 μg/ml ) in the presence of anti-CD107a mAb and brefeldin A ( eBiosience ) for 4 h . Cells were stained for surface markers followed by intracellular staining for IFN-γ using IC fixation and permeabilization reagents ( eBioscience ) following manufacturer’s instructions . Individual hybridoma clones were cultured in DMEM supplemented with 1 mM sodium pyruvate , 0 . 1 mM non-essential amino acids , 0 . 1 mM β-mercaptoethanol , 100 U/ml penicillin , and 100 μg/ml streptomycin . Culture supernatants were then centrifuged ( 10 , 000x g for 20 min ) and filtered through a 0 . 45 μm filter . Monoclonal antibodies ( mAb ) were purified using Protein G sepharose chromatography ( ExalphaBiologicals , Inc , USA ) . Monoclonal Ab was dialyzed against 1x PBS buffer ( pH 7 . 4 ) and then concentrated using an Amicon ultra-15 centrifugal filter unit with an ultracel-100 kDa membrane ( EMD Millipore Corporation , MA , USA ) . Monoclonal Ab concentration was determined by SDS-PAGE gel and by spectrophotometric measurement at 280 nm . To make 5E6 F ( ab' ) 2 fragments [31] , mAb was dialyzed twice against 100 mM sodium acetate solution ( pH 4 . 0 ) , digested using pepsin ( Sigma-Aldrich , Ontario , Canada ) , and then dissolved in 100 mM sodium acetate solution ( pH 4 . 0 ) at a 1:40 pepsin to mAb ratio for 10 h at 37°C . The digested mAb was dialyzed against 1x PBS buffer ( pH 7 . 4 ) and then concentrated using an Amicon ultra-15 centrifugal filter unit with an ultracel-50 kDa membrane ( EMD Millipore Corporation , MA , USA ) . F ( ab' ) 2 fragments were then purified using protein A affinity chromatography . 5E6 F ( ab' ) 2 fragment concentration was determined by SDS-PAGE gel and by spectrophotometric measurement at 280 nm . The purity of 5E6 F ( ab' ) 2 fragments was determined by SDS-PAGE gel . Lungs were collected from infected mice 7 days p . i . and fixed in 10% neutral buffered formalin ( 25 ml ) for 24 h . Subsequently , lungs were embedded in paraffin , sectioned at a thickness of 4 μm and stained with hematoxylin and eosin ( H&E ) . Slides were examined under a microscope to score histopathologic changes in the lungs by a pathologist blind to the experimental conditions . Statistical comparisons were made by a two-tailed Student’s t-test , one-way ANOVA with Bonferroni post-hoc test , or Kaplan Meier survival statistical analysis ( log rank test ) using GraphPad Prism software ( GraphPad , San Diego , USA ) . A p value <0 . 05 was considered statistically significant .
It is well established that influenza virus infection in vitro inhibits NK cell cytotoxicity by enhancing NK cell inhibitory receptor binding to MHC-I on infected human lung epithelial cells [11 , 12] . It is possible that MHC-I has direct negative effects on NK cell activity during influenza virus infection in mice as well . To test this , we first determined whether influenza virus infection could modulate MHC-I expression on mouse lung epithelial cells . WT ( B6 ) mice were infected with 600 PFU of mouse-adapted influenza strain A/FM/1/47 ( H1N1 ) ( FM-MA ) intranasally . MHC-I expression was determined on EpCAM+ ( CD326 ) lung epithelial cells on day 5 post-infection ( p . i . ) by flow cytometry . In uninfected mice , low levels of MHC-I expression were detected on lung epithelial cells ( Fig 1A ) . However , upon infection with influenza virus , lung epithelial cell expression of MHC-I was dramatically increased ( Fig 1A ) . As a control , infected and uninfected MHC-I-deficient mice ( B2m-/- mice ) were included , and as expected , displayed no increase in MHC-I staining upon infection . Influenza virus infection is known to upregulate type I IFN ( IFN-I ) production , which in turn drives the upregulation of a variety of immunomodulatory proteins including MHC-I . To determine whether this upregulation is mediated by IFN-I production , we infected WT ( B6 ) and mice deficient in the receptor for IFN-α and IFN-β ( Ifnar1-/- ) with 600 PFU FM-MA virus intranasally . Lungs were harvested on day 5 p . i . Interestingly , expression levels of MHC-I were similar between WT ( B6 ) and Ifnar1-/- mice ( Fig 1B ) , confirming this upregulation was independent of IFN-I . These data demonstrate that influenza virus infection induces upregulation of MHC-I expression on lung epithelial cells . Increased expression of MHC-I on lung epithelial cells upon influenza virus infection may have implications for the inhibition of NK cells through interaction with inhibitory Ly49 receptors . Recently , we reported that Ly49-deficient ( NKCKD ) mice exhibit uncontrolled tumor growth and metastases [22] . Lacking licensed NK cells renders these mice highly susceptible to tumor formation , despite having otherwise normal mature NK cells . Flow cytometry analysis of lung lymphocytes showed that NK cells in the lungs of NKCKD mice were mostly devoid of Ly49 expression ( Fig 1C ) , but were otherwise predominantly mature cells ( CD11b+ CD27low ) with normal expression of the activating receptors NKp46 and NKG2D ( Fig 1C and 1D ) . To determine whether Ly49 interaction with MHC-I molecules is relevant to influenza infection in vivo , we inoculated WT and NKCKD mice with 1050 PFU of FM-MA intranasally . The animals were observed daily for over two weeks , and sacrificed when moribund . Death due to infection began occurring on day 8 p . i . ( Fig 1E ) . Two weeks p . i . , almost 90% of the WT mice had succumbed to the infection , while unexpectedly , only 35% of the NKCKD mice succumbed ( Fig 1E , **p = 0 . 0072 ) . Next , we depleted NK cells in our mouse model to confirm the role of NK cells during influenza virus infection . This depletion was performed with anti-AsialoGM1 treatment instead of the standard NK1 . 1 treatment , as these mice express an allele of NKR-P1C that is not recognized by the NK1 . 1 antibody . Anti-AsialoGM1 is known to deplete some activated macrophages and CD8+ T cells at high doses [32 , 33] . We used a dose that showed total NK depletion without T cell depletion ( S1A Fig ) ; however , depletion of activated T cells and activated macrophages remains a possibility . Depletion of NK cells from NKCKD mice using anti-AsialoGM1 treatment ( S1B Fig ) resulted in a complete loss of protection ( Fig 1F ) , indicating that NK cells from NKCKD mice play a key role in protecting those mice from influenza . Very low ( 50 PFU ) or high ( 5000 PFU ) doses were too extreme to note any survival differences , but 600 PFU gave similar results to 1050 PFU ( S2A Fig , Fig 1E and 1F ) , therefore , 600 PFU infection dose was used for the remainder of the study . NKCKD mice lack Ly49Q , which plays an important role in IFN-α production by plasmacytoid dendritic cells [29 , 34] . To confirm that this survival advantage is not due to a loss of Ly49Q alone , we inoculated Ly49QKO mice with influenza virus . Like WT mice , Ly49QKO mice died 10 days post influenza virus infection ( Fig 1E , **p = 0 . 0049 ) . NKCKD mice also express lower levels of NKG2A/C/E . However , the inhibitory NKG2A is not believed to be involved in influenza protection by NK cells [35] , and its expression is not altered in B2m-/- mice ( S3A Fig ) , which are also protected from influenza virus ( described below ) . In addition , while NKG2A/C/E expression is decreased in NKCKD mice , it is increased in B2m-/- mice ( S3B and S3C Fig ) . These data show that NKCKD mice survive influenza virus infection better than WT mice in a Ly49Q- and NKG2A/C/E-independent manner . Better survival of NKCKD mice compared to WT mice indicates a possible role for Ly49:MHC-I interactions in the pathogenesis of influenza virus in mice . Our data implicate NK cells from NKCKD mice , which are unlicensed due to a lack of Ly49 receptors , in better protection against influenza . To determine the responsiveness of unlicensed NK cells during influenza virus infection , we compared the function of different NK cell subsets in WT mice following infection . NK cells in WT mice can be divided into four populations based on their expression of Ly49C/I and Ly49G . While the licensed Ly49C/I+ G- population dominates the lung microenvironment in the steady-state ( Fig 2A , left ) , following infection , the unlicensed Ly49C/I- G+ population shows more dramatic expansion ( Fig 2A , right ) . This can be attributed in part to the greater number of proliferating cells in this subpopulation following infection ( Fig 2B ) . NK cells were activated ( IFN-γ+ and CD107a+ ) upon infection with influenza virus , however , we observed equal levels of intracellular IFN-γ levels in all of these four NK cell subsets ( S4 and S5 Figs ) . While we see an outgrowth of the Ly49G+ cells in the WT mice following infection , we observe no similar outgrowth from the residual Ly49G+ cells present in NKCKD mice ( Fig 2C ) , indicating that it is likely a lack of Ly49C/I , and not the presence of Ly49G , that is conferring the survival advantage in these animals . Ly49-negative NK cells are not subject to inhibition via interaction with MHC-I molecules . A comparison of intracellular IFN-γ levels and CD107a ( LAMP1 ) expression on this subset of NK cells in the WT and NKCKD mice revealed a similar level of response in both mice following influenza infection ( S6A and S6B Fig ) . However , a significantly higher number of IFN-γ+ and CD107a+ NK cells lacking expression of Ly49 are present in the NKCKD mice , due to the disruption of Ly49 expression in these mice ( Fig 2D and 2E ) . Therefore , the presence of a larger number of activated NK cells , which are not inhibited via interaction with the increased expression of MHC-I molecules on the lung epithelial cells in the influenza-infected mice ( Fig 1A ) , may confer a survival advantage to NKCKD compared to the WT mice , during an influenza virus infection . Influenza virus infection causes severe lung pathology , leading to respiratory distress and mortality [36] . To examine lung pathology in WT and NKCKD mice , lungs were collected 7 days p . i . with 600 PFU of FM-MA viruses . Microscopic examination of H&E stained lung sections showed more severe alveolar damage , leukocyte infiltration , and pulmonary edema in WT mice compared to NKCKD mice ( Fig 3A–3D ) . Similar results were obtained with 1050 PFU ( S2B Fig ) . From these data , we attribute the increased mortality in WT mice over NKCKD to more severe influenza-induced—and possibly immune-mediated—lung pathology . Next , we asked whether NKCKD mice can eliminate influenza virus-infected cells more efficiently than WT mice , possibly leading to less viral burden and inflammation . To avoid survivor bias and ensure that the observed effect is due to innate immune responses , lungs were collected from infected WT and NKCKD mice 5 days p . i . , and viral titers were determined . Interestingly , viral titer in NKCKD mice was significantly lower than in WT mice on day 5 p . i . ( Fig 3E ) , while viral loads were equivalent between the two on days 3 and 8 p . i . ( S7A Fig ) . This result indicates that NKCKD mice are better than WT mice at controlling lung viral loads early during infection . Better control of the virus may lead to lower levels of inflammation and decreased lung injury following influenza virus infection in NKCKD mice . In addition to injury resulting from influenza virus replication , pro-inflammatory cytokines and chemokines have been suggested to play a pathogenic role in humans and animals infected with influenza virus [37 , 38] . We have demonstrated previously that , similar to WT mice , NK cells from NKCKD mice produce normal levels of cytokines upon stimulation with tumor cell lines , anti-NKp46 mAb , and after murine CMV ( MCMV ) infection [20] . To address the role of pro-inflammatory cytokines and chemokines in the pathogenesis of influenza viruses , we determined the cytokine and chemokine profile in the lungs of influenza-infected WT and NKCKD mice by bead array flow cytometry , 5 days p . i . The majority of cytokines had similar baseline levels in WT and NKCKD mice ( S8 Fig ) . However , the most striking differences occurred in the levels of TNF-α , IFN-γ , IL-17 , MCP-1 , MCP-3 , and MIP-1β , which were elevated significantly in lung homogenates of WT mice compared to the NKCKD mice ( Fig 4A–4F ) . These remarkable changes in pro-inflammatory cytokines and chemokines in the lungs of influenza virus-infected WT mice suggest their involvement in lung pathology . Furthermore , while we noticed a trend toward elevated IFN-γ and TNF-α in the WT lung NK cells themselves ( Fig 4G and 4H ) , it is likely that other immune cells in the microenvironment are contributing to the cytokine profile we observe in the bulk lung extracts . To demonstrate that the observed significant differences in cytokine levels and influenza virus load in the lungs of WT mice at day 5 p . i . is dependent on an NK cell response , we quantified the immune cell subsets responding to viral infection . We examined the NK cell and the virus-specific CD8+ and CD4+ T cell responses in the lungs of influenza virus−infected WT and NKCKD mice 5 and 7 days p . i . Assessment of the percentage and absolute number of these lymphocyte subpopulations after influenza virus infection showed that a protective response to the infection within the first 5 days directly correlated with NK cell expansion ( Fig 5A–5F ) , and not that of virus-specific CD4+ or CD8+ T cells . Substantial CD4+ and CD8+ T lymphocyte counts were only observed 7 days post-influenza virus infection ( Fig 5A–5D ) ; however , flow cytometry detected an expansion of the NK cell population in all mice 5 days p . i . ( Fig 5E–5F ) . Notably , the percentage and the number of NK cells from both WT and NKCKD mice show that there is no statistically significant difference between both groups 5 days p . i . However , the number of NK cells was significantly higher in WT mice ( **p = 0 . 0031 ) compared to NKCKD mice ( Fig 5F ) 7 days p . i . , most likely as a result of the reduction in virus load along with cytokine levels in the lungs of NKCKD mice at day 5 . These data strongly suggest that activation and increase in NK cell numbers within the first 5 days post-influenza virus infection enhances the antiviral response mediated by NK cells , and as a result , plays a substantial role in the initial control of influenza virus , especially by NKCKD mice . To test the direct contribution of self-MHC-I-specific Ly49 inhibitory receptors in the pathogenesis of influenza virus , we introduced the inhibitory self-MHC-I-specific Ly49I transgene into NKCKD mice through breeding as previously described [20] . We inoculated WT , NKCKD , and NKCKD-Ly49Itg mice with 600 PFU FM-MA virus intranasally . The animals were observed daily for over two weeks and sacrificed when moribund . Death due to infection began occurring on day 8 p . i . ( Fig 6A ) . In agreement with our previous results , 90% of WT and 40% of NKCKD mice succumbed by two weeks p . i . ( Fig 6A , *p = 0 . 0147 ) . Remarkably , 100% of NKCKD-Ly49Itg mice died ten days p . i . ( Fig 6A , ***p<0 . 0001 ) . Thus , Ly49 deficiency was definitively protective in NKCKD mice , likely via the loss of MHC-I-specific NK cell inhibition . This finding raises an intriguing question as to whether MHC-I-deficient mice would also be protected from lethal influenza virus infection . To answer this question , we infected four groups of mice: WT ( B6 ) and B2m-/- ( MHC-I deficient ) , and WT and B2m-/- mice treated with anti-NK1 . 1 mAb to deplete NK cells . Remarkably , 50% of B2m-/- mice survived the infection , whereas all B6 mice and NK cell-depleted B6 mice died ten days p . i . ( Fig 6B , ***p<0 . 0001 ) . B2m-/- survival advantage was observed despite a similar viral load measured in the lungs of these mice on day 5 and 7 p . i . compared to the B6 mice ( S7B Fig ) . NK cell activity ( IFN-γ+ and CD107a+ ) following in vitro stimulation was also similar in influenza virus-infected B2m-/- and WT ( B6 ) mice ( S6C and S6D Fig ) ; however , NK cell inhibition via Ly49:MHC-I interaction is disrupted in B2m-/- mice . Interestingly , all NK cell-depleted B2m-/- mice succumbed to the infection as well ( Fig 6B , ***p = 0 . 0015 ) , indicating a direct role for MHC-I-unlicensed NK cells in controlling influenza virus infection in these mice . To validate this proof-of-concept , several groups have shown that blocking the interaction between Ly49C/I and their ligands enhances NK-mediated anti-cancer cytotoxic functions [39] . Likewise , it is possible that a functional blockade of Ly49C/I:MHC-I interactions might protect WT ( B6 ) mice during influenza virus infection . To determine whether Ly49 interactions with MHC-I molecules are relevant to influenza infection in vivo , WT ( B6 ) mice were treated two days prior to FM-MA infection , at the day of infection , and every two days after until day 10 p . i . with 200 μg of 5E6 F ( ab' ) 2 mAbs , previously reported to block Ly49C/I:H-2Kb interactions [39] . Blockade of Ly49C/I inhibitory receptors resulted in a significant increase in mouse survival when compared with untreated B6 mice ( Fig 6C , *p = 0 . 0313 ) . These results demonstrate that preventing Ly49 inhibitory receptor interactions with their cognate MHC-I ligands is protective in mice against influenza virus infection . Our data strongly suggest that an early increase in NK cell numbers in the lungs during influenza virus infection enhances the antiviral response and the initial control of influenza virus , especially in NKCKD mice . NK cells can directly limit virus replication by killing infected cells via the release of cytotoxic granules containing perforin and granzymes , or indirectly by producing IFN-γ , which plays a role in macrophage activation and in inhibiting viral replication . We wanted to address whether perforin or IFN-γ protects NKCKD mice against influenza virus infection . To determine whether the absence of perforin can be fatal for NKCKD mice , we crossed NKCKD mice to perforin-deficient mice ( Prf-/- ) on the C57BL/6 background to obtain NKCKD-Prf-/- mice . We inoculated NKCKD and NKCKD-Prf-/- mice with influenza virus intranasally . Interestingly , in contrast to NKCKD mice , all NKCKD-Prf-/- mice succumbed to the influenza infection ( Fig 7A ) . Similarly , to address the role of IFN-γ , two groups of NKCKD mice were inoculated with influenza virus intranasally . We neutralized IFN-γ in one group of influenza-infected mice using anti-IFN-γ mAb injections two days prior to and at the day of influenza infection , and every two days until day 10 p . i . As a control , a group of uninfected NKCKD mice were treated with anti-IFN-γ mAb as well . Almost 90% of influenza virus-infected NKCKD mice that were treated with anti-IFN-γ mAb died from the infection ( Fig 7B ) . Both the NKCKD-Prf-/- and anti-IFN-γ mAb-treated NKCKD mice had higher lung viral titers compared to the NKCKD control group on day 5 and 7 , prior to the engagement of an adaptive immune response ( Fig 7C ) . These data show that protective NK cell-mediated antiviral defenses in NKCKD mice during influenza virus infection require both IFN-γ and perforin .
In the current study , we demonstrate for the first time that disrupting the interaction between MHC-I and Ly49 inhibitory receptors on NK cells protects mice from lethal influenza virus infection . Here , we found that the proportion of lung epithelial cells that express MHC-I increased dramatically following influenza virus infection , in an IFN-I independent manner . We also provide evidence for the relevance of this upregulation in the severity of influenza virus infection , in which upregulation of MHC-I on lung epithelial cells in the presence of Ly49:MHC-I interactions allows influenza virus to escape recognition by NK cells , facilitating viral replication and mouse mortality . This demonstrates the direct contribution of Ly49:MHC-I interactions in the severity of influenza virus infection . The human analogues of the mouse Ly49 receptors are the KIRs . Previous studies have reported that patients with either mild or severe pandemic influenza A H1N1/2009 virus infections had significantly higher levels of inhibitory KIR2DL5 gene expression in comparison to healthy individuals [40] . The presence or absence of interactions between NK cells expressing the KIR2DL5 inhibitory receptor and targets expressing its MHC-I ligand could be responsible for the severity of influenza virus infection in human patients . Moreover , another group observed that inhibitory KIR2DL2 and KIR2DL3 allotypes and their cognate ligands , HLA-C1 and HLA-C2 , respectively , are significantly enriched in H1N1/2009 intensive-care unit patients in comparison to healthy individuals [41] . Taken together , these reports suggest that potential associations of specific inhibitory KIRs and their MHC-I ligands during severe influenza A virus infections may result in decreased NK function and increased virus pathogenicity . In vitro studies have shown that impairment of human NK cell activity is directly mediated by the accumulation of MHC-I molecules on the surface of influenza virus-infected cells , which in turn increases binding of the NK cell inhibitory receptor KIR2DL1 to the infected cells and inhibits human NK cell-mediated cytotoxicity against them in vitro [11 , 12] . MHC-I upregulation by influenza virus may act as a mechanism to evade NK cell recognition by inhibiting NK cell functions via inhibitory Ly49 engagement . Disrupting this interaction between MHC-I and inhibitory NK cell receptors in humans may interfere with this and reduce the severity of an infection . Ly49-defecient mice are able to survive an influenza virus infection specifically due to the lack of Ly49 inhibitory receptors; transgenic expression of a self-MHC-I-specific inhibitory Ly49 receptor in NKCKD mice restores their influenza susceptibility to WT levels . This demonstrates the direct contribution of Ly49:MHC-I interactions in the severity of influenza virus infection . The presence of Ly49I restored the Ly49:MHC-I interaction in NKCKD NK cells , and lead to significant mortality . While licensed NK cells are protective against MHC-I-deficient tumors [22] , we show that unlicensed NK cells are chiefly responsible for protection from influenza , which may also be true with other viral infections . This could be explained by the lack of inhibitory Ly49 receptors which , if engaged by their MHC-I ligands on the virus-infected cells , could otherwise result in NK cell inhibition ( Fig 7D ) . This hypothesis is supported by our F ( ab’ ) 2 fragment-mediated Ly49C/I blockade , which was found to be protective in WT mice against influenza virus infection , possibly by disrupting the Ly49C/I:MHC-I interaction . Protection conferred by the blocking of Ly49C/I receptors was smaller compared to the lack of Ly49 receptors ( NKCKD ) or MHC-I ligands ( B2m-/- ) , possibly because of a suboptimal blocking in these experiment or the involvement of other inhibitory MHC-I receptors in influenza immunoevasion . Our results from influenza virus infection in NKCKD mice indicate that survival is associated with decreased cytokine and chemokine levels . It would be reasonable , therefore , to assume that unlicensed NK cells are ‘protective’ by being hyporesponsive , leading to a reduced immune response and consequentially less immune-mediated tissue damage . If this were true , a reduction in NK cell activity would be protective in mice challenged with influenza . However , our data show that reducing the activity of NK cells in NKCKD mice , either by deleting perforin or by neutralizing IFN-γ , is detrimental to the animal’s survival . One caveat to interpreting these data is that neither model is NK specific , and T cells will also be affected by a loss of perforin or IFN-γ . As we have shown , however , NK depletion in NKCKD mice is sufficient to disrupt their viral resistance , making it difficult to discount NK-derived perforin and IFN-γ . Furthermore , although infected with equal amounts of virus , perforin-deficient and IFN-γ-neutralized NKCKD mice show a significantly higher viral titer compared to the NKCKD control mice on days 5 and 7 p . i . , before a T cell response is established . In addition , the NKCKD mice show a significantly reduced viral titer compared to the WT mice on day 5 p . i . , implying that the NK cells in these mice are more efficient in clearing the virus . Taken together , these results indicate that the protection enjoyed by NKCKD mice is not due to a decrease in NK cell activity but is in fact due to an increased efficiency at clearing the virus . This is also supported by increased viral lethality following the depletion of NK cells in B2m-/- mice . This increased efficiency of unlicensed NK cells against influenza agrees with previous results obtained with MCMV infections [28] . Our survival data indicate that the NK cells in the NKCKD mice are more protective than in WT , suggesting that what we have called ‘unlicensed’ NK cells do in fact possess an efficient anti-viral activity . This could explain why , in a normal mouse immune system , up to half of the NK cells are ‘unlicensed’ with regard to self-Ly49 expression [18 , 25] . If the two compartments were understood to provide immunity under different circumstances ( tumor versus infection ) , there is a rationale for maintaining such a large pool of NK cells that , until now , were believed to be dysfunctional . We propose that the increased efficiency of NK cells in clearing influenza virus in NKCKD mice results in less need for an extended and toxic immune response , reducing the amount of inflammatory cytokines known to cause severe lung damage and a high rate of mortality during influenza virus infection [38 , 42] . Survival of MHC-I-deficient mice infected with influenza virus was previously attributed to the lack of cytotoxic CD8+ T cells and hence reduced pulmonary damage in these mice [43] . Our data , on the other hand , show a dominant role for NK cells in the protection of these mice against influenza virus infection , since depletion of NK cells makes them susceptible . Consistent with our observations is the reported reduced frequency of NK cells in patients with severe H1N1/2009 infections; CD8+ effector T cells were detected at normal levels in these patients [44] . Our study identifies the mechanisms by which influenza virus escapes recognition by NK cells . Disrupting this interaction between MHC-I and inhibitory NK cell receptors in vivo interferes with this evasion mechanism and , thus , alters the severity of the infection . Additionally , we have demonstrated a previously unidentified role for unlicensed NK cells of the innate immune system . These observations open up a new field of investigation related to NK cell education and encourage more precise elucidation of the nature of each of these functional NK subsets .
|
Influenza virus has developed a number of immune-evasion mechanisms to prolong its survival within the host . Development of functional NK cells is dependent on multiple factors such as the interaction between MHC-I and Ly49 receptors . NK cells that develop in the absence of these interactions are referred to as ‘unlicensed’ and represent up to half of the total number of NK cells . We show that significant MHC-I upregulation on lung epithelial cells following influenza virus infection most likely allows influenza virus to evade detection by licensed NK cells . Importantly , we demonstrate that unlicensed NK cells play a major role in protecting mice from influenza infection . Both Ly49- and MHC-I-deficient mice , which possess unlicensed NK cells , exhibit better survival than WT mice when infected with a lethal dose of influenza virus . Survival of Ly49-deficient mice is associated with reduced viral titers and lung pathology , compared to the infected WT mice . Moreover , disrupting the interaction between MHC-I and inhibitory Ly49 receptors protects WT mice from a lethal influenza virus infection . These results suggest that the so-called unlicensed NK cells , previously characterized as being hyporesponsive , actually possess potent antiviral activity , and are crucial for protection from influenza virus and possibly other viral infections .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"cytokines",
"influenza",
"pathogens",
"immunology",
"microbiology",
"orthomyxoviruses",
"epithelial",
"cells",
"viruses",
"animal",
"models",
"developmental",
"biology",
"model",
"organisms",
"rna",
"viruses",
"molecular",
"development",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"animal",
"cells",
"medical",
"microbiology",
"microbial",
"pathogens",
"t",
"cells",
"biological",
"tissue",
"mouse",
"models",
"spectrophotometry",
"immune",
"system",
"cytophotometry",
"cell",
"biology",
"anatomy",
"nk",
"cells",
"influenza",
"viruses",
"viral",
"pathogens",
"physiology",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"viral",
"diseases",
"organisms"
] |
2016
|
Influenza Virus Targets Class I MHC-Educated NK Cells for Immunoevasion
|
The inoculation of a low number ( 104 ) of L . amazonensis metacyclic promastigotes into the dermis of C57BL/6 and DBA/2 mouse ear pinna results in distinct outcome as assessed by the parasite load values and ear pinna macroscopic features monitored from days 4 to 22-phase 1 and from days 22 to 80/100-phase 2 . While in C57BL/6 mice , the amastigote population size was increasing progressively , in DBA/2 mice , it was rapidly controlled . This latter rapid control did not prevent intracellular amastigotes to persist in the ear pinna and in the ear-draining lymph node/ear-DLN . The objectives of the present analysis was to compare the dendritic leukocytes-dependant immune processes that could account for the distinct outcome during the phase 1 , namely , when phagocytic dendritic leucocytes of C57BL/6 and DBA/2 mice have been subverted as live amastigotes-hosting cells . Being aware of the very low frequency of the tissues' dendritic leucocytes/DLs , bone marrow-derived C57BL/6 and DBA/2 DLs were first generated and exposed or not to live DsRed2 expressing L . amazonensis amastigotes . Once sorted from the four bone marrow cultures , the DLs were compared by Affymetrix-based transcriptomic analyses and flow cytometry . C57BL/6 and DBA/2 DLs cells hosting live L . amazonensis amastigotes do display distinct transcriptional signatures and markers that could contribute to the distinct features observed in C57BL/6 versus DBA/2 ear pinna and in the ear pinna-DLNs during the first phase post L . amazonensis inoculation . The distinct features captured in vitro from homogenous populations of C57BL/6 and DBA/2 DLs hosting live amastigotes do offer solid resources for further comparing , in vivo , in biologically sound conditions , functions that range from leukocyte mobilization within the ear pinna , the distinct emigration from the ear pinna to the DLN of live amastigotes-hosting DLs , and their unique signalling functions to either naive or primed T lymphocytes .
Leishmania ( L . ) amazonensis perpetuates in South and Central America , its main location being the wet forests of the Amazon basin . The perpetuation of this Leishmania species relies successively on two hosts which cohabit more or less transiently within this ecosystem: blood-feeding sand flies and mammals , including wild rodents and humans . A broad spectrum of clinical manifestations , ranging from single cutaneous lesions to multiple , disfiguring nodules [1] , [2] , [3] assess the durable establisment of L . amazonensis as intracellular amastigotes in the dermis . As model rodents , the laboratory mice of different inbred strains can be subverted as hosts by L . amazonensis , the establishment of parasites in the dermis being more or less rapid . In C3H , BALB/c and C57BL/6 mice high parasite loads , coupled to non healing skin-damages are displayed at site of L . amazonensis inoculation and in multiple skin sites reached by parasites emigrating from the primary inoculation site [4] , [5] , [6] , [7] , [8] . By contrast , in DBA/2 mice , at the inoculation site , the L . amazonensis population size is rapidly controlled , a process coupled to a controlled inflammatory process with limited parasite dissemination in distant tissue ( s ) , if any [9] . Knowing that once in the dermis of the mouse , amastigotes are hosted by mononuclear phagocytes including macrophages and dendritic leukocytes ( DLs ) [10] , [11] , [12] , [13] , [14] , [15] , we have addressed the following question: could the DLs harbouring live amastigotes contribute to the distinct phenotypes observed in C57BL/6 and DBA/2 mice ? Since the frequency of DLs hosting live Leishmania amastigotes within the skin and skin-draining lymph nodes ( DLNs ) remains very low [16] , [17] we decided to first conduct an in vitro study relying on bone marrow-derived DLs ( BMD-DLs ) from C57BL/6 and DBA/2 mice exposed or not to live L . amazonensis amastigotes . Based on flow cytometry ( FCM ) , genechip ( Affymetrix Mouse GeneChip ) and real-time quantitative PCR ( RT-qPCR ) analyses performed on sorted DLs hosting live DsRed2-expressing L . amazonensis transgenic amastigotes [17] many distinct features have been highlighted . DBA/2 DLs displayed transcriptional signatures and markers that could be related to the early phenotype observed in vivo , in contrast to live amastigotes-hosting C57BL/6 DLs . The data are consistent with rapid and sustained immune regulatory functions accounting for the remodeling of the DBA/2 ear as L . amazonensis protective niche . All together this study provides , for the first time , a solid base for exploring i ) the inflammatory processes that maintain the amastigote population under control in DBA/2 mice and ii ) the inflammatory processes coupled to extended parasite dissemination and to poor parasite population control in C57BL/6 mice .
Six week old female DBA/2 , C57BL/6 and Swiss nu/nu mice were purchased from Charles River ( Saint Germain-sur-l'Arbresle , France ) . All animals were housed in our A3 animal facilities in compliance with the guidelines of the A3 animal facilities at the Pasteur Institute which is a member of Committee 1 of the “Comité d'Ethique pour l'Expérimentation Animale” ( CEEA ) - Ile de France - Animal housing conditions and the protocols used in the work described herein were approved by the “Direction des Transports et de la Protection du Public , Sous-Direction de la Protection Sanitaire et de l'Environnement , Police Sanitaire des Animaux under number B75-15-28 in accordance with the Ethics Charter of animal experimentation that includes appropriate procedures to minimize pain and animal suffering . TL is authorized to perform experiment on vertebrate animals ( licence 75-717 ) issued by the Paris Department of Veterinary Services , DDSV ) and is responsible for all the experiments conducted personally or under his supervision as governed by the laws and regulations relating to the protection of animals . DsRed2-transgenic L . amazonensis strain LV79 ( WHO reference number MPRO/BR/72/M1841 ) amastigotes were isolated from Swiss nude mice inoculated 2 months before within a BSL-2 cabinet space as described previously [17] . These amastigotes did not present any antibodies at their surface [18] . Promastigotes derived from amastigotes were cultured at 26°C in complete M199 medium . The metacyclic promastigote population ( mammal-infective stage ) was isolated from stationary phase cultures ( 6 day-old ) on a Ficoll gradient . Ten thousand metacyclic promastigotes in 10 µl of PBS were injected into the ear dermis of C57BL/6 and DBA/2 mice . Increased ear thickness was measured using a direct reading Vernier caliper ( Thomas Scientific , Swedesboro , NJ ) and expressed as ear thickness . DLs were differentiated from bone marrow cells of DBA/2 or C57BL/6 mice according to a method described previously [18] , [19] . Briefly , bone marrow cells were seeded at 4×106 cells per 100 mm diameter bacteriological grade Petri dish ( Falcon , Becton Dickinson Labware , Franklin Lakes , NJ ) in 10 ml of Iscove's modified Dulbecco's medium ( IMDM; BioWhittaker Europe , Verviers , Belgium ) supplemented with 10% heat-inactivated foetal calf serum ( FCS; Dutscher , Brumath , France ) , 1 . 5% supernatant from the GM-CSF producing J558 cell line , 50 U/ml penicillin , 50 µg/ml streptomycin , 50 µM 2-mercaptoethanol and 2 mM glutamine . Cultures were incubated at 37°C in a humidified atmosphere with 5% CO2 . On day 6 , suspended cells were recovered and further cultured in complete IMDM supplemented with 10% of the primary culture supernatant before seeding on day 10 in hydrophobic 6-well plates ( Greiner , St Marcel , France ) at a concentration of 9×105 cells/well in 3 ml complete IMDM . On day 4 post the distribution of DLs in the 6 well plate culture , DLs were exposed or not to freshly isolated DsRed2-LV79 amastigotes or to live BCG at micro-organism-DL ratios of 5∶1 and 10∶1 , respectively . DL cultures were placed at 34°C and sampled at 24 hours post micro-organism addition . Recovered DLs were incubated first in PBS-FCS supplemented with 10% heat-inactivated donkey serum for 15 minutes , second in PBS containing 10% FCS and 0 . 01% sodium azide in presence of antibodies directed against surface antigens . Extracellular staining procedures were performed with specific monoclonal antibodies ( mAbs ) directed against MHC class II molecules ( M5/114 clone ) conjugated to PE-CY5 ( 0 . 2 µg/ml ) and either of the following biotinylated mAbs directed against CD86 ( GL1 clone ) , CD80 ( K-10A1 clone ) , CD54 ( 3E2 clone ) , CD11c ( HL3 ) and IgG control ( B81-3 clone ) at 0 . 5 µg/ml ( eBioscience , San Diego , USA ) . Biotinylated mAbs were revealed using 1 . 5 µg/ml Streptravidin conjugated to Phycoeythrin ( Molecular Probes , Cergy Pontoise , France ) . PE-conjugated mAb directed against CXCR-4 ( 2B11 clone ) was purchased from eBioscience . Analysis was performed on the FACSCalibur . DLs were selected on FSC-SSC parameters ( to excluded debris ) , and on the basis of MHC class II expression to discard the fraction of “contaminating” cells expressing no surface MHC class II molecules . Intracellular staining of amastigotes was performed after fixation in PBS containing 1% paraformaldehyde ( PFA ) for 20 minutes at 4°C with the 2A3-26 mAb which was shown to strictly bind to the L . amazonensis amastigote [18] . DLs were washed in Perm/Wash solution from the BD Cytofix/Cytoperm™ Plus Kit ( BD Bioscience ) and incubated with 5 µg/ml of Alexafluor 488- conjugated 2A3-26 mAb in Perm/Wash buffer for 30 minutes at 4°C in the dark . Then DLs were washed in Perm/Wash buffer and fixed with in PBS −1% paraformaldehyde ( PFA ) . DLs were exposed or not to freshly isolated DsRed2-LV79 amastigotes at a parasite -DL ratio of 5∶1 . DL cultures were placed at 34°C and sampled at 5 , 24 and for 48 hours post parasite addition . Detached DLs were centrifuged on poly-L-lysine-coated glass coverslips and incubated at 34°C for 30 minutes . Cells were then fixed with 4% PFA for 20 minutes , permeabilised with saponin and incubated with 10 µg/ml of the amastigote-specific mAb 2A3-26-AlexaFluor 488 and 1 µg/ml of biotinylated-mAb ( M5/114 ) directed against MHC class II molecules . The revelation was performed using 1 . 5 µg/ml streptravidin conjugated to Texas Red ( Molecular Probes , Cergy Pontoise , France ) . Finally , they were mounted on glass slides with Hoechst 33342-containing Mowiol . Incorporation of Hoechst into DNA allowed the staining of both host cell and amastigote nuclei . Epifluorescence microscopy images were acquired on an upright microscope Zeiss Axioplan 2 monitored by the Zeiss Axiovision 4 . 4 software . DsRed2-LV79 amastigotes were added or not to cultures of C57BL/6 and DBA/2-DLs . Twenty four hours later , three samples collected from three distinct cultures of either unexposed DLs or DLs exposed to DsRed2-LV79 amastigotes were carefully sorted as previously described by Lecoeur et al . [20] . Briefly cells were first incubated in PBS-FCS containing 0 . 2 µg/ml of the anti-MHC class II mAb ( M5/114 ) conjugated to PE-Cy5-conjugated mAb ( eBioscience ) . After two washes , cells were resuspended at 5×106 cells/ml in PBS containing 3% FCS and 1% J558 supernatant . The cell sorting was performed using a FACSAria ( BD Biosciences , San Jose , CA ) equipped with completely sealed sample injection and sort collection chambers that operate under negative pressure . PE-Cy5 and DsRed2 fluorescences were collected through 695/40 and 576/26 bandpass filters respectively . FSC and SSC were displayed on a linear scale , and used to discard cell debris with the BD FACSDiva software ( BD Biosciences ) [17] . L . amazonensis amastigote-hosting DLs were sorted by selecting cells expressing both surface MHC Class II molecules and DsRed2 fluorescence and immediately collected for RNA extraction by using the RNeasy Plus Mini-Kit ( Qiagen ) as previously described [21] . Whatever the readout assays-Affymetrix or RT-qPCR - the RNA populations used were prepared from the same samples . The quality control ( QC ) and concentration of RNA were determined using the NanoDrop ND-1000 micro-spectrophotometer ( Kisker , http://www . kisker-biotech . com ) and the Agilent-2100 Bioanalyzer ( Agilent , http://www . chem . agilent . com ) . Two hundred ng of total RNA per sample were processed , labelled and hybridized to Affymetrix Mouse Gene ST 1 . 0 arrays , following Affymetrix Protocol ( http://www . affymetrix . com/support/downloads/manuals/expression_analysis_technical_manual . pdf ) . Three Biological replicates per condition were run . Following hybridization , the arrays were stained and scanned at 532 nm using an Affymetrix GeneChip Scanner 3000 which generates individual CEL files for each array . Gene-level expression values were derived from the CEL file probe-level hybridization intensities using the model-based Robust Multichip Average algorithm ( RMA ) [22] . RMA performs normalization , background correction and data summarization . An analysis is performed using the LPE test [23] ( to identify significant differences in gene expression between parasite-free and parasite-harbouring DLs , and a p-value threshold of p<0 . 05 is used as the criterion for significant differential expression . The estimated false discovery rate ( FDR ) was calculated using the Benjamini and Hochberg approach [24] in order to correct for multiple comparisons . A total of 1 , 340 probe-sets showing significant differential expression were input into Ingenuity Pathway Analysis software v5 . 5 . 1 ( http://www . ingenuity . com ) , to perform a biological interaction network analysis . The symbols of the modulated genes are specified in the text ( fold change [FC] values between brackets ) , while their full names are given in additional file 1 . MIAME-compliant data are available through GEO database http://www . ncbi . nlm . nih . gov/geo/ accession GSE Total RNAs from DLs cultures were reverse-transcribed to first strand cDNA using random hexamers ( Roche Diagnostics ) and Moloney Murine Leukemia Virus Reverse Transcriptase ( Invitrogen , Life Technologies ) . A SYBR Green-based real-time PCR assay ( QuantiTect SYBR Green Kit , Qiagen ) for relative quantification of mouse target genes was performed on a 384-well plate LightCycler 480 system ( Roche Diagnostics ) . Crossing Point values ( Cp ) were determined by the second derivative maximum method of the LightCycler 480 Basic Software . Raw Cp values were used as input for qBase , a flexible and open source program for qPCR data management and analysis [25] . Relative expression for 8 transcripts ( ccl2 , cl17 , ccl19 , ccr1 , ccr2 , cxcr4 , cd274 , tnfsf4 ) were calculated for sorted LV79-hosting DLs using sorted DLs from Leishmania unexposed cultures as calibrators . For normalization calculations , candidate control genes were tested ( pgk1 , h6pd , ldha , nono , g6pd , hprt , tbp , l19 , gapdh , rpIIe and ywhaz ) with the geNorm [26] and Normfinder programs [27] . Tbp and nono were selected as the most stable reference genes for the C57Bl/6 DLs . RpIIe and tbp were selected for the DBA/2 DLs . At day 4 and 7 post the inoculation of 104 metacyclic promastigotes , three mice were sacrificed , the abundance of some transcripts being determined by real time RT-qPCR . Control , naïve mice were analyzed in parallel . Whole ear pinnas and ears-DLN were removed and fragmented using the Precellys 24 System [21] . Total RNAs were extracted and processed for RT-qPCR as described above . Ldha and nono were selected as the most stable reference genes for the C57Bl/6 and DBA/2 ears . tbp and nono were selected for the as the most stable reference genes for C57Bl/6 DLNs while ywhaz and nono were selected for the DBA/2-DLNs . The experimental procedure for quantifying Leishmania in tissues was done as previously described by de La Llave et al [21] . Briefly , serial 10-fold dilutions of parasites ( from 108 to 101 ) were added to either ears or ear-DLN recovered from C57BL/6 or DBA/2 naive mice . Total RNAs were extracted and processed for RT-qPCR as described above . The primers for Leishmania gene target ( ssrRNA ) to quantify the number of parasites were F- CCATGTCGGATTTGGT and R- CGAAACGGTAGCCTAGAG [28] . A linear regression for each standard curve was determined: number of parasites against the relative expression of ssrRNA values . Two-sided Student's paired t-tests were used to compare FCM experiments ( 4<n<6 ) . A Mann-Whitney test was used to compare ear thickness measurements and number of parasites .
C57BL/6 and DBA/2 mice were given into the ear pinna dermis a low number ( 104 ) of L . amazonensis ( LV79 strain ) metacyclic promastigotes . The monitoring of ear macroscopic features up to 100 days post inoculation ( PI ) has evidenced mouse inbred strain-specific features ( Figure 1 ) . C57BL/6 mice did not display any significant inflammatory signs during the early phase ( ranging from day 0 to day 22 PI , phase 1 ) , whereas they later display sustained inflammatory signs ( after 22 days , phase 2; figures 1A , 1B ) . During the early phase , only a few parasites can be quantified in the ear pinna , the ear pinna-DLN displaying lower number of parasites ( <100 parasites/DLN; figure 1C ) . In contrast , in DBA/2 mouse ear pinna , a mild inflammatory process was observed immediately post the inoculation whereas a rapid increase of the amastigote population size was noted in both the ears and ears-DLN . The second phase was delineated by the persistence of inflammatory process ( Figure 1 ) coupled to the control of parasite load in the ear pinna and ear-DLN ( data not shown ) . We reasoned that early distinct DLs-dependent immune processes- promoting either rapid or slow remodeling of the dermis as amastigote-protective niches- could account for the distinct features displayed , over time , by the L . amazonensis amastigotes-hosting ear pinna of the C57BL/6 and DBA/2 mice . Being aware that , whatever the tissues , the DL frequency is very low , we considered biologically sound to start the comparative analysis with GM-CSF-dependent C57BL/6 or DBA/2 cultured DLs , once they were hosting , or not , live L . amazonensis amastigotes . Briefly , C57BL/6 and DBA/2 bone marrow cell suspensions were exposed or not to live DsRed2 L . amazonensis amastigotes and carefully sorted from otherwise heterogeneous cultures . The immunolabelling of surface MHC class II allowed us to exclude the low fraction of amastigote-hosting cells that did not express surface MHC class II . The subsequent step of such an approach was to first monitor , at the transcriptional level with the Affymetrix-based technology any potential distinct reprogramming of live L . amazonensis amastigotes-hosting DLs . We used a carefully designed in vitro model [20] based on cultures of mouse BMD-DLs in which more than 97% of cells expressed CD11c , CD11a and CD11b ( data not shown ) . When the presence/absence of surface MHC class II molecules was monitored on whole cell cultures by fluorescence microscopy and FCM , three phenotypically distinct cell subsets were evidenced ( Figures S1A–C ) . The population of cells that did not express surface and intracellular MHC class II molecules were considered as “Contaminating” Cells ( CC ) . The two other cell populations partition between i ) a majority of cells displaying a moderate surface MHC class II amount ( MHC IIlow; bona fide immature DLs ) and ii ) a minority of cells expressing very high levels of MHC II molecules ( MHC IIhigh; bona fide mature DLs ) . DsRed2 L . amazonensis/LV79 amastigotes were put in contact with BMD-DLs ( MOI of 5/1 ) and analysed 5 , 24 or 48 hours later ( Figure 2 ) . Intracellular amastigotes ( 2A3-26+ ) detected by immunofluorescence microscopy analysis were evidenced in all BMD-DL subsets with much higher number of amastigotes in CC ( data not shown ) . Low percentages of DLs hosting 2A3-26+ parasites were also documented by FCM analyses at 24 hours post amastigote addition ( 23 . 0%+/−12 . 6 and 26 . 0%+/−8 . 1 of 2A3-26+ cells in C57BL/6 and DBA/2 BMD-DLs , respectively , for n = 9 experiments ) . Interestingly , while the percentage of DLs housing amastigotes did not change from 5 hours to 24 hours ( Figure 2A ) , the number of intracellular amastigotes did slowly expand whatever the mouse genotype ( Figure 2B ) over the otherwise limited temporal window we did focus on . L . amazonensis amastigote-hosting DLs were sorted by selecting cells expressing both surface MHC Class II molecules and DsRed2 fluorescence ( see below ) .
|
The rapid and long term establishment of parasites such as L . amazonensis , otherwise known to strictly rely on subversion of macrophage and dendritic leucocyte ( DL ) lineages , is expected to reflect stepwise processes taking place in both the skin dermis where the infective form of the parasite and the skin-draining lymph node ( DLN ) were inoculated . Relying on mice of two distinct inbred strains—C57BL/6 and DBA/2—that rapidly and durably display distinct phenotypes at the two sites of establishment of L . amazonensis , we were curious to address the following question: could live L . amazonensis-hosting DL display unique signatures that account for the distinct phenotypes ? Based on flow cytometry , genechip and real-time quantitative PCR analyses , our results did evidence that , once subverted as cells hosting live L . amazonensis , DLs from C57BL/6 or DBA/2 do display distinct profiles that could account for the i ) distinct parasite load profiles , ii ) as well as the distinct macroscopic features of ear pinna observed once the L . amazonensis metacyclic promastigotes completed their four day developmental program along the amastigote morphotype .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] |
[
"genome",
"expression",
"analysis",
"immune",
"cells",
"antigen-presenting",
"cells",
"immunology",
"microbiology",
"host-pathogen",
"interaction",
"parasitology",
"immune",
"defense",
"t",
"cells",
"biology",
"pathogenesis",
"immune",
"response",
"immunity",
"genomics",
"genetics",
"and",
"genomics"
] |
2012
|
Distinct Transcriptional Signatures of Bone Marrow-Derived C57BL/6 and DBA/2 Dendritic Leucocytes Hosting Live Leishmania amazonensis Amastigotes
|
Neural crest cells are multipotent progenitor cells that can generate both ectodermal cell types , such as neurons , and mesodermal cell types , such as smooth muscle . The mechanisms controlling this cell fate choice are not known . The basic Helix-loop-Helix ( bHLH ) transcription factor Twist1 is expressed throughout the migratory and post-migratory cardiac neural crest . Twist1 ablation or mutation of the Twist-box causes differentiation of ectopic neuronal cells , which molecularly resemble sympathetic ganglia , in the cardiac outflow tract . Twist1 interacts with the pro-neural factor Sox10 via its Twist-box domain and binds to the Phox2b promoter to repress transcriptional activity . Mesodermal cardiac neural crest trans-differentiation into ectodermal sympathetic ganglia-like neurons is dependent upon Phox2b function . Ectopic Twist1 expression in neural crest precursors disrupts sympathetic neurogenesis . These data demonstrate that Twist1 functions in post-migratory neural crest cells to repress pro-neural factors and thereby regulate cell fate determination between ectodermal and mesodermal lineages .
Neural Crest Cells ( NCCs ) are multi-potent progenitor cells , which after delaminating from the dorsal lip of the neural tube and migrating throughout the developing embryo , can differentiate along either ectodermal or mesodermal lineages [1]–[4] . For example , a subset of NCCs , the cardiac NCCs ( cNCCs ) invades the aorticopulmonary cushions ( APC ) and septum ( AoPS ) of the developing cardiac outflow tract ( OFT ) , the conduit through which blood exits the ventricles [5]–[8] . There , these cells assume a mesodermal identity , differentiating into connective tissue and smooth muscle and septating the pulmonary trunk and aorta to divide systemic and pulmonary circulation . Alternatively , the trunk , vagal , and sacral NCCs assume ectodermal identities , differentiating into the sympathetic , parasympathetic , and enteric neurons of the autonomic nervous system [9] . The transcriptional mechanisms that regulate NCC fate choice between these ectodermal and mesodermal lineages are not known . NCCs are specified along a rostro-caudal axis into distinct subpopulations that have limited capacity to change their cell fate [10] . However , both the cardiac and the rostral-most vagal neural crest originate at the same axial level ( somites 1–3 ) , suggesting that additional mechanisms of cell fate determination beyond axial specification are necessary to distinguish these ectodermal and mesodermal lineages . Indeed , both differentiating sympathetic neurons and cNCCs have been shown to respond to local secreted signaling cues , notably Bone Morphogenetic Proteins ( BMPs ) , subsequently upregulating transcriptional effectors such as the Twist family bHLH proteins , Hand2 and Hand1 , and initiating differentiation programs [11] , [12] . The mechanisms that enable post-migratory NCCs to interpret these local signaling cues and to undergo either ectodermal or mesodermal differentiation programs are not understood . In Drosophila , the transcription factor twist is essential for mesoderm development . In vertebrates; however , despite the evident requirement for Twist1 in the development of multiple mesodermally-derived organ systems , an analogous molecular mechanism by which Twist1 might control mesenchymal cell fate choice has not been defined [13] . Previously , we have shown that Twist1−/− knockouts display cNCC phenotypes at E11 . 5 , including abnormally compacted cell aggregations in the APC ectomesenchyme . Cell-lineage and marker analyses show that these aggregations are NCC-derived and robustly express Hand1 and Hand2 [14] . We hypothesized that these mutant phenotypes reflect aberrant NCC fate choice , and sought to better define these NCC aggregates . To precisely examine Twist1 function during cNCC maturation , we conditionally inactivated Twist1 in NCCs at specific stages of Mus musculus ( mouse ) development using the Wnt1-Cre [15] and Hand1Cre [16] alleles , which respectively express Cre recombinase in pre-migratory and post-migratory cNCCs . Both cNCC-specific Twist1 ablation models display OFT defects; however , pre-migratory Twist1 deletion causes these defects with greater penetrance and severity , compared to post-migratory deletion . In contrast , the presence of NCC-derived nodules is evident in all Twist1 conditional knockouts ( CKOs ) , regardless of the timing of gene ablation . Expression analyses show that Twist1 CKO nodules are molecularly similar to sympathetic ganglion neurons . We find that , unlike that of bona fide sympathetic ganglion neurons , ectopic neuron formation in Twist1 CKOs is independent of Hand2 function , indicating that these neurons are a distinct cell population . Similar to Hand1 and Hand2 , Twist1 can molecularly interact with both the pro-neural Paired-like homeobox transcription factor Phox2b and HMG box transcription factor Sox10 . Sox10 is upstream of Phox2b in the Bmp-dependent sympathetic neuron transcriptional program [17] . Here , we confirm that Sox10 regulation of Phox2b is direct . We demonstrate that Sox10 trans-activates the Phox2b promoter , and that Twist1 represses this transactivation . Twist1 itself can bind to an evolutionarily conserved non-canonical E-Box in the Phox2b promoter . Mutation of a C-terminal domain known as the Twist-box disrupts the ability of Twist1 to molecularly interact with Sox10 , bind DNA , and transcriptionally repress Phox2b . Embryos harboring this Twist-box mutation ( the Charlie Chaplin Twist1 allele ) [18] similarly display ectopic neurons in their OFTs . Indeed , the appearance of these ectopic neurons is dependent upon Phox2b , as Phox2b ablation rescues ectopic neuron formation in Twist1 mutants . Finally , we show that ectopic Twist1 expression in NCCs , using a conditionally activatable CAG-CAT-Twist1 transgene activated by Wnt1-Cre , leads to sympathetic ganglia containing fewer neurons , in which Tyrosine Hydroxylase ( TH ) , Phox2b , and Hand2 expression is diminished or absent , demonstrating that Twist1 expression is sufficient to disrupt normal neurogenic developmental programs . Together , these data suggest that Twist1 represses neuronal cell fate choice in the cNCC by repressing transcription of Phox2b , and reveal a fundamental mechanism controlling ectodermal versus mesodermal cell fate choice in NCCs .
To assess Twist1 function during cNCC maturation , we conditionally inactivated Twist1 in pre-migratory NCCs using the Wnt1-Cre driver . Twist1 in situ hybridization of E10 . 5 embryos confirmed effective Wnt1-Cre-mediated NCC-specific gene deletion ( Figure S1 ) . Conditional ablation of Twist1 in pre-migratory NCCs produces abnormal cellular aggregates , indistinguishable from those observed in systemic Twist1 knockouts [14] , in the APCs of the cardiac OFT . Lineage trace analyses using the ROSA26R reporter allele confirms that these nodules are NCC-derived ( data not shown ) and establishes a cell-autonomous requirement for Twist1 in developing cNCCs . At mid-gestation ( E11 . 5 ) , normal NCCs in the APCs generate extracellular matrix ( ECM ) that can be visualized by Alcian Blue staining ( Figure 1A ) . The cNCC aggregations observed in Twist1;Wnt1-Cre CKO embryos are devoid of ECM ( arrowheads , Figure 1B ) . In situ hybridization analysis demonstrates that expression of Sox9 , a transcriptional regulator of ECM components , is abnormally absent from the aggregations ( arrowheads , Figure 1D ) , but is expressed normally within non-phenotypic cNCCs . In addition to Sox9 , PlexinA2 , Smad6 , Hey2 and Pdgfrα , which are normally expressed in APC mesenchyme , are also excluded from the NCC aggregates ( Figure S2 ) . Thus , the cNCC aggregates in the cardiac OFT are molecularly distinct from the phenotypically normal cNCCs fated to differentiate along a mesodermal lineage and to ultimately contribute to the AoPS and valves . As reported previously , the feature that distinguishes these cNCC aggregates from normal cNCC mesenchyme is the expression of Hand1 and Hand2 , which accounts for approximately 30–40% of the cNCC within the E11 . 5 OFT [14] . Indeed , neither Hand1 nor Hand2 expression is readily detectible within non-aggregated cNCCs in Twist1−/− embryos [14] . As the NCC aggregations do not express the expected ecto-mesenchymal markers , we assessed the expression of genes associated with other NCC cell fates . The Bmp-dependent gene Sox10 is expressed in NCC-derived neuronal progenitors [12] . In situ hybridization analysis demonstrates that Sox10 mRNA is not detected in the APCs of control embryos; however , it is expressed in the Twist1;Wnt1-Cre CKO aggregates ( Figure 1F ) . Immunohistochemical analyses revealed that these aggregates also express the pan-neuronal marker Tubulin β-III ( Tubb3 ) , confirming their identity as ectopic neurons ( Figure 1H ) . As NCCs contribute to a number of distinct neuronal populations , we sought to determine the cellular identity of the ectopic neurons in the Twist1;Wnt1-Cre CKO OFTs . Our previous data showed that Hand1 and Hand2 , both markers of cardiac NCCs , are readily detectable in the OFT aggregates of Twist1−/− mutant embryos [14] . We confirmed this finding in Twist1;Wnt1-Cre CKOs ( Figure 2B , 2D ) . Importantly , Hand1 and Hand2 are only robustly co-expressed in NCC-derived sympathetic ganglion neurons ( SGNs ) , and no other neuronal cell type [9] , [19] , [20] . We therefore hypothesized that these ectopic neurons may express other SGN markers . In situ hybridization revealed that NCC aggregates express transcriptional regulators of sympathetic neurogenesis , including the transcription factors Phox2a ( data not shown ) and Phox2b [21] ( Figure 2F ) , the bHLH transcription factor Ascl1 [22] ( Figure 2H ) and the zinc finger transcription factor Gata3 [23] ( Figure 2J ) . These cells also express the norepinephrine biosynthetic enzymes Tyrosine Hydroxylase ( TH; Figure 2L ) and Dopamine β Hydroxylase ( DBH; Figure 2N ) indicating that , like SGNs , these neurons are noradrenergic . Ectopic Phox2b- and Ascl1-positive NCCs are detectable in the Twist1;Wnt1-Cre CKO pharyngeal arches as early as E10 . 5 , and persist at least until E16 . 5 ( Figure S3 ) . To further confirm that these neurons most resemble sympathetic neurons , we assessed a panel of sensory neuron markers , including TrkA , Brn3a , NeuroD1 , and Runx1 . We found that ectopic OFT neurons do not express these sensory neuron markers ( Figure S4 ) . We also sought to distinguish these ectopic neurons , which are noradrenergic , from parasympathetic neurons , which are predominantly cholinergic . We therefore analyzed the cholinergic neuron markers Ret , VaCHT , and ChAT . Ret and VaCHT are expressed in both parasympathetic neurons , and in sympathetic neurons during their early development [24] , and both are detectable in ectopic OFT neurons ( Figure S4 ) . Conversely , the parasympathetic-specific marker ChAT is not detectable above background levels in ectopic OFT neurons ( Figure S4 ) . Given that Hand1 and Hand2 are not expressed within parasympathetic neurons [25] , [26] , we conclude that the ectopic neurons in the Twist1;Wnt1-Cre CKO APCs are most molecularly similar to sympathetic neurons . In addition to the formation of ectopic OFT sympathetic-like neurons , Twist1;Wnt1-Cre CKOs also display completely penetrant persistent truncus arteriosus ( PTA; Figure S5G , S5K; Table S1 ) . These structural OFT defects are presaged by the complete absence of detectable Semaphorin 3c ( Sema3c ) expression in Twist1;Wnt1-Cre CKO cNCCs ( Figure S6 ) . Sema3c is required for septation of the OFT [27] , [28] and may be directly regulated by Twist1 [29] . As PTA can result from either defective cell migration or NCC differentiation , a breakdown of either of these two mechanisms could cause the ectopic sympathetic-like neurons in the Twist1;Wnt1-Cre CKO APCs . To distinguish between these two possibilities , we deleted Twist1 in post-migratory cNCCs . NCC fate choice is largely determined by axial level of origin [2] , [4] , [10] . Twist1 is expressed in NCCs immediately following delamination from the neural tube and throughout subsequent migration . Loss of Twist1 disrupts cranial NCC migratory pathways [30] . To test whether loss of Twist1 either causes presumptive neural progenitor NCCs to similarly mis-migrate into the APC mesenchyme or alters cNCC fate choice , we examined Twist1 function during post-migratory cNCC maturation . To this end , we conditionally inactivated Twist1 using the Hand1Cre knock-in allele [16] , which expresses Cre recombinase in post-migratory NCCs ( Figure S5C , S6D ) . Twist1;Hand1Cre CKOs show a markedly reduced penetrance of PTA ( Figure S5L , S5M; Table S1 ) . This suggests that , for the OFT to septate properly , Twist1 function is required to regulate cNCC migration to the OFT during early development . In contrast , the formation of ectopic neurons in E11 . 5 Twist1;Hand1Cre CKO OFTs occurs with 100% penetrance ( Figure 3B and 3D , arrowheads ) . In situ hybridization and immunohistochemical analyses show that ectopic neurons express Sox10 , Tubb3 ( data not shown ) , TH ( Figure 3B ) and Ascl1 ( Figure 3D ) . Thus , when Twist1 function is ablated in Hand1-lineage cNCCs that have completed migration , a subset of these cells then differentiate into neurons . These data indicate that ectopic neurons form in Twist1 mutant embryos via a novel post-axial specification mechanism whereby post-migratory cNCCs trans-differentiate . Additionally , although Sema3c expression was also disrupted in a majority of Twist1;Hand1Cre CKO OFTs ( n = 3/4 ) , unlike Twist1;Wnt1-Cre CKOs , the penetrance was not 100% . Indeed , Sema3c expression was unaffected , even when ectopic Ascl1-positive cells were detectable ( Figure S6 ) . Collectively , these data demonstrate that OFT septation and repression of cNCC trans-differentiation are distinct Twist1-associated phenotypes , and that this Hand1-lineage-resticted trans-differentiation does not preclude the remaining cNCCs from maintaining characteristic cNCC gene expression ( Figure S2 ) and differentiating normally once they arrive within the OFT . To identify the transcriptional effectors of sympathetic neurogenesis that Twist1 regulates , we first explored genetic interactions with the related bHLH factor Hand2 . Sympathetic neurogenesis depends upon Hand2 function [26] , [31] , [32] . As Twist1 and Hand2 are functionally antagonistic in the developing limb [33] we investigated a similar mechanism in the OFT . As expected , at E11 . 5 , NCC-specific Hand2 ablation results in fewer DBH-positive neurons ( compare Figure 4A1 and 4C1 ) , reduced TH expression ( data not shown ) , and complete loss of Hand1 expression ( compare Figure 4E1 with 4G1 ) in the forming SGNs . Twist1;Hand2 double CKOs display similarly reduced SGN specification ( Figure 4D1 and 4H1 ) . Surprisingly , ectopic sympathetic-like neurons persist in the OFTs of E11 . 5 Twist1;Hand2 double CKOs ( Figure 4D and 4H ) . These ectopic neurons robustly express TH ( data not shown ) , DBH ( Figure 4D2 ) , and Hand1 ( Figure 4H2 ) . Endogenous Hand1 SGN expression is directly dependent upon Hand2 function [26] , [32] , [34] . Although the ectopic neurons in the Twist1 CKO APCs are molecularly similar to SGNs , their specification and expression of SG-specific markers is Hand2-independent , thereby distinguishing them as a separate neuronal cell population . Thus , Twist1 antagonism of Hand2 does not repress ectopic neurogenesis in developing cNCCs . We have identified a Hand2-dependent Hand1 cis-regulatory enhancer sufficient to drive reporter gene expression in the developing and adult SGNs [34] . To validate that the ectopic neurogenesis in Twist1 CKO APC mesenchyme is truly via a non-canonical mechanism , we assessed the expression of this SGN-specific enhancer ( Hand1SG+hsp68-lacZ ) on a Twist1;Wnt1-Cre CKO background . Although this SGN-specific reporter transgene is normally expressed by the endogenous E11 . 5 SGNs of both control and Twist1;Wnt1-Cre CKO embryos ( Figure 4 , 4I1 and 4J1 ) , the Hand1 SGN-specific enhancer is not transcriptionally active in Twist1;Wnt1-Cre CKO ectopic neurons , despite the presence of its direct regulators , Hand2 and Phox2b [34] Figure 4J2 ) . Thus , although the ectopic neurons in Twist1;Wnt1-Cre CKO APCs express Hand1 ( Figure 2B; Figure 4F2; [14] , this expression is driven by a Hand2-independent enhancer distinct from the one that drives SGN-specific expression . Thus , the molecular pathways regulating gene expression in the ectopic neurons of Twist1;Wnt1-Cre CKO APCs are distinct from those endogenously regulating sympathetic neurogenesis . Interestingly , Hand2-independent Hand1 expression in the Twist1;Wnt1-Cre CKO ectopic OFT neurons suggests that Hand1 could be functionally redundant with Hand2 , and that loss-of-function of both Hand1 and Hand2 would “rescue” the ectopic neurons in Twist1;Wnt1-Cre CKOs . To test this , we generated Twist1fx/−;Hand1fx/−;Hand2fx/−;Wnt1-Cre ( + ) triple CKOs ( Figure S7 ) . Triple CKO embryos were rarely viable past E10 . 5 . Nevertheless , histological analysis clearly shows ectopic OFT neurons , further confirming that , although the 30–40% of cNCC undergoing neuronal trans-differentiation are marked by Hand1 and Hand2 [14] , this aberrant NCC cell fate trans-differentiation caused by Twist1 loss-of-function is completely Hand factor independent . As Hand factors are dispensable for ectopic neuron formation in the Twist1 CKO APCs , we sought an alternative pro-neural factor ( s ) with which Twist1 interacts to repress neurogenesis . Hand2 and Phox2 proteins molecularly interact to synergistically affect critical transcriptional programs during sympathetic neurogenesis [26] , [31] , [32] , [35] . Phox2b is molecularly upstream of Hand2 , and these two factors auto-regulate each other via a feed-forward mechanism [21] , [22] , [36] . As Twist1 and Hand2 are closely related , we used co-immunoprecipitation to test for a possible interaction between Twist1 and Phox2b . Similar to Hand2 , Twist1 molecularly interacts with Phox2b ( Figure S8A and S8B ) . Twist1 molecularly interacts with both Runx2 [18] and Runx3 [36] through its carboxy-terminal Twist-box domain . Mutation of the Twist-box ( S192P ) impairs Runx molecular interactions . [18] Co-immunoprecipitation analyses using epitope-tagged Twist1 S192P showed that protein interactions between Twist1 and Phox2b are , in part , dependent upon the Twist-box domain ( Figure S8A and S8B ) . However , transactivation assays in HeLa cells revealed that Twist1 does not significantly inhibit Phox2b auto-regulation , and thus did not provide compelling evidence that Twist1-mediated antagonism of neurogenesis is mediated though direct inhibition of Phox2b protein function . We therefore sought additional pro-neural factors with which Twist1 might interact . Phox2b is not normally expressed in wild-type APCs ( Figure 2E ) . Its presence in the APCs of Twist1 mutant embryos suggests that Twist1 could either directly repress Phox2b expression or repress the activity of a key Phox2b transcriptional regulator . Recently , Twist1 was shown to interact with and repress the function of the transcription factor Sox9 [37] . The related factor Sox10 functions directly upstream of Phox2b in the Bmp-dependent pathway that drives sympathetic neurogenesis [17] . To assess whether Twist1 can also interact with Sox10 and potentially inhibit its transcriptional activity , we performed co-immunoprecipitation assays . Twist1 molecularly interacts with Sox10 , and although Twist1 S192P displays a comparatively reduced interaction with Sox10 ( Figure 5A , 5B , asterisk , 48 . 3%+/−5 . 1% , p-value = 0 . 002 ) , Twist1 T125;S127A , a Saethre-Chotzen Syndrome-associated Helix I dimerization mutant , does not ( Figure 5A , 5B , 102 . 7%+/−12 . 0% , p-value = 0 . 85 ) . Thus , as has been reported for Sox9 [37] , Sox10 also interacts with Twist1 in a Twist-box dependent manner . Sox proteins bind the consensus sequence WWCAAWG [38] . Bioinformatic analyses revealed putative HMG box cis-elements in the Phox2b 5′ promoter that are evolutionarily conserved ( Figure 5C , highlighted in blue ) . Sox10 is required for Phox2b expression in vivo [17] . Given the presence of evolutionarily conserved putative Sox10 cis-elements in the Phox2b promoter ( Figure 5C ) , we tested whether Sox10 might transactivate Phox2b directly . Indeed , Sox10 significantly upregulates gene expression from the Phox2b promoter by 48 . 9+/−2 . 2-fold ( Figure 5D , p-value = 0 . 0002 ) . Co-expression of Twist1 inhibits Sox10-mediaded transactivation , reducing its transcriptional activation to 26 . 3+/−1 . 9-fold ( Figure 5D , p-value = 0 . 0003 ) . Co-expression of Twist1 S192P significantly ameliorates this Sox10 inhibition , restoring Sox10 transactivation to 49 . 3+/−3 . 9-fold ( Figure 5D , p-value = 0 . 92 ) . Together , these data indicate that Twist1 molecularly interacts with Sox10 through its Twist-box domain , and partially represses trans-activation of the Phox2b promoter . Twist1 , through its Twist-box domain , inhibits DNA binding of both Runx2 [18] and Sox9 [37] . We therefore reasoned that Twist1 might similarly inhibit Sox10 binding to the Phox2b promoter . We therefore first confirmed Sox10 DNA binding to the conserved HMG boxes in the Phox2b 5′ promoter through electrophoretic mobility assays ( EMSA ) using radiolabeled oligonucleotides ( Figure 5E ) and in vitro translated Myc-tagged Sox10 ( Figure 5F ) . Sox10 binds to the proximal two HMG boxes , termed HMG 3 and HMG 4 ( Figure 5G , asterisks ) , but not the distal two HMG boxes , termed HMG 1 and HMG 2 ( data not shown ) . We conclude that Sox10 can bind to conserved HMG boxes in the Phox2b promoter , supporting that Sox10 directly regulates Phox2b expression . We then synthesized in vitro translated Myc-tagged Sox10 in the presence of increasing amounts of Myc-tagged Twist1 ( Figure 5F ) . EMSAs revealed that Twist1 does not disrupt Sox10 DNA binding to HMG 3 and HMG 4 ( Figure 5G , asterisks ) . However , Twist1 does bind within HMG 3 to a non-canonical cis-element ( Figure 5G , open circle ) . Given this surprising result , we then tested whether Twist1 can bind directly to other elements within the Phox2b promoter . Twist1 binds to consensus sites termed E-Boxes ( CANNTG ) [39] , [40] . Bioinformatic analyses uncovered one conserved E-Box and one conserved E-Box-like element within the Phox2b 5′ promoter ( Figure 5C , highlighted in yellow ) . We in vitro translated Twist1 and Twist1 S192P protein ( Figure 5H ) to assess Twist1 DNA binding to these E-Boxes . As controls , we included three mutant forms of Twist1 in which conserved arginines in the basic domain have been mutated ( Figure 5H ) . Of these three mutants , Twist1 R116W [41] , Twist1 R118H [42] , and Twist1 R120A [43] , the former two are associated with the congenital disorder Saethre-Chotzen Syndrome , and all are predicted to have impaired DNA binding capabilities [44] . Although Twist1 displayed no binding to the perfect consensus E-Box 2 ( data not shown ) , it robustly bound to the non-canonical E-Box 1 ( Figure 5I , open circle ) . Twist1 did not detectably bind to a mutated ( mut ) E-Box 1 oligo ( Figure 5I ) , in which the E-Box core was disrupted ( Figure 5I ) . As predicted , none of the Twist1 basic domain mutants bound to E-Box 1 . Twist1 S192P does bind to E-Box 1; however , it binds more weakly when compared to WT Twist1 . Given that equivalent amounts of Twist1 and Twist1 S192P were added to each EMSA , the strength of DNA-binding is quantitative ( see [39]; Figure 5H and 5I ) . Together , these data suggest that the Twist-box is critical not only for Twist1 protein interactions with Sox10 , but also for Twist1 DNA binding . Thus , Twist1 inhibition of Phox2b transcription is likely mediated through direct DNA binding of non-canonical , conserved E-Box elements in the Phox2b promoter . Moreover , Twist1 could directly interact with the potent Phox2b trans-activator Sox10 while both factors are bound to the promoter . As Phox2b is considered a master regulator of autonomic neurogenesis , [9] we hypothesized that disruption of the Twist-box in vivo would lead to aberrant Phox2b activation , and , consequently , the appearance of ectopic neurons in the developing OFT similar to those observed in Twist1 CKO embryos . The Charlie Chaplin Twist1 allele ( Twist1CC ) is a Twist1 S192P point mutation that specifically disrupts function of the Twist1 Twist-box domain . Embryos harboring this mutant allele exhibit craniofacial and limb abnormalities [18] . To test the hypothesis that Twist1-mediated repression of Phox2b is necessary to repress ectopic neurogenesis in the developing OFT , we examined the OFT phenotypes of both heterozygous and homozygous Twist1CC embryos . Twist1CC/CC mutants lack overt structural OFT phenotypes ( Figure S5I , S5M ) . Indeed , Sema3c expression is not disrupted in either Twist1CC/+ or Twist1CC/CC mutants ( Figure S6F , S6G ) . These data suggest that the Twist-box is dispensable for proper cNCC migration and OFT morphogenesis . However , as predicted by the necessity of the Twist-box for robust Sox10 interactions and binding to the Phox2b promoter , ectopic OFT neurons are detectable in the APCs of both Twist1CC/CC homozygous mutants and Twist1CC/+ heterozygotes . In Twist1CC/+ heterozygous embryos , dispersed Phox2b- ( Figure 6B ) and Ascl1-positive ( data not shown ) cells are detectable in the APC ectomesenchyme . Relatively few of these cells are TH-positive , however , suggesting that these ectopic cells fail to undergo complete neuronal differentiation ( Figure 6E , arrowhead ) . Twist1CC/CC homozygous mutants display more robust ectopic ganglia that are positive for Phox2b ( Figure 6C ) , Ascl1 ( Figure 7A ) , and TH ( Figure 6F ) . Collectively , these data show that Twist1 interacts with Sox10 and Phox2b at least partially via its Twist-box domain , and that this domain is required to inhibit cNCC trans-differentiation into ectopic sympathetic-like neurons . To confirm that Phox2b upregulation is absolutely required to modulate post-migratory NCC cell fate choice in Twist1 mutants , we assessed ectopic neurogenesis in E11 . 5 Twist1;Phox2b doubly mutant embryos . Robust ectopic aggregates of Ascl1-expressing cells were evident in Twist1CC/CC homozygous mutants ( Figure 7A ) . Interestingly , the size of these aggregates appeared reduced when Phox2b gene dosage was attenuated to heterozygosity ( Figure 7B ) . It is not technically possible to count individual staining-positive cells following in situ hybridization . We therefore validated our results using morphometric analyses . Quantification of the Ascl1 staining-positive area in each sequential section of these mutant embryos revealed that , while a reduction of Phox2b gene dosage had no appreciable effect on Ascl1 staining in 5 out of 11 Twist1CC/CC;Phox2blacZ/+ mutants examined , 4 out of 11 of these mutants showed reduced ectopic Ascl1 staining comparable to that seen in control ( Phox2blacZ/+ ) APCs ( Figure 7D ) . Ectopic neurons were not detectable in either Twist1CC/+ or Twist1CC/CC mutants when Phox2b function was completely ablated ( Figure 7C , data not shown ) . Collectively , these results validate the hypothesis that Twist1 ultimately represses the proneural activity of Phox2b in NCCs by inhibiting its transcription both via direct molecular antagonism of Sox10 activity and through binding to the Phox2b promoter . These findings provide the first evidence , to our knowledge , that the potential ectodermal and mesodermal cell fates of post-migratory NCCs can be regulated through functional antagonism between transcription factors . If Twist1 is a true repressor of ectodermal cell fate , then ectopic expression of Twist1 in NCCs should inhibit differentiation of endogenous SGNs . To test whether Twist1 can repress sympathetic neurogenesis , we used a conditionally activatable transgene ( CAG-CAT-MycTwist1 ) [40] to ectopically express Twist1 in the NCC progenitors of SGNs . TH expression , as revealed through immunohistochemistry of E12 . 5 embryos , was either absent in thoracic sympathetic chain ganglia ( Figure 8B ) , or was restricted to a few cells ( Figure 8C ) . Co-localization of Twist1 , visualized via a Myc epitope tag , and TH was not observed ( Figure 8C ) . Tubb3 immunohistochemistry and Phox2b in situ hybridization analyses confirm that the thoracic sympathetic ganglia in CAG-CAT-MycTwist1 ( + ) ;Wnt1-Cre ( + ) embryos are either absent ( Figure 8E , 8K ) , or markedly reduced ( Figure 8F , 8L ) . In the Control thoracic sympathetic chain , Sox10-expressing presumptive support cells surround the ganglia , which are mostly , but not entirely , Sox10-negative ( Figure 8G ) . In E12 . 5 CAG-CAT-Twist1 ( + ) ;Wnt1-Cre ( + ) transgenic embryos , the hypoplastic thoracic sympathetic chain ganglia are either entirely positive for Sox10 ( Figure 8K , arrowhead ) or display a markedly reduced core of Sox10-negative cells ( Figure 8I , arrowhead ) . Hand2 mRNA expression is also drastically reduced in E12 . 5 CAG-CAT-Twist1 ( + ) ;Wnt1-Cre ( + ) thoracic sympathetic chain ganglia , and is occasionally absent , even when neurons in adjacent sections express Phox2b , suggesting that normal SGN regulatory cascades are disrupted in Twist1 mis-expressing SGNs ( Figure 8N , 8O ) . These findings , in combination with Twist1 loss-of-function and genetic interaction analyses , demonstrate that Twist1 is a potent repressor of sympathetic neurogenesis , and that Twist1 antagonizes downstream Bmp targets to act as a novel post-migratory cell fate switch in the NCCs that populate the cardiac OFT .
Here , we describe a novel cell fate switch that regulates post-migratory NCC differentiation to either an ectodermal or a mesodermal cell fate . The bHLH transcription factor Twist1 is expressed in migratory and post-migratory cNCCs , but is not expressed in the NCC-derived SGNs . Loss of Twist1 in Wnt1-Cre-expressing NCCs results not only in structural OFT defects , but also the formation of dense aggregates of sympathetic-like ganglia in the APCs of the cardiac OFT . These neurons express Sox10 , Phox2b , Ascl1 , Gata3 , Hand1 , Hand2 , TH and DBH , which are all components of the BMP-induced SGN differentiation program , but not markers of sensory neurons or specific markers of parasympathetic neurons . Our earlier work shows that these ectopic neurons are marked by both Hand1 and Hand2 expression , and account for approximately 30–40% of E11 . 5 OFT cNCCs [14] . Deletion of Twist1 in the smaller , post-migratory Hand1Cre lineage-derived subpopulation of cNCCs reveals that this ectopic neurogenesis is a bona-fide cell fate trans-differentiation from a mesenchymal to neuronal cell type , and is not a consequence of the altered migration of trunk NCCs . The penetrance of OFT structural defects , such as PTA , is greatly reduced in embryos in which Twist1 is deleted using the Hand1Cre , indicating that NCC migration-dependent phenotypes are largely rescued in this model ( Table S1 ) . Thus , cells fated to contribute to the smooth muscle and valves of the OFT can be reprogrammed to an SGN-like fate in the absence of Twist1 . Both Wnt1-Cre and Hand1Cre Twist1 CKOs survive until birth and , in some Hand1Cre Twist1 CKOs , exhibit no significant OFT defects other than ectopic neurons . Furthermore , although they have completely penetrant ectopic neurons , Twist1CC mutants do not display either structural OFT defects or diminished Sema3c expression . In conjunction with normal marker analysis ( Figure S2 ) , these data suggest that Twist1 function is solely required in specific subpopulations of cNCCs , and that unaffected cNCCs correctly follow their developmental programs in the absence of Twist1 . Mechanistically , Twist can physically interact with other proteins using both its bHLH and Twist-box domains [18] , [37] , [45] , [46] . Twist1 antagonistically interacts with Hand2 [33] . Although loss of Hand2 in NCCs dramatically effects the development of the endogenous sympathetic chain , as previously reported [26] , [32] , the trans-differentiation of the SGN-like neurons in the OFT was not affected . Furthermore , Hand1 expression , which is directly dependent upon Hand2 function in SGNs [34] , is maintained , but its expression is not mediated through its identified SGN-specific enhancer ( Figure 4 ) . Although we have demonstrated that the Hand1 SG-enhancer is not auto-regulated by Hand1 [34] , we analyzed Twist1;Hand2;Hand1 triple CKO mice . Ectopic neurons remain histologically evident in these mutants , demonstrating that the trans-differentiation of cNCCs to an SGN-like fate is independent of Hand factors and , by consequence , represents a non-canonical transcriptional program . Hand factor expression , instead , appears to define which cNCCs convert to ectopic neurons , as opposed to those that retain their normal developmental programs . Indeed , these studies raise questions concerning the developmental origin of neurons innervating the heart . Although it is known that the NCCs contribute neurons to the arterial pole of the heart [47]–[49] , the origin of these neurons and the mechanisms regulating their development are poorly understood . The NCCs migrating to the arterial pole of the heart were thought to possess limited neurogenic potential [50] . Nonetheless , a subset of NCCs migrating to the caudal pharyngeal arches is thought to contribute to the cholinergic cardiac ganglia within the parasympathetic plexus [51] . These neurons are initially observed at E11 . 5 , but are not consistently and robustly detectable until E12 . 5 [50] . We observe noradrenergic ganglia , identified through their expression of TH and DBH , in the APCs of Twist1 mutant embryos at E11 . 5 . The presumptive NCC progenitors of these neurons , identified through Sox10 , Phox2b , Ascl1 , Hand2 and Hand1 expression , are evident in the E10 . 5 Wnt1-Cre CKO OFTs a day earlier , and in a much larger proportion of NCCs . It is notable that not all of the NCCs occupying the Twist1 mutant APCs trans-differentiate into neurons . As stated above , the only established distinguishing characteristic of this subpopulation of cells is that they express the two Hand genes , while the remaining , ostensibly unaffected cells do not ( Figure 2B; [14] ) . The significance of this observation is not clear , as Hand factors are not required for the differentiation of these ectopic neurons . Furthermore , it is unclear whether this trans-differentiation constitutes a corruption of endogenous neuronal differentiation pathways , or whether within a specific subpopulation of NCCs migrating to the APCs , which are competent to differentiate into neurons , Twist1 functions to repress neurogenic pathways , preventing them from differentiating in such a manner . It would thus be of interest to further explore the differences inherent in NCCs competent to differentiate into neurons in the APCs , and those that are not . The Twist-box domain of Twist1 interacts with Runx2 , Runx3 and Sox9 [18] , [37] , [46] . These interactions repress the respective functions of these non-bHLH factors by interfering with their DNA binding [18] , [37] , [46] . Here , we show that Twist1 ( via the Twist-box domain ) can both interact with the potent trans-activator of Phox2b transcription Sox10 and inhibit transcriptional activation of the Phox2b promoter ( Figure 5A–5D ) . This repression is not mediated through simple interference with Sox10 DNA-binding to the Phox2b promoter ( Figure 5G ) . Rather , Twist1 itself can bind to conserved , non-canonical E-box elements within the Phox2b promoter in a manner that is Twist-box dependent ( Figure 5I ) . This finding is the first evidence that Twist-box mutations can influence Twist1 DNA binding . Co-immunoprecipitation experiments indicate that the Twist1 S192P mutation does not compromise Twist1 homodimerization , ( J . Vincentz and A . Firulli , unpublished results ) . This suggests that this impaired DNA binding is not the result of defective Twist1 dimerization , but may instead result from a conformational change to the Twist1 secondary structure that impairs juxtaposition of the two required basic DNA-binding domains within the Twist1 S192P homodimer . It should also be noted that the cis-element to which Twist1 binds does not conform to a typical E-box sequence . The sequence , CAGGTT , is a putative binding site for the zinc-finger transcriptional repressor Snai1 ( Snail ) [52]–[54] . Snai1 and Twist1 can genetically interact in Drosophila and mouse [55] , [56] . Like Twist1 , Snai1 and the related factor Snai2 ( Slug ) is a crucial regulator of epithelial-mesenchymal transition ( EMT ) [12] . Indeed , Snai1/Snai2 , in the presence of Sox9 , is sufficient to induce an EMT in neural epithelial cells [57] . Thus , functional interactions between both Twist1 and Snai1/Snai2 with Sox proteins during NCC development would be an intriguing avenue of further study . Phox2b is both necessary and sufficient to drive the SGN cell fate program [22] , and Phox2b activation is dependent upon Sox10 [17] . The Twist1CC mouse model shows that disruption of the Twist-box is sufficient to promote cNCC trans-differentiation . Normal cNCC fate choice is partially restored when Phox2b gene dosage is reduced to heterozygousity and completely rescued when Phox2b function is completely ablated , demonstrating its necessity to initiate SGN cell fate . Additionally , Twist1 can interact with the bHLH factor Ascl1 , as determined via co-immunoprecipitation analysis ( data not shown ) . Collectively , Twist1 then interacts with at least five of the key SGN cell fate transcriptional regulators ( Figure S9 ) . Given that both cNCC and SGN NCC transcriptional programs are initiated by Bmp-signaling , and that these two programs , at least in part , share key transcriptional regulators ( Hand1 , Hand2 , Gata3; [58] , Sox4 [59] , and ( likely ) Sox11 [60] , [61] ) , it is obvious that a switch to ensure that the correct cell fate is specified and maintained is built into these developmental programs . Twist1 is a strong candidate to fulfill such a role . Indeed , when Twist1 is expressed throughout the Wnt1-Cre-expressing NCCs , we observe a reduction in the differentiation of SGNs as indicated by the mutual exclusivity of Myc-Twist1 and TH immunoreactivity , decreased expression of the pan-neuronal marker Tubb3 and loss of expression of both Hand2 and , more importantly , Phox2b . Additionally , the majority of the cells remaining in these hypoplastic ganglia express Sox10 ( Figure 8H , 8I ) . We can draw two conclusions from this observation . First , the presence of these Sox10-positive cells , along with that of Myc-Twist1-positive cells ( Figure 8B , 8C ) , confirms that NCC-derived progenitors retain the ability to migrate properly to the region proximal to the dorsal aorta when Twist1 is ectopically expressed . Second , as Sox10 is initially broadly expressed in NCC-derived neural precursors , but , by E12 . 5 , is largely restricted to the surrounding support cells , we can infer either that ectopic Twist1 expression in NCCs specifically disrupts sympathetic neurogenesis , but not gliogenesis , or that Sox10 is not downregulated in Twist1-overexpressing NCC-derived neural precursors , and these cells therefore fail to differentiate . In either case , this data clearly demonstrates that ectopic Twist1 expression in neuronal precursor cells fundamentally impairs sympathetic neurogenesis , and potentially maintains these cells in a precursor state . It is intriguing to speculate about the insight these findings may provide into the role of TWIST1 in cancer . Reactivation of TWIST1-regulated embryonic programs has been proposed to contribute to tumor progression [62] . TWIST1 induces EMT in tumor cells , and thus plays a dominant role in defining the metastatic potential of primary tumors [63] . In both embryonic and adult stem cells , Twist1 is also thought to prevent differentiation , thus promoting a stem cell-like phenotype [13] . Thus , Twist1 represents a single factor that can intimately link two cellular processes , EMT and maintenance of a stem cell-like fate , both integral to tumor cell progression . As our data demonstrates that Twist1 can functionally antagonize key regulators of a sympathetic neuronal identity , it would be of interest to closely examine the function of TWIST1 in neuroblastomas and other tumors affecting neuronal derivatives of NCCs . TWIST1 overexpression in N-Myc-amplified neuroblastomas has been shown to inhibit p53-dependent apoptosis [64] . Pheochromocytomas are neoplasms originating from NCC-derived adrenal chromaffin cells . Chromaffin cells have a molecular profile similar to that of SGNs . Like SGNs , chromaffin cells do not normally express Twist1 . TWIST1 is frequently upregulated in pheochromocytomas and , intriguingly , this aberrant expression is tightly associated with malignancy in these tumors [65] . Thus , the model presented here , by which Twist1 represses neuronal cell identity in NCC derivatives , may ultimately shed light upon the role of TWIST1 in cancer .
Genotyping of the Twist1tm1Bhr ( Twist1; [66] ) , Twist1tm2Bhr ( Twist1fx; [67] provided by James Martin ) , Tg ( Wnt1-cre ) 11Rth ( Wnt1-Cre; [68] ) , Hand1tm1 . 1 ( EGFP/cre ) Abfi ( Hand1eGFPCre; [16] ) , Hand2tm1Cse , ( Hand2fx; [32] ) , CAGCAT-Twist1; [40] , Hand1SG-hsp68-lacZ; [34] , Hand1tm2Eno ( Hand1lx; [69] provided by Eric Olson ) and Gt ( ROSA ) 26Sortm1Sor ( R26RlacZ; [70] alleles was performed as described . A 125bp region containing the mutated nucleotide in the Twist1Ska10 ( Twist1CC; [18] ) allele was amplified using the primers Twist1CC ( F ) , 5′-ACGAGCTGGACTCCAAGATG-3′ , and Twist1CC ( R ) , 5′-GGAGCTCCGCTGCTAGTG-3′ . Amplicons were then purified and sequenced . Phox2btm1Jbr ( Phox2blacZ; [71] provided by Michelle Southard-Smith and Jean-FranÇois Brunet ) mice and embryos were genotyped using the primers Phox2bEx2 ( F ) , 5′-GTTCAGTGGCCCTTCACATC-3′ , Phox2bEx2 ( R ) , 5′-TCCTCTCACGGGACACTTCT-3′ , and lacZ_5′_out , 5′-CGGAAACCAGGCAAAGCGCC-3′ to generate ∼500 bp WT and ∼250 bp mutant amplicons . Alcian Blue , Nuclear Fast Red , Hematoxylin and Eosin ( H&E ) staining were performed as described [14] , [72] . X-gal staining was performed as described [16] . 1Section in situ hybridizations were performed on 10 µm paraffin sections as described [14] , [72] . Antisense digoxygenin-labeled riboprobes were synthesized using T7 , T3 or SP6 polymerases ( Promega ) and DIG-Labeling Mix ( Roche ) using the following plasmid templates: Twist1 ( provided by Richard Behringer ) , Hand1 , Hand2 , Sema3c , Runx1 , Hey2 ( provided by Yasuhide Furuta ) , Sox9 ( provided by Benoit De Crombrugghe ) , Sox10 ( provided by Paul Trainor ) , Ascl1 ( provided by Xin Zhang ) , Phox2b , VaCHT ( provided by Peter Cserjesi ) , Gata3 , Ret , DBH ( provided by Jean-FranÇois Brunet ) , PlexinA2 , Pdgfrα , Smad6 ( provided by Jonathan Epstein ) , TrkA ( provided by David Ginty ) , Brn3a , NeuroD1 ( both provided by Eric Turner ) , and ChAT ( IMAGE clone #8734071 ) . Morphometric analyses of Ascl1 staining were performed as described [73] . Immunohistochemistry was performed as described [73] α-Tubb3 ( AbCam ) , α-TH ( AbCam ) , and α-Myc ( Sigma ) antibodies were used in combination with DyLight secondary antibodies ( Thermo Scientific ) . Co-immunoprecipitation experiments were performed in HEK 293 cells using α-Myc and α-Flag ( Sigma ) as described [73] . Rat Sox10 was affixed with an N-terminal 6X Myc-tag via cloning into pCS2+MT . Densitometry analyses were performed using BioRad Quantity One software . Luciferase assays were performed in HeLa cells using the dual luciferase assay kit ( Promega ) as per manufacturer's instructions . 2 µg total DNA ( 0 . 25 µg of either pCS2-FLAG , pCS2-FLAG+Phox2b , or pRK5-FLAG+Sox10 ( provided by Brian Black ) , 0 . 5 µg of either pcDNA3 . 1 , pcDNA3 . 1-FLAG+Twist1[WT] , or pcDNA3 . 1-FLAG+Twist1 S192P , 1 . 25 µg of PHOX2b ( HindIII/NcoI ) -pGL3b ( provided by Diego Fornasari ) , and 0 . 125 µg of pRL-CMV ) was transfected in 6-well plates using X-tremeGENE HP transfection reagent ( Roche ) . Cell lysates were read using a 96-well micro-titer plate luminometer ( Thermo Labsystems ) . Data represent four independent experiments . Error bars denote standard error . All sequences were obtained via Ensembl BLASTN search ( http://www . ensembl . org ) using the human PHOX2b 5′ promoter as a point of reference . PATTERNMATCH and CLUSTALW analyses were performed using the SDSC Biology WorkBench ( http://workbench . sdsc . edu ) , EMSAs were performed as previously described [74] with minor alterations . In vitro transcription and translation of Sox10 and Twist1 mRNAs were performed using pCS2-MT+Sox10 and pCS2-MT+Twist1 expression plasmids and the TnT rabbit reticulocyte lysate in vitro transcription system ( Promega ) as per manufacturer's instructions . 5 µL of TnT was used per reaction . Radiolabeled , annealed probes were purified using mini Quick Spin Oligo Columns ( Roche ) . 1 µg poly ( dG-dC ) was used as a nonspecific DNA-binding competitor . Reactions were incubated for 30 min at room temperature following addition of probe . The following oligos , annealed to their complements , were used: E-Box 1 , 5′-ACACTCTTACAAAACAGGTTTTCTATGACATCAAGGTTTCTTC-3′; E-Box 1 ( Mut ) , 5-ACACTCTTACAAAAtgGGacTTCTATGACATCAAGGTTTCTTC-3′; HMG 3 5′-TTTCCAAGTAGTGTGATTGAATTAAAGGGCAGGGA-3′; HMG 4 5′-TGGTATTAAATTCTAATTAGAGATGCAGGAATCAATGATAGGGAGGTTGGACAGCT-3′; E-box 2 , 5′-AAGACCAACCGCTTTGCTATTGTCCAAGTGGAAAGAGCCAAGTTTATTATGAGG-3′ . An oligo featuring a mutated HMG Box ( HMG 4 ( Mut ) 5′-TGGTATTAAATTCTAATTAGAGATGCAGGAATatcTGATAGGGAGGTTGGACAGCT-3′ ) was used as an unlabeled competitor . Animal work ( mouse ) was performed according to an approved animal protocol from the University of Indiana IACUC , which is an AAALAC accredited program . We strive to focus on the three Rs ( reduction/refinement/replacement ) when working with animal models .
|
During vertebrate development , a unique population of cells , termed neural crest cells , migrates throughout the developing embryo , generating various cell types , for example , the smooth muscle that divides the aorta and pulmonary artery where they connect to the heart , and the autonomic neurons , which coordinate organ function . The distinctions between neural crest cells that will form smooth muscle and those that will become neurons are thought to occur prior to migration . Here , we show that , in mice with mutations of the transcription factor Twist1 , a subpopulation of presumptive smooth muscle cells , following migration to the heart , instead mis-specify to resemble autonomic neurons . Twist1 represses transcription of the pro-neural factor Phox2b both through antagonism of its upstream effector , Sox10 , and through direct binding to its promoter . Phox2b is absolutely required for autonomic neuron development , and indeed , the aberrant neurons in Twist1 mutants disappear when Phox2b is also mutated . Ectopic Twist1 expression within all neural crest cells disrupts the specification of normal autonomic neurons . Collectively , these data reveal that neural crest cells can alter their cell fate from mesoderm to ectoderm after they have migrated and that Twist1 functions to maintain neural crest cell potency during embryonic development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"biology",
"molecular",
"cell",
"biology"
] |
2013
|
Twist1 Controls a Cell-Specification Switch Governing Cell Fate Decisions within the Cardiac Neural Crest
|
Mid-hindbrain malformations can occur during embryogenesis through a disturbance of transient and localized gene expression patterns within these distinct brain structures . Rho guanine nucleotide exchange factor ( ARHGEF ) family members are key for controlling the spatiotemporal activation of Rho GTPase , to modulate cytoskeleton dynamics , cell division , and cell migration . We identified , by means of whole exome sequencing , a homozygous frameshift mutation in the ARHGEF2 as a cause of intellectual disability , a midbrain-hindbrain malformation , and mild microcephaly in a consanguineous pedigree of Kurdish-Turkish descent . We show that loss of ARHGEF2 perturbs progenitor cell differentiation and that this is associated with a shift of mitotic spindle plane orientation , putatively favoring more symmetric divisions . The ARHGEF2 mutation leads to reduction in the activation of the RhoA/ROCK/MLC pathway crucial for cell migration . We demonstrate that the human brain malformation is recapitulated in Arhgef2 mutant mice and identify an aberrant migration of distinct components of the precerebellar system as a pathomechanism underlying the midbrain-hindbrain phenotype . Our results highlight the crucial function of ARHGEF2 in human brain development and identify a mutation in ARHGEF2 as novel cause of a neurodevelopmental disorder .
Brain development depends on spatiotemporally controlled gene expression . [1–3] Alterations in the expression pattern of such genes can result in neurodevelopmental disorders by impinging on key processes such as neural progenitor specification , cell division , and differentiation and the migration of newly born neurons from their site of origin to their final destination within the brain . [4–6] The latter is crucial for the formation of specific brain structures . [7–9] Factors that control localized gene function include Rho GTPase regulators . Here , we present evidence that the loss of function of Rho guanine nucleotide exchange factor 2 ( ARHGEF2 ) causes a human neurodevelopmental disorder characterized by intellectual disability , mild microcephaly , and midbrain-hindbrain malformation . ARHGEF2 ( synonym GEF-H1 , murine Lfc ) catalyzes the replacement of GDP to GTP bound to Rho-related proteins and thereby controls timing and localization of the activation of Rho GTPases such as RhoA . [10–14] ARHGEF2 connects microtubule and actin cytoskeleton dynamics . [14] In this context ARHGEF2 activity is reduced through microtubule binding and further controlled by upstream regulators . [15–25] ARHGEF2 is key for actin and microtubule reorganization and is required for mitotic spindle formation and orientation . [11] Inhibition of ARHGEF2 results in spindle disorientation and dysfunction , mitotic delay , accumulation of prometaphase cells , and further mitotic aberrations . [11 , 22] In mouse neocortex , Arhgef2 is expressed in neural precursor and immature neurons and regulates neurogenesis from cortical precursor cells . [24] Arhgef2 down-regulation by shRNA keeps radial precursors cycling , potentially by disrupted spindle plane orientation , and thereby inhibits neurogenesis . In contrast , Arhgef2 overexpression causes an increase of neurons in the cortical plate . [24–26] Arhgef2 also plays a role in neural tube closure by regulating morphogenetic movements . [27] Furthermore , Arhgef2 participates in the migration of non-neuronal cells and in Wnt-induced planar cell polarity , via the activation of RhoA . [28 , 29] Although evidence for a central function of Arhgef2 in cytoskeletal dynamics and critical signal transduction pathways exists and other ARHGEF genes have been linked with neurological disease , [30–32] little is known about ARHGEF2 function in humans and no disease phenotype associated with this gene has been reported .
We report that patients with a homozygous mutation in ARHGEF2 develop intellectual disability , mild microcephaly , and midbrain-hindbrain malformations . Two affected children of healthy , consanguineous parents of Kurdish-Turkish descent were born at term without complications after an uneventful pregnancy ( II . 1; II . 2 , Fig 1A ) . At birth , mild congenital microcephaly with occipitofrontal head circumferences ( OFC ) of -1 . 95 ( II . 1 ) and -2 . 33 ( II . 2 ) SDS ( standard deviation score ) but normal weight and height were apparent ( S1 and S2 Tables ) . In addition , the two boys variously displayed wide intermamillary distance , broad fingers , low posterior hairline , and facial dysmorphism with long philtrum , thin upper lip , high palate , downslanted palpebral eye fissures , long eyelashes , bilateral ptosis , and horizontal pendular nystagmus ( S1 Table ) . Ophthalmological examination revealed congenital strabismus , astigmatism , amblyopia ( II . 2 ) , optic disc pallor ( II . 2 ) , abnormalities of the retinal pigment epithelium ( II . 2 ) , and abnormal visual-evoked potentials further underlining optic nerve affection . Motor milestones were not severely delayed despite generalized muscular hypotonia observed in patient II . 1 and weak tendon reflexes in both children . Both children stumbled frequently and had a disturbance of fine motor movements ( S3 Table ) . Moderate intellectual disability ( IQ <50 ) and severe developmental speech delay despite normal hearing capacity were diagnosed in both patients ( S1 and S3 Tables ) . Cranial magnetic resonance imaging ( MRI ) revealed mild microencephaly , elongated midbrain , hypoplasia of the pons , ventral and dorsal longitudinal clefts ( grooves ) in pons and medulla , and inferior vermis hypoplasia ( Fig 1B ) . The finding of grooves and cerebellar hypoplasia ( in absence of the ‘dragonfly sign’ with hypotrophy of the cerebellar hemispheres rather than of the vermis ) argued against the differential diagnosis of pontocerebellar hypoplasia . Results of routine blood tests , extensive metabolic work-up , chromosome analysis , and cardiac function assessment were normal . To identify the genetic cause of this disease , we performed whole exome sequencing and bioinformatic analysis followed by Sanger sequencing in both affected individuals . We thereby identified a homozygous deletion of a single guanine ( G ) at the exon12-intron12 boundary ( Fig 2A and 2B ) , causing deletion of a G at the exon12-exon13 transition from normally spliced cDNA and ultimately a frameshift mutation in the affected children ( c . 1461delG , NM_004723 . 3 , Fig 2C ) . The mutation lies within a highly conserved region , results in predicted truncation of the protein ( p . D488Tfs*11 , NP_004714 . 2 , Fig 2D ) , and segregates with the disease phenotype , i . e . , was heterozygous in both parents and not observed in healthy controls . We failed to detect the deletion in various whole exome databases and did not detect biallelic ARHGEF2 mutations in six further patients , as detailed in the methods chapter . In lymphoblastoid cell lines ( LCLs ) , generated from the patients , ARHGEF2 mRNA levels were decreased as can be expected by partial nonsense-mediated decay ( Fig 2E , n = 3 , One-way ANOVA ) . Moreover , ARHGEF2 protein levels were virtually absent in patient cells and decreased to intermediate levels in cells from the heterozygous parents ( Fig 2F , n = 3 , One-way ANOVA ) . The predicted truncated form of ARHGEF2 protein was below detection levels in patient LCLs ( S1 Fig ) , indicating the total loss of ARHGEF2 in patient cells . To evaluate the biological effects of the identified ARHGEF2 mutation on brain development , we utilized both in vitro and in vivo approaches . We used a well-characterized cell culture model of E13 mouse neocortical cells recapitulating the timing of cortical precursor differentiation observed in vivo . [24 , 33 , 34] The large majority of these cells are actively dividing radial glial precursors positive for the neural progenitor marker nestin immediately following plating . These cells then differentiate to generate excitatory neurons ( DIV1-3 ) and subsequently astrocytes and oligodendrocytes ( DIV5-7 ) . We demonstrated previously , using this model , that knockdown of Arhgef2 causes a decrease in the proportion of βIII-tubulin positive neurons and a corresponding increase of cycling cortical precursors . [24] We thus expected the human ARHGEF2 frameshift mutation to have a similar impact on neurogenesis , leading to an increase of cycling ( and eventually apoptotic ) precursors and a decrease of neurons generated during brain development . In a first series of experiments , we transfected murine cortical radial precursors with an Arhgef2 shRNA plasmid and a nuclear EGFP reporter plasmid; the latter allows the visualization of transfected cells . As expected , Arhgef2 knockdown significantly decreased the proportion of double positive cells for EGFP and early neuron marker βIII-tubulin , whereas it increased the proportion of EGFP and proliferation marker Ki67 double positive precursors ( Fig 3A , n = 1378–1293 cells , One-way ANOVA ) . Next , we co-transfected cortical progenitors with the same combination of plasmids and an additional construct encoding for human wildtype ARHGEF2 . In this condition , we observed no significant change in the proportion of EGFP and βIII-tubulin double positive cells or EGFP and Ki67 as compared to untransfected cells , demonstrating that wildtype ARHGEF2 rescues the phenotype produced by shRNA ( Fig 3A ) . Last , we co-transfected cortical progenitors with Arhgef2 shRNA , the EGFP reporter , and a construct encoding mutant ARHGEF2 identified in the index patients . Here , the proportion of EGFP and βIII-tubulin double positive cells was reduced , whereas the number of EGFP and Ki67 remained increased ( Fig 3A ) . We conclude that the phenotype produced by Arhgef2 shRNA was rescued by coexpression of wildtype but not of mutant human ARHGEF2 , consistent with the interpretation that the patient mutation is a loss-of-function mutation and that this ARHGEF2 mutation impairs neurogenesis . To further analyze whether the identified ARHGEF2 loss-of function mutation affects brain development in vivo we in utero electroporated E13 . 5 mouse cortex with Arhgef2 or Con ( control ) shRNA constructs along with wildtype or mutant human ARHGEF2 plasmids as well as an EGFP vector . EGFP as a marker for electroporated cells and SatB2 as a marker for newborn neurons in murine cortices were quantified three days after electroporation . We identified a pronounced increase in the proportion of EGFP-positive cells in the ventricular , sub-ventricular , and intermediate zones ( VZ/SVZ/IZ ) and decreased proportions in the cortical plate ( CP ) of the Arhgef2 shRNA electroporated mice , as compared to mice that were electroporated with the EGFP reporter only ( Fig 3B and 3C ) . In line with our in vitro data , overexpression of wildtype , but not mutant , ARHGEF2 was able to effectively rescue the cell distribution in the cortex induced by Arhgef2 shRNA . Upon electroporation of Arhgef2 shRNA , we additionally observed a dramatic reduction in the proportion of SatB2/EGFP double positive cells localized to the cortical plate . This is consistent with impaired neurogenesis described above in cultured cortical precursor cells . Overexpression of wildtype ARHGEF2 was able to rescue this impairment of neurogenesis while mutant ARHGEF2 was ineffective ( Fig 3B and 3C , n = 3 , One-way ANOVA ) . These data further confirm that the patient mutation in human ARHGEF2 acts as a loss-of-function mutation also in vivo , results in dramatically decreased neurogenesis and is most probably disease-causative ( Fig 3D ) . ARHGEF2 has been associated previously with spindle plane orientation , [11 , 24] and the latter has been shown to play a crucial role for the fate of neuronal precursors , ‘deciding’ between self-renewal and differentiation . [35] To evaluate whether human mutant ARHGEF2 alters neurogenesis by impairing mitotic spindle orientation , we performed in utero electroporation with the experimental setup described above and determined spindle plane orientation . We determined the angle between the ventricular surface and a line connecting centrosomes in metaphase/anaphase cells in brain sections . Arhgef2 shRNA led to a significant shift of the spindle orientation towards a more horizontal orientation , characterizing putatively more symmetrically , self-renewing cell divisions . Again , overexpression of human wildtype but not mutant ARHGEF2 was able to rescue significantly this phenotype ( Fig 4A , n = 68–70 cells from 3–4 mice , One-way ANOVA ) . In the developing brain , ARHGEF2 reportedly controls the arrangement for symmetric or asymmetric cell divisions of apical progenitors through plane orientation . [24] Given our results from overexpressing mutant ARHGEF2 in the mouse brain , we concluded that loss of ARHGEF2 activity inhibits neurogenesis by favoring more symmetric divisions of neocortical progenitors ( Fig 4B ) . To further substantiate the effect of a loss of ARHGEF2 functions in human tissues and other cell types , we analyzed cell cycle apparatus in LCLs from our index patients and healthy controls . We detected an abnormal morphology of the mitotic spindle apparatus , decreased spindle pole distance , and decreased cell size in LCLs derived from the patients compared to controls , significantly ( Fig 5A and 5B , n = 100–200 cells , One-way ANOVA ) . There was , however , no cell cycle defect , increased radiosensitivity , nor abnormal centrosome ‘morphology’ in the patient LCLs ( S2 and S3 Figs ) . We concluded that regulation of the mitotic spindle apparatus is a primary function of ARHGEF2 in the cell cycle also in human tissues and other cell types . ARHGEF2 regulates various cellular processes through activation of Rho family GTPases , specifically RhoA and its downstream effectors , such as RhoA/ROCK/MLC pathway members . [12 , 22 , 28 , 36 , 37] Specifically , RhoA has been implicated in the regulation of neurogenesis and planar cell divisions . [38 , 39] We thus analyzed the activity of RhoA and its immediate downstream effectors in patient and control LCLs . We detected a significant reduction of active RhoA and , consistently , reduced levels of active phospho-myosin light chain ( MLC ) in the LCLs of affected individuals with an ARHGEF2 mutation compared to healthy individuals ( Fig 5C–5E , n = 3 , One-way ANOVA ) . This indicates that the RhoA/ROCK/MLC pathway is most likely impaired in humans with the ARHGEF2 mutation . Previous studies showed that Arhgef2 is expressed in the neural tube of mice from embryonic day 11 on and maintained at high levels during brain development . [24] Our in situ hybridization studies on embryonic mice showed strong expression of Arhgef2 transcripts in the neuroepithelium of telencephalon , diencephalon , and rhombencephalon of E11 mice ( S5 Fig ) . At the time of birth , Arhgef2 is maintained in the germinal zones of both the neocortex and the cerebellum , as well as in the pontine gray nuclei ( PGN ) ( S5 Fig ) . To further corroborate the pathogenicity of the identified ARHGEF2 mutation in brain development , we analyzed the phenotype of Arhgef2 deficient mice . [40] In adult Arhgef2 mutant mice , we observed a significant reduction in volume of the total brain size ( referred to as microencephaly ) , the cerebellum and the brainstem , as well as the striking absence of the pontine nuclei ( Fig 6 , n = 3–4 , Student’s t-test ) . We concluded that the phenotypes observed in the nervous system of Arhgef2 deficient mice recapitulate largely those malformations observed in the index patients . Our results showed a clear correspondence between regions displaying high levels of Arhgef2 expression and those pathologically affected in the index patients with ARHGEF2 mutation . To gain further insight in the function of ARHGEF2 in brain development , we set out to analyze the cerebral cortex , midbrain , cerebellum , and hindbrain of Arhgef2 mutant mice . First , no drastic anatomical abnormality was found in the neocortex of Arhgef2 mutant mice , as assessed by the measurement of its volume , thickness , and surface area ( S6 Fig ) . Though it should be noted that there was a trend towards a reduction in the analyzed parameters in the mutant mice , which could contribute to the mild microcephaly phenotype observed in Arhgef2 mutant mice . Similarly , no major alteration on the distribution of neurons in the cortical layers was found in Arhgef2 mutant mice ( S6 Fig ) . Next , we analyzed the integrity of midbrain structures by immunostaining the inferior and superior colliculus with antibodies against the transcription factor FoxP2 . This revealed no significant alteration in development of midbrain structures in Arhgef2 mutant mice ( S7 Fig ) . To address the cerebellar phenotype observed in Arhgef2 mutant mice and patients , we measured the size of Purkinje cells and the thickness of the cerebellar molecular layer in Arhgef2 mutant and control mice . Our analysis revealed no significant difference in these parameters between wildtype and mutants ( S8 Fig ) . This led us to hypothesize that the input to the cerebellum , from precerebellar nuclei , could be affected , and thus contributing to the observed reduction of the cerebellar size . During hindbrain development , precerebellar neurons emerge from the proliferative neuroepithelium of the rhombic lip and migrate tangentially to reach their final destinations in the pons and medulla . These neurons form five distinct nuclei that are located at different positions within the hindbrain: pontine gray nuclei ( PGN ) , reticulotegmental nuclei ( RTN ) , external cuneate nuclei ( ECN ) , lateral reticular nuclei ( LRN ) , and inferior olivary nuclei ( IO ) [41–44] . The PGN , RTN , ECN , LRN projects mossy fibers to the cerebellum , whereas IO projects climbing fibers , and thereby provide input to the cerebellum . [41–44] To tackle the hypothesis that deficits in development of precerebellar nuclei result in abnormalities in cerebellar size of Arhgef2 mutant mice , we immunostained sagittal and coronal sections taken from Arhgef2 homozygous and Arhgef2 heterozygous ( used as control ) mice with antibodies against the transcription factor Mbh2 ( which labels PGN , RTN , ECN , LRN ) and FoxP2 ( that marks IO ) . This analysis revealed that Arhgef2 mutant mice completely lack , in addition to PGN , the RTN and have a severe reduction in LRN neurons ( Figs 7A , 7B , 7D , 8A , 8B and 8D ) . Interestingly , the rostral part of the ECN was abnormally enlarged in Arhgef2 mutant mice when compared to controls , suggesting that some PGN , RTN and LRN might fail to migrate to their ventral positions and aberrantly locate in the ECN ( Figs 7C , 8C and 8D ) . The formation of the IO was not obviously disturbed in Arhgef2 mutant mice ( Fig 7D ) . Taken together , our anatomical analysis revealed that specific precerebellar nuclei are either absent or severely reduced in Arhgef2 mutant mice . These findings draw our attention to two possible interpretations on the role of Arhgef2 in development of precerebellar nuclei: ( 1 ) Arhgef2 modulates the generation of precerebellar neurons and/or ( 2 ) Arhgef2 regulates the migration of these cells from their progenitor niche towards their final localization . Precerebellar neurons arise from discrete progenitor domains located in the developing dorsal hindbrain , which are defined by the combinatorial expression of the transcription factor Olig3 with other bHLH ( basic Helix-Loop-Helix ) transcription factors , such as Atoh1 ( Protein atonal homolog 1 ) or Ptf1a ( Pancreas transcription factor 1 alpha ) . [45] Precerebellar neurons that form the PGN , RTN , LRN and ECN are generated from Olig3/Atoh1 positive ( known as dA1 ) progenitors , whereas those that form the IO emerge from Olig3/Ptf1a positive ( known as dA4 ) progenitors , [46 , 47] reviewed in [45] . Arhgef2 is unlikely to modulate generation of precerebellar neurons as its transcript becomes expressed only by E11 in mice , a time point when these cells have already been specified and have started their migration . We thus speculated that Arhgef2 rather modulates precerebellar neuron migration . To test this possibility , we co-stained transverse hindbrain sections , taken from wildtype mice at E11 , with an Arhgef2 in situ hybridization probe and antibodies against Olig3 and Mbh2 ( BarH-like homeobox 1 ) . This revealed that Arhgef2 is co-expressed largely with Mbh2-positive cells emanating from dA1 progenitors and throughout the anterior and posterior extramural streams precerebellar neurons in the mantle zone of hindbrain ( S9 Fig ) . Arhgef2 does not follow the expression pattern of Olig3 in dorsal hindbrain ( S9 Fig ) . It is interesting to note that mutation of Arhgef2 appears to selectively affect the migration of dA1-derived neurons , as other hindbrain centers such as the nucleus of the solitary tract ( a Olig3+/Ascl1+ dA3 derivative ) or the IO ( dA4 derivative ) were not affected ( Figs 7E and 8E ) . This report demonstrates that intellectual disability , mild microcephaly , and a midbrain-hindbrain defect can be caused by a homozygous mutation of the ARHGEF2 in humans . We highlight the importance of ARHGEF2 during brain development and , in particular , in neuronal progenitor cell division and differentiation , as well as in neuronal migration . We show that in both human and murine cells ARHGEF2 deficiency interferes with the normal orientation of mitotic spindles and cell fate choices . As expected , the frameshift mutation of ARHGEF2 deregulates the activity of its downstream effectors , i . e . the RhoA/ROCK/MLC pathway . ARHGEF2 thereby joins other members of the same protein family that have already been associated with neurologic disease such as non-syndromic intellectual disability ( ARHGEF6 ) , [30] epileptic encephalopathy ( ARHGEF9 ) , [31] and peripheral demyelinating neuropathy ( ARHGEF10 ) . [32] We show that Arhgef2 deficiency in mice severely impairs the migration of dA1 progenitors , which results in the incorrect location of precerebellar neurons in the hindbrain of Arhgef2 mutant mice . In keeping with our data , previous studies have shown that deficits in the RhoA/ROCK pathway affect migration of precerebellar neurons . [48] Thus , increasing evidence support the notion that the ARHGEF2-RhoA/ROCK pathway is essential for the migration and specification of precerebellar neurons . ARHGEF2 has been shown to be involved in Wnt-mediated planar cell polarity pathway through its interaction with the Daam and Dishevelled proteins , which are known to control migratory events . Wnt1 mutant mice have been reported to have a midbrain-hindbrain malformation , [49] and humans with biallelic WNT1 mutations displayed cerebellum and brainstem phenotype . [50] It is interesting to note that the major effects in the hindbrain of Arhgef2 mutant mice occur in the migration of dA1-derived neurons , but not other neuronal cell types that emerge from neighboring progenitor domains . Our data demonstrate thus a marked specificity in the molecular control of neuronal migration in the hindbrain . The combined midbrain-hindbrain malformation phenotype observed in humans was not fully recapitulated in Arhgef2 deficient mice , which lacked an obvious midbrain phenotype . This suggests that during evolutionary divergence of these species , additional molecular mechanisms coordinate the function of ARHGEF2 . Identifying additional families with ARHGEF2 mutations will help to consolidate the disease-causative role of ARHGEF2 in humans .
The human study was approved by the local ethics committees of the Charité ( approval no . EA1/212/08 ) , and the written consent form was received from the investigated individuals . All animal experiments were carried out in accordance to the guidelines of national ethic principles , and approved by Charité ( registration no . T0344/12 ) . DNA samples of the two affected individuals were subjected to whole exome sequencing . Five μg genomic DNA were enriched with the Agilent Human All Exon V3 kit ( Agilent , Santa Clara , CA , USA ) following the manufacturer’s protocol . The libraries were sequenced using Illumina HiSeq 2000 sequencer for single-end 101 bp . Coverage of coding regions was >92 . 6–93 . 4% with a minimal depth of 20-fold; the raw data were processed as described previously . [51] In brief , the raw sequences were aligned to human reference genome ( hg19 ) , and the alignment results were used to call variants of SNVs , indels , and CNVs . We defined from the exome data the regions of homozygosity by at least four high-quality ( genotype quality ≥ 20 , allele counts ≥ 10 , and allele percentage ≥ 0 . 9 ) consecutive homozygous SNVs uninterrupted by heterozygous SNVs . The overlapping regions of homozygosity ( >1 Mb ) between the two patients were listed in S4 Table . Subsequently , the variants were screened by known causal mutation databases ( OMIM and HGMD ) , and polymorphism databases ( 1000Genome , ESP6500 , dbSNP138 , and an in-house exome database with 721 individuals of whom >90% are of the Middle East origin ) with matching variants not exceeding the prevalence cutoff of 0 . 5% , and the pathogenicity of variants was further evaluated ( analysis pipeline: https://sourceforge . net/projects/merap/ ) . Sanger sequencing of the ARHGEF2 ( NM_004723 . 3 ) in our cases was performed to confirm the mutation in the patients , establish the genotype in the other family members , and evaluate the presence of potential mutations in additional six consanguineous pedigrees of Sri Lankan , Italian and Turkish descent with developmental delay and a similar radiological phenotype of pontine and cerebellar hypoplasia . Some of these patients also had facial dysmorphism , ataxia , autism , epilepsy , and further MRI findings including a migration or myelination defect . The healthy status of heterozygous parents argues against a haploinsufficiency mechanism underlying this disease phenotype . Two patients with heterozygous deletions encompassing the ARHGEF2 are listed in the Decipher database: ( i ) patient 276515 ( https://decipher . sanger . ac . uk/patient/276515; 2 . 76 Mb deletion covering 59 genes; arr 1q22;q23 ( 155 , 296 , 964–158 , 056 , 876 ) ) with abnormal facial features , intellectual disability , and speech delay and ( ii ) patient 255240 ( https://decipher . sanger . ac . uk/patient/255240; 0 . 92 Mb deletion spanning 24 genes; arr 1q22 ( 155 , 192 , 986–156 , 108 , 069 ) ) with facial dysmorphism , brachydactyly , intellectual disability , severe speech delay , muscular hypotonia , but normal cMRI findings . Apart from the ARHGEF2 candidate variant , there were three further homozygous variants shared by the two patients that we ranked as unlikely disease-causing: ( i ) EXTL1 ( NM_004455 , exostosin-like glycosyltransferase 1 ) , g . 1:26356156G>A , c . 939G>A , p . W313X . However , there were 16 heterozygotes in the ESP6500 database and one healthy individual in our in-house database with a homozygous variant . ( ii ) HMCN1 ( NM_031935 , hemicentin 1 ) , g . 1:186010193G>A , c . 6229G>A , p . D2077N . However , this gene has been reported to be associated with macular degeneration , a phenotype that does not fit the clinical features of our index patients . ( iii ) IGFN1 ( NM_001164586 , immunoglobulin-like and fibronectin type III domain containing 1 ) , g . 1:201193885G>A , c . 10369G>A , p . G3457S . However , there were 7 heterozygotes in the ESP6500 database , and this variant has a prediction score from SIFT ( tolerated ) and PolyPhen2 ( possibly damaging ) that makes it unlikely as well as a GERP conservation score of merely 1 . 101 . Because both patients were male , we checked on the X chromosome for hemizygous variant , but detected no rare deleterious variant shared by the two individuals . LCLs were established and cultured according to the protocol published by Neitzel et al . 1986 . [52] RNA extraction and cDNA synthesis were performed with established methods reported previously . [53] To specifically amplify and detect ARHGEF2 and RPII ( RNA polymerase II , reference gene ) cDNA , we designed sets of primers using the Primer3 online software ( www . primer3 . ut . ee ) . qPCR experiments were run in triplicate using Maxima SYBR Green/ROX qPCR Master Mix ( Thermo Scientific , Braunschweig , Germany ) according to the manufacturer’s protocol with primers specified in S5 Table . Quantification was performed as described previously , [53] and statistical calculations were performed on GraphPad Prism 5 Software ( GraphPad Software Inc . , La Jolla , CA , USA ) . Protein extraction and Western blots ( run in triplicates ) were performed with established methods reported previously;[54] antibodies are listed in S6 Table . LCLs briefly plated on poly-L-lysine ( Sigma-Aldrich , Taufkirchen , Germany ) -coated coverslips were fixed in 4% PFA . Coverslips were further incubated in staining buffer ( 0 . 2% gelatin , 0 . 25% Triton X-100 , 10% donkey normal serum ) for 30 min for permeabilization and blocking , followed by overnight incubation with primary antibodies and an 2 h incubation with the corresponding secondary antibodies ( antibodies listed in S6 Table ) . Nuclei were labeled with 4’ , 6-Diamidino-2-phenylindole ( DAPI , 1:1000 , Sigma-Aldrich ) . Fluorescently labeled cells were analyzed and imaged using a fluorescent Olympus BX51 microscope with the software Magnafire 2 . 1B ( 2001 ) ( Olympus , Hamburg , Germany ) , and images were processed using Adobe Photoshop and ImageJ . Peripheral blood lymphocytes ( PBL ) were isolated from heparinized blood samples by Ficoll density gradient centrifugation . DNA damage was introduced by exposure of PBLs to 6 MV X-ray photons . Matched cultures were set up from untreated and irradiated cells in RPMI 1640 medium , supplemented with 15% fetal bovine serum ( FBS , PAN Biotech , Aidenbach , Germany ) . Lymphocyte growth activation was achieved by phytohemagglutinin ( PHA HA16 , Thermo Scientific , Dartford , UK ) . The cell cycle assay was performed using 5-bromo-2′-deoxyuridine ( BrdU ) –Hoechst 33258 flow cytometry . Cells with replication fork-stalling types of DNA damage become delayed in the G2 phase of the cell cycle . LCLs were washed in PBS , fixed in 2% paraformaldehyde ( PFA ) , and permeabilized with 90% methanol for 30 min on ice . Phospho-histone H3-Ser10 ( pH3 ) was detected with anti-pH3 antibody and corresponding secondary antibody ( S6 Table ) . DAPI at 2 μg/ml final concentration was used as a counterstain for DNA content and cell cycle distribution . Fluorescence was recorded using the same LSRII flow cytometer ( Becton Dickinson , Franklin Lakes , NJ , USA ) as for cell cycle studies . Data analysis was done with WinMDI 2 . 9 software ( MicroSoft , Redmond , WA , USA ) . Cortices of CD1 mouse E13 embryos ( Charles River , Sherbrooke , Quebec , Canada; E , embryonic days ) were dissected in ice-cold Hank’s balanced salt solution ( HBSS , Gibco , Carlsbad , CA , USA ) and immediately placed in cortical precursor culture media ( CPCM ) , Neurobasal ( Gibco ) supplemented by 40 ng/ml bFGF ( BD Biosciences , San Jose , CA , USA ) , 2% B27 ( Gibco ) , 1% Penicillin/Streptomycin ( Lonza , Basel , Switzerland ) , and 500 μM L-Glutamine ( Gibco ) . Following mechanical tissue dissociation , cells were plated on Poly-D-Lysine ( Sigma , St . Louis , MO , USA ) and mouse laminin ( Corning , Corning , NY , USA ) -coated coverslips ( 150 , 000 viable cells per well of a 24-well plate ) and allowed to adhere for four hours in CPCM . Subsequently , cells were transfected with the appropriate plasmids overnight using Lipofectamine LTX ( Invitrogen , Carlsbad , CA , USA ) in Opti-MEM ( Gibco ) . Because , Lfc overexpression is known to cause cell death , following transfection , 50 μM of the caspase inhibitor ZVAD-FMK ( R&D Systems , Minneapolis , MN , USA ) was added to the culture media . [24] Two days later , on DIV3 , cells were fixed with 4% PFA and permeabilized with 0 . 2% NP-40 ( USB Corporation , Cleveland , OH , USA ) in PBS . Cultures were then stained with antibodies listed in S6 Table . EGFP-positive cells ( n = 75–100 per cover slip , prepared in four independent experiments ) were evaluated for co-staining with βIII-tubulin or Ki67 using a Zeiss Axiovert A . 1 fluorescence microscope . The nuclear EGFP ( pEF-EGFP ) expression plasmid and the Arhgef2 ( Lfc ) shRNA constructs in pG-SHIN2 vector were generated as reported previously . [24] Wildtype ARHGEF2 ( NM_001162383 . 1 ) and mutant ARHGEF2 ( c . 1462delG , position according to NM_004723 . 3 ) were cloned into a pcDNA3 . 1 ( Zeo ) ( + ) plasmid ( Invitrogen ) . Mutant ARHGEF2 was generated using the QuikChange II XL Site-Directed Mutagenesis kit ( Agilent , 200521 ) with primers given in S5 Table following the manufacturer’s protocol . Sanger sequencing confirmed the correct insertion of the mutation . In utero electroporation was performed as described previously[55] using a CUY21 EDIT electroporator ( TR Tech , Japan ) to deliver five 50 ms pulses of 40–50 V with 950 ms intervals per embryo . Per embryonic brain , 4 μg total of plasmid DNA was suspended in the tracer 0 . 5% trypan blue in a ratio of 37 . 5% shArhgef2 ( Lfc ) ( or shCon ) , 37 . 5% human ARHGEF2 ( wt ) /human ARHGEF2 ( mut ) /empty vector , and 25% pEF-EGFP plasmid . Embryos were collected 3 days following electroporation . Following an over-night fixation in 4% PFA , brains were cryopreserved and mounted . Coronal cryosections of 18 μm were incubated with antibodies listed in S6 Table . Images were collected from three brain sections per condition using a Quorom spinning disk confocal microscope system or the Zeiss Axio Imager M2 system . EGFP-positive cells were counted in either the combined ventricular zone/sub-ventricular zone/intermediate zone ( VZ/SVZ/IZ ) or the cortical plate ( CP ) and reported as proportion of EGFP-positive cells in each of these regions . SatB2/EGFP-double positive cells in the CP were also counted for each condition . For mitotic spindle plane analysis , the electroporated brain sections were stained with antibodies directed against Cdk5rap2 and corresponding secondary antibodies ( S6 Table ) . Nuclei were labeled with DAPI . Images of transfected cells in 10–15 μm z-stacks at an interval of 0 . 1 μm thickness were collected using a Zeiss Spinning Disc microscopy system CXU-S1 with ZEN 2012 software . The mitotic spindle plane was evaluated through measurement of the angle between the ventricular surface and a line connecting the centrosomes , using the ImageJ software . Activated RhoA was assessed using the Rho Activation Assay Biochem kit ( Cytoskeleton , Denver , USA ) according to the manufacturer’s protocol . Experiments were run in triplicate . Arhgef2 deficient mice were genotyped as previously reported . [40] P0 and adult brains were dissected and embedded in paraffin/OCT medium as reported previously . [54] Sagittal/coronal brain sections of 10 μm were collected on histologic slides and stained with Hematoxylin and Eosin ( H&E ) or Gallyas or 3 , 3'-Diaminobenzidine ( DAB ) staining following standard protocols . Whole brain , brain stem , cerebellum , and pontine nuclei volume were evaluated by measuring respective areas in every tenth section using ImageJ and multiplying by the thickness between sections . Immunohistology was performed on brain sections by blocking in staining buffer ( 0 . 2% gelatin , 0 . 25% Triton X-100 , 3% BSA ) for 30 min for permeabilization , followed by overnight incubation with primary antibodies and an 2 h incubation with the corresponding secondary antibodies ( antibodies listed in S6 Table ) . Nuclei were labeled with 4’ , 6-Diamidino-2-phenylindole ( DAPI , 1:1000 , Sigma-Aldrich ) . Images were obtained using a Zeiss Spinning Disc microscopy system CXU-S1 with ZEN 2012 software . A 516 bp Arhgef2 PCR product was generated with primers listed in S5 Table and subsequently cloned into a p-AL2-T vector . The RNA probe was generated by in vitro transcription . Chromogenic/fluorescence in situ hybridization was performed on 16 μm thick brain sections from E11/P0 mice post-fixed in 4% PFA for 15 min and permeabilized with Proteinase K for 2 . 5 min at RT . The reaction was stopped by applying 0 . 2% glycine in PBS 1x , followed by 20 min post fixation with 20% glutaraldehyde in 4% PFA . The sections were treated with the pre-hybridization buffer ( 50% formamide , 5x SSC pH 7 . 0 , 2 . 5 M EDTA , 0 . 1% Tween-20 , 0 . 15% CHAPS , 0 . 1 mg/ml Heparin , 100 μg/ml yeast tRNA , 50 μg/ml salmon sperm DNA , 1x Denhardt’s solution ) at 65°C for 2 hours followed by overnight incubation with Arhgef2 RNA probes at 65°C . Sections were then treated with RNase A for 30 min at 37°C , subsequently washed with SSC pH 4 . 5 in 50% formamide at 65°C and then with KTBT ( 100 mM NaCl , 50 mM Tris–HCl pH 7 . 5 , 10 mM KCl , 1% Triton X-100 ) at RT . Following blocking in 20% sheep serum and incubation with sheep anti-DIG antibody ( 1:1 , 000 , Roche , Mannheim , Germany ) at 4°C overnight , the sections were washed in KTBT and NTMT ( 50 mM NaCl , 100 mMTris–HCl pH 9 . 5 , 50 mM MgCl2 , 0 . 5% Tween-20 ) for 1 hour at RT . Development was achieved through addition of the chromogenic NBT/BCIP substrate ( 1:50 , Roche ) . CD4+ naive T cells were isolated from spleen of C57BL/6 ( WT ) and Arhgef2-/- mice by using CD4+CD62L+ T Cell Isolation Kit ( Miltenyi Biotec , San Diego , CA , USA ) and subsequently stained with CellTracer Violet Cell Proliferation Kit ( Thermo Fisher Scientific ) according to the instructions provided . Cells were then incubated with 2 μg/ml of anti-CD28 antibody ( BD , New Jersey , USA ) in anti-CD3e antibody ( BD , New Jersey , USA ) pre-coated wells for 4 days at the cell concentration of 1 × 106 cell/ml . 4 days later , cells were collected and stained with APC-conjugated anti-CD3e ( Biolegend , San Diego , CA , USA ) and BV786-conjugated anti-CD4 ( BD , New Jersey , USA ) . Cells were analyzed by FACS Aria II ( BD Bioscience , New Jersey , USA ) followed by using FlowJo software ( Tree Star ) . All cells were pre-gated on singlet cells ( area & width ) and on living cells ( 7AAD− ) .
|
During brain development , localized gene expression is crucial for the formation and function of specific brain regions . Various groups of proteins are known to regulate segmentation through controlled gene expression , among them , the Rho GTPase regulator family . In this study , we identified a frameshift mutation in the Rho guanine nucleotide exchange factor 2 gene ( ARHGEF2 ) in two children presenting with intellectual disability , mild microcephaly , and a midbrain-hindbrain malformation . This phenotype is also observed in Arhgef2 mutant mice , highlighting the importance of ARHGEF2 across development of distinct mammalian species . We show that loss of Arhgef2 affects neurogenesis and also cell migration . In addition , we extended the current knowledge of ARHGEF2 expression and its role in early central nervous system development , with special reference to the formation of the precerebellar system . In addition to extensive literature on ARHGEF2 , we now provide evidence for its significant role in neuronal migration in brain development and link the gene to human neurodevelopmental disease .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"nuclear",
"staining",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"brain",
"precursor",
"cells",
"neuroscience",
"mutation",
"developmental",
"biology",
"hindbrain",
"cerebellum",
"frameshift",
"mutation",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"animal",
"cells",
"neuron",
"migration",
"dapi",
"staining",
"cellular",
"neuroscience",
"cell",
"biology",
"anatomy",
"neurons",
"genetics",
"cell",
"migration",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"cerebral",
"cortex"
] |
2017
|
Homozygous ARHGEF2 mutation causes intellectual disability and midbrain-hindbrain malformation
|
Restless legs syndrome ( RLS ) is a sensorimotor disorder with an age-dependent prevalence of up to 10% in the general population above 65 years of age . Affected individuals suffer from uncomfortable sensations and an urge to move in the lower limbs that occurs mainly in resting situations during the evening or at night . Moving the legs or walking leads to an improvement of symptoms . Concomitantly , patients report sleep disturbances with consequences such as reduced daytime functioning . We conducted a genome-wide association study ( GWA ) for RLS in 922 cases and 1 , 526 controls ( using 301 , 406 SNPs ) followed by a replication of 76 candidate SNPs in 3 , 935 cases and 5 , 754 controls , all of European ancestry . Herein , we identified six RLS susceptibility loci of genome-wide significance , two of them novel: an intergenic region on chromosome 2p14 ( rs6747972 , P = 9 . 03 × 10−11 , OR = 1 . 23 ) and a locus on 16q12 . 1 ( rs3104767 , P = 9 . 4 × 10−19 , OR = 1 . 35 ) in a linkage disequilibrium block of 140 kb containing the 5′-end of TOX3 and the adjacent non-coding RNA BC034767 .
Restless legs syndrome ( RLS ) is a common neurological disorder with a prevalence of up to 10 % , which increases with age [1] . Affected individuals suffer from an urge to move due to uncomfortable sensations in the lower limbs present in the evening or at night . The symptoms occur during rest and relaxation , with walking or moving the extremity leading to prompt relief . Consequently , initiation and maintenance of sleep become defective [1] . RLS has been associated with iron deficiency , and is pharmacologically responsive to dopaminergic substitution . Increased cardiovascular events , depression , and anxiety count among the known co-morbidities [1] . Genome-wide association studies ( GWAs ) identified genetic risk factors within MEIS1 , BTBD9 , PTPRD , and a locus encompassing MAP2K5 and SKOR1 [2]–[4] . To identify additional RLS susceptibility loci , we undertook an enlarged GWA in a German case-control population , followed by replication in independent case-control samples originating from Europe , the United States of America , and Canada . In doing so , we identified six RLS susceptibility loci with genome-wide significance in the joint analysis , two of them novel: an intergenic region on chromosome 2p14 and a locus on 16q12 . 1 in close proximity to TOX3 and the adjacent non-coding RNA BC034767 .
We enlarged our previously reported [2] , [4] GWA sample to 954 German RLS cases and 1 , 814 German population-based controls from the KORA-S3/F3 survey and genotyped them on Affymetrix 5 . 0 ( cases ) and 6 . 0 ( controls ) arrays . To correct for population stratification , as a first step , we performed a multidimensional scaling ( MDS ) analysis , leading to the exclusion of 18 controls as outliers . In a second step , we conducted a variance components analysis to identify any residual substructure in the remaining samples , resulting in an inflation factor λ of 1 . 025 ( Figures S1 and S2 ) . The first four axes of variation from the MDS analysis were included as covariates in the association analysis of the genome-wide stage and all P-values were corrected for the observed λ . Prior to statistical analysis , genotyping data was subjected to extensive quality control . We excluded a total of 302 DNA samples due to a genotyping call rate <98 % . For individual SNP quality control , we adopted a stringent protocol in order to account for the complexity of an analysis combining 5 . 0 and 6 . 0 arrays . We excluded SNPs with a minor allele frequency ( MAF ) <5% , a callrate <98% , or a significant deviation from Hardy-Weinberg Equilibrium ( HWE ) in controls ( P<0 . 00001 ) . In addition , we dropped SNPs likely to be false-positive associations due to differential clustering between 5 . 0 and 6 . 0 arrays by adding a second set of cases of an unrelated phenotype and discarding SNPs showing association in this setup ( see Materials and Methods ) . Finally , we tested 301 , 406 SNPs for association in 922 cases and 1 , 526 controls . Based on a threshold level of a nominal λ-corrected PGWA<10-4 , a total of 47 SNPs distributed over 26 loci were selected for follow-up in the replication study ( Figure 1 , Table S1 ) . We genotyped these 47 SNPs together with 29 adjacent SNPs in strong linkage disequilibrium ( LD , r2 = 0 . 5–0 . 9 ) using the Sequenom iPLEX platform in seven case-control populations of European descent , comprising a total of 3 , 935 cases and 5 , 754 controls . Eleven SNPs with a call rate <95% , MAF<5% , and P<0 . 00001 for deviation from HWE in controls as well as 432 samples with a genotyping call rate <90% were excluded . A set of 47 SNPs , genotyped in 186 samples on both platforms ( Affymetrix and Sequenom ) , was used to calculate an average concordance rate of 99 . 24 % . The combined analysis of all replication samples confirmed the known four susceptibility loci and , in addition , identified two novel association signals on chromosomes 2p14 and 16q12 . 1 ( Table 1 ) . To address possible population stratification within the combined replication sample , we performed a fixed-effects meta-analysis . For four of the replication case-control populations , we included λ inflation factors which were available from a genomic controls experiment in a previous study in these populations [4] . These were used to correct the estimates for the standard error . Joint analysis of GWA and all replication samples showed genome-wide significance for these two novel loci as well as for the known RLS loci in MEIS1 , BTBD9 , PTPRD , and MAP2K5/SKOR1 with a nominal λ -corrected PJOINT <5×10−8 ( Table 1 ) . Depending on the variable power to detect the effects , the separate analyses of individual subsamples in the replication either confirmed the association after correction for multiple testing or yielded nominally significant results ( Tables S2 and S3 ) . The differing relevance of the risk loci in the individual samples is illustrated in forest plots ( Figure 2 ) . There was no evidence of epistasis between any of the six risk loci ( PBonferroni >0 . 45 ) . The association signal on 2p14 ( rs6747972: nominal λ-corrected PJOINT = 9 . 03×10−11 , odds ratio ( OR ) = 1 . 23 ) is located in an LD block of 120 kb within an intergenic region 1 . 3 Mb downstream of MEIS1 ( Figure 3 ) . Assuming a long-range regulatory function of the SNP-containing region , in silico analysis for clusters of highly conserved non-coding elements using the ANCORA browser ( http://ancora . genereg . net ) identified MEIS1 as well as ETAA1 as potential target genes [5] , [6] . The second locus on chromosome 16q12 . 1 ( rs3104767: nominal λ-corrected PJOINT = 9 . 4×10−19 , OR = 1 . 35 ) is located within an LD block of 140 kb ( Figure 3 ) , which contains the 5′UTR of TOX3 ( synonyms TNRC9 and CAGF9 ) and the non-coding RNA BC034767 ( synonym LOC643714 ) . TOX3 is a member of the high mobility box group family of non-histone chromatin proteins which interacts with CREB and CBP and plays a critical role in mediating calcium-dependent transcription in neurons [7] . GWAs have identified susceptibility variants for breast cancer in the identical region [8] . The best-associated breast cancer SNP , rs3803662 , is in low LD ( r2∼0 . 1 , HapMap CEU data ) with rs3104767 , but showed association to RLS ( λ-corrected nominal PGWA = 7 . 29×10−7 ) . However , logistic regression analysis conditioned on rs3104767 demonstrated that this association is dependent on rs3104767 ( rs3803662: PGWA/conditioned = 0 . 2883 ) . BC034767 is represented in GenBank by two identical mRNA transcripts , BC034767 and BC029912 . According to the gene model information of the UCSC and Ensembl genome browsers ( http://genome . ucsc . edu and http://www . ensembl . org/index . html ) , these mRNAs are predicted to be non-coding . Additional in silico analysis using the Coding Potential Calculator ( http://cpc . cbi . pku . edu . cn ) supported this by attributing only a weak coding potential to this RNA , suggesting a regulatory function instead [9] . We also searched for rare alleles with strong effects and performed a mutation screening by sequencing all coding and non-coding exons of TOX3 and BC034767 in 188 German RLS cases ( Table S4 ) . In TOX3 , a total of nine variants not listed in dbSNP ( Build 130 ) were found , three of which are non-synonymous . Only one of these is also annotated in the 1000 Genomes project ( November 2010 data release ) . Three additional new variants were located in putative exons 1 and 2 of BC034767 . Analysis of the frequency of these variants as well as all known non-synonymous , frameshift , and splice-site coding SNPs in TOX3 in a subset of one of the replication samples ( 726 cases and 735 controls from the GER1 sample ) did not reveal any association to RLS . For a power of >80% , however , variants with an OR above 4 . 5 and a MAF ≥0 . 01 would be required . For even lower MAFs , ORs ≥10 would be necessary for sufficient power . Furthermore , the described CAG repeat within exon 7 of TOX3 was not polymorphic as shown by fragment analysis in 100 population-based controls . According to publicly available expression data ( http://genome . ucsc . edu ) , in humans , BC034767 is expressed in the testes only , while TOX3 expression has been shown in the salivary glands , the trachea , and in the CNS . Detailed in-depth real time PCR profiling of TOX3 showed high expression levels in the frontal and occipital cortex , the cerebellum , and the retina [10] . To assess a putative eQTL function of rs6747972 or rs3104767 , we studied the SNP-genotype-dependent expression of TOX3 and BC034767 as well as of genes known to directly interact with TOX3 ( CREB-1/CREBBP/CITED1 ) and potential target genes of long-range regulatory elements at the locus on chromosome 2 ( MEIS1/ETAA1 ) in RNA expression microarray data from peripheral blood in 323 general population controls [11] . No differential genotype-dependent expression variation was found . To assess the potential for genetic risk prediction , we split our GWA sample in a training and a test set and determined classifiers for case-control status in the training set to predict case-control status in the test set . Training and test set were independent of each other – not only with respect to included individuals but also with respect to the genotyping procedure as we used genotypes generated on different genotyping platforms . As training set , we used those cases of the current GWA which had been genotyped on 500K arrays in a previous GWA and the corresponding control set [2] , in total , 326 cases and 1 , 498 controls . The test set comprised 583 cases and 1 , 526 controls , genotyped on 5 . 0/6 . 0 arrays as part of the current study . Prior to the analysis , we removed the six known risk loci and performed LD-pruning to limit the analysis to SNPs not in LD with each other . In the end , a total of 76 , 532 SNPs were included in the pruned dataset . We conducted logistic regression with age and sex as covariates . Based on these association results , the sum score of SNPs showing the most significant effects ( i . e . the number of risk alleles over all SNPs ) weighted by the ln ( OR ) of these effects was chosen as predictor variable in the test set . We then varied the P-value threshold for SNPs included in the sum score . For a P-value <0 . 6 , we observed a maximum area under the curve ( AUC ) of 63 . 9% and an explained genetic variance of 6 . 6% ( Nagelkerke's R ) , values comparable to estimates obtained for other complex diseases such as breast cancer or diabetes ( Table S5 ) [12]–[14] . Inclusion of the six known risk loci in this analysis resulted in a maximum AUC of 64 . 2% and an explained genetic variance of 6 . 8% . Additionally , we performed risk prediction in the combined GWA and replication sample including only the six established RLS risk loci . For this purpose , we used the weighted risk allele score resulting in ORs of up to 8 . 6 ( 95% CI: 2 . 46–46 . 25 ) and an AUC of 65 . 1% ( Figures S3 and S4 ) . By increasing the size of our discovery sample , we have identified two new RLS susceptibility loci . The top six loci show effect sizes between 1 . 22 and 1 . 77 and risk allele frequencies between 19 and 82% , and reveal genes in neuronal transcription pathways not previously suspected to be involved in the disorder .
Statistical analysis was performed using PLINK 1 . 07 ( http://pngu . mgh . harvard . edu/~purcell/plink , [21] ) . In the GWA sample , we applied logistic regression with age , sex , and the first four axes of variation resulting from an MDS analysis as covariates . P-values were λ-corrected with the λ of 1 . 025 from the EMMAX analysis . In the individual analysis of the single replication samples , we tested for association using logistic regression and correcting for gender and age as well as for population stratification where possible ( see Population Stratification ) . Each replication sample was Bonferroni-corrected using the number of SNPs which passed quality control for the respective sample . For the combined analysis of all replication samples , we performed a fixed-effects inverse-variance meta-analysis . Where available , we used λ-corrected standard errors in this analysis . Bonferroni-correction was performed for 74 SNPs , i . e . the number of SNPs which passed quality control in at least one replication sample . For the joint analysis of the GWA and the replication samples , we also used a fixed-effects inverse-variance meta-analysis and again included λ-corrected values as far as possible . For the conditioned analysis , the SNP to be conditioned on was included as an additional covariate in the logistic regression analysis as implemented in PLINK . Interaction analysis was performed using the –epistasis option in PLINK . Significance was determined via Bonferroni-correction ( i . e . 0 . 05/28 , as 28 SNP combinations were tested for interaction ) . Power calculation was performed using the CaTS power calculator [25] using a prevalence set of 0 . 08 and an additive genetic model ( Table S3 ) . The significance level was set at 0 . 05/74 for replication stage analysis and at 0 . 05/301 , 406 for genome-wide significance in the joint analysis of GWA and replication . For the rare variants association study , the significance level was set at 0 . 05/12 . All coding and non-coding exons including adjacent splice sites of TOX3 ( reference sequence NM_001146188 ) and BC034767 ( reference sequence IMAGE 5172237 ) were screened for mutations in 188 German RLS cases . Mutation screening was performed with high resolution melting curve analysis using the LightScanner technology and standard protocols ( IDAHO Technology Inc . ) . DNAs were analyzed in doublets . Samples with aberrant melting pattern were sequenced using BigDyeTerminator chemistry 3 . 1 ( ABI ) on an ABI 3730 sequencer . Sequence analysis was performed with the Staden package [26] . Primers were designed using ExonPrimer ( http://ihg . gsf . de ) or Primer3plus ( www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi ) . All identified variants were then genotyped in 735 RLS cases and 735 controls of the general population ( KORA cohort ) on the MassARRAY system , as described above . In addition , fragment analysis of exon 7 of TOX3 was performed to screen for polymorphic CAG trinucleotide repeats . DNA of 100 controls ( 50 females , 50 males ) was pooled and analyzed on an ABI 3730 sequencer with LIZ-500 ( ABI ) as a standard . Primers were designed using Primer3plus , the forward Primer contains FAM for detection . Analysis was performed using GeneMapper v3 . 5 . Associations between MEIS1/ETAA1 RNA expression and rs6747972 and between TOX3/BC034767/CREB-1/CREBBP/CITED1 expression and rs3104767 were assessed using genome-wide SNP data ( Affymetrix 6 . 0 chip ) in conjunction with microarray data for human blood samples ( n = 323 general population controls from the KORA cohort , Illumina Human WG6 v2 Expression BeadChip ) [11] . A linear regression model conditioned on expression and controlling for age and sex was used to test for association .
|
Restless legs syndrome ( RLS ) is one of the most common neurological disorders . Patients with RLS suffer from an urge to move the legs and unpleasant sensations located mostly deep in the calf . Symptoms mainly occur in resting situations in the evening or at night . As a consequence , initiation and maintenance of sleep become defective . Here , we performed a genome-wide association study to identify common genetic variants increasing the risk for disease . The genome-wide phase included 922 cases and 1 , 526 controls , and candidate SNPs were replicated in 3 , 935 cases and 5 , 754 controls , all of European ancestry . We identified two new RLS–associated loci: an intergenic region on chromosome 2p14 and a locus on 16q12 . 1 in a linkage disequilibrium block containing the 5′-end of TOX3 and the adjacent non-coding RNA BC034767 . TOX3 has been implicated in the development of breast cancer . The physiologic role of TOX3 and BC034767 in the central nervous system and a possible involvement of these two genes in RLS pathogenesis remain to be established .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"movement",
"disorders",
"neurological",
"disorders",
"neurology",
"sleep",
"disorders"
] |
2011
|
Genome-Wide Association Study Identifies Novel Restless Legs Syndrome Susceptibility Loci on 2p14 and 16q12.1
|
Marine coccolithophorid phytoplankton are major producers of biogenic calcite , playing a significant role in the global carbon cycle . Predicting the impacts of ocean acidification on coccolithophore calcification has received much recent attention and requires improved knowledge of cellular calcification mechanisms . Uniquely amongst calcifying organisms , coccolithophores produce calcified scales ( coccoliths ) in an intracellular compartment and secrete them to the cell surface , requiring large transcellular ionic fluxes to support calcification . In particular , intracellular calcite precipitation using HCO3− as the substrate generates equimolar quantities of H+ that must be rapidly removed to prevent cytoplasmic acidification . We have used electrophysiological approaches to identify a plasma membrane voltage-gated H+ conductance in Coccolithus pelagicus ssp braarudii with remarkably similar biophysical and functional properties to those found in metazoans . We show that both C . pelagicus and Emiliania huxleyi possess homologues of metazoan Hv1 H+ channels , which function as voltage-gated H+ channels when expressed in heterologous systems . Homologues of the coccolithophore H+ channels were also identified in a diversity of eukaryotes , suggesting a wide range of cellular roles for the Hv1 class of proteins . Using single cell imaging , we demonstrate that the coccolithophore H+ conductance mediates rapid H+ efflux and plays an important role in pH homeostasis in calcifying cells . The results demonstrate a novel cellular role for voltage gated H+ channels and provide mechanistic insight into biomineralisation by establishing a direct link between pH homeostasis and calcification . As the coccolithophore H+ conductance is dependent on the trans-membrane H+ electrochemical gradient , this mechanism will be directly impacted by , and may underlie adaptation to , ocean acidification . The presence of this H+ efflux pathway suggests that there is no obligate use of H+ derived from calcification for intracellular CO2 generation . Furthermore , the presence of Hv1 class ion channels in a wide range of extant eukaryote groups indicates they evolved in an early common ancestor .
Coccolithophores represent a pan-global group of oceanic phytoplankton , often forming massive monospecific blooms in oceanic waters . These unicellular eukaryote algae produce highly intricate calcium carbonate scales , known as coccoliths , and are the most numerous calcifying organisms in our oceans . Globally abundant species such as Emiliania huxleyi and Coccolithus pelagicus spp braarudii [1] play fundamental roles in long-term carbon deposition , marine biogeochemical cycling , and atmospheric chemistry through their direct effects on surface ocean alkalinity and through ballasting of organic carbon fluxes to deeper waters [2] . Anthropogenic CO2 emissions are predicted to have a significant impact on calcifying organisms , due to a decrease in both ocean surface water pH and the saturation state of calcium carbonate . However , there is currently significant debate regarding the effects of elevated CO2 and decreased ocean pH on coccolithophores [3]–[7] , due in part to a lack of understanding of the cellular mechanisms underlying calcification . Improved knowledge of coccolithophore cell biology is therefore necessary for both predicting the physiological consequences of ocean acidification and identifying experimental versus physiological sources of variability observed in experimental manipulations on this ubiquitous group of phytoplankton . In contrast to other major marine calcifyers such as corals [8] and foraminifera [9] , coccolithophores produce calcite scales ( heterococcoliths ) entirely within intracellular compartments ( the coccolith vacuole , CV ) which are secreted to the cell surface forming the external coccosphere [10] . Intracellular calcification requires large sustained fluxes of Ca2+ and inorganic carbon ( Ci ) to the coccolith vacuole . The bulk of experimental work supports HCO3− as the primary Ci species transported into the cell to sustain calcification , resulting in 1 mole of H+ generated for every 1 mole of calcite precipitated [11] , [12] . Our calculations , based on published calcification rates , indicate that H+ production during calcification without H+ removal or consumption will cause rapid cytoplasmic acidification of ∼0 . 3 pH min−1 ( Table S1 ) . Metabolic pH balance may arise through photosynthesis [13] , though the degree to which this occurs has been questioned by studies that indicate no mechanistic dependence of photosynthesis on calcification [14] , [15] . It follows that in the absence of rapid metabolic H+ consumption , sequestration , or removal , the cytosol of calcifying coccolithophore cells would be subject to significant acidosis . Our data show that a plasma membrane voltage-activated H+ channel , novel for photosynthetic organisms , plays a crucial role in short-term cellular pH homeostasis which is in turn required for maintenance of calcification .
In order to understand the membrane transport processes which underlie the extraordinary process of intracellular calcification in coccolithophores , we applied the patch clamp technique following removal of the external calcite coccolith scales by brief treatment with the Ca2+ chelator ethyleneglycol-O , O'-bis ( 2-aminoethyl ) -N , N , N' , N'-tetraacetic acid ( EGTA ) ( see Materials and Methods; Figure S1 ) . Patch clamp recordings revealed a slowly activating plasma membrane ion current in C . pelagicus in response to depolarisations more positive than the equilibrium potential for H+ ( EH+ ) ( Figures 1A , B ) . Tail current analysis demonstrated a reversal potential ( Erev ) very positive of EK+ and ECl− , and closest to EH+ ( Figures 1C , D ) . A strong Nernstian relationship between Erev and transmembrane pH gradient ( ΔpH ) in the presence of various bath ( pHο 6 . 5–8 . 0 ) solutions showed that the current is selective for H+ ( Figure 1D ) . The outward current was depressed and voltage activation shifted more positive in response to decreased pHο ( Figure 1E , F ) , a characteristic of animal H+ currents [16] , [17] . Reducing pHi from 7 . 5 to 6 . 5 ( with pHo held at 8 . 0 ) resulted in a greater outward current amplitude at all membrane potentials ( Figure S2 ) , although it was not possible to determine accurately the activation potential of the H+ current at pHi 6 . 5 as EH+ was too close to the activation potential for the inward Cl− current [18] . C . pelagicus H+ currents were inhibited by Zn2+ ( Figure 1G , H ) , which is characteristic of animal H+ currents [19] , [20] and were sensitive to the trivalent cation Gd3+ ( Figure S3 ) . To confirm that K+ or Cl− was not contributing significantly to the outward current , tail current analysis was performed using a range of pipette solutions . In all cases the reversal potentials of C . pelagicus outward currents were consistently close to EH+ regardless of large changes in EK+ or ECl− ( Figure 2 ) . The small deviation observed from EH+ may be due to incomplete pH buffering by the pipette solution [21] , [22] or a small contribution from an additional unidentified conductance . Application of the Goldman , Hodgkin , Katz equation for relative permeabilities of the ions in the pipette solutions ( H+ , K+ , and Cl− ) from the data in Figures 1 and 2 gives permeability ratios in excess of 106 for gH+ relative to K+ and Cl− . The extremely high selectivity for H+ is consistent with other reported values for H+ channels ( e . g . , [16] ) . The observed H+ current magnitudes in C . pelagicus are comparable to the large currents found in activated granulocytes [21] and more than adequate to dissipate calcification associated H+ production in the absence of metabolic consumption or sequestration ( Table S1 ) . While a range of H+ transport and homeostatic mechanisms are likely to contribute to pHi regulation , the H+ efflux channel identified here possesses the transport capacity and kinetics that would enable rapid short-term regulation of potentially large pHi fluctuations . The outward H+ conductance in C . pelagicus shares many characteristics with those produced by the Hv1 class of voltage-gated H+ channels identified in animals [17] , [19] , [20] . Similarity searches using animal Hv1 sequences identified a single putative open reading frame coding for 339 amino acids ( EhHVCN1 ) within the E . huxleyi genome ( Joint Genome Institute; Read et al . , unpublished ) . We subsequently identified a further putative homologue in a collection of C . pelagicus ESTs ( CpHVCN1 , von Dassow et al . , unpublished ) . The coccolithophore sequences exhibit a weak overall similarity to mammalian Hv1 channels at the amino acid level ( EhHv1 has 19% identity , 33% similarity to human Hv1 ) , but have similar organisation including four predicted transmembrane domains and conserved features including the critical voltage sensing arginine residues in transmembrane domain S4 ( Figure 3A–C ) . Notable differences are the non-conservation of histidine residues required for Zn2+ inhibition in mammalian Hv1 channels [19] and extension of the putative extracellular loop between the S1 and S2 domains . RT-PCR confirmed that EhHVCN1 and CpHVCN1 were expressed in calcifying strains of E . huxleyi and C . pelagicus , respectively . Sequence similarity searches of currently available genomic datasets using the coccolithophore proteins identified further putative Hv1 homologues in several evolutionarily distant eukaryotes including the diatoms , Phaeodactylum tricornutum and Thalassiosira pseudonana , and the social amoebozoan , Polysphondylium pallidum , indicating that the Hv1 class of proteins may have a broad taxonomic distribution and an ancient evolutionary origin ( Figure 3A , C ) . Human HEK293 cells transfected with either EhHVCN1 or CpHVCN1 exhibited robust voltage-dependent outward currents significantly greater than endogenous outward currents known to occur in this cell type ( Figure 4A , B ) [17] . Further characterisation of EhHv1 expressed in HEK293 cells indicated that the magnitude and Erev of the current was pH dependent ( Figure 4C–F ) and sensitive to Zn2+ ( Figure 4G , H ) . Analysis of H+ current activation kinetics in response to +50 mV depolarisation shows that the currents generated by heterologously expressed EhHv1 and CpHv1 had faster activation kinetics than the C . pelagicus native currents ( τ = 107 ms ±30 . 4 SE , 22 . 9 ms ±5 . 4 SE , and 220 ms ±39 . 3 SE for EhHv1 , CpHv1 in HEK cells , and C . pelagicus native conductances , respectively ) . This probably reflects the different cellular context of the native and HEK cell expression since it is well documented that activation time constant can be strongly influenced by physiological conditions [23] . As the external histidine residues associated with Zn2+ binding in human Hv1 ( H140 , H193 ) are not conserved in coccolithophore Hv1 proteins , we hypothesised that Zn2+ must bind alternative residues in order to inhibit the H+ conductances in these channels . We identified two histidine residues in the predicted S1–S2 extracellular loop region which are completely conserved across the available coccolithophore and diatom Hv1 sequences and used site-directed mutagenesis of EhHv1 to examine whether they contributed to inhibition by Zn2+ . Replacement of either histidine residue ( H197 or H203 ) with an alanine residue resulted in a very substantial reduction in the extent of inhibition induced by 500 µM Zn2+ ( from 84% to 27% or 28% , respectively , Figure S4 ) , suggesting that whilst the individual histidine residues concerned are not conserved between animal and algal Hv1 proteins , the mechanism of inhibition by Zn2+ may be similar . We conclude that coccolithophores express functional homologues of mammalian voltage gated H+ channels which exhibit highly similar biophysical properties to the outward H+ conductance observed in C . pelagicus , strongly suggesting that this conductance is generated by CpHv1 . As the biophysical characteristics of the outward conductance in C . pelagicus and Hv1 channel homologues are consistent with those described for animal Hv1 channels [17] , [19]–[21] , we hypothesized a role in rapid H+ efflux during pH homeostasis . Using simultaneous patch clamp and pH imaging , we verified that direct activation of the C . pelagicus H+ current induced significant changes in cytoplasmic pH ( Figure 5A ) . The application of a 10 s depolarization more positive of EH+ activated the H+ current and induced significant reversible cytoplasmic alkalinisation ( Figure 5A ) . The mean increases in pHi following voltage steps from −50 mV to +20 or +70 mV were 0 . 22 ( ±0 . 04 SE ) and 0 . 36 ( ±0 . 04 SE , n = 7 ) pH units , respectively . These results are consistent with H+ efflux through a voltage-sensitive conductance as described in animal systems ( e . g . , [21] ) . Following the depolarising pulses , pHi recovered approximately exponentially over 30–60 s , which is faster but not substantially different from recovery times reported in animal cells [24] . The faster pH recovery reported here may reflect a number of factors , including the rate of diffusion of buffer from the patch electrode and a significantly lower relative cytoplasmic volume of coccolithophore cells which contain large chloroplasts , vacuoles , and coccolith vacuole . Hyperpolarizing pulses that activate a large inward Cl- current [18] caused no significant pHi change ( Figure S5 ) . These observations bear key similarities to results from a number of animal cell types in which H+ currents are known to mediate pH homeostasis and charge balance [17] , [25] , [26] . They provide strong evidence that the C . pelagicus outward H+ current can specifically and significantly contribute to regulation of cytoplasmic pH . In order to understand how the outward H+ current operates to regulate cytoplasmic pH , we determined the resting membrane potential ( Vm ) in C . pelagicus . Using sharp microelectrodes , Vm of intact C . pelagicus cells was measured at −45 . 7 mV ( ±4 . 8 SE , n = 10 ) in ASW at pHo 8 . 0 . At pHo 6 . 5 Vm depolarised to −29 . 0 mV ( ±3 . 1 SE , n = 10 ) . The measured Vm at pHo 8 . 0 is very close to EH+ ( −48 mV , assuming a resting pHi of 7 . 2 typical of eukaryote cells ) , suggesting that the H+ conductance will be close to its activation potential under normal conditions . In a model whereby channel-mediated H+ efflux regulates pHi , treatments that inhibit the H+ current may cause cytosolic acidification in calcifying cells . Accordingly , the vast majority of C . pelagicus cells ( 85% , n = 44 , 7 independent experiments ) showed a strong dependence of pHi upon pHo , exhibiting reversible acidification when external pH was rapidly shifted from pH 8 . 0 to pH 6 . 5 ( Figure 5B ) . This is consistent with the sensitivity of the H+ conductance to the H+ electrochemical gradient across the plasma membrane . The direct effect of pHo on pHi also implies that C . pelagicus is not able to maintain intracellular pH in the face of transient shifts in pHo , which is consistent with evolution in the open ocean where pH is relatively stable . To address the requirement for the H+ conductance during calcification , we examined the effect of Zn2+ on pHi in actively calcifying cells . Treatment with 30 µM free Zn2+ for 2 . 5 min caused an immediate decrease in cytosolic pH in calcifying cells ( mean ΔpH = −0 . 13±0 . 02 SE , n = 62 , Figure 5C ) . Conversely , in cells where calcification was inhibited by incubation in Ca2+-free artificial seawater [15] Zn2+ did not induce a large decrease in pHi ( mean ΔpH = −0 . 02±0 . 02 SE , n = 29 , Figure 5C ) . The inward Cl− rectifier is also sensitive to Zn2+ [18] but is unlikely to influence the observed acidification as inhibition of this conductance would act to depolarise Vm resulting in cytoplasmic alkalinisation . We conclude that the plasma membrane H+ conductance plays an important role in release of H+ from the cytosol in actively calcifying cells . To address further the role of pHi homeostasis during calcification , we manipulated pHi whilst measuring calcification rates with a non-invasive in vivo method ( Figure 6A , Figure S6 ) . A reversible decrease in pHi induced by a brief reduction in extracellular pHo ( 10 min at pHo 6 . 5 , Figure S7 ) caused a 69 . 0% ± 11 . 4% inhibition of calcification rate ( Figure 6A , B ) . As changes in pHo also affect extracellular Ci speciation , we used a pulse of NH4Cl ( 10 mM , 10 min , Figure S7 ) to induce intracellular acidification while maintaining constant pHo [27] . This resulted in a 67 . 0% ± 8 . 7% inhibition of calcification ( Figure 6B ) . Inhibition of calcification continued for up to 2 h post-treatment , suggesting that down-regulation of the calcification machinery operates in response to disruption of pHi ( Figure 6B ) . Whilst indirect effects of pHo and NH4Cl treatments may contribute to the inhibition of calcification , the similar effects of different treatments imply a direct relationship between pHi homeostasis and calcification . The sensitivity of calcification to fluctuations in pHi highlights the requirement for efficient regulation of pHi , which is in turn dependent on the voltage-gated H+ conductance , and strongly supports a role for the coccolithophore Hv channels as a key component in the calcification process ( Figure 7 ) .
In combination with the emerging genomic information , our data provide clear evidence for physiological features that are novel for photosynthetic eukaryotes . C . pelagicus expresses a homologue of animal voltage-gated H+ channels and exhibits an H+-selective conductance that is activated by depolarization and dependent upon the H+ electrochemical gradient . As with metazoan H+ channels [17] , [28] , the properties of the C . pelagicus H+ conductance appear ideally suited to mediating rapid H+ efflux during metabolic acidosis [28] . Our data also support a functional link between the coccolithophore H+ conductance and sustained intracellular calcification via regulation of pHi . Our electrophysiological and molecular analyses lead us to propose the following model ( Figure 7 ) . Calcification in coccolithophores is likely to generate H+ at a relatively constant rate , whereas the rates of H+ consumption by metabolic processes ( e . g . , photosynthesis ) are likely to fluctuate rapidly ( e . g . , with changes in light intensity ) . The net H+ load resulting from calcification will therefore vary constantly . In calcifying coccolithophores , two dominant conductances are expressed in the plasma membrane: an inward Cl− rectifier activated by hyperpolarisation and dependent upon ECl [18] and an outward H+ conductance activated by depolarisation and dependent upon EH+ ( this study ) . Excursions of Vm more negative of resting Vm will have no effect on pHi because only the inward Cl− current is activated ( as evidenced in Figure S5 ) and those positive of EH+ will induce outward H+ flow and changes in pHi ( see Figure 5 ) . Our calculations suggest that the H+ conductance will be very close to its activation potential under normal conditions ( i . e . , at pHo 8 . 0 and pHi 7 . 2 , Vm is −46 mV and EH+ is −48 mV ) . Cytoplasmic acidification would also activate H+ efflux by shifting EH+ more negative ( i . e . , independent of any change in Vm ) . Therefore a calcifying cell in which H+ production is not balanced by metabolic H+ consumption will generate an acid load on the cytosol which will trigger activation of the H+ outward current . Combined depolarisation of Vm and acidification would act synergistically to induce potentially rapid H+ efflux and subsequent membrane hyperpolarization . This efflux may be sustained by activation of the sensitive inward Cl− current ( Cl− efflux ) at more negative Vm [18] , which will balance the charge and facilitate maintenance of H+ removal during calcification . Thus , we propose that outward rectifying Hv1 channels and inward rectifying Cl− channels work together to sustain H+ efflux . Hv1 homologues are not universally present in marine algae , being absent from the genomes of prasinophytes ( both Ostreococcus and Micromonas ) and the brown macroalga , Ectocarpus siliculosus , suggesting that these channels play specialised cellular roles in coccolithophores and diatoms . Interestingly , a protein exhibiting weak similarity to EhHv1 is also present in the genomes of the moss Physcomitrella patens and other land plants , although in these predicted proteins a conserved arginine in S4 ( corresponding to human R205 ) is replaced by a threonine residue . Moreover , a plasma membrane voltage-gated H+ conductance has not been described in land plants , indicating that such homologues play distinctly different functional roles in these organisms . Comparative studies of Hv1 proteins from such divergent eukaryote taxa have the potential to provide critical insight into the conserved features required for the novel mechanisms of H+ conductance within this group of ion channels ( see Figure 3A–C ) [17] . As Hv1 proteins lack a classic pore , the mechanism of H+ permeation through Hv1 proteins is not fully understood . Ionisable residues in the transmembrane domains may contribute to H+ conduction via a hydrogen bonded chain mechanism [17] , although recent evidence from a combined mutagenesis and structural modelling approach suggests that H+ may be conducted via an internal water wire , rather than the ionisable side chains of transmembrane residues [29] . Analysis of the transmembrane domains of coccolithophore Hv1 proteins indicates that the acidic residues are broadly conserved , along with the arginine residues associated with voltage gating in S4 . Physiological studies so far have not allowed a clear distinction between HCO3− and direct CO32− uptake for calcification . However , this knowledge is critical for understanding the impacts of reduced ocean surface pH on calcification , because predicted changes in ocean chemistry will bring about significant reductions in ocean surface [CO32−] [30] . The presence of an effective mechanism to dissipate excess H+ is consistent with HCO3− as the primary seawater Ci substrate for calcification and suggests there is no obligate requirement for H+ derived from calcification to be utilized for intracellular CO2 generation . The effects of elevated CO2/decreased seawater pH on coccolithophores vary according to species , strain , and experimental conditions [3]–[5] , [31] . Understandably , much attention has been paid to the effects of a decrease in calcite saturation on the dissolution of extracellular coccoliths . However , our studies identify a mechanism through which predicted ocean pH scenarios [32] , [33] may also have a direct impact on the intracellular production of coccoliths [4] , [5] , [34] via the pH-dependent properties of the plasma membrane H+ channel . An understanding of the combined effects of ocean acidification on both calcite saturation and intracellular pH homeostasis is likely to be critical for unravelling the factors underlying the variation seen in laboratory studies of coccolithophore responses to ocean acidification . Based upon molecular phylogenetic evidence , coccolithophores evolved ∼250 MYA , with the earliest heterococcolith fossils dating between 204 and 217 MYA [35] , [36] . Over this time period the oceans have remained supersaturated with regard to calcite , although surface ocean pH likely varied within the range of pH 7 . 6–8 . 2 [37] , suggesting ancestral coccolithophores have previously experienced and survived in significantly lower ocean pH . The gating dependence of the H+ current on membrane ΔpH implies that its activity may be negatively impacted by reduced seawater pH . The expression and biophysical properties of the coccolithophore H+ conductance are likely to be important factors in determining how coccolithophores may themselves respond to predicted future changes in ocean pH . The sensitivity of calcification to transient changes in cytoplasmic pH is also evident from our results . This suggests an additional level of control of calcification whereby the calcification machinery may shut down under conditions where rapid control of cytoplasmic pH is compromised , relieving the cell of additional H+ load . Whilst the H+ conductance is strictly dependent on the transmembrane H+ gradient , the gating properties of the H+ conductance and the magnitude of H+ flux described here are also under tight control of membrane potential . This may impart a degree of tolerance and physiological plasticity to the calcification process since effects of decreased pHo on H+ efflux may be countered by slight adjustments of membrane voltage to maintain H+ efflux .
Batch cultures of the unicellular haptophyte alga Coccolithus pelagicus ssp braarudii ( PLY 182G , from the Plymouth Culture Collection ) were grown in either filtered seawater ( FSW ) or artificial seawater ( ASW: 450 mM NaCl , 30 mM MgCl2 , 16 mM MgSO4 , 8 mM KCl , 10 mM CaCl2 , and 2 mM NaHCO3 ) , supplemented with nutrients , trace metals , and vitamins [18] . Cultures were maintained at 15°C under 100 µmol m2 s−1 light on a 16∶8 h light:dark cycle . Under these growth conditions the cell diameter was between 10 and 15 µm . Before electrophysiological and optical recordings , cells were decalcified with brief EGTA treatment followed by trituration as previously described [18] , which removed the external calcite coccosphere and body scales . The resultant protoplasts ( 7–12 µm in diameter , Figure S1 ) allowed patch clamp recording and improved the optical properties of the cells for imaging . This brief treatment did not affect subsequent cellular calcification and growth rates [10] , [18] . Whole cell patch clamp recordings were conducted at 20°C as previously described [18] . The recording chamber volume was 1 . 5 cm3 , and solutions exchanged using gravity-fed input and suction output at a rate of 5 cm3 min−1 . All pipette solutions contained EGTA , HEPES , or PIPES buffer and sorbitol to balance the osmolarity to between 1 , 000 and 1 , 200 mOsmol kg−1 . Specific ionic compositions of bath and pipette solutions were chosen to give optimal buffering or pH responses and are given in Table S2 and in the figure legends . Liquid junction potentials were calculated using the junction potential tool in Clampex ( Molecular Devices , Sunnyvale , CA ) and corrected off-line . Whole cell capacitance and seal resistance ( leak ) were periodically monitored during experiments by applying a <5 mV test pulse . Currents were linear leak subtracted in Clampfit ( Molecular Devices , Sunnyvale , CA ) using the pre-test seal resistance . Current voltage relations were determined on leak subtracted families by measuring the maximum steady state amplitude ( averaging between 10 and 50 ms of the current trace ) . Reversal potentials were determined by manually measuring the peak tail currents of leak subtracted families of traces and calculating a linear regression versus test voltage . Series resistance was monitored throughout the experiments and whole cell currents were analysed only from recordings in which series resistance varied by less than 15% . Decalcified C . pelagicus cells were either loaded by incubation of between 20 and 40 min at 20°C with 5 µM BCECF-AM ( Invitrogen , Paisley , UK ) or by inclusion of 300 µM BCECF free acid in the patch clamp pipette ( solution P1b , Table S2 ) . Changes in intracellular pH were monitored by the ratio of the fluorescence emission at 525±25 nm when excited sequentially with 488 and 458 nm ( LSM 510 confocal microscope , Zeiss , Jena , Germany ) . For each excitation wavelength , the average fluorescence intensity was determined for a region of interest encompassing the whole cell and used to calculate the ratio . Background fluorescence was minimal and was not subtracted . In patch clamp experiments , pH was calibrated using mean steady state fluorescence ratio ( F488/F458 ) at pHi 7 . 5 and pHi 6 . 5 ( n = 10 and n = 5 , respectively ) . We were unable to achieve a satisfactory calibration for ester-loaded cells using the nigericin technique as BCECF fluorescence was not stable in C . pelagicus in the presence of this protonophore . Absolute pH values are therefore not given for ester-loaded cells; rather ΔpH values are given , calculated from the calibration performed in patch-loaded cells with pHi set by the pipette solution . The magnitude of ΔpH calculated in this manner closely matched ΔpH calculated from an in vitro calibration curve using 25 µM BCECF free acid in “cytosolic” buffer solutions ( 15 mM HEPES , 15 mM MES , 1 mM MgCl2 , 130 mM KCl , pH 6 . 5–8 . 5 ) . Location of intracellular coccoliths was determined by imaging in reflectance mode ( 633 nm excitation ) of the confocal microscope . Chloroplasts were visualised with 488 nm excitation and emission >600 nm . EhHVCN1 ( JGI protein ID: 631975 ) was identified by sequence similarity searches using animal Hv1 to query the E . huxleyi genome ( Joint Genome Institute , http://genome . jgi-psf . org/Emihu1/Emihu1 . home . html ) . CpHVCN1 was identified in a collection of ESTs generated from C . pelagicus spp braaudii strain LK1 by Peter von Dassow and co-workers at the Station Biologique de Roscoff , France ( Genbank accession no . HM560965 ) . To confirm expression and the coding sequence of these genes , 1 . 1 kb cDNAs corresponding to the open reading frame of EhHVCN1 or CpHVCN1 were amplified by reverse transcription-polymerase chain reaction from E . huxleyi strain CCMP1516 or C . pelagicus strain PLY 182 g using the primers EhHVCN1_F3 TCATCCCTCTCTTTGCGATG with EhHVCN1_R3 GGTCTTTGGAAACGGTTAGC and CpHVCN1_F7 GCAAATATTTTAGAAGGATGAGG with CpHVCN1_R4 GAGATTTGAACACGCGAAT . Due to the high GC content ( E . huxleyi ) and unusual codon usage ( both species ) , we synthesised codon-optimised versions of the transcripts for characterisation in mammalian expression systems ( GenScript , Piscataway , NJ ) . These inserts were subcloned into pcDNA3 . 1 via HindIII and XbaI . For electrophysiology , HEK293 cells were transiently co-transfected with 1 . 0 µg of pcDNA3 . 1 HVCN1 plus 0 . 4 µg of pCDNA3 . 1-eGFP using Lipofectamine LTX ( Invitrogen ) . To confirm EhHVCN1 and CpHVCN1 were expressed in HEK293 cells , these genes were subcloned into a pcDNA3 . 1-eGFP construct via HindIII and BamHI sites to generate green fluorescent protein ( GFP ) fusions ( EhHVCN1-GFP , CpHVCN1-GFP ) . HEK293 cells transfected with the GFP-fusions demonstrated localisation to both the plasma membrane and endomembranes for both proteins and exhibited very similar currents to the non-fusion proteins during patch-clamp recordings . For all further characterisations with the exception of the site-directed mutagenesis , EhHVCN1 or CpHVCN1 alone were used 24–48 h after transfection . Site-directed mutagenesis of histidine residues in EhHVCN1-GFP was performed using a QuikChange II Site-Directed Mutagenesis kit ( Stratagene , La Jolla , CA ) . All mutagenesis products were confirmed by DNA sequencing . Amino acid sequences of proteins were aligned using ClustalW . Accession numbers and protein length for Hv1 proteins are as follows: Homo sapiens ( NP_001035196 , 273aa ) , Mus musculus ( NP_083028 , 269aa ) , Gallus gallus ( NP_001025834 , 235aa ) , Danio rerio ( NP_001002346 , 235aa ) , Xenopus laevis ( NP_001088875 , 230aa ) , Ciona intestinalis ( NP_001071937 , 342 aa ) , Emiliania huxleyi ( JGI v1 . 0 prot ID: 631975 , 339aa ) , Coccolithus pelagicus ( HM560965 , 325aa ) , Phaeodactylum tricornutum ( XP_002180795 , 338aa ) , Thalassiosira pseudonana ( XP_002293360 . 1 , 293aa ) , Polysphondylium pallidum ( EFA75681 . 1 , 280aa ) , and Physcomitrella patens ( XP_001767834 . 1 , 198aa ) . Accession numbers for additional proteins used are as follows: Drosophila melanogaster - Shaker ( NP_728122 . 1 ) , Homo sapiens - Kv2 . 2 ( AF450111 ) , Arabidopsis thaliana - KAT1 ( NP_199436 . 1 ) , Aeropyrum pernix K1 - KvAP ( NP_147625 . 1 ) , and Bacillus halodurans C-125 - NaChBac ( NP_242367 . 1 ) . For the phylogenetic analysis , an alignment was constructed based on the conserved residues surrounding the four transmembrane domains . Sequences were aligned using MUSCLE and then manually corrected to ensure only unambiguous residues were compared . Maximum likelihood phylogenetic analysis was performed using PhyML within the Bosque software package [38] , based on the JTT substitution matrix [39] . One hundred bootstrap replicates were performed . HEK293 cells ( Health Protection Agency Culture Collection , Salisbury , UK ) were maintained in Dulbecco's Modified Eagle Medium at 37°C in 5% CO2 . Whole cell patch clamp recordings were performed at 20°C . HEK293 cells transfected with pCDNA3 . 1-eGFP alone were used as a control . The intracellular and extracellular solutions were based on those used by Sasaki et al . [20] ( see legends for Figures 4 , S4 and Table S2 ) . Calcification rate in C . pelagicus cells was quantified in vivo by monitoring the degree of birefringence of calcite using cross-polarized light microscopy . Cells were initially perfused with Ca2+-free ASW supplemented with 20 mM EGTA until decalcified , followed by perfusion with f/2 FSW and recovery for 2–3 h prior to imaging . Image capture was performed using a Nikon Diaphot microscope equipped with an Orca-100 cooled CCD camera ( Hamamatsu Photonics , Shizuoka , Japan ) . All recordings were performed at 15°C , light intensity 110 µmol m−2 s−1 . Time-lapse images were captured at a frame rate of 20 images h−1 . The change in grey scale image intensity , which is proportional to production of birefringent calcite , was determined using LSM Image Examiner software ( Zeiss ) .
|
The production of calcium carbonate structures by marine organisms has a major influence on the Earth's carbon cycle and is responsible for the eventual formation of sedimentary rocks such as chalk and limestone . The major contributors to marine calcification are the coccolithophores , a family of unicellular algae which surround themselves in calcified plates known as coccoliths . Unlike many other calcifying organisms , coccolithophores produce their calcified structures inside the cell , enabling precise control of this process . However , the other product resulting from the calcification reaction , H+ , must be rapidly removed to maintain the pH inside the cell . In this study , we show that coccolithophores possess a voltage-gated H+ channel , which removes H+ rapidly from the cell during calcification and helps maintain a constant pH . We identify the gene encoding this H+ channel , HVCN1 , and find that it is a distant relative of those recently identified in animal cells , suggesting that H+ channels may be present in many other types of eukaryote organism . As calcifying organisms may be affected by ocean acidification , the identification of an H+ channel in coccolithophores gives us an important mechanistic understanding of cellular pH regulation during the calcification process , and may give insight into the response of coccolithophores to future changes in ocean pH .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"physiology",
"cellular",
"stress",
"responses",
"molecular",
"cell",
"biology",
"phycology",
"cell",
"biology",
"electrophysiology",
"physiology",
"marine",
"and",
"aquatic",
"sciences",
"biology",
"anatomy",
"and",
"physiology",
"marine",
"biology"
] |
2011
|
A Voltage-Gated H+ Channel Underlying pH Homeostasis in Calcifying Coccolithophores
|
The analysis of methylation patterns is a promising approach to investigate the genealogy of cell populations in an organism . In a stem cell–niche scenario , sampled methylation patterns are the stochastic outcome of a complex interplay between niche structural features such as the number of stem cells within a niche and the niche succession time , the methylation/demethylation process , and the randomness due to sampling . As a consequence , methylation pattern studies can reveal niche characteristics but also require appropriate statistical methods . The analysis of methylation patterns sampled from colon crypts is a prototype of such a study . Previous analyses were based on forward simulation of the cell content of the whole crypt and subsequent comparisons between simulated and experimental data using a few statistics as a proxy to summarize the data . In this paper we develop a more powerful method to analyze these data based on coalescent modelling and Bayesian inference . Results support a scenario where the colon crypt is maintained by a high number of stem cells; the posterior indicates a number greater than eight and the posterior mode is between 15 and 20 . The results also provide further evidence for synergistic effects in the methylation/demethylation process that could for the first time be quantitatively assessed through their long-term consequences such as the coexistence of hypermethylated and hypomethylated patterns in the same colon crypt .
Most tissues are renewed during the life of the organism through a continuous replacement of their differentiated cells by new mature cells that originate from tissue-specific stem cell lineages . Genetic and epigenetic somatic variations having long-term consequences are believed to preferentially affect these cell lineages . Besides its intrinsic interest , understanding the structure of the stem cell populations is therefore expected to help gene therapy , cancer therapy , and aging research . Our knowledge about the adult tissue-specific stem cells is still sparse , but the experimental results accumulated during the past decades on several tissues from a few organisms suggest the existence of stem cell niches [1 , 2] . Each niche contains a small self-renewing population of proliferating cells , the stem cells , whose progeny commit to differentiation processes that span several rounds of cell division . In some tissues , such as Drosophila ovary and testis , cells that enter differentiation processes have been shown to be the result of asymmetric stem cell divisions giving rise to one stem cell and one cell committed to differentiation [3] . In this context , rare symmetric stem cell division events producing either two cells committed to differentiation or two stem cells may compensate for occasional gain or loss of one stem cell [4] . Cell turnover is particularly fast in the gastrointestinal epithelium where the mature differentiated cells live only a few days . In this epithelium , the self-renewing unit is a small and morphologically well-identified structure known as the intestinal crypt . Probably owing to this relatively simple anatomical structure and its rapid cell turnover , the intestinal epithelium is one of the tissues where the stem cells have been the most studied in mammals . Results have been recently reviewed in [5–7] . Each crypt contains about 2 , 000 cells in Human and exhibits a strong polarity with mature epithelial cells located at the extremity of the crypt opening onto the gastrointestinal tract . Phenotype markers allowing an accurate identification of intestinal stem cells are still lacking , although some progress is being made [8] . Staining experiments of cell lineages in mouse using tritiated thymidine show that cells migrate away from the base of the crypt toward the lumen while they differentiate [9 , 10] . The stem cell niche is therefore believed to be located at the base of the colon crypt . In the small intestine , the situation is slightly different as Paneth cells occupy the very base of the crypt ( a type of mature differentiated cell absent in the colon crypt ) and the stem cell niche is believed to be located just above the Paneth cells . Gastrointestinal stem cells have a short cell cycle: they seem to divide daily in the mouse small intestine and may divide about weekly in the human colon [5 , 6 , 11] . Staining experiments with tritiated thymidine in mouse reveal asymmetric segregation of the DNA between daughter cells during stem cell division: stained template strands are retained through rounds of cell divisions at the likely location of the stem cell niche [12] . This suggests that mechanisms reducing the probability of mutations in the stem cell lineages may exist . It also provides strong support for a central role of asymmetric stem cell division in the stability of the intestinal stem cell populations . Experiments with chimeric mice and with mutagen agents have shown that crypts initially polyclonal for one marker eventually become monoclonal , which suggests that symmetric division occasionally occurs and leads to a niche succession by stochastic extinction of stem cell lineages . The time needed for a single stem cell lineage to replace all other lineages , or “clonal stabilization time , ” was measured as approximately 24 weeks in the mouse small intestine but only four weeks in the mouse colon . Studies are much harder in human where mutagenesis experiments cannot be undertaken . Making use of radiotherapy-induced mutations , one study suggested that a significant fraction of the somatic mutations in human colon stem cells are lost within one year [13] . Most of the insights we have about the number of stem cells in intestinal crypts come from mouse studies where the number of crypts surviving increasing doses of radiation is measured . The underlying hypothesis of these studies is that the number of stem cells can be estimated assuming a simple model for dose-dependent cell death if crypts regenerate , as long as one stem cell survives the treatment . However , the estimated number of stem cells increase from about six to 36 as higher doses of radiation are used . It may be that a cell that would differentiate in a normal context can be recruited to regenerate the crypt [5] , but alternative explanations cannot be ruled out [14] . This also emphasizes the difficulty of estimating the actual number of stem cells in normal physiological conditions by this approach . Another limitation of this approach is that it cannot be used in humans . We recently proposed to investigate the properties of the human intestinal stem cell populations through the analysis of methylation pattern polymorphism that occurs naturally within the crypts in some CpG islands [15 , 16] . Methylation at CpG sites serves as cell lineage markers in these analyses as inheritable methylation changes occur somatically [17] . Compared with genetic DNA sequences that are widely used in population studies at the organism level , epigenetic methylation patterns have the advantage of evolving much faster . Even a small number of CpG sites can carry information about the underlying genealogy of cells sampled from closely related lineages . Several aspects of the approach make it particularly attractive: it does not require any treatment that could disturb the normal self-renewal dynamic of the crypt; it provides accurate discrete data richer than binary phenotype polymorphism; and it is practicable in humans . Our previous analyses of the methylation pattern polymorphisms were based on forward simulations of the whole crypt content under a simple stem cell–niche model [15 , 16] . Although it provided support for a stem cell–niche model with multiple stem cell lineages whose genealogy shows coalescent events leading to stochastic niche succession during life , this approach allows only rather crude estimate of the model parameters . In particular , it did not give us an estimate of the number of stem cells . This study further analyzes these data in light of the simple stem cell–niche model using a more sophisticated methodological framework . We present a full probabilistic model of the methylation patterns sampled from a crypt together with a Markov chain Monte Carlo ( MCMC ) algorithm allowing Bayesian inference of its parameters that include the number of stem cells , the depth of the genealogical tree ( niche succession time ) , and the rate of the methylation/demethylation process . The fit of the model is assessed by comparing the observed values of several statistics summarizing the data with their posterior distribution under the model . We compare the estimates of the model parameters in either an unconstrained model or a model with short niche succession time .
In this study , we analyze methylation at nine CpG sites in a 77-bp locus within a CpG island upstream of the human biglycan ( BGN ) gene on chromosome X [15] . Our data consist of methylation patterns at loci randomly sampled from individual colon crypts from seven male patients between 40 and 87 years old . The total of 57 crypts studied here can be divided into eight-pattern and 24-pattern datasets . The eight-pattern data comprise the sequences of between five and 14 methylation patterns for 37 crypts isolated from five patients whom we described in a previous study [15] . The 24-pattern data are new , and correspond to the sequences of between 20 and 24 methylation patterns for another 20 crypts sampled from three patients ( one patient is common to both the eight-pattern and 24-pattern datasets ) . The patterns sampled from these 20 crypts are shown in Figure 1 . All patients are males and therefore haploid for the BGN locus . This simplifies the modelling of the genealogy of the sampled sequences , as it ensures a one-to-one correspondence between sequences and cell lineages . Polymorphic methylation patterns arise from methylation and demethylation events that take place in the genealogy of the sampled cells . We propose a probabilistic model of the observed polymorphism in terms of biologically meaningful parameters that govern both the shape of this genealogy and the methylation process . Events taking place in stem cell lineages are believed to play a crucial role in the generation of the observed patterns , and we first present how the model accounts for the genealogy of those lineages along with their methylation process . We next explain how the model relates the sampled patterns to the patterns of the stem cell lineages . The shape of the stem cell genealogy is described by two parameters , N and τ . N is the number of stem cells in a crypt and τ corresponds approximately to the average number of years before niche succession within a crypt . Formally , τ/2 is the average waiting time for the coalescence of a pair of lineages , whereas , from a biological point of view , τ/ ( N−1 ) is the average lifespan of a stem cell . Methylation and demethylation in the stem cell genealogy is modelled as a point process whose rates ν are expressed in terms of expected number of events per CpG site in time τ ( before niche succession time ) . In agreement with experimental observations [15] , CpG sites at the BGN locus are modelled as initially unmethylated at the birth of the patient . Two models were compared for the somatic methylation/demethylation process . In a first model , we distinguish methylation and demethylation rates , but sites are independent . In a second , more complicated , model we allow for interactions between sites through rates that depend on the current level of methylation of the locus . We later refer to this model as the context-dependent model . Sampled methylation patterns are related to those found in the stem cell lineages through three parameters , g , α , and ɛ . The parameter g can be interpreted as the number of cell cycles of the cell differentiation process . It describes the shape of the genealogy of cells sampled from the progeny of the same stem cell ( illustrated in Figure 2 ) . Higher values of g correspond to genealogies that tend to have longer terminal branches ( and so are more star-like ) . The parameter α reflects the ratio between the amount of methylation and demethylation in a cell lineage during the cell differentiation process and in a stem cell lineage in time τ . More precisely , α = η/ν , where η is expressed in terms of the expected number of events in a single lineage during the few cell divisions of the differentiation process . The parameter ɛ corresponds to the rate of sequencing error per site per sequence . Inference is carried out in a Bayesian framework using an MCMC algorithm designed to sample the posterior distribution of the parameters given the data . An uninformative prior is used . Parameters N , τ , g , and ɛ can take values in the intervals ( 2 , 50 ) , ( 0 . 5 , 200 ) , ( 5 , 10 ) , and ( 0 , 1 ) , respectively . Parameters ν , α can take any positive values . The inference framework was validated on simulated datasets . The methodology developed by Gelman et al . [18] served to assess the model adequacy . It consists in checking , through a set of statistics , the extent of the discrepancies between the data and datasets simulated with the model using parameters sampled from their posterior . Here data are summarized by the intercrypt average and standard deviation of five within-crypt statistics . These summary statistics are the number of distinct patterns , the number of polymorphic sites , the average pairwise distance between patterns ( the average number of sites with distinct methylation status between pairs of patterns ) , the number of entirely unmethylated patterns , and the number of singletons ( those patterns that appear only once in the crypt ) . Expected distributions and empirical values of the five statistics are plotted in Figure 3 . The fit of the model with context-dependent methylation rate is much better than the fit of the model with independent sites . This can be seen for instance in the average and standard deviation of the number of distinct patterns per crypt; in the standard deviation of the average distance between patterns of the same crypt; or in the average number of unmethylated patterns . However , the model with a context-dependent methylation rate still shows some lack of fit . In particular , intercrypt variation in the number of singletons is higher than expected for the 24-pattern datasets . A detailed patient-by-patient analysis of the summary statistics ( Table S1 . 1 in Protocol S1 ) reveals that this discrepancy may be explained by variability between patients: the observed intercrypt average of the number of singletons by patient falls above the upper limit 99% confidence interval of its expected distribution for patient X ( 4 . 0 singletons in average ) and appears just at the lower limit of the 95% interval for patient X ( 0 . 86 singletons in average ) ( see Figure 1 for visual inspection ) . This observation encouraged us to ignore data from patient X . The results then show a remarkable match between observed and simulated values ( Figure 3 , right , and Table S1 . 1 in Protocol S1 ) . Such fit could not be achieved by exclusion of either patient Y or M ( Table S1 . 2 in Protocol S1 ) . Besides , data from patient X were found to shift the posterior of number of stem cells to higher values of N , probably due to the high number of singletons ( Figure 4 ) . The consequences of excluding either patient Y or patient M on the posterior of N were also explored ( Figure S1 . 1 in Protocol S1 ) , and patient X was found to impact the most on the posterior of N . Ignoring patient X actually softens our conclusions on the number of stem cells and can therefore be regarded as a conservative choice which further illustrates the relevance of a careful assessment of the model fit . We were also able to show that enhanced rates of methylation/demethylation are a possible hypothesis to explain the data from patient X ( Figure S1 . 2 and Table S1 . 3 in Protocol S1 ) . The marginal posterior of each parameter of the model with context-dependent methylation rate obtained after removing data from patient X is shown in Figure 5 ( N and ν ) and Figure 6 ( τ , g , α , and ɛ ) . The posterior distribution of N , τ , and g give a clear picture of the main features of the shape of the genealogy . The posterior distribution of the number of stem cells , N , reaches its mode between 15 and 20 , gives little support for values of N below 8 , and seems to exclude any value of N smaller than 6 . The posterior distribution of τ suggests that the actual value of this parameter , which corresponds approximately to the average time before the stem cell population finds its most recent common ancestor , is located between 15 and 40 years . Finally , there seems to be very little information on parameter g that accounts for the shape of the genealogy of the cells sampled from the progeny of the same stem cell: the posterior distribution of g closely matches its continuous uniform prior on ( 5 , 10 ) . The posterior distributions of the parameters that explain the polymorphism given the genealogy are also enlightening . Parameter ν reveals synergistic methylation/demethylation across the sites of the BGN locus . The methylation rate is found to be highly dependent on the number of already methylated sites . The rate is very low when no sites are methylated and shows a more than fivefold increase when one site is already methylated ( the median of ν moves from 0 . 05 to 0 . 35 methylation events per site in time τ ) . It is then relatively constant up to seven methylated sites and then increases again . In contrast , demethylation dynamics does not seem to depend on the current level of methylation . The posterior distribution of the parameter α suggests that methylation/demethylation events during cell differentiation contribute little to the observed polymorphism: its density decreases sharply between zero and 0 . 04 , where it becomes negligible . Finally , the posterior distribution of ɛ indicates that sequencing errors are extremely rare ( rate smaller than 0 . 004 and more likely between zero and 0 . 002 ) . It is worth emphasizing the overall coherence of the picture that emerges of this posterior inference: the genealogical trees of the stem cell lineages up to their most recent common ancestor are rather deep ( high τ ) , and most methylation/demethylation events occur in those lineages ( small α ) . Furthermore , the value of α seems compatible with the same probability of methylation/demethylation events per cell cycle during differentiation and in the stem cell lineages . A value of τ of about 20 years suggests about 1 , 000 rounds of cell divisions in the stem cell lineages before these lineages find their most recent common ancestor ( stem cells are thought to divide about weekly in the human colon [6 , 11] ) , whereas the number of rounds of cell divisions during differentiation is certainly smaller than ten . Therefore , we expect a value of α smaller than 0 . 01 = 10/1 , 000 . The comparison with the posterior obtained from the full dataset including patient X ( Figures S1 and S2 ) reveals that the data from patient X not only impact on the posterior of N but also to a lesser extent on the distribution of α , ɛ , and the demethylation rates of hypomethylated sequences . In each case , patient X shifts the distribution toward higher values of the parameters . Inspection of the posterior obtained when considering only the 24-pattern data without patient X ( Figures S3 and S4 ) indicates that most of the information on N , g , α , and ɛ is contained in the 24-pattern datasets while the eight-pattern datasets greatly contributed to the information on τ and to a lesser extent to the information on ν . Unconstrained posterior inference suggests that the high level of intracrypt polymorphism is due to the existence of many stem cells in each crypt ( high N ) . An alternative explanation for this high level of polymorphism could be a significant amount of methylation/demethylation events taking place in differentiation lineages ( high α ) . Although this hypothesis does not receive support from our analysis , a closer look at the posterior reveals a negative correlation between N and α shown in Figure 7 . The posterior of N for values of α below 0 . 01 virtually excludes values of N below ten , but a number of stem cells below ten becomes likely when α increases . These observations allow a better interpretation of the posterior of α and N obtained in the unconstrained analysis: a number of stem cells between eight and ten is unlikely but may be compatible with the data if α > 0 . 01 . In the context of a niche succession time of 20 years , this would indicate that the rate of methylation/demethylation is enhanced during cell differentiation . The model and its associated inference procedure provide an invaluable tool to further explore the hypothesis of a high amount of pattern changes during cell differentiation . A relatively high value of α is particularly realistic in a scenario with short niche succession time ( small τ ) . As an illustration , we investigate here what could be the consequence of a niche succession time of about one year ( τ = 1 ) . We can see in Figure 8 that setting the parameter τ to 1 has no impact on the posterior of N . We also found that the constraint τ = 1 has no impact on the estimate of α ( unpublished data ) and g ( Figure 8 , right panel ) . From a biological point of view , however , a short niche succession time does not seem compatible with a very small value of α . Indeed , a τ of one year suggests that niche succession may be reached after only 50 cell cycles , compared with a minimum of five cell cycles for cell differentiation . We thus expect α greater than 0 . 1 = 5/50 . The posterior of N was therefore investigated subject to this constraint . Results are presented in Figure 8 and reveal the deep impact of such a high value of α on the posterior of N , which now indicates a value of N between four and 12 , whereas the value of g is close to five . This impact of α prompted us to understand why the data do not support a high value of α under our model assumptions . The posterior of g concentrated close to its lower bound suggest that g > 5 might be incompatible with α > 0 . 1 . This hypothesis is confirmed after examination of the posterior of g when allowing g to take values smaller than five ( see Figure 8 , right panel ) . The density decreases sharply between zero and three and excludes values of g higher than four . Data seem , therefore , incompatible with a scenario where a significant fraction of the methylation/demethylation events take place during cell differentiation across star-like genealogies .
The analysis of methylation patterns is a promising approach to investigate the structure of cell populations in an organism [15 , 19–21] . In a stem cell–niche scenario , sampled methylation patterns are the stochastic outcomes of a complex interplay between niche structural features such as the number of stem cells within a niche and the niche succession time , the methylation/demethylation process , and the randomness due to the sampling . As a consequence , methylation pattern studies can reveal niche characteristics but also require appropriate statistical methods . The analysis of methylation patterns sampled from colon crypts is a prototype of such a study . Previous analyzes were based on forward simulations of the whole cell content of the crypt and subsequent comparisons between simulated and experimental data using a few statistics as a proxy to summarize the data ( number of distinct patterns per crypt , average methylation , and intracrypt distance ) . In this paper we develop an alternative inference framework based on likelihood computations that make full use of the data rather than making use of the values of a few summary statistics . Our assumptions about the biological mechanisms underlying the methylation patterns we observe are essentially the same as in previous works but we reformulate the model backward in time as a coalescent process . Coalescent modelling is a starting point for carrying out likelihood inference that makes use of the full data . It is also interesting by itself as it permits direct simulations of the small part of the whole crypt history that is relevant for explaining observed samples of methylation patterns , the history of the sampled cell lineages . We developed an MCMC algorithm that allows inference of all the parameters of the model . It is worth emphasizing that the inference relies on a number of model assumptions that we can summarize in four points: ( 1 ) the stochastic process generating the methylation pattern we sample is the same in all crypts , except for the age of the crypts , which differ among patients; ( 2 ) the number of stem cells is stable and the loss of a stem cell is compensated by the symmetric division of one of the other stem cells . This justifies modelling the stem cell genealogy as a simple coalescent process ( this supposes , for instance , that stem cell losses are not compensated by new stem cells originating from another layer of stem cells ) ; ( 3 ) the genealogy of differentiation is star-like when branch lengths are measured through the accumulation of methylation/demethylation events; ( 4 ) appropriate modelling of methylation/demethylation can be achieved using a simple continuous-time point process across lineages . These assumptions are intended to be as reasonable as possible , and the demonstration that the data are compatible with this simple model is one merit of this work . Careful assessment of the model fit , nevertheless , revealed the need for partial relaxation of hypotheses ( 1 ) and ( 4 ) . Data from patient X were ignored owing to too high a number of singletons , and a context-dependent model allowing methylation and demethylation rates to vary with the number of already methylated sites was introduced . In the future , additional data may reveal the need for further refinement of the model , and we can envision modelling parameter variation across crypts and patients or introducing distinct parameters for the methylation processes in stem cell and differentiation lineages . The need for a context-dependent model of the methylation process brings another piece of evidence for synergistic methylation processes that are also supported by a number of experimental studies . However , this seems to be the first study that directly assesses those effects on the basis of their long-term consequences such as the transition between hypomethylated and hypermethylated sequences that translate into the coexistence of both types of sequences within the crypt . These effects are likely to rely on interactions between maintenance methyl-transferase ( Dnmt1 ) and de-novo methyl-transferase ( Dnmt3a/b ) . It is believed that de-novo methylation by Dnmt3a/b ( de-novo methyl-transferase ) is stimulated by Dnmt1 ( maintenance methyl-transferase ) acting to maintain the methylation status through the methylation of the newly synthesized DNA strand at hemimethylated sites after replication [22 , 23] . This mechanism could explain the increase in the rate of methylation while the rate of demethylation remains stable . Development of models that can effectively account for both dependencies between sites and variation in methylation/demethylation rates across CpG sites will be an interesting challenge . Our results show that the average distance between patterns and the number of distinct patterns unambiguously call for context-dependent effects . These statistics are known to be important indicators of the effective population size in population genetics [24 , 25] . However , examination of the level of methylation at each site of the BGN locus suggests the existence of some differences between sites [15] . The main challenge in the estimation of the number of stem cells from methylation patterns is to distinguish polymorphism due to methylation/demethylation events in stem-cell lineages and the polymorphism due to events during cell differentiation . The framework proposed here allows us to address this issue in a quantitative manner . The results suggest that the methylation changes in differentiation lineages are rare while the number of stem cells is higher than eight and reaches its posterior mode between 15 and 20 . The small contribution of the events taking place in differentiation lineages to the diversity of methylation patterns is coherent with our estimate of a relatively long niche succession time of more than 15 years . This “high N–high τ” scenario can account for the rapid decline in the number of partially mutant crypts reported by Campbell et al . [13] , as it is compatible with a short life of individual stem cells ( average τ/ ( N−1 ) ) . The almost negligible amount of methylation events in the few cell divisions of the differentiation process is also in agreement with small rates of methylation/demethylation experimentally measured in some human cell lineages around 0 . 001 change/site/generation [26–28] , but contrasts with the high rates found in other lineages or other loci where errors in the replication of the methylation pattern attain 0 . 01 to 0 . 15 change/site/generation [22 , 28 , 29 , 30] . Changes are believed to occur when sites are hemimethylated at the time of DNA replication after either de novo or incomplete maintenance methylation . In the future , experimental assessment of the frequency of those hemimethylated sites by double-strand DNA methylation pattern sequencing [28] in cells sampled from the crypt may help to further calibrate the amount of methylation changes in the differentiation lineages . Sequence redundancy artifacts caused by bisulfite treatment and PCR amplification have been detected in the context of bisulfite sequencing , although in a distinct experimental setting [31] . In this study we limited our analysis to a short 77-bp locus , sampled a relatively small number of times ( fewer than 24 ) compared with the number of copies available from the biological sample ( more than 1 , 000 ) , and amplified in four independent PCRs . All these precautions certainly decrease the chances of artifacts [32 , 33] . In addition , we have previously reported pattern redundancy at the BGN locus between samples from both parts of bisected crypts [15] , and this suggests that a substantial fraction of the pattern redundancy reported here reflects genuine biological redundancy . Finally , our main biological conclusions are probably robust against limited sequence redundancy artifacts . Little is known about those artifacts , but intuitively we think they are more likely to decrease than to increase the apparent number of stem cells , and they can hardly be invoked in place of synergistic methylation to explain the coexistence within the same crypts of a substantial proportion of fully unmethylated patterns and a diversity of related methylation patterns . We nevertheless acknowledge the need for better verified data; molecular barcoding will be the method of choice for this purpose [29 , 31] . The inference framework proposed in this study will be an invaluable tool to address questions related to the design of future experiments . We could , for instance , wonder whether it will be better to assess more cells or to get higher resolution in the description of the cell lineages by sequencing more CpG sites . As sequencing longer patterns could prove technically hard , an intermediate route could consist in sequencing additional loci sampled from the same crypt but from independent cells . Our approach could rather easily be extended to handle this kind of multilocus data as well as diploid locus data . We also envision the development of less computationally demanding inference methods based on summary statistics , for which our work provides both a simulation tool and a benchmark .
Individual crypts were isolated from fresh colectomy specimens , and , after extracting the crypt DNA content , unmethylated cytosines were converted into uracil by bisulfite treatment . The bisulfite-treated DNA was further amplified by quantitative PCR , and the BGN locus of a relatively small number of molecules ( five to 24 ) was sampled and sequenced . Patients had colectomies for adenocarcinoma , but the normal colon crypts examined in this study were taken at least 10 cm away from the tumors , and are unlikely to be directly involved with tumorigenesis . Details of the experimental protocol can be found in [15] . We model the size of the stem cell population in a crypt as identical in all crypts , in all patients , and kept at a constant value N throughout the life of the patients . A stem cell is said to die if it gives birth to two cells committed to differentiation . The death of one stem cell is assumed to be instantly compensated by a symmetric division of one of the remaining stem cells ( chosen at random ) that produces two stem cells . We model the lifespan of a stem cell as an exponential random variable with rate γ ( mean 1/γ ) . Under these assumptions , when looking backward in time the stem cell genealogy follows a particular coalescent process [34] , a version of the Moran model [35] . The waiting time for the coalescence of two stem cell lineages is an exponential random variable with rate 2γ/ ( N−1 ) . When considering k lineages , the waiting time for the first coalescence event of two lineages is exponential with rate γk ( k−1 ) / ( N−1 ) . Furthermore , the two lineages that coalesce are chosen at random with probability 2/k ( k−1 ) . This representation would be clearly inadequate if used for the entire cell population of the colon crypt , as it would not account for the very different behavior of the stem cells and differentiating cells [36] . We explain later how our model accounts for the properties of the lineages of differentiating cells . In agreement with our previous experimental observations [15] , all the CpG sites that we analyze here are supposed unmethylated at the birth of the patient , and the observed methylation patterns are supposed to result from methylation and demethylation events that took place across the genealogy of the sampled cells . In a first simple model , we assume that every CpG site evolves independently at the same constant rate . Such a model has two parameters μ = ( μ+ , μ− ) , where μ+ is the rate of methylation per site and μ− is the rate of demethylation per site . We also introduce a second , more sophisticated , model that accounts for context-dependent effects on the evolution rate . In this model , the rate of methylation and demethylation is allowed to vary with the number of already methylated sites . To avoid introducing a number of parameters as large as twice the number of CpG sites , we consider only four sets of methylation/demethylation rates that apply on four distinct ranges of numbers of methylated sites . Range boundaries will be estimated . In the general case , the use of models where sites do not evolve independently requires prohibitively heavy computation . However , when all sites are supposed to evolve according to the same model , we show how to take advantage of the model symmetry to considerably speed up the computations ( see Protocol S2 , section 1 , “Calculations in the context-dependent model” ) . To our knowledge , it is the first use of this context-dependent model for modelling biological sequences . Methylation patterns sampled within the crypt are not directly sampled from stem cell lineages . Most of them come from differentiated and differentiating cells that are the product of the few rounds of cell division that take place during cell differentiation . The next two paragraphs explain how our model relates the sampled methylation patterns to the stem cell methylation patterns . We suppose that the cell content of the crypt is composed of N equal-sized subpopulations , each one corresponding to the progeny of one of the stem cells . As we know that differentiated cells are short-lived and that the differentiation process spans only a few generations of cells , it is attractive to assume that all cells sampled from the same subpopulation share the methylation pattern of their most recent common stem cell ancestor . Under this hypothesis , we do not need to keep track of the precise genealogical process that goes back in time from the sampled cell to the most recent ancestral stem cell . All we need is to model which sampled cells come from the same stem cell lineage . A model of sampling with replacement is fairly reasonable for this purpose as we believe that subpopulations are large compared with the number of sampled patterns and DNA is amplified by PCR before sequencing . In this model , when we sample n sequences , we actually sample a random number M ≤ N of stem cell lineages . The probability mass function for Γ , the random variable that defines M and the partitioning of the n sequences into M groups , is given by Preliminary numerical experiments suggested that estimates of N based on this simple model would be misleading even for a limited amount of methylation/demethylation during cell differentiation . Therefore , we preferred a more general model that can account for methylation and demethylation during cell differentiation . For this purpose , the model needs to describe the genealogy of the cells sampled from the same subpopulation ( Equation 1 ) back in time until its ancestral stem cell lineage . Under our model , the genealogy of q cells sampled from the same subpopulation results from a coalescent process between random pairs of lineages whose times ( sq , sq−1 , … , s2 ) are drawn according to the probability density function where 0 ≤ sq ≤ sq−1 ≤ … ≤ s2 <1 and ωg is the first derivative of Ωg , an integrated rate function defined by Ωg ( t ) = −log ( ( 2g−2gt ) / ( 2g−1 ) ) . The parameter g controls the star-likeness of the genealogy of the subsample back in time until the lineages of all the differentiated cells of the crypt are stem cell lineages . The form of the integrated rate function Ωg is justified in Protocol S2 ( section 2 , “Genealogy of cells sampled from the progeny of the same cell” ) as a continuous-time approximation of the discrete genealogical process of the subsample under a simplistic model where each subpopulation is the result of g rounds of cell divisions . The timescale of this process is expressed in arbitrary units that we do not try to compare with the timescale of the genealogy of the stem cell lineages . Rather , the methylation/demethylation process across the branches of this genealogy has its own rate denoted η = ( η+ , η− ) that is proportional to μ = ( μ+ , μ− ) , the rate of the methylation/demethylation process in the stem cell lineages . Possible sequencing errors were also accounted for through a parameter ɛ which corresponds to the probability of error at one CpG site of a methylation pattern . Although we tried to introduce as few parameters as possible in our model , any estimate of its parameters will clearly be associated with a relatively high level of uncertainty due to the limited amount of data . The Bayesian statistical framework provides a straightforward approach to account for this uncertainty by allowing us to compute the posterior distribution of the parameters given the available data . However , the choices of a parametrization and a prior distribution of the parameters are important issues in the Bayesian context . Without reliable a priori information , it is natural to look for an uninformative prior . To our knowledge , current Bayesian methodology provides few guidelines that may be useful in the context of our study . The approach we decided to adopt consists of finding a parametrization that minimizes the dependencies between the parameters according to the posterior distribution . Several motivations justify this approach: it makes reasonable the use of independent priors for each parameter , it facilitates the interpretation of the posteriors , and it helps in designing efficient MCMC algorithms to explore the posterior . We chose a uniform distribution for the number of stem cells , N , which is the primary focus of our interest . N is also the only parameter that impacts on M , the random number of stem cell lineages sampled in a particular crypt . The speed of the coalescent process modelling the stem cells genealogy is a function of the ratio γ/ ( N−1 ) . We denote the inverse of this ratio by τ = ( N−1 ) /γ , which corresponds approximately to the expected number of years before the entire stem cell population finds a common ancestor ( without considering the truncating effect of the birth on this coalescent ) , and chose a uniform prior on ( 0 . 5 , 200 ) for it . This seems a better choice than direct modelling of γ , the expected lifespan of a stem cell , as when n is small enough compared with N then M is equal or close to n and the only effects we observe are those of the ratio γ/ ( N−1 ) . Under these conditions , the posterior distribution of γ ( but not τ ) will be highly correlated with that of N . The parameter g that accounts for the star-likeness of the genealogy of the cells sampled from the progeny of the same stem cell was chosen from a uniform density on the interval ( 5 , 10 ) . Concerning the methylation process , it is worth mentioning that the rate of the coalescent and the overall speed of the methylation process are distinguishable only if the methylation pattern in the most recent common ancestor of the stem cell population does not follow the stationary distribution of the methylation process . When the stationary distribution of the methylation process is reached in this ancestor , the data carry information only about the relative speed of the methylation compared with the depth of the genealogical tree ( μτ ) . As a consequence , μ and τ can be highly correlated under their joint posterior while ν = μτ and τ would be relatively independent . We therefore preferred a prior that models ν independent of τ rather than μ independent of τ . Looking for an uninformative prior on ν , we chose a log-normal distribution such that log ( ν ) is normally distributed with mean zero and standard deviation σ . The potential difficulty of the choice of σ was bypassed by modelling σ as an exponential random variable with mean one ( a strategy known as hyper-prior modelling ) . The rate η of the methylation process relative to the arbitrary timescale of the genealogy of the differentiation lineages was chosen as η = αν where α follows an exponential distribution with mean one . The parameter α corresponds to the relative amount of methylation/demethylation events taking place before a sampled cell finds its stem cell ancestor , compared with the number of events occurring in the lineage of this stem cell up to the most recent common stem cell ancestor . Finally , we modelled the probability of sequencing errors , ɛ , as a continuous uniform on ( 0 , 1 ) . The posterior distribution of the parameters has been investigated using an MCMC algorithm [37] . Denoting X the observed methylation patterns , the purpose of the algorithm is to sample from the joint posterior distribution of the parameters θ = ( N , τ , g , σ , ν , α , ɛ ) given X ( we use bold fonts to emphasize where there is a random variable per crypt analyzed ) . An MCMC algorithm creates a sample of dependent realizations from the target distribution by updating in turn the components of θ , each update preserving the target distribution . For practical reasons , the MCMC algorithm samples an augmented space much larger than θ . It consists of ( θ , Λ , Y ) , where Λ denotes the genealogies of the methylation patterns ( topology and coalescent times ) and Y stands for the methylation patterns in the nodes of Λ . Updating N in our model is difficult and we propose an original strategy to solve the problem . As explained in Protocol S2 ( section 3 , “A slightly modified model for the number of stem cells” ) , our approach consists of embedding our model in a slightly more general model that allows N to differ by one between crypts . Another challenging task of the algorithm is to explore the huge space of possible genealogies Λ . This is performed efficiently using the “branch-swapping” strategy introduced by Wilson and Balding [38] . In this scheme , the sequences Y at the nodes of Λ serve to propose relevant modifications of the genealogy . Further details of the algorithm are given in Protocol S2 ( section 4 , MCMC algorithm ) . We obtained all the results presented here by running the algorithm for 5 , 000 , 000 iterations and recording ( N , τ , g , ν , σ , α , ɛ , Λ , Y ) every 100 iterations once past the first 500 , 000 iterations . Systematic visual checking of the samples did not reveal convergence problems . Each run of the MCMC algorithm takes approximately ten days on a 3-GHz Intel Pentium processor when the model with context-dependent methylation rate is used . Software and data can be downloaded at http://genome . jouy . inra . fr/~pnicolas/mcmcniche/ . Bayesian methodology ensures that posterior distributions are meaningful “on average” for values of the parameters drawn from their prior , although the data are generally compatible only with a small fraction of the parameter values allowed by the prior . It is therefore a good idea to check that the prior gives relevant results for combinations of parameters having some level of compatibility with the data . We validated our inference framework on ten synthetic datasets . Five were simulated with N = 6 , whereas the other five were simulated with N = 24 . Two sets of values for τ , ν , α , g , and ɛ were chosen after running the MCMC algorithm with N constrained either to six or 24 . Both sets of parameters are given in Protocol S2 ( section 5 , “Parameters used to generate simulated datasets” ) . They differ mostly in the value of α , which reflects the relative contribution to the polymorphism of the methylation/demethylation events taking place during cell differentiation compared with those occurring in stem cell lineages . The six and 24 stem cell datasets were simulated with α = 0 . 082 and α = 0 . 018 , respectively . Posterior distributions were able to distinguish between both series of datasets ( Figure 9 ) . For all but one dataset simulated with N = 6 , the posterior shows a clear peak around five or six while the last dataset is less informative as the posterior gives a similar support for any value of N greater than five . On the other hand , posterior distributions obtained on datasets simulated with N = 24 were all found to increase slowly between N = 5 and N = 15 and to be relatively flat for N greater than 15 . Two kinds of effects combine to explain that posteriors of N obtained for large N are flatter than posteriors found for small N . First , the posterior mechanically becomes flatter as N increases because models with neighbor values of N tend to look more and more alike . Second and less obvious , the data carry an amount of information on N that is limited by the number of cells sampled from each crypt , as all the information on N comes from the fact that some patterns are sampled from the progeny of the same stem cell ( see Equation 2 ) . When N becomes large compared with the number of cells sampled in each crypt , each cell tends to belong to the progeny of a different stem cell , and at best we may be able to say that N is large compared with the number of patterns sampled .
|
The dynamics of the stem cell populations in human colon crypts are of interest to cancer researchers and stem cell biologists alike . One approach to studying stem cell divisions would be to adopt methods from population genetics: cells are sampled from crypts , DNA markers such as single nucleotide polymorphisms are identified , and a model of how these mutations arose is used to infer aspects of the ancestry of the sample . Because cells within an individual are being studied , mutations of this sort are extremely rare , and an alternative marker has to be used . Methylation patterns provide a feasible alternative , containing information similar to that obtained from short DNA sequences . The present study shows how such data can be used to infer aspects of stem cell dynamics , including inference about the likely number of stem cells in a crypt . In addition , biological aspects of methylation and demethylation are also studied .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"mathematics",
"gastroenterology",
"and",
"hepatology",
"computational",
"biology",
"homo",
"(human)"
] |
2007
|
The Stem Cell Population of the Human Colon Crypt: Analysis via Methylation Patterns
|
Bacterial endosymbionts of insects play a central role in upgrading the diet of their hosts . In certain cases , such as aphids and tsetse flies , endosymbionts complement the metabolic capacity of hosts living on nutrient-deficient diets , while the bacteria harbored by omnivorous carpenter ants are involved in nitrogen recycling . In this study , we describe the genome sequence and inferred metabolism of Blattabacterium strain Bge , the primary Flavobacteria endosymbiont of the omnivorous German cockroach Blattella germanica . Through comparative genomics with other insect endosymbionts and free-living Flavobacteria we reveal that Blattabacterium strain Bge shares the same distribution of functional gene categories only with Blochmannia strains , the primary Gamma-Proteobacteria endosymbiont of carpenter ants . This is a remarkable example of evolutionary convergence during the symbiotic process , involving very distant phylogenetic bacterial taxa within hosts feeding on similar diets . Despite this similarity , different nitrogen economy strategies have emerged in each case . Both bacterial endosymbionts code for urease but display different metabolic functions: Blochmannia strains produce ammonia from dietary urea and then use it as a source of nitrogen , whereas Blattabacterium strain Bge codes for the complete urea cycle that , in combination with urease , produces ammonia as an end product . Not only does the cockroach endosymbiont play an essential role in nutrient supply to the host , but also in the catabolic use of amino acids and nitrogen excretion , as strongly suggested by the stoichiometric analysis of the inferred metabolic network . Here , we explain the metabolic reasons underlying the enigmatic return of cockroaches to the ancestral ammonotelic state .
In 1887 , Blochmann first described symbiotic bacteria in the fatty tissue of blattids [1] . Later , Buchner [2] suggested that symbionts are involved in the decomposition of metabolic end-products from the insect host . A classic example is the cockroach . Several pioneering studies correlated the presence of cockroach endosymbionts with the metabolism of sulfate and amino acids [3] , [4] . These endosymbionts were classified as a genus Blattabacterium [4] , belonging to the class Flavobacteria in the phylum Bacteroidetes [5] and they live in specialized cells in the host’s abdominal fat body . Apart from cockroaches , they were only found in the primitive termite Mastotermes darwiniensis [6] . Phylogenetic analyses for the Blattabacterium-cockroach symbiosis supported the hypothesis of co-evolution between symbionts and hosts dating back to an ancient feature of more than 140 million years ago [7] , [8] . Recently , genome sizes of the Blattabacterium symbionts of three cockroach species , B . germanica , Periplaneta americana , and Blatta orientalis were determined by pulsed field gel electrophoresis as approximately 650±15 kb [9] . Similarly , the authors demonstrated the sole presence of Blattabacterium strains in the fat body of those cockroach species by rRNA-targeting techniques . Phylogenetic analyses based on 16S rDNA also confirmed the affiliation of these endosymbionts to the class Flavobacteria [9] . Therefore , they are phylogenetically quite distinct from the majority of intensively studied insect endosymbionts that belong to the phylum Proteobacteria , mainly class Gamma-Proteobacteria . Recently , the highly reduced genome of “Candidatus Sulcia muelleri” ( from now S . muelleri ) , an insect endosymbiont belonging to the class Flavobacteria has been also completely sequenced [10] . Primary endosymbionts such as Buchnera aphidicola or Wigglesworthia glossinidia complement the metabolic capacity of aphids or tsetse flies , respectively that feed on different nutrient-deficient diets [11] . There are also examples of metabolic complementation between two co-primary endosymbionts and their hosts . This is the case of S . muelleri , living in the sharpshooter Homalodisca vitripennis , which coexists with another Gamma-Proteobacteria endosymbiont , “Candidatus Baumannia cicadellinicola” ( hereafter B . cicadellinicola ) . Both have developed a metabolic complementation to supply the host with the nutrients lacking in the limited xylem diet [12] . Another example is the case of B . aphidicola and “Candidatus Serratia symbiotica” , co-primary endosymbionts of the cedar aphid Cinara cedri that complement each other in the provision of essential nutrients [13] , [14] . Omnivorous insects also harbor endosymbionts . It is the case , for example , of ants of the genus Camponotus and their primary endosymbionts , the Gamma-Proteobacteria “Candidatus Blochmannia floridanus” [15] and “Candidatus Blochmannia pennsylvanicus” [16] ( from now B . floridanus and B . pennsylvanicus , respectively ) . In this association endosymbionts play an important role in nitrogen recycling [17] . Evolutionary convergences are generally considered as evidence of evolutionary adaptation . The study of endosymbiont evolution could provide examples of evolutionary convergences if we were able to show that very distant phylogenetic groups present similar functional repertoires and metabolic capabilities when they have evolved endosymbiosis in organisms having similar feeding behaviors . This may be the case of Blochmannia ( a gamma-proteobacterium ) and Blattabacterium ( a flavobacterium ) that have independently evolved in carpenter ants and cockroaches , two omnivorous insects . In this study , we determine the genome sequence of an endosymbiotic flavobacterium , Blattabacterium strain Bge , primary endosymbiont of the German cockroach B . germanica . We have also inferred the metabolism to try to understand why cockroaches excrete ammonia , instead of being uricotelic like other terrestrial invertebrates , thus breaking the so-called “Needham's rule” [18] , a question that has puzzled physiologists for a long time . Finally , we compare the inferred metabolism with the corresponding one of B . floridanus , the primary endosymbiont involved in nitrogen recycling in the carpenter ant Camponotus floridanus , an insect that has also a complex diet .
The general features of the genome of Blattabacterium strain Bge ( CP001487 ) and their comparison with those of other selected bacteria are shown in Table 1 . The size of the circular chromosome is 637 kb , and the G+C content is 27 . 1% . Only 23 . 4 kb are not-coding and they are distributed in 480 intergenic regions with an average length of 49 bp . The overall coding density ( 96 . 3% ) is the highest among insect endosymbionts known to date , indicating a highly compact genome . It is surprisingly higher than the most reduced insect endosymbiont “Candidatus Carsonella ruddii” ( 93 . 4% ) [19] . In addition , 1 . 5 kb correspond to 139 overlapping regions with an average length of 11 bp . Of these overlaps , 94 ( 67 . 6% ) are between genes on the same strand and 1 to 70 bp long . The other 45 cases ( 32 . 4% ) involve two genes on opposite strands and are between 2 and 50 bp long . Of these , only in one case the two genes overlap with their start regions , whereas in the rest the overlap is in the terminal region of the genes . On the other hand , in “Ca Carsonella ruddii” 92% of the 126 overlaps are in tandem orientation , and thus on the same strand , and only five cases are between opposite strands , involving the termini and starts of the overlapping genes . Assembly of the pyrosequencing data gave highly reliable contigs that combined with the data from Sanger sequencing resulting in a single contig , representing the entire genome . Probably due to the formation of a secondary structure , only a 33 bp stretch in an intergenic region upstream of the GroEL gene was not covered by pyrosequencing data but only by Sanger reads . Furthermore , annotation of the ORFs allowed a clear assignation of protein functions even in cases with only weak similarities with existing database entries . Not a single case of a possible host gene incorporated in the symbiont genome was found . Neither had we found coding sequences affiliated with Blattabacterium strain Bge outside the genome that could have been assigned to the host genome . A total of 627 putative genes have been assigned ( Figure S1 ) , 586 of which are protein coding genes ( CDS ) , 40 are RNA-specifying genes ( 34 tRNAs , 3 rRNAs located in a single operon , one tmRNA , and the RNA components of RNase P and the Signal recognition particle ) . The only pseudogene found corresponds to the protein component of RNase P . This gene coding for 118 amino acids is disrupted by an in-frame stop codon at amino acid position 53 . The RNase P proteins of the free-living F . psychrophilum [20] , Flavobacterium johnsoniae ( http://genome . jgi-psf . org/flajo/flajo . info . html ) and Gramella forsetii [21] contain a lysine residue at that position . Therefore , it is possible that the stop codon has been generated by an A–T point mutation in position 157 of the nucleotide sequence . Despite this mutation , the RNase P could be functional as it has been described that in vitro the RNA component can act enzymatically without a functional protein component [22] . Regarding the coding genes , it is interesting that , despite the compactness of the genome , there are eight gene duplicates: miaB , rodA , serC , lpdA , ppiC , argD , hemD , and uvrD . No specific sequence of the origin of replication ( oriC ) , such as dnaA boxes , was found in the genome [23] . Likewise dnaA , which codes for the protein that initiates replication by binding to such sequences , was also absent . Thus , the putative origin of replication was determined by GC skew analysis . The transitional region where the GC skew changes from negative to positive one ( Figure S2 ) showed the position of replication origin to be in the gene dapB . It is worth mentioning that neither dnaA nor any of the genes normally adjacent to the replication site in bacteria ( dnaN , hemE , gidA , hemE , and parA ) have been found in this genome . However , Blattabacterium strain Bge , has retained recA , which could trigger replication by an alternative mechanism [15] , [23] . We have inferred the metabolism of Blattabacterium strain Bge from its complete genome ( Figure 1 ) . Blattabacterium strain Bge possesses a limited capacity for nutrient uptake with only one ABC-type transport system , which may be specialized in fructose transport because this bacterium , contrary to the other sequenced endosymbionts , seems unable to use glucose as a nutrient . On the other hand , Blattabacterium strain Bge also codes for a glycerol uptake facilitator that enables transport of solutes , such as O2 , CO2 , NH3 , glycerol , urea , and water . Therefore , it is possible that Blattabacterium strain Bge obtains carbon from glycerol as a supplementary source . A sodium/drug antiporter , NorM , is also encoded by this genome . This system of efflux drug transport is common among enterobacteria but not among flavobacteria . In this group it is only known for the free-living bacteria F . psychrophilum and G . forsetii . This system can act as a multidrug transport as well as transporting oligosaccharidyl lipids and polysaccharide compounds . There is an array of metal ion homeostasis transporters . In Blattabacterium strain Bge , there is a Trk transport system , a uniporter of the monovalent potassium cation , which requires a proton motive force and ATP in order to function . Only W . glossinidia has a similar transport system , although the encoded subunits differ: trkA and trkB in Blattabacterium; trkA and trkH in W . glossinidia . Other solutes are also transported by symport systems . Blattabacterium strain Bge is able to uptake glutamate and aspartate via a proton symporter . Both metabolites play an important role in the metabolism of this bacterium ( see below ) . A phosphate/sodium symporter is also present . Regarding electron transport , the encoded NADH-dehydrogenase ( ndh ) oxidizes NADH without proton translocation . There is also a succinate dehydrogenase ( sdhABD ) . Electrons are transferred to a membrane-bound menaquinone ( MQ ) and a molybdenum-oxidoreductase , which accepts electrons from the MQ . With these elements , a proton motive force can be generated . Blattabacterium strain Bge seems to be able to reduce intracellular sulfate to sulfite . A number of genes required for sulfur assimilation present in the genome , include those encoding for the two subunits of the sulfate adenylyltransferase , cysN and cysD , the adenosine phosphosulfate ( APS ) reductase cysH and the sulfite reductase proteins cysI , J . There is a missing step for the conversion of adenosine-5′-phosphosulfate ( APS ) into 3′-phospho adenosine-5′-phosphosulfate ( PAPS ) . The generated sulfite is reduced to sulfide further on and assimilated into the sulfur-containing amino acids L-cysteine and L-methionine . Blattabacterium strain Bge is able to synthesize its own cell wall and plasma membrane . However , it has lost the entire pathway required for lipopolysacharide ( LPS ) biosynthesis , like all sequenced Buchnera strains and B . cicadenillicola . This property explains why Blattabacterium strain Bge , similarly to these bacteria , are surrounded by a host vacuolar membrane , as shown in the electron-microscopy images ( Figure S3 ) . Regarding amino acid biosynthesis , Blattabacterium strain Bge has the genes encoding biosynthetic enzymes needed to synthesize 10 essential ( His , Trp , Phe , Leu , Ile , Val , Lys , Thr , Arg , and Met ) and 7 nonessential ( Gly , Tyr , Cys , Ser , Glu , Asp , and Ala ) amino acids . Thus , the endosymbiont metabolism relies on Pro , Gln and Asn supplied by the host . Also present is the complete machinery to synthesize nucleotides , fatty acids , and the cofactors folic acid , lipoic acid , FAD , NAD , pyridoxine , and riboflavin . Finally , genes encoding enzymes for the synthesis of siroheme and menaquinone were also identified . With respect to the metabolism of carbohydrates , genome analysis of Blattabacterium strain Bge indicates the presence of a truncated glycolysis pathway , since the genes that encode for phosphofructokinase ( pfkA ) and pyruvate kinase ( pyk ) are missing , as well as any sugar phosphorylating system except for fructose . Therefore , the pathway begins with fructose-1 phosphate and continues with the canonical enzymatic steps until the synthesis of phosphoenolpyruvate ( PEP ) . Given the lack of pyruvate kinase genes , Blattabacterium strain Bge must produce pyruvate via the malic enzyme ( NADP+-dependent malate dehydrogenase ) . Additionally , a complete non-oxidative pentose phosphate pathway is encoded in Blattabacterium strain Bge . As it is the case with Wigglesworthia , the glycolytic enzymes seem to be involved in gluconeogenesis rather than glycolysis complementing the non-oxidative pentose phosphate pathway [24] . In summary , although Blattabacterium strain Bge genome shows a strong reduction in gene number in all the functional categories , compared to their free-living relatives ( see below ) , the core of essential functions and pathways is particularly well preserved . The protein genes of Blattabacterium strain Bge were classified according to COG categories ( Figure 2 , Table 2 ) . This distribution was compared with those of twelve selected bacteria: four Flavobacteria , which included three free-living species ( F . psychrophilum , F . johnsoniae and G . forsetii ) and the endosymbiont S . muelleri , and eight Proteobacteria endosymbionts , seven Gamma-Proteobacteria ( B . floridanus , B . pennsylvanicus , B . cicadellinicola , B . aphidicola Aps , B . aphidicola Cce , S . glossinidius , and W . glossinidia ) and one Alfa-Proteobacterium ( Wolbachia sp . from Drosophila simulans ) . Taking the observed distribution of COG categories for Blattabacterium strain Bge as the expected distribution followed by each of the other bacteria examined , the hypothesis of equal distribution was rejected in all but the carpenter ant endosymbionts , Gamma-Proteobacteria B . floridanus and B . pennsylvanicus ( Table 2 ) . These results suggest that it is the hosts’ diet ( cockroaches and carpenter ants are both omnivores ) rather than phylogenetic closeness which is more strongly linked with the type of genes retained . This appears to be a clear case of functional evolutionary convergence in a broad sense . The proximity between the endosymbionts from omnivorous hosts was also confirmed when a dendrogram was created using the matrix of Kulczynski phenetic distances ( Figure 3A ) . To locate the phylogenetic position of Blattabacterium strain Bge and compare it with the COG-based functional analysis , we used a phylogenetic tree based on 16S rDNA gene sequences ( Figure 3B ) . As expected , the 16S rDNA gene analysis clearly separate Bacteroidetes from Proteobacteria phyla . Blattabacterium strain Bge clusters monophyletically within the Bacteroidetes phylum . The functional clustering differs clearly from the phylogenetic one . A striking trait of this genome is the presence of a complete urea cycle ( Figure 4 ) . This feature has been described in few bacteria , and in only one member of the Bacteroidetes phylum , the cellulolytic soil bacterium Cytophaga hutchinsonii [25] . Moreover , to date , there are no reports of a complete urea cycle in an endosymbiont . The Blattabacterium strain Bge genome also retains the genes for the catalytic core of urease and we have detected urease activity in endosymbiont-enriched extracts of cockroach fat body ( see below ) . The genome of Blattabacterium strain Bge has two urease genes , ureAB and ureC , coding for the catalytic subunits , but lacks all genes for the accessory proteins supposedly required to produce an active enzyme in most bacteria . The ureAB fusion is not a novel situation since fused urease genes have also been described in other bacterial genomes , as it is the case of the free-living Flavobacterium C . hutchinsonii [25] . Regarding the lack of accessory genes , a similar situation is found in Bacillus subtilis cells expressing urease activity , which are able to grow with urea as sole nitrogen source [26] . To corroborate the presence of an active urease in Blattabacterium strain Bge , we performed an enzymatic assay on crude extracts of the endosymbiont-enriched fraction of the B . germanica fat body . Figure S4 shows a representative result for the urease assay . Although the detected specific activity under our experimental conditions was low ( 2 mU mg−1 protein; 1 U of urease corresponds to the formation of 1 µmol of ammonia per min ) , it was reproducible . Urease activity was also reproducibly detected in endosymbiont extracts from P . americana fat body ( data not shown ) . To further study the inferred metabolism in relation to nitrogen economy , we carried out a stoichiometric analysis of the reactions involved in the Krebs and urea cycles as well as other directly related reactions , such as urease , the malic enzyme , and their links to amino acid utilization ( Figure 1 and Figure 4 ) . Our results strongly suggest a key involvement of the endosymbionts in nitrogen metabolism and excretion in the German cockroach , in addition to their role in providing essential amino acids and coenzymes to the host . It is also worth mentioning that the endosymbiont metabolism relies on a supply of Gln from the host to cater for all its biosynthetic needs , including the urea cycle . Stoichiometric analysis shows that eleven out of fourteen elementary modes produce ammonia ( Table S1 ) . It follows that the metabolic network of Blattabacterium strain Bge could potentially use amino acids efficiently as energy and reducing-power sources , generating nitrogen waste in the form of ammonia ( Figure 4 ) . Urease genes are also present in the Blochmannia endosymbiont genome [15] and the biochemical function of the urease in the carpenter ant endosymbionts is completely different from Blattabacterium . Studies of gene expression [27] and feeding experiments with 15N-labelled urea [17] in carpenter ants corroborate the role of urease in the transfer of nitrogen from dietary urea into the hemolymph amino acid pool . This requires an endosymbiont glutamine synthase to act as an essential step in nitrogen conservation during amino acid anabolism . Thus , although carpenter ants are omnivorous , their bacterial endosymbionts may upgrade their diet via an efficient nitrogen economy [17] . German cockroaches are also omnivorous; however , their endosymbionts lack genes encoding a glutamine synthase-like activity , a clear indication that the metabolic function of urease is not the same in the German cockroach and carpenter ant endosymbionts because generated ammonia cannot be re-assimilated . Therefore , although we have revealed a functional convergence between the cockroach and carpenter ant endosymbionts , which is probably due to their hosts’ omnivorous diets , they differ greatly from a metabolic viewpoint in detail , particularly in terms of nitrogen metabolism . Traditionally , Blattabacterium endosymbionts have been postulated to be involved in the metabolism of uric acid in cockroaches . For instance , uric acid accumulation has been observed in aposymbiotic cockroaches [28] , [29] . Metabolic use of nitrogen derived from fat body urates has been observed in B . germanica under certain conditions ( e . g . , in females on low-protein diet [30] and consumption of empty spermatophores by starved females [31] ) . Interestingly , fat body endosymbionts have been involved in uric acid degradation to CO2 in experiments with the wood cockroach Parcoblatta fulvescens injected with 14C-hypoxanthine [32] . Although involvement of gut microbiota cannot be completely ruled out , endosymbiont metabolism seemed more likely [33] . However , our results show that the endosymbiont genome does not code for any activity related to either the synthesis or the catabolism of urates . Therefore , and contrary to early reports based on putative cultured endosymbiotic bacteria [29] , Blattabacterium strain Bge cannot participate in the metabolism of this nitrogen compound directly . Since uricase activity has been detected in the fat body of the cockroach [28] , [34] , [35] , the host could contribute with uric-derived metabolites to the nitrogen economy of the endosymbiont which , in turn , would produce ammonia and carbon dioxide as final catabolic products . The genome sequencing , metabolic inference , detection of a urease in the endosymbiont and the stoichiometric analysis of the central pathways of Blattabacterium strain Bge shed light on a whole series of hitherto unexplained classical physiological studies on ammonotelism in cockroaches [33] , [36] , [37] . Contrary to the speculation that some terrestrial invertebrates , like gastropods , annelids [36] and isopods [38] , exploit ammonia excretion as “a return to the cheapest way” [38] to eliminate nitrogen , the case of the German cockroach and its bacterial endosymbionts indicates that this might not be the case . The evolution of terrestrial-living metazoa has favored the emergence of uricotely ( e . g . the majority of insects ) and ureotely ( e . g . mammals ) as water-saving strategies . Meanwhile , ammonotely , the ancestral character present in aquatic animals , has classically been considered maladaptive for terrestrial animals [18] . Symbiosis seems to play a role in this “return” of cockroaches to ammonotely by providing new enzymes required for this new nitrogen metabolism . Thus the metabolic capabilities acquired by symbiogenesis [39] afford to explore new ecological niches and dietary regimes .
B . germanica ( Blattaria: Blattellidae ) was reared in the Entomology laboratory ( Cavanilles Institute for Biodiversity and Evolutionary Biology , University of Valencia ) . The cockroaches were kept in the laboratory at 25°C and fed with a mixture of dog food ( 2/3 ) and sucrose ( 1/3 ) . The bacterial endosymbionts were extracted from the fat body of B . germanica females . To do so , cockroaches were killed by a 15 to 20 min treatment with ethyl acetate and the bacterial cells were separated from the fat body as in [15] . An enriched fraction of bacteriocytes is then obtained that is used to extract total DNA following a CTAB ( Cetyltrimethylammonium bromide ) method . The complete genome sequence of Blattabacterium strain Bge was obtained by a hybrid sequencing approach based on ABI 3730 sequencers and the pyrosequencing system ( 454; Life Science ) . To construct shotgun libraries , DNA fragments were generated by random mechanical shearing with a sonicator and posterior separation in a pulsed field gel electrophoresis . Insert sizes of 1–2 kb and 3–5 kb were purified and cloned into vector from XL-TOPO PCR cloning kit . Plasmid DNA was extracted using 96-well plates ( Millipore ) with the PerkinElmer MULTIPROBE II robot according to the manufacturers . DNA sequencing was performed on an ABI PRISM 3730 Genetic Analyzer ( Applied Biosystems ) . In the initial random sequencing phase 9 , 227 sequences were obtained with 1 . 5-fold sequence coverage . Given the lack of joining between sequences , which may have been due to a large number of sequences from the host , a strict sequence analysis was performed with a specific bioinformatic tool called a Categorizer . It carries out a sequence classification method based on n-mers composition to correctly distinguish between Blattabacterium strain Bge and contaminating host sequences . This classifier was trained with sets of sequences identified from Blattabacterium strain Bge and the host . With these sets , we constructed a feature vector or model representing the 4- to 7-mers usage pattern of each organism . Then the n-mers composition of each read was compared with these generated models with a k-nearest neighbor clustering algorithm ( KNN ) . Although the number of retrieved host sequence reads was higher than the one of Blattabacterium strain Bge sequences for both sequencing approaches , the pyrosequencing approach generated enough sequences to close the gaps identified with the first method . The tool Gap4 from Staden Package [40] was used for the total assembly . Fat body of B . germanica was isolated and prefixed in a 2 . 5% paraglutaraldehyde fixative mixture buffered with 0 . 1 M phosphate at pH 7 . 2 ( PB ) . Prefixation was performed at 4°C for 24 h and then rinsed several times in PB . To avoid the loss of this dispersed tissue , the fat body was placed in agar ( 2% ) forming small blocks . After prefixation , these blocks were fixed in 2% osmium tetroxide for one hour , dehydrated in graded alcohol and propylene oxide , stained in a saturated uranyl acetate solution 2% and embedded in araldite to form the definitive blocks . Thin sections ( 0 . 05 µm ) were made using the Reichert-Jung ULTRACUT E ( Leica ) ultramicrotome , and then were stained with uranyl acetate and lead citrate . A JEOL-JEM 1010 electron microscope was used for the analysis . The putative coding regions ( CDSs ) in the Blattabacterium strain Bge genome were identified with the GLIMMER3 program [41] . This program was first trained with closely related organism sequences from the Flavobacteria group . The coding sequence model obtained was then used by GLIMMER3 to scan the genome to predict potential coding regions by considering the putative existence of initiation codons and ORF length . Start and stop codons of each putative CDS were curated manually through visual inspection of the Blattabacterium strain Bge Genome Browser , a database specially designed for this symbiont . The putative coding proteins were initially analyzed by reciprocal best hits to determine orthology between genes of the Blattabacterium and those from bacteria belonging to the Flavobacteria group . According to these criteria , two genes are orthologs when a gene in one genome matches as the best hit with a gene in the other genome . Sequences that could not be assigned to any function in comparison with flavobacterial genomes were identified by searching a non-redundant protein database using BLASTX [42] . Final annotation was performed using BLASTP comparison with proteins in the NCBI and Pfam domains identified using the Sanger Centre Pfam search website . Non-coding RNAs were identified by different approaches . The tRNAscan program was used to predict tRNAs , as well as other small RNAs , like tmRNA , the RNA component of the RNase P . Signal Recognition Particle RNA were identified by programs like ARAGORN , BRUCE and SRPscan , as well as consulting the Rfam database [43]–[45] . In the absence of a diagnostic cluster of DnaA boxes , the origin of replication was identified by GC-skew calculated as ( C−G ) / ( C+G ) using the program OriginX [46] . The origin is located in the transitional region where the GC-skew changes from negative to positive values . The ORFs orthologous to known genes in other species were catalogued based on non-redundant classification schemes , such as COG ( Clusters of Orthologous Groups of Proteins ) . A metabolic network was reconstructed using the automatic annotator server from KAAS-KEEG [47] . According to our genome annotation , each pathway was examined checking the BRENDA [48] and EcoCyc databases [49] . Comparison between the COGs distribution of each species with that of the Blattabacterium strain Bge was carried using chi-square tests . To avoid the problem of multiple testing , we applied the Bonferroni correction so that for each individual test the significance level was 0 . 05/12 = 0 . 0042 . That is , if the p-value is lower than 0 . 0042 then the hypothesis is rejected . The first p-value corresponds to the standard chi-square test ( Chi2 p-value , df = 19 ) . Due to the asymptotic nature of this test , expected frequencies should be higher than 5 . However , we might expect some frequencies with low values . To correct this situation we also performed a Monte-Carlo version of this test ( MC p-value ) . We performed 19 , 999 simulations under the null hypothesis , which together with the observed Chi2 statistics constituted a set of 20 , 000 values . The MC p-value cannot be lower than 1/20 , 000 = 5 . 00E-5 . The Kulczynski distance between species 1 and 2 is given by 1−0 . 5 ( Σjmin ( y1j , y2j ) /Σjy1j + Σjmin ( y1j , y2j ) /Σjy2j ) where j ( from 1 to 20 ) refers to the corresponding normalized COG categories ( from 0 to 1 ) . The dendrogram was derived from the corresponding distance matrix by applying a complete clustering method in which the distance between clusters A and B is given by the highest distance between any two species belonging to A and B , respectively . The statistical significance of the clusters of the dendrogram was evaluated by bootstrap analysis based on 100 , 000 replicates . The sequences of 16S rDNA were aligned with MAFFT ( v6 . 240 ) [50] program . The positions for the phylogenetic analysis were derived by Gblocks v0 . 91b [51] . In total , 1530 nucleotides were selected . The phylogenetic reconstruction was carried out by maximum likelihood using the PHYML program [52] . The best evolutionary model chosen by MODELTEST [53] was a GTR + Gamma ( G ) + I ( Proportion invariant ) . Bootstrap values were based on 1000 replicates . Abdominal fat bodies from dissected B . germanica adult females were homogenized with a Douce homogenizer adding a 50 mM HEPES buffer containing 1 mM EDTA , pH 7 . 5 . The crude extract was centrifuged for 25 min at 6000 rpm at 4°C , and the pellet was resuspended with the homogenization buffer . The supernatant and a crude extract of cockroach heads ( host tissue without endosymbionts ) were used in control experiments . The resuspended pellet or bacteria-enriched fraction was treated with lysozyme ( 3 . 5 U mL−1 ) for 30 min at 4°C and sonicated for 5 sec . Urease activity was determined incubating the extract at 37°C with 110 mM urea . At different time intervals the reaction was stopped by adding 1 vol . 10% trichloroacetic acid and the produced ammonia was measured by the colorimetric Berthelot method [54] as described in [55] . The protein content was measured with a Nanodrop ND1000 equipment . Stoichiometric analysis ( using METATOOL ) [56] was performed on the central pathways directly involved in amino acid catabolism , including the Krebs and urea cycles . Information about the reversibility of reactions was checked in the BRENDA database [48] . The input file for METATOOL is available upon request to the corresponding author . The genome was sent to GenBank and has been assigned accession number CP001487 .
|
Bacterial endosymbionts from insects are subjected to a process of genome reduction from the moment they interact with their host , especially when the symbiosis is strict ( the partners live together permanently ) and the endosymbiont is maternally inherited . The type of genes that are retained correlates with specific metabolic host requirements . Here , we report the genome sequence of Blattabacterium strain Bge , the primary endosymbiont of the German cockroach B . germanica . Cockroaches are omnivorous insects and Blattabacterium cooperates with their metabolism , not only with essential nutrient metabolism but also through an efficient use of amino acids and the nitrogen excretion by the combination of a urea cycle and urease activity . The repertoires of functions that are maintained in Blattabacterium are similar to those already observed in Blochmannia spp . , the primary endosymbiont of carpenter ants , also an omnivorous insect . This constitutes a nice example of evolutionary convergence of two endosymbionts belonging to very different bacterial phyla that have evolved a similar repertoire of functions according to the host . However , the current set of genes and , more importantly , those that were lost in the process of genome reduction in both endosymbiont lineages have also contributed to a different involvement of Blattabacterium and Blochmannia in nitrogen metabolism .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/genome",
"projects",
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics"
] |
2009
|
Evolutionary Convergence and Nitrogen Metabolism in Blattabacterium strain Bge, Primary Endosymbiont of the Cockroach Blattella germanica
|
The FANTOM5 consortium utilised cap analysis of gene expression ( CAGE ) to provide an unprecedented insight into transcriptional regulation in human cells and tissues . In the current study , we have used CAGE-based transcriptional profiling on an extended dense time course of the response of human monocyte-derived macrophages grown in macrophage colony-stimulating factor ( CSF1 ) to bacterial lipopolysaccharide ( LPS ) . We propose that this system provides a model for the differentiation and adaptation of monocytes entering the intestinal lamina propria . The response to LPS is shown to be a cascade of successive waves of transient gene expression extending over at least 48 hours , with hundreds of positive and negative regulatory loops . Promoter analysis using motif activity response analysis ( MARA ) identified some of the transcription factors likely to be responsible for the temporal profile of transcriptional activation . Each LPS-inducible locus was associated with multiple inducible enhancers , and in each case , transient eRNA transcription at multiple sites detected by CAGE preceded the appearance of promoter-associated transcripts . LPS-inducible long non-coding RNAs were commonly associated with clusters of inducible enhancers . We used these data to re-examine the hundreds of loci associated with susceptibility to inflammatory bowel disease ( IBD ) in genome-wide association studies . Loci associated with IBD were strongly and specifically ( relative to rheumatoid arthritis and unrelated traits ) enriched for promoters that were regulated in monocyte differentiation or activation . Amongst previously-identified IBD susceptibility loci , the vast majority contained at least one promoter that was regulated in CSF1-dependent monocyte-macrophage transitions and/or in response to LPS . On this basis , we concluded that IBD loci are strongly-enriched for monocyte-specific genes , and identified at least 134 additional candidate genes associated with IBD susceptibility from reanalysis of published GWA studies . We propose that dysregulation of monocyte adaptation to the environment of the gastrointestinal mucosa is the key process leading to inflammatory bowel disease .
Inflammatory bowel disease ( IBD ) comprises a group of complex syndromes that arise from a dysfunctional interaction between the microbiota of the intestinal lumen and the immune system [1] . Loci associated with the heritability of susceptibility to IBD are shared in some measure with other chronic inflammatory diseases [2 , 3] . Extensive genome-wide association studies ( GWAS ) have identified more than 200 risk loci for IBD , with significant overlaps between the two major forms , Crohn’s disease and ulcerative colitis [4–6] . The analysis of candidate genes within susceptibility loci based upon apparently shared biological function has emphasised three major pathways: the activation of Th17 T cells , autophagy , and the response to mycobacteria [7 , 8] . The emphasis on T cell activation derives in part from a clear association of susceptibility with certain MHC haplotypes , and the observed activation of T cells in inflamed mucosa [9] . An alternative view is that IBD is primarily initiated by functional dysregulation in cells of the macrophage lineage [10 , 11] and prioritisation of candidates on that basis may be informative [12] . By contrast to other tissue macrophage populations which have a significant capacity for self-renewal , lamina propria macrophages of the gut are renewed continuously from the circulating monocyte pool [13 , 14] . The proliferation and differentiation of the monocyte-macrophage lineage is controlled by the growth factor , macrophage colony-stimulating factor ( CSF1 ) , which signals through a tyrosine kinase receptor , CSF1R [15–17] . The replenishment of the resident cells of the lamina propria requires the continuous exposure to CSF1 and lamina propria macrophages are rapidly depleted in mice treated with a blocking anti-CSF1R antibody [18 , 19] . Human macrophages are commonly generated by cultivation of isolated CD14+ monocytes for 4–5 days in CSF1 [20 , 21] . Several groups have published detailed transcriptomic analysis of the differentiation of monocytes in CSF1 [20–23] . Based upon the known biology ( derivation from monocytes , dependence upon CSF1 ) , we suggest that the monocyte-derived macrophage actually approximates an in vitro model of the differentiation of intestinal macrophages from incoming monocytes . Alongside CSF1 , monocytes entering the lamina propria of the lower GI tract are immediately exposed to microbial products of which the archetype is lipopolysaccharide ( LPS ) from gram-negative bacteria . Monocytes must rapidly down-modulate their response to bacteria in the lumen , to avoid initiating an inflammatory response in the gut wall [18] . In response to LPS , macrophages initiate a complex feed-forward and feed-back cascade of induction and repression of transcription factors and autocrine regulators [24 , 25] leading to a new steady state . The response of mouse macrophages to LPS has been studied in detail at many levels from the mechanisms of signaling [26 , 27] through transcriptional networks [28–30] to the underlying alterations in chromatin structure [31–33] . Dense time course data are required to enable inference of the sequence of transcriptional events in this response . The FANTOM consortium established tag sequencing of genome-scale 5’-RACE ( CAGE ) as an expression profiling tool and used the approach to create a comprehensive human promoter-based expression atlas [34] . Deep sequencing of CAGE libraries also detected the bidirectional transcripts ( eRNAs ) derived from active enhancers enabling genome-wide quantification of enhancer activity [35] . CAGE-based delineation of transcription start sites was used in the comparison of the promoters of LPS-inducible genes in mouse and human macrophages [21] as well as a recent detailed analyses of monocyte subsets [36] . Extensive analysis of cells undergoing state change showed a clear temporal pattern in induction of expression; enhancers are expressed first , then transcription factor genes , and finally the genes they regulate [37] . In the current study , we analysed CAGE data generated within FANTOM5 to dissect the transcriptional changes that occur in human macrophages grown in CSF1 , and their subsequent response to LPS , as a model for events that occur when monocytes adapt to the gut environment . We use these data , combined with extensive data on the gene expression of blood monocytes generated by the FANTOM5 consortium , to reassess the known candidate intervals associated with IBD . Our analysis supports the hypothesis that IBD is primarily a macrophage-initiated pathology and provides the basis for identification of alternative candidate genes within many IBD susceptibility loci that have been identified by genome-wide analysis ( GWA ) .
The aim of this study was to examine the hypothesis that dysregulation of intestinal macrophage responses to bacterial antigens is an important component of susceptibility to IBD . Initially , as part of the FANTOM5 project , we produced a detailed quantitative analysis of promoter utilization and gene expression across a dense time course of the LPS response of human monocyte-derived macrophages ( MDM ) from three separate donors . The time course focused on very early events ( at 15 minute intervals from initiation of the response ) , as well as the later change in differentiation state up to 48 hours after stimulation . For comparison , the FANTOM5 dataset contains three additional unstimulated MDM samples from different donors , obtained commercially , three separate CD14+ monocyte populations , each in triplicate , and cultured monocytes stimulated for 2 hours with a range of agonists including gamma interferon ( IFNγ ) LPS and live salmonella . The monocytes differed in their method of isolation . One set , also purchased commercially , was clearly “activated” and expressed many inflammatory cytokines . Another set was divided into the three known monocyte sub-populations based upon relative expression of CD14 and CD16 , as described in detail elsewhere [36 , 38] . The entire FANTOM5 dataset , including an extensive mouse CAGE-based transcriptomic dataset is available on the ZENBU browser ( see below ) . All references in the text to the level of expression , or the temporal expression profiles of individual genes within the text can be confirmed by accessing this browser and entering the gene name . Previous studies of mouse macrophage response to LPS revealed sequential induction and repression of numerous transcription factors [30] . To gain an overview of the transcription regulatory cascade of human macrophage response to LPS , we first utilized the data visualisation tool Biolayout Express3D [39] to generate a pairwise correlation matrix based upon aggregated expression of the promoter activity of annotated transcription factors identified previously [34] . The sample-to-sample graph in Fig 1 , based solely upon the set of transcription factors , clearly shows that the response to LPS can be visualized as a progressive and profound change in transcription factor milieu . Gene-to-gene analysis showed that this involves at least 200 distinct transcription factors . Fig 1 shows the averaged expression profiles of the transcription factor genes within selected clusters , identified by the gene-to-gene analysis , emphasizing that the transcription factor genes can be classified based upon the peak time of induction and whether or not induction was sustained . The detailed list of the transcription factor genes within all of the co-regulated clusters is provided in S1 Table . Many of these transcription factors are discussed below . CAGE is essentially genome-scale 5’ RACE . Unlike microarrays , CAGE tag sequencing resolves the separate and independent utilization of alternative promoters for the same gene , and is significantly more sensitive across a large dynamic range [34] . For example , the CAGE data reveal that SERPINA1 has at least 4 promoters . Macrophages and other myeloid cells profiles by the FANTOM5 consortium selectively utilize individual promoters p2 , p3 and p4 of the SERPINA1 gene , whereas p1 is liver-specific . Because it sequences only the 5’ ends , CAGE does not require normalisation for the length of the transcript , which is necessary for RNA sequencing ( RNAseq ) data , and is considerably more cost-effective than RNAseq . Extensive validation of the CAGE methodology has been published elsewhere , showing the strong correlation of CAGE tag peaks with tissue-specific chromatin marks and DNase hypersensitive sites , and the precise colocation of the 5’ ends of the CAGE tags with the 5’ ends of full length transcripts and promoter-associated motifs such as the TATA box [34–36] We created a pairwise correlation network of individual promoters ( defined as cluster of transcription start sites , or CTSS ) of all expressed genes ( Fig 2 ) . Because most genes have several alternative promoters , the same gene can appear in multiple coexpression clusters . The gene-to-gene ( or more correctly , promoter-to-promoter ) clusters , with graphical displays of their average expression profiles can be accessed at http://coexpression . roslin . ed . ac . uk/lps , and the gene names within selected clusters discussed in the text are also provided in S2 Table , together with a summary of the pattern of expression . In keeping with the regulation of transcription factors in Fig 1 , the analysis revealed clusters of coexpressed promoters that have peak activity at different times during the time course . The complex pattern is shown in a network graph produced with the software Biolayout Express3D in S1 Fig . The network graph groups sets of transcripts with related temporal profiles . The positions of clusters on the graph progressively diverge from each other as sets of transcripts are induced or repressed compared to the initial state . The clusters characterized by either early or late expression profiles are the most similar to the unstimulated state . So , the network graph forms a circle . The data are presented in more conventional heat map form in Fig 2 . The heat map highlights the cascade of transient regulation of gene expression with time in response to LPS , and the fact that no new steady state is reached even after 48 hours . Within the first 15–30 minutes , there was a set of genes that was rapidly induced and another set that was repressed . Thereafter , there were waves of transient induction of thousands of promoters , each of which peaks at different time points , ranging from 30–45 mins until 36–48 hours . In addition to those clusters that were induced transiently , there were others that were induced in a sustained manner , so that at the 48 hour time point , the transcriptional state of the stimulated cells was still changing . These coexpressed promoter clusters have been numbered in order based upon the number of annotated genes they contain . The largest clusters , are slowly-repressed ( Cluster 0 ) or slowly-induced ( Clusters 1 , 2 , 3 , 4 ) peaking at different times late in the time course . Cluster 0 contains many known cell cycle genes , and presumably reflects in part the cessation of cell division in a subset of cells growing in CSF1 . Clusters 1–4 are all enriched for innate immune genes . Together , these slow-response clusters paint a picture of the progressive differentiation of the cells , which parallels the progressive change in transcription factor milieu shown in Fig 1 . Immediate early genes are a set of transcriptional regulators induced in many different cell types as they transition from one cellular state to another . Cluster 5 ( S1 Fig , Fig 2 , S2 Table ) contains promoters with detectable increased expression within 15 mins , peaking at 45 mins , and declining rapidly thereafter . This cluster of the earliest inducible promoters contains genes encoding many immediate early transcription factors ( EGR1 , EGR2 , EGR3 , FOS , FOSB , JUN ) [37] . One of the most inducible , and the most highly-expressed , transcriptional regulators in the LPS response is NFKBIZ , which encodes IκB-ζ . NFKBIZ has two promoters , both of which were LPS-inducible . Induction of IκB-ζ is likely to be critical for subsequent events in the transcriptional cascade [40] . In mouse macrophages , it interacts with AKIRIN2 , to bridge NFκB and the chromatin remodeling SWI/SNF complex on proinflammatory genes including IL6 and IL12B in macrophages [41] . The early-inducible transcription factors included genes encoding factors more commonly associated with lymphocyte functions; NFIL3 , IKZF1 , KLF2 and PRDM1 . Two smaller clusters , 30 and 42 ( S2 Table ) , were induced marginally later , peaking around 1hr 20 minutes . This group contained IFNB1 , a target of NFκB signaling which is known to act in an autocrine manner to induce the interferon target genes ( see below ) . Other inducible feed-forward activators include the gene encoding the adaptor protein TRAF1 , which interacts directly with signaling molecule MAP3K5 ( also known as ASK1 ) which in turn is involved in LPS signaling [42] and the gene encoding EGR4 , which along with EGR3 , can interact with NFκB and probably contribute to activation of NFκB target genes [43] . Genes for the well-known inflammatory cytokines ( IL1B , TNF , IL6 , IL10 ) and chemokines that are the hallmark of the response to LPS form parts of Clusters 6 and 7 ( S2 Table ) . A previously unreported member of this group was TNFSF9 , or CD137L , which was massively-induced by LPS and has been proposed as a growth and differentiation factor for myeloid cells [44] . Expression of the proinflammatory genes differed from immediate early response genes in terms of the duration of the expression . For example , TNF mRNA detected by CAGE was induced 10-fold within 15 mins , but was still somewhat elevated after 5–6 hours . The small MAF transcription factor gene , MAFF , was strongly induced alongside the inflammatory cytokines . The other members of this family ( MAFG , MAFK ) were also induced by LPS , but later in the time course . The small MAF transcription factors form heterodimers with many other transcription factors , notably the NFE2 family ( NFE2 , NFE2L1 , NFE2L2 , NFE2L3 genes ) [45] all of which are induced by LPS later in the time course . Hence , induction of MAFF is likely to be part of a feed forward cascade . Interferons ( IFNs ) and their targets are a key part of innate immune defence mechanisms , and chronic over-expression of IFN target genes is a feature of many inflammatory diseases [46–49] . The immediate early induction of IFNB1 in response to LPS initiates an autocrine loop that leads in turn to downstream induction of the targets of IFN signaling [27 , 33 , 47 , 50] . Many known IFN targets were induced by LPS and formed parts of coexpressed clusters , but distinct clusters containing these genes followed distinct temporal profiles . Based upon a search of the Interferome database [51] , a total of 123 genes within Clusters 2 and 3 ( the large late-response clusters noted above ) have previously been shown to respond directly to IFN signaling in human monocytes or monocyte-derived macrophages . [51] . Several smaller clusters of known IFN-response genes were induced significantly earlier . One , Cluster 10 ( S2 Table ) , peaked at 2–3 hours , and included the IFIT2 , 3 and 5 genes that derive from a single genomic region as well as genes for known virus/IFN-induced chemokines , CCL3 and 4 and CXCL10 [52] . Other clusters showing an early peak in expression contained genes for the known inducible feedback regulators of IFN-induced JAK-STAT signaling , SOCS1 and SOCS3 [53 , 54] . Cluster 11 peaked around 1 hour later than Cluster 10 , and Cluster 17 peaked slightly later again , and declined more slowly , similar to Clusters 33 and 49 . [53 , 54] . The diversity of responses of IFN-inducible genes probably reflects the complex regulation of the transcription factors that bind to interferon-response elements ( IRFs ) . The canonical MyD88-independent , TRIF/TRAM ( TICAM1/2 ) -dependent pathway of IFN regulation determined from studies of mice involves the interactions of TRAF3 and the kinase TBK1 to phosphorylate IRF3 [55] . However , this pathway is not conserved in humans [21] . In the human MDM , IRF3 mRNA was almost undetectable , and both TRAF3 and TBK1 were very low . Genes encoding the adaptor molecule TICAM1 , and the downstream RIK kinases ( RIPK1 , RIPK2 ) were induced by LPS , peaking after 2–3 hours with Cluster 13 . The most-rapidly-induced IRF family member in the LPS-stimulated human MDM was IRF1 . This observation is consistent with recent direct evidence of LPS-induced IRF1 binding to promoters and enhancers of inflammatory cytokine genes ( TNF , IL6 , IL12B ) in human MDM , and its role in the priming response to IFNγ [56] . IRF1 mRNA was induced detectably within 15 mins , peaked around 2 hours and was sustained at greatly elevated levels for the remainder of the time course . IRF4 was part of cluster 17 , maximally induced at 3–4 hours and declining thereafter . IRF7 was induced marginally later and remained elevated throughout , IRF8 was induced transiently , peaking at 3 hours and gone by 12 hours , IRF2 and IRF9 were induced in late response clusters . Other components of the IFN signaling pathway , notably STAT1 , STAT2 and STAT4 , and JAK2 and JAK3 were also induced transiently . It is likely that each of these factors contributes specifically to transcription regulation of the clusters in which they are coexpressed possibly by interacting with other induced transcription factors . For example , IRF8 is known to bind uniquely to a specific subset of IRF recognition sequences [29] . Members of the FANTOM Consortium have produced a cellular interactome linking coexpressed ligands and receptors [57] . The progressive response of MDM to LPS emphasized in Figs 1 and 2 and S1 and S2 Tables is based upon part upon numerous autocrine loops in which induction of ligands is followed by induction of the receptors . The most obvious is the response to endogenous IFNB1; the genes for receptors for type 1 interferon , IFNAR1 and IFNAR2 , were expressed in MDM and further induced starting around 5–6 hours . TNF has also been shown to initiate an autocrine loop in mouse macrophages responding to LPS , so that induction of some late response genes is ablated in TNF knockout mice [58] . Both TNF receptor genes TNFR1 ( TNFRSF1A ) and TNFR2 ( TNFRSF1B ) were induced later in the response to LPS . Aside from TNF , five other members of the TNF superfamily , TNFSF15 , TNFSF8 ( encoding CD30L ) , TNFSF10 ( TRAIL ) , TNFSF14 ( LIGHT ) and TNFSF9 ( CD137L ) were each induced ahead of induction of their respective receptors TNFRSF6B , TNFRSF8 , TNFRSF10A and 10B and TNFRSF9 ( CD137 ) . The rapid induction of prostanoid synthesis is a hallmark of LPS signaling , associated with the early induction of PTGS2 ( COX2 ) [26] . PGE receptors ( PTGER3 , 4 ) and the prostacyclin receptor ( PTGIR ) were strongly induced at later time points in the response . The gene for the vitamin D3-generating enzyme , CYP27B1 , was strongly-induced by LPS in MDM and the vitamin D3 receptor gene ( VDR ) was induced later in the response to LPS . Another set of potential autocrine loops was generated through the induction by LPS of CSF1 , GMCSF ( CSF2 ) and GCSF ( CSF3 ) , commencing around 2–3 hours after LPS treatment . Their receptors , CSF1R and CSF3R , were expressed constitutively and CSF2RB was strongly-induced by LPS around 4–5 hours . The cytokine IL15 was induced substantially in MDM by 2 hours , whilst IL15RA was induced from undetectable levels starting at 2–3 hours . Genes for both IL27 subunits ( IL27 and EBI3 ) were strongly induced , and MDM expressed IL27RA constitutively . Finally , many of the inducible CC chemokines probably produce autocrine signals through CC receptors; some were expressed constitutively , and some ( e . g . CCR1 , CCRL2 , CCR5 , CCR7 ) were themselves LPS-inducible with distinct individual time courses . One striking feature of the response to LPS is the transience of each successive wave of inducible gene expression ( S1 Fig , Fig 2 ) . Each wave probably contains the seeds of its own destruction , inducible repressors that promote decay of the signal and degradation of the induced transcripts and proteins . For example , the set of induced autocrine loops elicited by LPS includes progressive accumulation of the repressive cytokine IL10 , alongside its receptor , IL10RB . The LPS time course data supports many known and suggests several novel feedback loops . Amongst the earliest inducible transcription factors in Cluster 5 , PRDM1 ( also known as BLIMP1 ) was the most highly-expressed . In mice , the PRDM1 protein has been attributed functions as a transcriptional repressor of cytokine induction [59] . IKZF1 ( ikaros ) has also been attributed roles in feedback repression of LPS signaling in mice [60] and was associated with a trans-acting expression quantitative trait locus ( eQTL ) in a large survey of gene expression in human peripheral blood leukocytes [61] . NR4A1 ( encoded by NR4A1 ) has also been implicated as a transcriptional repressor of NFκB signaling in mouse macrophages [62] . Alongside these inducible feedback repressors of transcription , Cluster 5 also contains promoters for feedback regulators of MAP kinase signaling , DUSP1 and DUSP2 [63] . All of the inhibitors of NFκB ( NFKBIA , NFKBIB , NFKBIE , NFKBIZ , BCL3 ) were rapidly induced by LPS , as were genes for molecules such as TNFAIP3 ( A20 ) which mediates feedback inhibition of NFκB through regulated ubiquitination . OTUD1 encodes a novel inducible deubiquitinase ( DUB ) enzyme , in the same structural class as TNFAIP3 but not previously implicated in control of the LPS response . Other induced genes for feedback repressors of the initial signaling cascade include the protein tyrosine kinase LYN , multiple members of the TRIM family of E3 ubiquitin ligases which promote degradation of signaling molecules ( TRIM5 , TRIM10 , TRIM25 , TRIM35 , TRIM36 , TRIM38 ) , the TRAF inhibitor TANK , multiple inhibitor microRNAs ( notably miR146A , miR155 , miR21 , miR3648 , miR4741 ) , IER3 , each of the GADD45 family members , which amongst many other targets , probably inhibit p38 MAP kinases [64] , the caspase inhibitor TNFAIP8 and the transcription factor ATF3 [28] . PPP1R15A ( also known as GADD34 ) , a regulator of protein phosphatase 1 is identified as a feedback regulator of TLR-induced phosphorylation of TAK1 [65] . One gene of unknown function that was clearly induced was ZBTB10 ( RINZF ) , most likely also encoding a transcriptional repressor [66] . Many proinflammatory genes encode short-lived mRNAs with AU-rich elements in their 3’UTR , subject to degradation by the ZFP36 ( tristetraprolin ) gene product [62 , 67] . Two related genes , ZFP36L1 and ZFP36L2 , have also been implicated in control of mRNA stability in LPS-stimulated mouse macrophages [68] . However , by contrast to mouse macrophages , in human MDM , ZFP36 is repressed by CSF1 ( compared to high levels in monocytes ) , and none of the ZFP36 family genes was highly-inducible by LPS . Instead , ZC3H12A , another gene encoding a novel ribonuclease that controls stability of other inflammatory cytokine mRNAs , notably IL6 and IL12p40 [69] , was highly-induced by LPS , within a similar time course to TNF , but remained elevated even after 48 hours . Interestingly , PARP14 , encoding a member of a family of intracellular proteins that generate ADP-ribose posttranslational adducts , was strongly induced by LPS , commencing around 2 hours , and peaking at 12 hours ( within Cluster 3 ) . PARP14 forms a complex with ZFP36 and the AU-rich element in the mRNA 3' untranslated region of the tissue factor ( TF ) gene [70] . The neighbouring PARP15 and PARP9 genes were also induced by LPS . PARP9 shares a shares bidirectional promoter with the E3 ubiquitin ligase gene DTX3L , and the two proteins interact to control , amongst other things , the function and expression of IRF1 [71] . Other members of the PARP family , PARP7 , PARP10 and PARP12 , which control protein translation [72] and/or feedback inhibit NFκB signaling [73] were also LPS-inducible at later time points . Based upon the function in mice , the inducible expression of the signaling molecule gene , IRAK2 [74] and the regulator of intracellular trafficking , optineurin [75] probably contributes positively to the sustained induction of inflammatory genes . HCAR2 and HCAR3 ( also known as GPR109A and GPR109B ) are neighbouring duplicated genes in the human genome , and encode receptors for butyrate and niacin . Both genes were induced in parallel by LPS . In mice , GPR109A has been associated with feedback regulation of the LPS response and suppression of macrophage reactivity to gut luminal contents [76] . PELI1 ( Pellino1 ) encodes an E3 ubiquitin ligase that is required for TRIF-dependent signaling from TLR3 and TLR4 [77] and probably enables subsequent induction of the IFN target genes . XBP1 lies downstream of the gene for the ER stress sensor kinase , IRE1alpha ( ERN1 ) , and in mice XBP1 was found to be required for optimal and sustained cytokine production by macrophages responding to LPS [78] . The gene for another target of ER stress , PPP1R15B [79] was induced in parallel with XBP1 . MSC , encoding musculin or activated B cell factor-1 ( ABF-1 ) [80] , a repressor of bHLH transcription factors in muscle and B cells which has not previously been reported in macrophages , was also within this cluster . One novel finding was the marked induction of PDSS2 . PDSS2 encodes prenyl ( decaprenyl ) diphosphate synthase , subunit 1 , an enzyme involved in the synthesis of coenzymeQ ( CoQ ) . It is not known whether the activity of this enzyme limits flux through the pathway , but heterozygous mutation in another gene in the CoQ pathway , Mclk1 ( Cog7 ) in mice , produced increased levels of TNF in macrophages , and hypersensitivity to LPS [81] . Hence , this gene probably also contributes to feedback inhibition of the LPS response . The autocrine interferon response induces its own set of feedback regulators , notably the obvious suppressors of cytokine signaling ( SOCS1 , SOCS3 and SOCS6 ) which were each induced with distinct time courses ( S2 Table ) . PLEKHF2 is linked to control of interferon production [82] . LPS-inducible GPR183 ( also known as EBI2 , or EB virus induced 2 ) is a feedback regulator of type 1 interferon pathways [83] . The ligand for EBI2 is 7α , 25-dihydrocholesterol and the enzyme that synthesises it , CYP7B1 , was also induced strongly by LPS . Finally , the induction of the metal ion responsive transcription factor , MT1 , is linked to the recently described role of zinc in feedback regulation of NFκB activation and inflammatory transcription [84] . Indeed , genes for the zinc transporters , SLC39A8 ( ZIP8 ) , SLC39A14 ( ZIP14 ) and SLC30A1 ( ZNC1 ) , were each strongly induced by LPS , commencing from around 4–5 hours after induction , and rising continuously . LPS stimulation of mouse macrophages was reported to induce transcription from some 3000 enhancer loci , preceding modification of histone methylation [31 , 85] . eRNAs are relatively unstable , and are degraded by the RNA-exosome complex [86] . However , CAGE enables their quantitative detection , and the activation of eRNA transcription can be correlated with subsequent activation of promoter activity in putative target genes in the chromosomal vicinity [35] . Many enhancers described by the FANTOM5 consortium came from stimulated monocytes . The genes encoding limiting exosome components , EXOSC3 and EXOSC10 were low in MDM , and repressed transiently between 1 and 2 hours after LPS stimulation . CAGE tags derived from the genes encoding several other exosome components , EXOSC1 , EXOSC2 , EXOSC5 and EXOSC7 , were almost undetectable in monocytes in culture . The relative lack of the exosome complex may facilitate the detection of active enhancers in monocytes and macrophages . S2 Fig shows profiles of enhancer activation at a selected subset of inducible genes in which there was robust activation of transcription that clearly preceded the appearance of transcripts from associated promoters . In every case , the enhancers that have been identified in the time series are robustly supported by evidence of inducible bi-directional transcription in the LPS-stimulated monocytes . The data strongly support the view that enhancer transcription precedes activation of target promoters [35] . In many of these cases there were numerous individual regulated enhancers apparently associated with a single inducible promoter . For example , the IL6 locus has recently been dissected in detail in human monocyte-derived macrophages responding to LPS [56] , showing the roles of STAT1 and IRF1 in establishing permissive chromatin architecture in the vicinity of the gene , and identifying sites up to 50kb upstream of the transcription start site . The FANTOM5 data for the LPS induction series revealed LPS-inducible bidirectional transcription up to 150 kb upstream of IL6 . Fig 3 shows a time course of detection of transcription of each of these elements; nine separate elements had detectable activity in LPS-stimulated MDM in advance of the peak of IL6 promoter activity . These elements were more readily detectable in LPS-stimulated monocytes , where almost all of the 25 detected enhancers had significant transcriptional activity . The chemokine genes CCL3 and CCL4 are coregulated by LPS . Here again , there was bidirectional promoter activity associated with enhancers outside , and between , the two inducible genes , more readily detected in the LPS-stimulated human monocytes ( Fig 4 ) . Two enhancers downstream of CCL4 ( to the right of the panel ) were induced ahead of coordinated induction of CCL3 and CCL4 promoter activity . By contrast , the neighbouring CCL18 gene was much more slowly induced , and four enhancers appeared to show more prolonged activation . The entire region shows evidence of bidirectional promoter activity in LPS-stimulated monocytes . Finally , the region surrounding the TNFAIP3 locus contained an array of enhancers extending over 400kb , at least 25 of which had detectable induction of transcription in advance of the activation of the promoter ( Fig 5 ) . A recent study identified around 120 long non-coding RNA ( lncRNA ) that were induced by LPS in human monocytes based upon RNAseq [87] . We reexamined these loci individually in the FANTOM5 data . All but one of the top 20 most-inducible candidate lncRNA were contained within large clusters of enhancers ( so-called super-enhancers ) identified by the FANTOM5 data , and were actually associated with bidirectional transcription in LPS-stimulated monocytes . The potential target loci identified included IL6 and TNFAIP3 ( discussed above ) , as well as IDO1 , miR155 , ACSL1 , IRF2 , HS3ST3B1 , TNFSF8 , DDX58 , CD38 and SLAMF7 . The only exception is NONCO3094 , which is driven by a strongly-LPS-induced promoter that is antisense to IL7 ( which is not itself expressed in monocytes/macrophages ) . Based upon the extensive linked CTSS associated with enhancers in these regions , there is some question as to whether the proposed lncRNA could actually be artificially concatenated over-lapping short RNAs . When sets of co-regulated genes share a particular motif in their promoters , we can infer that DNA binding protein ( s ) that recognize the motif regulate that set of genes . We utilized Motif Activity Response Analysis ( MARA ) [88] to identify transcription factor binding motifs associated with the response of the MDM to LPS . Fig 6 shows the motif activities for the most active motifs . There was a temporal separation of motif activities associated with enhancers ( red lines in Fig 6 ) , which peaked before the activation of promoters through the same motifs ( blue lines in Fig 6 ) . The exception to this pattern was the macrophage-specific transcription factor , SPI1 ( PU . 1 ) , which showed early motif activity for enhancers , but no subsequent activation on promoters . This finding is consistent with the proposed role for PU . 1 as a pioneer transcription factor that defines “latent enhancers” that are subsequently bound by other factors in response to agents such as LPS [32 , 89] . The earliest response detected by MARA was associated with the Fos/Jun , or AP1 motif , consistent also with these factors being amongst the earliest targets of LPS activation ( see above ) . Subsequently , the increased activity of the NFKB/REL/RELA motif peaked around 2 hours , but declined only slowly ( Fig 6 ) . An initial burst of activity probably derives from the translocation of a preexisting complex subsequent to activation by IKB kinases [27] . Subsequently , from around 2 hours , genes encoding all of the members of the NFκB family , REL ( c-rel ) , RELA ( p65 ) , RELB , NFKB1 and NFKB2 , were induced and presumably contributed to both replacement of the cytoplasmic pool of NFκB and transcriptional regulation of a distinct late-response set of target promoters containing the NFKB/REL/RELA motif . The progressive increase in motif activity of IRF1/2 , IRF7 and STAT2 , 4 , 6 was initiated later in the time course , and is entirely consistent with the induction of IFNB1 , and of the IRFs themselves , as discussed above . The MARA analysis reinforces the likely importance of the stress response in LPS action . NFE2 ( also known as NRF ) sites were previously identified as active motifs in LPS-stimulated mouse macrophages [30] . The stress response factors , NFE2L1 ( NRF1 ) and NFE2L2 ( NRF2 ) are the likely occupiers of the NRF sites , being expressed constitutively in MDM but further inducible by LPS . The motif annotated as NFATC ( 1…3 ) is most likely bound by NFATC1 , which was induced by LPS , peaking at 2 hours . NFATC1 has two promoters encoding alternative 5’ exons in both mouse and human , as discussed in a recent review [90] . The more distal promoter was stimulated by LPS . NFATC1 is itself a downstream target of both AP1 and NFκB factors [90] . MARA also implies a function for the MEF2 family . MEF2A , MEF2C and MEF2D , which recognise the MEF2 motif , were all expressed constitutively in MDM , as described by others [91] and the latter two were down-regulated by LPS late in the time course . Like NFκB these proteins undergo regulated nuclear translocation [92] . MEF2 most likely underlies the motif activity attributed to the serum response factor SRF ( another MADS box protein ) . Although there have been reports of the activity of SRF in mouse macrophages [93] SRF was expressed at barely-detectable levels in MDM , similar to the Ets components of the serum-response element complex ( ELK1 , ELK3 , ELK4 ) . The absence of detectable motif activity based upon MARA does not constitute evidence that any particular factor does not contribute to transcriptional regulation . The GGAA/T core motif bound by ETS family factors is relatively uninformative , and overlaps with the purine-rich PU . 1 motif found in many myeloid-specific promoters [16] . Genes encoding ETS factors with related binding motifs , including ETS2 , ELF4 , FLI1 , ETV3 , ETV5 and ETV7 ( TEL ) were each up-regulated by LPS with distinct kinetics . Sweet et al . reported previously that ETS2 is not only induced , but is phosphorylated on the pointed domain , required for effective trans-activation , in response to LPS [94] . MARA is also rather insensitive for the detection of activity of factors that bind GC-rich motifs within CpG island promoters . For example , the “KLF4” motif ( CaCaCCC ) showed little change in activity across the LPS time course . However , as evident from the clusters in Fig 1 , multiple members of the KLF family were regulated in response to LPS . KLF2 was one of the immediate early genes . KLF7 has two promoters , one induced by 2 hours , another induced later in the time course . KLF5 was induced even later ( around 3–4 hours , cluster 43 ) . KLF3 also has two promoters , one induced at 7–8 hours , one at 20 hours . Each of these KLF family members could have a distinct function and distinct target genes . In mice , KLF4 was suggested to be a feedback regulator , and to polarise macrophages towards M2 phenotype [95] . In humans , KLF4 was constitutively expressed in monocytes and almost completely repressed in MDM . It was transiently , but relatively weakly-induced by LPS . KLF6 did not cluster in the coexpression analysis , having an idiosyncratic expression pattern . It was induced rapidly , peaking by around 80 mins in all samples , but unlike most early response genes , it declined slowly and remained somewhat elevated even after 48 hours . KLF6 is in a rather gene poor-region of the genome , and is surrounded by numerous enhancers identified in the FANTOM5 dataset . Of these , at least 4 were detectably and transiently-induced ahead of the peak of accumulation of KLF6 . KLF6 is required for optimal LPS-induced gene expression in mice [96] . Based on the observations described above , we hypothesised that IBD-associated genes , of which NOD2 is an archetype , are likely to be specifically-inducible in monocytes in response to inflammatory stimuli AND down-regulated during differentiation to macrophages . In order to test this hypothesis , we quantified the evidence for disease association in the regions surrounding promoters meeting these criteria . 2413 promoters exhibited this expression pattern , out of a total of 201 , 801 promoters in the FANTOM5 dataset . After correcting for linkage disequilibrium , regional co-regulation , and the genomic distribution of variants ( see Methods ) , we found that variants associated with Crohn’s disease by GWA were very strongly enriched in the regions surrounding this set of promoters ( Fig 7 ) . Less enrichment was seen for genes putatively associated with ulcerative colitis , and no significant enrichment was observed for control sets of loci/SNPs associated by GWA with the other traits ( rheumatoid arthritis ( RA ) , LDL cholesterol , and height ) . The association between CD loci and regulated transcription in macrophages can also be visualized by examining the expression profiles of promoters that lie within 2kb of a putative Crohn’s disease-associated SNP ( p<10e-6 ) . This subset is displayed in Fig 2B . To enable visualization of the results of the CAGE data with the location of disease-associated SNPs , we created a genome viewer that displays locations of FANTOM5 promoters and enhancers together with the P values for association with CD and UC and other traits ( http://gsht . baillielab . net; available for review; this resource is currently being transferred to a permanent server , which will be available at the time of publication . We also provide a direct link to visualize the association data on the ZENBU browser ( see Methods ) . The most extensive meta-analysis of IBD GWA data to date [4] identified 163 loci that met genome-wide significance thresholds for CD , UC or IBD . A subsequent study identified an additional 38 loci that underlie shared genetic risk across different ethnicities [5] . Most dissections of candidate genes in GWAS focus on the nearest gene to the peak of association , or the genes within the genomic interval of association , and assume that actual causal variants are in high LD with the SNP ( s ) . The genomic intervals are not always precise , and of course they may contain enhancers that alter transcription of genes outside that interval . The criteria we tested ( specifically inducible in monocytes; down-regulated during differentiation to macrophages ) identify plausible alternative candidates in many known IBD loci . For example , the MUC19 gene on Chr12 is annotated by Jostins et al . [4] as the only gene with the IBD GWA candidate interval defined by SNP rs11564258 . However , Lacour et al . [97] identified an extended haplotype that includes the neighbouring LRRK2 gene , a clear candidate for which there is already functional evidence of a role in IBD [98] . LRRK2 is identified by our criteria , and shares transcriptional regulation with NOD2 , being expression and further induced by IFNγ in monocytes and ablated by CSF1 in monocyte-derived macrophages . S3 Table provides a detailed manual curation of genes within >200 genomic intervals from published IBD datasets , inspected using the viewer described above , and gene expression derived from the FANTOM5 data . All but 12 of the intervals within the boundaries of significant GWA peaks for IBD contained promoters/genes for which expression was strongly monocyte-macrophage-enriched and/or regulated by CSF1 , LPS or both . For some regions identified by GWA , no candidate gene had been proposed but the CAGE data implicate a novel candidate that was expressed and/or regulated specifically in monocytes or macrophages or both . For others , our analysis identified alternative candidates to those that are currently emphasized . For example , NOD2 was regulated in parallel with two flanking genes , SNX20 and CYLD ( cylindromatosis ) . These three genes are close to monocyte-specific , IFNγ -inducible enhancers , detected by CAGE , located in the interval between NOD2 and SNX20 . All three genes were expressed highly in monocytes and almost absent from MDM grown in CSF1 . SNX20 has been implicated in control of P-selectin location , and might therefore also contribute to extravasation of monocytes [99] . CYLD encodes a deubiquitinating enzyme that exerts feedback control on both NFκB and MAP kinase pathways and has itself been implicated in inflammation control [100] . So , even in this archetypal CD susceptibility locus , NOD2 , the gene that has been most studied , is not necessarily the best or only candidate , notwithstanding evidence of protein-coding variation . Another well-studied candidate IBD susceptibility gene , ATG16L1 , was expressed ubiquitously in the FANTOM5 atlas , providing no support for a gut-specific or inflammatory pathology . By contrast , the neighbouring gene , INPPP5D ( inositol polyphosphate 5-phosphatase , also known as SHIP1 ) , shares tight co-regulation with NOD2 , being high in monocytes , induced by IFNγ , down-regulated in MDM , and further ablated by LPS . Mice deficient in SHIP1 have chronic inflammatory disease , including a Crohn’s-like colitis [101] . A recent detailed analysis of the ATG16L1 association with IBD revealed that the commonly-studied SNP is in linkage disequibilibrium with SNPs across an extended haplotype and additional variants could implicate any region of the locus [102] . The CAGE data also identified at least four monocyte-specific , regulated enhancers in the 40kb interval between ATG16L1 and INPPP5D , including one only 2 . 5 kb upstream of the ATG16L1 locus ( Chr2:234156397 ) . Immediately downstream of ATG16L1 , another candidate within the genomic interval identified by linage to CD is DGKD , diacylglycerol kinase delta , also expressed highly in monocytes and ablated completely in MDM . The hypothetical involvement of dysregulated autophagy in IBD [4] , is based largely on association with ATG16L1 and is cast into doubt if this gene is not the only , or even the most likely , candidate gene in the region . The association of IBD susceptibility with autophagy also rests in part upon genetic linkage to IRGM , which has been linked in turn to the functions of the large family of inducible GTPases in the mouse . The original study of human IRGM [103] refers to the gene as an ortholog of mouse Irgm1 , but the location of human IRGM in the genome is actually not syntenic with mouse ( http://www . ensembl . org ) . Humans have lost the inducible GTPase family as a mechanism of host defense , and IRGM was shown not to be interferon-inducible in humans [104] . The evidence that the human IRGM locus actually encodes an expressed transcript or a functional protein that is detected in any cell type is equivocal [105] . Strongly indicating that it does not , the FANTOM5 CAGE data detects no expression of IRGM mRNA in any cell population or tissue in any state of activation . The IRGM SNP associated with IBD susceptibility is in strict LD with a 20kb deletion polymorphism , 2 . 5kb upstream of the putative TSS of IRGM [106] . One alternative explanation for the association is that the IRGM region contains distal enhancers for macrophage-expressed genes . Chromatin immunoprecipitation sequencing ( ChIPseq ) data derived from monocytes and macrophages suggests this is the case , with clear peaks for binding of PU . 1 and enhancer marks [20] corresponding to a cluster of enhancers identified by CAGE , which can be visualised on the ZENBU browser . Amongst the neighbouring genes , SMIM3 , encodes a small integral membrane protein about which little is known , but it is up-regulated in MDM and further induced by LPS . The most likely candidate is TNIP1 , encoding a regulator of NFKB signalling , which has been implicated in genetic susceptibility to other inflammatory diseases [107] . The association of the NKX2 . 3 locus on chromosome 10 with IBD susceptibility has been widely-reported and replicated , but no function of the gene product associated with inflammation has been described . Within the same region , SLC25A28 encodes a mitochondrial iron transporter ( also known as mitoferrin 2 ) which showed the same pattern of regulation as NOD2 , down-regulated in MDM compared to monocytes , but very strongly-induced by LPS . Similarly , no mechanistic link has been made between the SNP associated with the IL23R locus and IBD susceptibility , but there has been considerable emphasis in the literature on the central role of Th17 T cells , activated by IL23 , in disease pathogenesis [4] . There is an alternative myeloid-expressed candidate gene , MIER , within the linked candidate interval surrounding IL23R . Previous data have revealed that promoters and enhancers identified by CAGE are strongly enriched for informative SNPs in GWAS [35] . One specific example is the IL12B locus , where the SNP rs6871626 , associated with CD susceptibility , is located within one of a cluster of LPS-inducible upstream enhancers . Interestingly , susceptibility to tuberculosis in an African cohort has been associated with a separate SNP cluster 3’ of the IL12B locus and of the neighbouring UBLCP1 gene ( [108] ) . The UBLCP1 gene is expressed ubiquitously , but these remote enhancers were activated by LPS in both monocytes and MDM . Similarly , Jiang et al [109] identified multiple allelic variants in the PTGER4 locus on chromosome 5 , that associated independently with CD . The FANTOM5 data revealed around 50 LPS-inducible enhancers in the 0 . 5Mb upstream of PTGER4 . Confirming this pattern , the subset of CTSS within 2kb of a putative CD-associated promoter has a distinct pattern in the LPS time course data . Prior to the adoption of population GWAS , family-based studies identified six IBD loci with very high relative risks amongst related individuals . IBD1 contained NOD2 , and IBD3 , the well-known association with the complex HLA locus . IBD5 , located around chromosome 5q31 , was located to a missense mutation in the organic cation transporter , SLC22A4 [110] and a promoter variant in the neighbouring SLC22A5 gene . The combined haplotype conferred a 7 . 5 fold risk of Crohn’s disease [111] . Consistent with our hypothesis , SLC22A4 was highly-expressed in CD14+ monocytes , and completely down-regulated by CSF1 . The IBD6 locus was located on Chromosome 19p13 , by analysis of patients/families that did not possess NOD2 variants [112] . Subsequent analysis revealed multiple variants associated with the linked EMR1 , EMR2 , EMR3 and CD97 cluster of genes that contributed to the overall IBD susceptibility . All four of these genes were expressed at high levels in monocytes , and strongly down-regulated by CSF1 . The strong expression of EMR1 contradicts previous report claimed that human EMR1 is expressed only in eosinophils [113] . Emr1 in mice encodes the F4/80 antigen , widely used as a macrophage marker [113] . The expression of EMR1 in humans is clearly different in mice , where the mRNA and protein are actually induced by CSF1 and are retained on tissue macrophages . However the phenotype of a knockout of the gene in mice implies a function in oral and peripheral tolerance [114] . To further explore candidate loci , we considered one individual component of the criteria described above that is central to our hypothesis: the large set of promoters that is expressed in freshly-isolated blood monocytes ( excluding those that were isolated by adherence and culture in vitro ) and either up or down-regulated at least 5-fold in monocyte-derived macrophages grown in CSF1 . These promoters and their expression values in all samples analysed are shown in S4 Table; in many cases we identified multiple promoters from the same locus , as well as distal enhancers that have not been ascribed the gene name , but may be connected based upon their apparent strict coexpression [35] . Excluding existing candidate genes such as NOD2/SNX20 , amongst the promoters down-regulated by CSF1 in MDM , and associated with specific gene names , at least 88 genes were identified as functional candidates lying within 200kb intervals with a peak p value of at least 10−6 . These are annotated in S4 Table . Amongst these new candidates , several transcription factors ( JUN , FOS , FOSL2 , JUND , ETS1 , ETS2 , NFKB1 , NFKBIZ , NFKBIA , CREM , SMAD3 , BATF3 , NFAT5 , NR1D1 , RARA , NFACTC4 , FOXP1 , PRDM1 , NFIL3 , KLF3 ) are of particular interest since they are likely to produce trans-acting impacts . Note that CSF1 strongly down-regulates many of the non-MHC genes within the HLA region on chromosome 6 ( and to a lesser extent , most Class II MHC ( HLA-D ) transcripts ) providing an alternative explanation for some of the well-documented HLA association with CD . Amongst the CSF1-inducible genes annotated in S4 Table there were a further 46 candidates . This set did not include any transcription factors , and in the large majority of cases , the association was specific to either CD or UC . In passing , we noted that the CSF1-inducible set separately identifies many genes , including CSF1 itself , with loci associated with lipid traits ( e . g . LDL cholesterol ) , and as previously noted [115] , CSF1 induced many genes associated with lipid metabolism . The IBD loci identified by Jostins et al . [4] do not include any on the X chromosome , which is consistently under-represented in GWAS [116] . Nevertheless , there is evidence for X-linked IBD susceptibility [117] . Amongst the genes on the X chromosome , TLR8 is strongly monocyte-specific , and is part of an extended haplotype associated with IBD [118] . ARHGEF6 , also linked to CD [116] , is also strongly monocyte-enriched and expression was abolished in MDM . In summary , on the sole basis of regulation in monocytes by CSF1 as a biological prior , we identify a further 134 candidate genes associated with IBD susceptibility , in addition to novel candidates within IBD loci identified from traditional linkage analysis . S4 Table contains many additional regions that are not associated with a gene name , and the larger set of 2400 promoters analysed in Fig 7 ( S5 Table ) contains 392 named genes that will undoubtedly expand the list of candidate functional loci still further .
We analysed the response of human macrophages , differentiated in CSF1 , to the TLR4 agonist LPS in order to determine whether this mirrors the development of gut macrophages and hence whether gut macrophages might be involved in the development of IBD . The direct relevance of this system to IBD is reinforced by the recent identification of a functional missense variant in the LPS receptor , TLR4 , associated with Crohn’s disease [2] . We have produced an unprecedented analysis of the transcriptional events during the response of macrophages to LPS . Figs 1 and 2 and S1 Fig overview the complex transcriptional cascade of feed-forward and feedback regulation of sets of genes so that the transcriptome is not stable even after 48 hours . The earliest events detected were the induced transcription of enhancers ( [37] ( S2 Fig ) . As a class these regions are opened by binding of pioneer transcription factors . In the case of macrophages , the dominant pioneer is PU . 1 [20 , 89 , 119 , 120] . The observation that enhancer transcription precedes activation of target promoters is known from other systems , and indeed one might use this correlation to infer the likely connection between distal enhancers and promoters [35] . What is less obvious in all of the examples shown in Figs 3 , 4 , 5 and S2 , is that the induction of enhancer activity appears transient , even when the putative target gene is induced relatively stably . The transcripts are presumably rapidly degraded by the exosome complex . This suggests that the act of transcription of an enhancer is associated with its activation , but the transcript per se is less likely to be functional . Subsequent to early enhancer activation , wave after wave of transcription factors was induced ( Fig 1 ) followed by their putative target genes ( Fig 2 ) which were identified by enrichment of target motifs in their promoters ( Fig 6 ) . Genes for the inducible transcription factors , such as the members of the IRF family , formed clusters with their likely target genes , which contained binding site motifs detected by MARA . The temporal data provide indications of distinct and specific roles . For example , distinct members of IRF family were found in different clusters whereas in a previous analysis of monocyte eQTL data , IRF7 and IRF9 clustered together [47] . The endpoint of differentiation of mucosa-associated macrophages is a population that is longer responsive to restimulation with LPS or other microbial challenges [121] . The FANTOM5 data demonstrate that MDM grown in CSF1 , in common with mucosal macrophages , strongly down-regulated most pathogen recognition molecules expressed at high levels in monocytes , notably TLR1 , 2 , 4 and 6 , NOD2 and other Nod-like receptors ( NLR ) and non-NLR inflammasome activators ( NLRC3 , NLRC5 , NLRP1 , NLRP3 , NRLP6 , MEFV , PYCARD ) , C-type lectins CLEC7A ( dectin1 ) , CLEC5A , CLEC4D , CLEC4E and CLEC12B; SIGLECs 3 , 5 , 9 , 10 and 14 , formyl peptide receptors FPR1 and FPR2 , recently described cytoplasmic LPS receptors CASP4 and CASP5 [36] and the nucleic acid detector CGAS . There is clearly the potential for dysregulation of any of these individual pattern recognition receptors , which could explain why it has been difficult to link IBD incidence to particular microbial challenges common to all patients [1 , 122] . Compared to monocytes , MDM also shared with human gut macrophages the suppression of CD14 ( the LPS coreceptor ) , and adaptor and signaling molecules such as MYD88 , TRIF , TRAF6 and IRAK1 . So , although they clearly do respond to LPS , the response of MDM is greatly attenuated compared to the response of blood monocytes . The FANTOM5 data assays the blood monocyte response to LPS at only a single time point , but based upon the quantitative measure of CAGE tag frequency , the classical inflammatory cytokines , IL1B and IL6 were each induced by LPS to levels 10–100 fold greater in monocytes than at any time point in the MDM time course . By contrast , the monocyte attractants , CCL2 , CCL3 and CCL4 were induced to comparable levels in MDM and monocytes . We suggest that once inflammation has been initiated through the activation of hyper-responsive incoming monocytes , or resident macrophages , accelerated recruitment of further monocytes would induce a self-perpetuating hyperinflammatory state . TGFB1 has also been implicated in mucosal macrophage differentiation , and strongly implicated in IBD [123] . The modulated response of MDM to LPS probably involves this factor . The MDM are most likely autocrine for TGFB1 signaling , since they express high levels of TGFB1 mRNA , and both receptors , TGFBR1 and TGFBR2 . The rigorous feedback control that is documented here has previously been seen as a way to produce robustness and stringent control [24] . If this were true , one might expect the system to be resilient in the face of varied expression or function of individual components . This is clearly not the case . Most of the literature on the function of inducible repressors derives from studies of inbred mice , in which knockout of any one these regulators produces dysregulated LPS responses [24 , 25 , 83] . Parnas et al . [124] reported a genome-wide CRISPR screen that revealed hundreds of genes apparently required for the production of TNF by mouse macrophages in response to LPS . The structure of the screen did not reveal as many genes involved in negative regulation . To gain further insight into the scale of feedback control of the response , we searched PubMed to identify mouse knockouts that impact the response to LPS . S6 Table identifies >180 additional mouse loci where a null mutation is viable but produces a global change in macrophage responsiveness to LPS in vitro and/or in vivo . The list does not include all of the inflammatory cytokines and chemokines , and essential signaling molecules . Almost 2/3 of these mutations produced hypersensitivity to LPS administration or sepsis in vivo , and many also produced spontaneous colitis . The set of LPS susceptibility genes in S6 Table is annotated with the expression from the FANTOM5 data , and excludes the obvious signaling molecules and inflammatory cytokines and chemokines in the KEGG Toll-like receptor signaling pathway ( hsa04620 ) . In every case , the FANTOM5 data confirms that the genes are also expressed/enriched in human monocytes and/or regulated by CSF1 , LPS , or both . All of this discussion indicates that there may be hundreds of different genomic variants that can each contribute to an increased susceptibility to IBD . A major limitation of GWAS is the assumption that common variants underlie the risk of disease in different families . The hypothesis-free GWA approach can drive understanding of new areas of biology . Conversely , we have shown clearly in Fig 7 that a biological prior , based upon transcriptional co-expression , in this case based upon expression and regulation in monocytes , can focus analysis on a subset of genomic regions and specific candidate genes within those regions , thereby reducing the false discovery rate and the genome-wide significance threshold . Our hypothesis reduced the number of candidate genes/promoters analysed by two orders of magnitude ( around 2000 of 200 , 000 promoters ) . A previous study aimed at identifying gene features linked to CD susceptibility noted the differential expression of genes associated with “M1” versus “M2” macrophage activation states as a criterion [12] . The definition of those activation states is somewhat problematic [125 , 126] . We have taken a more nuanced view , with a much more extensive dataset , to emphasise the specific importance of CSF1 and monocyte differentiation . NOD2 is amongst the most studied of loci underlying Crohn’s disease susceptibility , and there are several protein-coding variants apparently linked to disease . However , there is little consensus about NOD2 function in the GI tract and the mechanistic link between sequence variants and disease susceptibility [127] . NOD2 mRNA is undetectable in either colon or small intestine in the CAGE data , supporting the view that it is rapidly down-regulated during monocyte differentiation to macrophages . NOD2 is believed to be expressed in Paneth cells , and to influence their functions including production of defensins . If NOD2 is expressed in Paneth cells , it is likely to be at a very low level , since Paneth cell markers such as DEFA6 , and stem cell markers such as LGR5 , were easily detected in the total intestine mRNA . Our data suggest that NOD2 is more likely to have a function in monocyte differentiation . Furthermore , we suggest that NOD2 is not necessarily responsible for the genetic linkage on 16q , since the neighbouring genes , CYLD and SNX20 are tightly co-regulated and lie within the interval of maximal association . The linkage to NOD2 is seen even in families where there is no coding variation in NOD2 , and an extended haplotype includes CYLD and the shared enhancer/promoter region between SNX20/NOD2 [128] . Similarly , the case for biological roles of autophagy in Crohn’s disease is also based upon linkage to a coding variant , T300A , in a key gene , ATG16L1 , but in this case also the link to function has been elusive [129] and the expression and functional data suggest a more likely candidate in the neighbouring INPPP5D ( SHIP1 ) gene . If macrophage differentiation in response to CSF1 is a key event in mucosal macrophage anergy to intestinal flora , CSF1 itself , the alternate CSF1R ligand , IL34 , and the receptor CSF1R , would also be obvious candidate genes for IBD susceptibility . Two recent studies have focused on elevated IL34 expression in IBD [130 , 131] . Both claim that IL34 is expressed in uninflamed intestinal mucosa , but the FANTOM5 data indicate that the levels are very low . The IL34 gene has two promoters , one expressed highly in skin and spleen , and the other in brain , consistent with evidence from the mouse IL34 knockout that IL34 controls the development of Langerhans cells and microglia [132] . Based upon the data herein , the elevated expression of IL34 mRNA and protein [130 , 131] in inflamed mucosa might actually be part of a feedback control to dampen inflammation and initiate repair . As noted above , CSF1R lies within 0 . 5MB of the IRGM locus linked with CD susceptibility by GWA , and might potentially share upstream regulatory elements . Heterozygous mutation in Csf1r has been shown to protect against pathology in a colitis model in mice [133] . There is no evidence of association with IBD in the more immediate vicinity of CSF1R in the GWA data , but one report based upon direct sequencing in a Acadian American population indicated linkage to an intron 11 SNP [134] . Whole genome sequence data from 1000 genomes and other sources ( http://www . ensembl . org ) , and the recent human exome paper [135] also reveal the existence of numerous likely loss-of-function alleles , with allele frequencies of 1/1000 or more , within the intracellular tyrosine kinase domain of CSF1R . So , there may be rare/private mutations in CSF1R that impact on IBD susceptibility . CSF1 was strongly induced in MDM compared to monocytes , suggesting that these cells become autocrine . Variation in the vicinity of an upstream enhancer at the CSF1 locus is very strongly linked to Paget’s disease [136] , and there is some evidence of association with UC at the same interval ( S4 Table ) . Given the complexity of the transcriptional regulation in monocytes and macrophages , and the lack of robustness discussed above , it is not at all surprising that there are hundreds of ways in which genetic variation can alter the sensitivity to environmental challenge , on the one hand underlying infectious disease resistance , and on the other producing susceptibility to IBD . Some of the genes discussed herein are expressed in both T cells and macrophages . For some , it is clear that expression in macrophages is most relevant to the gut . Macrophage-specific conditional deletion of at least two IBD susceptibility loci , Il10ra [137] and Stat3 knockout [138] has been shown to generate spontaneous colitis in mice . By contrast , in the case of Ship1 ( Inppp5d ) , conditional deletion studies in mice suggest that ileitis present in the constitutive knockout involves both myeloid and T cell dysfunction [101] . Many of the genes implicated in Mendelian very early onset inflammatory bowel diseases ( VEOIBD ) ( reviewed in [139 , 140] ) such as all the components of the phagocyte NADPH oxidase system , are also highly-expressed in blood monocytes . They include some , such as TTC7A , XIAP , BTK and MEFV that are down-regulated by CSF1 and therefore meet our criteria . The FANTOM5 data provide a novel insight into some VEOIBD loci . For example , both STXBP2 and DOCK8 genes contain distinct monocyte-specific promoters , and the latter also shows evidence of linkage to CD . Current GWA hits still account for only a small proportion of the heritability in IBD . There are likely to be epistatic interactions between susceptibility loci , which could explain some of the apparently missing heritability [141] . On the other hand , rare coding variants that are specific to particular populations , that might have much larger effect , are invisible to GWA . A recent example is a frame-shift mutation in the CSFR2B gene [142 , 143] , also highly-expressed in monocytes and regulated by CSF1 and LPS , associated with Crohn’s disease in Ashkenazi Jews . Several groups have sought evidence of rare protein-coding variation within larger populations , with success limited in large measure by the lack of family-based inheritance data and the lack of a biological hypothesis to prioritise variants [144–146] ( reviewed in [6] ) Identification of candidate genes within genomic intervals identified by GWA or direct sequencing currently relies on programs such as GRAIL ( Gene Relationships Across Implicated Loci ) . Van Limbergen et al . [8] in a recent review , suggested that knowledge of the genetic architecture of IBD has uncovered biological processes involved in IBD pathogenesis , but that new insights would require knowledge of regulation of transcription of key genes and cell-type specific experiments . Our detailed reanalysis of all of the susceptibility loci thus far identified in GWA support the view that the most relevant cell type to study is the monocyte-macrophage and the prevalent candidate mechanism underlying specific IBD susceptibility is a dysregulation of their differentiation in response to CSF1 and their subsequent response to TLR agonists . Westra et al [61] identified candidate genes based upon an eQTL analysis of blood transcriptional profiles . This is somewhat less sensitive than our approach because monocytes are only a subset of the total blood cells . Nevertheless , supporting our hypothesis , 15/16 separate genes identified as both eQTL and associated with IBD by these authors were also repressed in MDM compared to monocytes and/or induced by LPS in our dataset . Similarly , Fairfax et al . [47] described an eQTL analysis of the response of human monocytes to LPS or IFNγ in which >80% of genes showed genetic association under some condition , and associated a subset of those eQTL to CD susceptibility loci; notably CARD9 ( which is completely repressed by CSF1 ) . As noted in analysis of the entire FANTOM5 dataset [34 , 35] , SNPs associated with promoters and enhancers are strongly enriched for disease associations . The dataset we have created therefore provides a resource for identification and prioritization of SNPs that are likely to have direct effects on gene expression and causal links to disease .
This monocyte-macrophage dysregulation hypothesis for IBD is based upon the specific premise that the differentiation of monocytes in response to CSF1 , and their subsequent response to LPS , is unique to the gut and the response to the gut microbiome . Clearly , there are also aspects of T cell differentiation that are specific to the gut , for example the Th17 pathway , but there is currently limited direct evidence that any variation in these functions is causally linked to IBD susceptibility . Some of the T cell-associated genes and pathways are controlled by macrophages and their products , including the many cytokines discussed above , so that there is clearly the potential for pleiotropy . In general , variants that affect T cell activation would seem likely to influence multi-system inflammatory disease susceptibility . Similarly , not all variants that affect monocyte differentiation need impact solely on the gut . The obvious role of monocytes in all forms of inflammation could provide an explanation for shared susceptibility loci between IBD and other chronic inflammatory diseases [2] , but Fig 7 shows clearly that the CD-specific enrichment is not shared with RA . Our data indicate that the specific association with differentiation in response to CSF1 and/or TLR ligands exemplified by LPS provides an explanation for specific association with inflammation in the unique environment of the GI tract . Of course , the advantage of this hypothesis is that is testable by comparing the transcriptional profiles of MDM from patients and unaffected siblings . That will be a future direction .
A full description of CAGE protocols was provided previously [34] . Ethical approval for the isolation of cells from anonymous donors was granted by the University of Edinburgh Research Ethics Committee . This study was reviewed by the University of Edinburgh College of Medicine Ethics Committee ( 2009/01 ) and subsequently renewed by the Lothian Research Ethics Committee ( 11/AL/0168 ) . Written informed consent was received from all volunteers participating in the study . Blood samples from three donors were used . Peripheral blood mononuclear cells were isolated from a single donation of 320ml of whole blood anticoagulated with acid-citrate-dextrose , by Ficoll density gradient centrifugation . Monocytes were extracted using selection for CD14-positive cells using antibodies on magnetic beads ( MACS , Miltenyi Biotech ) . Monocytes were plated in 6-well plates at 800 , 000 cells per well in complete media ( RPMI containing 10% FCS , 20U/ml penicillin , 20 μg/ml streptomycin , and 2mM l-glutamine ( Invitrogen ) ) with 100ng/ml recombinant human CSF-1 ( a gift from Chiron Corp , Emeryville , CA ) on tissue culture plastic for 7 days . Supplemental medium ( 50% of the volume in each well , containing 300ng/ml CSF-1 ) was added at day 5 , and cells were used on day 7 . Cells were treated with 10ng/ml bacterial lipopolysaccharide ( LPS ) from salmonella Minnesota R595 as described previously [21] , and then harvested at time points from 15 minutes to 48 hours after treatment by cell lysis , before RNA extraction using the Qiagen RNeasy kit ( Qiagen , UK ) . CAGE library preparation and sequencing was carried out on each RNA sample , and clusters of transcription start sites ( CTSS ) were identified by decomposition-based peak identification as described [34] . The tag counts associated with CTSS meeting the FANTOM5 ‘robust’ criteria ( >200 , 000 CTSS ) were quantified and normalized as described in [34] . As discussed previously , many human genes have multiple promoters , and they are numbered in the FANTOM5 dataset in order of relative abundance of detected CAGE tags . The majority of the analysis herein is based upon quantification of individual promoters . Primary access to all of the FANTOM5 data , including comparable mouse data , is available at http://fantom . gsc . riken . jp/zenbu/ All of the primary data , including the normalized tag counts for each individual replicate at every time point , can be visualized in Table form by accessing the ZENBU browser for the human data directly at: http://fantom . gsc . riken . jp/zenbu/gLyphs/#config=dXO5cTaJBZiiw73fJq2oGD;loc=hg19::chr17:34413143 . . 34438026+ The link opens a specific locus in a genome browser . A search window can be used to move to other loci based upon gene names or genomic interval . One track shows only the LPS time course data . Another shows the entire FANTOM5 data set . The expression of individual promoters can be highlighted and entire set of expression profiles across the primary FANTOM5 data is available for download from the site . In order to reduce noise arising from technical variation and biological differences between cells from different volunteers , the expression value for each CTSS was compared to the expression values of time points immediately before or after it , and the corresponding values from the other two donors . Values deviating widely ( >3SD ) from the mean of this pool of 8 values were replaced with the average of the pool . An average expression value for each CTSS from the three volunteers at each time point was then calculated . CTSS with a minimum expression level of 10 tags per million in at least one timepoint , and with a coefficient of variation > 0 . 5 , were included in expression analysis . A pairwise coexpression network was generated using the Pearson correlation between expression profiles for each CAGE-defined CTSS across the LPS timecourse . Correlations were included where r ≥ 0 . 9 . Coexpressed clusters were detected using the clustering algorithm , MCL ( MCLtools , www . micans . org; inflation value = 1 . 6 ) . Heatmaps and line graphs of expression data were created using the matplotlib library for Python 2 . 7 . 2 . For the set of transcription factor genes identified previously [34] , total expression associated with all the promoters was aggregated , the three replicates were averaged and only samples that had at least one time point with ≥ 20 tags per million ( TPM ) were included in the analysis . Gene-based expression data were clustered using BioLayout Express3D ( http://biolayout . org ) . The network was again based on correlations of r ≥ 0 . 9 between genes . A sample-to-sample Pearson correlation for the pattern of transcription factor expression was calculated , and the resultant graph with correlations of r ≥ 0 . 95 and an MCL inflation value of 2 . 2 was also displayed using Biolayout Express3D Expression profiles of transcripts of named genes arising near genetic variants significantly associated with Crohn’s disease were examined manually and are described in the text . Following the observation that many of these named transcripts shared a similar pattern of expression with NOD2 , numerical criteria were determined , to describe the characteristics of this pattern . Subsets of cell types were defined according to the groups in S2 Table . The following criteria were determined a priori to describe the expression pattern observed: To assess the association of these candidates with IBD susceptibility , Crohn’s disease summary p-values were obtained from the International Inflammatory Bowel Disease Genetics Consortium ( ftp://ftp . sanger . ac . uk/pub4/ibdgenetics/cd-meta . txt . gz . ) . For comparison , we applied the same criteria to loci putatively associated with ulcerative colitis [ftp://ftp . sanger . ac . uk/pub4/ibdgenetics/ ucmeta-sumstats . txt . gz] , rheumatoid arthritis [http://plaza . umin . ac . jp/~yokada/datasource/software . htm] [147] and to a comparable size set of genes associated with unrelated pathologies: the level of LDL cholesterol in blood [http://csg . sph . umich . edu/abecasis/ public/lipids2013/] [148] , height [http://portals . broadinstitute . org/collaboration/giant/index . php/GIANT_consortium_data_files] [149] . The ZENBU Browser cited above also shows individual tracks for the p values for association of SNPs with each of the chosen traits , so that they can be precisely aligned with enhancers and promoters within chosen loci . For each of these studies , GWAS variants with p-values stronger than a permissive threshold ( p<1e-6 ) were counted in 1000 equally-sized bins above and below all TSS in the input set ( observed counts ) , and compared to the absolute count of all variants genotyped in the same study in the same distance bins from all TSS in the input set ( expected counts under the null hypothesis that no enrichment exists ) . For Crohn's disease , ulcerative colitis and the mixed phenotype of inflammatory bowel disease , linkage disequilibrium score regression analyses reveal strong evidence of polygenicity , supporting our decision to draw inferences from variants that do not meet stringent criteria for genome-wide significance . These analyses were conducted by Bulik-Sullivan et al [150] using the same summary p-value data that we use here . The ratio of observed:expected ratio was plotted for each bin ( Fig 7 ) . Summary p-values for this set of loci were computed using PASCAL [151] to calculate the total “burden of significance” within a range of +/-100kb from each TSS region , for each GWAS study , and then to quantify the probability of such a signal arising by chance using the PASCAL pathway enrichment feature to compensate for linkage disequilibrium .
|
Macrophages are immune cells that form the first line of defense against pathogens , but also mediate tissue damage in inflammatory disease . Macrophages initiate inflammation by recognising and responding to components of bacterial cells . Macrophages of the wall of the gut are constantly replenished from the blood . Upon entering the intestine , newly-arrived cells modulate their response to stimuli derived from the bacteria in the wall of the gut . This process fails in chronic inflammatory bowel diseases ( IBD ) . Both the major forms of IBD , Crohn’s disease and ulcerative colitis , run in families . The inheritance is complex , involving more than 200 different regions of the genome . We hypothesised that the genetic risk of IBD is associated specifically with altered regulation of genes that control the development of macrophages . In this study , we used the comprehensive transcriptome dataset produced by the FANTOM5 consortium to identify the sets of promoters and enhancers that are involved in adaptation of macrophages to the gut wall , their response to bacterial stimuli , and how their functions are integrated . A reanalysis of published genome-wide association data based upon regulated genes in monocytes as candidates strongly supports the view that susceptibility to IBD arises from a primary defect in macrophage differentiation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Conclusions",
"Methods"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"gene",
"regulation",
"regulatory",
"proteins",
"immunology",
"dna-binding",
"proteins",
"dna",
"transcription",
"gastroenterology",
"and",
"hepatology",
"transcription",
"factors",
"inflammatory",
"bowel",
"disease",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"gene",
"expression",
"genetic",
"loci",
"biochemistry",
"cell",
"biology",
"monocytes",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"macrophages"
] |
2017
|
Analysis of the human monocyte-derived macrophage transcriptome and response to lipopolysaccharide provides new insights into genetic aetiology of inflammatory bowel disease
|
Recent studies suggest that deep Convolutional Neural Network ( CNN ) models show higher representational similarity , compared to any other existing object recognition models , with macaque inferior temporal ( IT ) cortical responses , human ventral stream fMRI activations and human object recognition . These studies employed natural images of objects . A long research tradition employed abstract shapes to probe the selectivity of IT neurons . If CNN models provide a realistic model of IT responses , then they should capture the IT selectivity for such shapes . Here , we compare the activations of CNN units to a stimulus set of 2D regular and irregular shapes with the response selectivity of macaque IT neurons and with human similarity judgements . The shape set consisted of regular shapes that differed in nonaccidental properties , and irregular , asymmetrical shapes with curved or straight boundaries . We found that deep CNNs ( Alexnet , VGG-16 and VGG-19 ) that were trained to classify natural images show response modulations to these shapes that were similar to those of IT neurons . Untrained CNNs with the same architecture than trained CNNs , but with random weights , demonstrated a poorer similarity than CNNs trained in classification . The difference between the trained and untrained CNNs emerged at the deep convolutional layers , where the similarity between the shape-related response modulations of IT neurons and the trained CNNs was high . Unlike IT neurons , human similarity judgements of the same shapes correlated best with the last layers of the trained CNNs . In particular , these deepest layers showed an enhanced sensitivity for straight versus curved irregular shapes , similar to that shown in human shape judgments . In conclusion , the representations of abstract shape similarity are highly comparable between macaque IT neurons and deep convolutional layers of CNNs that were trained to classify natural images , while human shape similarity judgments correlate better with the deepest layers .
Recently , several studies compared the representations of visual images in deep Convolutional Neural Networks ( CNN ) with those of biological systems , such as the primate ventral visual stream [1–4] . These studies showed that the representation of visual objects in macaque inferior temporal ( IT ) cortex corresponds better with the representations of these images in deep CNN layers than with representations of older computational models such as HMAX [5] . Similar findings were obtained with human fMRI data [6–10] . The images used in these studies were those of real objects in cluttered scenes , which are the same class of images as those employed to train the deep CNNs for classification . Other single unit studies of IT neurons employed two-dimensional ( 2D ) shapes and observed highly selective responses to such stimuli ( for review see [11] ) . If deep CNNs provide a realistic model of IT responses , then the CNNs should capture also the selectivity observed for such two-dimensional shapes in IT . To our knowledge , thus far there has been no comparison between the 2D-shape representation of IT neurons , measured with such reduced stimuli , and that of deep CNN models . It is impossible to predict from existing studies that compared deep CNN activations and neurophysiology whether the deep CNNs , which are trained with natural images , can faithfully model the selectivity of IT neurons for two-dimensional abstract shapes . Nonetheless , such correspondence between CNN models and single unit selectivity for abstract shapes is critical for assessing the generalizability of CNN models to stimuli that differ markedly from those of the trained task but have been shown to drive selectively IT neurons . Previously , we showed that a linear combination of units of deep convolutional layers of CNNs trained with natural images could predict reasonably well the shape selectivity of single neurons recorded from an fMRI-defined body patch [4] . However , in that study , we adapted for each single unit the shapes to the shape preference of that neuron , precluding a comparison between the shape representation of the population of IT neurons and deep CNNs . To perform such a comparison , one should measure the responses of IT neurons to the same set of shapes . Furthermore , the shape set should include variations in shape properties IT neurons were shown to be sensitive to . Also , the IT response selectivities for such shapes should not trivially be explainable by physical image similarities , such as pixel-based differences in graylevels . Kayaert et al . [12] measured the responses of single IT neurons to a set of shapes that varied in regularity and the presence of curved versus straight boundaries ( Fig 1 ) . The first group of stimuli of [12] was composed of regular geometric shapes ( shown in the first two rows of Fig 1 and denoted as Regular ( R ) ) that all have at least one axis of symmetry . These shapes are simple , i . e . , have low medial axis complexity [13] . The stimulus pairs in each column of these two rows ( denoted by a and b ) differed in a non-accidental property ( NAP ) . NAPs are stimulus properties that are relatively invariant with orientation in depth , such as whether a contour is straight or curved or whether a pair of edges is parallel or not . These properties can allow efficient object recognition at different orientations in depth not previously experienced [14–16] . NAPs can be contrasted with metric properties ( MPs ) , which vary with orientation in depth , such as aspect ratio or the degree of curvature . The three other groups are all ‘Irregular’ . They differed from the Regular shapes in that they do not have a single axis of symmetry . The two shapes in each row of the three Irregular groups differed in the configuration of their concavities and convexities or corners . The shapes in the Irregular Simple Curved ( ISC ) set all had curved contours . The Irregular Simple Straight ( ISS ) shapes were derived from the ISC shapes by replacing the curved contours with straight lines . Thus , the corresponding stimuli in the ISS and ISC shapes differed in a NAP . Last , the Irregular Complex ( IC ) group was more complex in that the shapes in that group had a greater number of contours . Kayaert et al . [12] found that anterior IT neurons distinguished the four groups of shapes . Importantly , the differences in IT responses amongst the shapes could not be explained by pixel-based gray level differences , nor by HMAX C2 unit differences . In fact , none of the tested quantitative models of object processing could explain the IT response modulations . Furthermore , the IT response modulations were greater for the Regular shapes and when comparing the curved and straight Irregular Simple shapes than within the 3 Irregular shape groups , suggesting a greater sensitivity for NAPs than for MPs ( see also [17 , 18] ) . We reasoned that this shape set and corresponding IT responses was useful to examine to what degree different layers of deep CNNs and IT neurons represent abstract shapes similarly . We employed deep CNNs that were pretrained to classify ImageNet data [19] , consisting of images of natural objects in scenes . Hence , the CNNs were not exposed during training to silhouette shapes shown to the IT neurons . Deep CNNs have a particular architecture with early units having small receptive fields , nonlinear pooling of units of the previous layer , etc . Such a serial , hierarchical network architecture with increasing receptive field size across layers may result in itself , i . e . without training , in changes in the representational similarity across layers . To assess whether potential correlations between IT and CNN layer response modulations resulted from classification training or from the CNN architecture per se , we also compared the activations of untrained CNNs with the IT response modulations . Kayaert et al . [12] had also human subjects sort the same shapes based on similarity and found that human subjects had a pronounced higher sensitivity to the difference between the curved and straight simple irregular shapes ( relative to the regular shapes ) than the IT neurons . We examined whether a similar difference in response pattern between macaque IT neurons and human similarity judgements would emerge in the deep CNNs . We expected that deeper layers would resemble the human response patterns while the IT response pattern would peak at less deep layers .
Kayaert et al [12] recorded the responses of 119 IT neurons to the 64 shapes shown in Fig 1 . The 64 shapes are divided in four groups based on their regularity , complexity and whether they differed in NAPs . We presented the same shapes to 3 deep CNNs: Alexnet [20] , VGG-16 , VGG-19 [21] and measured the activations of the units in each layer of the deep nets . These deep nets differ in their number of layers , the number of units in each layer and the presence of a normalization stage , but each have rectifying non-linearity ( RELU ) and max pooling stages ( Fig 2 ) . We employed deep nets that were pre-trained in classification of a database of natural images , which were very different in nature from the abstract shape stimuli that we employ here to test the models and neurons . The aim was to compare the representations of the shapes between IT neurons and each layer of the deep nets . To do this , we employed representational similarity analyses [22 , 23] , following the logic of second order isomorphism [24 , 25] , and examined the correlation between the neural IT-based similarities and CNN-based similarities in responses to shapes . We are not trying to reconstruct the shapes based on IT neuron or CNN unit outputs but we are examining whether shapes that are represented close to each other in the neural IT space are also represented close to each other in the CNN layer space . In a first analysis , we computed the pairwise dissimilarity between all 64 stimuli using the responses of the IT neurons and the activations in each of the CNN layers . We employed two dissimilarity metrics: Euclidean distance and 1 –Spearman rank correlation ρ . The dissimilarity matrices computed with the Euclidean distance metric for the IT neurons and for 5 layers of the trained CNNs are illustrated in Fig 3B and 3C , respectively . In this and the next figures , we will show only the data for Alexnet and VGG-19 , since VGG-16 and VGG19 produced similar results . In addition , Fig 3A shows the pixel-based dissimilarities for all image pairs . Visual inspection of the dissimilarity matrices suggests that ( 1 ) the pattern of dissimilarities changes from the superficial to deep layers in a relatively similar way in the CNNs , ( 2 ) the dissimilarity matrix of the first layer ( e . g . conv1 . 1 ) resembles the pixel-based similarities ( Fig 3A ) and ( 3 ) the deeper layers resemble more the IT neural data ( Fig 3B ) . We quantified the similarity between the IT shape representation and that of each layer by computing the Spearman Rank correlation between the corresponding pairwise dissimilarities of IT and each layer . Thus , we could assess to what degree stimuli that produce a very different ( similar ) pattern of responses in IT also show a different ( similar ) pattern of activations in a CNN layer . We found that for both dissimilarity metrics the similarity between IT neuronal responses and trained CNN layer activations increased significantly with the depth of the layer . This is shown using the Euclidean distance metric for Alexnet and VGG-19 in Fig 4 ( see S1 Fig for the data of both distance metrics and the 3 networks ) . In the VGG nets , the similarity peaked at the deepest convolutional layers ( Fig 4 ) and then decreased for the deepest layers . In fact , the Spearman correlations for the last two fully connected layers did not differ significantly from that of the first convolutional layer in each CNN ( Fig 4 ) . The decrease in similarity for the deepest layers was weaker in Alexnet . The peak similarity was similar amongst the 3 nets , with ρ hovering around 0 . 60 , and were larger for the correlation ( mean peak ρ = 0 . 64 ) compared with Euclidean distance metric ( mean peak ρ = 0 . 58 ) . To assess the degree to which the models explained the neural data , we computed the reliability of the neural-based distances giving the finite sampling of the IT neuron population . This noise ceiling was computed by randomly splitting the neurons into two groups , computing the dissimilarities for each group , followed by computation of the Spearman rank correlation between the dissimilarities of the two groups . This split-half reliability computation was performed for 10000 random splits . Fig 4 shows the 2 . 5 , 50 ( median ) and 97 . 5 percentiles of the Spearman-Brown corrected correlations between the two groups . The correlations between ( some ) CNN layers and neural responses were close but still below the estimated noise of the neural dissimilarities . In order to assess to what degree the similarity between neural data and the CNN layers reflects the architecture of the CNNs versus image classification training , we computed also the similarity for untrained networks with random weights . Fig 3C illustrates dissimilarity matrices computed using Euclidean distances for 5 untrained layers of two CNNs . Visual inspection suggests little change in the dissimilarity matrices of the different layers of the CNNs , except for fc8 . Furthermore , the pattern of dissimilarities resembled the pixel-based dissimilarities shown in Fig 3A . Both observations were confirmed by the quantitative analysis . The Spearman correlations of the neural data and untrained CNNs increased only weakly with depth , except for a marked decrease in correlation for the last two fully connected layers . Except for the deep convolutional and the last two layers , the trained and untrained networks showed similar Spearman correlations of the neural and CNN distances ( Fig 4 ) . This suggests that overall the similarity between the IT data and the shallow CNN layers are unrelated to classification training but reflect merely the CNN architecture . Significant differences between trained and untrained CNNs were observed for the deeper convolutional layers ( Fig 4 ) , suggesting that the similarity between IT and the deep convolutional layers depends on classification training . The similarities for the first fully connected layer ( fc6 and relu6 in Fig 4 ) did not differ significantly between the trained and untrained layers ( except for the correlation metric in AlexNet ( S1 Fig ) . The deepest two ( fully connected ) layers showed again a significantly greater similarity for the trained compared with the untrained networks . However , this can be the result of the sharp drop in correlations for these layers in the untrained network . Overall , these data suggest that the shape representations of the trained deep convolutional layers , but not of the deepest layers , shows the highest similarity with shape representations in macaque IT . Receptive field ( RF ) size increases along the layers of the CNNs , allowing deeper layer units to integrate information from larger spatial regions . The difference in IT-CNN similarity between untrained and trained layers shows that the increase in RF size cannot by itself explain the increased IT-CNN similarity in deeper layers , since untrained CNN also increase their RFs along the layer hierarchy . Also , the decrease in similarity between IT responses and the fully connected layers argues against RF size being the mere factor . Nonetheless , although not the only contributing factor , RF size is expected to matter since arguably small RFs cannot capture overall shape when the shape is relatively large . Hence , it is possible that the degree of IT-CNN similarity for different layers depends on shape size , with smaller shapes showing a greater IT-CNN similarity at earlier layers . We tested this by computing the activations to shapes that were reduced in size by a factor of two in all layers of each of the 3 trained CNNs . Fig 5 compares the correlations between dissimilarities of the trained Alexnet and VGG-19 networks and IT dissimilarities for the original and reduced sizes , with dissimilarities computed using Euclidean distances . The stars indicate significant differences between the similarities for the two sizes ( tested with a FDR corrected randomization test; same procedure as in Fig 4 when comparing trained and untrained correlations ) . In each of the CNNs ( S2 Fig ) , the IT-CNN similarity increased at more superficial layers for the smaller shape . The overall peak IT-CNN similarity was highly similar for the two sizes in the VGG networks and occurred at the deep convolutional layers . For Alexnet , the overall similarity was significantly higher for the smaller shapes in the deep layers . This analysis indicates that shape size is a contributing factor that determines at which layer the IT-CNN similarity increases , but that for the VGG networks , peak similarity in the deep layers does not depend on size ( at least not for the twofold variation in size employed here ) . Note that also for the smaller size the IT-CNN similarity drops markedly for the fully connected layers in the VGG networks . Thus , the overall trends are independent of a twofold change in shape size . In the preceding analyses , we included all units of each CNN layer . To examine whether the similarity between the CNN layers and the IT responses depends on a relatively small number of CNN units or is distributed amongst many units , we reran the representational similarity analysis of deep CNN layers and IT neurons for the whole shape set for smaller samples of CNN units . We took for each network the layer showing the peak IT-CNN similarity and for that layer sampled 10000 times at random a fixed percentage of units . We restricted the population of units to those that showed a differential activation ( standard deviation of activation across stimuli greater than 0 ) since only those can contribute to the Euclidean distance . Fig 6A plots the median and 95% range of Spearman rank correlation coefficients between IT and CNN layer dissimilarities for the whole shape set as a function of the percent of sampled units for two CNNs . We found that the IT-CNN similarity was quite robust to the number of sampled units . For instance , for Alexnet , the IT-CNN similarity for the original and the 95% range of the 10% samples overlap , indicating that 315 Alexnet units can produce the same IT-CNN similarity as the full population of units . Note also that the lower bound of the 95% range is still above the IT-CNN similarities observed for the untrained network ( median Spearman rho about 0 . 40; see Fig 4 ) . This indicates that the IT-CNN similarity does not depend on a small subset of units , since otherwise the range of similarities ( Spearman rho correlations ) for the 10% samples would be much greater . The same holds for the other CNNs ( S3 Fig ) , except that these tolerated even smaller percent sample size ( for VGG19 even 0 . 1% , which corresponds to 100 units ) . The above analysis appears to suggest that the activations of the CNN units to the shapes are highly correlated with each other . To address this directly , we performed Principal Component Analysis ( PCA ) of unit activations of the same peak CNN layers as in Fig 6 and computed Euclidean distance based dissimilarities between all stimulus pairs for the first , first two , etc . principal components ( PCs ) , followed by correlation with the neural dissimilarities as done before for the distances computed across all units of a CNN layer . For both the Alexnet and VGG-19 layer , the first 10 PCs explained about 70% of the variance in CNN unit activations to the 64 stimuli ( Fig 7B ) . Only the first 3 ( Alexnet ) or 5 ( VGG-19 ) PCs were required to obtain a similar correlation between the model and neural distances as observed when using all model units of the layer ( Fig 7A; about 7 PCs were required for VGG-16; see S4 Fig ) . This analysis shows that the neural distances between the abstract shapes relate to a relatively low dimensional shape representation in the CNN layer , with a high redundancy between the CNN units . In the above analyses , we compared the overall similarity of the shape representations in IT and CNN layers . However , a more stringent comparison between the shape representations in IT and the CNNs involves response modulations for the shape pairs for which Kayaert et al [12] observed striking differences between predictions of pixel-based models or computational models like HMAX and the neural responses . The average response modulations ( quantified by pairwise Euclidean distances ) for the different group pairs comparisons are shown in Fig 8 for the IT neural data , the HMAX C2 layer and the pixel differences . Kayaert et al [12] showed that the mean response modulation in IT ( Fig 8A ) was significantly greater for the regular shape pairs ( 1–8 in Fig 1 ) than for the 3 irregular shape group pairs , despite the pixel differences between members of a pair being , on average , lower or similar for the regular group than for the 3 irregular groups ( Fig 8D ) . In addition , the response modulation to ISC vs . ISS was significantly greater than the modulations within IC , ISC and ISS , although the average pixel-difference within the ISC vs . ISS-pairs was much lower than the pixel-differences within the other pairs . This differential neural response modulation to ISC vs ISS was present for both members of the ISC and ISS pairs ( a and b members: “ISCa vs ISSa” and “ISCb vs ISSb” ) and thus was highly reliable . Note that the difference between ISC vs . ISS and the IC and ISS shape groups that are present in the neural data is not present for the HMAX C2 distances ( Fig 8C ) . Kayaert et al . [12] reported also a relatively higher sensitivity to the straight vs . curved contrast of the ISC vs . ISS comparison compared with the regular shapes in human similarity ratings ( Fig 8B ) , compared with the IT neural data . In other words , human subjects appear to be more sensitive to the curved versus straight NAP difference than macaque IT neurons . In a second analysis , we determined whether the marked differences in IT response modulations and human judgements shown in Fig 8 are present in the dissimilarities for the different layers of the deep CNNs . Fig 9 illustrates the results for 8 layers of VGG-19 . The left column of the figure plots the distances for the trained network . The dissimilarities for the first convolutional layer fits the pixel-based distances amongst the shape pairs ( Fig 8D; Pearson correlation between pixel-based distances and first layer distances = 0 . 966 ) , but differ from those observed in IT and for human judgements . Similar trends are present until the very deep convolutional layers where the dissimilarities became strikingly similar to those observed in macaque IT ( e . g . compare trained conv5 . 4 or pool5 of Fig 9 with Fig 8A ) . The dissimilarities for the last two layers ( e . g . trained relu7 and fc8 in Fig 9 ) are strikingly similar to those observed for the human judgements ( Fig 8B ) , and differ from the pattern seen in macaque IT neurons . Indeed , as noted above , the human judgements differ from the IT responses in their sensitivity for the ISC vs ISS comparison relative to that for the regular shape pairs: for the human judgement distances , the ISC vs ISS distances are greater than for the regular shape distances while for the neural distances both are statistically indistinguishable ( Kayaert et al . [12] ) . Therefore , we tested statistically for which CNN layer the ISC vs ISS distances were significantly greater than the regular shape distances ( Wilcoxon test ) , thus mimicking the human distances . We found a significant difference for the very deep VGG19 layer fc8 ( p = 0 . 039 ) and VGG16 layers fc7 ( p = 0 . 039 ) , relu7 ( p = 0 . 023 ) , and fc8 ( p = 0 . 023 ) . Although the deepest Alexnet ( fully connected ) layers showed the same trend , this failed to reach significance . These tests showed that only the very deep CNN layers mimicked the human judgements . None of the untrained CNN layers showed a dissimilarity profile similar to that observed in monkey IT or in human judgements ( Fig 9 , right column ) . In fact , the untrained data resembled more the pixel-based distances ( see Fig 8D ) . Indeed , the Pearson correlation between the pixel-based distances and the conv1 . 1 distances was 0 . 999 for the untrained VGG-19 . We quantified the correspondence between the neural response dissimilarities of Fig 8A and the CNN layer dissimilarities ( as in Fig 9 ) by computing the Pearson correlation coefficient between the dissimilarity profiles ( n = 6 dissimilarity pairs ) . The same quantification was performed for the human judgements ( Fig 8B ) and the CNN dissimilarities ( n = 5 pairs ) . These correlations are plotted in Fig 10A and 10B as a function of layer for two CNNs , trained and untrained . For the neural data , the correlations are negative for the shallow layers and highly similar for the trained and untrained CNNs . The negative correlations are a result of the nearly inverse relationship between neural and low-level ( pixel ) differences between the shapes ( Fig 8D ) . This was not accidental , but by design: when creating the stimuli , Kayaert et al [12] ensured that the NAP differences ( e . g . between ISC and ISS ) were smaller than MP differences . For both VGG networks ( S5 Fig; Fig 10B ) , there was a sharp increase in correlations at the trained deep 5 . 1 convolutional layer , followed by a decrease of the correlations for the fully connected layers . This trend was similar , although more abrupt , to that observed for the global dissimilarities of Fig 4 . For Alexnet , the increase of the correlations with increasing depth of the trained convolutional layers was more gradual , but like the VGG networks , high correlations were observed for the deeper trained convolutional layers . For the human judgement data , the correlations were already higher for the trained compared with the untrained CNNs at the shallow layers , although still negative . Like the neural data , there was a marked increase in correlation at the very deep trained layers . Contrary to the neural data , the correlations for the human judgements continued to increase along the trained fully connected layers , approaching a correlation of 1 at the last layer . These data show that the average response modulations of IT neurons for the shape groups of Fig 1 correspond nearly perfectly with those of the deeper layers of CNNs , while the differences in human similarity judgements between the groups are captured by the later fully connected layers of the CNNs . This holds for Alexnet and VGG nets . Note that the deep CNN layers performed better at predicting the neural IT and human perceptual dissimilarities than the HMAX C2 layer output ( Fig 10C ) . As for the representational similarity analysis for all shapes ( Fig 6A ) , we computed also the Pearson correlation coefficients between the dissimilarity profiles ( n = 6 dissimilarity pairs ) of the same peak CNN layers and the IT distances for the 6 shape groups ( as in Fig 10 ) for smaller samples of units . As shown in Fig 6B , we observed similar IT-CNN correlations for the within-group distances up to the 1% and 0 . 1% samples compared with the full population of units for Alexnet and VGG , respectively . Again , this suggests that IT-CNN similarity does not depend on a small number of outlier CNN units . The greater tolerance for percent sample size for the VGG units is because the VGG layers consisted of a larger number of units per se ( total number of units are indicated in the legend of Fig 6 ) . In addition , we computed the mean distances for the same layers and their correlation with the mean neural modulations as a function of retained PCs ( Fig 7B ) . Up to 30 PCs were required to obtain a similar correlation between neural and CNN layer distances for the six groups of shapes as when including all units of the layer ( Fig 7B ) . This suggests that the close to perfect modeling of the mean response modulations across the 6 shape groups required a relatively high dimensional representation of the shapes within the CNN layer .
The particular set of shapes that we employed in the present study was designed originally to test the idea that the shape selectivity of IT neurons reflects the computational challenges posed when differentiating objects at different orientations in depth [12 , 14] . Here , we show that deep CNNs that were trained to classify a large set of natural images show response modulations to these shapes that are similar to those observed in macaque IT neurons . We show that untrained CNNs with the same architecture than the trained CNNs , but with random weights , demonstrate a poorer IT-CNN similarity than the CNNs trained in classification . The difference between the trained and untrained CNNs emerged at the deep convolutional layers , where the similarity between the shape-related response modulations of IT neurons and the trained CNNs was high . Unlike macaque IT neurons , human similarity judgements of the same shapes correlated best with the deepest layers of the trained CNNs . Early and many later studies of IT neurons employed shapes as stimuli ( e . g . [26–31 , 22 , 32–37] ) , in keeping with shape being an essential object property for identification and categorization . Deep CNNs are trained with natural images of objects in cluttered scenes . If deep CNNs are useful models of biological object recognition [38] , their shape representations should mimic those of the biological system , although the CNNs were not trained with such isolated shapes . We show here that indeed the representation of the response modulations by rather abstract , unnatural shapes is highly similar for deep CNN layers and macaque IT neurons . Note that the parameters of these CNN models are set via supervised machine learning methods to do a task ( i . e . classify objects ) rather than to replicate the properties of the neural responses , as done in classic computational modeling of neural selectivities . Thus , the same CNN model that fits neural responses to natural images [1–4] also simulates the selectivity of IT neurons for abstract shapes , demonstrating that these models show generalization across highly different stimulus families . Of course , the high similarity between deep CNN layers and IT neurons activation patterns we show here may not generalize for ( perhaps less fundamental ) shape properties that we did not vary in our study . Kubilius et al . [39] showed that deep nets captured shape sensitivities of human observers . They showed that deep Nets , in particular their deeper layers , show a NAP advantage for objects ( “geons” ) , as does human perception ( and macaque IT [18] ) . Although we also manipulated NAPs , our shapes differed in addition in other properties such as regularity and complexity . Furthermore , our shapes are unlike real objects and more abstract than the shaded 3D objects employed by Kubilius et al . [39] when manipulating NAPs . In both the representational similarity analysis and the response modulations comparisons amongst shape groups , we found that the correspondence between IT and deep CNN layers peaked at the deep convolutional layers and then decreased for the deeper layers . Recently , we observed a similar pattern when using deep CNN activations of individual layers to model the shape selectivity of single neurons of the middle Superior Temporal Sulcus body patch [4] , a fMRI-defined region of IT that is located posterior with respect to the present recordings . The increase with deeper layers of the fit between CNN activations and neural responses has also been observed when predicting with CNN layers macaque IT multi-unit selectivity [40] , voxel activations in human LO [9] and the representational similarity of macaque and human ( putative ) IT [8 , 10] using natural images . However , the decrease in correlation between CNNs and neural data that we observed for the deepest layers was not found in fMRI studies that examined human putative IT [8 , 10] , although such a trend was present in [6] when predicting CNN features from fMRI activations . The deepest layers are close to or at the categorization stage of the CNN and hence strongly dependent on the classifications the network was trained on . The relatively poor performance of the last layers is in line with previous findings that IT neurons show little invariance across exemplars of the same semantic category [41 , 42] , unlike the deepest CNN units [43] . The question of what the different layers in the various CNN models with different depths represent neurally remains basically unanswered . Shallow CNN layers can be related to early visual areas ( e . g . V1; V4 ) and deeper layers to late areas ( e . g . IT ) . However , different laminae within the same visual area ( e . g . input and output layers ) may also correspond to different layers of CNNs . Furthermore , units of a single CNN layer may not correspond to a single area , but the mapping might be more mixed with some units of different CNN layers being mapped to area 1 , while other units of partially overlapping CNN layers to area 2 , etc . Finally , different CNN layers may represent different temporal processing stages within an area , although this may map partially to the different laminae within an area . Further research in which recordings in different laminae of several areas will be obtained for the same stimulus sets , followed by mapping these to units of different layers in various CNNs , may start to answer this complex issue . In contrast with IT neurons , human similarity judgements of our shapes matched to a greater extent the last rather than the less deep convolutional layers . In particular , the deepest layers showed a similar enhanced sensitivity for straight versus curved irregular shapes . The untrained CNNs did not show such straight versus irregular bias for the irregular shapes . Thus , it appears that a system , be it artificial like the CNNs or a biological system like humans , that is required to classify natural images of objects develops such bias for curved versus straight contours , indicating that this shape property must be highly informative for object categorization . Whether this relates to straight versus curved being a NAP [14] is unclear . Kayaert et al . [12] employed a sorting task to rate shape similarity . In this task , subjects were required to sort the shapes into groups based on their similarity . Although this is not the same as labeling an object , the task for which the CNNs were trained , higher order classification can intrude the sorting task judgements . This may explain why the human sortings of Kayaert et al . [12] resembled that closely the activation pattern seen at the deepest CNN layers , which are strongly category label driven . Interestingly , even for the shallow convolutional layers , the correlations between the human judgements and the CNN activations were higher for the trained compared with the untrained CNNs . This contrasted with the equal correlations for trained and untrained shallow layers for the IT data . This suggests that the trained shallow CNN layers show already some , albeit weak , bias for higher order category-related information . Previous studies that compared deep CNNs and neural responses rarely included untrained CNNs as control ( e . g . [8] , [40] ) . We found the untrained CNNs helpful in interpreting our data . The comparison with untrained CNNs can inform to what extent neural responses reflect features that can be picked up by untrained CNNs ( because of CNN architectural properties such as tiling of local RFs in shallow layers and non-linear pooling ) . Indeed , we found that most layers of the untrained CNNs represented rather closely the pixel-based graylevel differences between the shape groups , which assisted to interpret the representational similarity of the trained CNNs at shallow layers . Thus , we advise that future studies use untrained CNNs as control or benchmark . Currently , deep CNNs are the best models we have of primate object recognition , providing the best quantitative fits of ventral stream stimulus selectivities and primate recognition behavior [38] . However , recent studies show that CNNs have their limitations , especially when stimuli are noisy or partially occluded . For instance , the commonly used deep CNNs tolerate less image degradation than humans [44] , can be fooled by unrecognizable images [45] or show a sensitivity to imperceptible stimulus perturbations ( “adversarial examples”; [46] ) . Our data show that training CNNs in object categorization produces at least some shape selectivities ( that are thought to reflect fundamental aspects of shape processing [14] ) similar to those that are observed in neural IT data and human similarity judgements . This does not imply that CNNs can explain all shape or stimulus selectivity in IT and there is still considerable room for model improvement ( e . g . recurrent connectivity etc . ) . In conclusion , deep CNN layers that were trained to classify complex natural images represented differences among relatively simple abstract 2D-shapes similar to macaque IT neurons . Human sorting of the same shapes corresponded better with the deepest layers of the CNNs . The similarity between IT neurons and the deeper convolutional layers is greater for trained compared to untrained CNNs , suggesting a role of image classification in shaping the shape selectivity of macaque IT neurons . The latter likely occurs during ontogenetic development , but may not result from the same supervised learning algorithm as employed to train the CNNs . Indeed , independent of the particular training protocol ( e . g . supervised versus unsupervised ) , any biological object classification system may have similar shape representation biases that are inherently useful for performing invariant object classification .
Two male rhesus monkeys served as subjects . The animals were housed individually with visual and auditory contact with conspecifics . During the recording weeks , they had controlled access to fluids but food was available at libitum . All procedures were in accordance with the Weatherall report on “The use of non-human primates in research” and were approved by the Animal Ethics committee of the KU Leuven ( protocol number: P631/2002 ) . The 64 shapes were identical to the first stimulus set employed by Kayaert et al . [12] and are shown in Fig 1 . The Regular shapes R were created with Studio MAX , release 2 . 5 , while the Irregular shapes were made with Fourier Boundary Descriptors , using MATLAB , release 5 . The Irregular Simple Straight ( ISS ) stimuli were made by replacing the curves of the Irregular Simple Curved ( ISC ) shapes by straight lines while preserving the overall shape . The increase in complexity of the Irregular Complex ( IC ) shapes compared to the simpler ISC shapes was produced by increasing the number and frequency of the Fourier Boundary Descriptors . Each group contains 8 pairs of stimuli ( one stimulus in row a and one in row b in Fig 1 ) . The columns of Fig 1 comprise a set of 4 pairs ( one for each group ) that were matched in overall size and aspect ratio , both within and between groups . The averaged pixel-based graylevel differences between the members of the pairs were balanced across groups ( see [12] for more details ) . The members of the pairs within the Regular shapes differ in a NAP , such as parallel vs . nonparallel sides , or straight vs . curved contours . The differences among the members of an irregular pair were created by varying the phase , frequency or amplitude of the Fourier Boundary Descriptors . For the single unit recordings and the human behavioral study , all stimuli were filled with the same random dot texture pattern . The number of black and white dots was required to be equal for 2*2 squares of pixels , so the texture patterns were highly uniform . Stimuli were presented on a gray background . In the single unit study , they extended approximately 7 degrees and were shown at the center of the screen . We employed the identical shapes for the CNN modeling , except that the noise pattern was replaced by a uniform white surface ( see Fig 1 for the actual stimuli presented to the CNNs ) . The single unit data have been published before [12] and the procedures have been described in detail in that paper . Therefore , we will summarize here only briefly the experimental procedures . The IT recordings were made while the two monkeys performed a passive fixation task . Eye movements were measured with the scleral search coil technique or with a noninvasive eye tracker ( ISCAN ) . During the recordings , their head was fixed by means of an implanted head post . We employed the standard dorsal approach to IT and recording sites were verified with MRI and CT scans with the guiding tube in situ . We lowered a tungsten microelectrode through the guiding tube that was fixed in a Crist grid , which was positioned within the plastic recording chamber . The signals of the electrode were amplified and filtered using standard single-cell recording equipment . Single units were isolated on line and their timing was stored together with stimulus and behavioral events for later offline analysis . The stimuli were presented during fixation for 200 ms in a randomly interleaved fashion . In the present study , the response of a neuron was defined as the average firing rate in spikes/s during a time interval of 250 ms , starting from 50 to 150 ms after stimulus onset . The starting point of this time interval was chosen for each neuron to best capture its response , by inspection of the peristimulus time histograms averaged across the stimuli . Responses were averaged across presentations per stimulus . The minimum number of presentations per stimulus was 5 ( median = 10 ) . The data set consisted of 119 anterior IT neurons ( 76 in monkey 1 and 43 in monkey 2 ) that showed significant response selectivity to the stimuli of the set ( ANOVA , p<0 . 05 ) . The data were pooled across animals . The neurons were located in the lower bank of the Superior Temporal Sulcus and the lateral convexity of anterior IT ( TEad ) . As described by Kayaert et al . [12] , printed versions of the shapes were given to 23 naive adult human subjects who were asked to sort the stimuli in groups based on shape similarity . No further definition of similarity was given and they could freely choose the number of groups . This is a classical task to measure image similarities [47] . Dissimilarity values between pairs of stimuli were computed by counting the number of subjects that put the two members in different groups . In order to compare the shape representation of the IT neurons’ population with deep CNN layers , we extracted stimulus features for each processing stage ( layer ) of three deep models: Alexnet [20] , VGG-16 and VGG-19 [21] . We used the pretrained networks , which are available through the MatConvNet toolbox [48] in MATLAB , and their untrained versions . The pretrained CNNs were trained on ~1 . 2 million natural images divided in 1 , 000 classes for the ImageNet Large Scale Visual Recognition Challenge 2012 ( ILSVRC2012 ) . The untrained versions of these networks have the same architecture , but did not undergo any training , thus no update of their weights took place after initialization . Their layer weights were initialized by sampling randomly from a normal distribution , using the opts . weightInitMethod = 'gaussian' setting in the cnn_imagenet . m function of the MatConvNet toolbox . The stimuli shown to the CNNs were black and white images with pixel values ranging from 0–255 ( 0 for black and 255 for white ) . Before feature extraction , the mean of the ILSVRC2012 training images was subtracted from each stimulus , since this was also part of the preprocessing stage of the networks’ training procedure . In addition , the stimuli were rescaled accordingly to match each network’s input requirements ( 227x227 pixels for Alexnet and 224x224 pixels for VGG-16 & VGG-19 ) . In all analyses , we employed as distance metric the normalized Euclidean distance between the neuronal responses or deep CNN unit activations: ( Σin ( Ri1−Ri2 ) 2n ) 12 , where Ri1 is the response of neuron or deep CNN unit i , to stimulus 1 , and n is the number of neurons or the number of deep CNN units in a specific layer . For the representational similarity analyses , we computed also a second distance metric: 1- Spearman’s correlation coefficient . The Spearman rank correlation coefficient ρ was computed between the neural responses or CNN units’ activations for all stimulus pairs . To compare neuronal data to CNN layers , we performed representational similarity analysis [49] , using both distance metrics . We constructed representational dissimilarity matrices ( RDMs ) for the whole stimulus set ( n = 64 stimuli ) for both IT neurons and each deep CNN layer ( trained and untrained separately; for examples see Fig 3 ) , by arranging all possible pairwise distances in 64x64 RDMs . We extracted all values above the diagonal ( upper triangle of the RDM , excluding the diagonal ) of the symmetrical RDMs , and computed for each layer the Spearman rank correlation coefficient between the distances of the corresponding pairs of the neural and CNN matrices . We computed 95% confidence intervals of the Spearman correlation coefficient between neural and CNN distances by resampling with replacement 10 , 000 times 119 neurons out of our pool of IT neurons and correlating each time the resulting neural distance matrix with each deep CNN layer for the trained and untrained versions of the same network . The confidence intervals corresponded to the 2 . 5 and 97 . 5 percentiles of the bootstrapped correlation coefficient distributions . To assess whether the trained deep CNN layers significantly differed from the untrained , we computed for each CNN layer the distribution of the paired differences of trained minus untrained layer correlations across the 10 , 000 iterations ( one difference per bootstrapped neuronal sample ) . For each layer , we computed the percentile in the corresponding distribution of the zero difference value and these defined the p values of the test . For each of the 3 CNNs , we corrected the p values for multiple comparisons ( n = number of CNN layers ) using the Benjamini and Hochberg [50] False Discovery Rate ( FDR ) procedure . A difference between the trained and untrained CNNs was judged to be significant when FDR q < 0 . 05 . The same procedure was used to assess the significance of the difference in IT-CNN correlations between the original and reduced shape size for each of the CNN layers ( Fig 5 ) . We employed a similar procedure to test the significance of the difference in Spearman rank correlation coefficients of the neural and CNN distances between the first layer and each subsequent layer . Thus , we computed the pairwise difference between the correlation for the first and a subsequent layer for each of the 10 , 000 bootstrapped neural samples and then obtained the percentile of the zero difference in that distribution of differences . The p values were corrected for multiple comparisons using the FDR procedure and significance was defined when q < 0 . 05 . In the second analysis , using only the original , non-bootstrapped distances , we compared the pairwise Euclidean stimulus distances amongst the 4 stimulus groups R , IC , ISC , ISS . For each group , we included only the stimulus pairs numbered 1–8 in Fig 1 , i . e . for each group the members of the a and b rows of Fig 1 . In addition , we selected the distances for the “ISCa vs . ISSa” and “ISCb vs . ISSb” pairs of Fig 1 , e . g . the column-wise distances between row a of ISC and row a of ISS in Fig 1 ( likewise for the b rows ) . This produced twice 8 distances for the ISC versus ISS comparison , which we analyzed separately , unlike in Kayaert et al . [12] . For each of the 4 groups and the two ISC versus ISS comparisons , we computed the mean distance ( and standard errors of the mean ) across the 8 pairs per group or comparison . To quantify the relationship between the mean distances across groups for the neural data and each CNN layer , we computed the Pearson correlation coefficient between the mean neural distances and the mean distances of the CNN layers ( n = 6 pairs of distances per layer ) . A similar analysis was performed comparing the CNN layer distances and the distances based on the human ratings . However , for this analysis , the available human rating data consisted of the distances that were computed by Kayaert et al [12] , having an ISC versus ISS comparison of 8 stimulus pairs ( for selection of those pairs , see [12] ) instead of twice 8 pairs as above . We compared those distances with the average of the “ISCa vs . ISSa” and “ISCb vs . ISSb” pairs of the CNN layers . Note that our average neural distances for the “ISCa vs . ISSa” and “ISCb vs . ISSb” pairs were highly similar to those for the 8 “ISC vs . ISS” pairs selected by Kayaert et al . [12] , justifying this procedure . We compared neural dissimilarities also with dissimilarities based on pixel graylevels and the HMAX model [5] , employing the same procedures as in Kayaert et al . . We computed the Euclidean distance between the gray-level values of the pixels for all image pairs ( Fig 3B ) . In addition , we computed the Euclidean distances between the outputs of C2-units of the HMAX model as described by Riesenhuber and Poggio [5] and presented in [12] . The HMAX C2 units were designed to extract moderately complex features from objects , irrespective of size , position and their relative geometry in the image . HMAX-based dissimilarities were computed as the Euclidean distance between the output of the 256 C2 units .
|
The primate inferior temporal ( IT ) cortex is considered to be the final stage of visual processing that allows for object recognition , identification and categorization of objects . Electrophysiology studies suggest that an object’s shape is a strong determinant of the neuronal response patterns in IT . Here we examine whether deep Convolutional Neural Networks ( CNNs ) , that were trained to classify natural images of objects , show response modulations for abstract shapes similar to those of macaque IT neurons . For trained and untrained versions of three state-of-the-art CNNs , we assessed the response modulations for a set of 2D shapes at each of their stages and compared these to those of a population of macaque IT neurons and human shape similarity judgements . We show that an IT-like representation of similarity amongst 2D abstract shapes develops in the deep convolutional CNN layers when these are trained to classify natural images . Our results reveal a high correspondence between the representation of shape similarity of deep trained CNN stages and macaque IT neurons and an analogous correspondence of the last trained CNN stages with shape similarity as judged by humans .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"functional",
"magnetic",
"resonance",
"imaging",
"neural",
"networks",
"visual",
"object",
"recognition",
"vertebrates",
"social",
"sciences",
"neuroscience",
"animals",
"mammals",
"learning",
"and",
"memory",
"magnetic",
"resonance",
"imaging",
"primates",
"perception",
"cognitive",
"psychology",
"cognition",
"brain",
"mapping",
"memory",
"vision",
"neuroimaging",
"old",
"world",
"monkeys",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"imaging",
"techniques",
"monkeys",
"animal",
"cells",
"macaque",
"cellular",
"neuroscience",
"psychology",
"eukaryota",
"diagnostic",
"medicine",
"radiology",
"and",
"imaging",
"cell",
"biology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"sensory",
"perception",
"cognitive",
"science",
"amniotes",
"organisms"
] |
2018
|
Representations of regular and irregular shapes by deep Convolutional Neural Networks, monkey inferotemporal neurons and human judgments
|
In man , infection with South American Andes virus ( ANDV ) causes hantavirus cardiopulmonary syndrome ( HCPS ) . HCPS due to ANDV is endemic in Southern Chile and much of Argentina and increasing numbers of cases are reported all over South America . A case-fatality rate of about 36% together with the absence of successful antiviral therapies urge the development of a vaccine . Although T-cell responses were shown to be critically involved in immunity to hantaviruses in mouse models , no data are available on the magnitude , specificity and longevity of ANDV-specific memory T-cell responses in patients . Using sets of overlapping peptides in IFN-γ ELISPOT assays , we herein show in 78 Chilean convalescent patients that Gn-derived epitopes were immunodominant as compared to those from the N- and Gc-proteins . Furthermore , while the relative contribution of the N-specific response significantly declined over time , Gn-specific responses remained readily detectable ex vivo up to 13 years after the acute infection . Tetramer analysis further showed that up to 16 . 8% of all circulating CD3+CD8+ T cells were specific for the single HLA-B*3501-restricted epitope Gn465–473 years after the acute infection . Remarkably , Gn465–473–specific cells readily secreted IFN-γ , granzyme B and TNF-α but not IL-2 upon stimulation and showed a ‘revertant’ CD45RA+CD27−CD28−CCR7−CD127− effector memory phenotype , thereby resembling a phenotype seen in other latent virus infections . Most intriguingly , titers of neutralizing antibodies increased over time in 10/17 individuals months to years after the acute infection and independently of whether they were residents of endemic areas or not . Thus , our data suggest intrinsic , latent antigenic stimulation of Gn-specific T-cells . However , it remains a major task for future studies to proof this hypothesis by determination of viral antigen in convalescent patients . Furthermore , it remains to be seen whether Gn-specific T cells are critical for viral control and protective immunity . If so , Gn-derived immunodominant epitopes could be of high value for future ANDV vaccines .
The family Bunyaviridae is comprised of five genera of tri-segmented negative-stranded RNA viruses , which are responsible for a considerable burden of zoonotic disease in man . While most are tick- or mosquito-borne , members of the genus Hantavirus are transmitted from chronically- and asymptomatically-infected rodents to humans via aerosols , which may derive from urine , feces or saliva . Globally hantaviruses may cause as many as 200 , 000 cases of human disease per year . In man , two clinical conditions may arise: hemorrhagic fever with renal syndrome , caused by the Asian and European strains ( e . g . Hantaan , HTNV and Puumala , PUUV ) or hantavirus cardiopulmonary syndrome ( HCPS ) , which is caused by Sin Nombre virus ( SNV ) and Andes virus ( ANDV ) , among others in the Americas . HCPS is an emerging infectious disease in North- and South America [1]-[5] and , currently , Chile represents among the most endemic regions for HCPS with more than 580 cases since 1995 [6] . As for ANDV , transmission to man is followed by infection of lung endothelial cells and , after an incubation period of 7 to 39 days [7] , the development of a vascular leakage syndrome , eventually leading to massive pulmonary edema , shock and , in many cases , death . The high case-fatality ratio ( mean 36% ) , the absence of a proven antiviral treatment or a vaccine , their mode of transmission and their potential use as weapons for bioterrorism , have rendered HCPS-causing hantaviruses Category A pathogens within NIAID's biodefense program [8] . Importantly , ANDV is the only hantavirus for which person-to-person transmission has been repeatedly documented [9]–[11] . The hantavirus virion contains a lipid-bilayer envelope into which both constituents , the Gn and Gc antigens of the heteromeric glycoprotein , are inserted via transmembrane domains . In the viral core , there are three nucleocapids each consisting of the RNA-binding N or nucleocapsid protein in complex with one of the genomic RNAs . These mRNAs encode the RNA-dependent RNA polymerase or L protein on the large or L segment ( 2153aa ) , the Gn ( 650aa ) and Gc ( 490aa ) glycoproteins on the middle or M segment , and the N protein ( 430aa ) on the S segment [12] . Currently , there is a big discrepancy regarding the role of T cells in either pathogenesis or immunity of hantavirus infections . On one hand some studies in SNV-infected patients describe a correlation between the severity of HCPS and either the frequency of SNV-specific CD8+ T-cells [13] or the HLA-B35 haplotype [14] , suggesting a T-cell driven pathogenesis of HCPS . On the other hand , several early reports highlight the importance of lymphocytes for immunity of mice towards hantaviruses , such as HTNV [15]–[17] . Likewise , clearance of HTNV in newborn mice was dependent on TNF-α production and cytotoxic activity of specific CD8+ T cells [18] . In addition , ( HTNV- ) N-protein-specific memory T-cells conferred partial protection and cross-protection towards N-expressing vaccinia virus [19] or hantaviruses [20] in mice and Syrian hamster [21] , [22] , respectively . In line with these findings we have recently reported that clearance of ANDV-RNA from peripheral blood cells of a patient was closely related to the appearance of cytotoxic CD8+ T cells about two months after the acute infection [23] . This observation together with the finding that we were unable to detect memory T-cells in many of the survivors of ANDV-induced HCPS ( see below ) led us to the hypothesis that T-cells may be crucial for protection and immunity towards ANDV rather than the pathogenesis in ANDV-infected patients . Concisely , knowledge of targeted epitopes and functional properties of ANDV-specific T-cells in ANDV-survivors may be important for both future studies in acutely ill patients and possibly for vaccine development . Despite its importance , the knowledge of the human cellular immune response to hantaviruses is limited . To date , only few studies have assessed SNV- , HTNV- or PUUV- specific T cells [13] , [24]–[28] in rather small patient cohorts and based on individual in-silico-predicted peptides and/or T-cell lines that had been expanded in vitro . Thus , the overall magnitude of human T-cell responses in vivo and the epitope-hierarchy within the memory T-cell pool in convalescent patients remains uncertain . Also , phenotype , effector functions and longevity of specific memory T-cells in humans remain to be elucidated . Greater knowledge of these matters would likely to be of high value to potential vaccine developers . In an effort to gain insight into human cellular immunity to ANDV and to establish an immuno-hierarchy among ANDV-antigens , we carried out a study of the viral protein-specific T cell responses in 78 Chilean patients with past ANDV-infection . Our findings on the immuno-hierarchy among major structural hantavirus proteins and the frequencies as well as the functional features of CD8+ memory T cells may be of special interest for vaccine development since all attempts to induce long-lasting neutralizing humoral immunity have been unsuccessful so far [29] .
In order to quantify circulating ANDV-specific T-cells ex vivo and to determine the immunodominant epitopes of ANDV , we first challenged PBMC of 78 Chilean convalescent patients ( between 4 months and 13 . 2 years after hospitalization due to infection ) with 310 overlapping peptides ( distributed in 13 pools ) spanning the entire N- and Gn/Gc precursor proteins [30] in IFN-γ ELISPOT assays . Based on the criteria we used to score a sample as “positive” ( see Material and Methods ) , 51 ( 66% ) of the 78 patients showed significant responses against epitopes of at least one of the three viral antigens ( Fig . 1A ) . Among these patients , 33/51 ( 65% ) showed significant responses against epitopes of the N-protein , while 13/51 ( 25% ) showed Gc-specific T-cells ( Fig . 1B ) . However , 80% ( 41/51 ) of the positive patients launched significant responses towards Gn-derived epitopes . Moreover , while mean responses among the 51 individuals reached the sum of 1809 Spot Forming Units ( SFU ) /106PBMC when considering all viral antigens , Gn-specific responses accounted for more than half of the total response , at 973 SFU/106PBMC ( Fig . 1C ) as compared to 697 and 139 SFU/106 PBMC for N- and Gc-epitopes , respectively . Since all patients were BCG-vaccinated twice during childhood , we determined BCG-specific T cells ( n = 10 ) , resulting in a mean of 162 SFU/106 PBMC . Thus , Gn is the immunodominant antigen in ANDV-convalescent individuals . We next asked whether differences in the longevity of each of the specific T-cell categories could account for the relative immunodominance of Gn among ANDV antigens , e . g . , whether Gn-specific cells might persist longer than did cells responsive to the other antigens . We therefore considered the numbers of circulating N- , Gn- and Gc-specific T cells of each patient in relation to the time between the patient's hospital admission due to HCPS and the timepoint at which T cells were applied to ELISPOT assays ( Fig . 2A–C ) . Although this approach does not allow to draw conclusions on the slope of antigen-specific responses at the level of a given individual , it is possible to directly compare the different antigen-specific responses within the overall cohort over time . Similar approaches have been previously performed on a cohort of smallpox vaccinees in order to estimate longevity and half-life of cellular and humoral memory responses [31] . Interestingly , only the N-specific response ( Fig . 2A ) exhibited a negative and significantly descending linear regression slope over time ( r = −0 . 11 , p<0 . 05 ) , as when compared to Gn- and Gc-specific responses ( Fig . 2B , C ) ( r = 0 . 07 and r = 0 . 02 , respectively ) . We next segregated the 51 patients with positive T-cell responses into six groups according to the time past since hospitalization due to ANDV-infection ( <1 , 1–2 , 2–4 , 4–6 , 6–9 and >13 years , respectively ) . As can be seen in Figure 2D , up to four years after infection N-specific responses were predominant , whereas afterwards Gn-specific relative contributions to the overall T-cell response increased from approximately 40% to more than 80% ( Fig . 2D ) . Taken together , these data suggest that Gn- and Gc-specific T-cell responses are more stably maintained as compared to N-specific responses . However , in light of the limited value of the cross-sectional data available to us , future prospective studies assessing individual T-cell responses from the acute to the convalescent phase in individual patients would be needed to determine the absolute half-life of the N- , Gn- and Gc-specific T-cell responses . Subsequently , results from IFN-γ ELISPOT assays revealed that regions Gn1-230 and Gn221–450 elicited a mean of 102 and 209 SFU/106 PBMC in 34% and 38% of all responsive patients , respectively ( Fig . 3A ) . However , Gn441–650 elicited a mean response of 623 SFU/106PBMC ( range 0–5506 SFU/106PBMC ) in 24/51 ( 46% ) patients . These data clearly indicate that epitopes within the carboxyl-terminus of Gn of ANDV are responsible for the immunodominance of Gn among ANDV-convalescent individuals . In contrast to the strong Gn-specific responses , Gc641–815 and Gc806–980 elicited a response in 24% ( 77 SFU/106PBMC ) and 11% ( 25 SFU/106PBMC ) of all patients , whereas Gc971–1140 was targeted by only 4% ( 30 SFU/106PBMC ) of patients ( data not shown ) . We next stimulated cryopreserved PBMC from patients with Gn441–650-specific response using peptide pools representing Gn441–505 , Gn496–560 , Gn551–615 and Gn606–650 , respectively . This approach revealed that region Gn441–505 comprised the major epitopes of Gn441–650 ( data not shown ) . We then challenged cryopreserved PBMC of 11 patients with individual peptides spanning region Gn441–505 . As shown in Figure 3B , only three patients ( p30 , p40 , p59 ) exclusively recognized peptides Gn451–465 and Gn456–470 , whereas p17 and p57 additionally recognized Gn461–475 and Gn466–480 . By contrast , six patients ( p6 , p10 , p15 , p28 , p32 , p53 ) showed exclusive recognition of Gn461–475 and Gn466–480 . Thus , a total of five patients recognized Gn451–465/Gn456–470 , whereas eight patients recognized Gn461–475/Gn466–480 . Subsequently , four additional individuals with exclusive and significant responses towards Gn461–475/Gn466–480 were identified ( data not shown ) . Together with a previous report from our lab [23] , these epitopes are the first described within the Gn-region of hantaviruses . In addition , we determined immunodominant regions of ANDV N-protein , which included N1–70 ( 24 . 3% of responsive patients elicited a significant response ) and N121–190 and N181–250 ( 21 . 2% and 19 . 5% , respectively ) eliciting mean responses of 113–147 SFU/106PBMC ( data not shown ) . Downmapped individual epitopes within the N-protein are summarized in Table 1 . We next wondered , which state of differentiation was expressed by the IFN-γ producing memory CD8+ T cells . Intracellular cytokine staining showed that IFN-γ+CD8+CD45RO+ T-cells expressed varying levels of CD45RA but consistently expressed a CCR7−CD28−CD27− effector memory phenotype ( Fig . 3C ) . In line with this terminally differentiated phenotype , 25% of all IFN-γ+ T cells secreted granzyme B upon stimulation with their cognate peptide ( Fig . 3D ) whereas no IL-2 could be detected ( data not shown ) . In addition , up to 45% of these IFN-γ+CD8+ memory T cells co-expressed TNF-α as determined by ICS ( data not shown ) . These findings suggest that even years after acute ANDV infection ( e . g . p32 and p40 were investigated 5 . 4 and 13 . 2 years after hospitalization , respectively ) , high frequencies of cytolytic memory CD8+ T-cells are maintained in the periphery . As both the Gn451–465/Gn456–470 and Gn461–475/Gn466–480 epitopes share 10 amino acids in sequence within the pairs , we reasoned that one CD8+ T-cell epitope may be located within each overlapping sequence , respectively ( e . g . Gn456–465 and Gn466–475 ) . In support of this hypothesis , the analysis of HLA-A , -B , -DR and -DQ alleles revealed that 5/5 patients with response towards Gn451–465/Gn456–470 exclusively shared the HLA-A*24 allele ( data not shown ) , whereas the HLA-B*35 allele was the only allele shared by all 12 patients recognizing Gn461–475/Gn466–480 ( Fig . 4A ) . These data suggest the existence of two separate but neighboring CD8 T-cell epitopes in the carboxyterminal region of Gn that are restricted by HLA-A*24 and HLA-B*35 , respectively . Indeed , as shown in Figure 4B , a significant response in a Gn461–475–specific T-cell line could only be detected when the HLA-B*35 allele was present on heterologous APCs ( B-LCL ) . It was previously suggested that severe HCPS due to SNV is associated with the HLA-B*35 allele [14] and with CD8 T-cell responses restricted to it [13] . We therefore were interested in the relation of memory T-cell responses and outcome of their ANDV infection in HLA-B*35-positive and negative patients ( Fig . S1 ) . Among all 78 patients , that is patients with ( n = 51 ) and without ( n = 27 ) significant memory T-cell responses , no differences in overall T-cell responses could be observed when comparing HLA-B*35-negative patients with mild or severe HCPS ( Fig . S1A ) . By contrast , we found an about 3-fold higher overall T-cell response in HLA-B*35-positive patients with mild HCPS as compared to both HLA-B*35-positive patients with a history of severe disease and either group of HLA-B*35-negative patients ( Fig . S1B ) . Likewise , 10/12 ( 83% ) HLA-B*35-positive patients with significant responses to Gn461-475 had a history of mild HCPS ( Fig . S1C ) . Thus , these data suggest that HLA-B*35-restricted memory T-cell responses are related to mild rather than to severe disease outcome . We next sought to determine the optimal epitope of Gn461–475/Gn466–480 ( Fig . 4C ) . Because we consistently observed stronger immune responses towards Gn461–475 ( SLFSLMPDVAHSLAV ) than towards Gn466–480 ( MPDVAHSLAVELCVP ) , we reasoned that Leucine at position 465 may increase either the binding affinity or the TCR-recognition of the overlapping sequence Gn466–475 ( MPDVAHSLAV ) . We therefore decided to generate Gn461–475–specific T-cell lines from HLA-B*3501 individuals and then challenged these cells with cleaved peptides of Gn465–475 ( LMPDVAHSLAV ) . As shown in Figure 4C , cleavage of the carboxyterminal Leucine at position 473 led to a complete loss of epitope recognition . Similarly , elimination of the aminoterminal Methionine at position 466 was critical for epitope recognition . Most interestingly , virtually identical results were obtained when cells from HLA*B3501 , HLA-B*3502 and HLA-B*3505 individuals were challenged with cleaved peptides , indicating that the Gn466–473 epitope is equally immunogenic among different HLA-B*35 subtypes . We next were interested in comparing the phenotype of Gn465–473 restricted T cells and with other HLA-B*3501-restricted virus-specific T cells in seven HLA-B*3501 positive ANDV-convalescent patients ( Fig . 5A , B ) . The mean abundance of Gn465–473 specific T-cells was higher than those specific for N131–139 and Gc664–673 –epitopes , described by Kilpatrick et al . [13] , the latent EBV-epitope EBNA3A458–466 [32] , the Influenza A NP418–426 epitope [33] , or the Rv2903c201–209 epitope of Mycobacterium tuberculosis , known to be recognized by BCG-vaccinated individuals [34] . In the seven HLA-B*3501-positive individuals we had detected between 0 and 4394 SFU/106 PBMC ( 0%–0 . 0044% ) following stimulation with peptide Gn461–475 in IFN-γ ELISPOT assays . When normalizing the results by the percentages of CD3+CD8+ cells in these individuals ( range 9 . 9–33 . 9% of PBMC , mean 19 . 7% ) , one would have expected between 0% ( p45 ) and 0 . 036% ( p17 ) of CD3+CD8+ cells being tetramer positive . However , we found 0 . 3% and 5 . 9% of all CD3+CD8+ T cells being positive for Gn465–473:HLA-B*3501 tetramer complexes , respectively . In addition , the highest frequencies for tetramer-positive cells were found in patient 10 ( 16 . 8% of CD3+CD8+ cells ) , whereas only 0 . 0111% of his CD3+CD8+ cells produced IFN-γ in ELISPOT assays . This discrepancy between both detection methods is in line with previous reports [35] . We next determined the state of differentiation of Gn465–473-specific T-cells , where a clear dichotomy was observed ( Fig . 6A , B ) . In patients with positive responses towards Gn461–475 in IFN-γ ELISPOT assays ( IFN-γ ++ ) , Gn465–473 specific T cells were mostly CD45RA+CCR7− and significantly more of a differentiated CD28−CD27− phenotype as compared to IFN-γ− samples . In addition , we found significant differences with regard to the IL-7Rα ( CD127 ) , which is crucially involved in maintenance of memory T-cells in the periphery in the absence of cognate antigen [36] . Patients with IFN-γ+ ELISPOT results showed mainly CD127− Gn465−473 T cells , whereas T cells of IFN-γ− patients were mostly CD127+ ( Fig . 6A , C ) . Thus , Gn465−473-specific CD8+ T cells showed a phenotype that is clearly distinct relative to that described for other self-limited diseases such as those caused by influenza A and respiratory syncytial virus but more resembled the pattern associated with latent infections , such as past exposure to CMV [37] . Because a CD28−CD27−CD127− phenotype was previously described to be a result of ongoing antigen-stimulation , as found in latent CMV infection , we next determined the expression of activation markers , such as of KLRG-1 , CD69 , CD38 and CD25 on Gn465–473 and Influenza A-specific T cells within the seven HLA-B*3501+ ANDV-convalescent ( Fig . 6D ) . No significant differences could be observed between IFN-γ+ and IFN-γ− Gn465–473-specific populations or between Gn465–473- and Influenza A NP418–426-specific T-cells . In a next step we assessed whether re-exposure to viral antigens could have led to a boost in the donor's immune response . We therefore compared memory T-cell responses in patients who got infected during recreation ( R-patients ) with those of residents in endemic areas ( E-patients ) ( Fig . 7A ) . No significant differences were observed between the two groups in those responses , although endemic patients revealed about double as many Gn-specific memory T-cells than recreational patients ( mean 765 vs 361 SFU/106 PBMC ) for unclear reasons . However , these results are not in line with the hypothesis of repeated viral exposure in patients who reside in endemic areas , since N- and Gc-specific responses were virtually identical in both groups . In addition , we identified seven individuals , which had been infected during recreation ( R-patients , Fig . 7B , C ) and ten individuals residing in endemic areas ( E-patients , Fig . 7D , E ) , for all of which two prospective serum samples were available . The time period between sample 1 and sample 2 was 0 . 3–6 . 9 years and 1 . 2–4 . 1 years in R- and E-patients , respectively . Surprisingly , in R- and E-patients anti-N titers raised four- to 64-fold between samples 1 and 2 in 4/7 and 5/10 patients , respectively . Most intriguingly , however , also neutralizing antibody ( NAb ) titers rose two- to eight-fold in 4/7 and 6/10 of R- and E-patients , respectively . Importantly , NAb titers , measured by a blinded worker , increased two- to four-fold in patients R1 , R3 , R5 , E1 and E4 between sample timepoint 1 and 2 , although in all cases sample 1 was taken months to years after the acute phase . Taken together , these results suggest that re-exposure to extrinsic , environmental virus is not responsible for the observed rise in NAb titers or high frequencies of memory T cells . Finally , we sought to prospectively study Gn-specific T-cells in three patients . When Gn465–473-specific T cells were phenotyped over a time period of two years ( Fig . 8A–C ) , no dynamic changes of the CD27− population could be observed , indicating that differentiated Gn465–473-specific T cells are able to stably persist at high frequencies without the need for B7:CD28- or CD70:CD27-mediated survival signals . However , in patients E9 and E2 ( Fig . 8A , C , respectively ) , Gn465–473-specific T cells actually increased over time , paralleling the two- to eight-fold increase in NAb titers observed in these two individuals ( Fig . 7E ) .
Infection with ANDV is the predominant cause for HCPS in South America . Case-fatality rates of currently 36% , person-to-person transmission and the absence of a proven effective antiviral treatment urge the development of a vaccine . Although the protective potential of neutralizing antibodies against the hantavirus surface glycoproteins Gn and Gc , but not the N-protein , was established in vitro [38]–[40] and in animal models [41]–[43] , efforts to induce long-lasting neutralizing antibodies in human volunteers have been unsuccessful so far [29] , [44] or remain to be proven effective and long lasting [45] . On the other hand , several early reports highlight the importance of lymphocytes for immunity of mice towards hantaviruses , such as HTNV [15]–[17] . Likewise , appearance of virus-specific CD8+ T cells with cytotoxic activity and the ability to produce IFN-γ and TNF-α was associated with clearance of HTNV in newborn mice . In contrast , HTNV infection was not cleared when TNF-α production and cytotoxic activity of specific CD8+ T cells were impaired [18] . In another report , ( HTNV- ) N-protein-specific CD8+ memory T-cells , induced by a DNA vaccine , conferred partial protection against re-challenge with a vaccinia virus expressing the N-protein [19] . One study in mice showed that N-protein specific T cells rather than antibodies mediated protection and cross-protection upon re-challenge with homologous and heterologous hantaviruses [20] . Finally , ANDV infection of Syrian hamsters - the sole animal model for human HCPS–could be prevented for at least 10 months by previous vaccination with ANDV N-protein [21] , again indicating that protection can be achieved independently of neutralizing , Gn/Gc-specific antibodies . Most recently Safronetz et al . confirmed these findings in Syrian hamsters vaccinated with Gn-protein expressing Adenovirus vectors . Interestingly , these animals were protected from lethal ANDV infection independently of neutralizing antibodies and showed no or very low levels of ANDV-RNA up to 9 days after ANDV infection [22] . As for ANDV-infection in man , we recently showed that clearance of ANDV-RNA from peripheral blood cells was closely related to the appearance of cytotoxic CD8+ T cells , but not NAb , in a patient about two months after the acute infection [23] . Taken together , these reports suggest that cytotoxic T cells are crucially involved in clearance and protection from hantaviruses . Conversely , establishment of hantavirus-specific cytotoxic memory CD8+ T cells prior to infection , e . g . by a vaccine , may provide protective , albeit not sterilizing , immunity to the host . However , limited information on human cellular immunity to hantaviruses is available and , to date , only one study addresses ANDV-specific T-cell responses [23] . Using a panel of 310 overlapping peptides spanning the entire N- , Gn- and Gc-protein of ANDV allowed us to study most , existing T-cell epitopes in 78 convalescent survivors of ANDV infection in a non-HLA-restricted manner . In contrast to other reports with a similar approach , the majority of responses were specific for Gn- but not N-protein-derived epitopes . Thus , our results in ANDV-infected patients seem to contradict the current dogma of N-protein being the principal T-cell immunogenic hantavirus antigen [19] , [20] , [22] , [24] , [25] , [27] , [46]-[48] . However , in previous studies , epitope-specific T cells were detected either by in vitro expansion prior to testing or using individual peptides or tetramer complexes for ex vivo detection in a small and HLA-selected patient populations [13] , [24]–[28] . Thus , differences in the experimental design , rather than its elevated immunogenicity , may explain , why we found Gn being the immunodominant antigen of ANDV , whereas no single Gn-epitope had been described for other hantaviruses . In fact , 92–96% of the amnio acid sequence of the two Gn-epitopes described herein , are conserved within the PUUV and SNV sequence , respectively . An alternative explanation may derive from the differences between our study and other studies in the timing of T-cell testing after the acute phase or differences in infection kinetics between the different hantaviruses . Specifically , as can be seen in Figure 2D , N-derived epitopes seemed to be relatively predominant up to four years after the acute infection , whereas Gn-derived epitopes were predominant in patients with a longer convalescence phase . In addition , the kinetics of NAb titers in our patients suggest that viral antigen may be present for months or years after the acute infection , a phenomenon which has not been described for other hantaviruses . Thus , differences in epitope avidity and/or precursor expansion over time may have contributed to the relative predominance of Gn-specific T-cells in our study . Of note , 80% of all patients with detectable T-cell responses recognized Gn-derived epitopes ( Fig . 1B ) . To our surprise , however , responses against Gn were not broad but rather focused to the carboxyterminus of Gn , namely the region of aa 451–480 . Interestingly , the cytoplasmic tail of Gn has been shown to contain important virulence factors as it suppresses type 1-interferon responses in infected cells [49] , [50] . On the other hand , the carboxyterminal 142 residues of pathogenic ( namely of ANDV and HTNV ) , but not non-pathogenic hantaviruses , prone the C-terminal tail of Gn towards degradation by the proteasome , which then leads to the presentation of epitopes by MHC I molecules to CD8+ T-cells [51] . This mechanism could explain the relative immunodominance of Gn-derived epitopes seen in our study and also may represent a virulence factor of ANDV suggesting that T cells are causative for HCPS . Nonetheless , an early and vigorous cytotoxic T-cell response towards epitopes of the C-terminal Gn may also be able to restrict the virulence of ANDV infection . Indeed , among HLA-B*35-positive patients mild disease outcome seemed to be associated with stronger responses towards the Gn-carboxyterminus than in patients with severe HCPS ( Fig . S1B , C ) . In line with this finding , a recent study in 87 Chilean ANDV-infected patients found that the HLA-B*35 allele was the most frequent allele among patients with mild disease and almost twice as frequent as in patients with severe disease [52] . Although these data seem to contradict previous reports describing both the expression of the HLA-B*35-allele [14] as well as HLA-B*35-restricted T-cell responses [13] as risk factors for severe HCPS by SNV , it remains speculative whether our results indicate an pivotal role for T cells in disease outcome . The size of the memory T-cell pool is only indirectly linked to the effector T-cell response by the original burst size [53] and therefore may not reflect the size and composition of the effector T-cell pool during the acute phase . Moreover , 33% of HLA-B*35-negative patients and 48% of patients without any detectable memory T-cell responses had a history of mild HCPS ( data not shown ) . Likewise , 52% of patients with severe HCPS did not show memory T-cell responses ( data not shown ) . Both seem to argue against an exclusive role of T cells for disease outcome . In addition , other hand , we also showed a discrepancy between ELISPOT and tetramer-derived T-cell frequencies ( see above ) , which indicates the existence of IFN-γ-negative ANDV-specific T-cells . In fact , three of the seven studied HLA-B*3501 positive donors did not show significant responses in initial IFN-γ ELISPOT assays ( Fig . 1 ) but showed substantial numbers of tetramer-positive CD3+CD8+ T cells . This is in line with previous reports comparing determination of T-cell frequencies by ELISPOT and tetramer analysis [35] . In this report tetramer analysis revealed on average ten-fold higher frequencies than IFN-γ ELISPOT assays of T cells specific for a HLA-A2-restricted HIV Gag-derived peptide . This report as well as our data suggests that the vast majority of virus-specific T cells may not readily secrete IFN-γ when stimulated by peptides in ELISPOT assays . It is also possible that these T cells were functional during the acute phase but not during convalescence . Alternatively , differences in the infectious dose ( e . g . low versus high infectious dose ) or the kinetics and doses of the evolving neutralizing antibodies may interfere with the functional quality of the memory T-cell pool . Taken together , additional studies with HLA-tetramers in acutely ANDV-infected patients will be necessary to better understand the role of HLA-B*35 and T-cell kinetics for ANDV disease outcome . Another interesting aspect of our study concerns the longevity of the memory T cells in association with their highly differentiated phenotype . After more than 13 years after infection we still detected 564–2152 SFU/106 PBMC by ELISPOT , most of which were directed towards Gn-derived epitopes . With regards to the longevity of CD8 memory T cells , our results nicely confirm a previous study by Van Epps et al . in PUUV-convalescent patients , in which up to 100–300 SFU/106 PBMC of N-epitope-specific CD8 T cells were found in three patients up to 15 years after acute infection [24] , [25] , [27] . However , by tetramer analyses we detected still-higher frequencies with up to 16 . 8% of all CD3+CD8+ T cells proving positive for the single epitope Gn465–473 while displaying a late effector memory phenotype ( CD127−CD28−CD27−CCR7− ) . This phenotype was in line with the cells' ability to readily secrete granzyme B and TNF-α without IL-2 . Surprisingly , we also found that up to 5% of tetramer+ cells , which , again in accordance with their CD28+CD27+CCR7+CD127+ phenotype , did not exert any immediate effector functions , such as IFN-γ secretion upon stimulation with their cognate epitope . While these data again show that screening by ELISPOT underestimates the real proportion of Gn-specific T cells [35] , it is also tempting to speculate , albeit virtually impossible to prove , that patients with either of the observed phenotypes differ at their stage of immunity towards a possible ANDV re-challenge . In absence of antigenic stimulation as well as of autocrine or paracrine IL-2 and co-stimulation via CD28:B7 and CD27:CD70 interaction , late effector memory T-cells are heavily prone to apoptosis . However , we herein were able to show in three HLA-B*3501+ patients that peripheral Gn465–473-specifc T cells were maintained in the periphery for at least two years , despite a consistent CD27− phenotype , thereby lacking the receptor for crucial anti-apoptotic signals provided by CD70 . While the former perception was that senescent end-differentiated CD28−CD27− T cells were unable to divide , recent evidence suggests , that highly differentiated granzymeB+CD8+ memory T cells are actually dividing upon stimulation equally well as naïve CD8+ T cells [54] . In man , the CD45RA+CD28−CD27−CCR7− late effector memory phenotype has been mainly described in patients with latent CMV infection [37] , [55] . In this model repetitive or latent antigen stimulation is supposed to drive CD8+ memory differentiation and/or the recruitment of new memory T-cells . However human hantavirus infections are not known to cause latent or chronic infections and we were not able to detect viral RNA in plasma or peripheral blood cells of patients with sustained and high T-cell responses ( data not shown ) . Also , we failed to detect a fingerprint of recent antigenic stimulation of tetramer-positive cells through assessment of the expression of additional activation markers , including CD69 , CD38 and CD25 . In addition , expression of IL-7Rα ( CD127 ) was described to be a critical factor for long-term survival of CD8+ memory T cells in absence of their cognate antigen . Since CD127 is usually downregulated upon antigen exposure and rapidly re-expressed after antigen clearance , it is consistent that mainly virus-specific CD127+CD8+ memory T-cells are found in studies on Influenza- , respiratory syncytial virus- and HBV- specific T-cells . By contrast , in persistently HIV- , CMV- or EBV-infected individuals T cells are maintained despite their lack of CD127 expression [56] , [57] . Finally , KLRG1 is mainly expressed on antigen-experienced T-cells with immediate effector functions [55] . Thus , considering ANDV a self-limiting transient infection in man , a CD127+KLRG1− phenotype would have been expected years after the infection . However , IFN-γ + , but not IFN-γ − , Gn465–473-specific T cells expressed substantially less CD127 than their Influenza A virus ( NP418–426 ) -specific counterparts , whereas no clear pattern could be observed regarding the KLRG1 expression . However , although no difference could be observed between ANDV ( Gn465–473 ) and Influenza ( NP418–426 ) –specific T-cells with regards to CD25 , CD38 and CD69 expression , the lack of CD127 suggests persistent antigenic stimulation in individuals with IFN-γ+ T cells . When employing BLAST , we were not able to identify other organisms that share the amino acid sequence of Gn465–473 , thereby making cross-reactivity an unlikely explanation . Consistently , no further serology testing was performed in convalescent patients . We next hypothesized that residents of endemic areas might have received intermittent antigen-boosters due to viral re-exposure and therefore should show somewhat higher T-cell responses to all viral antigens . However , we did not find significant differences in ANDV-specific T cell numbers when comparing patients who reside in endemic areas and those , which got during recreation got infected in an endemic region ( Fig . 7A ) , while residing in non-endemic areas . Moreover , in the majority of prospective serum samples of ten patients from endemic regions and of seven patients from non-endemic regions , we surprisingly found an increase in both anti-N and NAb titers despite the fact that the second sample was taken years after the first samples in most cases . The fact that this was also observed in patients who never had returned to endemic regions since their primary ANDV infection , suggests that re-exposure to extrinsic ( environmental ) virus does not account for high antibody titers and , conversely , not for high ANDV-specific T-cell frequencies . Regarding the increase in NAb titers between sample 1 and 2 , we cannot exclude that NAb titers continued to rise after sample 1 was drawn during or shortly after the acute phase . Therefore , it is possible that in these patients ( Fig . 7 ) titers of sample 2 in fact were identical or even lower than the maximum titer achieved during the acute phase . However , NAb titers of 13/17 individuals were still relatively high ( ≥1∶400 ) at the timepoint of sample 2 , that is 1 . 2–11 . 3 years after the acute infection . This argues for continuous antigen-exposure in both E- and R-patients , since NAb titers , in contrast to non-neutralizing antibodies , strictly depend on the presence of their cognate antigen . Specifically , in absence of antigen , murine NAb titers fall below the detection limit after 100–200 days [58] . Most importantly , however , we also found that NAb titers increased two- to four-fold in five patients ( R1 , R3 , R5 , E1 and E4 , Fig . 7C , E ) in which sample 1 was taken months to years after the acute phase . Since not only maintenance [58] but also kinetics of NAb titers heavily depend on the presence of viral surface antigens [59] , these results support the hypothesis that re-exposure to viral surface ( Gn/Gc ) antigen is responsible for high and rising NAb titers in both R- and E-patients . Due to the fundamental differences between R- and E-patients in their risk for re-exposure to extrinsic virus , this is turn suggests that intrinsic viral antigen is responsible for the relative “immune-inflation” and the terminal differentiation of Gn-specific CD8+ T-cells . Thus , it may be that intermittent release of low doses of viral antigen from intrinsic virus ( e . g . that never completely cleared from tissue reservoir ( s ) ) is sufficient to maintain and boost of NAb titers and T-cell frequencies , whereas changes in activation marker expression on ANDV-specific T-cells are too short-lived ( with the exception of low CD127 expression ) to be consistently different ( e . g . KLRG1 ) from that of Influenza A virus-specific T cells . However , as long as viral antigen or genome cannot be detected in convalescent patients as those described herein , the concept of latent or persistent ANDV infection in convalescent patients remains speculative . Future studies should therefore focus on antigen detection in tissues ( e . g . surgical or post-mortem specimen ) from solid ( e . g . lung , kidney ) and immuno-privileged ( e . g . brain ) organs . Our data suggest that long-lived effector memory T cells can be maintained at high numbers in the periphery over years independently of IL-7 . Notably , our findings resemble those found in murine Sendai and Influenza A virus infections , where epitope-specific T-cell clonal expansions occurred in absence of antigen throughout the CD8 memory pool [60] . As in our study , but in contrast to models of persistent infection , clonally expanded effector memory T-cells in these studies retained potent functionality despite their highly differentiated phenotype . Although we could not formally show clonal expansion for all our patients ( with the exception of two individuals , see Fig . 8A , C ) due to our non-prospective study design , similar underlying , but yet undefined , mechanisms may explain our findings in human ANDV infection . The induction of a highly differentiated , resting e . g . Gn465–473–specific , memory T-cell subset might be of major interest in the context of vaccine development for several reasons . First , years after infection high numbers of these CCR7− Gn465–473–specific cells remain available for immuno-vigilance in the periphery . Second , as shown , this subset possesses the ability to readily secrete antiviral ( e . g . IFN-γ , TNF-α ) as well as lytic ( granzyme B ) effector molecules . Third , although most human studies seem to focus on epitopes restricted to HLA-A*02 because of its wide distribution among the Caucasian population ( about 25% , [61] ) , it should be noted that within the Amerindian population , the frequency of HLA-B*35 positive individuals is about 70% higher than in Caucasians [61] . In fact , 25% ( range 22–30% ) of the inhabitants in ANDV endemic regions in Southern Chile express the HLA-B*35 and/or the HLA-A*02 allele , respectively [52] . Thus , for this population , HLA-B*35-restricted epitopes , like Gn465–473 , might be of similar impact as HLA-A*02-restricted epitopes . However , it first has to be established in future studies ( e . g . Syrian hamster models ) whether and to which extend Gn-derived T-cell epitopes may contribute to protective immunity . Furthermore , although hantaviruses are not known to mutate , additional epitopes have to be identified in future studies in order to prevent failure of a T-cell based vaccine due to mutations within the Gn465–473 epitope . Taken together , our results suggest that infection with ANDV may lead to a strong highly differentiated effector memory response . The findings concerning the predominant immunogenicity of ANDV-Gn protein may have implications for the understanding of immunity not only to ANDV , but also to other hantaviruses .
A total of 78 patients were enrolled between 4 months and 13 . 2 years after hospitalization due to either mild or moderate/severe HCPS . All patients had a previous confirmed hantavirus diagnosis done in Chilean reference laboratories by IgG serology to SNV and ANDV antigens by enzyme-linked immunosorbent assay ( ELISA ) , as previously described [62]–[64] . Mild HCPS was defined by the sole support of the patient by symptomatic therapy , including respiratory support by an oxygen mask . On the other hand , ANDV-infected patients who required intensive care by mechanic ventilation and/or anti-shock treatment with vasoactive drugs were defined as moderate/severe HCPS . All patients included were Chilean citizens and volunteered to participate without receiving monetary incentive . Prior to enrollment all patients enrolled signed informed consent , which was previously informed by IRB committees of Clínica Alemana de Santiago , the Chilean Ministry of Health and regional IRB committees . Before enrollment , patients were extensively informed about the intention of the study by the local study nurse . Upon enrollment patients did not suffer from any signs of active disease and were only enrolled if considered healthy donors . Samples consisted in 45 cc of peripheral venous blood , using tubes containing Sodium Heparin ( BD vacutainer ) . Samples were shipped within 24 hours to our laboratory and were processed immediately upon receipt . PBMC were isolated by Ficoll-Hypague gradient and fresh PBMC were applied to ELISPOT assays . PBMC , which were not used immediately were cryopreserved in liquid nitrogen . 96-well filterplates ( Millipore ) were coated with 5 µg/ml of anti-hIFN-γ ( Endogen , clone M700A ) or 15 µg/ml anti-granzyme B ( mabtech , clone GB10 ) at 4°C overnight one day prior to the assay . For granzyme B assays , prior to coating membranes were activated by incubation of the wells with 15 µl/well of 35% Ethanol for 1 minute . After washing and blocking of the plate , fresh or cryopreserved PBMC or T-cell lines were applied and incubated for 20 hours in an incubator ( Nuraire ) at 37°C and 5% CO2 in the presence of 310 overlapping 15mer peptides ( Mimotopes , Australia ) organized in 13 pools of 12 to 44 peptides ( final concentration 1 µg/ml , each ) of continuous sequence spanning the entire genome of the N and GPC protein of the Chilean ANDV [30] . For mapping experiments , cells were incubated with 10 µg/ml of each individual peptide . As negative controls , corresponding dilutions of DMSO ( Sigma ) were used , whereas a 1∶100 dilution of PHA ( M form , Invitrogen ) was used as positive control . After the incubation period , plates were washed and incubated with biotinylated secondary antibodies ( IFN-γ: clone M701B , granzyme B: GB11 ) according to the manufacturer's manual . After incubation with Streptavidine-Alkaline phospahtase ( Vector , at 1∶1000 for 2 hours ) , plates were incubated with NCIP/BPT substrate ( BioRad ) , and analyzed using the ELI . Scan ( A . EL . VIS GmbH ) analyzing unit . Results were expressed as Spot Forming Units ( SFU ) , representing the numeric difference between specific spots and the spots in the negative control ( DMSO ) . All FRNT studies were carried out in an approved ( C20041018-0267 ) biosafety level 3 laboratory . Plasma samples from the patient were serially diluted in fourfold increments , mixed with equal volumes of approximately 60 focus forming units ( f . f . u . ) of a human Chilean virus isolate [65] before incubation on Vero E6 cells , processed and analyzed as described before [66] . The neutralization activity of an antibody was expressed as the highest plasma dilution capable of reducing the number of foci by at least 80% . Cryopreserved PBMC of four different patients were challenged in vitro for 1 . 5 hours by a previously determined individual immunogenic peptide ( 10 µg/ml ) , the corresponding DMSO dilution or PMA/ionomycin ( 500/50 ng/ml ) in the presence of 1 µg/ml anti-CD49d ( clone 9F10 ) and anti-CD28 ( clone CD28 . 2 ) ( both BD Pharmingen ) , respectively , and cultured for 4 . 5 additional hours in the presence of GolgiStop/monensin . Finally , intracellular cytokine staining was performed by fixation , permeabilization of cells and subsequent staining for surface markers and intracellular IFN-γ and TNF-α according to the manufacturer's protocol ( BD , California ) . 2×105 PBMC/well were stimulated with 10 µg/ml of the Gn461–475 in the presence of 10 ng/ml IL-7 and 300 pg/ml IL-12 ( R&D systems ) . On day 2 after setup and every 3–4 days 10 U/ml and 150 µg/ml IL-15 were added to the culture . On day 7 and 14 , T-cells were re-stimulated using irradiated ( 30Gy ) PBMC or irradiated autologous ( 100Gy ) B-LCL . On day 21 T-cells were assayed in IFN-γ ELISPOT assays using truncated peptides as indicated . Was performed on a 4-colour FACSCalibur ( Becton Dickinson ) or CyAn ( Dako ) and using either CellQuest ( Becton Dickinson® ) or Summit 4 . 0 ( Dako ) analyzing software . We used the following antibodies ( all BD Pharmingen ) : CD3-FITC ( UCHT1 ) , CD4-FITC ( SK3 ) , CD45RA-FITC ( HI100 ) , CD27-FITC ( M-T271 ) , TNF-α-FITC ( Mab11 ) , CD45RO-PE ( UCHL1 ) , CD28-PE ( L293 ) , CD8-PercP ( SK1 ) , IFN-γ-APC ( B27 ) , CD3-APC ( UCHT1 ) . FITC- , PE- and APC- ( MOPC-21 ) as well as PercP- ( X40 ) conjugated mouse IgG1κ were used as isotype controls . Anti-KLRG-1-Alexa488 was kindly provided by Prof H . P . Pircher ( University of Freiburg , Germany ) . Tetramer complexes were custom-synthesized by the NIAID tetramer facility ( Gaithersburg , MD ) according to the published protocol ( http://research . yerkes . emory . edu/tetramer_core/protocol . html ) , and Gn465–473:tetramers were either APC- or PE-labeled . All other tetramer complexes were APC-labeled . Table 2 shows a summary of HLA-B*3501 tetramer complexes used in the present study . The patients were genotyped for the HLA loci A , B , DRB1 and DQB1 , using the SSP PCR ( Sequence Specific Primer–Polymerase Chain Reaction ) technique . Low and high resolution SSP kits from Dynal ( Oslo , Norway ) and Invitrogen Corporation ( USA ) were used . For analysis of ELISPOT results for each patient an unpaired Student's t-test was applied in order to calculate significant results as compared to the internal negative ( DMSO ) control . To be evaluated as positive a sample ( that is response to a individual or a pool of up to 40 ANDV-derived 15mer peptides ) had to fulfill three criteria: ( i ) a significant difference ( p<0 . 05 ) between sample and negative control , ( ii ) specific SFU had to be superior of 50/106 PBMC , ( iii ) value had to be above a cut-off , defined as a mean + 2xSD , which was previously established in 20 healthy controls for each peptide pool . For differences in frequencies of tetramer cell populations an unpaired Student's t-test was applied . We further studied the association of time since ANDV infection until blood sampling with the T-cell responses against N , Gn and Gc as determined by IFN-γ ELISPOT at the time point of blood sampling . A decreasing response with increasing time since infection corresponds to a loss of T-cell memory in time and is reflected by a negative correlation . We reject the null-hypothesis of no association between time and the IFN-γ-ELISPOT response at the two-sided alpha level of 0 . 05 . Due to skewness of the response data , we log-transformed ELISPOT responses and fitted our linear regression models on the log-transformed responses and assessed the model assumptions by inspecting residual values against time .
|
In man , hantavirus cardiopulmonary syndrome ( HCPS ) caused by Andes Virus ( ANDV ) is endemic in the Southern cone of Chile and Argentina but cases of HCPS are being increasingly reported all over South America since 1995 . HCPS is characterized by fulminant pulmonary edema which progresses to shock and death in about 36% of patients with HCPS . Nevertheless , to date , neither antiviral treatments nor vaccines inducing neutralizing antibodies ( NAb ) have proven effective against HCPS-causing hantaviruses . We set out for the first study on human cellular immunity towards ANDV in 78 convalescent survivors of ANDV infection . We found that Gn-specific responses were predominant as compared to N- and Gc-specific responses , even up to 13 years after the infection . Surprisingly , most of the Gn-specific responses were restricted to two neighboring epitopes within the Gn carboxyterminus . Interestingly , among HLA-B*3501+ patients , Gn465−473-specific CD8+ T-cells showed highly differentiated but resting phenotype and functions . It remains to be seen in future studies whether the immunodominace of Gn-specific T-cells is crucial for protective immunity . Most intriguingly , titers of neutralizing antibodies increased in 10/17 individuals months to years after the acute infection and independently of whether they were residents of endemic areas or not . Thus , our data suggest viral persistence or latency in part of ANDV-convalescent patients . However , it remains a major task for future studies to proof the concept of latent/persistent human ANDV infection by the determination of viral antigen in convalescent patients .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"immunology/immune",
"response",
"virology/emerging",
"viral",
"diseases",
"infectious",
"diseases/viral",
"infections",
"infectious",
"diseases/respiratory",
"infections",
"immunology/immunity",
"to",
"infections",
"virology/host",
"antiviral",
"responses"
] |
2010
|
Highly Differentiated, Resting Gn-Specific Memory CD8+ T Cells Persist Years after Infection by Andes Hantavirus
|
The accurate identification of the route of transmission taken by an infectious agent through a host population is critical to understanding its epidemiology and informing measures for its control . However , reconstruction of transmission routes during an epidemic is often an underdetermined problem: data about the location and timings of infections can be incomplete , inaccurate , and compatible with a large number of different transmission scenarios . For fast-evolving pathogens like RNA viruses , inference can be strengthened by using genetic data , nowadays easily and affordably generated . However , significant statistical challenges remain to be overcome in the full integration of these different data types if transmission trees are to be reliably estimated . We present here a framework leading to a bayesian inference scheme that combines genetic and epidemiological data , able to reconstruct most likely transmission patterns and infection dates . After testing our approach with simulated data , we apply the method to two UK epidemics of Foot-and-Mouth Disease Virus ( FMDV ) : the 2007 outbreak , and a subset of the large 2001 epidemic . In the first case , we are able to confirm the role of a specific premise as the link between the two phases of the epidemics , while transmissions more densely clustered in space and time remain harder to resolve . When we consider data collected from the 2001 epidemic during a time of national emergency , our inference scheme robustly infers transmission chains , and uncovers the presence of undetected premises , thus providing a useful tool for epidemiological studies in real time . The generation of genetic data is becoming routine in epidemiological investigations , but the development of analytical tools maximizing the value of these data remains a priority . Our method , while applied here in the context of FMDV , is general and with slight modification can be used in any situation where both spatiotemporal and genetic data are available .
Predicting the most likely transmission routes of a pathogen through a population during an epidemic outbreak provides valuable information , which can be used to inform intervention strategies and design control policies [1] , [2] . In principle , studying transmission routes during past epidemics is likely to be broadly informative of how the same pathogens spread through similar populations in future outbreaks . Estimating a set of connected transmission routes from a single case is synonymous with estimating the transmission tree corresponding to the outbreak . Uncovering the transmission routes between individual hosts or other relevant infectious units ( for example farms or premises ) can provide valuable epidemiological information , such as the factors associated with source and target individuals , dissemination kernels and transmission modes . Unfortunately , reconstructing these transmission trees with available data can be an exceptionally hard task , as the problem is typically underdetermined: the precise number of cases is often unknown , and dates and times of infections are rarely known with precision , making it difficult to distinguish between a large number of alternative scenarios [3] . With knowledge of location and timing of disease incidence it is possible to sample transmission trees that are consistent with the space-time data , and when these samples of trees share emergent statistical or structural properties , they can lead to epidemiological insights . For example , Haydon et al . [4] generated transmission trees corresponding to the 2001 Foot-and-Mouth Disease Virus ( FMDV ) epidemics in the UK , and used these trees to estimate the reproductive number during different weeks of the epidemic . These trees could be pruned to investigate the consequences of different or earlier interventions on the final size of the epidemics . However , the data were consistent with very large numbers of different trees and so the approach was not suited to identifying with confidence “who infected who” . For pathogens with high mutation rates that fix mutations across their genome during the course of a single outbreak , genetic data can provide critical additional information regarding the relationships between isolates . The last few years have witnessed a revolution in our ability to generate genomic data relatively cheaply and in an automatised fashion [5] . Pathogen genome sequences collected during epidemics , if sufficiently diverse , can then be used to discriminate between alternative transmission routes . Several attempts to reconstruct transmission pathways have tried to combine genetic and other epidemiological data , many by adding spatial or temporal information to the process of phylogenetic reconstruction [6]–[11] . However , Jombart et al . point out that a “phylogenetic” approach attempts to infer hypothetical common ancestors among the sampled genomes , and may not be appropriate for a set of genomes containing both ancestors and their descendants [12] . Cottam et al . [13] identified a large set of transmission trees that were consistent with available genetic data , and ranked the likelihood of these trees using data on their relative timings , to find the most likely transmission tree . Ypma et al . [14] moved this approach forward by constructing an inference scheme that uses spatial , temporal and genetic data simultaneously , but assumed these data are independent of each other . Genetic and epidemiological data are evidently correlated , and a rigorous inference scheme should estimate the likelihood of a transmission tree accounting for these correlations . In this work , we present a novel framework , based on a bayesian inference scheme , able to reconstruct transmission trees and infection dates of susceptible premises , integrating coherently genetic and spatiotemporal data with a single model and likelihood function . Our scheme uses epidemiological data ( times of reporting and removal from the susceptible population of infected , spatially-confined hosts , their locations , and estimates of the age of an infection based on clinical signs ) together with pathogen sequences obtained from infected hosts to estimate transmission trees and infection dates during outbreaks . The genetic information is incorporated considering the probability distribution of the number of substitutions between sequences during the time durations separating them , and computing the likelihood of observing these sequences for a given transmission tree and the estimated infection dates . Each host generates an isotropic infectious potential responsible for transmission between hosts , whose strength is estimated from the data; the dynamical progression of the disease , from latency to infectiousness is part of the estimation scheme ( for a visual representation see Fig . 1 ) . As an illustration of the method , we concentrate on the case of FMDV , an infectious disease affecting cloven-hoofed animals , which has severely affected the UK in 2001 and , on a smaller scale but still contentiously , in 2007 . The infectious agent is single-stranded , positive-sense RNA virus , belonging to the genus Aphthovirus in the Picornaviridae family , and its small genome ( 8 . 2 kb ) is easily sequenced . Its high substitution rate ( per nt per day as measured over part of the 2001 UK epidemic [13] ) , implies that the number of mutations accumulate during infection of host individuals on a single premise is sufficient to be reasonably confident of distinguishing between infected premises . Upon infection by FMDV , a host individual first experiences a non-infectious latent period with lesions appearing on peripheral epithelia subsequently . The virus can spread through aerosol dispersal , on fomites , or through direct contact . Importantly , a visual exam of the clinical state of the lesions on infected hosts can provide valuable information about the age of the infection . For this application , premises comprising populations of spatially-confined hosts will be considered as the unit of infection ( the centroids of premises will be used as geographical coordinates ) , and complete FMDV genomes sampled from each premise will be used for the inference; the removal of a premise from the population corresponds to its culling . As the time course of FMDV infection within an individual host follows empirically characterised distributions [13] , when transmission events are inferred between premises infected at very different times and therefore with correspondingly long and unrealistic apparent latency durations , we interpret these as an indication of the presence of one or more unsampled infected premises , that epidemiologically linked the observed premises . After testing our method on simulated data , we considered two real datasets from two different FMDV epidemics: the 2007 UK epidemic ( 8 premises ) [15] and the Darlington cluster within the 2001 UK epidemic ( 15 premises ) [13] . For the former case , we confirmed the role of IP5 as the link between the two phases of the epidemics , whereas for the latter , our scheme highlights the presence of premises outside our sample that were part of the transmission process . While in this paper we discuss results related to FMDV , our method is in principle general and can be applied to epidemics generated by other pathogens , for which genetic and epidemiological data are both available .
Prior to applying our method to real data , we first used our model to simulate data for an outbreak infecting 20 premises whose locations are known in a 22×11 km area . The model was fitted to the observable data , that is , for each premise , the time at which the virus was detected , a 8000 bp DNA sequence sampled at , an assessment of the lesion age , and the time at which the premise was culled ( see Fig . 1 for a visualisation ) . More information on this dataset can be found in Text S1 . In Fig . 2 ( top left ) , the size of the dots corresponds to the posterior probabilities of pairwise transmissions , while the circles represent the true transmissions as they occurred in the simulation . Fig . 2 ( top right ) shows the tree with highest posterior probability . We note that only one true transmission ( ) is not reconstructed accurately , the algorithm instead identifying . However , the transmission has a high posterior probability and is included in the tree with the second highest posterior probability ( see Fig . S2 ) . The posterior probabilities for the mean latency duration and the mean transmission distance include the true values in the 95%-posterior intervals ( bottom panels of Fig . 2 ) . Posterior distributions for other model parameters and latent variables are provided in the Figs . S3 , S4 . In order to test our method for a large dataset , we considered an upscaled simulation of an outbreak infecting 100 premises . Results are described in Text S1 . Having established the validity of the inference scheme , we applied it to a dataset corresponding to the 2007 outbreak of FMDV in the UK , which infected 8 premises in Surrey and Berkshire [15] . Genetic sequences and epidemiological collected on each premise are available in the Dataset S1 and S2 , respectively . The most likely reconstructed scenario ( Fig . 3 , top right ) comprises two phases: IP1b was infected by an external source , and transmitted the virus to the neighbouring premise IP2b and to IP5 further away; the virus remained contained and undetected on IP5 until it spread to a closeby premise IP4b; finally the virus spread from IP4b to the other premises . While the link made by IP5 between the two phases is highly supported , the estimation of the other transmissions was more uncertain: within the two clusters ( IP1b , IP2b , IP5 ) and ( IP5 , IP4b , IP3b , IP3c , IP6b , IP7 , IP8 ) several other transmission scenarios have non-negligible posterior probabilities ( Fig . 3 , top left and Fig . S5 ) . The mean estimated latency duration has a posterior median of 14 days and a 95%-credible interval of ( 6 , 49 ) ( as shown in Fig . 3 , bottom left ) ; the long delay between the infection of IP5 and the subsequent transmissions is responsible for this result ( posterior distributions of latency durations of every premises are shown in Fig . S7 ) . The long distance between IP5 and its source ( IP5 is 18 . 2 km away from IP1b ) explains the large mean transmission distance ( Fig . 3 , bottom right ) , whose posterior median is 17 km and 95%-posterior interval is ( 5 , 58 ) . Posterior distributions of other model parameters and latent variables are provided in Figs . S6 , S7 , while a phylogenetic tree , based on statistical parsimony tree , implemented in the software package TCS [16] is represented in Fig . S14 . For a more complex scenario , we considered the FMDV epidemic that occurred in the UK in 2001 , and in particular a group of 12 premises within the so-called “Darlington cluster” ( Durham county ) , for which one virus sequence per premise is available [13] . This spatial cluster comprises 3 additional premises that were not epidemiologically linked to the rest of the cluster and which we exclude ( we discuss the choice of the subgroup of premises in the Text S1 ) . Genetic sequences and epidemiological data for this cluster can be found in the Datasets S3 and S4 , respectively . Our method allowed us to reconstruct a transmission scenario with little ambiguity , accounting for over 99% of the posterior probability , where premise K plays the role of a hub and only two chains of transmissions of length greater than two are found ( Fig . 4 , top panels ) . When premises become infectious approximately at the same time , they have a very low probability of mutual infection , even if the collected genomes are very close and share substitutions ( premises M and D , or L and E , for example ) . Premise K , on the other hand , became infectious very early on and is then estimated to have seeded the infection to the many premises that were observed at later times . Interestingly , some premises infected by the hub share mutations that are not found on the other premises , suggesting that different unsampled strains evolved on the hub and went on to infect distinct clusters of farms ( see the statistical parsimony network in Fig . S14 ) . However , another hypothesis can be formulated: the virus fixed the common substitutions while replicating on an unsampled premise , which constitutes a missing node in the transmission tree . This “ghost premise” went on to infect the premises we observed . The missing node scenario is supported by the distribution of the mean latency duration estimated for this dataset , which has a median of 24 days , and a 95%-posterior interval of ( 17 , 35 ) ( Fig . 4 , bottom left ) . These values are inconsistent with a typical latency period of FMDV of 5 days ( 95% confidence interval of 1–12 ) [17]–[19] . In particular , the premises infected by the hub all display high mean latency values ( Fig . S11 ) . We propose that these unrealistically long latency periods indicate the existence of missing premises intermediate in the chain of infection and so in our model , latency should be considered as an aggregated parameter , corresponding to the the sum of the real latent period and the time the virus spent on the unsampled premise . We will return to this point in the Discussion . The comparison of our results with those found by Cottam et al . on the same dataset [13] highlights that our method strengthens the role of infecting hubs in the network ( premise K ) , and therefore infers a lower number of long transmission chains . Details about the individual differences between the most likely trees inferred by the two methods can be found in Text S1 , while transmission trees with higher posterior probabilities and posterior probabilities of other paramteres can be found in Figs . S9 , S10 . The estimates of the transmission kernel for the two real data sets are similar: the 95%-posterior intervals of the mean transmission distance ( defined as ) overlap , ranging from 5 to 58 km for the 2007 outbreak and ranging from 9 to 72 km for the 2001 epidemic ( Figs . 3 and 4 , bottom right panels ) . On the other hand , the posterior distributions we obtained are related to the range of distances covered in the data sets ( up to about 24 km for 2007 and 16 km for 2001 ) , and cannot be used to extrapolate long distance transmission events: despite the large values of the mean transmission distance , the lengths of the average inferred transmission in the trees with the highest posterior probabilities are 4 . 3 km for the 2007 outbreak and 5 . 8 km for the 2001 epidemic . In the inference scheme , we used vague priors for model parameters . When we estimated the interval from the end of latency to detection , however , we used a more informative prior , centered over the estimated lesion age ( Eq . ( 8 ) in Materials and Methods ) . We investigated the effect on the most likely transmission tree of ( i ) using a flatter prior ( thus believing less than we did previously in the veterinarian assessment ) and ( ii ) using a more peaked prior ( thus believing in it more ) . The trees are illustrated in Fig . S12 , and the priors in the Fig . S13 . For the 2007 outbreak , the tree differed only by one transmission in case ( i ) , and by three transmissions in case ( ii ) . Remarkably , in all cases , the identification of the link between the two phases in IP5 maintained a posterior probability of one . For the 2001 epidemic , the star-like shape ( with K as a hub ) of the tree was strengthened in case ( i ) , where premise K now infected 9 premises , while more chains of length greater than two were inferred in case ( ii ) . Constraining the inference less around the estimates of the lesion ages relaxes the timing constraints and increases the weight accorded to genetic similarity in the transmission inference . As a result , transmissions mirror more closely the phylogenetic structure of the dataset , leading to a reduced hub role of premise K . In conclusion , we remark that the tree structure is robust and does not crucially depend on the specific choice of the prior for the values of the time intervals between the end of latency and detection ( lesion ages ) . Our method relies on one approximation: we do not reconstruct the genomes transmitted at the times of infection , and therefore we obtain a pseudo-posterior probability for the genetic data , where the similarity between isolates only depends on the Hamming distance between the sequences , and not on the full genetic network ( see Materials and Methods for details ) . We checked whether the use of a pseudo-posterior distribution led to appropriate inference by applying the estimation algorithm to three series of 100 simulations ( one for the test outbreak and two for the FMDV datasets ) generated using our model . For the first series , we used the parameter values that were used in the test simulation . For the two other series , we used the posterior medians of the parameters estimated previously . We were especially interested in the fraction of correctly predicted pairwise transmissions: for each premise , between 79% and 93% of the simulations reproduced the source with the highest posterior probability in the original inference ( Table 1 ) . Given the challenging nature of the data sets ( closely spaced premises becoming infectious almost simultaneously in the test data , and an abnormally long period of time between infection and transmission between two waves of infection in the 2007 data ) , these results suggest the approximation is performing well . Moreover , the mean of the posterior probability of each true transmission ( the proportion of iterations in the chain at which a premise is infected by the estimated source ) is also reproduced in about 80% of the cases . Performances vary slightly across datasets depending on the characteristics of the epidemics ( e . g . number of premises and parameter values ) , but are broadly compatible . For example , in the second phase of the 2007 outbreak , several scenarios have high posterior probabilities , lowering the fraction of correctly estimated transmissions . Further performance estimators are listed in Table S1 .
We propose here a new bayesian inference scheme , with which we estimate transmission trees and infection dates for an epidemic outbreak using genetic and epidemiological data . Our scheme is general , and with slight modification can be applied to rapidly evolving pathogens affecting spatially-confined hosts . To illustrate how this approach can be used to generate new insights and deliver statistically formal measures of confidence ( in particular transmission links ) , we applied it to the case of an RNA virus ( FMDV ) infecting premises whose spatial location is known . The knowledge of complete viral sequences , timing of reporting and culling of premises and estimates of the age of an infection made this case an ideal benchmark . After testing our method on simulated data ( 20 premises ) , we applied it to two pre-existing datasets: the still disputed 2007 FMDV outbreak in the UK ( 8 premises ) [15] and the Darlington cluster within the larger 2001 epidemic ( 12 premises ) [13] . The method proved successful in reconstructing the transmission network on the test dataset , and highlighted the role of IP5 as a relay between the two phases of the 2007 outbreak . The results for the Darlington cluster are intriguing , as they highlight the likely incompleteness of the dataset , and suggest the presence of unobserved premises in the transmission tree . The performance of the algorithm was evaluated through simulations , which showed the inference scheme to be consistent and accurate and able to deal successfully with clusters of infections . The power of this inference platform relies on a number of simplifying assumptions . In this application we have made two in particular that require further consideration . The first postulates that the epidemics are generated by a single introduction of the pathogen to a single premise . While this may often be adequate for small or early stage outbreaks , it is likely to be inadequate for more complex cases . For example , the Darlington dataset is a small subset of the 2001 epidemic , in which it was first considered to be an isolated cluster of infected premises . Previous analysis on the whole cluster [13] demonstrated two independent introductions . Trying to estimate “polyphyletic” transmission trees assuming only a single root would strain this formulation of the model and lead to unrealistic results . In order to solve this problem , the MCMC should be able to explore a parameter space where independent introductions range from one to the number of the premises ( each of them being independently infected by an external source ) and compute their likelihood . Moreover , the genetic data can be used to discriminate between a situation where a single external source infects several spatially-confined hosts in a cluster , and the presence of multiple external sources , characterised by distinct genomes . In practice , we could proceed by ( i ) describing the external source ( s ) as a set of genetic sequences varying in time ( and possibly in space ) , ( ii ) specifying the probability of transmission of the infection from the external source ( s ) to any of the premises and ( iii ) updating the transmission tree at each iteration of the MCMC by comparing this probability with the probability of transmission from one of the infectious premises in the cluster considered . The second assumption is that the epidemic has been completely observed and that there are no missing nodes in the transmission tree . When this assumption is likely to be violated , as in the case of the Darlington cluster , our method inferred unrealistically long latency times for some premises , an indication that a missing intermediate infected premise , where virus might have replicated extensively , may have been involved in the transmission chain . This situation is particularly likely in large epidemics , where perfect knowledge of every case is unlikely , or in epidemics arising in areas or countries where host or premise identification is ambiguous and comprehensive collection of data not feasible . In the 2007 outbreak , where no infected premises were missing , the premise linking the two phases showed a mean latency duration of over 25 days . In this case , the observation results from the real time the virus spent on the farm prior to its detection and reporting: by the time it was observed , the animals had started to heal and dating the lesions was more difficult . The long latency times could also account for the time virus spent in a non-replicative state ( e . g . on fomites ) : this case would be indicated by a slow rate of evolution on the premise where the virus is observed . In conclusion , extended latency times are valuable “alarm bells” , as they suggest a discrepancy between the observations and the actual course of the disease . A substantial improvement to the scheme would be to include in the inference additional sources of data , such as the locations of premises that may have maintained infections that were not detected , or premises that were infected but were removed prior to being confirmed as infected . We leave this development for future work . We only mention here that the solution given in the paragraph above to deal with multiple introductions could be adapted to deal with missing premises: any infectious premise could generate a set of genetic sequences describing possible missing premises . This set of sequences could then be used to compute a new probability of transmission from missing premises , to be compared with the probabilities of transmission from internal and external sources . We leave this for future work . Other minor assumptions in our model can be readily eased . We hypothesized that all premises have the same infection potential; however , it would be straightforward to make the infectiousness parameter in the model a function of the specific characteristic of the premise , like size or composition ( for example , for FMDV sheep are considered to be less infectious than cows , which are in turn less infectious than pigs [17] ) . Moreover , we note that the infectious potential felt by a premise at time is the sum of the contributions deriving from all the other premises that are infectious at that particular time . As unsampled premises could also contribute to this potential , the temporal dynamics of infection could be modeled in a more complex manner than the step function adopted here . The estimation of the age of an infection from clinical signs is used as a prior distribution in our scheme: an accurate knowledge of this quantity makes the inference computationally more efficient , but it is not essential , and the method can be applied to cases where this quantity is not available . The model used for the mutations of the virus is very simple and does not account for the specific characteristiscs of the FMDV genome , or for some well-known mutation biases ( like the transition/transversion bias observed in [20]: we decided once more to go for the simplest and more general assumption , while more detailed and pathogen-specific mutation models could easily be incorporated in our framework . Our “hosts” do not necessarily correspond to single animals/humans but were interpreted in a wider sense as “infectious units” . These units do not constitute a limitation to our method: even in the case of an infection where the units are individuals , the genetic divergence between sequencing results from an unknown number of viral replications in the donor individual post sampling ( but prior to transmission ) and in the recipient prior to sampling . In the case of a higher-order unit of infection , the genetic divergence between sequences from sequential samples will be just the result of a larger unknown number of generations . It is conceivable that multiple pathogen strains circulated on a single premise remained unsampled and went on to infect other premises . For example , FMDV is known to generate independent populations within single animals [20] and different genomes could circulate on a premise . Ideally , several sequences from each premise should be obtained and these data incorporated into the model . Finally , for the specific pathogen considered here , we have used a fixed substitution rate for both the Darlington cluster and the 2007 outbreak . Independent estimates obtained for the whole 2001 epidemic [21] and for 2007 outbreak yield very similar values , which do not change substantially the likelihoods of observing the sequenced genomes . In other applications , the substitution rate may be poorly known . In these cases , it could be viewed as an unknown parameter and estimated in the MCMC simulation . Computation time is a key element for a method that is expected to be useful in real-time during an outbreak . The computation time was strongly reduced by using a conditional pseudo-distribution of observed sequences instead of the exact conditional distribution . Clearly , it would be ideal to run the Bayesian estimation using the exact conditional distribution of observed sequences . To do so , one could incorporate in the MCMC the unknown transmitted genetic sequences as augmented data ( see Eq . ( 3 ) below ) , initialize using for example statistical parsimony [16] and determine a proposal distribution for based on a stochastic algorithm estimating genetic networks [22] . Unfortunately , this strategy is at present unfeasible on standard computing resources . However , despite the use of a pseudo-distribution , the running time of our inference algorithm strongly increases with the number of premises . We stress that the main focus of this work was to combine epidemiological and genetic data in a coherent framework , rather than producing an optimised code . Basic optimization procedures should dramatically increase the efficiency of the code . In particular , we suggest three directions worth pursuing: ( i ) use a conditional pseudo-distribution of the genetic sequences which can be computed faster , but still yielding a good approximation of the posterior distribution of the unknowns; ( ii ) parallelize the MCMC [23] and code it in a lower-level language; ( iii ) use alternative algorithms , such as sequential Monte Carlo [24] . Our bayesian inference scheme is a rigorous general platform on which different models can be implemented and tested . It is a useful tool that could be used in real time to detect the presence of missing links in inferred chains of transmission , and to assign confidence values to each inferred transmission event . The specific model we chose for FMDV contains a representation of the dynamics of FMD infections . Different models could be implemented to describe the dynamics of different pathogens , or the specific characteristics of a particular outbreak , while still maintaining rigorous estimation based on genetic and epidemiologic data . Previous work was initiated by Cottam et al . [13] , and significantly extended by Jombart et al . [12] and Ypma et al . [14]: all these studies considered the likelihood of the transmission tree given temporal , spatial and genetic data ( here denoted by the generic vectors , and ) as a product of three independent likelihoods: . Cottam et al . assumed a binary ( ) and a uniform ( their estimation does not depend on the location of the premises ) ; Jombart et al . designed a less “ad hoc” approach by introducing a maximum parsimony strategy to weight genetic similarity , while spatial and temporal information were considered only when several possible ancestors were genetically indistinguishable; finally Ypma et al . had more complicated forms for these likelihood functions . Our method can be considered as the “next step” on this road , as we relax the assumption of independence between the information sources , and we estimate the likelihood of transmission trees given all the sources of information simultaneously . Although some specific aspects of our inference scheme can be refined , expressing the likelihood of a transmission tree as a joint likelihood , depending on both epidemiological and genetic data , significantly advances this form of analysis .
The test data sets analyzed in the Results section were simulated under the model presented below and in Text S1 . In these data sets , the outbreak spread over 20 premises ( F1 , … , F20 ) , randomly and uniformly located in a rectangular 20×10 km region . Values of transmission and latency parameters were and . Observed sequences had length and substitution rate . In Text S1 , we analyzed an upscaled test data set with 100 premises , with the same premise density as above , and same values for parameters , , and . The data corresponding to the 2007 FMDV outbreak in the UK and to the Darlington cluster within the 2001 epidemic can be found in Refs . [15] and [13] , respectively , and are incudedin the Datasets S1 , S2 , S3 , S4 . In particular , FMDV sequence length was and the substitution rate per nt per day [13] . Consider a cluster of infected hosts ( in this case premises ) whose centroids are located at Longitude-Latitude coordinates . Let be the function defining the transmission tree: a given premise is infected by a source , which consists of either another premise , , or an external source denoted by 0 . For each premise , we consider four timing variables as illustrated by Fig . 1: premise is infected by at time , is infectious at time , where is the latency duration for premise , is detected as infected at time and is removed from the infectious population at time . The duration from infectiousness to detection , , is assessed by experts on the base of clinical signs: let denote this assessment . At time , the pathogen is sampled on premise and the genomes are collected for sequencing: let denote the observed consensus sequence . Among these variables , only , , , and are observed . The others are latent variables to be reconstructed with the bayesian inference scheme . In this section we briefly describe the essence of the model . The complete specification of the model is provided in the following sections . For a full description of the symbols , we refer to Table 2 . Our model for the dynamics of an infection takes into account the dependence between timing , space and genetics . It includes ( i ) the delays between infection and observation of infection and ( ii ) the difference between transmitted and observed genetic sequences of the pathogen . The direct acyclic graph ( DAG ) in Fig . 5 shows the structure of the model . Upper case letters are used for latent and observed variables , while Greek letters denote unknown parameters . Lower case letters are used for fixed parameters . Observation times and observed consensus sequences are viewed as response variables . They depend on the transmission tree and on the temporal dynamics ( infection times , latency durations and detection durations ) . The model assumes that the epidemic starts with the infection of a single premise from an external source . Then , transmissions and infection times depend on the infection potential generated by previously infected premises . The infection potential depends on the transmission parameters , the spatial location of premises and the times at which infected premises exit from latency and are removed from the infectious population: an infected premise is infectious between and , and the probability of infecting premise decreases exponentially with the distance . The parameter appears in the transmission kernel and quantifies the decrease with distance of the infection potential of each infectious premise , while quantifies the infection strength of each infectious premise . The mean transmission length , defined here as , is a function of the distances between farms and of the transmission kernel we used . Latency durations and durations from infectiousness to the time that virus is sampled are assumed to be independent . The distribution of is parametrised by its expectation and its variance ; is the vector of latency parameters . The distribution of is centered around the empirical estimate but has a variance increasing with , equal to , where . The premise index is sorted with respect to increasing infection times . We aim to assess the joint posterior distribution of the transmission tree , infection times , latency durations , durations from infectiousness to detection , and parameters , given the data . Data are observed sequences , pathogen observation times , observed durations from infectiousness to detection , removal times and premise locations : ( 1 ) where means “proportional to” ( the multiplicative constant does not depend on the unknowns ) . In this decomposition , are viewed as response variables ( or model output ) , as latent variables and as explanatory variables . The term is the complete likelihood of the model and the term is the conditional complete likelihood of the model given observation times . In the following sections , we specify the terms appearing in the last two lines of Equation ( 1 ) . Assumptions: ( a ) there is only one sequence per infected premise; ( b ) sequences in all the premises evolve at a constant rate ( is the substitution rate per day per nucleotide ) . The model for is based on the probability distribution of the number of substitutions between two sequences during the evolutionary durations separating the sequences . Let denote the number of substitutions and the evolutionary duration ( is the sum of time intervals computed along the transmission tree ) . The conditional distribution of given is a Binomial distribution taking into account the Jukes-Cantor's correction ( see Text S1 ) :and the probability of given is: ( 2 ) Therefore , does not depend on :and can be written as a multiple sum of products of binomial probabilities . The sum is computed over the unknown transmitted genetic sequences , say , at time ( the initial sequence of the root of the tree is not needed ) : ( 3 ) In Equation ( 3 ) , is the set of all possible sequences ( the size of is , where is the length of the sequence ) ; is the number of substitutions between and ; is the probability given by Equation ( 2 ) with and . The subscript denotes the premise whose node of infection belongs to the tree path from the root of the tree to the observation of ( at time ) and whose infection is just preceding the observation of . The node of infection of a given premise is defined as the point on the tree at which “the branch leading to the observation of ” and “the branch leading to the observation of the infecting premise ” diverged . The tree path from one point of the tree to another is defined as the most direct path on the graph conncting the two points . If did not infect any other premise , then is itself . In the particular case where was infected after the observation of the infecting farm and did not infect any other premise between and ” , the subscript coincides with , and . In the most frequent other cases , denotes the premise whose node of infection belongs to the tree path from the root of the tree to the infection of ( at time ) and whose infection is just preceding the infection of ; in these cases , and . In other words , the first series of factors in Equation ( 3 ) accounts for the probabilities of the number of substitutions between an observed sequence and the immediately preceding unobserved , transmitted sequence , while the second series of factors accounts for the probabilities of the number of substitutions between each transmitted sequence and the transmitted or observed sequences immediately preceding in time . Equation ( 3 ) is written in the Supporting Text S1 ( Equation ( 2 ) ) for the simple transmission tree drawn in Supporting Fig . S1 . The conditional distribution for ( Eq . ( 3 ) ) was written as a distribution depending solely on the genetic distances for pairs of sequences . However , in each pair , there is at least one unobserved transmitted sequence . Therefore , exploiting Equation ( 3 ) would lead us to consider extra latent variables ( or augmented data ) , namely the unobserved sequences . In order to reduce the complexity of the posterior , we preferred not to include these extra latent variables , but rather to use a conditional pseudo-distribution of , . In our method , replaces which is the conditional complete likelihood of the model given observation times . Thus , is a conditional complete pseudo-likelihood given observation times and we refer to it as a conditional pseudo-distribution . It follows that the posterior distribution that we assess is actually a pseudo-posterior distribution . With index being sorted with respect to increasing infection times , can be written: ( 4 ) where is the set of observed sequences of premise . We considered the sequence of the first infected premise as arbitrary . Thus , was discarded in the pseudo-distribution . Moreover , to compute exactly appearing in Equation ( 4 ) , we should write this probability as a sum over the unknown transmitted genetic sequences ( as done in Equation ( 3 ) ) . In order to avoid the inclusion of unknown transmitted sequences as augmented data , we replaced , for , the conditional probability of given past sequences ( ) by the product of the conditional probabilities of given each past sequence ( ) :where denotes the infection time at which the chain of infection leading to and the chain of infection leading to diverged ( is one of the latent variables in , also called “augmented data” ) and is the evolutionary duration separating the observation of and . Thus , the conditional pseudo-distribution of satisfies: ( 5 ) The right hand side of Equation ( 5 ) replaces in Equation ( 1 ) . Equation ( 5 ) is written in Equation ( 3 ) in Text S1 for the simple transmission tree drawn in Fig . S1 . We tested another form for , described in Text S1 . The form given by Equation ( 5 ) above led to the best reconstruction of the transmission tree . satisfies the relation . Therefore , the conditional distribution of is simply: ( 6 ) where 1 is the indicator function ( 1 if event occurs , zero otherwise ) . Assumptions: ( a ) Only one premise is infected by an external source , while the others premises in the dataset are infected by previously-infected premises within the dataset; ( b ) any premise may infect other premises after the latency period and before the culling time ; ( c ) infectious premises have same infection strength , considered constant; ( d ) the infection risk of a susceptible premise by an infectious premise decreases exponentially with the distance separating both premises , this distance being measured by the distance between the centroids of the premises; ( e ) the presence of unsampled premises in the area ( premises for which genetic or epidemiological data is not available ) is ignored . With the index being sorted with respect to increasing infection times , the probability can be written: ( 7 ) where and . Each premise has the same chance ( ) to be infected first ( by an external source ) , and its infection time is assumed to be greater or equal than a minimum infection time ( in this work we used ) , and less than or equal to the minimum removal time :Subsequent infections occur with the following probabilities:where the term is the probability that premise has not been infected until time by the previously infected premises , and the term is the probability density that premise has been infected by at time . The function is an exponential transmission kernel , defined for all distance asFor transmissions modelled using the exponential transmission kernel , the mean transmission distance ( mean length of transmissions ) is : this measure depends on the distances between farms as well as on the transmission kernel we used . Other transmission kernels , such as those presented in [25] , [26] could be tested . The selection of the best transmission kernel will be crucial for datasets with large number of premises and large spatial extent . In our applications , where the number of premises is limited and the spatial extent is much smaller than the dispersal capacity of the pathogen , there are enough data to infer the transmission parameters , but not enough to carry out a significant model selection about the transmission kernel . Assumptions: ( a ) a priori , latencies and durations from infectiousness to detection are independent; ( b ) characteristics of the latency distribution ( expectation and variance ) do not depend on time and premise; ( c ) the expectation ( resp . variance ) of the duration from infectiousness to observation is equal to ( resp . is proportional to ) the estimate provided . We chose gamma distributions for latency durations , with shape and scale parameters and , respectively , so that and . We refer to as mean latency duration . We chose gamma distributions for detection durations with shape and scale parameters and , respectively , so that and . Thus , the joint distribution of the vectors of latent variables and satisfies: ( 8 ) where is the gamma function . The four components of have independent exponential priors with mean parameters : ( 9 ) We have used the values . We built a Monte Carlo Markov Chain ( MCMC ) algorithm to assess the posterior distribution of , coded in the R language [27] . Details of this algorithm are provided in Text S1 . We recall that , in order to reduce the complexity of the algorithm , we replaced the conditional distribution of observed consensus sequences appearing in the posterior distribution by a pseudo-distribution . This replacement allowed us to remove some of the latent variables , namely the unobserved pathogen sequences transmitted at the infection times . Therefore , the MCMC algorithm assesses a pseudo-posterior distribution of . Vague priors were used for parameters and ( see above ) . In the cases considered in this study , iterations of the MCMC algorithm were enough to assess the posterior distributions of the unknowns . Running iterations took about two days for the simulation with 20 premises and one month for the simulation with 100 premises on an Intel Xeon Quad Core processor with clock speed 2 . 93 GHz and 48 Gb of RAM memory . The components of the algorithm which are especially computationally costly are ( i ) the search of the most recent ancestral premises appearing in the pseudo-distribution of the observed genetic sequences given in Equation ( 5 ) , ( ii ) the computation of the joint distribution of and in Equation ( 7 ) which is based on a convolution between the transmission kernel and the sources of infection , and ( iii ) the verification that timing constraints are satisfied when infection times are updated ( see proposal distributions in Text S1 ) . We generated data sets using the model described above and the location of the premises . The spread of the disease was first simulated using the conditional distributions of , , , , and , with previously inferred parameters , thus obtaining the complete dynamics of the infection and a transmission tree . Subsequently , genetic distances between the observed sequences were generated using the binomial distributions described in Equation ( 2 ) . We note that in this case we generated the unobserved transmitted sequences as well .
|
In order to most effectively control the spread of an infectious disease , we need to better understand how pathogens spread within a host population , yet this is something we know remarkably little about . Cases close together in their locations and timing are often thought to be linked , but timings and locations alone are usually consistent with many different scenarios of who-infected-who . The genome of many pathogens evolves so quickly relative to the rate that they are transmitted , that even over single short epidemics we can identify which hosts contain pathogens that are most closely related to each other . This information is valuable because when combined with the spatial and timing data it should help us infer more reliably who-transmitted-to-who over the course of a disease outbreak . However , doing this so that these three different lines of evidence are appropriately weighted and interpreted remains a major statistical challenge . In our paper we present a new statistical method for combining these different types of data and estimating trees that show how infection was most likely transmitted between individuals in a host population . Because sequencing genetic material has become so affordable , we think methods like ours will become very important for future epidemiology .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"systems",
"biology",
"computer",
"science",
"computer",
"modeling",
"veterinary",
"epidemiology",
"ecology",
"evolutionary",
"modeling",
"theoretical",
"ecology",
"biology",
"computational",
"biology",
"veterinary",
"science"
] |
2012
|
A Bayesian Inference Framework to Reconstruct Transmission Trees Using Epidemiological and Genetic Data
|
Inner ear sensory hair cell death is observed in the majority of hearing and balance disorders , affecting the health of more than 600 million people worldwide . While normal aging is the single greatest contributor , exposure to environmental toxins and therapeutic drugs such as aminoglycoside antibiotics and antineoplastic agents are significant contributors . Genetic variation contributes markedly to differences in normal disease progression during aging and in susceptibility to ototoxic agents . Using the lateral line system of larval zebrafish , we developed an in vivo drug toxicity interaction screen to uncover genetic modulators of antibiotic-induced hair cell death and to identify compounds that confer protection . We have identified 5 mutations that modulate aminoglycoside susceptibility . Further characterization and identification of one protective mutant , sentinel ( snl ) , revealed a novel conserved vertebrate gene . A similar screen identified a new class of drug-like small molecules , benzothiophene carboxamides , that prevent aminoglycoside-induced hair cell death in zebrafish and in mammals . Testing for interaction with the sentinel mutation suggests that the gene and compounds may operate in different pathways . The combination of chemical screening with traditional genetic approaches is a new strategy for identifying drugs and drug targets to attenuate hearing and balance disorders .
Hearing loss and vestibular dysfunction are among the most common disorders requiring medical attention . Globally , over a third of older adults suffer from these conditions . Studies of both laboratory animals and humans reveal tremendous variation in hearing loss due to ageing as well as exogenous challenges such as ototoxic drugs and noise exposure , and show that this variability can be at least partially understood using genetic methods [1]–[5] . Rapid progress has been made using genetics to understand the molecular basis for congenital deafness [6] , but adult-onset hearing loss is poorly understood despite its overwhelming prevalence . There are several examples where genes underlying familial adult-onset hearing loss have been identified [7]–[9] , but these are rare diseases that account for a very small fraction of the enormous variation of acquired or age-related hearing and balance problems . Understanding how hair cell death is genetically modified by intrinsic and extrinsic challenges should lead to identification of new therapeutic targets for prevention of inner ear damage . The initial cellular basis for most hearing loss and a significant proportion of balance problems is injury and loss of the mechanosensory hair cells that reside in the inner ear and transduce mechanical signals into electrical signals that are sent to the brain via the VIIIth cranial nerve . Treatments with aminoglycoside antibiotics or the cancer chemotherapeutics , cisplatin and carboplatin , often cause irreversible hearing loss [10]–[12] by killing hair cells . As with other forms of hearing loss , the effects of aminoglycoside exposure in humans and other outbred mammalian populations are widely variable and influenced by genetic factors [13] . For example , patients with mutations in mitochondrial genes , including mitochondrial 12S ribosomal RNA , show greatly enhanced sensitivity to aminoglycoside exposure [14] . However , these mutations also have variable penetrance , and are influenced by nuclear genes [15] . Mutations in mitochondrial rRNA are consistent with a model that aminoglycoside ototoxicity is the result of effects on mitochondrial translation similar to the antibiotic effects of prokaryotic translation inhibition [16] . Pharmacological approaches toward the prevention of hearing loss due to therapeutic drugs or chronic exposure to noise have centered primarily on antioxidants and cJUN kinase ( JunK ) inhibitors . While several studies support the idea that antioxidants or JunK inhibitors can limit aminoglycoside toxicity and cisplatin ototoxicity , the literature is complex and often the protection is dose dependent [11] , [17] . Target based drug discovery is limited , however , by our understanding of the cellular pathways contributing to the inner ear pathology , and by the lack of methods to do broad screening of potential candidates . The lateral line system of aquatic vertebrates is composed of mechanosensory organs on the surface of the head and body , and is used to detect variations in water pressure . Lateral line hair cells and their underlying support cells are organized into rosette-like clusters called neuromasts [18] . Zebrafish lateral line hair cells show structural , functional and molecular similarities to the mammalian inner ear hair cells ( reviewed in [19] , [20] ) . Like mammalian inner ear hair cells , the lateral line hair cells of zebrafish are killed by exposure to chemicals including aminoglycosides and cisplatin in a dose-dependent manner [21]–[25] . The accessibility of lateral line hair cells to visualization and manipulation , along with the cellular and molecular properties shared with inner ear hair cells , makes this system a good model for investigating genetic and pharmacological modulation of hair cell sensitivity to potentially ototoxic agents [26] . In this report , we describe a new approach for the identification of genes and pharmacological agents that modulate the sensitivity of hair cells to ototoxic agents such as aminoglycosides . We use this approach to identify 2 new pharmacological agents and 5 new mutations that protect against aminoglycoside-induced hair cell death . We describe a screen for small drug-like molecules that protect zebrafish lateral line hair cells and validate effectiveness of these newly discovered protective compounds in the mammalian inner ear . We report the initial results of an in vivo genetic screen for modulators of hair cell susceptibility to ototoxic drug exposure , including the identification of one such gene . These mutations provide an entry point for determining which molecular pathways can be modulated to alter drug response in the hair cells . Variation in these molecules may underlie differential susceptibility to drugs clinically and suggest likely points of regulation for prophylactic treatments in the future .
To screen for small molecule modifiers , we pretreated 5 day post-fertilization ( dpf ) larvae with a chemically diverse library of 10 , 960 compounds before exposing them to 200 µM neomycin . Screening was initially carried out by labeling hair cells of 5 dpf larvae with a combination of a nuclear dye and a cytoplasmic dye ( Yo-Pro-1 and FM 1-43 , respectively ) , then pretreating larvae in 96-well tissue culture plates for 1 hour to a cocktail of five compounds and then exposing them to 200 µM neomycin . When protection was observed , the 5 potential contributors were evaluated singly to determine the active compound . Two compounds exhibited reliable and robust protection of hair cells from neomycin . An example of this protection is shown in Figure 1D , compared to treatment with neomycin alone ( Figure 1B ) . Both compounds were benzothiophene carboxamides ( Figure 1E and 1F ) , suggesting specific selection from the diverse library . We have named these compounds PROTO-1 and PROTO-2 . We next compared the neomycin dose-response relationship in larvae pretreated with the compounds and controls ( Figure 2 ) . Figure 2A and 2B show that at a concentration of 10 µM both compounds show significant protection of hair cells over a broad range of neomycin concentrations , from 25 µM to 400 µM ( p<0 . 0001 by two-factorial ANOVA ) . We also determined the dose-dependent effects of PROTO-1 and PROTO-2 to a fixed ( 200 µM ) level of neomycin ( Figure 2C and 2D ) . Pretreatment with 1 and 10 µM PROTO-1 resulted in significant protection of hair cells exposed to 200 µM neomycin compared to neomycin alone ( p<0 . 0001 , unpaired t-test ) . There was no significant difference in the protection provided by 1 and 10 µM PROTO-1 ( p>0 . 10 ) . Although exposure to 50 µM and 100 µM PROTO-1 alone did not alter viability , in combination with 200 µM neomycin these doses were lethal to larvae . Pretreatment with PROTO-2 provided significant protection of hair cells at all doses ( p<0 . 0001 , unpaired t-tests ) with no dose-dependent difference ( p>0 . 20 ) . PROTO-2 was not lethal at any of the tested doses with or without neomycin . Aminoglycosides are used clinically , despite their known ototoxicity , because of their broad spectrum of antibacterial actions . Compounds that could be used to limit their ototoxicity must not limit the intended therapeutic functions . We therefore had the University of Washington Clinical Microbiology Laboratory test the bacteriostatic and bactericidal activity of neomycin in the presence of PROTO-1 and PROTO-2 . The minimum inhibitory concentration ( 3 . 25 µM ) and minimum bactericidal concentration ( 6 . 5 µM ) for E . coli ATCC 25922 was unchanged with or without 10 µM of either compound . This indicates that at least under standard in vitro assay conditions benzothiophene carboxamides do not inhibit aminoglycoside antibacterial activity . To identify genetic modifiers of aminoglycoside-induced hair cell death , a standard F3 screening paradigm was used . Males were mutagenized with ethylnitrosourea following standard protocols [27] , then crossed to wildtype females to produce F1 progeny . Mutagenesis was assessed by specific locus testing against unpigmented mitfa mutant animals [28] , with a rate of about 1∶300 . F2 families were produced from F1 individuals , and F3 larvae produced by pairwise intercrosses within each family . F3 larvae were treated at 5 dpf with either high ( 200 µM ) or low ( 25 µM ) concentrations of neomycin for 30 minutes to identify mutants that exhibit protection or heightened susceptibility of hair cells , respectively . Hair cells were then assessed with the vital dye DASPEI , which is differentially taken up by neuromast hair cells [29] , [30] . Figure 3 shows untreated and neomycin-exposed wildtype animals , and two mutants with altered susceptibility . In contrast to the wildtype subject ( Figure 3B ) , persephone mutants ( Figure 3C ) show robust staining indistinguishable from an untreated animal ( Figure 3A ) . Animals homozygous for the sentinel mutation also retain robust staining; in addition they display a linked morphological phenotype , a variable sinusoidal morphology that begins to be apparent by 3 dpf ( Figure 3D ) . While persephone mutants are homozygous viable , the sentinel mutation is lethal at approximately 10–12 dpf . To date , we have identified 5 mutations that confer resistance and behave as simple recessive alleles . Complementation testing demonstrated that they affect different genes . We identified 5 additional mutations that confer resistance with more complex genetics , showing semi-dominant effects and/or interactions with modifying background loci . All mutations were transmitted to the next generation . We were surprised that all loci identified to date confer resistance , suggesting that affected genes normally act to promote cell death . The 5 simple recessive loci can be separated into two classes , mutations that have no apparent secondary phenotype ( persephone , trainman , bane ) and those with additional phenotypes ( sentinel , merovingian ) . Animals homozygous for the merovingian mutation show reduced ear size and small otoliths ( not shown ) . We have found that mutations differ dramatically in the relative resistance they confer against neomycin exposure ( Figure 4 ) . In wildtype animals , 200 µM neomycin exposure reliably reduces DASPEI staining to less than 10% of control untreated animals ( Figure 4A ) . To examine the variability in phenotypes , we crossed heterozygous parents to produce offspring in typical Mendelian ratios ( 75% wildtype: 25% mutant progeny ) . Distributions show bimodality with robust DASPEI staining in 1/4 of the neomycin-treated progeny from crosses of heterozygous individuals for all 5 simple recessive mutations , as expected ( Figure 4B–4F ) . Linked phenotypes cosegregate with resistance as shown for snl mutations ( Figure 4D ) . Examples of mutations that confer near total resistance are shown in Figure 4B and 4C , and examples that confer only partial effects are shown in Figure 4D–4F . Partial protection may indicate that affected genes alter only one of several mechanisms involved in neomycin-induced cell death or that identified alleles may be hypomorphic and display only partial loss-of-function . The linked morphological features associated with sentinel mutants have allowed us to more fully characterize mutant phenotypes , since homozygous affected animals could be prospectively identified before directly testing the response to neomycin . We next tested whether sentinel mutants show altered response over the range of the aminoglycoside dose-response function ( Figure 5 ) . Animals were sorted by body phenotype as either wildtype or sinusoidal , and then exposed to different doses of neomycin . Animals homozygous for the sentinel mutation show robust , but partial , protection at all doses tested . Animals with wildtype body shape , including heterozygous mutants , are no different than the background *AB strain , demonstrating there are no effects of gene dosage . We also determined whether sentinel mutants show protection at later stages of development , since there are age-dependent differences in the dose-response to neomycin [22] . There is no change in the relative levels of protection by sentinel at 8–9 dpf ( Figure S1 ) , demonstrating that the mutation does not specifically confer protection by a general developmental delay . To determine the genetic location of the sentinel gene , we isolated 694 snl mutants and 234 snl+ siblings based on the neomycin response phenotype of their hair cells . We detected cosegregation of the sentinel phenotype with markers [31] on chromosome 23 of zebrafish . Analysis of recombinant chromosomes revealed a 41 kb linked genomic region containing one candidate gene ( Figure 6A ) predicted to encode a 1541 aa protein with 38 exons . The predicted exon and intron boundaries are shown in Figure 6B . The boundaries of the linked region are positioned within the coding region ( within introns 8 and 33 ) of this novel gene . Sequence of the coding regions and exon-intron junctions in sentinel cDNAs revealed a stop codon in exon 14 ( Figure 6B , red asterisk , and Figure 6D ) in place of a tryptophan . This alteration is predicted to truncate the protein at amino acid 491 with loss of 68% of the protein and is likely to lead to loss of function . The sentinel transcript is expressed ubiquitously in wildtype zebrafish ( Figure S2 ) . We observed attenuated expression in sentinel mutants , perhaps indicating that nonsense-mediated decay of the transcript occurs ( Figure S2C ) . Alignment of the zebrafish genomic region reveals homology to human ( KIAA1345 , 56% identity , 73% similarity ) and mouse ( RIKEN 5730509K17 , 59% identity , 76% similarity ) as well as to other vertebrates ( Figure S3 ) . The intron-exon structure between the zebrafish and mammalian orthologs is conserved with a few minor exceptions . We note looser homology to loci in Drosophila melanogaster , Aedes aegpytii , Caenorhabditis elegans , Trichomonas vaginalis and Paramecium tetraurelia genomes , suggesting that this is an ancient gene . The phylogenetic relationship between the predicted proteins is shown in Figure 6C . The Drosophila ortholog is annotated as two loci ( CG18432 and CG18631 ) corresponding to the predicted N-terminal and C-terminal end of the zebrafish protein , indicating that they may encode a single transcript or be derived from a single ancestral locus . The predicted Sentinel protein contains a putative C2 domain [32] in the C-terminus ( Figure 6E ) . The N-terminal third of the Sentinel protein is highly charged with two glutamine-rich acidic clusters flanking a lysine-rich basic cluster ( Figure 6E ) . There is a notable absence of other recognizable domains . To begin elucidating possible molecular pathways regulating susceptibility , we tested for an interaction between sentinel mutants and PROTO-1 . Both PROTO-1 treatment and snl loss of function result in substantial but incomplete protection against neomycin exposure . We tested whether exposure of PROTO-1 conferred any additional protection to snl mutants when exposed to 100 µM or 200 µM neomycin . Figure 7 provides these results for siblings ( left ) and sentinel mutants ( right ) , comparing hair cell counts in control animals and fish exposed to neomycin with or without pretreatment for 1 hr in 10 µM PROTO-1 . At both doses of neomycin , treatment with PROTO-1 provides a small amount of additional protection , over and above that provided by the sentinel mutation . Analyses by one-way ANOVA followed by pair-wise comparisons ( Fisher's PLSD test ) revealed that at both doses the additional protection provided by PROTO-1 was statistically reliable ( p<0 . 01 ) , but that even the combined effect did not provide complete protection ( p<0 . 01 ) . Attenuation of drug-induced hair cell death could result from a number of causes that are not directly linked to the activation of cell death or cell survival pathways . Some examples include the well-established link between mechanotransduction-dependent activity and aminoglycoside uptake and susceptibility [33]–[35] , the relative resistance seen in young animals [23] , and abnormalities of aminoglycoside uptake . Rapid uptake of the vital dye FM 1–43 is commonly used as an indicator of sensory hair cell mechanotransduction [36]–[38] . We compared the uptake of FM1-43FX in control ( wild-type ) fish , in sentinel mutants and in wild-type fish treated with PROTO-1 and PROTO-2 ( Figure 8; Figure S4 ) . Rapid entry of FM1-43FX into the hair cells of sentinel mutants ( Figure 8B and 8D ) is comparable to that of wildtype hair cells ( Figure 8A and 8C ) . Similarly , PROTO-1 and PROTO-2 did not alter FM1-43FX uptake ( Figure S4A , Figure S4B , Figure S4C ) , indicating that mechanotransduction-associated events appear intact with these modulators . In addition , examination of the neuromasts in sentinel mutants by light microscopy ( compare Figure 8A and 8C to Figure 8B and 8D ) reveals that hair cells are organized in the stereotypical rosette pattern found in wildtype animals . Together these results suggest that these modifiers do not act by blocking hair cell transduction or slowing development . To test whether these modifiers alter drug entry , we evaluated whether fluorescently-tagged aminoglycosides [39] enter hair cells in the presence of modifiers . Both the aminoglycosides gentamicin ( Figure 8E and 8F ) and neomycin ( not shown ) tagged with Texas Red fluorophore enter sentinel hair cells with a rapid , 45-second , exposure . Similarly , PROTO-1 and PROTO-2 did not alter labeled gentamicin uptake ( Figure S4D , Figure S4E , and Figure S4F ) . While these results do not rule out subtle changes in aminoglycoside uptake , they do show that there are no dramatic differences that might account for the broad range of protection seen . Hence , it appears most likely that modifiers affect steps in toxicity that occur after aminoglycoside entry . Although the initial mechanism of hair cell death induced by aminoglycosides and cisplatin may be quite different , the later general cell death events are thought to be similar . To test whether these modulators alter cisplatin toxicity , we tested the effects of a range of cisplatin doses on sentinel mutants and on animals treated with PROTO-1 . The response of sinusoidal sentinel mutants to cisplatin mirrored wildtype strains and siblings with wildtype body shape ( Figure 9A ) . Thus , sentinel mutants are not protected against cisplatin-induced hair cell toxicity . Similarly , PROTO-1 did not protect against cisplatin-induced cell death ( Figure 9B ) . The observation that sentinel mutants and fish exposed to PROTO-1 are relatively resistant to aminoglycoside-induced cell death but remain normally sensitive to cisplatin-induced cell death suggests that general cell death mechanisms are intact . We hypothesize that the sentinel mutation and PROTO-1 may abrogate aminoglycoside targets or early events in aminoglycoside-induced cell death that are not shared by cisplatin-induced cell death . Finally , we sought to determine whether modifiers we discovered in the zebrafish lateral line hair cell assay also confer protection to hair cells in the murine inner ear . While mutants for the mouse ortholog of sentinel are not yet available , a validated in vitro mammalian preparation of the mature mouse utricle has been used extensively to test protection of chemical modifiers [40]-[42] . We used the mouse utricle preparation to compare hair cell loss due to neomycin exposure between control utricles and utricles pretreated with PROTO-1 or PROTO-2 . Figure 10 shows the neomycin dose-response relationship of striolar and extrastriolar hair cells in control utricles and utricles pretreated with PROTO-2 . A two-factorial ANOVA ( compound pretreatment×neomycin ) showed significant protection using PROTO-2 ( p<0 . 0001 ) in both the striolar and extrastriolar hair cell populations . PROTO-1 protection against neomycin was tested at 4 mM neomycin and showed significant striolar ( p<0 . 0001 ) , but not significant extrastriolar , protection . These results suggest that modifiers that can be rapidly identified and validated in the zebrafish lateral line system can have application in understanding ototoxicity in the mammalian inner ear .
Mechanosensory hair cells in the inner ear are susceptible to a wide variety of environmental insults . However , the large amount of variation in hearing and balance problems resulting from environmental or age-related challenges among normal individuals is neither well documented nor well understood . There is even large variance among individuals with the A1555G mutation in the mitochondrial 12S rRNA that increases susceptibility to neomycin toxicity [15] . We hypothesize that alterations in unidentified components of the network of cellular pathways involved in cell death and cell survival would confer resistance to ototoxic compounds . The absence of secondary phenotypes in some of our mutants supports the idea that variation affecting drug response can exist without other outward manifestation . Identification of the human orthologs of these genes may provide candidates involved in the variability underlying human hearing and balance disorders . Our data suggest that hair cell death after neomycin treatment can involve multiple signaling pathways . Several mutations confer only partial protection against neomycin exposure . Although in some cases this might result from mutations that cause only partial loss of function , in the case of sentinel we suspect that the mutation is a functional null . The mutation introduces a stop codon early in the coding sequence and before the highly conserved regions . In addition , mRNA levels are reduced in sentinel mutants , suggesting nonsense-mediated decay . Together these observations suggest that gene function is completely lost , while protection against hair cell loss is only partial . Similarly , only partial protection is observed after treatment with maximal doses of PROTO-1 or PROTO-2 . The idea that there are several possible responses to aminoglycosides is consistent with our previous observed variations in ultrastructural changes after aminoglycoside exposure [43] . The sentinel mutation also genetically distinguishes between aminoglycoside-induced and cisplatin-induced death; mutant animals are resistant to neomycin but still sensitive to cisplatin . Both aminoglycoside and cisplatin exposure have been proposed to result in oxidative stress [11] , [44] , raising the possibility that ototoxic compounds share similar mechanisms . If such shared mechanisms occur , the sentinel gene product must act upstream of these events . Treatment with PROTO-1 also offered no protection against cisplatin , suggesting that its cellular target acts specifically during aminoglycoside toxicity . Inactivation of sentinel and treatment with PROTO-1 similarly alter the response of hair cells to neomycin treatment . Both modulators offer only partial protection against neomycin , offer no protection against cisplatin , and do not affect entry of FM1-43 or labeled aminoglycoside . Together these results suggest they work in common pathways . To test this idea , we performed epistasis experiments treating wildtype and mutant animals . While the effects of sentinel and PROTO-1 are not additive , there is a small but significant increase in protection when combined , suggesting that they may be accessing different cellular pathways to promote cell survival . Understanding similarities and differences among possible pathways will await the identification of the cellular targets of PROTO-1 . The identification of the sentinel gene highlights one strength of forward genetic screening , as it would be difficult or impossible to choose this gene a priori as a candidate regulator of mechanosensory hair cell death . No functional information is known about any of the sentinel orthologs . The only functional domain of note , the C2 domain , has been associated with calcium regulation and interaction with phospholipid membranes in signaling proteins such as protein kinase C or membrane trafficking proteins like Synaptotagmin [32] . However , the function of this domain has been demonstrated in only a few of the many proteins that contain it . Intriguingly , the D . melanogaster ortholog CG18631 was identified in a comparative bioinformatics screen as being associated with compartmentalized cilia-bearing organisms suggesting it may have a role in regulation of cilia [45] . Other members of this group include molecules related to intraflagellar transport ( IFT ) proteins and Bardet-Biedl syndrome ( BBS ) -related proteins associated with auditory function . In addition , the C . elegans K07G5 . 3 ortholog is enriched in ciliated neurons by SAGE analysis and localizes to ciliated sensory neurons [46] . Hair cells of the zebrafish lateral line and inner ear are also ciliated , bearing a microtubule-based kinocilium in addition to the actin-based stereocilia either throughout life ( lateral line and vestibular epithelia ) or during development ( auditory epithelia/cochlea ) . However , the broad distribution of sentinel mRNA and lack of hair cell functional defects in mutants suggest that the gene product does not have a role specific to hair cells . In addition to identifying possible therapeutic approaches , unbiased small molecule screening may reveal new molecular pathways that regulate hair cell death . This approach has been taken previously in a small molecule screen for compounds affecting zebrafish blood development; by identifying several compounds that affected prostaglandin metabolism , PGE2 was revealed as a regulator of haematopoiesis [47] . PROTO-1 and PROTO-2 are related benzothiophene carboxamides , suggesting that they may have the same molecular targets . Other benzothiophene carboxamides have previously been identified as HIV inhibitors , having effects on casein kinase , calcineurin and p53 [48]–[50] . Further work will be needed to determine whether any of these pathways modulate hair cell death . Perhaps the most important contribution here is the suggestion that our screens can serve as templates for other research programs to identify other gene-drug interactions . Individuals respond remarkably differently to environmental exposures and drug treatment in most disease conditions . Efforts to understand population variation have centered on epidemiological and pharmacogenomic approaches [51] . However there are only a few cases in which the genes responsible for this phenotypic variability have been identified , such as for VKORC1-warfarin response or PON1-organophosphate toxicity [52] , [53] . Genetic analysis may provide a systematic method to identify new molecules involved in cellular responses to drugs or disease .
Zebrafish embryos ( Danio rerio ) were produced by paired matings of adult fish in the University of Washington zebrafish facility by standard methods [54] . The *AB and WIK wildtype strains are maintained individually as inbred lines . Three to six-week-old CBA/CaJ mice were obtained from the Jackson Laboratory ( Bar Harbor , ME ) and maintained in the University of Washington Animal Care facility . All animal protocols were approved by the University of Washington Animal Care Committee . Larvae were transferred manually to baskets in 6-well culture plates containing defined E2 embryo medium . Baskets were constructed from the tops of 50 ml Falcon tubes in which the center of the lids were replaced with meshing . All treatment and wash volumes are 6 ml unless otherwise indicated . Hair cells of larvae were labeled with the following dyes: 1 ) FM 1-43FX ( n- ( 3 , 3-ammoniumpropyl-dimethyl ) ammoniumpropyl ) -4- ( 4- ( dibutylamino ) styryl ) pyridinium trichloride ) , an aminated derivative of FM1-43 ( n- ( 3-triethylyammoniumpropyl ) -4- ( 4- ( dibutylamino ) -styryl ) pyridinium dibromide Invitrogen Molecular Probes , Eugene , OR ) by immersing free swimming larvae in 3 µM FM 1-43FX in embryo medium for 30 or 45 s , followed by three successive rinses in embryo medium; 2 ) Yo-Pro-1 ( Invitrogen Molecular Probes ) at 3 µM for 1 hour followed by 3 rinses to selectively stain hair cell nuclei; or 3 ) DASPEI ( 0 . 005% final concentration , ( 2-{4- ( dimethylamino ) styryl}-N-ethylpyridinium iodide , Invitrogen Molecular Probes ) in the final 15 minutes of the recovery period , and rinsed twice to brightly label mitochondria-rich hair cell cytoplasm . Larvae were anesthetized with MS222 ( 3-aminobenzoic acid ethyl ester , methansulfoneate salt , Sigma-Aldrich , St . Louis , Missouri ) at a final concentration of 0 . 02% prior to imaging . Neomycin ( Sigma-Aldrich , catalog no . N1142 ) was diluted in defined E2 embryo medium . Animals were treated with drug or embryo media ( mock-treated controls ) for times indicated , subsequently washed rapidly three times in fresh embryo medium and allowed to recover for one hour . For cisplatin treatment , zebrafish larvae were exposed to 0–400 µM cisplatin ( Sigma-Aldrich , catalog no . P4394 ) for 4 hours , rinsed several times in embryo medium and held 3 hours in the same media prior to DASPEI staining and visualization . Larvae were stained with Yo-Pro-1 and FM 1–43 and then dispensed into 96-well glass bottom plates ( Nunc , Rochester , New York ) containing embryo medium ( 1–2 fish per well ) . Drug-like compounds from the Diverset E library ( ChemBridge , San Diego , California ) , dissolved in 0 . 05% DMSO to a final concentration of 10 µM , were aliquoted into each well . Fish were incubated at 28 . 5°C for 1 hour . Neomycin was then introduced into each well at a final concentration of 200 µM and fish were incubated for an additional hour . Larvae were anesthetized with MS222 for immobilization . Visual assessment of hair cell integrity was performed in vivo using an inverted epifluorescent microscope . This allowed examination of the whole animal on the side of its body facing the objective and thus rapid evaluation of many neuromasts ( ∼20 ) . In each row of the 96-well plate both positive ( neomycin treated only ) and negative ( no neomycin ) control animals were used for comparison to compound treatment . The entire plate of 96-well plate with 80 test wells and 16 positive or negative control wells was evaluated within one hour . Although intermediate responses were observed for some drugs , only those exhibiting robust protection were pursued for continued evaluation at this time . To quantify changes in the hair cell response , hair cell survival was determined by counting the surviving hair cells from four neuromast , SO1 , SO2 , OC1 and O1 for 10–20 fish ( i . e . 40–80 neuromasts ) . The percentage of surviving hair cells following treatment was calculated relative to mock-treated controls ( no drugs or neomycin exposure ) . Determination of the minimum inhibitory concentration ( MIC ) and the minimal bactericidal concentration ( MBC ) of neomycin alone and in the presence of 10 µM PROTO-1 or PROTO-2 were performed at the Clinical Laboratory of Microbiology at the University of Washington Medical Center as described by the National Clinical and Laboratory Standards Institute [55] , [56] . Adult males from the *AB wildtype strain were mutagenized with 3 mM ethylnitrosourea ( ENU ) using standard procedures [27] ) . To assess the effectiveness of the mutagenesis , we performed a specific locus test of mutagenized males with homozygous nacre females; mutation of the nacre ( mitfa ) gene results in lack of pigment , which is readily apparent [28] . The ratio of progeny with a nacre-like pigment phenotype to total progeny was 1/300 . Mutagenized males were then crossed to wildtype *AB females to produce F1 progeny . F2 families were derived from pairwise matings of F1 progeny of different mutagenized males . For each family screened , three to twelve F2 pairs were crossed and their progeny were examined for altered aminoglycoside response . Neomycin doses of 25 µM or 200 µM were used to screen for heightened susceptibility or protection , respectively . Ten neuromasts were evaluated on each fish for DASPEI staining and each neuromast was assigned a score of 0 for no/little staining , 1 for reduced staining , 2 for full staining [21] , resulting in a final score of 0–20 for each fish . Scores were averaged and normalized to mock-treated controls . For initial analysis , 12–50 fish were assessed for typical and atypical responders ( i . e . 120–500 neuromasts ) . Results were tabulated and chi-squared analysis was done to identify potential mutant strains of interest . Putative mutants were retested to confirm phenotype , outcrossed to *AB fish and tested again in the next generation to confirm transmission . Heterozygous mutant carriers were outcrossed to the wildtype zebrafish from the polymorphic WIK strain for mapping . Hybrid *AB/WIK carriers of the hair cell modulator were then identified and crossed to produce progeny for marker analysis . At 5 dpf larvae were exposed to neomycin as described for the initial screen . To ensure accurate phenotyping , only individuals with the highest and lowest DASPEI staining scores after 200 µM neomycin treatment were retained as mutant and wildtype , respectively . For bulk segregant analysis , DNA was pooled from 20 wildtype or mutant individuals . Distribution of markers was compared to DNA from fin clips of *AB/WIK parents and founder grandparents . Microsatellite markers for each chromosome [31] were amplified by PCR and evaluated for cosegregation with mutant phenotypes . Linked markers were further evaluated with individual DNAs from 694 mutant fish and 234 wildtype fish ( including both heterozygous and homozygous wildtype siblings ) . After determining initial linkage to chromosome 23 , fine mapping identified Z3794 and Z44679 as flanking markers . A contiguous genomic sequence was then assembled using whole genome shotgun trace sequences produced by the Zebrafish Sequencing Group at the Sanger Institute ( http://www . sanger . ac . uk/Projects/D_rerio/ ) . Additional markers were developed to better define the linked region in sentinel mutants based on genomic sequence . snp3 amplifies a single nucleotide polymorphism and sat3334 is a sequence length polymorphism . They are amplified by the primers: RNA was isolated from whole embryos at 62 hpf using Trizol according to manufacturer's specifications ( Invitrogen , Carlsbad , California ) . Oligonucleotide primers were designed based on in silico genomic sequence . cDNA was synthesized using First Strand cDNA synthesis kit ( Invitrogen ) using oligo DT primers . The following primer pairs were used to amplify portion of cDNA spanning the recombination breakpoints: Amplified products were cloned into pCR4 vectors using Topo TA cloning kit ( Invitrogen ) . cDNA and genomic regions were sequenced from the vector T3 or T7 sites using Big Dye terminator v3 . 1 cycle sequencing chemistry ( Applied Biosystems , Foster City , California ) . Zebrafish cDNAs were aligned to known ESTs , cDNAs and genomic sequence from this region using Sequencher software ( Gene Codes , Ann Arbor , Michigan ) . BLAST alignments of our cDNA sequences align with predicted cDNA ( Genbank XM_693709/gi:125851476 ) amino acids 75-1040 . Orthologs were identified from Genbank using BLAST and the corresponding predicted protein sequences were aligned with the Danio rerio predicted protein ( XP_698801/gi:125851477 ) : Mus musculus ( NP_758478 . 1/gi:26986583 ) , Homo sapiens ( NP_001073991/gi:122937494 ) , Pan troglodytes ( XP_001159814 /gi:114593231 ) , Canis familiaris ( XP_536233/gi:73951827 ) , Bos taurus ( XP_595408/gi:119894226 ) , Monodelphis domestica ( XP_001369774/gi:126331991 ) , Gallus gallus ( XP_420777/gi:118090694 ) , Caenorhabditis elegans ( NP_492026/gi:U17508151 ) , Drosophila melanogaster ( NP_611229 and NP_611230/gi:24654454/28573534 ) , Aedes aegyptii ( EAT41051/gi:108876826 ) , Trichomonas vaginalis ( XP_001323414/gi:123480792 ) , Paramecium tetraurelia ( CAK70738/gi:124405296 ) . ClustalW multiple sequence alignment software was used to align predicted proteins of orthologous genes using the Gonnet 250 matrix [57] . We used the Phylip 3 . 66 phylogeny software [58] to create a bootstrapped data set from the original alignment using Seqboot , then Protml to evaluate these datasets using the maximum likelihood method with a Jones-Taylor-Thorton model of amino acid substitution . A consensus tree was determined with Consense software by extended majority rule . Phylogenetic trees were draw with TreeView software [59] . Protein motif searching was performed using the Eukaryotic Linear Motif server ( elm . eu . org ) . 4 . 4 ml of gentamicin sulfate ( Sigma-Aldrich , 50 mg/ml ) and 0 . 6 ml succinimidyl esters of Texas Red ( Molecular Probes , Eugene , Oregon; 2 mg/ml in dimethyl formamide ) were agitated overnight to produce the conjugate solution [39] . The conjugated solution was diluted in embryo media to a final concentration of 200 µM gentamicin . Because neomycin contains six amino side groups , neomycin conjugation was performed similarly except that the ratio of neomycin to Texas Red was adjusted to 3:1 to ensure that on average one molecule of dye or less labeled each aminoglycoside molecule . To assess aminoglycoside entry , 5 dpf larvae were immersed in aminoglycoside-Texas Red conjugate for 45 seconds and rinsed in embryo medium four times before immediate imaging . Images were collected using Zeiss LSM5 Pascal confocal microscope . Z-stack images of neuromasts were collected . Utricles were dissected and cultured in basal medium EAGLE supplemented with Earle's balanced salt solution and 5% fetal bovine serum following established procedures [40] . Neomycin sulfate stock solution ( Sigma-Aldrich ) prepared in sterile water was added directly to culture wells at the desired concentrations . The utricles were incubated for 4 hours in the compounds diluted with 0 . 05% DMSO or 0 . 05% DMSO alone for controls followed by a 24 hour incubation with neomycin . Utricles were fixed for 1 hour at 4°C in 4% paraformaldehyde in phosphate buffer . Following fixation , otoconia were removed by gently “washing” the surface with buffer through a 26 gauge syringe needle . Utricles were then incubated in blocking solution ( 2% bovine serum albumin , 0 . 4% normal goat serum , 0 . 4% normal horse serum and 0 . 4% Triton-X in PBS ) for 3 hours at room temperature . Hair cells were double labeled in whole-mount preparations with a monoclonal antibody against calmodulin ( Sigma-Aldrich ) and polyclonal antibody against calbindin ( Chemicon , Temecula , California ) at 4°C diluted in blocking solution , 1∶250 . The utricles were then rinsed and incubated for 2 hours at room temperature in secondary antibody diluted in blocking solution with biotinylated horse anti-mouse IgG ( 1∶200 ) and Alexa 594-conjugated goat anti-rabbit IgG . Utricles were mounted with Fluoromount-G ( EMS , Hatfield , Pennsylvania ) and coverslipped . The density of mouse utricular hair cells was determined by counting the number of hair cells in three randomly chosen nonstriolar regions and the number of striolar hair cells in three randomly chosen striolar regions from each utricle . Counts were made at high magnification in areas of 900 µM2 , converted to density , and averaged over the three sampled areas of each region for each utricle . Ten utricles were analyzed in this way for each treatment group . Data were normalized relative to mock-treated controls ( no PROTO drug , no neomycin ) .
|
Loss of sensory hair cells in the inner ear is observed in the majority of hearing and balance disorders , affecting the health of more than 600 million people worldwide . Exposure to environmental toxins and certain pharmaceutical drugs such as aminoglycoside antibiotics and some cancer chemotherapy agents account for many of these hearing and balance problems . Variation in the genetic makeup between individuals plays a major role in establishing differences in susceptibility to environmental agents that damage the inner ear . Using zebrafish larvae , we developed a screen to uncover genes leading to differences in antibiotic-induced death of hair cells and to identify compounds that protect hair cells from damage . The combination of chemical screening with traditional genetic approaches offers a new strategy for identifying drugs and drug targets to attenuate hearing and balance disorders .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/animal",
"genetics",
"neuroscience/sensory",
"systems",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"genetics",
"and",
"genomics/disease",
"models",
"otolaryngology/ear",
"pathologies"
] |
2008
|
Identification of Genetic and Chemical Modulators of Zebrafish Mechanosensory Hair Cell Death
|
Neuronal activity in cortex is variable both spontaneously and during stimulation , and it has the remarkable property that it is Poisson-like over broad ranges of firing rates covering from virtually zero to hundreds of spikes per second . The mechanisms underlying cortical-like spiking variability over such a broad continuum of rates are currently unknown . We show that neuronal networks endowed with probabilistic synaptic transmission , a well-documented source of variability in cortex , robustly generate Poisson-like variability over several orders of magnitude in their firing rate without fine-tuning of the network parameters . Other sources of variability , such as random synaptic delays or spike generation jittering , do not lead to Poisson-like variability at high rates because they cannot be sufficiently amplified by recurrent neuronal networks . We also show that probabilistic synapses predict Fano factor constancy of synaptic conductances . Our results suggest that synaptic noise is a robust and sufficient mechanism for the type of variability found in cortex .
Cortical neurons respond to repeated presentations of the same stimulus in a remarkably idiosyncratic way , and no identical responses are observed twice [1] , [2] , [3] , [4] . Although the spike count responses are on average reproducible ( ibid . ) , they display high variability . It is well established that in evoked conditions the variance of the spike count over time windows of a few hundreds of milliseconds is closely proportional to the mean spike count , which in turn implies that the Fano factor –variance to mean ratio– is approximately constant as a function of firing rate [2] , [3] , [5] , [6] , [7] . Approximate Fano factor constancy is not only found for a small range of evoked firing rates , but rather it holds for the whole observable dynamic firing range of cortical neurons , which covers from a few to hundreds of spikes per second [2] , [5] , [7] , [8] . In addition , Fano factor constancy is not just a property of the distribution of a neuronal population , but every single neuron in the population displays Fano factor constancy over its whole dynamical range [3] , [7] . This single-cell , whole dynamical range property is referred here to as Poisson-like firing , in analogy to the Poisson process , whose Fano factor is rate-independent . Theoretical neuronal network models often invoke a balance between excitatory and inhibitory inputs to describe the high spiking variability observed in cortex , a mechanism that leads to complex or chaotic firing behavior that is Poisson-like at low firing rates [3] , [9] , [10] , [11] . These networks can also be adapted to display bistable dynamics to model working memory tasks , and it has been shown that they can generate Poisson-like firing during persistent activity even at moderately high firing rates [12] , [13] , [14] , [15] , [16] , [17] . Relatively less attention , however , has been paid to study the origin of Poisson-like variability over a full continuum ( non-discrete ) of firing rates ranging from a few to hundreds of spikes per second [18] , as observed experimentally in sensory areas [3] , [9] , [10] , [11] . As we show below , although balanced excitatory and inhibitory networks are well suited to generate Poisson-like firing at low rates , balanced networks fire with low variability as their firing rate increases continuously unless connectivity parameters are fine-tuned [12] or inputs to the network are themselves Poisson-like [18] . Introducing Poisson-like inputs to obtain Poisson-like outputs is a valid solution to the problem of how variability is generated in cortex . However , this solution might seem unsatisfactory because it does not address the problem of how and where Poisson-like inputs are originated in the first place . Moreover , the notion of Poisson-like inputs to sensory cortical areas fails to find strong experimental support , because LGN spike train inputs to V1 display Fano factors decreasing by two-fold or more as a function of firing rate [19] , [20] . In summary , although there is a solid understanding of how Poisson-like variability arises from the chaotic balanced dynamics of neuronal networks at low rates [3] , [9] , [10] , [11] , or in discrete high rate persistent states [12] , [13] , [14] , [15] , the mechanisms underlying single-cell Poisson-like variability over a broad continuum of rates have not yet been elucidated . Other sources of noise in neuronal networks that have so far been largely neglected might be responsible for cortical spiking variability , in particular at high firing regimes . A well-documented source of variability in the central nervous system is synaptic transmission failures [21] , [22] , [23] . Synaptic vesicle recovery and release has complex time history dependences [24] , [25] , [26] , [27] , but at the finest level synaptic transmission is fundamentally probabilistic [21] , [22] . In this paper , we show that amplification of synaptic noise generated by realistically small postsynaptic potentials through recurrent connections is sufficient to generate Poisson-like spiking over several orders of magnitude in the firing rate . Other variability-inducing mechanisms , such as random synaptic delays or intrinsic spike generation jittering [28] , [29] , constitute negligible sources of spiking variability at high rates .
To understand under what conditions Poisson-like firing can be generated , we simulated a single leaky integrate-and-fire neuron with various types of input white noise . We first considered the case where the mean input variance is constant as the mean input rises ( Fig . 1 , dashed lines ) . In correspondence to the responses of sensory cells to stimuli with increasing intensity ( e . g . contrast in V1 [5] , [30] ) , boosting the mean input drive of the neuron increases its firing rate and mean membrane potential ( Fig . 1a , d ) . Although the Fano factor is close to one at low input drive ( Fig . 1b , dashed line ) , when the input drive is above the threshold ( vertical line ) the Fano factor drops to very small values . This in turn implies that the Fano factor decreases to low values at high rates ( >50 Hz ) ( Fig . 1c , dashed line ) . At low rates the mean current is below the threshold current ( sub-threshold regime ) and spiking is induced by membrane potential fluctuations ( Fig . 1e , light green trace ) , while at high rates the mean current is above the threshold current ( supra-threshold regime ) and spiking is mainly induced by voltage threshold crossings around the mean membrane potential trajectory ( dark green trace ) , leading to a very regular spike train with low Fano factor . As constant input noise was not able to keep high the Fano factors at high rates , we considered next the scenario where the input variance grows in proportion to the mean input current ( ) . This manipulation corresponds to the case where inputs are Poisson-like because a rate-independent Fano factor in the input spike trains implies proportionality between input variance and mean [31] . In this scenario , the output Fano factor remains approximately constant even at very high rates ( >100 Hz ) ( Fig . 1c , solid line ) . Like the constant input noise case , at low rates spikes are induced by small membrane potential excursions around the mean membrane potential trajectory that sporadically reach spiking threshold ( Fig . 1e , light blue trace ) . However , unlike the constant input noise case , at high rates the membrane potential undergoes large fluctuations that cause bursts of spikes followed by silence periods ( dark blue trace ) , leading to high spiking variability even at elevated firing . The two scenarios described above are a priori possible in recurrent networks . If the Fano factor of the afferent spike trains to a neuron in the network decays with firing rate as ( as in dashed line of Fig . 1c ) , the input noise becomes approximately constant because [31] . In this scenario we find that the Fano factor of the output spike train is , and therefore the Fano factor displays the same firing rate scaling as that in the inputs . If , in contrast , the Fano factor of the afferent spike trains is constant with firing rate , , then the input noise becomes Poisson-like because . In this scenario the Fano factor of the output spike train is approximately independent of firing rate , ( solid line in Fig . 1c ) , like the Fano factor in the input , and it again displays the same firing rate scaling as that in the inputs . Therefore the two scenarios are potentially self-consistent in the sense that the same type of variability that is introduced in the inputs is recovered in the outputs . We sought to determine what type of neuronal variability is self-consistent and stable in recurrent networks , that is , whether the Poisson-like input noise scenario or the constant input noise scenario described above is stable in a recurrent network . We simulated a balanced recurrent spiking network [9] , [10] , [32] that generated strong excitatory and inhibitory currents . We also stimulated the network with external inputs . The external inputs were designed to be non-Poisson-like because the central question is whether Poisson-like variability can be self-generated by neuronal networks when the external inputs are not in the same Poisson-like family ( Fig . 2a , top ) . The results shown below correspond to networks with non-Poisson-like inputs modeled with constant variance to enforce the experimental constraint that the input Fano factor decreases with firing rate [19] , [20] . As the input drive increases , the mean firing rates ( Fig . 2b , dashed lines ) for the excitatory ( red ) and inhibitory ( green ) populations increase accordingly . At low firing rates the Fano factor is high ( Fig . 2c , d , dashed lines ) , a standard property in neuronal networks in the balanced regime [9] , [10] , [32] . However , the Fano factor drops monotonically to very low values as the mean population rate increases ( above 50 spikes per second ) . As long as the network holds a single stable state ( see Methods ) , the breakdown of approximate Fano factor constancy at high rates occurs regardless of the connectivity matrix of the network , including sparsely , densely and fully connected networks ( see below ) ; it also occurs regardless the overall connection strength , that is , the type of synaptic strength scaling used as the network becomes large , and regardless of the intensity of the constant noise . If the network is multi-stable , transitions between different states can exist , but conditioned on each state , the Fano factor is very low . These results are shown analytically for an even broader family of spiking neuronal networks in the Methods ( see eqs . ( 11 ) and ( 22 ) ) . In particular , for sparse and randomly connected balanced networks the dynamics displays elevated Fano factors at low rates [9] , [10] , [33] , but at high rates ( >50 spikes per second ) Fano factors fall off ( Fig . 3a ) . This is because the neurons in the network enter in the supra-threshold regime , in which firing is mainly induced by mean membrane potential threshold crossings; as a result variability becomes progressively lower as firing rates increases ( Fig . 3a ) . The same results hold when the network is randomly but densely connected and when the network is fully connected ( see Fig . 2d , dashed lines ) . In addition , if the value of the reset membrane potential is raised , the variability increases at low rates , but the Fano factor does not remain steady at high rates ( Fig . 3b ) . In conclusion , the previous analysis manifests that under a broad range of situations and network designs , the constant input noise scenario ( corresponding to Fano factor decreasing monotonically with firing rate ) becomes the only stable scenario in recurrent networks . The previous results suggest that , at least at high rates , additional sources of variability are required to account for Poisson-like spiking . A well-known source of noise in cortex is probabilistic synaptic transmission . Neurotransmitter release at a synapse upon arrival of an action potential is fundamentally stochastic [21] , [22] and thus it will result in spiking variability ( Fig . 2a , bottom ) . However , it is not obvious that this source of noise can account for most of the spiking variability observed in cortex . The average number of contacts that a cortical neuron makes on postsynaptic targets is 2–6 [22] and synaptic release is independent across contacts . Therefore it could occur that synaptic noise is mostly averaged out , leaving very little room for its contribution to spiking variability . Whether strong amplification of synaptic noise can be achieved with realistic neurophysiological parameters and whether probabilistic synapses can give rise to Poisson-like variability is unknown . We studied a balanced recurrent network with probabilistic synapses where the probability that an action potential generated a post-synaptic current underwent stochastic short-term-depression ( STD ) ( see Methods ) . The network can generate high spiking variability for its full dynamical range when the connections are sufficiently strong even when the external input to the network is noiseless ( Fig . 2c , solid lines; see also Fig . 2e , left panel ) . In the network , a presynaptic spike caused postsynaptic potentials between 0 . 2 and 1 mV on average , within the neurophysiological range [34] , [35] . Therefore weak , independent noise across synaptic contacts can be amplified by strong synapses to generate high fluctuations at the spiking level . The network was not only able to generate high variability , but the Fano factor was approximately constant for at least two orders of magnitude range in firing rate ( Fig . 2d , solid lines ) . Importantly , the Fano factor was not only constant on average over the population ( white dots ) , but also individually for each neuron ( black dots ) as a function of firing rate . The neurons' Fano factors increase with the strength of the recurrent connections , but in all cases they remain constant at high rates ( shown analytically for general neuronal networks in the Methods ) . The Fano factor was high and sustained for a broad region of scaling factors of the synaptic strength and input drives ( Fig . 4 , top panel ) , but this region vanished at moderately high input drives when the network lacked probabilistic synapses ( lower panel ) . These results hold when synapses display STD dynamics as long as synaptic transmission does not saturate for a very broad range of firing rates [23] , [36] ( see Methods ) . When STD is modeled without stochastic release [26] , [27] the Fano factor decreases monotonically to very low values at high rates ( Fig . 3c ) . Although high reset and STD are required in some models of delayed persistent activity to generate high variable binary attractor states [13] , these mechanisms do not guaranty high variability for a broad continuum of rates , as it has been shown above ( Fig . 3b , c ) . Finally , Poisson-like variability also holds for a stochastic model of synaptic transmission without STD ( Fig . 3d; see eqs . ( 11 ) and ( 22 ) in Methods ) . In summary , the probabilistic nature of synaptic transmission is sufficient to robustly generate Poisson-like firing at high rates . At elevated rates ( blue dots in Fig . 2b–d ) , as it is the case at low rates , the network generates strong excitatory and inhibitory currents ( Fig . 2e , red and green traces in the middle panel , respectively ) that approximately cancel , leading to a balanced net input current ( black trace ) that wanders around zero ( yellow line ) . However , the net input current is on average above zero and close to the threshold current ( mean ; threshold current ) . For the rightmost point in the solid lines of Fig . 2b–d , the mean current is supra-threshold . As it has been shown for single neurons ( see Fig . 1 ) , supra-threshold currents or currents around threshold generate low Fano factors unless the input noise is Poisson-like . Here , at elevated rates the net current is close or within the supra-threshold regime , and therefore the network must have generated spontaneously a current that is in the Poisson-like family ( see next section ) . Further support for our theory arises from the puzzling dependences of other statistical measurements of variability with firing rate . It is well established that the coefficient of variation ( ) of the inter-spike-intervals ( ISIs ) of cortical neurons ( s . d . to mean ratio ) decreases at high firing rates [3] , [4] . The rate dependence of seems to be at odds with the Fano factor constancy at the same high firing rates . In fact , if spike trains were renewal processes , one would expect that the Fano factor were equal to . Although renewal point processes with absolute spiking refractory periods could explain the drop of at very high rates [4] , they cannot explain why the Fano factor does not decay at high rates in the same way . In our recurrent networks with probabilistic synapses , the increases with the mean ISI ( Fig . 2f ) , implying that it decreases as a function of the firing rate . This is so even when the Fano factor remains approximately constant for the whole range of firing rates , particularly at high rates ( see Fig . 2d ) . The reason for this behavior is that the network dynamics generates temporal correlations that make the spike trains non-renewal , with experimentally consistent ISI distributions and auto-correlations functions ( Fig . 2g , h ) . Therefore , a network effect that cannot be understood at the single neuron level gives rise simultaneously to approximate Fano factor constancy and the drop of at high rates ( equivalently , at short mean ISIs ) . Finally , although no fit of the experimental data was performed , the dependence of the as a function of the mean ISI followed well the values and the mean ISI dependence previously reported [4] , with values close to one above mean ISIs of 30 ms , and a reduction of variability up to a value of around 0 . 6 at mean ISIs shorter than 20 ms . To understand how Poisson-like firing arises from networks with probabilistic synapses , we used a simplified network model where transmission probability is time-independent ( Fig . 5a; see Methods ) . A neuron in the network ( pre ) receives a barrage of spikes per presynaptic neuron with spike count and variance and generates an output spike train whose spike count has variance . This spike train in turn evokes post-synaptic currents ( PSCs ) with variance on postsynaptic cells ( post ) . The evoked PSCs are replicas of the same presynaptic spike train that has been diluted by a fraction p , corresponding to the probability of synaptic transmission . There are two competing forces that affect the variability of the spike trains and series of PSCs ( Fig . 5b ) . The first one is the integration step of the neuron , which tends to lower the input variance . And the second one is the probabilistic synaptic step , which increases the variance . These two forces have to cancel out precisely when the network reaches equilibrium , because at equilibrium should equal . More precisely , it can be shown ( see Methods ) that is proportional to the input variance at fixed input firing rate ( dashed red line , Fig . 5c ) and that the effect of probabilistic synapses is shifting this line upwards regardless of the value of the input variance ( solid red line ) . The point at which the input and output variances are the same ( red dot ) corresponds to the equilibrium state of the network . The crucial question is to determine how this equilibrium point depends on the rate of the network . If the firing rate of the network increases , the crossing point moves at higher values linearly with the population rate ( Fig . 5d ) . Because the spike count variance is proportional to the output variance at equilibrium ( see dashed line in Fig . 5c ) , the spike count variance increases linearly with population rate ( Fig . 5e ) . Therefore , the ratio between variance and mean in the spike count is constant in this network , leading to Fano factor constancy . The same Poisson-like generation mechanism takes place in more biophysically realistic networks ( Fig . 2 ) , and holds exactly for networks of spiking neurons with probabilistic synapses with constant transmission probability ( Fig . 3e; see Methods ) . This mathematical exercise ( see details in Methods ) shows that the presence of both excitation and inhibition is not strictly necessary for Poisson-like variability , since it is possible to obtain high sustained variability in large networks with pure excitation and sufficiently weak synapses . However , although balancing strong excitation with inhibition is not required per se for Poisson-like variability at moderately high rates , the presence of both excitation and inhibition is required to avoid runaway excitation in networks with more realistically strong excitatory synapses [9] , [37] ( see Fig . 2 ) . Finally , other types of noise , such as random synaptic delays [1] with arbitrary distributions or ion-channel noise that jitters randomly the timing of the evoked action potentials [38] , [39] do not lead to Poisson-like variability ( Fig . 3d; see also Methods ) . Although random synaptic delays , a broad static distribution of synaptic delays [40] , or spike jittering can improve the stability of Poisson-like variability , these types of noise are not sufficiently amplified by recurrent neuronal networks at high rates , and therefore they constitute a negligible source of noise . Because probabilistic synapses introduce multiplicative noise at the synaptic level , the membrane potential and synaptic conductances of neurons must show some characteristic statistical properties . We studied these statistical properties in recurrent networks of conductance-based spiking neurons in the high-conductance regime [11] , [41] . We found that the standard deviation of the neuron membrane potential is approximately constant as a function of firing rate ( Fig . 6a ) for both networks with ( full line ) and without ( dashed ) probabilistic synapses , consistent with experimental observations [5] , [30] . The fact that the constancy of the standard deviation of the membrane potential naturally arises in neuronal networks can be used as a justification of Gaussian rectification models of single cell spiking variability , where the standard deviation is assumed to be constant with firing rate [5] . Both excitatory and inhibitory synaptic conductances increased linearly with rate ( Fig . 6b ) . Interestingly , the Fano factor ( FF , variance to mean ratio ) of the conductances was approximately constant as a function of firing rate for networks with probabilistic synapses , but it was much smaller and decreasing rapidly with firing rate for networks without probabilistic synapses ( Fig . 6c ) . These results show that the size constancy of the membrane potential fluctuations arises as a result of the shunting effect of mean conductances on the conductance fluctuations [11] , [41] , [42] , while the FF constancy of the synaptic conductances is a natural consequence of the multiplicative noise introduced by probabilistic synapses . The FF constancy of synaptic conductances is an experimentally testable prediction of our theory .
We have shown that spiking networks endowed with probabilistic synapses lead to Poisson-like variability for several orders of magnitude in firing rate , in line with extensive experimental observations in sensory areas [2] , [3] , [4] , [6] , [7] . Poisson-like spiking variability naturally arises from the multiplicative nature of synaptic noise and its amplification through strong recurrent connections and does not require fine-tuning of the network parameters . The multiplicative noise implies that that the size of membrane potential fluctuations is relatively constant with firing rate while the size of synaptic conductance fluctuations grows in proportion to their means . Other sources of variability could also contribute to evoked cortical spiking variability . Experimentally uncontrolled external variables might artificially introduce spiking variability that is not a property of the system per se . However , even when eye movements are controlled [6] or paralyzed [43] , or when the statistical properties of the stimulus are fixed [44] , cortical neurons still respond with high Poisson-like variability at all registered neuron firing rates . Photon noise and intrinsic receptors' noise can partly explain cortical spiking variability [45] , but at high stimulus intensities this variability is unlikely to represent a major contribution . Internal variables such as attention and arousal might also be at place , but even when they are controlled experimentally , spiking responses are still highly variable [46] , [47] . Therefore , the hypothesis that variability is intrinsically generated by neuronal networks with probabilistic synapses is favored against other less specific alternatives in view of the very little explanatory power that the presence of uncontrolled external or internal variables has on the type of spiking variability that is observed in cortex . At the mechanistic level , a balance between strong excitatory and inhibitory inputs that sets the membrane potential below threshold has become the prevalent model for high cortical spiking variability [3] , [9] , [10] . Experimental evidence supports that cortical networks are in the balanced regime [48] . In evoked conditions , sensory stimulation drives individual cortical neurons to a state where the mean membrane potential increases with contrast and firing rate is high [5] , [30] . As it has been shown ( see Fig . 1 ) , in this condition Fano factors are low even in the balanced regime unless input spike trains are themselves Poisson-like , raising the question as to how input Poisson-like variability is generated in the first place and whether this type of inputs is realistic . Precise cancellation of the input currents can potentially clamp the membrane potential to a value below threshold for a broad range of firing rates , but this exquisite cancellation requires fine-tuning of the network parameters for very large networks [12] . It has also been suggested that to produce in vivo high spiking variability , presynaptic spikes need to be synchronous [49] , but it was unknown how large input variability caused by synchrony can be generated in recurrent networks . Previous models have also explored the role of synaptic noise in neuronal computations [50] , [51] , [52] , [53] , in up-down state transitions [54] and in the spiking variability of single cells or pairs of cells [50] , [55] , [56] , [57] , but the role of probabilistic synapses on Poisson-like variability in large recurrent networks or over a broad continuum of firing rates was not studied . As we have shown here , probabilistic synapses generate multiplicative noise that is amplified by recurrent connections without fine-tuning of the network parameters . This mechanism underlies a sufficient requisite for large multiplicative input fluctuations that guarantees Poisson-like spiking for several orders of magnitude in firing rate . Probabilistic synapses have also the potential to explain at the mechanistic level the origin of high spiking variability in a much broader context than the one that we have considered here . High activity states during delayed persistent activity in working memory tasks are characterized by high spiking variability [16] , [17] . Although bistable attractor networks have been shown to display high variability at both spontaneous and moderately high rate persistent activity states [12] , [13] , [14] , [15] , the contribution of probabilistic synapses to spiking variability in these networks has not been studied . We have shown that probabilistic synapses stabilize Poisson-like firing for a broad continuum of firing rates in single-attractor networks because this type of noise introduces multiplicative noise . Clearly , probabilistic synapses have also potential to account by itself for the high variability observed during persistent activity in working memory tasks . Therefore in future studies it will be important to elucidate the role of probabilistic synapses on spiking variability and stability of working memory states in bistable attractor networks . It has been recently shown that stimulus onset reduces the average Fano factor across a broad variety of cortical areas and conditions [58] , a reduction that is specific to the transition from spontaneous to evoked activity . This reduced variability has been hypothesized to arise because of the redirection of the system to a particular state configuration during stimulation [58] . It is important to realize however that despite the reduction of variability relative to spontaneous activity , the responses in evoked conditions are still highly variable and the Fano factor is approximately constant with neuron's firing rate , as it has been shown by many previous studies [2] , [3] , [4] , [6] , [7] . Simulated neuronal networks based on balanced inputs with weak multi-attractor states can account for the finding that variability is reduced at stimulus onset [12] , [33] , [59] , [60] , but they leave unanswered why the Fano factor remains approximately constant in a broad range of firing rates in evoked conditions . In the general condition as in the specific networks studied in those works , increasing the input will eventually guide neurons to the supra-threshold regime , where firing is due to quasi-deterministic membrane potential threshold crossings and Fano factors decrease with increasing firing rate ( see Fig . 1b ) . As we have demonstrated , balanced neuronal networks with probabilistic synapses can generate Fano factor constancy for a wide range of firing rates in evoked conditions even in the supra-threshold regime because synaptic noise is multiplicatively scaled up with firing rate . Finally , injecting noise in the brain with probabilistic synapses might seem harmful at a first glance . Therefore it can appear that we have presented a “solution” to the Poisson-like variability problem , but we have “created” a new one: boosting neuronal variability . However , noisy systems can have an advantage against deterministic systems in detecting sub-threshold stimuli [61] , learning more quickly [62] , and displaying larger memory capacity [63] . Injecting noise through probabilistic synapses is particularly relevant in view of the new computational capabilities that neuronal networks with Poisson-like firing acquire , allowing neuronal codes to be in the appropriate format to perform optimal cue combination [64] and sampling cortical states over the whole dynamical range [65] . Therefore , synaptic noise is not only a robust and sufficient mechanism for the type of variability found in cortex , but it can also provide cortical circuits with computational tools to perform probabilistic inference under noisy and ambiguous conditions .
We consider a network of leaky integrate-and-fire ( LIF ) neurons with cells , of which are excitatory and are inhibitory [10] , [66] , [67] , [68] . The membrane potential of neuron i below the spiking threshold obeys ( 1 ) where is the membrane capacitance , is the passive leak conductance and is the resting state potential . The membrane time constant of the neuron is defined as . The neuron emits a spike when the membrane potential reaches the threshold , after which the potential is reset to . The total synaptic current delivered to the neuron is ( 2 ) where the first term corresponds to the currents generated by other cells in the network , while the two last ones correspond to the current generated by external sources to the network . is the connectivity strength of contact k between the presynaptic neuron j and the postsynaptic neuron i . We typically consider 2–6 contacts per pair of connected neurons . Auto-synapses are not included in the network , i . e . for all i . The sum over m corresponds to the spikes times of each presynaptic cell j , denoted , . Each spike from neuron j can potentially generates a stereotyped current after a delay on the postsynaptic cell proportional to the synaptic kernel , such that ( 3 ) With this choice , the total charge injected in the neuron due to a presynaptic spike at time is determined by , where is the synaptic variable that specifies the amount of neurotransmitter released at contact k between postsynaptic neuron i and presynaptic neuron j at the time of the presynaptic spike . The dynamics of the synaptic variables are described by the end of this section . As synaptic kernel , we choose ( 4 ) where is the synaptic decay time constant of the postsynaptic current , which can be excitatory with time constant or inhibitory with time constant . The external current consists of a deterministic component ( mean drive ) , and a white noise process with variance . Here is a white noise process with zero mean and unit variance independent across neurons , i . e . and , where is the Kronecker's delta , and is the Dirac's delta function . We endowed synapses with a probabilistic transmission model where the synapses evoke successfully postsynaptic currents with a fixed probability upon presynaptic spike arrival if a vesicle is ready to be released , and the replenishment of the vesicle is stochastic with an exponential distribution over time [50] . This model is based on deterministic models of short-term-depression ( STD ) in vitro [26] , [27] . We further modified the deterministic models of STD for in vitro slices to incorporate the lack of depression observed at high rates in vivo [23] , [25] , [36] and in some neuronal population in vitro [69] as follows: upon vesicle release , a new vesicle is immediately ready to be released with the same probability , but with a lower neurotransmitter load . This model creates effectively a lower bound in synaptic efficacy , allowing for non-saturating responses at high firing rates . Other biophysical implementations of non-saturation and stochastic release at high rates ( ∼100–200 Hz ) are also possible ( such as , simply , very fast vesicle replenishment times ) , but the results of our work do not depend on the particularities of this implementation . Neuronal networks with STD models without the experimentally motivated non-saturating synapses cannot display firing above ∼50 Hz due to synaptic exhaustion with standard in vitro parameters , a firing rate condition in which Poisson-like variability is commonly observed in sensory areas [2] , [4] , [8] . Full details of the model are given next . We first specify the dynamics of the synaptic variable , defined as the amount of neurotransmitter released at each synaptic contact between the postsynaptic neuron i and the postsynaptic neuron j at the arrival of the m-th spike from neuron j , . Associated to this variable , there is a neurotransmitter availability variable that specifies how much neurotransmitter is ready for release at any time . The stochastic model of synaptic transmission at each synaptic contact is as follows and independent across contacts: ( 1 ) upon arrival of a spike at time a vesicle from a readily releasable pool fuses the membrane and releases its content with probability . If neurotransmitter is released , the synaptic variable equals the amount of neurotransmitter that is released by the vesicle , , and otherwise . ( 2 ) Immediately after release , a vesicle from a readily releasable pool with low neurotransmitter load becomes available . It has an amount of neurotransmitter , . ( 3 ) The time it takes this vesicle to be replaced by a vesicle with high neurotransmitter load , , is a random variable following an exponential distribution with mean . With the above choice of the maximum value , the synaptic strength at each contact is quantified by . The dynamics of probabilistic synapses has been simulated as follows: for each synaptic contact that has been partially depleted to the value , a random exponentially distributed time was generated as where is uniformly distributed in the interval [0 , 1] . Once this time has elapsed , the synaptic contact was replenished to its maximum value . When synaptic transmission is permitted immediately after a successful transmission . However , the neurotransmitter that can be immediately released is smaller than the maximum allowed value . The choice also ensures that the currents generated by the network do not saturate below 20–50 Hz . We start by describing a recurrent network with non-leaky integrate-and-fire neurons ( nLIF ) and probabilistic synapses . In an nLIF neuron the leak term ( see eq . ( 1 ) ) has been dropped and the voltage obeys ( 5 ) For simplicity in the expression we have taken the membrane capacitance . We also normalize the spiking threshold such that the reset membrane potential is defined to be at zero . The nLIF neuron is an excellent approximation for a LIF neuron when inputs are strong and firing rate is high , precisely the situation where Poisson-like firing breaks down in LIF networks . Therefore , showing that networks of nLIF neurons with probabilistic synapses give rise to Poisson-like firing will mean that the same property holds for LIF networks with probabilistic synapses . In the main text we show that the qualitative results derived for networks of nLIF neurons also apply to networks of LIF neurons . The total synaptic current delivered to the neuron is ( 6 ) identical to eq . ( 2 ) but where we use instead a simplified model of probabilistic synapse . Specifically , each synaptic variable in eq . ( 6 ) becomes one with probability p upon spike arrival , and otherwise it is zero , independently across contacts and time . Therefore , in this model the temporal dynamics of synapses is neglected , but the probabilistic nature of synaptic transmission is preserved . The theory that is presented below is valid for any arbitrary synaptic kernel with the properties described in eq . ( 3 ) . It is worth emphasizing that neglecting the temporal dynamics of synapses modifies the precise values of the steady-state neurons' firing rates and their Fano factors , but the qualitative effects about Poisson-like variability naturally extend to the more realistic case with synaptic dynamics . In the next section we compute exactly the mean activity and covariance of the spike counts across neurons in the network , required to show that a nLIF neuronal network with probabilistic synapses display exactly Poisson-like variability . First , we rewrite eqs . ( 5 ) and ( 6 ) in a more convenient way that will highlight the effect of membrane potential resetting . Since the effect of a spike emitted by neuron i is to decrease its membrane potential instantaneously from threshold to the reset values , eq . ( 5 ) can be expressed as ( 7 ) where denotes the spike times of neuron i . Eqs . ( 5 ) and ( 6 ) can be rewritten in matrix notation as ( 8 ) where and are diagonal matrices with entries and . In the expression , , ( recurrent part of the total current ) and are vectors with i-th components , ( 9 ) and , , respectively . The advantage of using eq . ( 8 ) instead of eq . ( 5 ) is that the non-linear resetting mechanism of the cells is transformed into a term indistinguishable from self-inhibition or negative-feedback . Now we move to compute the mean spike counts over the neurons in the network . In the following we assume that there is a single attractor state of the system . We start by taking expected values ( over all realizations of the white noise processes and initial conditions of the network leading to the same set of active neurons ) in the two sides of eq . ( 8 ) to obtain ( 10 ) Since the average membrane potential does not change in the stationary regime if the firing rates of the neurons are positive , the l . h . s . of the equation is zero . Noting that are random variables independent of spike times and both across contacts and synapses , and using that , where is the population firing rate vector , if the rates are non-negative we find that eq . ( 10 ) is equivalent to the constraint over the population firing rate vector ( 11 ) where the effective connectivity matrix has diagonal entries and off-diagonal entries . The matrix explicitly shows the self-inhibitory effect of the reset mechanism . If is invertible , eq . ( 11 ) can be readily solved to give an expression for the population firing rate ( 12 ) Eq . ( 11 ) , written more generally to include cases where the firing rates can be zero , becomes ( 13 ) where is the linear rectified function ( if , and otherwise ) . Although we have assumed the presence of a single attractor , this equation allows for multiple solutions in general . In those cases , multi-stability develops in the network , and each state obeys an equation like eq . ( 11 ) where the connectivity matrix becomes the original one but where the columns and rows corresponding to the inactive neurons have been removed . The firing properties described below hold for each state but in addition stochastic transitions between the states are possible . Eq . ( 11 ) expresses the required balance between excitatory ( E ) , inhibitory ( I ) and external inputs in the stationary regime . We can rewrite this equation for the simple but illustrative case where all neurons in population connects with all other neurons but itself in the population with strength , the mean input currents to the E and I populations are and , respectively , spiking threshold for all neurons is , and there is a single contact per neuron pair with the same transmission probability p for all synapses . In this case eq . ( 11 ) it is equivalent to the set of linear equations ( 14 ) Where and are the firing rates of E and I neurons . The first ( second ) equation describes the mean input currents to an E ( I ) neuron , and it shows that the mean excitatory , inhibitory and external mean currents and the effective depolarizing current proportional to cancel precisely in such a way that the sum of them is precisely zero . Therefore , in the stationary regime the network settles down in a set of firing rates that satisfies precisely this balance equation [9] . Note that if the connectivity strengths are large ( ) , the cancelation mainly occurs between a large excitatory mean drive by a large inhibitory mean drive . Fig . 2e shows how the mean currents are dynamically cancelled , leading to a state of low firing rates . As a next step , it is useful to show that the average activity over realizations of the white noise process equals its temporal average . We first integrate eq . ( 8 ) from time zero to a long time as ( 15 ) where is the spike count population vector in the time window with components , . The integrated current has components ( 16 ) where is the number of successful synaptic transmissions at contact k from neuron j to i . This number is a random number that depends on the number of actual spikes from neuron j , , as ( 17 ) where is a normally distributed variable independent across synaptic contacts . The first term in the equation corresponds to the average number of successful spike transmissions at the synaptic contact k from neuron j to i , while the second term correspond to the fluctuations of the number of successful transmissions around the mean , which becomes a Gaussian variable for long . Note that the mean and variances of the Gaussian are proportional to the mean and variances of a Bernoulli process with success probability . Strict equality of eq . ( 17 ) holds for . Now it is easy to extract useful information from eq . ( 15 ) . For long the terms proportional to and to the spike counts dominate . Therefore , for and positive spike counts eq . ( 15 ) reduces to ( 18 ) This equation is identical to eq . ( 11 ) if the expression is divided by , hence showing the equivalence between realization and temporal averages . Finally , we compute the covariance matrix of the spike counts across pairs of neurons using the previous equations . We start by splitting the recurrent term in eq . ( 15 ) into deterministic and fluctuation terms and lumping together the normally distributed random variables across contacts at each synapse . It can be shown that eq . ( 15 ) can be rewritten in a more convenient way as ( 19 ) where is a random matrix with entries ( are i . i . d . normally distributed variables ) , and is a vector whose components are the squared root of the spike counts up to time , , . Taking expectations in both sides of eq . ( 19 ) we find ( 20 ) Subtracting eq . ( 20 ) from eq . ( 19 ) , and defining the membrane potential and spike count fluctuations as and respectively , we obtain ( 21 ) This equation is the basis to obtain the covariance matrix of the spike counts . Solving for in eq . ( 21 ) , we find after some laborious algebra that the covariance of the spike count can be written up to order as ( 22 ) where is a rate-dependent diagonal matrix with entries ( 23 ) and “tr” stands for matrix transpose . We note that in eq . ( 22 ) the noise introduced by probabilistic synaptic transmission is multiplicative with the rate ( see eq . ( 23 ) ) , and that it enters as a diagonal matrix that is further amplified and transformed by the recurrent connections . These results fully and exactly describe the first and second-order firing statistical properties of nLIF recurrent networks , opening in turn the door to study correlations in spiking recurrent networks with probabilistic synapses . The probabilistic synaptic model that we have considered so far does not have variability in the amplitude of the synaptic strength . It is possible to include this source of variability in the present formalism by replacing in eq . ( 23 ) by ( 24 ) where is the variance of across successful synaptic transmissions at each contact , and by replacing in eq . ( 22 ) by a matrix with diagonal entries and off-diagonal entries . The expression for the firing rate and covariance of the spike counts , eqs . ( 11 ) and ( 22 ) , have been derived for the case where delays are fixed and there is not jittering during the generation of the spikes . However , it is possible to show that the same expressions hold when the delays and jitters are random with finite first and second order moments . This shows that noise introduced by random synaptic delays and spike generation jittering constitute a negligible source of noise in nLIF networks . From the equation of the mean firing rates , eq . ( 12 ) , and the expression for the covariance matrix of the spike counts , eq . ( 22 ) , it follows that that the Fano is constant for all firing rates . If the input drive is scaled by a factor , , then according to eq . ( 12 ) the firing rates are scaled up by the same factor , . Similarly , if the noise from external sources is small , then the covariance matrix of the spike counts is approximately scaled up by the same factor , because the noise introduced by probabilistic synapses is multiplicative . Since the Fano factor is defined as , scaling up the input drive can modulate the firing rate of individual neurons by several orders of magnitude while their Fano factor remains constant . This finally shows that Poisson-like variability arises in spiking networks with probabilistic synapses by virtue of the multiplicative nature of synaptic noise . In this section we provide details for Fig . 3 in the main text . Let us assume that the output spike train of the presynaptic neuron , described by the spike count , has a mean count and variance , where is the firing rate of the neuron and is the length of the time window . When this spike train passes through a probabilistic synapse with a probability of successful transmission , a sequence of PSCs is generated with mean count and variance . Note that the output firing rate has been diluted by a fraction , and that the variance contains two terms arising from a doubly stochastic process: the first term comes from the extra variability introduced by the probabilistic synaptic transmission , while the second term is a diluted version of the presynaptic spike train variability . It is crucial to realize that the first term is proportional to the firing rate in the network , while the second term is rate-independent . To close the loop , we need to specify the way that input mean count and variability are transformed into the mean spike count and variability of . Assuming an homogeneous network of neurons with connectivity strength , then a LIF neuron with threshold generates a spike train with mean count and variance , where is the mean external input drive to the neurons in the network . To derive the expression for the mean and variance , we have assumed that spike trains across neurons are approximately independent and that the firing rate is high . Finally , using the relationship between and both rate and , and the fact that input and output variances should be equal in a self-consistent recurrent network , we arrive to the expression . Note that this expression predicts that the spiking variability of the neurons is Poisson-like because it is proportional to the firing rate in the network . Note also that the Fano factor increases with the connectivity strength , and with the number of connections per neuron at fixed connectivity strength . Finally , we provide details about the parameters used in each figure . The parameters for Fig . 1 are as follows . A single neuron obeying eqs . ( 1 ) – ( 2 ) was simulated . The membrane capacitance , leak conductance and leak potential were , and respectively . With those choices , the membrane time constant was . The spiking threshold and reset membrane potentials were set at and respectively [11] . For the case with constant input noise ( dashed lines ) , the standard deviation of the noise was set at . For the case of Poisson-like inputs ( solid lines ) , the variance of the noise grew proportionally with the mean input drive as , where we chose . The parameters of the network with probabilistic synapses described in Fig . 2 ( solid lines ) were as follows . A total of 2000 neurons were simulated , 80% of which were excitatory and 20% were inhibitory . The connectivity was all to all . The membrane capacitance , leak conductance , leak potential , and resting potential were as in Fig . 1 , while . The connectivity strength of contact k between the presynaptic neuron j in population E and the postsynaptic neuron i in population E took the values , between E and I neurons , between I and E neurons , and between I and I neurons . There were a fixed number of 4 contacts between all pairs of neurons . With these values , a successfully transmitted E presynaptic excitatory spike generates an EPSP of 0 . 66 mV on E postsynaptic neurons [34] , [35] . The synaptic decay time constant of excitatory and inhibitory PSC ( see eq . ( 4 ) ) was and respectively . The probability of release was [22] , the recovery time constant of vesicles took the value [26] and the minimum vesicle neurotransmitter fractional load was set at [25] , [69] , identical for all contacts . The mean input current in eq . ( 2 ) was the same for all the neurons , and the input variance was taken to be zero . Simulations were run for 200 s with a one-step Euler method with time step . Fano factors were computed using time windows of 2 s . None of the results presented depend critically on the values of the parameters chosen . For the network without probabilistic synapses in Fig . 2 ( dashed lines ) , the parameters were as above except for the following . Neurons were driven by noise with constant standard deviation of the noise . We set and , and therefore probabilistic synapses and STD temporal dynamics were absent . To produce comparable rates to those in the network with probabilistic synapses we compensated the larger by reducing by half all synaptic strengths . In Fig . 3a the parameters were as in Fig . 2 for the network without probabilistic synapses . The network connectivity was random and sparse , with every neuron in the network receiving connections from a small fraction , , of pre-synaptic excitatory and inhibitory neurons randomly chosen . The strength of the connections was as before . In Fig . 3b , the parameters were as in Fig . 2 for the network without probabilistic synapses , with the exception that the reset potential was set at a higher value , . In Fig . 3c the parameters were as in Fig . 2 for the network with probabilistic synapses , but with and . In Fig . 3d , parameters are as in Fig . 3c with the addition of a random delay independently for each spike ( spike jitter ) and uniformly distributed between 0 and 10 ms . In Fig . 3e the parameters were again as in Fig . 2 for the network with probabilistic synapses , except that , and the synaptic weights were divided by half . In Fig . 4 parameters were identical as in Fig . 2 when the synaptic scaling factor g is 1 . For the cases of scaling factor different from one , all connectivity strengths of the network in Fig . 2 were multiplied by g , keeping fixed all other parameters . For the conductance-based network with probabilistic synapses in Fig . 6 , parameters were as in Fig . 1 , with and . Synaptic current were modeled as , where ( k = E , I ) is the synaptic conductance and the reversal potentials are and . The synaptic conductances follow an equation identical to eq . ( 2 ) with , , and . Probabilistic synapses without STD were studied with . For the network without probabilistic synapses ( ) all synaptic strengths were reduced three-fold to keep firing rates close to those from the network with probabilistic synapses . In addition , external constant noise was added to each neuron with standard deviation . In both networks , there were 100 neurons with a single contact between all pairs of neurons , of which 80 were excitatory and 20 were inhibitory . The synaptic decay time constant of excitatory and inhibitory PSC ( see eq . ( 4 ) ) was and respectively . External current-based inputs were excitatory and identical to all neurons .
|
Neurons in cortex fire irregularly and in an irreproducible way under repeated presentations of an identical stimulus . Where is this spiking variability coming from ? One unexplored possibility is that cortical variability originates from the amplification of a particular type of noise that is present throughout cortex: synaptic failures . In this paper we found that probabilistic synapses are sufficient to lead to cortical-like firing for several orders of magnitude in firing rate . Moreover , the resulting variability displays the property that variance of the spike counts is proportional to the mean for every cell in the network , the so-called Poisson-like firing , a well-known property of sensory cortical firing responses . We finally argue that far from being harmful , probabilistic synapses allow networks to sample neuronal states and sustain probabilistic population codes . Therefore , synaptic noise is not only a robust mechanism for the type of variability found in cortex , but it also provides cortical circuits with computational properties to perform probabilistic inference under noisy and ambiguous stimulation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"computational",
"neuroscience",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"neuroscience",
"coding",
"mechanisms"
] |
2014
|
Poisson-Like Spiking in Circuits with Probabilistic Synapses
|
Intestinal L-cells sense glucose and other nutrients , and in response release glucagon-like peptide 1 ( GLP-1 ) , peptide YY and other hormones with anti-diabetic and weight-reducing effects . The stimulus-secretion pathway in L-cells is still poorly understood , although it is known that GLP-1 secreting cells use sodium-glucose co-transporters ( SGLT ) and ATP-sensitive K+-channels ( K ( ATP ) -channels ) to sense intestinal glucose levels . Electrical activity then transduces glucose sensing to Ca2+-stimulated exocytosis . This particular glucose-sensing arrangement with glucose triggering both a depolarizing SGLT current as well as leading to closure of the hyperpolarizing K ( ATP ) current is of more general interest for our understanding of glucose-sensing cells . To dissect the interactions of these two glucose-sensing mechanisms , we build a mathematical model of electrical activity underlying GLP-1 secretion . Two sets of model parameters are presented: one set represents primary mouse colonic L-cells; the other set is based on data from the GLP-1 secreting GLUTag cell line . The model is then used to obtain insight into the differences in glucose-sensing between primary L-cells and GLUTag cells . Our results illuminate how the two glucose-sensing mechanisms interact , and suggest that the depolarizing effect of SGLT currents is modulated by K ( ATP ) -channel activity . Based on our simulations , we propose that primary L-cells encode the glucose signal as changes in action potential amplitude , whereas GLUTag cells rely mainly on frequency modulation . The model should be useful for further basic , pharmacological and theoretical investigations of the cellular signals underlying endogenous GLP-1 and peptide YY release .
Glucose sensing by a variety of specialized cells located , for example , in the pancreas [1] , the brain [2] and the ingestive tract [3] , plays a crucial role in the control of body weight and blood glucose levels , and dysfunctional glucose sensing is involved in the development of obesity and diabetes [2] . The various glucose-sensing cells rely on different molecular mechanisms for monitoring glucose levels . The prototype mechanism operating in pancreatic β-cells involves glucose-uptake by GLUT transporters and closure of ATP-sensitive potassium ( K ( ATP ) - ) channels , which leads to cell depolarization and action potential firing with subsequent insulin release [1] . However , for example the enteroendocrine L-cells use the electrogenic sodium glucose co-transporter 1 ( SGLT1 ) to link glucose stimulus to electrical activity and secretion [4–9] with a possible minor role for K ( ATP ) -channels [4 , 9] . Similarly , SGLTs are involved in glucose sensing in the hypothalamus [10] , and play a role in pancreatic α-cells [11] in addition to K ( ATP ) -channels [1] . Glucagon-like peptide 1 ( GLP-1 ) is an insulinotropic hormone released from intestinal L-cells in response to food ingestion [12] . It is , together with other hormones , responsible for the so-called incretin effect , i . e . , the fact that glucose ingested orally elicits a greater insulin response than glucose administered intravenously , even when glucose concentrations in plasma are matched . In addition , GLP-1 inhibits glucagon secretion , slows gastric emptying , regulates appetite and food intake , stimulates β-cell neogenesis and proliferation , and promotes β-cell survival both in vitro and in vivo [12] , and deficient incretin signalling has been suggested to be a major reason of insufficient insulin release and excessive glucagon release in type-2 diabetics [13] . The beneficial effects of GLP-1 have led to incretin-based therapies , and GLP-1 mimetics and inhibitors of GLP-1 degradation are already available [14] . Recently , alternative treatments , aiming at enhancing endogenous secretion from the intestinal L-cells directly , are under investigation [3 , 15 , 16] . However , the nutrient sensing mechanisms and the secretory pathways in L-cells remain still incompletely understood [17–19] . The GLP-1 secreting cell line GLUTag [20] has been widely used to obtain insight into the cellular mechanisms leading to GLP-1 release . GLUTag cells use the electrogenic SGLT1 [21] and K ( ATP ) -channels [22] to sense glucose . Electrical activity then promotes Ca2+ influx and release of GLP-1 [23] . Subsequent studies using transgenic mice with fluorescent L-cells [4] confirmed that primary L-cells rely on similar mechanisms to transduce glucose sensing to GLP-1 secretion [4 , 17] . However , differences in the electrophysiological properties of GLUTag [23] and primary L-cells [24] have emerged , which could underlie the variation in secretory responses in GLUTag versus L-cells . In particular , primary L-cells appear to rely mainly on SGLT1 for glucose sensing , in contrast to GLUTag cells , which use both SGLT1 and K ( ATP ) -channels to transduce glucose stimuli to GLP-1 secretion [4–9 , 21 , 22] . Related to the relative roles of SGLT1 and K ( ATP ) -channels is the debate on how SGLT1 and GLUT2 glucose transporters contribute to glucose sensing in L-cells [8] . As mentioned above , the electrogenic SGLT1 transporters could directly induce electrical activity , whereas glucose entering via GLUT2 should be metabolized to increase the ATP levels and reduce K ( ATP ) -channel activity to promote action potential firing . SGLT1 transporters are located on the luminal , apical side of the L-cells , and are therefore exposed directly to glucose in the intestine [8] . In contrast , GLUT2 is located on the vascular side of the L-cells [8] . It has been suggested that GLUT2 can be inserted into the luminal membrane of enterocytes in response to glucose by a SGLT1-dependent mechanism [6 , 25 , 26] , though other studies have cast doubt on this hypothesis [8] . Thus , a better understanding of the glucose-sensing mechanisms leading to electrical activity might also shed new light on the relative roles of SGLT1 and GLUT2 transport underlying GLP-1 secretion . The subtle differences in ion channel characteristics between the two GLP-1 secreting cell types complicate intuitive reasoning on the interplay of the various currents underlying GLP-1 release . In this context , a mathematical model could be useful to get a deeper insight into the stimulus-secretion pathway . Mathematical modeling has been used to study glucose sensing in pancreatic β-cells [27–29] and α-cells [30–32] , and we recently modelled human β-cells to investigate how species differences and cellular heterogeneity in electrophysiological properties are reflected in electrical activity [33–35] . Here , we present a mathematical model of electrical activity underlying GLP-1 release built directly from experimental data . A single model for primary L-cells and GLUTag cells is presented but with two sets of parameters to represent the two cell types . Thus , we investigate how the differences in ion channel characteristics translate into different electrophysiological responses in primary L-cells and GLUTag cells with particular focus on glucose sensing by SGLT1 and K ( ATP ) -channels . Further , we discuss how the simulations based on data from cultured cells can give insight into L-cells in situ with preserved physiological polarization .
Experimentally , it is possible to stimulate the two glucose-sensing mechanisms individually by using different sugar types . For example , alpha-methyl-D-glucopyranoside ( αMG ) is a non-metabolizable glucose analogue that is co-transported by SGLT1 and can depolarize the cell by a SGLT1-associated current without inducing K ( ATP ) -channel closure . Fructose , on the other hand , does not enter via SGLT1 , but is metabolized and the resulting ATP increase closes K ( ATP ) -channels [21 , 36] . We note that in contrast to β-cells , which utilize fructose poorly [37 , 38] and are unresponsive to fructose alone [38 , 39] , GLUTag cells efficiently metabolize fructose [36] , which triggers electrical activity [21] . Finally , a glucose stimulus might be sensed by the two different pathways simultaneously , and the model could help in differentiating the contribution of each pathway . Instead of fructose , the K ( ATP ) -channel blocker tolbutamide is commonly used to target K ( ATP ) -channels without affecting the SGLT1-associated current . Glucose-induced changes in K ( ATP ) -channel conductance , gK ( ATP ) , in physiological settings is the consequence of glucose transport , mainly via GLUT2 [7] , and its subsequent metabolization . To simplify the notation , in the following the term SGLT1-substrate will represent any substance that is cotransported by SGLT1 and induces the associated current . In the model described in the Methods , a SGLT1-substrate corresponds to the parameter GSGLT1 , which represents the extracellular concentration of e . g . glucose or αMG . Physiologically , this would be the major glucose stimulus from the intestine , since SGLT1 is located on the luminal side of the L-cells and is pivotal for physiological GLP-1 secretion [6 , 8 , 9] . To investigate more closely how the different membrane currents contribute to create and shape action potentials in the two cell types , we plotted the different currents during an action potential . In GLUTag cells ( Fig 5A and 5B ) , the sustained , inward SGLT1-current and a small Ca2+ current depolarize the membrane potential up to ∼40 mV . At this voltage Na+ and Ca2+ channels activate , which causes the rapid upstroke of the action potential . Inactivation of the Na+ current , and activation of the A-type and the delayed rectifier K+ current , as well as the transient , outward part of SGLT1-current , contribute to controlling the peak of the action potential . The delayed rectifier K+ current is the major current responsible for repolarization . Note that the ATP-sensitive K+ current is relatively big . In primary L-cells ( Fig 5C and 5D ) the T-type Ca2+ current plays a crucial role in depolarizing the membrane potential , which leads to , first , activation of Na+ channels , and , second , activation of HVA Ca2+ channels . Inactivation of Na+ and T-type Ca2+ channels in addition of activation of the delayed rectifier K+ current and the transient SGLT1 currents control the action potential amplitude and cause repolarization . The K ( ATP ) -current is small compared to the other currents . This insight can explain how changes in GSGLT1 and gK ( ATP ) mainly control action potential frequency in our simulations of GLUTag cells , but amplitude in primary L-cells . The effect of GSGLT1 on ISGLT1 is twofold ( Fig 2 ) : it increases the sustained , inward current , but reduced the transient outward current . Since the sustained current is important for depolarization in GLUTag cells , an increase in GSGLT1 and consequently in the sustained SGLT1 current will reduce the interspike interval , i . e . , increase the firing frequency . This effect is not present in primary cells , where the T-type Ca2+ currents is playing the main role in the depolarization . Similarly , reduced K ( ATP ) -channel conductance has a big influence on the interspike interval in GLUTag cells , since the K ( ATP ) -current is one of the dominant currents between action potentials ( Fig 5B ) . In primary L-cells , changes in the tiny K ( ATP ) -current does not affect the interspike interval , which is controlled by T-type Ca2+ channels ( Fig 5D ) . In contrast , the transient , outward SGLT1 current is more important for controlling the action potential height in primary cells because of the lack of the A-type K+ current . The role of the A-type K+ current in GLUTag cells appears to be to control the amplitude of the action potential , since it activates rapidly during the upstroke , and then inactivates ( Fig 5B ) . Reduced K ( ATP ) channel conductance has a bigger effect on peak voltage in primary L-cells ( Fig 4D ) than in the GLUTag cell line ( Fig 1D ) since IK ( ATP ) is more important during the upstroke in primary L-cells ( Fig 5D ) compared to GLUTag cells where the A-type K+ current activates ( Fig 5B ) . Thus , it is the presence or absence of complimentary currents , notably the A-type K+ current and the T-type Ca2+ current , that determines the effect of changes in GSGLT1 and gK ( ATP ) .
The relative contribution of SGLT1 and GLUT2 glucose transporters to glucose sensing in the intestinal L-cells has been a matter of debate [8] . Whereas SGLT1 transporters are electrogenic and could promote electrical activity on their own , glucose transported via GLUT2 should be metabolized to increase the ATP/ADP ratio and close K ( ATP ) -channels , which could lead to action potential firing as in pancreatic β-cells . Glucose entering via SGLT1 or GLUT2 could also reduce K ( ATP ) -channel activity . However , accumulating evidence support the main role of the SGLT1-mediated current in primary L-cells [5–9] , whereas both SGLT-1 currents and K ( ATP ) -channel closure contribute to stimulus-secretion coupling in GLUTag cells [7 , 21] . Elevations in intracellular glucose levels could also have effects downstream of electrical activity and Ca2+ influx , as has been shown in GLUTag cells [7] , and resembling the ‘amplifying pathway’ operating in pancreatic β-cells [42] . The theoretical analyses presented here provide new insight into how the electrophysiological differences between primary L-cells and the GLUTag cell line lead to their diverse glucose-sensing mechanisms . The model further suggests that the two cell types encode the glucose signal in electrical activity in different ways: primary L-cells appear to use action potential amplitude ( cf . Fig 4D and 4E ) to transduce glucose sensing to Ca2+ influx and exocytosis , while the model predicts that GLUTag cells rely mainly on changes in firing frequency ( Fig 1D and 1E ) as found experimentally [22] . We explained this difference by the presence of A-type K+ currents in GLUTag cells and T-type Ca2+ channels in primary L-cells ( cf . Fig 5 ) . Related , we note that small changes in SGLT-1 substrate lead to rapid action potential firing in primary L-cells ( Fig 4E ) but not in GLUTag cells ( Fig 1E ) . This difference might be related to the lower αMG sensitivity in GLUTag cells , which show little secretion to 5 mM αMG [21] , whereas primary L-cells have a EC50-value of ∼0 . 2 mM for αMG-triggered GLP-1 secretion [4] . A limitation of the current version of the model is that it is based on data from cultured cells that have lost their natural polarization , and possibly other characteristics . Future modeling of electrophysiology , Ca2+ dynamics and secretion based on mechanistic data from L-cells in situ will likely shed further light on the relative importance of action potential amplitude and frequency for GLP-1 secretion . In the isolated intestine , reflecting the situation in vivo , SGLT1 transporters are located on the luminal , apical side of the L-cells , and are therefore exposed directly to glucose in the intestine [8] . In contrast , the GLUT2 transporters are located on the basolateral , vascular side of the L-cells [8] , where they allow glucose to pass between the cytosol of L-cells and the plasma . There are reports of GLUT2 protein being transported to and inserted in the luminal membrane of enterocytes in response to glucose entering via SGLT1 [6 , 25 , 26] ( but see [8] ) . However , even in experiments where luminal GLUT2 expression increased , SGLT1-mediated glucose transport still predominated [6] . GLUT2 knock-out mice have been reported to have ∼50% less GLP-1 release than wild-type animals [43] , while another study found unchanged GLP-1 release in GLUT2 knock-out mice [8] . Of note , GLP-1 content is reduced by ∼50% compared to control animals [43] , which complicates reasoning on whether GLUT2 plays a role in glucose sensing in L-cells based on studies in GLUT2 knock-out animals . In contrast , luminal GLUT2 inhibition by phloretin has been shown to reduce but not abolish GLP-1 secretion in the perfused rat small intestine [9] . GLUT2 inhibition also abolished a SGLT1-independent component of GLP-1 secretion in isolated loops of small rat intestine [26] , but of note this SGLT-1 independent component was not observed in isolated rat small intestine [9] or in vivo in mice [5] , where the SGLT1 inhibitor phloridzin abolished glucose induced GLP-1 secretion . In summary , there is an ongoing debate of the role of apical GLUT2 in intestinal glucose absorption , which might be due to differences in experimental procedures [8 , 25] . Further , whether the mechanisms postulated for enterocytes are operating in L-cells still need to be shown directly . The presented model is unable to provide further insight into these question , mainly since it was build from data from cultured cells; mechanistic experimental results from L-cell in situ are needed before we can investigate these questions theoretically . Interestingly , vascular perfusion with high glucose concentrations in the presence of 3 . 5 mM luminal glucose triggered GLP-1 secretion in the isolated porcine intestine [44] , but in the isolated rat intestine vascular glucose did not lead to GLP-1 release in the absence of luminal glucose [9] . Besides species differences , these conflicting results can be explained as follows . In the presence of 3 . 5 mM luminal glucose , the SGLT1 current is operating , and vascular glucose can augment GLP-1 secretion by entering the L-cells via basolateral GLUT2 leading to metabolism and closure of K ( ATP ) channels . In contrast , in the absence of intestinal glucose and SGLT1 current , the closure of K ( ATP ) channels is insufficient to trigger electrical activity and GLP-1 secretion . In terms of our model , the presence of luminal glucose allows the L-cells to operate further to the right in Fig 4D and 4E where small downward movements due to reduced K ( ATP ) -conductance more easily lead to electrical activity . These various experiments point to a mechanism where SGLT1 is the major glucose-sensing component in primary L-cells , but glucose metabolism leading to K ( ATP ) -channel closure might play a modulating role . The theoretical results presented here support this picture . Pharmacological modulation of K ( ATP ) -channels can overwrite glucose-sensing , i . e . K ( ATP ) -channel closure by tolbutamide can trigger electrical activity and secretion in primary L-cells even in the absence of glucose [4 , 9 , 17] , and the K ( ATP ) -channel agonist diazoxide abolishes glucose-stimulated GLP-1 secretion [9 , 24] , which can be explained with the model as follows . Pharmacological modification of K ( ATP ) -channel conductance can push the system in or out or the area with activity , independently of glucose-sensing by SGLT . Such modulation of K ( ATP ) -channel activity corresponds to large vertical moves in Fig 4D and 4E such that horizontal movements ( SGLT1-mediated sensing ) are ineffective . In the basal state the L-cells have a K ( ATP ) -conductance of <10 pS [4] , which corresponds to just a single K ( ATP ) -channel being open on average [40] . Thus , tolbutamide would have very little K ( ATP ) -channel conductance to act upon . Nonetheless , our simulations showed that a further reduction in K ( ATP ) -channel conductance is sufficient to allow electrical activity . In contrast , the GLUTag cells have K ( ATP ) -channel conductance an order of magnitude larger than the primary L-cells [22] . This fact means that physiological modulation of K ( ATP ) -channel activity becomes a more reliable glucose-sensing mechanism in the GLUTag cell line , as highlighted by the findings that stimulated metabolism by fructose [21] or glucokinase activators [7] stimulate secretion in GLUTag cells but not in primary L-cells . During an oral glucose tolerance test , tolbutamide does not trigger further GLP-1 release [45] . In this condition , the luminal glucose concentration is high , meaning that the L-cells are active and operating far to the right in Fig 4D and 4E . A reduction in K ( ATP ) -channel conductance because of tolbutamide application , corresponding to a downward movement in Fig 4D and 4E , will therefore have very little effect . We are unaware of any results showing whether tolbutamide in the absence of ingested glucose stimulates GLP-1 release in vivo . However , it has been shown that fructose , which enters via non-electrogenic GLUT5 transporters and most likely act via K ( ATP ) -channel closure , stimulate GLP-1 secretion in humans in vivo [36] . While the role for Na+-channels in the generation of action potentials is clear in both GLUTag [23] and primary L-cells [24] , their importance for GLP-1 secretion is—surprisingly—less evident . The addition of the Na+-channel blocker TTX does not change glucose-stimulated GLP-1 secretion from GLUTag cells [23] , while both basal and glutamine-stimulated GLP-1 secretion from primary L-cell cultures are lowered slightly and to the same extent by TTX [24] . Another Na+-channel blocker , lidocaine , did not lower glucose-stimulated GLP-1 secretion from perfused rat intestines [9] . Our model simulations showed that , in line with experiments , glucose was able to depolarize GLUTag cells in the presence of TTX , but demonstrated also that the mean Ca2+ current was smaller in the presence of TTX than during electrical activity in the absence of TTX ( Fig 6 ) . In simulated primary L-cells , Ca2+ influx was either reduced or unaffected by TTX , depending on the conditions ( Fig 6 ) . If the modest glucose-induced elevation in Ca2+ current in the presence of TTX is sufficient to trigger maximal secretion , for example because of depletion of the pool of releasable secretory vesicles , then secretion in presence or absence of TTX would be similar , as seen experimentally in GLUTag cells [23] , rather than slightly reduced as reported for cultured primary L-cells [24] . However , this interpretation is at odds with the fact that glucose-evoked Ca2+ elevations in GLUTag cells were unaffected by TTX [23] . It might be that the small Ca2+ current evokes Ca2+-induced Ca2+ release ( CICR ) , which then is responsible for triggering exocytosis , suggesting that it is the depolarization of the base-line rather than action potential firing that causes GLP-1 secretion . Interestingly , CICR is an important component of glucose-sensing in pancreatic δ-cells [46] , and glucose amplifies GLP-1 secretion in GLUTag cells downstream of Ca2+ influx [7] . Clearly , the importance of Na+ channels and electrical activity in L-cells needs further investigation . So which of the two cell types investigated here , GLUTag or cultured mouse colonic L-cells , resemble human physiology the most ? As human , rat and mouse L-cells in vivo [36] , GLUTag cells release GLP-1 in response to fructose [21 , 36] , and GLUT2 inhibition reduce GLP-1 release from these cells [7] , similarly to the perfused rat small intestine [9] , and isolated loops of small rat intestine [26] . These properties point to a role of K ( ATP ) -channels and/or metabolism-dependent ‘amplifying pathways’ , which augment secretion at a given Ca2+ level [7] , in GLUTag cells . Thus , in these aspects GLUTag cells surprisingly resemble in vivo physiology more than primary cultured L-cells , which are unaffected by GLUT2 inhibition [7] and respond poorly to fructose ( personal communication , F . Gribble and F . Reimann , University of Cambridge , U . K . ) . However , cultured primary mouse L-cells clearly depend more strongly than GLUTag cells on SGLT1 , since the SGLT1 blocker phloridzin virtually abolish GLP-1 secretion from primary cell cultures , but only lowers release from GLUTag cells by 40–50% [7 , 21] . This strong dependence of SGLT1 in primary L-cell cultures resembles more physiological settings [5 , 6 , 9] . In summary , cultured primary L-cells are preferable for investigations on SGLT1 and electrophysiology , whereas GLUTag cells appear more similar to in vivo physiology with respect to metabolism . Hopefully improvements in isolation and culture procedures , and advanced studies on L-cells in situ will allow investigations on primary cells with maintained physiological characteristics . The model presented here should be valuable also for understanding glutamine-stimulated GLP-1 secretion , since glutamine is co-transported with Na+ by the electrogenic glutamine co-transporter [47] , and the stimulus pathway is therefore similar to glucose-sensing by SGLT1 investigated here . Further developments of the model will take into account the spatial organization of L-cells , in particular the role of SGLT1 co-transport in the apical membrane in contrast to GLUT2 transport at the basolateral membrane , where GLP-1 is also secreted . Inclusion of GLP-1 vesicle dynamics and stimulation by proteins and fat will also be interesting to study based on the present model as new data emerges . In this context , experiments on primary L-cells , preferably in situ with their polarization preserved will in our opinion be necessary to provide further insight . Mathematical modeling can and should be used in interpreting such more physiological data , which in turn will guide the evolution of the model developed here .
A single mathematical Hodgkin-Huxley-type model that , depending on the parameters , describes electrical activity in primary mouse L-cells or in the GLP-1 secreting cell line GLUTag [20] was developed . The model and the two parameter sets were based on patch clamp data from primary colonic L-cells [24] and GLUTag cells [23] , respectively . The model includes ATP-sensitive K+-channels ( K ( ATP ) -channels ) , voltage-gated Na+- , K+- and Ca2+-channels , and the electrogenic sodium glucose co-transporter SGLT1 . The evolution of the membrane potential V is driven by the contribution from the different currents ( normalized by cell capacitance ) to be described in details below , d V d t = - I N a + I C a T + I C a H V A + I K v + I K A + I K , h y p e r + I S G L T + I K ( A T P ) . ( 1 ) Voltage-gated membrane currents are modelled as I X = g X m X h X V - V X , ( 2 ) where X stands for the channel type , VX is the associated reversal potential , gX the maximal whole-cell channel conductance , and mX and hX describe activation and inactivation of the channel , respectively . Activation ( similarly inactivation ) is described by d m d t = m X , ∞ ( V ) - m X τ m X , ( 3 ) where mX , ∞ is the steady-state voltage-dependent activation function , and τmX is the time-constant of activation , which in some cases depends on the membrane potential . Steady-state voltage-dependent activation ( inactivation ) functions were described by the Boltzmann equation m X , ∞ = 1 1 + e ( V - V m X ) / k m X . ( 4 ) Reimann et al . [23] reported non-normalized currents for GLUTag cells . In order to normalize these currents , we estimated the cell capacitance C to be ∼7 pF from the results by Gribble et al . [21] , who reported that 100 mM alpha-methyl-D-glucopyranoside ( αMG ) induced a current of ∼5 pA/cell or ∼0 . 7 pA/pF in GLUTag cells . Parameters for membrane currents can be found in Table 1 , whereas parameters for the SGLT1 model are given in Table 2 . Simulations were performed in XPPAUT [48] with the cvode solver . Computer code can be found in the Supplementary Material .
|
Metabolic diseases are to a great extent because of disturbances in hormone secretion . Endocrine cells releasing hormones with a role in metabolism typically possess a refined molecular system for nutrient sensing , which allows them to respond in an appropriate manner to changes in e . g . glucose levels . The gut is the largest endocrine organ of the human body due to a range of endocrine cells that are strategically located to sense nutrient levels in response to food intake . The intestinal L-cells secrete glucagon-like peptide 1 ( GLP-1 ) , peptide YY and other hormones with anti-diabetic and weight-reducing effects , but the stimulus-secretion cascade in L-cells is still only partly understood . Here we dissect glucose sensing underlying GLP-1 secretion using mathematical modeling of electrical activity in primary L-cells and the GLP-1 secreting GLUTag cell line . We cast new light on the differences in glucose-sensing between the two cell types , and we propose that primary L-cells encode the glucose signal as changes in action potential amplitude , whereas GLUTag cells rely mainly on frequency modulation . Our results should be of general interest for understanding glucose-sensing in various cell types .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Mathematical Modeling of Interacting Glucose-Sensing Mechanisms and Electrical Activity Underlying Glucagon-Like Peptide 1 Secretion
|
During meiosis , the Msh4-Msh5 complex is thought to stabilize single-end invasion intermediates that form during early stages of recombination and subsequently bind to Holliday junctions to facilitate crossover formation . To analyze Msh4-Msh5 function , we mutagenized 57 residues in Saccharomyces cerevisiae Msh4 and Msh5 that are either conserved across all Msh4/5 family members or are specific to Msh4 and Msh5 . The Msh5 subunit appeared more sensitive to mutagenesis . We identified msh4 and msh5 threshold ( msh4/5-t ) mutants that showed wild-type spore viability and crossover interference but displayed , compared to wild-type , up to a two-fold decrease in crossing over on large and medium sized chromosomes ( XV , VII , VIII ) . Crossing over on a small chromosome , however , approached wild-type levels . The msh4/5-t mutants also displayed synaptonemal complex assembly defects . A triple mutant containing a msh4/5-t allele and mutations that decreased meiotic double-strand break levels ( spo11-HA ) and crossover interference ( pch2Δ ) showed synergistic defects in spore viability . Together these results indicate that the baker's yeast meiotic cell does not require the ∼90 crossovers maintained by crossover homeostasis to form viable spores . They also show that Pch2-mediated crossover interference is important to maintain meiotic viability when crossovers become limiting .
Meiosis produces haploid gametes from diploid progenitor cells . This reduction in ploidy results from the segregation of homologous chromosomes at the first meiotic division ( Meiosis I ) . In most organisms , the accurate segregation of chromosomes during Meiosis I requires crossing over between homologs . These crossovers provide physical linkages between homologs that enable their proper positioning at metaphase I through spindle microtubule generated forces [1] . Disruption of these forces by the loss of chromosome arm cohesion facilitates the Meiosis I division [2] . Failure to achieve at least one crossover per homolog pair results in non-disjunction of the homolog pair , leading to the production of aneuploid gametes ( reviewed in [3] ) . Meiotic crossing over is initiated in meiotic prophase by the formation of Spo11-dependent DNA double strand breaks ( DSBs; [4] ) . Meiotic DSBs can be repaired as either crossovers or non-crossovers through distinct repair pathways [5] , [6] . In Saccharomyces cerevisiae , approximately 60% of the 140–170 DSBs that form in meiosis ( estimated from a whole genome microarray analysis of dmc1Δ and dmc1Δ rad51Δ mutants ) are processed as crossovers [7] , [8] . A single S . cerevisiae cell in meiosis forms approximately 90 crossovers distributed over sixteen homolog pairs [9]–[11] . In contrast , in C . elegans meiosis , only a single crossover forms between each homolog pair that ensures Meiosis I disjunction [12] . The majority of meiotic crossovers in baker's yeast display interference . Interference ensures that a crossover designation for one DSB site makes a non-crossover fate more likely at adjacent sites , and leads to the formation of widely and evenly spaced crossovers [13]–[15] . In the interference-dependent crossover pathway , DSBs are processed to form single end invasion intermediates ( SEIs ) that result from the invasion of a DSB end into an intact homolog . These intermediates are then thought to undergo second-end capture with the intact homolog to form double Holliday junctions ( dHJs ) that are ultimately resolved to form crossovers [16]–[19] . A crossover homeostasis mechanism was identified in baker's yeast that ensures crossovers are preferentially formed at the expense of non-crossovers when the number of initiating DSBs is reduced [20] . Thus crossover interference and homeostasis ensure formation of at least one crossover on all homolog pairs [20] , [21] . The presence of at least one crossover per homolog pair is known as the obligate crossover . Barchi et al . [22] further define the obligate crossover “as one of the outcomes of the process ( es ) through which most crossovers form , not as a special type of crossover . ” Control mechanisms that ensure the obligate crossover are likely to act during the crossover/non-crossover decision , an event that takes place at or just prior to SEI formation [5] , [6] . It is important to note that previous work in baker's yeast suggested that ∼20% of crossovers on a large chromosome and ∼50% of crossovers on a small chromosome involved interference-independent crossovers that occurred through a distinct Mms4-Mus81 pathway [23] , [24] . The ZMM proteins ( Zip1-4 , Spo16 , Mer3 , Msh4-Msh5 ) act as pro-crossover factors in the interference-dependent crossover pathway by coordinating crossing over with formation of the synaptonemal complex , a zipper-like structure that connects homologous chromosomes in late stages of meiotic prophase [24]–[34] . Msh4-Msh5 attracted our attention because strains defective in this complex show strong defects in Zip1 polymerization during synaptonemal complex formation [27] , [29] . Msh4 and Msh5 each contain domains II–V found in the bacterial MutS family of mismatch repair proteins , but lack the N- terminal domain I that is required to interact with domain IV for mismatch DNA binding ( Figure 1A; [25] , [26] , [35] , [36] ) . S . cerevisiae msh4Δ and msh5Δ mutants display reduced crossing over ( ∼2 . 5 fold decreased ) and spore viability ( 30–40% ) . Tetrads obtained from these mutants display an excess of zero and two viable spores compared to wild-type . This phenotype is consistent with a Meiosis I disjunction defect [24]–[27] . The equivalent mutations in male and female mice result in sterility as a consequence of chromosome pairing and synapsis defects [37]–[39] . The residual crossovers seen in yeast msh4/5Δ mutants lack genetic interference [24] , [27]; however in msh4Δ mutants , Zip2 foci , which mark crossover designation sites , still display a pattern indicating that they are subject to interference [40] . These and other data suggest that Msh4-Msh5 acts after the crossover/noncrossover decision [16] , [40] . Consistent with the above data , biochemical and molecular studies showed that Msh4-Msh5 is required to stabilize SEIs and is capable of specifically binding to Holliday junctions as multiple sliding clamps [16] , [41] . Additional cell biological observations , primarily in the mouse , have led to a model in which Msh4-Msh5 interacts with the MutL mismatch repair homologs Mlh1-Mlh3 to resolve Holliday junctions [25] , [41]–[47] . In mouse spermatocytes in zygotene , Msh4/5 foci are present at high levels ( ∼140 per nucleus ) but decrease until mid pachytene , where they are present at roughly twice the number of crossover sites . At this stage , roughly half of Msh4/5 foci interact with Mlh1/3 foci , which localize to sites of crossing over [48]–[50] . The presence of a large number of Msh4/5 foci in zygotene suggest the possibility of early roles for Msh4/5 in meiosis; consistent with this idea is work in Sordaria which show an early role for Msh4-Msh5 during interhomolog interactions , at a time prior to when it is required for recombination progression [51] . The above information encouraged us to systematically mutagenize Msh4-Msh5 to study its role in implementing the crossover decision . We identified a class of msh4/5 threshold ( msh4/5-t ) mutants that displayed high spore viability despite 1 . 5 to 2 fold reductions in crossing over that occurred primarily on large ( XV , VII ) and medium ( VIII ) sized chromosomes . msh4/5-t mutants displayed Msh5 foci similar to wild-type; however , they showed defects in Zip1 polymerization during synaptonemal complex formation . This phenotype is consistent with defects in a crossover maturation process that occurs after Msh4-Msh5 loading onto chromosomes . A triple mutant containing a msh4/5-t allele and mutations that decreased DSB levels ( spo11-HA ) and crossover interference ( pch2Δ ) showed preferential loss of crossovers on the small chromosome III and a synthetic spore viability defect , suggesting that crossover interference is critical to maintain meiotic viability when crossovers become limiting .
Msh4 and Msh5 amino acid sequences from S . cerevisiae , H . sapiens , M . musculus , A . thaliana , and C . elegans were aligned using clustalW and CLC free Workbench software ( Figure 1 , Figure S1; data not shown ) . We selected four different classes of conserved residues to alter by site-specific mutagenesis ( Figure 1B ) . Class 1 ( Msh4/5-specific ) residues were conserved in Msh4 and Msh5 but were not conserved in other Msh family members such as Msh2 , Msh3 , and Msh6 . Class 2 ( Msh4-specific ) and Class 3 ( Msh5-specific ) were conserved only in Msh4 and Msh5 , respectively ( Figure 1B; Table 1 ) . Previous work by Pochart et al . [52] showed that mutations in the ATP binding domain of Msh5 conferred a null phenotype . Based on these observations , we also mutagenized ATP and DNA binding residues conserved among all Msh family members ( Class 4 ) . Eight of these Class 4 mutations were in homologous positions in Msh4 and Msh5 ( Figure S1 ) . In total 57 residues were mutated , 29 from Msh4 and 28 from Msh5 ( Table 1 ) . All residues were mutated to alanine , with the exception of one residue in the Msh4/5 ATP binding domain that was mutated to tryptophan to allow comparison with an amino acid substitution in a homologous position in Msh2 that affected Msh2-Msh6 ATP hydrolysis [53] . All alleles were integrated into the congenic SK1 strain EAY1108 ( EAY background , [24] ) . msh4 and msh5 alleles were analyzed as heterozygotes over their respective deletion mutations in the SK1 congenic strain EAY1112 [24] . The mutant diploid strains were sporulated and assessed for spore viability and genetic map distances on chromosome XV ( Table 1; Figure 1C ) . The mutations are presented relative to Thermus aquaticus MutS domains II , III ( linker ) , IV ( DNA binding ) and V ( ATPase ) [35] . The spore viability profiles of msh4 and msh5 mutants indicated that the Msh5 subunit was more sensitive to mutagenesis ( Figure 2A ) . A larger proportion of msh5 mutants showed ≤50% spore viability compared to msh4 ( 9 of 28 for msh5 versus 2 of 29 of msh4; p = 0 . 02 , Fisher's exact test ) . This difference was also seen in an analysis of mutations in domain IV ( DNA binding ) ; 5 of 12 msh5 mutations conferred ≤50% spore viability compared to 0 of 11 msh4 mutations ( p = 0 . 03 , Fisher's exact test ) . Five of the eight mutations in homologous positions in Msh4 and Msh5 conferred subunit-specific phenotypes . Both msh4-G639A and msh5-G648A strains contain mutations ( Walker motif A ) predicted to disrupt ATP binding; both of these strains displayed null phenotypes [35]–[36] , [52]–[55] . In contrast , a predicted ATP hydrolysis mutation in Msh4 , msh4-R676W , conferred wild-type spore viability but the corresponding mutation in Msh5 , msh5-R685W , conferred a null phenotype ( Figure 2B; Table 1 ) . Similar asymmetries between Msh4 and Msh5 were observed at four residues in the DNA binding domain IV ( Figure 2B; Table 1 ) . msh4-N532A , msh4-Y485A , msh4-L493A , and msh4-L553A had spore viabilities of 89 , 95 , 75 , and 95% , respectively; corresponding mutants msh5-D527A , msh5-Y480A , msh5V-488A , and msh5-L548A had significantly lower spore viabilities ( 30 , 67 , 40 , and 50% , respectively ) . Most msh4 and msh5 mutants with significant spore viability and/or crossover defects could not form stable Msh4-Msh5 complexes as assessed in the two-hybrid assay ( Table 1 ) . The only exceptions were msh4-E276A ( domain II ) , msh4-R676W ( ATP hydrolysis ) , msh5-D539A ( domain IV ) , msh5-G648A ( ATP binding ) , and msh5-R685W ( ATP hydrolysis ) mutants that displayed poor spore viability or crossover defects but formed stable complexes with a wild-type partner . Inability to form a stable complex in the two-hybrid assay can be explained by the disruption of an interaction domain or a loss in protein stability . Because most mutations were created in highly conserved residues that lie outside of putative interaction domains in Msh proteins [35] , [36] , [54] , a defect in the two-hybrid assay is likely to reflect a disruption of protein structure . Spore viability was plotted as a function of genetic map distance for all msh4 and msh5 mutants ( Figure 3 ) . This plot shows that crossing over could be reduced by up to two-fold on the large chromosome XV without affecting spore viability . msh4/5 mutations ( msh4/5-t ) near the threshold limit for crossovers included msh4-E276A , msh4-F491A , msh4-N532A , msh4-R676W , msh5-D76A , msh5-D250A , msh5-S416A , msh5-Y486A , and msh5-D539A ( Table 1 ) . The phenotypes conferred by these mutations were independent of their ability to disrupt the Msh4-Msh5 complex as measured in the two-hybrid assay ( Table 1 ) . A second class of msh4/5 mutants showed greater than two-fold decreases in crossing over on chromosome XV . This below-threshold class ( msh4/5-bt; msh4-Y143A , msh4-F194A , msh4-R456A , msh4-L493A , msh5-R436A , msh5-Y480A , msh5-D532A , msh5-L548A , msh5-D680A ) showed spore viabilities between 50 and 76% . These mutants were all defective in their ability to form stable Msh4-Msh5 complexes in the two-hybrid assay ( Table 1 ) . The wild-type spore viability profile for the msh4/5-t mutants suggested they were able to properly segregate all sixteen homolog pairs in Meiosis I ( Table 1; Figure 3 , Figure 4 ) . We further examined the phenotype of a subset of msh4/5-t mutants ( msh4-E276A , msh4-R676W , msh5-S416A , msh5-D539A; all but msh5-S416A showed wild-type two-hybrid interactions ) in the SK1 isogenic NHY strain background . msh4 and msh5 alleles were analyzed as heterozygotes over their respective deletion mutations . The NHY diploid strains allowed us to measure genetic map distances in large ( VII ) , medium ( VIII ) , and small ( III ) chromosomes ( Figure 5A; [23] ) . Smaller chromosomes have higher map distances per physical distance and weaker interference relative to larger chromosomes ( [40] , [56] , [57] but see [58] ) . Thus we used this strain set to determine if msh4/5-t mutations altered crossover patterns on representative small , medium , and large chromosomes . All four msh4/5-t mutants displayed wild-type spore viability but decreased crossing over ( ∼1 . 5-fold for the sum of map distances in three chromosomes; Figure 4 , Figure 5B; Table 2 ) . The spore viabilities of wild-type and one msh4/5-t mutant , msh4-R676W , were unaffected by raising the sporulation temperature to 33°C , a condition shown previously in the SK1 background to cause coordinated defects in the formation of recombination intermediates and crossover products in msh5Δ ( data not shown; [16] ) . This observation provides another indication that msh4/5-t alleles confer sufficient Msh4-Msh5 function in meiosis . The sum of genetic map distances calculated from tetrads ( similar values were obtained from total spores ) in wild-type was 147 cM; map distances for msh4-E276A , msh4-R676W , msh5-S416A and msh5-D539A were 101 , 109 , 99 , and 100 cM , respectively . As shown in Figure 6 , msh4/5-t mutants displayed a chromosome size-dependent loss of crossovers . For three intervals on the smallest chromosome III , the four msh4/5-t mutants showed 73 to 92% of wild-type crossover levels ( determined from tetrad data ) . In contrast these mutants showed 63 to 76% of wild-type levels for the two intervals on a medium sized chromosome VIII , and 61 to 66% of wild-type levels for the three intervals on a large chromosome ( Chromosome VII ) . The loss of crossovers on the large chromosome VII approached that seen in msh4/5Δ strains . For the msh4Δ and msh5Δ mutants , the sum of genetic map distances calculated from tetrads was 68 and 56 cM , respectively ( 2 . 2 to 2 . 6-fold drop in crossovers over three chromosomes , Figure 5; Table 2 ) . The values from total spores were 87 and 75 cM for msh4Δ and msh5Δ , respectively . The differences in map distance calculated by spore and tetrad data were likely due to the high rate of gene conversion seen in msh4Δ and msh5Δ mutants ( see below ) . Based on tetrad data msh4Δ crossovers levels were 36 , 42 and 54% of wild-type on chromosomes III , VIII , and VII , respectively . For msh5Δ crossover levels were 26 , 34 and 47% of wild-type on chromosomes III , VIII , and VII , respectively ( Figure 6 ) . Previously Stahl et al . [15] and Abdullah et al . [59] reported a greater loss of crossovers on larger chromosomes ( VII ) compared to smaller ones ( III ) in msh4Δ/msh5Δ mutants . These groups analyzed crossing over in wild-type , msh4Δ and msh5Δ strains in two intervals ( HIS4-LEU2 and LEU2-MAT ) on chromosome III ( small ) and two ( TRP5-CYH2 and CYH2-MET13 ) on chromosome VII ( large ) in the congenic RHB strain background . They found that the crossover defect in msh4Δ and msh5Δ mutants was stronger on chromosome VII ( 23% and 27% of wild-type , respectively ) compared to chromosome III ( 39% and 34% of wild type , respectively ) . We performed our analysis in the NHY SK1 isogenic strain . We do not have a good explanation for why our data differ from the Stahl et al . [15] and Abdullah et al . [59] studies . One possibility is that genetic mapping information from a limited number of intervals may yield a pattern due to localized recombination effects that is not seen when a larger number of intervals is examined . We then looked at crossover distribution in a msh4/5-bt mutant ( msh5-D532A ) . This msh4/5-bt mutation conferred similar spore viability levels in the NHY and EAY strain background ( 65% in EAY vs 69% in NHY; Figure 4 ) . Interestingly , the sum of genetic map distances for chromosomes III , VII , and VIII in msh5-D532A ( 69 cM ) was similar to msh5Δ ( 56 cM ) and msh4Δ ( 68 cM ) ( Figure 5 ) ; however , msh5-D532A showed a preferential retention of crossovers on the small chromosome III . Crossovers in this mutant were 56 , 39 , and 48 percent of wild-type for chromosomes III , VIII and VII , respectively ( determined from tetrads; Table 2; Figure 6 ) . Gene conversion events were analyzed at eleven marker sites in a subset of msh4/5 mutants , ( msh4-E276A , msh4-R676W , msh5-S416A , msh5-D532A , msh5-D539A ) . The frequency of gene conversion in these strains was similar to wild-type ( Table 3 ) . As seen previously , msh4/5Δ mutants displayed an elevated frequency of gene conversions compared to wild-type [21] , [25] , [60] . Lastly , crossover interference was analyzed in a representative msh4/5-t mutant ( msh4-R676W ) by measuring the coefficient of coincidence ( COC , ratio of observed double crossovers to those expected by chance; Table 4; [61] ) and the NPD ratio ( Table 5; [62]–[63] ) . Lack of interference yields COC and NPD values of 1 while strong interference yields values significantly less than 1 . On the whole crossover interference appeared similar in wild-type and msh4-R676W . In COC analysis the msh4-R676W mutant showed a lack of interference for two intervals on chromosome III; wild-type showed a lack of interference for only one of these intervals ( Table 4 ) . For chromosomes VII and VIII , msh4-R676W and wild-type both showed crossover interference at two intervals and the absence of interference at another . NPD ratios , calculated for intervals where at least eight NPD events were expected , were determined using Stahl's “better way” calculator . This method performs a chi square test to determine if there is a significant difference between the observed PD , TT and NPD tetrad classes and those expected by random crossing over . This analysis showed the presence of interference in both wild-type and msh4-R676W in three intervals on chromosomes VII and VIII ( Table 5 ) . pch2Δ mutants display elevated crossing over on medium and large chromosomes , and are defective in crossover interference , yet display wild-type spore viability [21] , [64]–[66] . In addition , initial genetic analyses showed that pch2Δ mutants displayed an increased ratio of crossovers to non-crossovers [21] . These observations , combined with cytological analyses indicating that Pch2 promotes domainal axis organization in meiosis [66] , [67] , led Zanders and Alani [21] to propose that Pch2 acts in early steps in crossover control to promote crossover interference at the crossover versus non-crossover decision . To test if msh4/5-t mutants showed increased sensitivity to early defects in crossover control , we made double and triple mutant combinations involving the msh4/5-t , spo11-HA , and pch2Δ mutations in the NHY strain background . The spo11-HA mutation was examined because strains bearing this allele display a 20% reduction in meiosis specific DSBs but show wild-type levels of crossing over and spore viability due to crossover homeostasis [20] . pch2Δ spo11-HA strains , however , display a significant loss in spore viability ( 73% ) . One explanation for this phenotype is that when DSBs become limiting , the proper distribution of crossovers becomes even more critical to ensure that every chromosome receives at least one crossover [21] , [66] . As shown in Figure 4 , Figure 5B , and Table 2 , msh4-R676W spo11-HA and msh4-E276A spo11-HA double mutants displayed wild-type spore viability ( 89 and 91% , respectively ) and cumulative map distances ( 113 and 106 cM , respectively , from tetrads ) . These values were similar to those seen in msh4-R676W ( 109 cM ) and msh4-E276A ( 101 cM ) single mutants . However , compared to msh4-R676W and msh4-E276A single mutants , msh4-R676W spo11-HA and msh4-E276A spo11-HA double mutants showed a decrease ( ∼30% ) in crossing over in the small chromosome III that was accompanied by modest increases in crossing over in the medium and large chromosomes ( Figure 6; Table 2 ) . We do not have a good explanation for this phenotype; one possibility is that the spo11 hypomorphs confer mutant phenotypes in addition to lowering DSBs ( see Discussion; [21] ) . msh4-R676W pch2Δ and msh4-E276A pch2Δ double mutants also showed wild-type spore viability ( 93% for both , Figure 4 ) ; however the pch2Δ mutation conferred an increase in crossing over in msh4-R676W and msh4-E276A strains that appeared specific to the medium- ( VIII ) and large-sized ( VII ) chromosomes ( Figure 5B , Figure 6 ) . The cumulative map distances from tetrads in msh4-R676W pch2Δ ( 194 cM ) and msh4-E276A pch2Δ ( 190 cM ) , were higher than wild-type ( 147 cM ) but lower than pch2Δ ( 226 cM; Figure 5B ) . pch2Δ msh5Δ mutants were previously shown to have higher crossover frequencies than the msh5Δ mutant [21] . The wild-type spore viability profile seen in msh4/5-t spo11-HA suggested that crossover interference and homeostasis can distribute a smaller pool of crossovers to all 16 homolog pairs . In contrast , the wild-type spore viability profile seen in msh4/5-t pch2Δ can be explained by an increased number of crossovers compensating for interference defects [21] . Such explanations predict that compromising crossover interference ( pch2Δ ) and limiting DSB's ( spo11-HA ) would decrease spore viability because a random distribution of crossovers will favor large chromosomes ( Figure 6; [21] ) . These effects are likely to be more pronounced in a msh4/5-t pch2Δ spo11-HA mutant that is predicted to be compromised for DSB formation , crossover interference , and crossing over . To test this we created the msh4-R676W pch2Δ spo11-HA triple mutant and analyzed its phenotype with respect to spore viability , crossover distribution , and chromosome III non-disjunction . As shown in Figure 4 , the msh4-R676W pch2Δ spo11-HA triple mutant displayed 55% spore viability , which was lower than spo11-HA pch2Δ ( 72% spore viability ) . The cumulative crossover level from tetrads for chromosomes III , VII and VIII in this mutant was 135 cM , which was lower than wild-type ( 147 cM ) and pch2Δ spo11-HA ( 165 cM ) , but significantly higher than msh4-R676W ( 109 cM ) , which displayed high spore viability ( Table 2; Figure 4 , Figure 5B ) . msh4-R676W pch2Δ spo11-HA also showed a greater reduction in crossing over on chromosome III compared to pch2Δ spo11-HA mutants ( Figure 6 ) . Although crossover levels on chromosome III in msh4-R676W pch2Δ spo11-HA were similar to msh4-R676W spo11-HA , the medium ( VIII ) and large chromosomes ( VII ) in msh4-R676W pch2Δ spo11-HA showed specific increases in crossing over compared to msh4-R676W spo11-HA as predicted by the model ( Figure 6 , Figure S2 ) . Consistent with this , the triple mutant displayed a spore viability profile indicating a Meiosis I disjunction defect ( Figure 4 ) . The triple mutant showed a higher frequency of non-mater two-spore viable tetrads in the triple mutant ( 12 . 7% , n = 71 two spore viable tetrads; 1 . 9% of total tetrads ) compared to both pch2Δ spo11-HA ( 6 . 9% , n = 130; 0 . 96% of total tetrads ) and msh4-R676W ( 6 . 8% , n = 44; 0 . 37% of total tetrads ) . Such tetrads are indicative of nondisjunction of chromosome III because the two viable spores carry both yeast mating types ( MATa and MATalpha ) . In addition , 82% of the two spore viable tetrads in the triple mutant were sister spores compared to 68% in pch2Δ spo11-HA and 50% in msh4-R676W . These data are suggestive of non-disjunction of other chromosomes . Together this information is consistent with the triple mutant being unable to distribute at least one crossover between all homolog pairs ( see Discussion ) . msh4Δ and msh5Δ mutants show strong defects in Zip1 polymerization during synaptonemal complex formation [27] , [29] . Our data below indicate that fully functional Msh4-Msh5 is required for complete Zip1 polymerization along homologs . Immunostaining of Msh5 and Zip1 was performed on a subset of the msh4/5-t ( msh4-E276A , msh4-R676W , msh5-S416A , msh5-D539A ) and msh4/5-bt ( msh5-D532A ) mutants in the NHY strain background four hours after induction into meiosis ( Figure 7 ) . The number and distribution of Msh5 foci on meiotic chromosomes for wild-type , msh4/5-t , and msh5-D532A mutants were similar . The average number of Msh5 foci per nucleus ( n = 30 ) was 122 for wild-type , 120 for msh5-D532A , and 130 for msh5-D539A . However , all mutants showed a partial defect in Zip1 elongation and accumulated Zip1-specific polycomplexes . This phenotype is reminiscent of that displayed by spo16 and zip4 null mutants with the exception that spo16 and zip4 null mutants display poor spore viability [29] . One explanation for these observations is that the msh4/5 mutants present fewer crossover sites to initiate Zip1 polymerization; thus these mutants , while capable of loading Msh4-Msh5 onto meiotic chromosomes , appeared defective in steps required to implement crossing over at designated sites . Thus complete Zip1 polymerization may require feedback from Msh4-Msh5 that is delayed or does not occur in the msh4/5 mutants . We also measured by DAPI staining the percent of cells that completed at least Meiosis I ( MI/MII ) for all of the strains examined by immunofluorescence . As shown in Figure S3 , wild-type and one msh4/5-t threshold mutant , msh4-E276A , displayed similar timing and efficiencies of meiotic divisions . The msh4Δ , msh5Δ , three msh4/5-t mutants ( msh4-R676W , msh5-S416A , msh5-D539A ) , and one msh4/5-bt mutant ( msh5-D532A ) all showed about a 1 . 5 to 2 hr delay relative to wild-type .
S . cerevisiae maintains a high level of crossing over , an average of 5 . 6 per homolog pair [8]–[11] , [20] . In most organisms that display crossover interference ( C . elegans , A . thaliana , Zea mays , D . melanogaster , Mus musculus and Homo sapiens ) , the ratio of crossovers in meiosis to homolog pairs is less than or equal to three ( reviewed in [68] ) . Why does S . cerevisiae enjoy such a high level of crossing over when a single crossover per homolog pair appears sufficient to promote Meiosis I disjunction [12] , [15] ? One possibility is that high crossover levels improve fitness by reducing mutational load through the segregation of deleterious alleles [69] . Consistent with this idea are simulation studies suggesting that meiotic crossover rates in S . cerevisiae are optimized for mutational robustness [69] . Another possibility is that excess crossovers are needed to ensure crossover formation on small chromosomes [8] , [11] , [21] . Consistent with the latter explanation is work in yeast showing that a small chromosome ( I , 230 KB ) has a higher than average recombination rate . Chromosome I also showed a frequency of non-disjunction ( 0 . 2–0 . 4% ) that was lower than expected ( 5% ) if it had recombined at the average rate [56] , [57] , [70] . The enhanced recombination rates on smaller chromosomes in S . cerevisiae are likely to result from DSBs that occur at a higher than average density and weak crossover interference [40] , [57] , [58] , [71] , [72] . msh4/5 mutants displayed high spore viability and a higher retention of crossovers on a small chromosome ( III ) compared to larger chromosomes ( VIII , VII and XV ) . We entertain two models to explain this phenotype . Both of these are based on work showing that Msh4-Msh5 is required to stabilize SEI recombination intermediates and can bind to Holliday junctions [16] , [41] . In one model , msh4/5-t mutants are defective in converting all SEI and Holliday junction intermediates into crossovers with equal probability . Such a model predicts that crossover interference would not be affected in msh4/5-t mutants , and that msh4/5-t mutants would show defects in synaptonemal complex formation . Both of these phenotypes were seen in this study . This model predicts that msh4/5-t mutants would show high spore viability despite a decrease in crossing over because smaller chromosomes have a higher frequency of crossovers and the number of crossovers in yeast is much greater than the number of chromosomes . A drawback of this model is that it cannot fully explain why msh4/5 null mutants displayed more severe crossover defects on the smaller chromosome III . Such a pattern is unexpected if crossovers on small chromosomes are present at higher density and occur primarily through a non-interfering pathway [23] . It also cannot explain how msh4/5-t pch2Δ mutants make excess crossovers . We cannot rule out the possibility that the small number of intervals examined on chromosome III is not representative of the overall pattern . In the future we would like to test this model further by examining additional intervals on this chromosome as well as on another small chromosome such as chromosome I . In addition , we would like to examine the effect of the msh4/5-t mutations on early recombination intermediates such as SEIs . We considered a second model that proposes a prioritization mechanism for the distribution of crossovers amongst chromosomes . This model is somewhat similar to that proposed by Kaback and colleagues [56] , [57] . We suggest that msh4/5-t phenotypes reflect a temporal order of crossover designation that favors a crossover on every homolog pair before additional interference-dependent crossovers are made . Such a pattern can be presented within the context of a stress relief model for crossover initiation and distribution . In this model “crossover designation with accompanying interference can be explained by imposition , relief , and redistribution of compression stress and stress relief along chromosome axes” [13] . Crossover initiation on every homolog pair would lead to the release of mechanical stress along the homolog axis of every chromosome . For shorter chromosomes , interference created from stress relief at the crossover initiation site would extend to the end of the chromosome , leading to fewer interfering crossovers as was seen experimentally [57] . For large chromosomes , interference created by stress relief that accompanies obligate crossover designation would prevent additional crossovers until mechanical stresses are re-distributed . We suggest that this redistribution of stress delays additional crossover designations on larger chromosomes . In this model the msh4/5-t phenotype can be explained if mutant Msh4-Msh5 complexes can participate in initial stress relief to form an obligate crossover but are defective , perhaps due to stability issues , in subsequent crossover initiations that are subject to interference . This model could explain the synapsis defects seen in msh4/5-t mutants if the defect is specific to long chromosomes; a single synapsis initiation site on a small chromosome could be sufficient to allow polymerization along the entire chromosome . This model , however , does not account for why Msh5 focus formation appears wild-type in msh4/5 mutants . One possibility is that subsequent crossover initiations require functions that occur after Msh4-Msh5 loading onto chromosomes . The temporal order model outlined above predicts that spore viability would be maintained in msh4/5-t mutants due to formation of the obligate crossover and that interference would appear stronger on larger chromosomes . Such an idea is consistent with previous studies in yeast showing that multiple interfering crossovers occur more frequently on large chromosomes and with models that explain the distributions of interfering crossovers seen on different sized chromosomes ( e . g . [13] , [15] , [40] , [57] , [73] ) . While we have shown that msh4/5-t mutants maintain high spore viability and display crossover interference on large chromosomes ( Figure 4; Table 4 , Table 5 ) , our data are not robust enough to test whether interference becomes stronger on these chromosomes . A caveat in this model is that msh4/5-t mutants display crossover levels on large chromosomes that are higher than wild-type in the pch2Δ mutant background . Thus msh4/5-t mutants do not appear limited in their ability to form crossovers . One way to explain this observation is that Pch2 acts as a general factor to repress recombination that increases the temporal window over which a mutant Msh4-Msh5 complex must execute crossover decisions . Alleviation of this repression results in increased crossing over in msh4/5-t pch2Δ mutants . Crossovers in msh4-R676W pch2Δ spo11-HA triple mutants appear to be randomly distributed , thus leading to more crossing over on larger chromosomes compared to the msh4-R676W single mutant , and increased non-disjunction on a small chromosome . Previous studies have suggested that Pch2 is essential for proper meiotic axis organization following crossover designation and that crossover distribution is mediated by changes in meiotic axis organization/assembly ( e . g . [13] , [67] , [74] ) . We suggest that the triple mutant phenotype can be explained in the second model if the pch2Δ mutation disrupts stress/stress relief mechanisms so that crossover designations occur without interference and no crossovers show a temporal delay . In this scenario Pch2 maintains meiotic viability when crossovers are limiting ( i . e . msh4/5-t , spo11 hypomorph mutations ) because it imposes a delay on additional interfering crossovers . This delay ensures that every homolog pair has received at least one crossover . One way to test this idea in yeast is to perform a genome wide analysis of crossing over in the msh4/5-t mutant versus the triple mutant [8] , [11] . The Msh family of mismatch repair proteins display asymmetric roles with respect to DNA binding and ATP hydrolysis . In MutS , residues in domain I of subunit A specifically stack with the mismatch while domain IV of subunit B makes non-specific contacts with the DNA backbone [35] , [36] . Similarly in MutSα , domain I in Msh6 specifically interacts with the mismatch while domain IV in Msh2 makes non-specific contacts with DNA [54] , [75] , [76] . Msh subunits also display different affinities for ATP and ADP [77]–[79] . For example in the Msh2-Msh6 mismatch repair complex , Msh6 and Msh2 contain high affinity binding sites for ATP and ADP , respectively [80] . Such asymmetries in ATP binding by Msh subunits are thought to be important to induce coordinated conformational changes in Msh-mismatch DNA complexes that signal downstream repair factors [80]–[84] . Three observations support the presence of asymmetries in Msh4-Msh5 analogous to those seen for the Msh mismatch recognition factors . 1 . Snowden et al . [85] reported that the Msh4 subunit of human Msh4-Msh5 appears to have reduced ATP binding activity . 2 . We identified different spore viability phenotypes for matched sets of msh4 and msh5 mutations that map to the ATP and DNA binding domains ( Figure 2B ) . 3 . We also found that on the whole , msh5 mutations conferred more severe meiotic phenotypes than the equivalent msh4 mutations , though this could indicate different structural organizations for the two proteins rather than asymmetric functions . Msh4-Msh5 binds to both single end invasion and symmetric double Holliday junction substrates [41] , [85] . Based on studies performed with Msh and Mlh mismatch repair factors , it is easy to imagine that asymmetric Msh4-Msh5 interactions with its DNA substrate will involve analogous signaling steps that activate downstream factors such as Mlh1-Mlh3 . Biochemical analysis of some of the mutant complexes presented in this study can provide evidence to support or refute these ideas .
S . cerevisiae SK1 yeast strains were grown on either yeast extract-peptone-dextrose ( YPD ) or synthetic complete media at 30°C [86] . When required , geneticin ( Invitrogen , San Diego ) and nourseothricin ( Werner BioAgents , Germany ) were added to media at prescribed concentrations [87] , [88] . Sporulation medium was prepared as described in Argueso et al . [24] . msh4 , msh5 mutants were analyzed in either the congenic EAY1108/EAY112 background ( “EAY” ) described in Argueso et al . [24] or the isogenic NHY942/NHY943 background ( “NHY” ) described in de los Santos et al . [23] . 28 msh5 and 29 msh4 point mutants were introduced in the EAY1108 background by transformation of EAY1281 and EAY2409 with integration plasmids bearing these mutations using standard techniques [89] . A smaller subset of these msh4 , msh5 point mutants were made in the NHY background by transformation of EAY2844 and EAY2848 respectively . Double and triple mutants bearing different combinations of msh4 , msh5 , pch2Δ and spo11-HA were made in the NHY background by crossing single or double mutant strains followed by tetrad dissection . All strains used in this study are listed in Table S1 . Msh4 amino acid sequence from S . cerevisiae ( YFL003C ) , A . thaliana ( NM_117842 ) , C . elegans ( AF178755 ) , M . musculus ( BC145838 ) , H . sapiens ( NM_002440 ) and Msh5 amino acid sequences from S . cerevisiae ( YDL154W ) , A . thaliana ( EF471448 ) , C . elegans ( NM_070130 ) , M . musculus ( NM_013600 ) , H . sapiens ( BC002498 ) were aligned using ClustalW software ( www . ebi . ac . uk/clustalw ) and CLC free workbench . A Msh4 , Msh5 consensus sequence was generated using CLC and aligned against S . cerevisiae Msh2 ( YOL090W ) , Msh3 ( YCR092C ) , Msh6 ( YDR097C ) to check if residues conserved across Msh4 , Msh5 in all five species are conserved in the other Msh family members . The SK1 MSH4 open reading frame with 600 bp upstream sequence and 400 bp downstream sequence was amplified with pfu DNA polymerase and cloned into pRS416 with a 1 . 5 kb KanMX fragment inserted 90 bp downstream of the MSH4 stop codon to create the single step integrating plasmid pEAA427 . The SK1 MSH5 open reading frame with 500 bp upstream sequence and 400 bp downstream sequence was similarly amplified with pfu DNA polymerase and cloned into pRS416 with a 1 . 5 kb KanMX fragment inserted 45 bp downstream of the stop codon to create the single step integrating plasmid pEAA424 . The MSH4 and MSH5 SK1 sequences in these plasmids were confirmed by Sanger DNA sequencing . pEAA424 and pEAA427 were mutagenized using Quick Change site directed mutagenesis method ( Stratagene , La Jolla , CA ) to create 28 msh5 and 29 msh4 point mutations . The entire open reading frame of MSH4 , MSH5 was sequenced to ensure only the desired amino acid change was introduced . Table S1 shows a list of plasmids bearing the msh4 , msh5 point mutations . Full length SK1 MSH4 , MSH5 and point mutant derivatives were amplified by pfu DNA polymerase and cloned into pGAD424 ( prey ) and target pBTM116 ( target ) vectors kindly provided by Nancy Hollingsworth . The entire open reading frame of MSH4 , MSH5 was checked by DNA sequencing to ensure that no additional mutations were created . The L40 strain [90] was co-transformed with the Prey and Target vectors and expression of the LACZ reporter gene was determined by the ortho-nitrophenyl-β-D-galactopyranoside ( ONPG ) assay [91] . All msh4 and msh5 point mutations integrated into EAY1108 or NHY943 were mated to null strains bearing corresponding msh4Δ ( EAY2411 , EAY background; EAY2843 , NHY background ) and msh5Δ ( EAY1280 , EAY background; EAY2846 , NHY background ) alleles . The resulting diploids were sporulated using the zero growth mating protocol [92] . Briefly , the haploid strains were patched together on synthetic complete media for four hours and then spread on sporulation media and incubated for 2 days at 30°C . Tetrads were dissected on synthetic complete media for the EAY background and on YPD media supplemented with amino acids for the NHY background . Spore clones were replica plated onto selective media or minimal drop out plates and incubated overnight . Segregation data were analyzed using the recombination analysis software RANA to determine genetic map distances for tetrads and recombination frequencies for spores [24] . Time course , DAPI , and immunostaining analyses of meiotic progression were performed as described using antibodies to Zip1 and Msh5 [29] , [93] . Stable SK1 isogenic diploid strains used in the time courses were created by mating the haploid strains shown in parentheses: Wild-type ( NHY942×NHY943 ) ; msh4Δ ( EAY2843×EAY2844 ) ; msh4-E276A ( EAY2849×EAY2843 ) , msh4-R676W ( EAY2851×EAY2843 ) ; msh5Δ ( EAY2846×EAY2848 ) : msh5-S416A ( EAY2855×EAY2846 ) ; msh5-D539A ( EAY2857×EAY2846 ) ; msh5-D532A ( EAY2785×EAY2846 ) .
|
In meiosis , sex cells that become eggs or sperm undergo a single round of DNA replication followed by two consecutive chromosomal divisions . In most organisms , the segregation of chromosomes at the first meiotic division is dependent upon at least one genetic exchange , or crossover event , between homologous chromosome pairs . Matched chromosomes that do not receive a crossover frequently undergo non-disjunction at the first meiotic division , yielding gametes that lack chromosomes or contain additional copies . Such missegregation events have been linked to Down syndrome and human infertility . This paper focuses on Msh4-Msh5 , a complex required for the proper segregation of homologous chromosomes during the Meiosis I division . We performed a mutational analysis of the baker's yeast Msh4-Msh5 complex to study its role in implementing the decision to make a crossover . We identified a class of mutants that are functional in meiosis despite significant reductions in crossing over that occurred primarily on larger chromosomes . In combination with mutations ( pch2Δ , spo11-HA ) that disrupted early steps in crossover placement , this msh4/5 class of mutants displayed poor spore viability . Together , these data support the presence in yeast of a robust crossover distribution mechanism .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"biology/recombination",
"genetics",
"and",
"genomics/nuclear",
"structure",
"and",
"function",
"molecular",
"biology/chromosome",
"structure",
"genetics",
"and",
"genomics/chromosome",
"biology",
"genetics",
"and",
"genomics",
"molecular",
"biology/dna",
"repair"
] |
2010
|
Genetic Analysis of Baker's Yeast Msh4-Msh5 Reveals a Threshold Crossover Level for Meiotic Viability
|
In 2010 , the World Health Organization released a new cholera vaccine position paper , which recommended the use of cholera vaccines in high-risk endemic areas . However , there is a paucity of data on the burden of cholera in endemic countries . This article reviewed available cholera surveillance data from Uganda and assessed the sufficiency of these data to inform country-specific strategies for cholera vaccination . The Uganda Ministry of Health conducts cholera surveillance to guide cholera outbreak control activities . This includes reporting the number of cases based on a standardized clinical definition plus systematic laboratory testing of stool samples from suspected cases at the outset and conclusion of outbreaks . This retrospective study analyzes available data by district and by age to estimate incidence rates . Since surveillance activities focus on more severe hospitalized cases and deaths , a sensitivity analysis was conducted to estimate the number of non-severe cases and unrecognized deaths that may not have been captured . Cholera affected all ages , but the geographic distribution of the disease was very heterogeneous in Uganda . We estimated that an average of about 11 , 000 cholera cases occurred in Uganda each year , which led to approximately 61–182 deaths . The majority of these cases ( 81% ) occurred in a relatively small number of districts comprising just 24% of Uganda's total population . These districts included rural areas bordering the Democratic Republic of Congo , South Sudan , and Kenya as well as the slums of Kampala city . When outbreaks occurred , the average duration was about 15 weeks with a range of 4–44 weeks . There is a clear subdivision between high-risk and low-risk districts in Uganda . Vaccination efforts should be focused on the high-risk population . However , enhanced or sentinel surveillance activities should be undertaken to better quantify the endemic disease burden and high-risk populations prior to introducing the vaccine .
Cholera was first reported in Uganda in 1971 , when 757 cases were reported to the World Health Organization ( WHO ) . During the subsequent years up to 1993 , Uganda reported cholera cases every 2–4 years to the WHO . From 1994 to 1998 , cholera was reported annually in Uganda [1] . In 1998 , Uganda reported almost 50 , 000 cases with incidence throughout the country [2] . The reported incidence has fluctuated between 250 and 5 , 000 cases every year since 2000 ( Figure 1 ) . The reported case fatality ratio has decreased from 4–7% in the late 1990s to about 2–3% during 2004–2010 . Cholera in Uganda appears to be largely an epidemic disease . However , endemic cholera occurs in high-risk areas along the southwestern border with DRC and in Kampala city slums . Endemic cholera is commonly noted before and during the rainy season , from December through March . Epidemic cholera can occur any time , but is often associated with extreme rain events or water supply disruptions . The frequency of reported cholera cases varies among districts in Uganda . The highest risk areas include the border areas with the Democratic Republic of Congo ( DRC ) , Sudan , and Kenya as well as urban slums in Kampala . Displaced populations and their neighboring communities are at elevated risk . The ongoing migration of people into and within Uganda can lead to rapid spread of the disease . The African Great Lakes , including Lake Albert and Lake Victoria on the border of Uganda , may provide a reservoir for cholera bacteria . Further , increases in incidence among the nations bordering these lakes have been shown to be correlated with El Niño warm weather events [3] . The WHO [4] has recently revised its guidelines and states in a position paper that cholera vaccines should be used in combination with other prevention and control strategies in areas where the disease is endemic . “Endemic” is defined as areas with occurrence of culture-confirmed cholera in at least three of the previous five years . The WHO also recommends that cholera vaccines should be considered for preemptive use in areas at risk for epidemic cholera as long as vaccination does not interfere with efforts to treat cholera cases , improve water and sanitation , and mobilize communities . The vaccine may also be considered for reactive use if local infrastructure is sufficient to conduct mass campaigns depending on the current and historical epidemiology . There is a dearth of information about the burden of cholera in low-income countries such as Uganda . A more accurate picture of this burden is particularly important because it can be used to inform cholera prevention and control intervention questions: whether or not to introduce vaccination as a complement to other cholera prevention and control interventions , where and when it would be most effective to do so , and what demographic population should be targeted . This article presents available disease burden data for Uganda that may help inform such questions .
This is a retrospective study in which we collected data from Uganda's health information management system and Diarrheal Disease Control program . District-specific data were used to classify districts as endemic or non-endemic based on the WHO criterion and to identify high-incidence districts . A convenience sample of more detailed data from individual cholera outbreaks were summarized to estimate the age distribution of reported cholera cases and to develop epidemic curves . Because the Ugandan surveillance system is designed primarily to identify and respond to cholera outbreaks , a sensitivity analysis was performed to explore the potential limitations of the existing surveillance system to identify cases outside of recognized outbreaks . These data were used to compile national statistics and for reporting to the World Health Organization's ( WHO ) Weekly Epidemiological Record [5] . The cholera case definition was based on WHO criteria , depending on whether or not cholera is endemic in the area: The identification of such cases should have triggered laboratory investigation . A cholera outbreak was confirmed when Vibrio cholerae O1 or O139 was isolated from at least one stool sample . Only cases meeting the standard case definition above were investigated and included in the official cholera data . Summary laboratory data were obtained from the Head of the Central Public Health Laboratory from the Ministry of Health . Prior to analysis , stool samples from suspected cholera patients were transported from the field in Cary Blair transport media . Culture plates were set at 37°C overnight ( for 18–24 hours ) using three culture media: TCB , XLD and MacConkey . Biochemical identification of cholera organisms were based on Oxidase or Indole tests . Polyvalent antisera were used to differentiate between the Inaba and Ogawa serotypes , and specific monovalent tests further confirmed which of the Inaba , Ogawa or O139 ( Bengal ) strains caused disease . Isolates were refrigerated at −80°C and sent to the WHO's collaborating laboratory ( Unité de La Rage , Institut Pasteur , Paris , France ) for quality control . District-specific data were abstracted from the Uganda Ministry of Health , Health Management Information System disease surveillance database for the period 2005–2010 . The 2005–2010 period was chosen based on the WHO criteria [4] for identifying endemic cholera ( i . e . areas in which cholera has been reported in three of the previous five years . ) The districts shown are based on the 2002 district boundaries , which were in existence during the most recent census . Cases in new districts created after 2005 were apportioned back to the 2002 districts . Age-specific morbidity and mortality data are stored at the district level . These ‘line list’ records include: patient age , outcome of treatment ( i . e . discharge or death ) , and date of admission or death ( for suspected cholera patients who die prior to seeking treatment ) . We were able to obtain these data from 15 outbreaks , which occurred in 12 districts spanning the time period from 2002–2010 . In total , the line list data included records of 6 , 125 cholera cases with at least 154 deaths . The actual number of deaths could not be ascertained because some of the records lack data on patient outcomes . These records included seven instances in which death occurred in the community , ( i . e . prior to receiving treatment ) . In addition , there were 923 records with data on inpatient and outpatient treatment and the duration of inpatient treatment . For this retrospective analysis , the study team compiled data from samples that were previously collected and analyzed as part of routine surveillance activities . The incidence of hospitalized cholera was estimated by district based on the annual average number of cases reported over the six-year period from 2005–2010 . The district-specific reporting does not include data by age group . Thus , the age distribution of cases was estimated based on the 15 line lists . It was assumed that these 15 outbreaks were representative of the age distribution of cholera incidence in Uganda . The numbers of cases by age group were calculated from the product of total cases and the national average percentage distribution of cases by age from the line list data . Age-specific incidence rates were then calculated by dividing the age-specific cases by the age-specific populations ( 2010 UN population data ) . All analyses and graphs were produced with Microsoft Excel ( Microsoft , Redmond , WA ) and maps were created with ArcGIS ( ESRI , Redlands , CA ) . Cholera case fatality rates were estimated from the number of reported cases and deaths by age group from the line list data . The Fisher's exact test was used to compare case fatality rates across age groups . Statistical analyses were performed using STATA software ( Version 8 , College Station , TX , USA ) . In addition to hospitalized cases , we also estimated the number of non-hospitalized or non-reported cholera cases in Uganda . In a recent analysis , Kirigia et al . [7] estimated that 10% of cholera cases could be classified as severe and require hospitalization . In addition , Poulos et al . [8] reported that 22–38% of cholera patients were hospitalized during multi-site surveillance studies conducted in Jakarta , Indonesia and Kolkata , India . In Uganda , patients with mild diarrhea often do not seek formal seek care , but do receive oral rehydration therapy at home . In this analysis , we assumed that the official statistics include 25% of cholera patients with severe cholera who would seek treatment and be reported in official statistics and that 75% took oral rehydration therapy at home . This assumption is greater than that assumed by Kirigia et al . , but falls at the lower end of the actual data presented by Poulos et al . Thus , we estimated that there were three non-hospitalized cholera cases requiring treatment at home per one officially reported case . Estimation of asymptomatic cholera infections were omitted from this analysis . Since the reported numbers of deaths were based on individually-identified cholera patients , these reports should be a lower bound . While deaths that occurred outside treatment facilities were included in official reports when identified , it remains likely that some cholera deaths were missed and not reported in official statistics . In a recent study in neighboring Kenya , an active case finding exercise identified a 200% increase in the number of cholera deaths that occurred during a 2008 cholera outbreak [9] . This is a worst-case-scenario , since the outbreak occurred during a chaotic period of post-election violence . However , in addition to deaths that were missed during outbreaks , isolated cholera deaths that occur outside of recognized outbreaks may also contribute to underreporting in official statistics . As an upper bound estimate of the annual number of cholera deaths in Uganda , we applied the 200% correction factor from the Kenya study to the number of cholera deaths identified in Uganda . In addition , for an upper bound estimate of the number of hospitalized cases , we assumed that a number of severe acute watery diarrheal cases that occurred outside of recognized outbreaks were the result of infection by Vibrio cholerae . These endemic cholera cases have frequently been omitted from totals in other cholera endemic countries [10] , [11] . Thus , we assumed the number of reported hospitalized cases may have only been about 50% of the actual cases although this rate is difficult to estimate in the absence of sentinel surveillance data .
The estimated annual number of cholera cases by age and by district risk group is shown in Table 1 . Inclusive of unreported cases treated at home , our estimated annual average number of cases was around 11 , 000 , with around 81% of the cases occurring in the high risk districts . On average , about 61 cholera deaths were reported per year during 2005–2010 . Using the 200% correction factor reported from the Kenyan study [9] , the potential range of annual cholera mortality is 61–182 deaths per year . The epidemic curves for fifteen cholera outbreaks that occurred in Uganda between 2002 and 2010 are shown in Figure 6 . The average duration of the 15 outbreaks was about 15 weeks from the identification of the first case through the identification of the last case . The range of outbreak duration was between 4 weeks and 44 weeks ( Table 2 ) . Almost half of the observed cases ( 43% ) occurred within six weeks of the first case . It should be noted that the cases reported for Kasese were more likely to be representative of endemic disease , as this district is one of the few that report cases on an ongoing basis . Arua district reported four outbreaks between 2005 and 2008 , but weekly cases declined from the peak observed in early 2008 .
This retrospective analysis showed that there is a clear subdivision between high-risk districts and low-risk districts in Uganda with about 24% of the population residing in high-risk districts accounting for 81% of the average reported cholera burden . These high-risk districts may be considered for preventive cholera vaccination campaigns in combination with other cholera control activities . Cholera affects all age groups in Uganda . The age distribution of cases matched the population distribution . This may be due to low levels of background immunity , so that the entire population is equally susceptible . This age distribution deviates from age distributions in other cholera-endemic areas , where young children tended to be at greater risk when systematic surveillance was conducted [14] . Systematic sampling of diarrheal cases from endemic areas has never been attempted in Uganda and may reveal that outbreak-based surveillance findings are not representative of the true cholera burden . A comparison of age-specific cholera incidence rates from Bangladesh demonstrated that the average age of cholera infection was much higher during outbreaks [15] than for endemic cholera [16] . Outbreak data from the sub-district level suggests that there may be considerable heterogeneity of cholera incidence . Thus , surveillance efforts and reporting should be improved to facilitate better epidemiological characterization of cholera incidence and improved targeting of interventions to reach those at greatest risk . The estimate of 61 deaths per year involves accreditation of all cholera deaths to specific individuals , either at treatment centers or in the community . This is a relatively small fraction of the estimated 30 , 000 diarrheal deaths per year in Uganda ( exclusive of deaths attributed to cholera and bloody diarrhea ) [17] . It is certainly possible that a significant proportion of these 30 , 000 deaths was caused by unrecognized cholera than would be estimated from individually identified deaths . Although the outbreak response focus of cholera surveillance in Uganda may be insufficient to accurately estimate the numbers of cases and deaths caused by cholera , these data are very useful for identifying areas to target for surveillance in consideration of future vaccine introduction . In order to better quantify the burden of cholera in Uganda , sentinel site surveillance should be undertaken in at least two regions with districts at high risk for cholera for a period of at least two years . It would be better to continue surveillance for at least five years , since cholera incidence is highly variable from year to year . The Ministry of Health is participating in the AFRICHOL cholera surveillance in Africa project ( www . africhol . org ) , led by Agence de Médecine Préventive ( AMP ) and the African Field Epidemiological Network ( AFENET ) , which is an African-based non-government organization working to improve field epidemiology and public health laboratories in sub-Saharan Africa . As part of this project , enhanced cholera surveillance is being conducted in five districts in Eastern Uganda ( Mbale , Tororo , Manafwa , Butaleja and Busia ) and , whenever outbreaks occur , throughout the country . Such data may be combined with the available national reporting statistics to better model cholera burden within Uganda , which in turn may be used to conduct economic analyses ( e . g . , cost effectiveness or cost benefit studies ) of the potential use of cholera vaccines in Uganda . Given the health challenges facing Uganda , the decision to pursue cholera vaccination must be weighed against the introduction of other health interventions that may have a greater impact on mortality ( e . g . , pneumococcal conjugate vaccines , rotavirus vaccines , future malaria vaccines or other interventions ) . In addition to targeting high-risk endemic populations , Uganda may consider using cholera vaccines from a recently established international stockpile to mitigate epidemic cholera . A review of 15 epidemic curves showed that about 57% of the cases occurred after six weeks across all outbreaks . This 57% figure may represent an upper bound on the number of cases that could be averted via reactive use of cholera vaccines from a global stockpile , assuming it would take at least three weeks to diagnose an outbreak and prepare for a mass vaccination campaign plus three weeks to generate immunity from the two-dose vaccine . This stockpile may also be used to prevent the spread of cholera to neighboring districts , such as when it was used in an Adjumani district refugee camp in 1997 [18] . While cholera incidence in Uganda has been manageable over the past decade , elimination of the disease is likely to take time especially given the slow progress on provision of safe water and sanitation among other risk factors . Most of the areas with the highest incidence rates either border countries with political instability and endemic cholera ( e . g . , DRC and Sudan ) or contain semi-nomadic populations . For these districts , it would be difficult to prevent cholera-infected persons from crossing borders , achieve high vaccination coverage rates , or to construct reliable water and sanitation infrastructure for semi-nomadic populations [19] . Some global trends in cholera disease burden may lead to an increase in the number of cases and should be considered in cholera control planning . At present , cholera is more prevalent in rural areas than in urban areas within Uganda . This may change if present urbanization trends continue and the maintenance and expansion of water and sanitation infrastructure cannot keep pace with the rapidly growing urban population . The urban population in Uganda is projected to increase more than seven-fold from 4 . 5 million in 2010 to 31 million by 2050 [20] . Some studies have found multidrug resistant V . cholerae in Uganda , including strains resistant to trimethoprim , sulfonamides , ampicillin , tetracycline , chloramphenical and streptomycin [21] . In addition , there is evidence that the severity of clinical cases of cholera in Asia and Africa is increasing , especially during outbreaks . Some scientists attribute the increase in severity of cholera cases seeking treatment to the emergence of a new altered strain of V . cholerae O1 El Tor that secretes the classical cholera toxin , making it more virulent [22] . It is not presently known if this strain is present in Uganda . However , it has been isolated from recent African outbreaks in Mozambique and Zimbabwe [23] , [24] . Due to global warming , the average temperature in Uganda is estimated to increase by up to 1 . 5 degrees over the next 20 years [25] . Recent research suggests a strong correlation between increased rainfall and elevated temperatures with higher cholera incidence [26] , [27] , [28] , [29] , [30] . This may pose an elevated risk for districts bordering Lake Albert and Lake Victoria , which may provide an endemic reservoir of V . cholerae [31] , [32] . There are also trends suggesting a reduced need for cholera vaccination in Uganda . The multidisciplinary cholera outbreak response activities have been effective in mitigating the severity of outbreaks , both in terms of morbidity and mortality . The cholera case fatality rate has steadily declined since the large , nationwide outbreak in 1998 ( refer to Figure 1 ) . While improving treatment does not reduce the incidence of cholera cases , it does reduce the social and economic burden of the disease . Improvements in access to improved water and sanitation would also lead to a concomitant decrease in cholera incidence . These cholera incidence data may also be used to target priority districts for improvements in water , sanitation , and hygiene efforts . Cholera incidence is likely to be associated with high prevalence of other enteric diseases , for which cholera vaccination would have no effect . Considering that an estimated 30 , 000 persons die from diarrheal disease every year in Uganda , improved water , sanitation , and hygiene are urgently needed even if cholera vaccine is deployed . In conclusion , the existing surveillance system is geared toward mitigating the impacts of cholera outbreaks , not quantifying the burden of endemic cholera . Cholera control activities have been effective in slowing the spread of cholera and reducing cholera fatalities . However , cholera cases continue to be reported on an annual basis . The combination of sentinel surveillance with national cholera incidence data could be used to develop an economic analysis to inform cholera vaccination policy .
|
Uganda has reported cholera cases to the World Health Organization every year since 1997 . Thus , the country may consider the introduction of a WHO-prequalified oral cholera vaccine . This article reviews cholera surveillance data from 1997–2010 with a focus on the 2005–2010 time period to identify high risk populations that may be targeted for preventive vaccination campaigns . We estimated that an average of about 61–182 deaths occur each year . Most cases ( 81% ) occurred in a relatively small number of districts comprising just 24% of Uganda's total population . While there is a clear distinction between low and high-risk districts , sentinel surveillance would help to better quantify the burden in endemic districts . An economic analysis should also be undertaken prior to making a decision to introduce a cholera vaccine .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2013
|
The Burden of Cholera in Uganda
|
Information processing is a major aspect of the evolution of animal behavior . In foraging , responsiveness to local feeding opportunities can generate patterns of behavior which reflect or “recognize patterns” in the environment beyond the perception of individuals . Theory on the evolution of behavior generally neglects such opportunity-based adaptation . Using a spatial individual-based model we study the role of opportunity-based adaptation in the evolution of foraging , and how it depends on local decision making . We compare two model variants which differ in the individual decision making that can evolve ( restricted and extended model ) , and study the evolution of simple foraging behavior in environments where food is distributed either uniformly or in patches . We find that opportunity-based adaptation and the pattern recognition it generates , plays an important role in foraging success , particularly in patchy environments where one of the main challenges is “staying in patches” . In the restricted model this is achieved by genetic adaptation of move and search behavior , in light of a trade-off on within- and between-patch behavior . In the extended model this trade-off does not arise because decision making capabilities allow for differentiated behavioral patterns . As a consequence , it becomes possible for properties of movement to be specialized for detection of patches with more food , a larger scale information processing not present in the restricted model . Our results show that changes in decision making abilities can alter what kinds of pattern recognition are possible , eliminate an evolutionary trade-off and change the adaptive landscape .
The evolution of behavior is to a large extent the evolution of information processing [1]–[4] . On short timescales individuals respond to local information in the environment . For instance in foraging , a basic local information processing is that animals detect food , turn and move to food , and eat . On the long term this generates behavioral patterns . The latter shapes how individual behavior relates to patterns in the environment ( e . g . resource distributions ) and affects aspects of Darwinian fitness ( e . g . foraging success ) . At present it is poorly known how local information processing mechanisms ( e . g . cognition ) determine larger scale pattern detection and evolve [3] , [5]–[8] . Here we study the evolution of local information processing and orientation to the environment , and its relation to environmental pattern detection . In evolutionary theory on foraging , the focus is often on how well individuals match ( fitness relevant ) patterns in the environment . In optimal search theory ( OST ) the main focus has been on what kinds of random turning strategies optimize search [9]–[11] . A second focus has been on the value of alternating between intensive searching , once a food patch is found , to extensive search when food has not been found for a while , using combinations of correlated random walks differing in turning rates [12] . Simulations show that such switching between search strategies can enhance foraging efficiency because it concentrates search effort in the right places ( i . e . it allows patches to be “detected” ) , so called area-concentrated search . This is true for models in which “continuous” patchy environments are assumed [12] , [13] , where resource items are only locally detectable , but aggregated on a scale that is beyond the perception of individuals , as apposed to models in which discrete and fully detectable patches are assumed ( e . g . the marginal value theorem [14] ) . Random-walk models have been used to statistically characterize animal movement trajectories , including bi-modal search patterns similar to area-concentrated search [15] , [16] . However , such model fitting does not necessarily reveal underlying movement mechanisms [6] , [17] . Interaction with , and orientation to , the external environment can generate similar movement patterns as those generated by internal turning strategies [6] , [17] , [18] . Moreover , Benhamou showed that local orientation via memory of where an individual last found a food item , can further improve foraging efficiency relative to “random” area-restricted search without such memory [19] , indicating the adaptive value of reacting to external cues . However , like the random-walk search models , an important assumption is that food is detected and consumed on the same range . Instead , if food can be detected beyond the range at which it can be eaten ( as is often the case ) , an animal will be able to approach foraging opportunities from some distance via direct visual cues . This is probably one of the most simple ways through which animals can orientate themselves relative to food . Important is that such opportunity-based adaptation ( or responsiveness ) stands in direct relation to feeding opportunities in the environment . Therefore , on longer timescales , behavioral patterns emerge that are “a reflection of complexity in the environment” [20] . To conceptualize how interaction of individuals with the environment can structure behavior , Hogeweg and Hesper [21] coined the TODO principle . This envisages behavior as multi-scale information processing [22] , [23] ( see Figure 1 ) : ( i ) TODO: individuals behaviorally adapt to local opportunities by “doing what there is to do” , and ( ii ) Pattern formation and detection: behavioral patterns self-organize on larger spatio-temporal scales through the continual feedback between behavior and local environmental contexts ( This use of the term “information processing” differs from that in behavioral ecology where it generally refers only to individual-level behavioral flexibility , often specifically in relation to energy-dependent behavioral choices ) . A simplistic example of TODO is that as food density declines individuals end up moving more and eating less , because there is no opportunity to eat . As such , the environment is like a “behavioral template” to which individuals can respond , allowing individuals to effectively “detect” patterns of opportunities in the environment beyond their own perception . In order to fit models to movement data and elucidate underlying mechanisms , requires a thorough understanding of how both internal and external structuring of behavior can generate foraging patterns . This can be done using pattern oriented modeling [24] and other multi-level modeling approaches [25] , where model fits are evaluated based on patterns on multiple levels: small scale movement decisions , mesoscale patterns such as trajectories and space use and more global patterns such as population distributions . The requirement of fitting models to multiple levels places the focus on the mechanisms that generate the inter-relation between small-scale processes and patterns on larger scales . A thorough understanding of how small scale behavior interactions generate behavioral patterns through TODO could be an important contribution to such modeling approaches . Essentially , TODO and the longer term behavioral patterns it generates , come to expression ( in models ) when individuals interact with the environment and need to make behavioral decisions based on local information . In this light , Hogeweg [26] showed that foragers with simple TODO rules could forage much more efficiently than those with much more complicated rules . This was because foragers with simple rules could react to local opportunities and therefore automatically adapt to larger-scale patterns in the environment ( i . e . generalize their behavior ) . More counter-intuitive and complex behavioral patterns emerge in models with more detailed environmental structure and multiple types of behavior . Examples include “self-structuring” explanations for social dynamics in bumblebee colonies [27] , grouping patterns in chimpanzees [28] , diet learning and cultural inheritance in group foragers [29] , [30] . At present , the role of pattern recognition through TODO is most likely underestimated in most approaches to the evolution of foraging behavior . For instance in OST the simple orientation mechanism of turning and moving to food is generally not included . Moreover , behavior is usually assumed to be continuous in that movement , search and food consumption occur in parallel ( although a trade-off between movement speed and search accuracy is often assumed [12] ) . Decision-making is therefore restricted to changes in direction . However , if movement , scanning for food and eating are at least partially mutually exclusive , then individuals must decide about what to do next ( e . g . search again at a certain location , or move on ) . Such foraging behavior can be referred to as pause-travel [31] , or intermittent search [7] . Here we focus on local orientation towards food in such a setting where individuals must make decisions , and study the role of TODO in the evolution of simple foraging behavior . We ask: how does local information processing evolve in order to determine how individuals “do what there is to do” ? More specifically , how does the responsiveness and orientation of individuals to feeding opportunities in the environment evolve in light of the larger spatio-temporal pattern recognition that this generates ? To address this question , we study the evolution of foraging behavior in a model with individuals that have to choose amongst alternative behavioral actions according to information they obtain through searching . This happens in a spatial environment with patchy and uniform patterns of feeding opportunities . To address how local information processing ( sensing and decision making ) affects information processing on larger spatio-temporal scales ( pattern recognition and genetic adaptation , see Figure 1 ) , we compare the evolution of decision making and properties of behavioral actions in two model variants . In a “restricted” model we limit information individuals can remember and use relative to an “extended” model . The comparison across environments is used to understand evolutionary adaptation to prevailing ecological conditions ( patchy or uniform ) . The comparison across models ( restricted versus extended ) is used to understand how differences in the evolutionary freedom ( or constraints ) for evolving decision making affect evolution . This has similarities to artificial neural network approaches to the evolution of behavior , where behavior is not predefined , but emerges from neural architecture and learning processes [32]–[35] . Such models have been used to show , for instance , that risk-averse foraging can emerge as a side-effect of an evolved reinforcement learning process [33] . In our case there is no learning , but the “architecture” of decision making can evolve such that non-predefined behavior can evolve . Therefore we do not prespecify a selection function , but only define that inter-birth intervals decrease with increased food intake , and allow natural selection to arise from competition in a world with finite resources . We then study how Darwinian fitness arises as an emergent property of how micro-scale interactions generate longer-term behavioral patterns . Thus , we study evolution as the interplay of information processing on multiple timescales ( Figure 1 ) , based on bioinformatic ( processes ) theory [22] , [23] , [36]–[38] . Using this approach , we show that local information processing and opportunity-based adaptation can play a significant role in detecting patterns of resources in the environment , and the evolution of foraging . In particular , we find that the differences in decision making capabilities affect how individuals interact with the environment ( TODO ) , and this can alleviate evolutionary trade-offs and allows for novel pattern recognition specializations .
Our model incorporates ( i ) individual foragers and ( ii ) a 2-dimensional environment with resource items in either a patchy or uniform distribution , adapted from van der Post and Hogeweg [29] . Individuals have a decision making algorithm which determines the sequence and context dependency of the following behavioral actions: MOVE , FOODSCAN , MOVETOFOOD and EAT . Each of these behavioral actions has specific properties ( such as distances , angles etc ) . Our model is event-based , which means that actions take time . When individuals complete an action they choose a new one . The individual with the shortest time to complete its action is next to choose a new action . We study two model variants ( “restricted” and “extended” ) which differ in the type of decision making algorithm that can evolve . Both the parameters of the decision making algorithm and the details of behavior are “genes” which change through mutation . This generates genetic variation , which may result in differences in foraging efficiency and rates of reproduction . Natural selection then arises from resource competition . For a full list of model parameters please see Table 1 and 2 . Next we discuss the model in more detail . Our environment is 5660 by 5660 lattice , where grid points are scaled to be 1 meter apart , giving 32 , 035 , 600 grid points ( 32 . 035 km squared ) . This size was chosen to support a population size ( about 100–150 individuals ) . This was the minimal population where: ( i ) parameters evolved , ( ii ) the population is self-sustaining , and ( iii ) simulations are completed in a reasonable time span . It also ensures that individuals need to move through space to find food , survive and reproduce . Resource items were placed on grid points . Resource items appeared at fixed , but randomly assigned time points within a year , and remained there until eaten . If eaten the resource item was depleted , and appeared again at its fixed time point in the year . Days are 720 minutes ( 12 hours of “daylight” ) and years are 365 days ( 262800 minutes ) . We implement a patchy and a uniform environment , where we keep the total number of food items constant and only vary the resource distribution . In the patchy environment we placed 8000 patches , each with about 2500 items depending on overlap of randomly positioned patches . Each patch is a circle with a radius of 20 meters . Within this circle , 2 resource items are placed at each grid point . All resource items in a patch appear at the same time point , and different patches appear at random fixed times in the year . In the uniform environment resources are placed with probability 0 . 535 per grid location to match the total number of resources placed in the patchy environment ( 17150000 items ) . In the uniform environment , resource items appear at randomly assigned fixed times throughout the year . The restricted and extended model differ in the decision making that can evolve . Figures 1a and b show the basic decision making algorithms: the behavioral actions that are possible ( ovals ) and in the case of FOODSCAN , the information this provides ( rectangles ) . Arrows indicate what can be done next , or what information is obtained ( after FOODSCAN ) , and an individuals last action ( + information obtained ) represents its “state” ( or memory ) . EAT and MOVETOFOOD can only occur after food is detected . EAT occurs when food is detected in range , otherwise individuals first MOVETOFOOD ( MTF ) and then EAT . Without any information about food , individuals can either MOVE or do FOODSCAN . As a starting condition , we set these to alternate so that individuals always do FOODSCAN after MOVE and vice versa . To allow decision making to evolve we define parameters which determine the probability of moving again after MOVE ( ) and scanning again after FOODSCAN ( ) ( Figure 2a ) , searching again after EAT ( ) , or searching again after NO FOOD ( ) ( Figure 2b , see also Table 2 ) . This is indicated by decision points ( black diamonds ) after MOVE , NO FOOD and EAT , where arrows split . For each of these probabilities , the alternative decision has a probability of . For the restricted model we only allow and to evolve , where is a general probability to do FOODSCAN again , irrespective of whether individuals have eaten or did not find food ( Figure 2a ) . Thus in the restricted model , the probability to do FOODSCAN again after EAT or after NO FOOD , is determined by the same parameter ( ) . For the extended model we allow , , and to evolve ( Figure 2b ) , where , and can be seen as context dependent forms of . In the extended model , the probability to do FOODSCAN again after EAT or after NO FOOD , can therefore evolve independently . Thus , in the restricted model individuals cannot remember and make use of the additional information “just ate” or “didn’t find food” to determine the probability to do FOODSCAN again , while in the extended model they can . Moreover , in the restricted model , we assumed individuals always MOVETOFOOD when food is out of reach . In the extended model we allowed this probability ( ) to evolve , and it always evolved to ( see section 2 in Text S1 and Figure S1 ) . The parameters of specific behavioral actions determine how individuals move and sense their environment ( see Figure 2c ) . Unless stated otherwise , we allow all these parameters to evolve: where is the area scanned ( ) , and where 1 second of scanning for 1 gives . The closest detected item is chosen for consumption . If there are multiple items equally close , a random closest item is chosen . This scanning algorithm therefore represents the case where individuals eat the first item they find . Note also that we assume that MOVE and FOODSCAN cannot occur at the same time , and thus we focus pause-travel foraging [31] or “intermittent search” behavior [7] . Individuals gain energy through food ( energy units per item ) which is added to their energy store ( with a maximum: ) . To survive , individuals must have energy ( ) , which means energy intake must compensate basal metabolism ( , which is subtracted from every minute ) . Because resources become locally depleted individuals must move to eat . We do not add explicit movement costs , but time spent moving cannot be spent eating . Individuals reproduce when . Energy is then halved and the other half goes to a single offspring . The time taken to get back to defines a birth interval . Individuals with shorter birth intervals achieve greater lifetime reproductive success . Individuals can die with a probability of 0 . 1 per year , and can reach a maximum age of 10 years . This adds some stochasticity in survival and limits lifespans to 10 years . Since resources are limited in the environment , the population grows until the reproduction is at replacement rate ( carrying capacity ) . Our model requires that the population is viable in relation to resource availability , thus energy and life-history parameters are chosen such that at low population sizes individuals can definitely gain sufficient energy to reproduce . Moreover , to focus on movement and foraging in differently patterned environments , we set the energy required to give birth in relation to energy per food time , and the density of food items in space , such that individuals have to move to and forage from many food patches and experience the full scale of environmental patterns during a reproductive cycle ( i . e . they cannot complete reproductive cycles within a single patch ) . Lifespan is set to allow multiple reproductive events per individual . We expect most parameter combinations that satisfy these qualitative relationships ( see section 1 in Text S1 for more detail ) , to give similar results . When individuals reproduce , the parameters of decision making and behavioral actions are inherited by offspring , with a probability of mutation of 0 . 05 per gene ( this rate of mutation was chosen after observing that natural selection lead to consistent evolutionary change with increases in foraging efficiency ) . We allow all action durations , distances and angles to evolve except and . The mutation “step” is defined by drawing the parameter value from a normal distribution with the mother’s parameter value as mean and standard deviation scaled to about 20% of the range of values that is relevant for that parameter ( see Table 2 ) . Moreover , in order to keep simulations running fast enough , we limited the minimal action duration to seconds . Most mutations are close the mother’s parameter value , but larger jumps are possible . This was chosen to make evolution of parameters possible without predefining their ranges . We cannot predict what parameter settings are viable and take a “zero” state ( all parameters zero ) as initial condition . To make sure the population does not die out initially , we use a birth algorithm in which the non-viable population is maintained at a minimum of 10 individuals , and let it evolve to a viable state . During this time , if the population drops below this minimum then an individual is chosen to reproduce according to a probability ( ) relative to its energy ( ) : ( 2 ) Energy costs of reproduction and energy of offspring as the same as before . Once the population grows above 10 individuals and becomes viable , this algorithm is not used anymore . At this point the population grows to carrying capacity and becomes stable . For our study we used the following types of simulations:
We find that in both models the population evolves to environment specific attractors . We refer to these evolved states as “specialists”: uniform specialists in the uniform environment , and patch specialists in the patchy environment . These four specialists differ from each other and these differences depend on the following parameters: ( i ) probabilities to SEARCH again ( , , ) , ( ii ) probability to MOVE again ( ) , ( iii ) MOVE distance ( ) , ( iv ) turning angle ( ) , and ( v ) FOODSCAN angle ( ) ( see Figure 3 ) . For ease of reference we name the specialists and summarize their distinguishing features as follows ( illustrated in Figure 4 ) . Parameter values shown are means of ancestor traces between year 800 and 900 ( see also Table S1 ) : Further analysis revealed that variation of both probability to repeat move ( ) and turning angles ( ) did not impact food intake significantly . For both parameters we found that evolved values result from evolutionary drift because of a very flat adaptive landscape ( for more detail see Text S1 section 2 and Figure S1 and Text S1 section 4 and Figure S3 and S4 ) . Moreover , other parameters did not differ between specialists: durations evolved to minimal values ( see section 2 in Text S1 and Figure S1 ) and food scan range ( ) converged to between 2–2 . 5 meters ( see sections 2 and 3 in Text S1 and Figure S2 ) . From here on we focus on those parameters that generated differences in foraging efficiency between the specialists , namely: , , , and . We use the means of evolved parameter values to characterize each specialist ( see Table S1 for a complete list of average evolved parameter values ) . The values of the evolved decision making parameters mean that in the extended model decision making evolves to: always do FOODSCAN after EAT , always MOVE after NO FOOD ( and , Figure 4c and d ) . This generates a clear differentiation of behavior in food and non-food contexts ( Figure 4c and d , blue and yellow loops respectively ) . Thus in a food context individuals continue to do FOODSCAN until they no longer find food ( blue loop ) . This generates efficient FOODSCAN - EAT - FOODSCAN - EAT sequences and allows systematic depletion of resources at a given location . During this time any movement is via MOVETOFOOD when food is out of range , always towards food . Only when no more food is found do individuals MOVE . Thus in a “no food” context , individuals switch behavior and no longer repeat FOODSCAN ( yellow loop ) . In the restricted model only the patch specialist ( R-Patchy ) has a certain degree of repeated scanning for food ( , Figure 4a ) . However this happens equally after EAT and NO FOOD , because differentiating behavior relative to FOOD and NOFOOD is not possible . This specialist therefore can only to a certain extent avoid MOVE in the presence of food , and is more limited in generating time efficient FOODSCAN-EAT sequences and to only MOVETOFOOD when food is beyond REACH . In contrast the uniform specialist ( R-Uni ) of the restricted model never repeats FOODSCAN ( Figure 4b ) . It only searches once per location and generates MOVE - FOODSCAN - EAT or MOVE - FOODSCAN - MOVETOFOOD - EAT sequences . For behavioral actions the most obvious difference between the specialists is that between the patch specialists of the different models ( illustrated in Figure 4a and c ) . R-Patchy’s maximum FOODSCAN angle in combination with its short move distance leads to a behavioral pattern with a large overlap in areas searched after each MOVE . In contrast , Ext-Patchy’s smaller FOODSCAN angle with long move distance generates a pattern with long distances in which it does not scan , followed by food directed movement when food is detected . The difference between the uniform specialists is more subtle ( Figure 4b and d ) . The shorter MOVE of R-Uni leads to considerable overlap in areas scanned after each MOVE . Ext-Uni’s longer MOVE leads to hardly any overlap in areas scanned after each MOVE . To qualitatively reveal larger-scale behavioral patterns , we visualize the movement trajectories of all evolved specialists in both environments using ecological simulations ( Figure 5 ) . Most striking is that it is difficult to distinguish between the specialists in the same environment , because they all adapt flexibly to both environments , whether they evolved there or not . This is because all specialists are responsive to opportunities in the environment , and have the same basic TODO ( “do what there is to do” ) : move when there is no food , turn and move to food when out or reach , and stop to eat . In the uniform environment this generates random-walk-like patterns reflecting the random encounters with food . In the patchy environment TODO generates a bi-modal pattern of straight movements between patches and frequent turning and remaining localized for some time within patches . Thus irrespective of genetic adaptations , through ( automatic ) opportunity-based adaptation all specialists are able to generalize their behavior to an environment in which they did not evolve . The large-scale behavioral patterns of individuals reflect patterns of feeding opportunities in the environment ( patchy or uniform ) . The more accurate this reflection , the better individuals “detect” resource patterns , and this affects their foraging success . An individual’s genotype determines how it responds to opportunities in the environment , and we find that the genetic adaptations of specialists increase their foraging success relative to the environment they evolved in ( Figure 6 ) . Overall , differences in food intake rates of evolved specialists , as measured in ecological simulations , are as follows: ( Figure 6a ) . ( Figure 6b ) . where represents a minor difference , and a large difference . In both environments , specialists from the extended model are the most successful foragers . Interestingly , Ext-Uni is not only the best forager in the uniform environment , but the second best in the patchy environment . In the uniform environment , Ext-Uni has about 9% greater food intake than R-Uni ( this difference is significant: Wilcoxon rank sum test , , . For Ext-Uni: ; ; . For R-Uni: ; ; ) . In the patchy environment , Ext-Uni has on average about 11% lower food intake than Ext-Patchy ( this difference is significant: Wilcoxon rank sum test , , . For Ext-Uni: ; ; . For R-Patchy: ; ; ) . However , Ext-Uni has nearly 2 times greater food intake than R-Patchy , even though it did not evolve in the patchy environment ( unlike R-Patchy ) . In contrast , Ext-Patchy is the least successful forager in the uniform environment , although average food intake is only about 3% lower than R-Patchy ( but this difference is significant: Wilcoxon rank sum test , , . For Ext-Patchy: ; ; . For R-Patchy: ; ; ) . Overall , differences in the patchy environment are greater ( 2 fold versus a 1 . 5 fold maximum difference in the uniform environment ) , indicating more room for specialization . To understand these results we look in detail at how changes in decision making and behavioral actions affect food intake . The difference in decision making capabilities of the two models has a profound effect on the evolutionary landscape . This is most clear in the patchy environment , where the enhanced information use in the extended model allows a trade-off on within- and between-patch behavior to be eliminated . Therefore , while we find that evolved parameters in both patch specialists reflect a tendency to maximize food intake by ( i ) trying to stay in patches , and ( ii ) minimizing inter-patch travel , how this is achieved depends on how the underlying decision making capabilities shape the evolutionary landscape . This is most clearly illustrated with a local adaptive landscape characterization around the evolutionary attractors relative to the probability to search again ( and ) and move distance ( ) . We consider how parameters affect yearly food intake ( “fitness” ) , and how this depends on inter-patch travel , patch visits time ( i . e . how much they manage to eat in a patch ) and size of patches visited ( Figure 7 ) . The comparison between the extended model ( top ) and the restricted model ( bottom ) reveals a significant shift in the location of the adaptive peak ( Figure 7a top and bottom , yellow zone ) , which coincides with evolved parameter values ( indicated by black circles ) . In the restricted model we can understand the location of the adaptive peak ( and evolved parameters ) in terms of a trade-off between inter-patch travel rate , and patch visit times . As one increases , the other declines ( compare Figure 7b and c bottom row ) . This is because in order to stay in patches ( and find food ) , individuals need short move distances and repeated food scans , otherwise they prematurely leave the patch . However , this slows down inter-patch travel with redundant search . The evolutionary attractor is therefore located where interpatch-travel time and intrapatch-travel time are such that food intake is maximized ( Figure 7a , bottom ) . As a result R-Patchy has the slowest inter-patch travel of all specialists ( see section 5 in Text S1 and Figure S5 ) . Moreover , this is also why R-Patchy has such a large food scan angle , because this allows it to “turn back” when it inadvertently leaves a patch ( see section 3 in Text S1 and Figure S2 ) , and why it does not evolve repeated moving ( see section 4 in Text S1 and Figure S3 ) . In the extended model this trade-off does not arise . Here decision making allows differentiation of behavior: food scanning is only repeated after eating and does not occur during inter-patch travel ( no food encountered ) . Repeated food scanning can therefore evolve to maximal values , which allows individuals to move systematically from one food item to the next within patches via MOVETOFOOD . This leads to longer patch visit times ( Figure 7c top ) and enhanced patch depletion . Unlike in the restricted model , MOVE is now used purely for inter-patch travel . Move distance ( ) is then freed from the trade-off between inter- and intra-patch travel because it no longer affects patch visit times . The enhanced decision making in the extended model therefore eliminates the trade-off , allowing both extended model specialists to be more efficient than R-patchy . As a consequence of the trade-off disappearing , move distance evolves to much longer distances ( Figure 3c ) because this allows individuals to bias foraging to larger patches ( Figure 7d top ) . ( Note that while we implement patches of a fixed size , partial depletion of patches generates smaller patches . ) In fact there are two feedbacks which affect that individuals bias their patch visiting to larger patches: ( i ) by extending patch visiting times , an individual visits on average larger patches longer , and ( ii ) by reduced scanning for food while moving during inter-patch travel ( i . e . due longer move distances ) individuals are less sensitive to each food item on their way . Thus they are more likely to find food and stop moving when local resource densities are higher . Effectively this allows individuals to “select” larger patches . Therefore , for the same time spent traveling , Ext-Patchy manages to find on average larger patches and eat more than Ext-Uni ( see section 5 in Text S1 and Figure S5 for more detail ) . Long move distances also generate more neutrality for repeated move and turning angles , allowing them to evolve ( see section 4 in Text S1 and Figure S3 and S4 ) . For the uniform specialists we also find a difference between the extended and restricted model . Both specialists tend to maximize food intake by ( i ) not wasting time searching depleted areas , and ( ii ) not moving too far and skipping food items on the way . However , in the extended model food intake peaks at maximal repeated search after finding food , while in the restricted model food intake peaks at minimal repeated search and slightly shorter move distances ( Figure 8a , top and bottom respectively ) . In both cases , local depletion of food causes that individuals who move further during MOVE , find a greater average density of food during their next food scan ( Figure 8b ) . However , the further individuals move the longer they travel between food items ( Figure 8d ) . By repeating food scans , travel between food items can be reduced because several food items can be eaten at a given location ( Figure 8d , see interaction between and ) . However , for the restricted model , redundant food scanning ( when no food is found ) rises quickly with repeated food scanning ( Figure 8c , bottom ) , because FOODSCAN also happens after not finding food . The best option is therefore not to repeat food scanning ( and therefore not systematically deplete a given location ) , but not move too far , as to not miss undepleted food items on the way . In the extended model , repeated food scanning only occurs after eating , and redundant food scanning is avoided , unless individuals do not move far enough ( Figure 8c , top ) . Here the best option is therefore to always repeat food scans , systematically deplete a given location and move somewhat further than in the restricted model , to avoid a larger depleted area . Overall Ex-Uni is more efficient than R-Uni ( Figure 6a ) . Both are more efficient than patch specialists in the uniform environment ( Figure 6a ) , because these either have too much redundant overlap in search ( R-Patchy , due to repeated search ) or skip too many resources on the way ( Ext-Patchy , due to long MOVE distance ) ( see section 5 in Text S1 and Figure S6 for more detail ) . To further evaluate our results we studied evolution in an intermediate patchy environment ( twice as many patches , but half the density of resources ) and a mixed environment ( half resources uniform half patchy , only with extended model ) . In the intermediate patchy environment we find that foraging parameters evolve to be qualitatively the same as our main patchy environment both in the extended ( parameter averages are: , , , ) and restricted model ( parameter averages are: , , , ) . This indicates that the behavioral adaptations in the patchy environment are relatively robust to this change in patchiness although the difference in search angles is less pronounced . It is however likely that much smaller patches would select for smaller move distances , because in the mixed environment we find that the extended model evolves to be most similar to Ext-Uni ( parameter averages are: , , , ) . This makes sense given that Ext-Patchy does much worse than Ext-Uni in the uniform environment compared to the performance of Ext-Uni relative to Ext-Patchy in the patchy environment ( see Figure 6 ) . Selection for generalizability in more heterogeneous environments will therefore probably lead to Ext-Uni type genotypes .
Much of foraging theory focuses on foraging efficiency , and uses optimality predictions to assess the foraging behavior of animals ( e . g . optimal foraging theory [4] , [39] , and optimal search theory [9]–[11] ) . Foraging optima are often specified relative to constraints ( e . g . body size , morphology , mode of locomotion , information processing abilities ) [4] . However , this does not necessarily give insight into why a species faces particular constraints , since “constraints” are also often evolvable . At present little is known about how constraints arise and change in the evolution of behavior , though presumably this has been a driving factor in the evolution of morphology and information processing abilities ( e . g . sensing and cognition ) . If we assume that in our model the change from restricted to extended decision making represents an evolutionary innovation in information processing , our results show how small evolutionary changes in decision making can lead to a “release from constraints” on a larger scale and shift the system to a new local optima ( i . e . going from the bottom to top landscape in Figure 7a ) . This reveals how the inter-relation between local information processing and larger scale behavioral patterns allows a small increment in memory ( i . e . remembering the outcome of a previous search event ) to generate a cascade of consequences: ( i ) differentiation of behavior , ( ii ) altering the adaptive landscape and eliminating trade-off constraints and ( iii ) allowing novel foraging specializations . Such insights are relevant for studying the evolution of cognition , which is likely to involve changes in constraints and behavioral opportunities [40] , [41] . Moreover , in light of evolving cognitive complexity our model provides a useful reference . For instance , to establish the impact of elementary spatial cognition such as “remembering where one last found food” , it is probably more appropriate to use TODO-based patch detection as a baseline , rather than random-walks ( as in [12] ) , if individuals can orientate towards food on a local scale without memory . This is also true in terms of model fitting to data to establish mechanisms used by animals during movement . An interesting study by Morales et al . [35] used a spatial grid based model to study movement behavior in elk , assuming that individuals perfectly know the vegetation state of 8 neighboring cells around an individual’s location , and know with less accuracy the state of cells 1 and 2 steps further . Their results show interesting similarities to movement patterns in real elk , and like in our study , shows how orientation to cues in the environment structure movement patterns . However , given the relatively coarse grained resolution of their lattice ( 28 . 5 by 28 . 5 meters ) , their model does not allow for smaller-scale processes via local visual cues , but assumes spatial cognition . In principle it is possible that if food availability patterns traverse the larger scale grid boundaries of Morales et al . ’s model , TODO-based processes could allow individuals to move from grid cell to grid cell according to food availability without using spatial memory . The point here is not to claim the elks couldn’t use spatial memory , but that pattern recognition via TODO could be underestimated . To address this requires models and data with a greater spatial resolution . Our results also have implications for understanding extensive and intensive search behavior . First , we show that a bi-modal search pattern easily self-organizes from TODO in patchy environments in all evolved specialists whether they evolved there or not . This bi-modal pattern is not an evolved strategy , but simply a reflection of the environment . Bi-modal movement patterns are therefore the default expectation in patchy environments . Secondly , in terms of the extended model , we show how a simple mechanism generating extensive and intensive search modes can be created by evolution . Here there is a difference with the model of Benhamou [12] , where bi-modal search is assumed as an adaptive strategy , and studied as a combination of random walks . We find that the regulation of switching between extensive and intensive search does not evolve as a specific strategy in the patchy environment , because it also evolves in the uniform environment ( Ext-Uni also shows intensive and extensive search ) . Instead we find that the specific adaptations in Ext-Patchy function to refine the self-organized extensive and intensive search in order to enhance a new kind of pattern detection: implicitly finding larger patches . This latter pattern detection is not usually considered in foraging theory , but may play an important role in foraging success . Given the focus of optimal search theory on internally-driven turning strategies [6] , [7] , [9] , [11] , it is surprising that we do not find any significant evolution of turning angles . This suggests that in some cases externally-driven turning behavior may pre-empt any need for internally-driven turning strategies and that opportunity-based orientation towards food may be an under-represented aspect in this field [6] , [17] . Moreover , we show how individuals can generalize their behavior across environments via TODO , while fixed internally-driven turning strategies are less robust because they need to be specified to a given environment . However , our results depend on the fact that individuals can detect food items from beyond their reach . This may often be the case in animals , but not always , especially if food items are very cryptic . Moreover , given our simplistic implementation of turning behavior , and other model assumptions ( e . g . random turning at environment boundary , intermittent searching ) , more work is needed to specifically address the relationship between internally- and externally-driven turning . In terms of the evolution of behavior , the value of our results lie in revealing how small changes in decision making and memory have profound influences on multiple scales relevant for individuals foragers . Clearly our foragers are simplistic ( especially cognitively ) and therefore it is unlikely that the local optima we find are directly relevant for a given animal species . However , we show that TODO can be a means through which animals could detect larger-scale environmental patterns , which should be taken into account . Moreover we find that extensive search modes can be used to implicitly detect larger food patches in the environment . These findings can be useful to consider when modeling foraging processes and its fitness consequences . Thus our results provide a useful baseline for understanding the evolution of behavioral flexibility and how evolutionary changes in cognition can alter trade-off constraints and adaptive landscapes .
|
Animals differ in how they sense and process information obtained from the environment . An important part of this information processing is used to find food . In terms of foraging , local decision making determines how successful individuals are at finding food on longer timescales . Using an artificial-world model , we studied different kinds of decision making to understand how local information processing affects larger scale behavioral patterns and their evolution . We compared a restricted decision making ( less memory ) to extended decision making ( more memory ) . We then compared the evolution of decision making and behavioral actions ( moving and scanning for food ) in patchy and uniform environments . Our results show that with restricted decision making individuals face a trade-off in the patchy environment: they try to stay in patches by not moving forward too far , but to do so they sacrifice how fast they travel between patches . With extended decision making this trade-off completely disappears because decision making allows moving forward to be avoided in patches . Instead moving forward can be used exclusively for faster traveling between patches and for selecting bigger patches . Our results show how changes in local decision making can significantly alter what evolutionary forces are faced and can eliminate evolutionary trade-offs .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"theoretical",
"biology",
"ecology",
"biology",
"computational",
"biology",
"evolutionary",
"biology"
] |
2011
|
Local Orientation and the Evolution of Foraging: Changes in Decision Making Can Eliminate Evolutionary Trade-offs
|
Trans-splicing of leader sequences onto the 5′ends of mRNAs is a widespread phenomenon in protozoa , nematodes and some chordates . Using parallel sequencing we have developed a method to simultaneously map 5′splice sites and analyze the corresponding gene expression profile , that we term spliced leader trapping ( SLT ) . The method can be applied to any organism with a sequenced genome and trans-splicing of a conserved leader sequence . We analyzed the expression profiles and splicing patterns of bloodstream and insect forms of the parasite Trypanosoma brucei . We detected the 5′ splice sites of 85% of the annotated protein-coding genes and , contrary to previous reports , found up to 40% of transcripts to be differentially expressed . Furthermore , we discovered more than 2500 alternative splicing events , many of which appear to be stage-regulated . Based on our findings we hypothesize that alternatively spliced transcripts present a new means of regulating gene expression and could potentially contribute to protein diversity in the parasite . The entire dataset can be accessed online at TriTrypDB or through: http://splicer . unibe . ch/ .
Trypanosoma brucei is a unicellular eukaryotic parasite with a digenetic life cycle alternating between the tsetse fly and a variety of mammalian hosts . Besides its importance as a human and veterinary pathogen it has been key to the discovery and understanding of general biological principles such as RNA editing , antigenic variation , GPI anchoring and trans-splicing [1] , [2] , [3] , [4] , [5] . The genome sequence of the 11 megabase-sized chromosomes of T . brucei revealed a compact structure containing about 9000 predicted genes , including 900 pseudogenes and 1700 genes specific to T . brucei [6] . The majority of protein coding genes in trypanosomes is organized in polycistronic units that are transcribed by RNA polymerase II ( Pol II ) [7] . Polycistronic RNA precursors are processed into mature monocistronic mRNAs by trans-splicing of a 39 nt leader sequence to the 5′ end and polyadenylation of the 3′ end [5] . With the exception of the actin promoter , no Pol II promoters for protein-coding genes have been identified , reviewed in [8] . Recently , an elegant study showed correlation between the position of histone marks and putative transcription start sites , suggesting that chromatin structure plays a major role in directing Pol II to its promoter sites [9] . While promoter structures are still elusive , it has long been known that transcription itself is regulated very little , if at all . The regulation of gene expression occurs mainly at the level of RNA stability , translation and protein stability ( for review see [8] ) . Four microarray analyses have shown the transcriptome of the organism to be rather static , with only 2–10% of the transcripts being stage-regulated [10] , [11] , [12] , [13] . This limited degree of regulation at the level of transcript abundance is surprising , given the fundamental changes in morphology and metabolism that occur during the development of the parasite , and especially in the light of the large differences that occur in the closely related species Trypanosoma cruzi , where approximately half of the genome is regulated at the level of transcript abundance [14] . To date , very little is known about the way in which trypanosomes regulate translation , but in many other eukaryotes the 5′ untranslated regions ( UTRs ) contain sequence elements that are responsible for regulating protein synthesis . In order to analyze the T . brucei genome for such elements it is crucial to delineate the 5′ UTR of each expressed gene . In the past , bioinformatics approaches have been used to predict 5′ splice sites in T . brucei , but few of these have been confirmed experimentally [15] . Using a novel high throughput parallel sequencing approach that we term spliced leader trapping ( SLT ) , we have now mapped the vast majority of 5′ UTRs and analyzed the developmental regulation of transcript abundance in bloodstream and insect form trypanosomes . SLT also provides a means of selectively analyzing the transcriptome of parasites without having to purify them from host tissue . Furthermore , since trans-splicing of a spliced leader has been identified not only in protozoa but also in cnidarians , nematodes , flatworms and ascidians , SLT could potentially be applied to a wide variety of organisms [5] , [16] , [17] , [18] , [19] , [20] .
Long slender and short stumpy bloodstream form trypanosomes are the proliferative and quiescent life cycle stages , respectively , in the mammalian host , while the procyclic form is a proliferative form in the midgut of the insect host . In order to map the 5′splice sites and quantify abundance of the corresponding transcripts in these three life cycle stages , poly ( A ) -RNA was purified and first strand cDNA was synthesized using random hexamers ( Figure S1 ) . In order to process multiple samples in one sequencing channel a four-nucleotide barcode was added to the 3′end of the cDNAs . After amplification and size fractionation to 120–160 base pairs the cDNA library was sequenced on a Genome Analyzer ( GA-II , Illumina ) using the Chrysalis 36 cycles v 3 . 0 sequencing kit and 76 cycles . Base calling was performed using the Genome Analyzer Pipeline and linker sequences were removed while separating reads according to identified barcodes . The estimated error probability of the retained bases was 1 . 1% . Inserts containing a barcode were up to 70 base pairs ( bp ) with an average insert size of 42 bp ( median 45 bp , σ 17 bp; Figure 1 ) . We obtained 4 . 6 million sequence tags that could be attributed to one of the libraries ( Table 1 ) . Of these 4 . 5 million tags ( 98% ) could be aligned to one of the 11 megabase-sized chromosomes of T . brucei , the Antat 1 . 1 variant surface glycoprotein ( VSG; 227 , 063 reads ) or the telomeric expression sites ( 40 , 333 reads ) . Depending on the library , 77% to 79% of all genes had at least one tag associated with their 5′ UTR or annotated coding sequence . The median number of tags per gene ranged from 12 in the long slender bloodstream form to 24 in the procyclic form ( Table 1 ) . The dynamic range of the SLT method is best described in the bloodstream library where it spanned more than 5 orders of magnitude from 1 tag for an individual gene to 205 , 410 tags or 11% of all tags in that library for the Antat1 . 1 VSG . The majority of genes without any tags were hypothetical unlikely ( 316 ) , hypothetical ( 118 ) , expression site associated genes ( 116 ) and hypothetical conserved genes ( 103 ) . Using information for the positions of snoRNAs , rRNAs and tRNAs , together with the direction of transcription , we estimated the number of polycistronic transcription units to be approximately 200 . Transcription profile analysis of the majority of individual transcription units containing five or more protein coding genes ( excluding VSG genes ) indicated large variations in transcript abundance between adjacent genes . Of 270 monocistronic mRNA transcription units that have been annotated , 76 had major splice sites less than 2kb upstream of the ATG and eight were differentially expressed . It has recently been shown that procyclin-associated genes ( PAGs ) are located preferentially at a subset of strand-switch regions , or between polycistrons that are transcribed by different polymerases , and that a region within PAG1 could silence transcription by Pol I [21] . Based on SLT analysis , mature transcripts from genes downstream of PAGs were not observed for five of the six annotated sites . In one case transcripts were absent from the procyclic form , while the bloodstream form produced spliced transcripts downstream of two PAG-like genes ( Tb11 . 01 . 6210-6220 ) . It has been documented that overlapping sense and antisense transcription can occur at a single locus , giving rise to processed mRNAs from both strands [22] . We identified 140 unique antisense spliced leader addition sites with ≥1 tag per million ( TPM ) , 60 of which were detected in more than one life cycle stage . Thirty-eight of the unique antisense splice leader addition sites were found on the reverse strand of a hypothetical gene at the end of a transcription unit and twenty-three were in strand-switch regions . In order to evaluate if the expression profiles were consistent with previously published data we first analyzed 20 well-studied genes and found good agreement between the direction and magnitude of change in transcript abundance between life cycle stages ( Table S1 ) . These included the phosphoglycerate kinases PGKB and PGKC [23] , the nuclear-encoded cytochrome oxidase complex subunits ( IV–X ) [24] , the terminal alternative oxidase [25] , invariant surface glycoproteins ( ISGs; [26] ) and the major surface glycoproteins , VSG and procyclins [27] . In total , 3554 genes or ∼40% of the genome significantly changed expression levels in at least one of the three life cycle stages ( statistical significance of <10−5; [28]; Table 2 ) . More than 2000 differentially expressed transcripts could be observed between the bloodstream and procyclic forms , while 1226 changes in expression level could be detected between the proliferative long slender and quiescent short stumpy bloodstream stages . ( Figure 2; Table 2 ) . Of the entire transcriptome 5472 transcripts ( 60% ) did not change abundance between life cycle stages; 919 of these transcripts were represented by ≥25 TPM in all stages . We further evaluated the SLT approach by comparing the data to a previously published tiling array study [11] . Transcripts more abundant in the procyclic form correlated with a coefficient of 0 . 77 or 0 . 93 to the SLT approach , depending on the statistics used by Koumandou et al . ( Figure S2 , [11] ) . Transcripts more abundant in the bloodstream form showed a correlation coefficient of 0 . 23 or 0 . 43 ( Figure S3 ) . When we compared our data to the most recent microarray study by Jensen et al . [12] , which identified 234 transcripts are being less abundant in the procyclic than the bloodstream form , 172 showed the same direction of change while 62 did not agree with our study . In addition , from the 317 transcripts that are increased in the procyclic form , 270 are in agreement with our study . Of the 551 transcripts that were significantly changed ≥2-fold between the two life cycle stages in the study by Jensen et al . ∼80% showed the same pattern in our study . Using the two data sets , we have compiled a list of 442 genes that show a robust change in expression pattern ( Table S2 ) . When we analyzed the metabolic pathways as annotated in KEGG we found the majority to be down regulated in the stumpy form when compared to the long slender form ( Figure S4 , S5 , S6 , Table S3 ) . In the glycolytic pathway for example , 8 of 11 transcripts are decreased in abundance in the short stumpy form , which is in good agreement with previous studies of the metabolic pathways in quiescent cells in general [29] and trypanosome short stumpy form in particular ( for review see [30] ) . We furthermore compared the expression profile of ten genes between bloodstream long slender and procyclic cells using SLT and RT qPCR and found positive correlation between the two methods ( Pearson correlation r = 0 . 97; p-value of p<10−5; Figure S7 , Table S4 ) . Additionally , we performed two RNAi experiments ( Alba1 and Alba 3/4 ) in procyclic forms ( Figure S8 ) . The efficiency of the knockdown was then measured by Northern blot analysis and SLT . For Alba1 both techniques showed a knockdown in message level to 6% compared to the uninduced cell line . For Alba 3/4 Northern blot analysis indicated a knockdown to 8% while the value measured by SLT was 13% . The overall correlation between the uninduced and the induced libraries was between 0 . 97 ( Alba1 ) and 0 . 94 ( Alba 3/4; Spearman rank correlation coefficient ) . Lastly we compared the expression profile from SLT with one run of regular RNAseq from poly ( A ) -RNA from procyclic forms ( 24 million sequence tags ) and found a positive correlation between the two techniques ( Spearman ρ = 0 . 69 , Figure S9 ) . In addition to the different life cycle stages of Antat 1 . 1 , we also analyzed the widely used monomorphic bloodstream form MITat 1 . 2 ( 221 ) and mapped the sequence tags to the active VSG expression site ( Figure 3; [31] ) . The most highly expressed gene was VSG 221 with more than 69 , 000 TPM . The second most highly expressed gene in the expression site was the ESAG 12 gene ( 761 tags ) followed by a hypothetical gene that has not been annotated previously , now designated ESAG13 ( 138 tags ) and ESAG 6 and 7 , which encode the two subunits of the transferrin receptor ( 80 and 43 tags; for review see [32] ) . We identified 29 , 406 splice sites in the T . brucei genome with a median of two splice sites per gene ( bloodstream form AnTat1 . 1; Table 1 ) . The major splice site was strictly conserved with 94% of the splice acceptor dinucleotides being AG preceded by an upstream polypyrimidine tract ( −14 to −43 , relative to the splice site; Figure 4A ) . Twenty percent of the minor splice acceptor sites contained a dinucleotide other than AG . GG occurred in 7% of these while TG , AA , GA and AC were found in 2% of the minor splice sites ( Figure 4B–C ) . The least abundant dinucleotide was CC . When we compared the major splice sites from this study to a previous genome-wide prediction we found that about 40% of the major sites and 6% of the minor sites had correctly been predicted by the model [15] . Using the position of the major splice sites we determined the mean length of all 5′ UTRs to be 104–138 nucleotides and the median length to be 32–47 nucleotides ( Figure 4D–E , Table 1 ) . In the procyclic life cycle stage we identified 588 transcripts with splice sites that would allow N-terminal extension of the annotated protein coding gene and , for 11 of these , peptides corresponding to the N-terminal region were recently detected by mass spectrometry ( Table S5; [33] ) . About 500 of the currently annotated genes only had internal splice sites 3′ of the predicted AUG start codon , and depending on the life cycle stage , 683–870 genes had major internal splice sites ( Table 1 ) . Of the transcripts with major internal sites ∼90% had a downstream AUG in the same reading frame within 400 nucleotides of the splice site , leaving about 10% of the transcripts without a likely translation start . In each life cycle stage >1200 transcripts exhibited an alternative splicing pattern ( Figure 5–6 Table 3 ) . A transcript was termed alternatively spliced if the major splice site contained less than 60% of all sequence tags . We identified four different types of alternative splicing , based on the potential functional consequences ( Figure 7 ) . Splicing type A renders the transcript potentially untranslatable since the transcript does not contain an in-frame AUG downstream of the alternative splice site ( Figure 7 , Figure S10 ) . Splicing type B leads to a 5′ truncation of the original open reading frame , with a potential downstream AUG start site ( Figure S11 ) . Splice Type C does not change the open reading frame , but it includes or excludes potential regulatory elements such as upstream open reading frames ( uORFs ) in the 5′ UTR ( Figure S12 ) . Splice type D potentially allows for the use of a novel open reading frame due to the inclusion of an AUG start codon in a different reading frame ( Figure S13 ) . We then analyzed if the major splice sites for a particular transcript changed between the life cycle stages and termed this differential splicing ( Figure 6 ) . Depending on the two life stages we compared , 158–458 differential splicing events affecting a total of 676 genes were identified ( Table 2 ) . We additionally verified the differential splicing patterns for three transcripts ( Tb927 . 1 . 790 , Tb927 . 6 . 4240 , Tb11 . 02 . 2700 ) by RT qPCR and found positive correlation between the SLT and the RT qPCR data ( Pearson r = 0 . 85; 95% confidence interval; Table S6 ) . Of the 23 currently annotated distinct AARS the Asp-RS , Lys-RS , Trp-RS and Tyr-RS are each encoded by two genes , one of which contains a mitochondrial targeting signal ( MTS ) at the N-terminus as determined by MITOPROT ( Table S7; [34] . Recently , Charrière and coworkers verified the cytosolic and mitochondrial localization of the Trp-RS and Asp-RS experimentally [35] , [36] . Five additional AARS ( Asn-RS , Pro-RS , Glu-RS , Gln-RS , Ser-RS ) contain an N-terminal MTS; for these AARS , however , there is no cytosolic isoform encoded in the genome . When we analyzed their expression and splicing patterns we found that the major 5′ splice site is internal to the currently annotated AUG start codon leaving only a small fraction of the splice sites upstream of the first AUG ( Figure 8 , Table S7 ) . Thus the major translation product would exclude the MTS . Additionally , 4 AARS ( Arg-RS , Cys-RS , His-RS and Leu-RS ) contain an internal MTS that is masked by 48–101 amino acids at the N-terminus ( Table S7 ) . However these MTS could become N-terminal if a downstream start codon would be used as the translation start site .
We have developed a cost effective method to analyze whole genome expression and splicing profiles from organisms employing leader trans-splicing . This was tested on T . brucei , where we obtained 85% coverage of all genes with fewer than 1 million sequence tags . A major improvement over traditional microarray technology is the possibility of sequencing directly from the spliced leader , which allows selective analysis of the transcriptome of intracellular parasites like the amastigote forms of Trypanosoma cruzi or forms that are closely associated with their host and extremely difficult to purify like the epimastigote form of T . brucei in tsetse salivary glands . Using the SLT approach we sequenced more than 1 million splice site tags from poly ( A ) -mRNA from each of three life cycle stages of T . brucei . Each sequence tag covered at least 24 nucleotides 3′ of the spliced leader/5′ UTR junction . Even though the analyzed strains ( Antat 1 . 1 and Mitat 1 . 2 ) were not identical to the genome strain ( TREU 927 ) mapping of the 24mer sequence tags onto the TREU 927 genome was very successful , given that the majority ( 98% ) could be aligned with high statistical significance using a combination of mismatch and sequence quality scores ( see materials and methods ) . In the process , this method revealed the existence of a new expression site associated gene in an active VSG expression site . In general , our data are in good agreement with the current genome annotation , and strongly support both the number and positions of putative transcription start sites identified by binding of specific histones [8] . In addition , 98% of the sequence tags from the procyclic form map in the sense orientation of annotated transcription units , and only 2% in the zones between transcription units , where the direction of transcription is not immediately obvious from the annotations . The splicing patterns indicated that transcription overlaps in several of the converging strand switch regions as has been suggested previously for converging transcription units transcribed by Pol I and Pol II , respectively [22] . Interestingly , bloodstream form cells show twice the number of sequence tags mapping to the zones between the transcription units compared to the procyclic form , indicating that the control of transcription initiation may be less specific , or degradation less efficient in this life stage . It is also possible that splice site recognition is different , or not equally efficient between the stages , leading to the increased number of processed RNAs from the regions between the transcription units . The large variation in transcript abundance that we detected within transcription units was expected and is in good agreement with previous studies , strengthening the notion that steady state RNA levels in trypanosomes are regulated mainly post-transcriptionally at the level of RNA processing and/or RNA stability . When we analyzed the monocistronic transcription units we found >70% unlikely to be expressed in the life cycle stages analyzed because we could not detect any splice sites within 2 kb upstream of the start codon . Of the remaining 76 transcripts , only 8 showed differential expression . Previous microarray studies of T . brucei using genomic arrays or single probe arrays have indicated a rather static transcriptome with relatively few changes between the bloodstream and procyclic forms . Estimates ranged from 2–6% of the genome being regulated at the transcript abundance level [9] , [10] , [11] , [12] , [13] . A more recent study by Jensen et al . using eight oligonucleotides per gene on a Nimblegen microarray found that up to 700 transcripts or ( 8% ) change expression between the two life cycle stages [12] . Considering these studies it would appear that a multiprobe array is more likely to detect a larger set of regulated genes . Furthermore , using arrays with only one probe in the 5′ region of the genes is likely to be affected by the misannotation of the start sites of open reading frames and/or by alternative splicing . Using SLT we found 30% of the transcripts from protein-coding genes to be significantly regulated between the long slender bloodstream form and the procyclic form ( Table 2 ) . The number of changes increased to 40% of all genes when extended to include the short stumpy form of the parasite . Even at a more conservative threshold ( ≥2-fold , significantly changed ) , 35% of all genes exhibited changes in transcript abundance . In conclusion , we have found that a much larger cohort of genes changes expression levels during development than was previously thought to be the case . This is in line with recent findings that about 50% of the genome of T . cruzi is regulated at the level of transcript abundance [14] . When we compared our results with the previous microarray studies we find different levels of correlation in gene expression depending on the study . While some part of the differences might be due to the higher sensitivity of the SLT approach we have to keep in mind that the low level of correlation might also be due to strains and culture conditions that vary considerably between the different studies . We also employed three different approaches to verify the results obtained by SLT . ( i ) Results from RT qPCR showed strong positive correlation with the SLT approach for the expression level changes between life stages and the differential splicing events ( Figure S7 , Table S4 , Table S6 ) . ( ii ) We also performed two RNAi experiments indicating that the abundance of the RNAi target transcript as quantified by SLT is in excellent agreement with quantification by Northern blots . The overall correlation between the induced and uninduced transcriptomes was on a par with the technical reproducibility of RNAseq , further supporting that SLT tag counts serve as a sufficient proxy for comparisons between different life cycle stages or cell lines of the same organism ( Figure S8 ) . ( iii ) We included the correlation to one run of regular RNAseq of poly ( A ) -mRNA ( 24 million sequence tags , Figure S9 ) . While not perfect , the correlation between SLT and RNAseq ( Spearman ρ = 0 . 69 ) is nearly on a par with that between a technical comparison of RNAseq and microarrays ( Spearman ρ = 0 . 75 e . g . [37] ) . During the revision of this manuscript Siegel et al . published a study describing the expression profile in bloodstream and procyclic forms of T . brucei using RNAseq [38] . Although the approaches used in both studies are different ( RNAseq vs . SLT ) many features found in both studies are well correlated; the mean number of splice sites per gene ( 2 . 6 versus 2 . 7–2 . 9 ) , the mean lengths of the 5′ UTRs ( 184 versus 105–135 ) , the number of genes with internal splice sites ( 488 versus 496–558 ) and the large dynamic range ( 105 to 106 ) . However , there are also a number of features that do not correlate so well , most striking of which is the difference in expression profile . We identified almost 40% of the genes as being regulated significantly while Siegel et al . found only about 10% . There are several reasons that could explain the differences observed: ( i ) strain differences and the number of life stages analyzed , ( ii ) growth conditions for bloodstream forms ( in vitro versus in vivo ) , ( iii ) cDNA preparation , with RNAseq being more likely to capture precursors and breakdown intermediates ( e . g . intron sequences ) , and ( iv ) scaling of the data ( tags/million versus constant median count/gene ) . The major advantage of SLT when compared to conventional RNAseq is the cost effective mapping of 5′ splice sites in splice leader bearing organisms , especially intracellular parasites where the purification of RNA is very difficult . According to our study , VSG transcripts account for 7–11% of spliced mRNAs in bloodstream forms trypanosomes , which is in excellent agreement with the value obtained from hybridization data [39] . The levels of most other transcripts , however , are at least two to three orders of magnitude lower . Assuming there to be approximately 40 , 000 mRNA molecules in a procyclic trypanosome , and approximately half that number in a bloodstream form ( Haanstra and co-workers and our calculations ) , a large number ( >5000 , 65% ) would be present at <1 mRNA per cell [40] . Similar results have been reported for yeast where more than 80% of all transcripts are present at ≤2 copies per cell [41] , [42] . The question if the small number of transcripts is distributed evenly across the population of trypanosomes , or accumulates in very few cells during a transcriptional burst , remains to be investigated . Recently it has been suggested that transcription in yeast is much more steady , with fewer transcriptional bursts than seem to occur in mammalian cells [43] , [44] , [45] . The data presented here describes the splice sites for 85% of the annotated genes in the T . brucei genome ( Table 1 ) . Spliced leader addition sites were very well conserved within and between life cycle stages . AG was the predominant splice acceptor dinucleotide of the major splice sites , however 20% of the minor splice sites used other dinucleotides , predominantly GG ( Figure 4A–C ) . The least abundant acceptor dinucleotide was CC followed by any pyrimidine combination . The major splice sites contained a 15 ( ±6 ) nucleotide polypyrimidine stretch at an average distance of 31 nucleotides ( ±19 ) upstream of the actual splice acceptor site . Both these findings are in very good agreement with previously published experimental data on individual splice sites [46] . The UTR length distribution indicated 5′ UTRs with a median length between 32 to 47 nucleotides , which is similar to yeast ( 50 nt; [47]; Figure 4 D–E ) . Interestingly we detected a shift towards longer UTRs in the long slender bloodstream form when compared to the procyclic form . This is at least in part due to the differential use of splice sites in the two life forms and might indicate differential regulation of translation . About half of the major splice sites in the different life stages could not be predicted using our current splice site recognition model , although the majority contained the signals that conform to our current understanding of splice sites . An obvious consequence of alternative splicing would be the change of N-terminal targeting sequences as has been shown for T . cruzi ( LYT1 ) and for alternative cis-splicing in other systems [48] , [49] , [50] , [51] . Our analysis indicated this likely to be the case for several AARS that are essential in the cytosol and mitochondrion [35] , [36] thus providing evidence that alternative splicing is a potential mechanism for dual localization of proteins similar to what has been reported for the LYT1 gene in T . cruzi [48] . More than 500 transcripts with splice sites exclusively 3′ of the annotated start site were identified by SLT . While we cannot exclude that a very small fraction is still spliced upstream of the annotated AUG , we consider it much more likely that the bona fide start codon is within the open reading frame . A second set of transcripts ( >600 ) indicated the possibility of 5′ extensions to the currently annotated open reading frames , which would effect changes in the N-terminus of the corresponding protein . The evaluation of these N-terminal extensions is much more difficult and will require additional experiments . However , by screening the recently published proteomics dataset from Panigrahi and co-workers , we were able to identify 11 candidates where peptides corresponding to a region upstream of the annotated start codon are expressed in the procyclic form trypanosomes [33] . Depending in the life cycle stage , we also identified 90–97 transcripts in which alternative splicing ablated the start codon , suggesting this form of splicing plays a minor but significant role in regulating gene expression . Most surprising was that a large number of transcripts showed differential abundance of alternative splice variants in the three life stages analyzed . A previous report indicated that T . brucei used different splice sites on an artificial construct , but it remained unclear if , and how frequently , this might occur in the T . brucei transcriptome [52] . We found more than 600 differentially spliced transcripts between the life stages , supporting the idea that alternative splicing has functional consequences for the regulation of parasite development . One open question is if the actual splicing event is regulated or the differential abundance is a result of altered stability of the RNA transcripts . So far , we have been unable to detect sequence elements in the vicinity of the alternative splice sites that would explain the differential regulation of splicing itself . It is worth noting , however , that transcripts encoding several of the core components of the spliceosome , such as SMD1 , SMD3 and SMG , are themselves differentially regulated during development; this may reflect an adaptation of the splicing machinery to differences in the major splicing targets , such as the VSG and procyclin transcripts , or to the subtleties of alternative splicing . These hypotheses should now be testable on a genome-wide scale , using the SLT approach in combination with RNA knockdown of specific splicing components .
T . brucei brucei AnTat 1 . 1 and MITat 1 . 2 ( 221 ) were used in this study . Late procyclic forms of AnTat 1 . 1 were cultivated in SDM79 supplemented with 10% FBS . Bloodstream forms were grown in mice . For bloodstream forms of AnTat 1 . 1 , mice were immunosuppressed with 260mg/kg cyclophosphoamide ( Sigma ) 24 hours prior to intraperitoneal injection of 106 parasites . Short stumpy parasites were harvested 5 days post-infection at a density of 2–4×108/ml blood . 75–80% of the cells showed a short stumpy phenotype as determined by light microscopy of blood smears after methanol fixation . Long slender forms were harvested from untreated mice at day three post-infection at a density of <5×107 cells/ml blood . MITat 1 . 2 bloodstream forms were grown in mice as described above for long slender bloodstream forms of AnTat 1 . 1 . Parasites were purified from whole blood using DE-52 anion exchange resin equilibrated to pH 8 with a bicine glucose buffer . After purification cells were centrifuged and resuspended in TriPure RNA isolation reagent ( Roche , Switzerland ) . RNA was extracted using TriPure ( Roche , Switzerland ) according to the manufacturer . Poly ( A ) mRNA was purified from 7 . 5 µg total RNA using Dynabeads oligo- ( dT ) beads according to the manufacturer ( Invitrogen , USA ) . First strand cDNA was synthesized from poly ( A ) RNA using random hexamers and Superscript II reverse transcriptase ( Invitrogen , USA ) in a final volume of 20 µl for 1 hour at 42°C . Half of the first strand mix was used for second strand synthesis ( 10 µl 1st strand mix , 1 µl RNaseH , 15 minutes at 37°C ) . Second strand synthesis was done using 2 µl 10× Thermopol buffer ( New England BioLabs ) , 1 µl dNTPs 10 mM , 1 µl 2nd strand primer 10 µM ( [BIOT]5′- AATGATACGGCGACCACCGAGATCTACAGTTTCTGTACTATATTG -3′ ) , 2 units Taq polymerase ( New England BioLabs , USA ) and 4 . 5 µl H2O by incubation for 5 minutes at 50°C and then 5 minutes at 72°C . The cDNA was purified on a Qiagen MinElute column ( Qiagen , USA ) and eluted in 10ul TE buffer . Adapter ligation to the purified dsDNA was done using ( 10 . 0 µl DNA , 2 . 5 µl 10× Ligase Buffer ( New England BioLabs ) , 10 . 0 µl H2O , 1 . 0 µl Fasteris customized bar-coded paired-end Illumina adapter , 600 units T4 DNA Ligase ( New England BioLabs , USA ) for 1 hour at room temperature . The ligation mix was purified from unligated linker with streptavidin beads ( Dynabeads ) . After incubation at room temperature for 15 minutes beads were separated on a magnetic stand , and washed twice and then resuspended in 20 µl 10 mM Tris buffer . 5 µl was used for PCR amplification with primers ( 5′- AATGATACGGCGACCACCGA -3′/5′- CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAAC -3′ ) following the standard Illumina mRNA-SEQ library amplification protocol . Fragments in the range 120–160 bp were separated on a 2% agarose-TBE gel and subsequently purified on Qiagen Gel Extraction MinElute columns ( Qiagen , USA ) . Quality control for insert size was performed by TOPO cloning and subsequent ABI sequencing . Sequencing on the Illumina Genome Analyzer was carried out using the following sequencing primer ( 5′- ACCGAGATCTACAGTTTCTGTACTATATTG -3′ ) . In total we sequenced three libraries from bloodstream form mRNA , ( long slender and short stumpy , both Antat1 . 1 and monomorphic Lister 427 ) , one from procyclic form mRNA ( Antat1 . 1 ) , 2×2 RNAi libraries ( uninduced and induced ) , where each pair could be considered a biological replicate . Lastly we prepared and sequenced one conventional RNAseq library from procyclic mRNA ( Antat1 . 1 ) . Base calling was performed using the Genome Analyzer Pipeline ( Illumina ) . Linker sequences were removed while separating reads according to identified barcodes . Only sequence reads of inserts with a length of at least 24 nucleotides were retained . A pipeline was set up using languages with open source interpreters ( bash , perl and R ) to automate the following analysis . The reads were mapped to the genome sequence of T . brucei TREU927 using maq ( [53] http://maq . sourceforge . net ) with n = 3 and an effective first read length of 24 . Single mapping reads were separated from multi mapping reads by an alignment quality threshold of 30 . Tags ending in the same position were joined together , and their counts were added ( using bioperl , [54] ) . Tag counts were normalized to the library size ( number of reads of length 24 or more ) and scaled linearly to reflect counts of tags per million ( TPM ) . Mapped tags were assigned to the annotated protein coding gene 3′ of the tag . Tags mapping internally to a coding sequence ( CDS ) were assigned to it as internal splice sites . Data was exported in tabular and GFF format ( http://www . sanger . ac . uk/Software/formats/GFF/ ) and then visualized using Gbrowse [55] . Alternative spliced leader addition sites were cataloged for each gene . Genes with ≤60% tags in the assigned major splice site were designated alternatively spliced . The 5′ UTR lengths were calculated and visualized using R [56] . Upstream open reading frames ( uORFs ) were detected on the mapped 5′ UTR , counted and assigned a note describing their length , whether they were terminated on the 5′ UTR , or overlapped the bona fide ( CDS ) start codon . Sequences surrounding the splice sites were extracted using bioperl and visualized as sequence logos using WebLogo [57] , [58] . Differences in expression levels of a gene in two stages was tested for significance according to Audic and Claverie with a threshold of P<10−5 . Scatterplots of the differential gene expression levels of all libraries were produced using R [28] . When a gene had different major splice sites in two stages , the normalized counts of these sites were tested for a statistically significant difference ( Fisher two-tailed test , P<10−5 ) . Significant differences were termed differential trans-splicing events . Expression levels were pooled over T . brucei specific KEGG pathways and visualized as heatmaps after log2 transform and hierarchical average linkage clustering of euclidian distances using MeV ( [59]; http://www . genome . jp/kegg; http://www . tm4 . org/mev . html ) . Anti-sense splice sites were detected using bioperl . In order to simulate gene coverage a subset of reads was drawn randomly without replacement from one library and the mapping and analysis pipeline executed for each subset . This was repeated five times for each subset size . Saturation curves were drawn for several parameters , with error bars given as confidence intervals assuming normal distribution . Comparison to splice model predictions were made according to Benz et al . with the mapped splice sites substituted for the previous small mapping EST set [15] . All bioinformatics tools , programs , pipelines used in this study will be provided upon request . All sequence data including the regular RNAseq data is available through our website and TriTrypDB . PCR was used to confirm the splice sites detected by SLT . 3 µg total RNA was used as a template for reverse transcription in 50 µl AMV reverse transcriptase buffer ( Promega , USA ) in the presence of 1mM dNTPs , 360 ng random hexamers , 80 U RNasIn ( Promega , USA ) and 60 U AMV reverse transcriptase ( Promega , USA ) . Subsequently 1µl of this reaction was used for PCR . For PCR the splice leader primer CGCTATTATTAGAACAGTTTCTGTAC-3′ ( Tm 55°C ) , and reverse primers 5′- GTTGCATCCGGTGTTCTTTT -3′ ( Tb11 . 02 . 2700 ) and 5′-AGCAGTCATCAATTCTTCCT-3′ ( Tb927 . 6 . 4240 ) were used . The reaction was performed in 25µl , ( 2mM MgCl2 , 2 . 5µl PCR 10× buffer , 400nM primer , 1 unit of Taq DNA polymerase ( Qiagen , USA ) , 0 . 2mM dNTP and 30ng/ul cDNA ) for 30 cycles . RT qPCR was done essentially as described previously . cDNA for RT-qPCR was prepared as described above . The primers were designed such they would amplify regions of 80 to 150 nucleotides using the online software tool for real time PCR from Genescript ( Table S8 ) . Real-time PCR was run on the GeneAmp 7000 ( Applied Biosystems ) using 30ng of cDNA , 400nM oligonucleotides in 25µl of the MESA GREEN qPCR Master mix for SYBR assay ( Eurogentec ) . Values were normalized to beta tubulin and the amplification efficiency was derived from a cDNA dilution series covering five logs . Average values and standard deviations of 3 RT qPCRs from one cDNA sample are shown . The sequence data can be accesed through GEO ( http://www . ncbi . nlm . nih . gov/geo/ ) with the accession number GSE22571 or through our website http://splicer . unibe . ch/ or in future also from TriTrypDB ( http://tritrypdb . org/tritrypdb/ ) . All animal work has been conducted in accordance with the regulations for the production of bloodstream form trypanosomes of the cantonal animal protection agency in Bern , Switzerland . Approved by the cantonal animal protection agency in Bern Switzerland ( form B , 14/07 ) .
|
Some organisms like the human and animal parasite Trypanosoma brucei add a leader sequence to their mRNAs through a reaction called trans-splicing . Until now the splice sites for most mRNAs were unknown in T . brucei . Using high throughput sequencing we have developed a method to identify the splice sites and at the same time measure the abundance of the corresponding mRNAs . Analyzing three different life cycle stages of the parasite we identified the vast majority of splice sites in the organism and , to our great surprise , uncovered more than 2500 alternative splicing events , many of which appeared to be specific for one of the life cycle stages . Alternative splicing is a result of the addition of the leader sequence to different positions on the mRNA , leading to mixed mRNA populations that can encode for proteins with varying properties . One of the most obvious changes caused by alternative splicing is the gain or loss of targeting signals , leading to differential localization of the corresponding proteins . Based on our findings we hypothesize that alternative splicing is a major mechanism to regulate gene expression in T . brucei and could contribute to protein diversity in the parasite .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"computational",
"biology/alternative",
"splicing",
"molecular",
"biology/mrna",
"stability",
"molecular",
"biology/rna",
"splicing",
"molecular",
"biology/bioinformatics",
"microbiology/parasitology",
"infectious",
"diseases/protozoal",
"infections"
] |
2010
|
Spliced Leader Trapping Reveals Widespread Alternative Splicing Patterns in the Highly Dynamic Transcriptome of Trypanosoma brucei
|
Extremely AT-rich DNA sequences present a challenging template for specific recognition by RNA polymerase . In bacteria , this is because the promoter −10 hexamer , the major DNA element recognised by RNA polymerase , is itself AT-rich . We show that Histone-like Nucleoid Structuring ( H-NS ) protein can facilitate correct recognition of a promoter by RNA polymerase in AT-rich gene regulatory regions . Thus , at the Escherichia coli ehxCABD operon , RNA polymerase is unable to distinguish between the promoter −10 element and similar overlapping sequences . This problem is resolved in native nucleoprotein because the overlapping sequences are masked by H-NS . Our work provides mechanistic insight into nucleoprotein structure and its effect on protein-DNA interactions in prokaryotic cells .
Transcription is initiated by binding of RNA polymerase to specific DNA sequences known as promoters [1] . Following promoter recognition the resulting complex undergoes a process of isomerisation . Hence , ∼14 base pairs ( bp ) of DNA , close to the transcription start site , are unwound [2] . RNA polymerase then engages in abortive cycles of initiation before escaping the promoter to form an elongation complex [3] . It has long been known that promoter unwinding is facilitated by the weak base stacking interactions associated with AT-rich DNA . Thus , the eukaryotic TATA box ( 5′-TATAAA-3′ ) is unwound during transcription initiation [4] . Similarly , the prokaryotic −10 hexamer ( 5′-TATAAT-3′ ) , recognised by Domain 2 of the RNA polymerase σ70 subunit , participates in DNA opening [5] . Because DNA elements recognised by RNA polymerase are AT-rich , chromosomal regions , where DNA AT-content is unusually high , prove particularly challenging templates for recognition . For example , the horizontally acquired sections of some bacterial chromosomes have an elevated AT-content . As a result , RNA polymerase may bind cryptic promoters [6] or initiate transcription promiscuously [7] . In Escherichia coli , gene regulatory regions are targeted by chromosome folding proteins [8] . Hence , in addition to their architectural role , these proteins can influence RNA polymerase-DNA interactions [9] . The Histone-like Nucleoid Structuring ( H-NS ) protein recognises AT-rich DNA and is associated with horizontally acquired genes [10]–[13] . The prevailing view is that , when bound at such regions , H-NS silences transcription [14] . However , the precise mechanism remains elusive; models proposing exclusion of RNA polymerase from , and trapping of RNA polymerase at , H-NS bound regions have both been proposed [15] . Since these models are not mutually exclusive a third possibility is that a myriad of different configurations exist . Interestingly , two recent studies have reported close association between RNA polymerase and H-NS [16] , [17] . In one case , H-NS stimulated rather than repressed gene expression [17] . In this work we describe an undocumented role for H-NS; facilitating the correct recognition of promoters by RNA polymerase . The ehxCABD operon from Shiga toxin-producing E . coli ( STEC ) has an unusually high AT-content . Consequently , the operon regulatory region contains multiple sequences that resemble −10 hexamers . We show that , despite the apparent ambiguity of this DNA template , RNA polymerase initiates transcription specifically from a single promoter in vivo . However , in vitro , RNA polymerase is unable to differentiate between this promoter and adjacent binding sites . We show that H-NS plays a critical role by blocking access of RNA polymerase to the adjacent binding sites . Thus , H-NS ensures correct positioning of RNA polymerase .
The ehxCABD operon is located on the pO157 plasmid and its derivatives . The operon encodes an enterohemolysin and proteins for its post-translational modification and export [18] . The 248 bp of regulatory DNA immediately upstream of the operon has an AT-content of 71% . H-NS has been implicated in regulating expression of the operon but a comprehensive molecular analysis is lacking [19]–[21] . As a first step we determined which section of the regulatory DNA contained promoter activity . Note that the ehxCABD regulatory DNA has an almost identical sequence in multiple E . coli serotypes and we arbitrarily used the ehxCABD regulatory sequence described by Rogers et al . [20] . We began by generating DNA fragments carrying discrete sections of the ehxCABD regulatory region ( illustrated in Figure 1Ai ) . The fragments encompass 248 bp of DNA adjacent to the first gene in the operon ( fragment F1 ) , the downstream part of this region ( fragment F2 ) or the upstream section of the locus ( fragment F3 ) . We assayed each fragment for promoter activity using two plasmid based systems ( illustrated in Figure 1Aii ) . Hence , pRW50 and pLux encode the reporter proteins β-galactosidase and Luciferase respectively . Note that pRW50 was used to report promoter activity in E . coli K-12 whilst pLux was used with E . coli O157:H7 as a control for effects of STEC encoded transcriptional regulators . The raw activity data , for each DNA fragment , in each plasmid , is summarised in Figure 1B . Our results show that the F1 and F3 fragments stimulate transcription , to similar levels , in all of the assays . No detectable transcription was driven by the F2 fragment . Therefore , the ehxCABD promoter must be located in the upstream portion of the regulatory region common in both F1 and F3 . Our next aim was to identify transcription start sites in the F3 fragment . To do this we conducted mRNA primer extension experiments . We used RNA extracted from E . coli JCB387 cells , carrying the F3 fragment cloned in plasmid pRW50 . Our analysis yielded two extension products of 155 and 154 nucleotides ( nt ) in length ( Figure 1C ) . The transcript start , corresponding to the more abundant 154 nt extension product , is labelled +1 in Figure 1D . A consensus extended promoter −10 element ( 5′-TGnTATAAT-3′ ) was found 8 bp upstream of the transcription start site . A four out of six match to a promoter −35 element ( 5′-TTGACA-3′ ) was observed further upstream . Throughout this work we refer to this promoter , highlighted green in Figure 1D , as PehxCABD . The two primer extension products , differing in length by a single nt , both likely originate from this promoter . Importantly , we confirmed that PehxCABD was the only promoter present in the F1 fragment . Thus , using RNA extracted from E . coli JCB387 cells carrying the F1 fragment cloned in plasmid pRW50 , we observed only primer extension products corresponding to PehxCABD ( Figure S1 ) . Our primer extension analysis shows that , in vivo , RNA polymerase initiates ehxCABD transcription with precision ( Figure 1 ) . This is remarkable given the abundance of potential −10 hexamer sequences in this regulatory region ( two such sequences are highlighted red in Figure 1D ) . To better understand how specificity is achieved we examined recognition of the naked F3 fragment by RNA polymerase . We utilised two in vitro DNA footprinting techniques . First , we exploited the properties of Fe2+ chelated Bromoacetamidobenzyl-EDTA ( FeBABE ) . FeBABE is a DNA cleavage reagent that can be attached to specific cysteine side chains in proteins . Once attached , FeBABE cleaves nucleic acids within a 12 Å radius of the attachment site . Thus , FeBABE conjugated with the RC461 derivative of E . coli σ70 cleaves promoter −10 elements [22] . Figure 2Ai shows the pattern of FeBABE cleavage observed with the F3 fragment . As expected , the PehxCABD −10 element was cleaved ( highlighted by green box in Figure 2Ai ) . However , we also observed DNA cleavage at additional sites overlapping PexhCABD ( highlighted by red stars in Figure 2Ai ) . In complementary experiments KMnO4 footprinting was used to detect DNA unwinding by RNA polymerase . We observed DNA melting at the PehxCABD −10 element ( highlighted by a green box in Figure 2Aii ) and at additional sites ( highlighted by yellow stars in Figure 2Aii ) . It did not escape our attention that the additional sites of FeBABE and KMnO4 reactivity align with each other and with sequences that resemble −10 hexamers highlighted in Figure 1D . Nevertheless , we were concerned that the additional FeBABE and KMnO4 reactivity signals might originate from RNA polymerase bound at PehxCABD . To exclude this possibility we ran identical reactions with unrelated cbpA P6 promoter DNA . In these experiments no DNA cleavage products were observed other than those at the cbpA P6 −10 hexamer . We conclude that the naked PehxCABD F3 fragment must contain multiple overlapping RNA polymerase binding sites . Factors present in vivo must influence RNA polymerase interactions with PehxCABD . Such factors may explain why the additional RNA polymerase binding sites observed in vitro do not generate transcripts in vivo . Our attention turned to H-NS , which is known to recognise AT-rich regulatory regions and influences ehxCABD expression [19]–[21] . Thus , we used chromatin immunoprecipitation ( ChIP ) to measure binding of RNA polymerase and H-NS to PehxCABD in vivo . Recall that , in ChIP experiments , a cell's nucleoprotein is cross-linked with formaldehyde , extracted , and then fragmented by sonication . Antibodies directed against the protein of interest are then used to select DNA fragments with which the protein is cross-linked . Finally , PCR is used to identify recovered DNA fragments . Figure 3A shows PCR analysis of DNA immunoprecipitated with anti-RNA polymerase ( β subunit ) or anti-H-NS . Control experiments , in which we analysed total cellular DNA , or DNA recovered from a mock immunoprecipitation , are also shown . The PehxCABD DNA is detected in the total DNA sample , the anti-β , and anti-H-NS immunoprecipitates . Importantly , the PehxCABD DNA was not detected in the mock immunoprecipitate . In a set of control PCR reactions we probed the lacZ and yabN loci . Note that these loci are not transcribed in the conditions used here and are not bound by H-NS . As expected , lacZ and yabN were not detected in the immunoprecipitates . We next reconstituted co-association of RNA polymerase , H-NS and PexhCABD in vitro . Electrophoretic Mobility Shift Assays ( EMSA ) were used to probe the complexes formed . The result is shown in Figure 3B . The data show that RNA polymerase ( lane 2 ) and H-NS ( at two different concentrations , lanes 3 and 5 ) form distinguishable complexes with the DNA . When H-NS and RNA polymerase are added in unison an additional complex can be detected ( boxed in lanes 4 and 6 ) . To confirm that this additional complex contained both H-NS and RNA polymerase the band was extracted , submitted to tryptic digest , and the resulting peptides analysed by mass spectrometry . Both RNA polymerase and H-NS were present in the excised band . To more precisely understand the ternary H-NS-RNA polymerase-DNA complex we repeated our σ70RC461-FeBABE analysis . The data show that , in the presence of H-NS , the signal for RNA polymerase binding at the PehxCABD −10 element is retained . Conversely , binding of RNA polymerase at adjacent sites is lost ( Figure 4A ) . In a complementary experiment we used DNAse I footprinting to locate H-NS binding in the absence of RNA polymerase . The data show that H-NS recognises the same AT-rich region , extending from +10 to −30 , as the transcriptional apparatus ( Figure 4B ) . Thus , the binding sites for H-NS and RNA polymerase overlap . To assess how H-NS effects RNA polymerase interactions with PexhCABD in vivo we repeated our primer extension analysis . We used RNA extracted from wild type E . coli K-12 and cells lacking hns . As described above , RNA from wild type cells yielded two extension products of 155 and 154 nt in length ( Figure 4C lane 5 ) . These extension products were also observed when we analysed RNA from Δhns cells ( Figure 4C lane 6 ) . Strikingly , RNA from Δhns cells yielded a further 9 extension products of between 138 and 194 nt in length . These additional primer extension products align with the additional sites of RNA polymerase binding observed in Figure 4A . Finally , it is noteworthy that , in order to observe the primer extension products in Lane 6 of Figure 4C , we had to “overload” the sample onto the gel . This suggests that the net result of reduced RNA polymerase binding specificity is a reduction in transcription . Consistent with this , we observed reduced expression from the F3 fragment , in cells lacking H-NS , using our LacZ reporter assay ( Figure S2 ) . Our data suggest that PehxCABD is flanked by at least two overlapping elements that can bind RNA polymerase . If this model is correct there should be competition between RNA polymerase molecules for binding the various targets . A logical consequence of this competition would be reduced transcription from PehxCABD . To test this model we disrupted either the PehxCABD −10 hexamer or the overlapping RNA polymerase binding elements . The mutations utilised are illustrated in Figure 1D . Figure 5A shows LacZ activity data from wild type E . coli cells carrying the various promoter::lacZ fusions . The -41G mutation increases LacZ expression that is further increased by the -7T-5T-4T mutations . Conversely , the -13G mutation , in the canonical PehxCABD −10 element , reduces LacZ expression . We next sought to confirm the stimulatory effect of H-NS on specific recognition of PehxCABD by RNA polymerase . Thus , we compared the effects of H-NS and the -41G mutation using in vitro transcription assays . The F3 DNA fragment was cloned upstream of the λoop terminator in plasmid pSR . In the context of this construct PehxCABD produces transcripts , of 178/179 nt in length , that can be quantified after electrophoresis . Additional transcripts , corresponding to the Δhns primer extension products in Figure 4C , should also be generated . On this basis , we expected to detect an abundant 162 nt transcript ( corresponding to the 138 nt extension product in Figure 4C ) and scarce transcripts sized between 183 nt and 218 nt ( equivalent to the primer extension products in the 159–194 nt range ) . The results of the analysis with and without H-NS are shown in Figure 5Bi alongside a set of “marker” transcripts ( Lane 1 ) . Lane 2 shows the result in the absence of H-NS . As expected we observed two intense bands corresponding to the 178/179 and 162 nt products . Note that because the bands in the 183–218 nt range are less abundant and poorly resolved in this assay they were not clearly visible . The 108 nt “RNAI” transcript is from the pSR replication origin and acts as an internal control . Addition of H-NS to the reactions specifically stimulated transcription from PehxCABD ( Lanes 2–5 ) . Figure 5Bi shows the effect of the -41G mutation , it is indistinguishable from the effect of H-NS . Note that both the addition of H-NS , and the introduction of the -41G mutation , resulted in a decrease in the relative abundance of the 162 nt transcript compared to the RNAI control transcript ( Figure 5B ) .
The data presented here demonstrate that nucleoprotein organisation , as well as primary DNA sequence , controls the specificity of regulatory DNA for RNA polymerase . In our model , RNA polymerase competes with itself for binding to AT-rich sequences overlapping PehxCABD ( Figure 6 ) . In the context of native nucleoprotein this self-competition is negated . This is because RNA polymerase has instead to compete with H-NS ( Figure 6 ) . Hence , evolution of RNA polymerase binding targets likely involves a trade-off between attaining the optimal DNA sequence for correct chromosome folding and precise transcription initiation . We note the PehxCABD has a consensus extended −10 element . Such sequences are incredibly rare , being found in only 3 of the 554 documented promoters in E . coli [23] . We speculate that , in very AT-rich gene regulatory regions , closer matches to the consensus RNA polymerase recognition elements are highly beneficial . Thus , in the presence of H-NS , RNA polymerase is able to recognise PehxCABD because of its close similarity to a consensus promoter . Conversely , adjacent AT-rich sequences are ignored . Interestingly , the net effect of H-NS on transcription from PehxCABD is positive and this results from correct positioning of RNA polymerase by H-NS ( Figures 4 and 5 ) . Park and co-workers [17] recently documented a mechanism for positive regulation of malT by H-NS . Although H-NS exerts its effect on malT by binding the malT mRNA there are some clear parallels with the mechanism described here . Hence , the incoming ribosome is unable to correctly recognise the 5′ end of the malT mRNA because the Shine Dalgarno sequence is ambiguous . H-NS corrects mispositioning of the ribosome by binding to an adjacent AU-rich element . We note that the effect of H-NS on binding of RNA polymerase to PehxCABD is similar to the effect of CRP on binding of RNA polymerase to the acsP2 promoter [24] . However , the molecular mechanisms underlying the effects are different . Hence , at acsP2 , CRP makes direct contacts with RNA polymerase that ensure it engages the promoter precisely . Rogers et al . [20] previously studied a 1338 bp DNA fragment carrying 126 bp of the ehxCABD gene regulatory region , the entire 516 bp ehxC gene , and 695 bp of ehxA . The fragment was fused to lacZ and , on detection of LacZ expression , it was concluded that a promoter must be located within the 126 bp regulatory section of the 1338 bp fragment . We show that , when examined in isolation , the 126 bp of DNA immediately upstream of ehxC is not able to promote transcription ( see the F2 fragment in Figure 1 ) . Similarly , no mRNA species were found to originate in this section of the regulatory region ( highlighted blue in Figure S1 ) . Thus , the only plausible explanation for the observations of Rogers et al . is that they unwittingly measured transcription from spurious promoters located within the AT-rich ehxCABD coding sequence . More recently , Iyoda and co-workers [21] examined the full ehxCABD regulatory region ( similar to our F1 fragment ) . The authors found that deleting the upstream part of the regulatory region greatly reduced transcription . Building on the assumptions of Rogers et al . ( 2009 ) the authors presumed that they had removed the binding site for a transcriptional activator . A speculative binding site for the activator was identified; this sequence aligns with the PehxCABD consensus extended −10 hexamer . Clearly , a more likely explanation is that Iyoda and co-workers had simply removed PehxCABD . Taken together , these data suggest that control of ehxCABD expression is more complex than previously thought . In particular , the possibility that additional promoters exist within the ehxCABD coding sequence is intriguing [20] . Should any such promoters be repressed by H-NS , as suggested by Rogers et al . [20] , this would further ensure specific transcription initiation from PehxCABD . We also speculate that small differences in the DNA sequence of the ehxCABD regulatory region , in different E . coli isolates , may provide information about how H-NS regulated promoter regions evolve . Further biochemical and genetic dissection of the ehxCABD locus should provide the necessary insight .
Wild type E . coli strains JCB387 and M182 have been described previously [25] , [26] . The Δhns M182 derivative ( JRG4864 ) is described by Wyborn et al . [27] . Plasmids pRW50 and pLux are described by Lodge et al . [28] and Burton et al . [29] respectively . Plasmid pSR is described by Kolb et al . [30] . More detailed descriptions of strains and plasmids are provided in Table S1 . H-NS and RNA polymerase were prepared as described previously [22] , [25] . DNA fragments for footprinting and EMSA experiments were derived from Qiagen maxi-preparations of plasmid pSR . Thus , the ehxCABD F3 fragment was excised from pSR by sequential digestion with HindIII and then AatII . After digestion fragments were labelled at the HindIII end using [γ-32P]-ATP and polynucleotide kinase . DNAse I and KMnO4 footprints were then performed as described by Grainger et al . [25] . FeBABE footprinting reactions were completed according to the methodology of Bown et al . [22] . Radio-labelled DNA fragments were used at a final concentration of ∼10 nM . Note that , apart from the KMnO4 reactivity assays , all in vitro DNA binding reactions contained a vast excess ( 12 . 5 µg ml−1 ) of Herring sperm DNA as a non-specific competitor . We checked that our reaction conditions were meaningful by comparing the affinity of H-NS for PehxCABD and the well-characterised H-NS target proU . We found that the affinity of H-NS for the two DNA fragments was similar in our conditions ( Figure S3 ) . Footprints were analysed on a 6% DNA sequencing gel ( molecular dynamics ) . The results of all footprints and EMSA experiments were visualized using a Fuji phosphor screen and Bio-Rad Molecular Imager FX . The in vitro transcription experiments were performed as described previously Savery et al . [31] using the system of Kolb et al . [30] . A Qiagen maxiprep kit was used to purify supercoiled pSR plasmid carrying the different promoter inserts . This template ( ∼16 µg ml−1 ) was pre-incubated with purified H-NS in buffer containing 20 mM Tris pH 7 . 9 , 5 mM MgCl2 , 500 µM DTT , 50 mM KCl , 100 µg ml−1 BSA , 200 µM ATP , 200 µM GTP , 200 µM CTP , 10 µM UTP with 5 µCi [α-32P]-UTP . The reaction was started by adding purified E . coli RNA polymerase . Labelled RNA products were analysed on a denaturing polyacrylamide gel . Luciferase assays were done as described by Burton et al . [29] using E . coli O157:H7 . β-galactosidase assays were completed using the protocols of Miller [32] with E . coli JCB387 , M182 or the Δhns derivative . All assay values are the average of three independent experiments and , in all cases , cells were grown aerobically , at 37°C , in LB media . The ehxCABD F1 fragment was synthesised by DNA2 . 0 ( USA ) . The F3 fragment was generated using overlapping oligonucleotides ( 5′-ggctgcgaattctatcttacaaatcaatcatctgagtgttataatataacttagctgtgatatgtgtaagaatgtttaggcaat-3′ and 5′-cgcccgaagcttcatctctcccaaccaaaacaacattagcgataataatatattgcctaaacattcttacacatatca-3′ ) . Similarly , F2 was generated using 5′-ggctgcgaattctgtttttagatgcttcttgcttaaaagaatataattcctgttcttttatatagagttctttaca-3′ and 5′-cgcccgaagcttcataatgtttaaacaaataagaaaattcagtaaatgtaaagaactctatataaaagaac-3′ . Mutations were introduced using derivatives of these oligonucleotides . All ehxCABD regulatory region sequences are numbered with respect to the transcription start point ( +1 ) and with upstream and downstream locations denoted by ‘−’ and ‘+’ prefixes respectively . Transcript start sites were mapped by primer extension , as described in Lloyd et al . [33] , using RNA purified from strains carrying the F3 DNA fragment cloned in pRW50 . The 5′ end-labelled primer D49724 , which anneals downstream of the HindIII site in pRW50 was used in all experiments . Primer extension products were analysed on denaturing 6% polyacrylamide gels , calibrated with size standards , and visualized using a Fuji phosphor screen and Bio-Rad Molecular Imager FX . Chromatin Immunoprecipitation was done exactly as described previously [8] , [34] . Briefly , formaldehyde crosslinked nucleoprotein , obtained from growing JCB387 cells carrying the F3 fragment in plasmid pRW50 , was fragmented by sonication . Some of this sample was retained as the “total DNA” fraction . DNA cross-linked with RNA polymerase or H-NS was then precipitated using a rabbit polyclonal antibody against H-NS or an antibody against the RNA polymerase β-subunit ( Neoclone ) . A control mock immunoprecipitation ( with no antibody ) was done in parallel . After immunoprecipitation the protein-DNA complexes were de-cross-linked and the DNA was recovered using a Qiagen PCR purification kit . Recovered DNA was resuspended in 50 µl of elution buffer and 1 µl of this solution was used as a template in a 50 µl PCR . The reactions were run for 28 cycles of amplification before 5 µl was loaded onto a 7 . 5% polyacrylamide gel . After electrophoresis PCR products were visualised with ethidium bromide . The oligonucleotide primers for amplification of the yabN [34] and lacZ [8] open reading frames , in their chromosomal context , have been described previously . To amplify PehxCABD we used 5′-ggctgcctcgagtatcttacaaatcaatcatctgagtgttataatataacttagctgtga-3′ and 5′-cgcccgggatcccatctctcccaaccaaaacacattagcg-3′ .
|
The information required to build and maintain a cell is written into an organism's DNA in the form of genes . When individual genes are “read , ” the DNA code is transcribed into an mRNA molecule by RNA polymerase . Hence , the DNA sequence adjacent to the start of a gene must contain a signal to recruit RNA polymerase . In certain instances this signal is difficult to differentiate from the background DNA sequence . For example , many bacterial chromosomes contain discrete sections of DNA with a high percentage of A and T nucleotides . Because RNA polymerase recognises an AT-rich signal sequence , these chromosomal regions can be ambiguous . In this paper we address the long-standing question of how RNA polymerase specifically recognises such DNA target sites . We show that a crucial factor is local nucleoprotein organisation . Hence , the manner in which DNA is folded , in conjunction with primary DNA sequence , facilitates specific RNA polymerase interactions with DNA .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genetics",
"biochemistry",
"biology",
"microbiology"
] |
2013
|
H-NS Can Facilitate Specific DNA-binding by RNA Polymerase in AT-rich Gene Regulatory Regions
|
Transposons are discrete segments of DNA that have the distinctive ability to move and replicate within genomes across the tree of life . ‘Cut and paste’ DNA transposition involves excision from a donor locus and reintegration into a new locus in the genome . We studied molecular events following the excision steps of two eukaryotic DNA transposons , Sleeping Beauty ( SB ) and piggyBac ( PB ) that are widely used for genome manipulation in vertebrate species . SB originates from fish and PB from insects; thus , by introducing these transposons to human cells we aimed to monitor the process of establishing a transposon-host relationship in a naïve cellular environment . Similarly to retroviruses , neither SB nor PB is capable of self-avoidance because a significant portion of the excised transposons integrated back into its own genome in a suicidal process called autointegration . Barrier-to-autointegration factor ( BANF1 ) , a cellular co-factor of certain retroviruses , inhibited transposon autointegration , and was detected in higher-order protein complexes containing the SB transposase . Increasing size sensitized transposition for autointegration , consistent with elevated vulnerability of larger transposons . Both SB and PB were affected similarly by the size of the transposon in three different assays: excision , autointegration and productive transposition . Prior to reintegration , SB is completely separated from the donor molecule and followed an unbiased autointegration pattern , not associated with local hopping . Self-disruptive autointegration occurred at similar frequency for both transposons , while aberrant , pseudo-transposition events were more frequently observed for PB .
Mobilization of transposable elements ( TEs ) is a DNA recombination reaction that can occur either via RNA ( retroelement/retrovirus ) or DNA intermediates ( DNA transposon ) . In non-replicative , ‘cut and paste’ DNA transposition , the excised transposon relocates from one genomic location to another . In contrast , the ‘copy and paste’ mobilization of a retroelement/retrovirus does not include the excision step , but the downstream events of retroviral integration are highly similar to DNA transposition [1] Many DNA transposons are bracketed by terminal inverted repeats ( IRs ) that contain binding sites for the recombinase , the transposase . The transposition process is catalysed by the transposase , and can be divided into four steps: ( i ) the transposase recognizes and binds to the ends of the transposon; ( ii ) the transposase and two transposon ends form a complex called synaptic or paired end complex; ( iii ) the transposon is excised from the donor site; and ( iv ) the excised transposon is transferred to a new location by the transposase reviewed in [2] . TEs are ubiquitous components of both prokaryotic and eukaryotic genomes [3] Even though TEs are best viewed as molecular parasites that propagate themselves using resources of the host cells , their long-term coexistence with their host has provided ample examples of mutual adaptation . The mobility of TEs is regulated by diverse molecular mechanisms , and can be achieved by self-limiting regulatory features intrinsic to the TE itself [4] or mechanisms provided by the host cell . For example , the RNA interference ( RNAi ) machinery in eukaryotes is probably the best-known cellular mechanism that evolved to control transposition [5] , [6] . Notably , generally little is known about the regulation of DNA transposons in eukaryotes . Indeed , our understanding of the mechanisms and the regulation of transposition in eukaryotes are mostly based on assuming analogies to bacterial transposons [2] , [7] , [8] . In the last decade , the DNA transposition of Sleeping Beauty ( SB ) , a resurrected fish transposon [9] was intensively studied [10]–[13] . Using SB as a model to study host-transposon interaction in eukaryotic cells , a series of evolutionarily conserved ( from fish to human ) cellular determinants has been identified . HMGB1 , a non-histone chromatin factor , is required for synaptic complex formation during SB transposition [11] . Factors of the non-homologous-end-joining ( NHEJ ) pathway of double strand DNA break ( DSB ) repair , including Ku70 and the DNA-dependent protein kinase ( DNA-PKcs ) are required for SB transposition by acting at repairing the transposon excision sites [10] . Through its association with Myc-interacting zinc finger protein 1 ( ZBTB17 or Miz1 ) , the SB transposase down-regulates cyclin D1 expression in human cells , resulting in a cell cycle slowdown [12] . A temporary G1 arrest enhances transposition , suggesting that SB transposition is favoured in the G1 phase of the cell cycle , where NHEJ is preferentially active [10] . The HMG-box transcription factor HMGXB4 ( HMG2L1 ) , a component of the Wnt-signaling pathway is involved in a feedback regulation of SB transposase expression [13] . These studies indicate that eukaryotic transposons can participate in a complex interactive regulatory platform involving evolutionary conserved cellular mechanisms . Although , SB is a relatively well-characterised eukaryotic transposon , one part of the transposition reaction , the step following excision but prior reintegration , is yet unexplored . In the process of productive transposition , the excised molecule integrates into a new genomic location . However , in principle , the excised transposon molecule could reinsert , in a self-disruptive process , into its own genome . This suicidal transposition event is called autointegration , self-integration or intramolecular transposition , and is well characterized in prokaryotes [14]–[16] . The best-understood example in bacteria is Tn10 transposition , in which regulation of transposition is a delicate interplay between the transposon and host-encoded factors [17]–[19] . These host factors , namely IHF ( integration host factor ) , HU ( heat unstable nucleoid protein ) and H-NS ( nucleoid structuring protein ) are among the most important regulatory factors in E . coli . IHF and HU stimulate the early steps of transposition prior to excision of Tn10 [19] . However , if they remain associated with the transpososome ( a DNA-protein complex minimally containing the excised transposon and the transposase ) , they promote autointegration [7] . By opposing the effects of IHF [20] and HU , H-NS inhibits autointegration and promotes productive transposition [18] , [19] . In eukaryotes , autointegration was reported in mariner transposition [21] , [22] Curiously , one third of the autointegration events mediated by Mos1 ( mariner ) were recovered from non-canonical target sites [22] . Self-disruptive autointegration has also been observed during retroviral integration [23]–[25] . A host-encoded protein , barrier-to-autointegration factor ( BANF1 or BAF ) has been identified by its ability to protect retroviruses from autointegration [23] . Two observations suggest that , similarly to bacterial transposons and retroviruses , autointegration could be a significant factor affecting productive DNA transposition in eukaryotes as well . First , similarly to certain bacterial DNA transposons [26] , [27] , transposition of SB from a genomic locus frequently occurs into sites that are close to the donor locus [28]; this phenomenon is termed “local hopping” . Obviously , the transposon itself is the closest target to integrate . In Tn10 transposition , the host factor IHF promotes ‘target site channelling’ close to the IR of the transposon [19] . Second , larger transposons are expected to be particularly attractive targets for autointegration . Indeed , it has been observed that , similarly to certain bacterial TEs , longer elements of SB tend to transpose less efficiently [29] , [30] . Thus , both ‘local hoping’ and size-sensitivity might be associated with vulnerability of SB transposition to self-integration . In the present study , we investigated the post-excision fate of two DNA transposons , SB [9] and piggyBac ( PB ) [31] in vertebrate cells . Although , both SB and PB belong to the superfamily of DDE/D transposases , characterized by a highly conserved catalytic domain [1] , they exhibit significant differences in their mechanisms of transposition [32] , [33] . For example , the activity of SB is essentially restricted to vertebrates [29] , [34] , with the exception of a chordate , Ciona intestinalis [35] . By contrast , PB seems to have an extremely wide host range as it can transpose in insects as well as in human cells [36]–[38] . In comparison to PB , SB was reported to exhibit a much stronger ‘local hopping’ phenotype [39] , [40] . Furthermore , SB , but not PB was reported to be sensitive to the size of the mobilized element . Specifically , the transposition of PB was reported to be independent on the size of the element below 14 kb [41] . Importantly , both SB and PB are valuable genomic tools for genome manipulation [42] , and mostly used in heterologous cellular environments , thereby offering a unique opportunity to investigate various survival strategies of DNA elements in eukaryotes . Indeed , we can model how these elements behave in naïve genomes , and adapt to their new environment . We have used a simple experimental setup , i . e . , transfection into cultured cells to monitor the process of establishing a host-parasite relationship in a heterologous environment . This strategy identified BANF1 as a host-encoded factor influencing this process . We propose that deciphering the mechanism and regulation of transposon reactions and translating this knowledge can be effectively used to derive transposon-based genetic tools for genome manipulation or for gene therapy .
To detect and characterise potential autointegration products , the following assay system was established . The test construct , SBrescue , is a plasmid comprising a replication origin ( Ori ) and an antibiotic resistance cassette for zeocin ( Zeo ) located between the IRs of the transposon ( Figure 1A ) . Outside of the transposon SBrescue contains the rpsL gene rendering bacteria sensitive to streptomycin [43] . SBrescue and the helper plasmid encoding for the transposase are co-transfected into cells . Plasmid DNA is recovered from the cells two days post-transfection and transformed into E . coli . Bacteria are subjected to double antibiotic selection of zeocin and streptomycin ( Figure 1B ) . Following transposon excision and circularization of the excised transposon , the rpsL is lost , thereby rendering bacteria StrepR ( Figure 1B ) . Autointegrative transposition events can be rescued in the form of either two deletion circles or a single inversion circle , depending on the topology of the strand attack ( Figure 1C ) . The assay can detect autointegration events occurring into regions designated A , B , C and IR ( Figure 1A ) . In addition , integration events into the rpsL gene would render bacteria resistant to streptomycin and recovered by the assay . In contrast , autointegration events into Zeo or Ori would not be detectable with the assay system , because these regions are required for plasmid propagation and maintenance . To identify conditions affecting autointegration of SB , the following factors were considered: ( a ) cell type specificity; ( b ) transposase activity; ( c ) target site distribution; ( d ) the size of the transposon; ( e ) host-transposon interaction . First , SBrescue was introduced into human HeLa cells with or without a helper plasmid expressing the hyperactive SB100X transposase [44] ( Figure 1B ) . Compared to the control ( 0 . 03% , 1 . 19×103/3 . 88×106 ) , significantly elevated numbers ( 0 . 45% , 4×103/9 . 09×105 ) of ZeoR/StrepR bacterial colonies were observed when SB100X transposase was present in the experiments ( Figure 1D ) . To characterize potential autointegration events and map the transposon insertion sites , the recovered products were subjected to DNA sequencing . Sequencing data confirmed that similarly to productive transposition , the autointegration events of SB transposition were targeted into TA dinucleotides within the mappable A , B , C and IR regions of the transposon ( Figure 1E ) . To investigate if cellular factors in various vertebrate species might differentially promote or protect against autointegration of SB , the assay was performed in cultured cells of different origin , including AA8 ( Chinese hamster , ovarian; 0 . 85% vs 0 . 03% , 1 . 51×103/1 . 77×105 vs 476/5 . 52×106 ) , MEF ( mouse , embryonic fibroblast; 0 . 05% vs 0 . 01% , 8 . 06×103/1 . 71×107 vs 332/9 . 54×105 ) and PAC2 ( zebrafish , fibroblast; 0 . 13% vs 0 . 08% , 1 . 03×103/8 . 42×105 vs 770/9 . 69×105 ) cells ( Figure 1D ) . Our results revealed that the SB-mediated autointegration events were detectable in all tested cell lines , including fish , the natural cellular environment of SB ( Figure 1D ) . Similarly to productive transposition , the frequency of autointegration varied in the different cell types [29] . The highest frequencies of autointegration were detected in HeLa and AA8 cells that generally support efficient transposition [29] , suggesting that the frequency of autointegration was primarily dependent on the activity of the transposase , rather than the cell type ( Figure 1D ) . Indeed , compared to the original SB10 transposase [9] , autointegration by the hyperactive SB100X transposase [44] was higher by one order of magnitude in human HeLa cells . Remobilization of the SB transposon from a genomic donor site exhibits a significant bias toward the donor locus ( local hopping ) [32] . Similarly , the reintegration of Tn10 transposons is not unbiased and targeted to the IRs of the transposon during autointegration , referred as ‘target site channelling’ [19] . In contrast , when launched from an extrachromosomal donor molecule , the genomic distribution of SB insertion sites is fairly random [45]–[47] . Target site selection during transposition of SB from an extrachromosomal plasmid is primarily determined on the level of DNA structure , as insertion sites tend to have a palindromic pattern and a bendable structure [45] . Accordingly , the insertion profile of the SB transposon can be modelled by determining the DNA-deformability scores , called Vstep for each potential TA target site , using the software ProTIS [48] . To determine the autointegration profile of SB , Vstep values were generated for the mappable regions of SBrescue and the observed insertion frequencies were compared to the calculated Vstep values ( Figure 2B ) . Altogether , 53 autointegration products were identified and mapped to the regions of IR , A , B and C . Most of the autointegration events occurred into region B that is farther away from the IRs , and relatively few into regions A and C that are closer to the transposon ends ( Figure 2 ) . In regions B and C , there was a correlation between insertion frequencies and Vstep scores ( Figure 2B ) . These results suggest that similarly to transposition from an extrachromosomal donor , insertion site selection during autointegration of SB is largely independent from the donor site and did not exhibit ‘target site channelling’ close to the IRs of the transposon . On the contrary , despite of the predicted high Vstep score , only a single insertion event was recovered from the IRs ( Figure 2 ) , suggesting that the transposon ends of SB , embedded in a paired end complex are limited in their abilities to target the IRs or sites close to the IRs during autointegration . Due to the linkage , the autointegration of SB was primarily intramolecular , and no insertions were detected from the rpsL region . Thus , the transposon was fully excised from the flanking donor DNA prior its integration into a new site . Next , we tested whether self-destructive autointegration could also occur during PB transposition . We have used a transposon donor construct that is identical to SBrescue , except that the SB IRs were replaced by PB IRs [49] ( PB2K in Fig . 3B ) , together with a mouse codon-optimized PB transposase ( mPB ) [50] . As shown in Figure 3A , autointegration of the PB transposon occurred at frequencies comparable to SB100X ( 0 . 49% , 3 . 2×104/6 . 4×106 ) in HeLa cells . As predicted and confirmed by DNA sequencing , autointegration of PB occurred into TTAA motifs , the canonical target site of PB [31] ( Supporting Figure S1 ) . Altogether , 23 integration sites were mapped and twelve were recovered from regions B and C ( Figures 3B , C ) . However , unlike with SB , a significant number of integration events ( 48% , 11/23 ) mapped outside of the transposon , in the rpsL gene ( Figures 3B , C ) . These non-canonical transposition events also targeted TTAA target sites , but involved only a single end of the transposon . The other IR was not separated from the donor molecule during the reaction . We refer to these non-canonical transposition events as single-ended transposition . To investigate the phenomenon of single-ended transposition of PB further , a reciprocal construct , PBsingle was generated , where the PB transposon carried an rpsL gene ( Figure 3D ) . In addition to single-ended transposition events detected by PB2K , the PBsingle assay system was suitable to capture various deletion products ( Supporting Figure S2 ) . Bacteria that gained StrepR could report on ( i ) double-ended excision products , ( ii ) single-ended integration events into either rpsL or ( iii ) the vector sequence flanking the transposon . The autointegration assay was performed as shown in Figure 1B , except bacteria were exposed to double selection of kanamycin and streptomycin . To capture single-ended events , 336 transposition products were pre-filtered by colony PCR , using primers flanking the PB excision site . Canonical excision products would appear as uniformly sized PCR products , while size difference would report on either single-ended transposition or non-transposase-mediated small deletions/insertion events generated by DNA repair . 31/336 pre-filtered PCR products were analysed further by DNA sequencing , and six out of 31 ( 19% ) products were clearly generated by PB transposase-mediated , single-ended transposition that occurred into TTAA either inside or outside of the transposon ( Figures 3D and 3E ) . In the ‘single-ended’ transposition of PB , only one of the IRs was mobilized . Still , true single ended events , when the second IR is not involved in any of the steps of transposition , cannot be convincingly demonstrated . In fact , alternative mechanisms can generate similar , hard-to distinguish products . For example , the canonical transposition reaction might fail at the final step , and only one end of the transposon is transferred ( lariat model ) , ( Supporting Figure S2 ) . Aberrant transposition might also occur by a mechanism that involves pseudo or cryptic sites mistakenly recognized as IRs . In addition , the ends of the transposon can also be derived from two separate molecules [51] ( bimolecular transposition ) . To explore the scenario of bimolecular transposition , truncated ‘solo’ transposons were generated . ‘Solo’ substrates , lacking either the left ( PBΔleft; SBΔleft ) or the right IRs ( PBΔright; SBΔright ) were tested in a cell culture-based transposition assay [9] . Molecular analysis of the resistant colonies revealed that neither PBΔright nor SBΔleft supported transposition ( Table 1 ) . In contrast , the analysis confirmed transposase-mediated transposition of the ‘solo’ substrates , PBΔleft ( 4 . 6% ) and SBΔright ( 0 . 56% ) [52] ( Table 1 ) , indicating that both transposases are capable of utilizing ‘solo’ substrates . In either cases , the IRs of the ‘solo’ transposons were properly integrated into respective target sites ( Supporting Test S1 ) . Notably , in clone PBΔleft#8 , we have identified a second right IR integrated into a same genomic locus , confirming that the transposase used the two IRs from separate molecules ( Supporting Text SF1 ) . As ‘solo’ transposition occurred ∼8-fold more frequently for PB , we monitored the PB system further in the ‘solo-mixing’ experiments . In this strategy , the PBΔleft and PBΔright constructs were transfected either alone or mixed in equimolar ratios , and tested in the colony forming , transposition assay . If transposition utilizes the IRs from separate molecules , one would expect elevated colony numbers when either PBΔleft or both ‘solo’ substrates are present in the assay , compared to PBΔright that does not support transposition alone ( Table 1 ) . The higher number of resistant colonies in the respective experiments indicated that the transposase was able to utilize the IRs from different copies of the transposon , supporting the bimolecular model ( Figure 4 ) . The efficacy of transposition was reported to depend on the size of the transposon [29] , [30] , [53]–[55] . One potential mechanism responsible for such size-dependence is that following transposon excision , self-disruptive autointegration competes with productive transposition . Since larger transposons have more target sites , they could be particularly attractive targets for autointegration . This hypothesis predicts that the size of the transposon does not affect the frequency of excision , but it shifts the ratio between autointegration and productive transposition . To test this assumption , a series of transposons of different size , ranging from 2679 bp to 7256 bp ( SB2K , SB3K , SB4K , SB7K ) and 2795 bp to 7319 bp ( PB2K , PB3K , PB4K , PB7K ) were generated for SB and PB , respectively . Frequencies of transposon excision , autointegration and productive transposition events were determined for the various transposons . Excision frequencies were estimated by quantitative PCR , autointegration was monitored as above . Productive transposition was determined in a cell culture-based assay [9] . Figure 5A shows that excision frequencies declined with increasing size , while autointegration frequencies elevated over 4 kb either moderately or sharply for SB and PB transposons , respectively ( Figure 5A ) . Accordingly , productive transposition frequencies dropped with increasing size of both SB and PB transposons . These results indicated that the size of the transposon affected transposition already at the excision step , thereby arguing against the hypothesis of autointegration being the sole factor that compromises productive transposition with increasing transposon size . Nevertheless , autointegration contributes as an additive element to the less efficient transposition of long transposons . Surprisingly , the two transposons behaved similarly in all three assays ( Figures 5A an 5B ) . Thus , in contrast to general assumptions , and similarly to SB , size affects PB transposition as well . A cellular protein , BANF1 ( BAF ) barrier-to-autointegration factor was identified by its ability to protect retroviruses from autointegration [23] . BANF1 binds to double-stranded DNA , including freshly transfected , extrachromosomal plasmid DNA [56] , in a non-specific manner [57] , [58] . Thus , in principle , BANF1 could affect DNA transposition as well , between the molecular steps of excision and reintegration , when the transposon exists as an extrachromosomal molecule in the cell . To test this assumption , we asked if BANF1 could protect DNA transposons from autointegration . We addressed this question by monitoring autointegration events in HeLa cells , where BANF1 was either knocked-down or transiently overexpressed ( Figure 6A ) . When BANF1 expression was knocked-down by RNA interference ( Supporting Figure S3 ) , the frequency of autointegration of SB was increased by two-fold compared to the control ( Figure 6A , left panel ) . In contrast , BANF1 overexpression decreased the frequency of autointegration to one third ( Figure 6A , left panel ) . Similar results were obtained by using the PB transposon ( Figure 6A , right panel ) . No significant effect of BANF1 was observed at the excision step of SB transposition ( not shown ) , suggesting the BANF1 acted specifically following excision . In addition to BANF1 , the effect of another host-encoded factor , the high-mobility group protein ( HMGB1 ) was tested on autointegration . Similarly to BANF1 , HMGB1 binds DNA in a non-specific manner [59] . In SB transposition , the transposase physically associates with HMGB1 and recruits it to the transposon DNA [11] . Autointegration was monitored in cells where HMGB1 was either transiently overexpressed or knocked-out [60] . Although , HMGB1 overexpression or deficiency was significantly affecting productive transposition [11] , it had no detectable influence on autointegration ( Supporting Figure S4 ) . These results indicate that despite their similar non-specific DNA-binding activity , BANF1 and HMGB1 have a clearly distinct effect on DNA transposition . Alternatively to a non-specific engagement , and similarly to retroviruses , BANF1 might be actively recruited to a preintegration complex of a transposon . In order to distinguish between these two scenarios , a high throughput immunoprecipitation experiment was designed to analyse a protein interactome forming around the SB transposase in mammalian cells . Affinity purification combined with mass spectrometry is a powerful strategy to detect protein-protein interactions among proteins in their native cellular environment [61] . This method is suitable to reveal the composition of entire protein complexes . If we use the analogy to retroviruses [62] , one should keep in mind that even if BANF1 is recruited actively to the preintegration complex , it might not be recruited directly by the transposase . To distinguish true interaction partners from non-specific contaminants , we needed an easy-to-detect , confirmed interacting partner of the SB transposase as bait . We can readily monitor interactions of HMGXB4 ( HMG2l1 ) with either the transposon or the transposase in vivo [13] . Thus , HMGXB4 was chosen as bait to analyse higher order complexes formed around SB . The experiments were run in parallel , in the presence and in the absence of the SB transposase . In the control experiment , it is not expected to detect interaction partners of the SB transposase . HEK293T cells were transiently transfected with HA-tagged HMGXB4 protein in the presence/absence of the SB10 transposase [9] . A SILAC pull-down experiment was performed . This experimental strategy identified BANF1 as an interaction partner of HMGXB4− in the presence , but not in the absence of the SB transposase ( Figure 6B ) . The presence of BANF1 was also detectable when the bait , HMGXB4 was used in a co-immunoprecipitation assay ( Figure 6C ) . This observation predicts that BANF1 can be actively recruited into a higher order protein complexes forming around the SB transposase in mammalian cells .
This study focuses on molecular events following the excision steps of two eukaryotic DNA transposons , SB and PB , derived from fish and insect genomes , respectively . The transposition reactions were performed in a heterologous host environment , phylogenetically distant from their natural hosts . The experimental setup mimics the scenario of introducing DNA transposons into a naïve eukaryotic host . We have shown that a significant portion of SB and PB transposon excision events is accompanied by suicidal integration into the transposon's own DNA . Although , different transposons may have different frequency of autointegration depending on the structure of the transpososome and the number of the integration target sites on the transposon , autointegration would influence the success of a transposon in a new environment . Neither SB nor PB was immune to the suicidal process of autointegration . Thus , in general , transposases/integrases in eukaryotes might not be able to distinguish between their own genome form foreign DNA . This would define autointegration as the lack of ability of self-avoidance upon integration . In contrast , certain prokaryotic transposons , including Tn7 and Mu exhibit ‘target immunity’ that prevents the transposon from transposing into its own genome [63] , [64] . Both Tn7 and Mu avoid integration into DNA molecules that already have a copy of the transposon . As an alternative to self-encoded ‘target immunity’ , some bacterial transposons and eukaryotic retroviruses recruit cellular host factors to protect against autointegration [19] , [23]–[25] . In Tn10 transposition a host protein , histone-like nucleoid structuring ( H-NS ) plays a role in promoting intermolecular and supressing self-destructive intramolecular integration events [19] . Similarly , DNA transposons in eukaryotes might also capture cellular factors to protect their genome against autointegration . This strategy could defend the invading molecule and contribute establishing a stable host-transposon relationship . BANF1 is involved in several critical processes , including host defence [65] , [66] . The usual mode of BANF1 is repressive , due to its propensity to coat DNA . For example , BANF1 acts as a potent inhibitor of virus replication , defending against poxvirus invasion [67] . Intriguingly , and in contrast to its original function in host defence , BANF1 is piggybacked by various retroviruses to protect their viral genome against autointegration . BANF1 inhibits autointegration of the Moloney Murine Leukemia retrovirus , MoMLV [23] , [68] , [69] or HIV-1 [62] . By physically protecting the retrovirus , BANF1 promotes productive viral integration into the host genome [62] . In our experimental setup , BANF1 was influencing the fate of the excised molecules of two DNA transposons of different origin , SB and PB . Thus , in addition to its reported activity to bind freshly transfected DNA [56] or retroviral cDNA [23] , BANF1 might influence the fate of DNA transposons as well . An important ramification of utilizing phylogenetically conserved cellular proteins by transposons might be the ability to survive and establish stable host-parasite relationship in a heterologous host environment . Accordingly , in addition to its role in Tn10 transposition , H-NS was reported to selectively bind the transpososomes of Tn5 , and is likely to modulate many other transposition processes in Gram-negative bacteria [70] . SB and PB are members of the superfamily of DDE/D transposases and retroviral integrases , utilizing the same strategy for target joining . Still , how reasonable it is to assume an interaction of BANF1 with both DNA transposons and retroviruses ? In fact , BANF1 might be an ideal cellular factor for integrating elements in higher eukaryotes . Due to its non-specific DNA-binding activity to double-stranded DNA [58] , a capacity to compact DNA and assemble higher-order nucleoprotein complexes , BANF1 could influence the fate of any extrachromosomal DNA molecule . As in retroviral integration [23] , [69] , BANF1 may compact the transposon genome to be a less accessible target for autointegration , and promote the integration step . Furthermore , similarly to retroviruses , BANF1 could be even actively recruited to preintegration complexes . The exact manner of recruitment might vary , providing specificity . BANF1 is recruited via physical interaction by the viral matrix protein gag to the retroviral preintegration complex of HIV-1 [62] . In SB transposition , BANF1 was enriched in a higher order complex containing the SB transposase and its interactor HMGXB4 . Thus , the enrichment was mediated via protein-protein interaction . Since the experimental setup did not include the transposon DNA , we could not faithfully simulate preintegration complex formation . Nevertheless , HMGXB4 is a specific interaction partner of both the transposon and the transposase of SB [71] . Therefore , it might be reasonable to assume that BANF1 associates with the preintegration complex . In sum , our strategy to model the process of establishing a host-transposon relationship in a naïve environment identified BANF1 as a host encoded factor influencing this process . Future work will have to clarify if a common role of BANF1 to protect integrating mobile elements in general exists . Traditional models predict that efficient integration must follow the excision of DNA elements . Strikingly , autointegration was estimated to be over 90% in mariner transposition in vitro , suggesting that under the standard reaction conditions , the vast majority of the excised transposon inserts into itself , rather than into another DNA molecule [21] . This high frequency would establish autointegration as a major factor affecting productive integration . Furthermore , as longer transposons present more potential target sites , autointegration would be a reasonable explanation for size-dependence of transposition , observed for both SB [29] , [30] and PB ( this work ) transposition . Still , the role of autointegration in counteracting productive transposition might be overestimated . We found that transposon excision , a step prior to integration , is already affected by the size of the transposons ( Figure 5A ) , indicating that a larger transposon might have difficulty to form a synaptic complex . Our data argue that competition between self-integration and productive transposition is unlikely to be the only factor responsible for sensitivity to size . If we assume that unproductive transposition equals suicidal autointegration , the gap between transposon excision and productive transposition could be a good estimate for the effect , and was reported to be around 25% in SB transposition in vivo [32] . In contrast to an earlier report [41] , we found that SB and PB transposons were affected similarly by the size of the transposon in three different assays ( Figure 5 ) . When the size of the transposon increased from 2683 to 7260 and 2795 to 7319 bps , the frequency of productive transposition dropped by 83% and 89 . 6% for SB and PB , respectively ( Figure 4B ) . In addition , SB and PB behaved similarly in assays monitoring either excision or autointegration ( Figure 5A ) . Therefore , our data argue against the general assumption that the PB transposon is not sensitive to size below 14 kb [41] . The different observation might be related to the fact that ( i ) the DNA fragment that Ding et al . used to increase the size of the transposon contained a higher density of TTAA target sites than the existing transposon . Actually , it is impossible to separate the true effects of length and numbers of target sites for a transposon that is highly specific in terms of integrating into a given sequence; ( ii ) Ding et al . estimated transposition frequencies in transgenic mouse experiments by counting transgenic embryos , regardless of the copy number of the integrated elements per embryo . Therefore , to compare productive transposition of SB and PB transposons , we have adjusted transgenic frequencies by the copy number of the integrated transgenes [72] . Importantly , small size does not seem to be an absolute requirement for mobilization in either case . Decreasing the distance outside the transposon ends of SB was reported to increase transpositional rates under experimental conditions [29] . Moreover , both PB and SB100X were reported to capable of mobilizing giant molecules of DNA , such as BACs ( bacterial artificial chromosomes ) [37] , [73] . These reports indicate that in contrast to viruses , DNA transposons have no strict ( if any ) upper limit regarding their cargo capacity . Autointegration of SB , likely due to physical constraints , avoided the IRs , suggesting that the captured events were rather intramolecular than intermolecular . Nevertheless , SB integration is not channelled to the terminal repeats of the transposon as it was observed for Tn10 [19] . Furthermore , the lack of linkage of autointegration sites to nearby regions at the donor DNA molecule would argue against an association between the ‘local hoping’ phenotype and autointegration . Our experimental approach gave us the opportunity to have a closer insight into the mechanism of both PB and SB transpositions . We have captured autointegration products at comparable frequencies for both SB and PB . We assume that the excision and reintegration steps of autointegration and canonical transposition are mechanistically not significantly different [32] , [41] , [74] ( Figures 1 and S1 ) . In addition to the autointegration products , our assays detected aberrant , pseudo-transposition events . In the ‘single-ended’ transposition products of PB , one IR of the transposon was clearly separated from the donor site , without obvious involvement of the other IR in the reaction ( Figure 3D ) . The liberated end of PB targeted either the transposon or the backbone DNA ( Figures 3C and 3E ) . SB did not display this feature in a similar assay system . By contrast , both transposons were capable of mobilizing substrates , lacking one of the IRs from separate molecules ( [52] and this work ) . These bimolecular transposition events were eight-fold more frequently detected for PB . How could aberrant transposition events be generated ? In fact , ‘true single ended’ transposition , when a transposase interacts with a single transposon end , performs the cleavage and integration steps without the involvement of another end has not been undoubtedly reported from any system . In fact , alternative mechanisms can generate hard-to distinguish , similar products . For example , the canonical transposition reaction could fail at the final step , and only one end of the transposon is transferred ( lariat model ) . In addition , our ‘solo’ experimental data support the ‘bimolecular model’ , when the ends of the transposon derive from separate molecules [51] . In addition to single ended events , small deletions at the donor sites of PB transposition are assumed to be associated with imprecise transposon excision , and involve non-homologous end joining [40] . These structures were reported following PB excision in Drosophila ( 4 . 3% ) , mouse ( 5% ) and in human cells [38] , [40] . Aberrant pseudo-transposition can be considered as a fidelity problem of the transposition reaction , and has been observed with P-element in Drosophila , Ds element in Arabidopsis , Ac/Ds elements in maize [51] , [75] or Tam3 in Antirrhinum majus [76]–[80] . Small sequence variations generated by NHEJ at the excision sites are unlikely to cause genome rearrangements . By contrast , pseudo-transposition events can generate difficult-to-repair lesions and be genotoxic . Aberrant transposition events were reported to induce deletions , insertions , chromosome translocations and could initiate McClintock's chromosomal breakage-fusion-bridge cycles [51] , [81] . Occasional mis-pairing between extrachromosomal molecules would not compromise the safety feature of a transposon-based transfer vector in a heterologous environment . However , fidelity problems could be problematic when the transposon is mobilized from the genome . Thus , cells subjected to PB-based genome manipulation techniques , e . g . , transgene-free iPS cells generated by PB excision [82] , should be carefully monitored for genome rearrangements . There seems to be a basic difference in the ways transposons in pro- and eukaryotes control their activity to minimize the potential genotoxicity generated by improper synapsis of the transposon ends . For all classical bacterial transposons characterized to date , including Tn5 transposition , the catalytic steps of the reaction are tightly coupled to the synapsis of the transposon ends [83] . In addition , the coupling of transcription and translation in bacteria also increases the probability of a proper synapsis as the transposase binds tightly to the first IR before searching for nearby ends . In contrast , eukaryotic transposases must search at random for transposon ends when they enter the nucleus . Therefore , regulatory mechanisms promoting accurate double-ended reactions from the same transposon molecule are crucial . Tc1/mariner transpositions , including SB , might have invented novel “built in regulatory checkpoints” to enforce synapsis prior catalysis [21] . A simple topological filter could also suppress promiscuous synapses of distant ends of the transposon [84] . Furthermore , certain transposition-like reactions , including V ( D ) J recombination , are also capable of filtering out unpaired reaction products . This regulatory mechanism , assisted by a cellular factor , HMGB1 , regulates a highly controlled , ordered assembly process [85] , [86] . Similarly to V ( D ) J recombination , HMGB1 was reported to assist paired end complex formation of SB [11] . In addition to HMGB1 , SB transposition requires various vertebrate-specific host factors [10] , [11] , [13] , [29] that render SB transposition restricted to vertebrates . In contrast , PB has an incredibly wide host range ( from yeast to human ) that could be associated with loose or no host factors requirement . In comparison to SB , PB transposition results in more frequent , aberrant transposition products in a heterologous environment . Why is it so ? If PB does not use host factors to enforce fidelity of the end pairing before excision , the reaction might be less precise by its nature . Alternatively , PB might utilize a host factor in its endogenous host ( insect ) that guarantees precise regulation . However , this factor is diverged or not available in mammalian cells . Finally , both PB and SB transposons have “built in regulatory checkpoints” that are most effectively filter out aberrant products under optimal conditions and in appropriate hosts . Notably , aberrant transposition events , including single-ended transposition of the Mos1 , mariner element were observed under suboptimal conditions [22] . In sum , when a transposon is transferred too far from its original host , the conditions in a new environment could be suboptimal , and the fidelity of the reaction could be compromised . The wide host range of PB can be explained by relative independence from host-encoded factors , perhaps a price to be paid for fidelity .
The IRs of the transposons were identical to the versions published earlier [49] , [87] and were not modified for the assays . All the primers used for construct cloning were listed in Supporting Table S1 . SBrescue: XmnI/BsaI fragment ( Klenow-filled ) containing ampicillin gene on pUC19 was replaced by PstI and SalI fragment containing zeocin gene from vector pZEO ( isolate SV1 , Invitrogen ) resulting in pUC19-zeo . Klenow-filled SapI/SspI fragment containing zeocin gene and replication origin was inserted into EcoRI site of PT2/HB to get PT2/SBzeo . The transposon was PCR-amplified with primer AATASB-IR from PT2/SBzeo and ligated to rpsL gene fragment , which was PCR-amplified with primers rps1F/rpslR from nNG639 [43] . SB2K: BspHI/EcoRI fragment containing zeocin gene on SBrescue was replaced by BsaI/BglII fragment containing zeocin and promoter sequences from pFP-Zeo [88] . SB3K , SB4K and SB7K: DNA fragments were PCR-amplified from bacteriophage lamda DNA , using primers lam1kF/lam1kR , lam1kF/lam2kR and lam1kF/lam6kR , respectively , and were inserted into XbaI site ( Klenow filled ) of SB2K . PB2K: Klenow-filled NotI/HindIII fragment containing zeocin gene from SBrescue was inserted into SpeI site of pUC19PBneo [72] resulting in PUC19XLzeo . PvuII fragment containing PB transposon was ligated to rpsL gene PCR-amplified with primers rps1F/rpslR from nNG639 . PB3K , PB4K , PB7K: The AatII/BglII fragments containing lamda DNA from SB3K , SB4K and SB7K were inserted into AatII/BglII sites of PBPr respectively . pcDNA3 . 1BANF1 ( BANF1 gene expressing vector ) : BANF1 coding sequence was PCR-amplified from pcDNA3 . 1/HiscBANF1 ( a gift from Katherine Wilson , Johns Hopkins University ) with primers BAFF/BAFR and cloned into EcoRV site of pcDNA3 . 1/Zeo ( + ) ( Invitrogen ) . BAF-RNAi: Oligos of BAF96F/BAF96R were annealed together and cloned into BglII/HindIII site of pFP-Neo-H1 [88] . To generate ‘solo’ substrates PB pUC19XLneo [69] was digested with BamHI to delete the right IR ( PBΔright ) or with KpnI to remove the left IR ( PBΔleft ) . For “solo” SB , pTneo was digested EcoRI to generate SBΔleft , while the digestion with BamHI yielded SBΔright . HeLa , AA8 and mouse MEF cells were cultured at 37°C with 5% CO2 in Dulbecco's modified Eagle's medium ( DMEM , Gibco/Invitrogen ) supplemented with 10% fetal calf serum ( FCS , PAA ) . The zebrafish PAC2 cells were grown at room temperature and atmospheric CO2 concentrations in Leibovitz L15 medium ( Gibco/Invitrogen ) supplemented with 15% FCS . Cells were transfected at 50–80% confluence with QIAGEN-purified plasmid DNA using jetPEI ( Polyplus transfection , for mammalian cells ) or FuGene6 ( Roche , for fish cells ) according to instructions of manufacture . Transfection efficacy of a ∼3 kb and a ∼7 kb plasmid containing GFP cassette was monitored and compared by FACS analysis , but no significant difference was found ( not shown ) . Cell culture and transfection was done as described [9] . Typically , 1 . 5×105 cell were subjected to transfection with plasmids containing the transposon ( 500–1000 ng ) and the transposase ( 60–100 ng ) . Two days post transfection plasmid DNA was recovered and transformed into bacteria ( Invitrogen , ElectroMAX DH10B Cells , Cat . No . 18290-015 , Genotype: F– mcrA Δ ( mrr-hsdRMS-mcrBC ) Φ80lacZΔM15 ΔlacX74 recA1 endA1 araD139 Δ ( ara leu ) 7697 galU galK rpsL nupG λ– ) . Bacteria were subjected to either zeocin ( to determine total number of plasmids ) or zeocin/streptomycin double selection ( to determine autointegration events ) . The number of autointegration events was normalized by total number of plasmids . To confirm autointegration events , individual bacterial colonies were cultured and recovered plasmid DNA was subjected to DNA sequencing using primers of psbLacR3 and PB-F or PB-R for SB- and PB transposon , respectively . For BANF1 overexpression or knockdown experiments , 300 ng of pcDNA3 . 1BANF1 or BAF-RNAi plasmid was cotransfected with the transposon and helper constructs . Cell culture and transfection was done as described [9] . Two days post transfection 105 cells were plated on 10 cm dishes and exposed to antibiotic selection ( 100 ng/ml zeocin , for two weeks ) . Resistant colonies were visualized by methylene blue staining [9] . Transgene copy number was normalized by using qPCR specific to zeocin . The plasmid DNA was prepared as described in autointegration assay and dissolved in 50 µl water . Excision frequencies of eight transposon plasmid constructs of various sizes ( four SB and four PB ) were estimated by using a quantitative , real-time PCR ( 7700 sequence detection system from ABI , Applied Biosystems , Foster City , CA ) . To determine the total number of parental plasmid DNA molecules , a ‘parental’ titration curve was established . PCR primers of rpsL-F/rpsL-probe/rpsL-R were used to amplify the rpsL gene on the construct of SBrescue . For the curve , dilutions of 10−2 , 10−3 , 10−4 , 10−5 , 10−6 ng of SBrescue plasmid DNA were subjected to a PCR reaction to amplify the rpsL gene ( rpsL-F/rpsL and probe/rpsL ) . To quantify the total number of parental plasmid molecules , total DNA extract was used ( 3 µl , diluted by 2000-fold , rpsL-F/rpsL-probe/rpsL-R ) . The excision products were PCR-amplified from the total extract DNA using nested PCR ( 1st round , primers of rpslexciF1/rpslexciR1 , 94°C for 30 s and 30 cycles of 94°C for 30 s , 58°C for 30 s , and 72°C for 30 s; 2nd round , rpslexciF2/rpslexciR2 , 1 µl , diluted by 100-fold , 94°C for 30 s and 35 cycles of 94°C for 30 s , 58°C for 30 s , and 72°C for 30 s ) . The amplified products ( 10−2 , 10−3 , 10−4 , 10−5 , 10−6 ng ) were used to establish a second titration curve , specific for the excision products . To quantify excision products , primers of SB-F/SB-probe/SB-R and PB-F/PB-probe/PB-R were used on a total DNA extract ( 5 µl ) , for SB and for PB , respectively . The excision frequency was calculated as the ratio of excision products normalized by the total number of parental plasmid molecules . qPCR was performed for each experimental sample in triplicates . Ct values were determined following recommendations by the manufacturer . Briefly , bacteria were picked by a pipette tip and directly subjected to a PCR assay using primers of PB-F and PB-R ( 5 pmol of each , Supporting Table S1 ) and Taq polymerase ( Takara ) in a total volume of 20 µl . PCR program: 94°C for 1 min; 30 cycles of 94°C for 30 s , 58°C for 30 s , and 2°C for 30 s; and 72°C for 2 min . A triple SILAC pull-down experiment was performed using anti-HA resin . HEK293T cells were transiently transfected with HA-tagged wild type or mutant HMGXB4 ( HMG2l1 ) [13] and SUMO1 in the presence/absence of Sleeping Beauty , SB10 [9] using Polyplus-transfection jetPEI transfection reagent with 3 µg of plasmids each . We compared proteins co-purifying with HA in cells expressing the empty vector ( “light” ) , HA-tagged HMGXB4− with mutated sumoylation site ( “medium” ) and HA-tagged wild-type HMGXB4 ( “heavy” ) . The cells were plated on a 15-cm dish and harvested 48 h post-transfection . Two dishes were used for each condition . Detection of interaction partners is performed by mass spectrometry and the results obtained were analyzed by MaxQuant computational platform [89] . Results presented show protein abundance ratios between cells transfected with HMGXB4− and the empty vector control . Whole-cell extracts were prepared using extraction buffer ( Tris-HCl 50 mM at pH 8 . 0 , NaCl 150 mM , 0 . 1% SDS ( Na-dodecylsulphate ) Triton X-100 1% and Na-deoxycholate 0 . 5% ) supplemented with protease inhibitor cocktail ( Roche , Mannheim , Germany ) . For immunoprecipitations , equal amounts of lysate ( containing 5 mg of total cellular protein from HEK293 cells ) were pre cleared with protein G-agarose beads ( Sigma , St Louis , MO ) . Pre-cleared extracts were incubated with EZview Red Anti-HA Affinity Gel ( Sigma-Aldrich , USA ) for 1 h at 4°C . Precipitates were washed extensively in extraction buffer . Bound complexes were eluted with 2× SDS–PAGE sample buffer and resolved by 7 . 5–15% SDS–PAGE . Immunoblotting was performed according to standard procedures and proteins detected with the indicated antibodies . Antibodies were detected by chemiluminescence using ECL Advance Western Blotting Detection Kit ( Amersham Bioscience ) .
|
Transposons ( “jumping genes” ) are ubiquitous , mobile genetic elements that make up significant fraction of genomes , and are best described as molecular parasites . During ‘cut and paste’ transposition , the excised transposon relocates from one genomic location to another . Here we focus on the molecular events following excision of two eukaryotic DNA transposons , Sleeping Beauty and piggyBac . Both transposons are primarily used in a cellular environment that is different from their original hosts , thereby offering a new model to study host-parasite interaction in higher organisms . In the last decade , they have been developed into a technology platform for vertebrate genetics , including gene discovery , transgenesis , gene therapy and stem cell manipulation . Despite the wide range of their application , relatively little is known about their molecular mechanism in vertebrates . We show that these elements are not capable of self-avoidance , as a significant portion of the excised transposons integrates into its own genome in a suicidal process . Despite mechanistic differences , both transposons are affected similarly , and larger transposons are particularly vulnerable . We propose that transposons might recruit phylogenetically conserved cellular factors in a new host that protects against self-disruption . Suboptimal conditions in a new environment could generate abnormal , genotoxic transposition reactions , and should be monitored .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"bioengineering",
"genetics",
"biology",
"evolutionary",
"biology",
"engineering"
] |
2014
|
Suicidal Autointegration of Sleeping Beauty and piggyBac Transposons in Eukaryotic Cells
|
Leprosy is remaining prevalent in the poorest areas of the world . Intensive control programmes with multidrug therapy ( MDT ) reduced the number of registered cases in these areas , but transmission of Mycobacterium leprae continues in most endemic countries . Socio-economic circumstances are considered to be a major determinant , but uncertainty exists regarding the association between leprosy and poverty . We assessed the association between different socio-economic factors and the risk of acquiring clinical signs of leprosy . We performed a case-control study in two leprosy endemic districts in northwest Bangladesh . Using interviews with structured questionnaires we compared the socio-economic circumstances of recently diagnosed leprosy patients with a control population from a random cluster sample in the same area . Logistic regression was used to compare cases and controls for their wealth score as calculated with an asset index and other socio-economic factors . The study included 90 patients and 199 controls . A recent period of food shortage and not poverty per se was identified as the only socio-economic factor significantly associated with clinical manifestation of leprosy disease ( OR 1 . 79 ( 1 . 06–3 . 02 ) ; p = 0 . 030 ) . A decreasing trend in leprosy prevalence with an increasing socio-economic status as measured with an asset index is apparent , but not statistically significant ( test for a trend: OR 0 . 85 ( 0 . 71–1 . 02 ) ; p = 0 . 083 ) . Recent food shortage is an important poverty related predictor for the clinical manifestation of leprosy disease . Food shortage is seasonal and poverty related in northwest Bangladesh . Targeted nutritional support for high risk groups should be included in leprosy control programmes in endemic areas to reduce risk of disease .
Leprosy is known as a disease of poverty . Only in the poorest areas of the world the infectious disease caused by Mycobacterium leprae is still endemic . A causal relationship between poverty and leprosy is difficult to demonstrate , and uncertainty exists about how leprosy and poverty are associated [1] , [2] . Bangladesh is one of the countries where the disease is still endemic . Despite reaching the ‘elimination’ target of less than one registered case per 10 , 000 inhabitants for the whole country in 1998 , the prevalence is still above target in some of the poorest areas of Bangladesh [3] , [4] . In the poverty stricken northwest part of the country , where The Leprosy Mission Bangladesh is operating a leprosy control programme , the new case detection rate was still 1 . 25 per 10 , 000 inhabitants in 2008 . To generate more knowledge about risk factors for leprosy and to assess the effect of new interventions , a research project was initiated in northwest Bangladesh in 2001: the COLEP study , a prospective ( sero- ) epidemiological study on contact transmission and chemoprophylaxis in leprosy [5] . The first results of the study indicated that prophylactic treatment with rifampicin is a promising way to prevent leprosy in contacts of patients [6] . Physical distance to a patient and the severity of the disease ( leprosy classification ) were identified as risk factors associated with transmission of Mycobacterium leprae to contacts of a patient . Furthermore , the host characteristics “blood relationship to the patient” and “age” were identified as risk factors for the development of clinically apparent disease , while a previous vaccination with BCG had a preventive effect [7] . These findings indicate that innate and acquired immunity affects the development of clinical signs of leprosy . Clinical disease occurs most probably in only 1–5% of persons infected with Mycobacterium leprae , after an incubation period of several years . The objective of this study , which is part of the COLEP project , was to assess the association between poverty and leprosy more closely , by measuring the effects of different socio-economic factors on acquiring clinical signs of leprosy disease .
A case-control study was carried out in August 2009 in the districts of Nilphamari and Rangpur in northwest Bangladesh . This large ( 3951 km2 ) - mainly rural - area has app . 4 . 5 million inhabitants and is one of the poorest parts of Bangladesh [8] , [9] . The first 110 new leprosy patients registered in 2009 in the study area were selected as cases . These patients were diagnosed by The Leprosy Mission Bangladesh or government facilities according to the national guidelines [10] . Only one patient per household was interviewed to avoid bias due to clustering . From the initially selected group , 10 people could not been reached , while one was excluded because he was living in the same household as another selected patient . Controls without leprosy were randomly selected from a referent group , representative for the general population in the area . This group was selected at the start of the COLEP study in 2002 by a multi-cluster sampling procedure [11] . Twenty clusters of 1000 people each were randomly selected from the 13 sub-districts in this area . In each of the sub-districts one to three clusters were allocated proportional to the population size . Within the sub-districts first unions and thereafter sub-unions were selected randomly by computerized sampling . In each of the thus created clusters , everyone willing to participate and available on the day of registration was included . Registration started at the northern border of the selected village or urban ward and continued until 1000 people were included in the cluster . For this study , 15 people were randomly selected from each of the 20 clusters by computerized sampling . The 15 selected candidates of each cluster were numbered one to fifteen . Interviewers started to contact the first person and continued following the numbering until 10 people were interviewed or everyone was contacted . Controls were excluded when they were ever diagnosed as leprosy patient or if they came from the same household as another participant in the study . Research staff of The Leprosy Mission Bangladesh carried out home visits to conduct interviews with pre-tested structured questionnaires . Besides questions on personal data and some details about their disease ( for patients only ) , participants were asked about their living circumstances and economic situation . They were asked about ownership of assets , including housing , drink water supply , sanitary facilities , livestock and land , while they were also questioned about educational level , job status , monthly household income , seasonal income variations , changes in economic and living situation due to the disease leprosy as well as over the last three years in general , and periods of food shortage in the previous year and ever in life . Food shortage was defined as a period in which a family had to reduce the number of meals a day or had to reduce the intake of foods other than rice , like vegetables , fruits , meat or fish . Data from the questionnaires were entered into an Access database . After data cleaning , analysis was performed using the statistical package STATA version 10 . 0 . Socio-economic status of the participants was estimated by an asset index . Factor analysis , principal components factor , as described by Filmer and Pritchett was used to construct an asset index to assign a wealth score to all participants [12] . Data on ownership of different assets in their household was used to calculate a wealth score by weighing the response for each asset of their household by the coefficient of the first factor as determined by application of the factor analysis , and summing the results ( Table 1 ) . Data regarding possession of a car , rickshaw , animal cart , and drink-water supply were not correlated with the wealth scores as calculated and therefore excluded from the final model . The control group was assigned to five wealth quintiles according to their final score . Cases were assigned to these quintiles according to the threshold values set by the control group . Logistic regression was used to compare cases and controls for the wealth score quintile and the other factors measuring aspects of socio-economic situation: income level , educational level of the highest educated person in the household , household size , crowding ( defined for this study as more than three people per sleeping room on average ) , food shortage ever and a period of food shortage in the last year . Univariate and multivariate logistic regression with a backwards elimination procedure was used to assess the association between these factors as well as the potential confounding factors age and sex . All participants received verbal information about the study and were asked to sign a consent form . Ethical approval for this study was obtained from the Bangladesh Medical Research Council ( under reference: BMRC/NREC/2007-2010/2107 ) .
Initially 99 patients ( cases ) and 199 controls were included in the study population . A deterioration of socio-economic or living condition due to the disease was mentioned by 9 ( 8 . 9% ) of the cases . All these patients had severe forms of leprosy; 6 had grade II disabilities , while the other 3 had the more severe MB form of leprosy . Because the objective of this study was to assess the socio-economic condition as a risk factor for developing clinical signs of leprosy disease , it was important to establish the situation around the time the disease became apparent . We therefore excluded for further analysis the 9 cases in which the economic situation had changed due to the disease , to avoid confusion about cause and effect . Of the 90 patients included for analysis , the sex ratio ( M/F ) of the was 1 . 2; 21 . 1% had the multibacillary ( MB ) form of the disease , while 6 . 6% was diagnosed with a grade II disability , according to the WHO classification ( Table 2 ) . The child rate ( <15 years of age ) was 15 . 6% . At the time of the interview , 58 . 9% of the cases were still on multidrug therapy ( MDT ) , while the other 41 . 1% had just completed their therapy and were released from treatment . Both the case and control populations were distributed randomly throughout the study area . The control group was representative for the general population in the area with respect to the household characteristics religion , household composition , educational level , and living area ( urban/rural ) , as compared to the national statistics , but males in the working age ( 20–39 years ) were slightly underrepresented in the control group [8] , [9] . The prevalence of leprosy decreased with an increased level of economic status , measured by the wealth score quintile ( test for a trend: OR 0 . 85 ( 0 . 71–1 . 02 ) ; p = 0·083 , Table 3 ) . Uni- and multivariate logistic regression analysis revealed only a statistically significant association of the socio-economic factor “a self reported period of food shortage in the last year” with leprosy disease ( OR 1 . 79 ( 1 . 06–3 . 02 ) ; p = 0 . 030 , Table 3 ) . None of the other socio-economic factors were associated with leprosy disease .
A recent period of food shortage and not poverty per se was identified as the only socio-economic risk factor significantly associated with clinical manifestation of leprosy disease in northwest Bangladesh . A decreasing trend in leprosy prevalence with an increasing socio-economic status as measured with an asset index is apparent , but not statistically significant . The strength of this case control study is that it takes into account recently diagnosed leprosy cases , while patients who reported changes in economic or living situation due to their disease were excluded . In this way the actual situation at the time of diagnosis could be measured , making it possible to draw conclusions about the association of leprosy and socio-economic situation as risk factor for acquiring clinical signs of leprosy disease . A limitation of the study is the use of self-reported data on income , educational level and food shortage as measured by a questionnaire , which is by definition subjective . The effect of this form of bias was reduced by asking cases and controls the same questions . Furthermore also an asset index as proxy to measure wealth was constructed , which is a more objective measure for socio-economic status of the household . Although objective , a limitation of the use of a wealth index is that the score of the index depends highly on the set of assets used [13] . Since the asset index used in the USAID sponsored Demographic and Health Survey , carried out in 84 developing countries , has a proven valuable for public health purposes we used a set of assets based the local version of the Demographic and health Survey for Bangladesh [14] , [15] . Another limitation of this method is that the index is relative and based on the assets of others in the group . The whole assessed group is divided into five equal quintiles based on their wealth score . Since the majority of people are very poor in the study area in northwest Bangladesh , people assigned to the higher quintiles have more assets and are somewhat better off than the households included in lower quintiles , but can not be considered as rich by any means in this poverty stricken area . It is likely that most people who reported “food shortage in the last year” in our study observed shortage of food in the yearly period of seasonal income shortage in rural Bangladesh which lasts from the end of September until November , just after the rainy season and before the main rice harvest in November/December . In this period there are few work opportunities , low household food stocks , and increased rice prices . The yearly period of food shortage roughly coincides with the start of symptoms of leprosy in the selected cases , as 70% of the patients reported start of their symptoms less than six months before they were registered ( between seven to twelve months before the interview , between September to December 2008 ) . In poor rural communities in Bangladesh seasonal income changes are common . In our study the reported income changed from a monthly average of 3000 BDT ( 43 US$ ) to 9000 BDT ( 130 US$ ) per household . Seasonal income changes are closely related to daily expenditure on food and influences the nutritional status of the people in rural Bangladesh [16] . In rural Bangladesh , chronic energy deficiency ( CED ) based on body mass index ( BMI ) is high ( between 60–70% ) in all age and sex groups , while seasonal differences in energy intake are substantial in all age and sex group as well [17] . The amount of rice consumed is quite stable , but expenditure on high nutritious and more expensive food decreases in months of low income in rural communities , likely causing micronutrient deficiencies . Studies in Bangladesh revealed an association between the proportion of expenditure on non-rice food and maternal underweight as well as child stunting [18] , and an association between a low BMI and increased mortality in adults [19] . The hypothesis that seasonal food deficiencies might be associated with leprosy is strengthened by the seasonal pattern in number of new leprosy cases registered per month over the last nine years ( 2002–2010 ) in the districts where the study was carried out . The number of newly registered cases is rising from February , about four months after the start of the seasonal low-income period , and reaches a maximum in June at the beginning of the monsoon period in Bangladesh and six months after the end of the low-income period ( figure 1 ) . However alternative explanations are possible . A study in a leprosy endemic area in India showed a strong seasonal pattern in Mycobacterium leprae bacteria detectable in the general population by nasal PCR and salivary ML-IgA positivity . The rates of PCR positive nasal swabs were high in the period immediately after the monsoon rains from July to November , while salivary ML-IgA titres were high in November at the end of the wet period . This indicates a seasonal pattern in exposure to Mycobacterium leprae [20] . “Food shortage in the last year” as assessed in this study represents a recent ( short ) period of poverty with limited expenditure on high nutritious food , likely causing nutritional deficit . In contrast , an asset index as a proxy to measure wealth gives an indication of the long-term economic status of a household , since people tend not to sell their assets in seasonal short periods of low income , but only in longer term poverty [12] , [21] . Although the general population sample ( referent group ) of the COLEP trial was selected almost seven years before this study , a selection of this group is still suitable to use as control group . Only three of the selected leprosy cases were born less than seven years before the start of the study , from which you can conclude that leprosy below this age is rare . Furthermore 80% of the selected people of the control group participated in the study , which indicates that the population in this area is not very mobile . However , due to the original selection method used for this referent group , men in the working age are underrepresented , since many of them were absent from their house at the time of registration . Therefore age and sex were included as potential confounders in the analysis . The actual association between poverty and leprosy might be stronger than indicated by this study , because only registered cases were included in the study . Registered cases receive leprosy treatment and have access to health services . Although the area has a long running active disease control programme in which treatment is given free of charge , there are still people who have no access to these services . In a study carried out in 2002 in northwest Bangladesh , the population prevalence of leprosy was found to be six times higher than the registered prevalence [11] . The fact that 11% of the original selected cases in our study had grade II disability , indicating late detection of the disease , suggests that there may be undetected leprosy cases in the area . Poverty is one of the reasons for limited access to leprosy care . Stigma , although less common due to the active control and health education activities in the area , and cultural defined limited access to health care for women might be of importance as well [22] . An association between food shortage and leprosy was also observed in Brazil [1] . However , in Brazil a period of food shortage at any time in life , as indicator of poverty in general , was found associated with leprosy , while in our study only a recent period of food shortage was associated with the disease . Although a higher percentage of leprosy cases also reported food shortage at any time of life in Bangladesh , this association was not statistically significant . Different case definitions of food shortage or differences in social norms regarding nutritional requirements between the countries could be an explanation for this difference . Food shortage however , may also be a less strong indicator of poverty in general in Bangladesh than in Brazil , since the percentage of people who reported food shortage ever was much higher in Bangladesh ( 66 . 7% of the cases and 61 . 8% of the controls ) than in Brazil ( 28% of the cases and 19% of the controls ) . Nutritional status is known to influence the development of other infectious diseases such as respiratory infections , infectious diarrhoea , measles and malaria . These diseases are observed more commonly in malnourished children . Malnutrition affects the immune system negatively , causing infected individuals to be more vulnerable for developing a clinically apparent infection [23] . In tuberculosis , which has similarities to leprosy since it is also caused by a mycobacterium , nutritional deficit has been identified as an important risk factor in the development of clinical symptoms of disease . This is based on historical reports of outbreaks during famines and wars , and on animal studies in which cell mediated immunity was diminished in malnourished guinea pigs . Cell mediated immunity , which is affected by both protein energy malnutrition and micronutrient deficiencies , plays an important role in host defense against tuberculosis and leprosy [24] . A recent period of food shortage as identified in our study as most important poverty-related factor associated with leprosy , very likely has reduced the cell mediated immunity of individuals incubating Mycobacterium leprae , causing the development of clinical leprosy disease . Targeted nutritional support to high risk groups should therefore be included in leprosy control programmes in endemic areas to reduce risk of disease . It would be useful to give contacts of leprosy patients , who are at high risk of developing leprosy themselves , dietary advices to prevent malnutrition . Because food shortage is seasonal and poverty related in northwest Bangladesh , extra attention and support should be given to the poorest families with leprosy patients . It is important to prevent malnutrition in these families to prevent clinical leprosy among contacts of patients .
|
Although intensive control programs reduced the prevalence of leprosy worldwide , new cases of this infectious disease are still detected in several of the poorest areas of the world . Therefore the disease is known as a disease of poverty . To be able to control the disease it is important to know which aspects of poverty play a role in transmission and acquiring clinical signs of disease . In this study socio-economic circumstances of recently diagnosed leprosy patients were compared with those of a control population in the poverty stricken northwest area of Bangladesh where leprosy is common . A recent period of food shortage was the only socio-economic factor that was found related to leprosy disease in this study and not poverty as such . Food shortage is seasonal and poverty related in northwest Bangladesh , while malnutrition is known to lower immunity and make people more vulnerable to infectious diseases . Therefore it was concluded that malnutrition as an aspect of poverty played an important role in the development of the clinical signs of leprosy . We therefore recommend that nutritional support for high risk groups should be included in leprosy control programmes to reduce risk of disease in areas where leprosy is common .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"socioeconomic",
"aspects",
"of",
"health",
"non-clinical",
"medicine",
"nutrition",
"neglected",
"tropical",
"diseases",
"leprosy",
"infectious",
"disease",
"control"
] |
2011
|
Recent Food Shortage Is Associated with Leprosy Disease in Bangladesh: A Case-Control Study
|
The extracellular matrix plays a critical role in orchestrating the events necessary for wound healing , muscle repair , morphogenesis , new blood vessel growth , and cancer invasion . In this study , we investigate the influence of extracellular matrix topography on the coordination of multi-cellular interactions in the context of angiogenesis . To do this , we validate our spatio-temporal mathematical model of angiogenesis against empirical data , and within this framework , we vary the density of the matrix fibers to simulate different tissue environments and to explore the possibility of manipulating the extracellular matrix to achieve pro- and anti-angiogenic effects . The model predicts specific ranges of matrix fiber densities that maximize sprout extension speed , induce branching , or interrupt normal angiogenesis , which are independently confirmed by experiment . We then explore matrix fiber alignment as a key factor contributing to peak sprout velocities and in mediating cell shape and orientation . We also quantify the effects of proteolytic matrix degradation by the tip cell on sprout velocity and demonstrate that degradation promotes sprout growth at high matrix densities , but has an inhibitory effect at lower densities . Our results are discussed in the context of ECM targeted pro- and anti-angiogenic therapies that can be tested empirically .
The physical properties of the ECM , such as density , heterogeneity , and stiffness , that affect cell behavior is also an area of current investigation . Matrigel , a popular gelatinous protein substrate for in vitro experiments of angiogenesis , is largely composed of collagen and laminin and contains growth factors , all of which provide an environment conducive to cell survival . In experiments of endothelial cells on Matrigel , increasing the stiffness of the gel or disrupting the organization of the cellular cytoskeleton , inhibits the formation of vascular cell networks [10] , [11] . Cells respond to alterations in the mechanical properties of the ECM , for example , by upregulating their focal adhesions on stiffer substrates [12] . For anchorage-dependent cells , including endothelial cells , increasing the stiffness of the ECM therefore results in increased cell traction and slower migration speeds [12] . Measurements of Matrigel stiffness as a function of density show a positive relationship between these two mechanical properties [13] . That is , as density increases , so does matrix stiffness . In light of these two findings , it is not surprising that this experimental study also shows slower cell migration speeds as matrix density increases [13] . Moreover , matrices with higher fiber density transfer less strain to the cell [14] and experiments of endothelial cells cultured on collagen gels demonstrate that directional sprouting , called branching , is induced by collagen matrix tension [15] . Thus , via integrin receptors , the mechanical properties of the ECM influence cell-matrix interactions and modulate cell shape , cell migration speed , and the formation of vascular networks . Understanding how individual cells interpret biochemical and mechanical signals from the ECM is only a part of the whole picture . Morphogenic processes also require multicellular coordination . In addition to the guidance cues cells receive from the ECM , they also receive signals from each other . During new vessel growth , cells adhere to each other through cell-cell junctions , called cadherins , and in order to migrate , cells must coordinate integrin mediated focal adhesions with these cell-cell bonds . This process is referred to as collective or cluster migration [16] . During collective migration , cell clusters often organize as two-dimensional sheets [16] . Cells also have the ability to condition the ECM for invasion by producing proteolytic enzymes that degrade specific ECM proteins [17] . In addition , cells can synthesize ECM components , such as collagen and fibronectin [11] , [18] , and can further reorganize the ECM by the forces they exert on it during migration [10] , [11] , [14] . Collagen fibrils align in response to mechanical loading and cells reorient in the direction of the applied load [14] . Tractional forces exerted by vascular endothelial cells on Matrigel cause cords or tracks of aligned fibers to form promoting cell elongation and motility [11] . As more experimental data are amassed , the ECM is emerging as the vital component to morphogenic processes . In this work , we extend our cellular model of angiogenesis [19] and validate it against empirical measurements of sprout extension speeds . We then use our model to investigate the effect of ECM topography on vascular morphogenesis and focus on mechanisms controlling cell shape and orientation , sprout extension speeds , and sprout morphology . We show the dependence of sprout extension speed and morphology on matrix density , fiber network connectedness , and fiber orientation . Notably , we observe that varying matrix fiber density affects the likelihood of capillary sprout branching . The model predicts an optimal density for capillary network formation and suggests matrix heterogeneity as a mechanism for sprout branching . We also identify unique ranges of matrix density that promote sprout extension or that interrupt normal angiogenesis , and show that maximal sprout extension speeds are achieved within a density range similar to the density of collagen found in the cornea . Finally , we quantify the effects of proteolytic matrix degradation by the tip cell on sprout velocity and demonstrate that degradation promotes sprout growth at high densities , but has an inhibitory effect at lower densities . Based on these findings , we suggest and discuss several ECM targeted pro- and anti-angiogenesis therapies that can be tested empirically .
We previously published a cell-based model of tumor-induced angiogenesis that captures endothelial cell migration , growth , and division at the level of individual cells [19] . That model also describes key cell-cell and cell-matrix interactions , including intercellular adhesion , cellular adhesion to matrix components , and chemotaxis to simulate the early events in new capillary sprout formation . In the present study , we extend that model to incorporate additional mechanisms for cellular motility and sprout extension , and use vascular morphogenesis as a framework to study how ECM topography influences intercellular and cell-matrix interactions . The model is two-dimensional . It uses a lattice-based cellular Potts model describing individual cellular interactions coupled with a partial differential equation to describe the spatio-temporal dynamics of vascular endothelial growth factor . At every time step , the discrete and continuous models feedback on each other , and describe the time evolution of the extravascular tissue space and the developing sprout . The cellular Potts model evolves by the Metropolis algorithm: lattice updates are accepted probabilistically to reduce the total energy of the system in time . The probability of accepting a lattice update is given bywhere is the change in total energy of the system as a result of the update , is the Boltzmann constant , and is the effective temperature corresponding to the amplitude of cell membrane fluctuations . A higher temperature corresponds to larger cell membrane fluctuation amplitudes . The energy , , includes a term describing cell-cell and cell-matrix adhesion , a constraint controlling cellular growth , an effective chemotaxis potential , and a continuity constraint . Mathematically , total energy is given by: ( 1 ) In the first term of Eq . 1 , represents the binding energy between model constituents . For example , describes the relative strength of cell-cell adhesion that occurs via transmembrane cadherin proteins . Similarly , is a measure of the binding affinity between an endothelial cell and a matrix fiber through cell surface integrin receptors . Each endothelial cell is associated with a unique identifying number , . is the Kronecker delta function and ensures that adhesive energy only accrues at cell surfaces . The second term in Eq . 1 describes the energy expenditure required for cell growth and deformation . Membrane elasticity is described by , denotes cell current volume , and is a specified target volume . For proliferating cells , the target volume is double the initial volume . This growth constraint delivers a penalty to total energy for any deviation from the target volume . In the third term , the parameter is the effective chemical potential and influences the strength of chemotaxis relative to other parameters in the model . This chemotaxis potential varies depending on cell phenotype ( discussed below ) and is proportional to the local VEGF gradient , , where denotes the concentration of VEGF . Cells must simultaneously integrate multiple external stimuli , namely intercellular adhesion , chemotactic incentives , and adherence to extracellular matrix fibers . To do so , endothelial cells deform their shape and dynamically regulate adhesive bonds . In the model , however , it is possible that collectively these external stimuli may cause a cell to be pulled or split in two . To prevent non-biological fragmentation of cells , we introduce a continuity constraint that preserves the physical integrity of each individual cell . This constraint expresses that it is energetically expensive to compromise the physical integrity of a cell and is incorporated into the equation for total energy ( Eq . 1 ) in the last term , where is a continuity constraint that represents the effects of the cytoskeletal matrix of a cell . is the current size of the endothelial cell with identifying number , and is a breadth first search count of the number of continuous lattice sites occupied by that endothelial cell . Thus , signals that the physical integrity of the cell has been compromised and a penalty to total energy is incurred . Cooperatively , the continuity constraint and the volume constraint implicitly describe the interactions holding the cell together . The amount of VEGF available at the right hand boundary of the domain is estimated by assuming that in response to a hypoxic environment , quiescent tumor cells secrete a constant amount of VEGF and that VEGF decays at a constant rate . It is reasonable to assume that the concentration of VEGF within the tumor has reached a steady state and therefore that a constant amount of VEGF , denoted , is available at the boundary of the tumor . We use constant boundary conditions for the left ( ) and right boundaries and periodic boundary conditions in the y-direction . A gradient of VEGF is established as VEGF diffuses through the stroma with constant diffusivity coefficient , decays at a constant rate , and is bound by endothelial cells , . A complete description of the biochemical derivation of the function for endothelial cell binding and uptake of VEGF ( ) has been previously published [19] . For more direct comparison to other mathematical models of angiogenesis models and to isolate the effects of ECM topology on vessel morphology , we assume that the diffusion coefficient for VEGF in tissue is constant . This is a simplification , however , because the ECM is not homogeneous and VEGF can be bound to and stored in the ECM . Realistically , the diffusion coefficient ( ) for VEGF in the ECM depends on both space and time . We address the implications of this assumption in the Discussion . Under these assumptions , the concentration profile of VEGF satisfies a partial differential equation of the form: ( 2 ) The inset in Figure 1A provides an illustration of the 166 µm×106 µm domain geometry . We initialize the simulation by establishing the steady state solution to Eq . 2 . The activation and aggregation of endothelial cells , and subsequent breakdown of basement membrane in response to VEGF [20] is a pre-condition ( boundary condition ) to the simulation . The breakdown of basement membrane allows endothelial cells to enter the extravascular space through a new vessel opening . Our simulation starts with a single activated endothelial cell ∼10 µm in diameter that has budded from the parent vessel located adjacent to the left hand boundary [20] . We use 10 µm only as an initial estimate of endothelial cell size [21] , [22] . Once the simulation begins , the cells immediately deform in shape and elongate . During the simulation , the VEGF field is updated iteratively with cell uptake information , for example as shown in Figure 1B , C . VEGF data is processed by the cells at the cell membrane and incorporated into the model through the chemotaxis term in Eq . 1 . From the parent blood vessel , endothelial cells ( red ) migrate into the domain in response to VEGF that is supplied from a tumor located adjacent to the right hand boundary . The space between represents the stroma and is composed of extracellular matrix fibers ( green ) and interstitial fluid ( blue ) . The physical meanings of all symbols and their parameter values are summarized in Table 1 . To more accurately capture the cell-cell and cell-matrix interactions that occur during morphogenesis , we implement several additional features to this model . One improvement is the implementation of stalk cell chemotaxis . Stalk cells are not inert , but actively respond to chemotactic signals [23] . As a consequence , cells now migrate as a collective body , a phenomenon called collective or cohort migration [24] . This modification , however , also makes it possible for individual cells , as well as the entire sprout body , to migrate away from the parent vessel , making it necessary to consider cell recruitment from the parent vessel . Cell recruitment is another added feature . During the early stages of angiogenesis , cells are recruited from the parent vessel to facilitate sprout extension [20] , [25] . Kearney et al . [26] measured the number and location of cell divisions that occur over 3 . 6 hours in in vitro vessels 8 days old ( a detailed description of these experiments is provided in our discussion of model validation ) . In these experiments , the sprout field is defined as the area of the parent vessel wall that ultimately gives rise to the new sprout and the sprout itself . The sprout field is further broken down into regions based on distance from the parent vessel and these regions are classified as distal , proximal , and nascent . The authors report that 90% of all cell divisions occur in the parent vessel and the remaining 10% occur at or near the base of the sprout in the nascent area of the sprout field . On average , total proliferation accounts for approximately 5 new cells in 3 . 6 hours , or 20 cells in 14 hours . This data suggests that there is significant and sufficient proliferation in the primary vessel to account for and facilitate initial sprout extension . This data does not suggest that proliferation in other areas of the sprout field does not occur at other times . In fact , it has been established that a new sprout can migrate only a finite distance into the stroma without proliferation and that proliferation is necessary for continued sprout extension [25] . We model sprout extension through a cell-cell adhesion dependent recruitment of additional endothelial cells from the parent vessel . As an endothelial cell at the base of the sprout moves into the stroma , cell-cell adhesion pulls a cell from the parent vessel along with it . In practice , a new cell is added to the base of the sprout when and where the previous cell detaches from the parent vessel wall ( left boundary of the simulation domain ) . We assume , based on the data presented in [26] , that there is sufficient proliferation in the parent vessel to provide the additional cells required for initial sprout extension while maintaining the physical integrity of the parent vessel . As in our previous model , once a cell senses a threshold concentration of VEGF , given by , it becomes activated . We recognize that cells have distinct phenotypes that dictate their predominate behavior . Thus , we distinguish between tip cells , cells that are proliferating , and non-proliferating but migrating stalk cells . Tip cells are functionally specialized cells that concentrate their internal cellular machinery to promote motility [23] . Tip cells are highly migratory pathfinding cells and do not proliferate [25] , [26] . To model the highly motile nature of the tip cell , we assign it the highest chemotactic coefficient , . The remainder of the cells are designated as stalk cells and use adhesive binding to and release from the matrix fibers for support and to facilitate cohort migration . Stalk cells also sense chemical gradients but are not highly motile phenotypes . Thus , the stalk cells in the model are assigned a lower , that is weaker , chemotactic coefficient than the specialized tip cell . Proliferating cells are located behind the sprout tip [23] , [26] and increase in size as they move through an 18 hour cell cycle clock in preparation for cell division [27] . Cells that are proliferating can still migrate [26]; it is only during the final stage of the cell cycle that endothelial cells stop moving and round up for mitosis ( personal communication with C . Little ) . As we assume that the presence of VEGF increases cell survivability , we do not model endothelial cell apoptosis . As described in our previous work [19] , we model the mesh-like anisotropic structure of the extracellular matrix by randomly distributing 1 . 1 µm thick bundles of individual collagen fibrils at random discrete orientations between −90 and 90 degrees . Unless otherwise stated , model matrix fibers comprise approximately 40% of the total stroma and the distribution of the ECM is heterogeneous , with regions of varying densities as can be seen in Figure 1A and Figure 7D . The cells move on top of the 2D ECM model and interact with the matrix fibers at the cell membrane through the adhesion term in Eq . 1 . To relate the density ( ) of this model fibrillar matrix to physiological values , we measure matrix fiber density as the ratio of the interstitium occupied by matrix molecules to total tissue space , , and compare it to measured values of the volume fraction of collagen fibers in healthy tissues [28] . In order to isolate and control the effects of the matrix topology on cellular behavior and sprout morphology we look at a static ECM , that is we do not model ECM rearrangement or dynamic matrix fiber cross-linking and stiffness . We do , however , consider endothelial cell matrix degradation in a series of studies presented in Results . No single model has been proposed that incorporates every aspect of all processes involved in sprouting angiogenesis , nor is this level of complexity necessary for a model to be useful or predictive . It is not our intention to include every bio-chemical or mechanical dynamic at play during angiogenesis . We develop this two-dimensional cell-based model as a step towards elucidating cellular level dynamics fundamental to angiogenesis , including cell growth and migration , and cell-cell and cell-matrix interactions . Consequently , we do not incorporate processes or dynamics at the intracellular level . For example , we describe endothelial cell binding of VEGF to determine cell activation and to capture local variations in VEGF gradients , but neglect intracellular molecular pathways signaled downstream of the receptor-ligand complex . Moreover , our focus is on early angiogenic events and therefore we also do not consider the effects of blood flow on remodeling of mature vascular beds . Numerical studies of flow-induced vascular remodeling have been given attention in McDougall et al . [29] , and Pries and Secomb [30] , [31] . As is the case in many other simulations of biological systems , when we do not have direct experimental measurements for all of the parameters , choosing these parameter values is not trivial . A list of values and references for our model parameters is provided in Table 1 . A parameter is derived from experimental data whenever possible , otherwise it is estimated and denoted ‘est’ . Fortunately , a sensitivity analysis ( discussed later ) shows that the dynamics of our model are quite robust to substantial variations in some parameters and tells us exactly which parameters are most critical . We can then choose from a range of parameter values that exhibits the general class of behavior consistent with experimental observations . See Table 1 for these parameter ranges and Table 3 for the effect of parameter perturbations , as well as , supplemental Figures S1 and S2 for examples of cellular behavior under different parameter sets . In the cellular Potts model , the relative value , not the absolute value , of the parameters corresponds to available physiological measurements and gives rise to a cell behavior observed experimentally . For example , the Young's modulus for human vascular endothelial cells is estimated at 2 . 01*105 Pa [32] . The Young's modulus of a collagen fiber in aqueous conditions is between 0 . 2–0 . 8 GPa [33] . However , the modulus of a collagen gel network is much lower and is measured at 7 . 5 Pa [34] . Although interstitial fluid compressibility ( water ) is estimated to be 2 . 2 GPa [35] , indicating it's hard to compress under uniform pressure , it deforms easily , that is , the shear modulus is low and is measured at 10−6 Pa [36] . The qualitative parameters corresponding to these quantitative measurements are where . Thus , the elastic modulus of endothelial cells>matrix fibers>interstitial fluid ( 0 . 2 MPa>7 . 5 Pa>10−6 Pa ) and is reflected in the relative values of the corresponding parameters , , and . In a similar manner , the coupling parameters , , describe the relative adhesion strengths among endothelial cells , matrix fibers , and interstitial fluid . For instance , choosing reflects that fact that endothelial cells have a higher binding affinity to each other , via cadherin receptors and gap junctions for example , than they do to matrix fibers [37] , [38] . The chemotactic potential , , is chosen so that its contribution to the change in total energy is the same order of magnitude as the contribution to total energy from adhesion or growth . The difference between the concentration of VEGF at two adjacent lattice sites is on the order of 10−4 . Therefore , to balance adhesion and growth , must be on the order of 106 . We calibrate this parameter to maximize sprout extension speeds . Similarly , the parameter for continuity , , is chosen so that cells will not dissociate . This is achieved by setting greater than the collective contribution to total energy from the other terms . By equating the time it takes an endothelial cell to divide during the simulation with the endothelial cell cycle duration of 18 hours , we convert Monte Carlo steps to real time units . In the simulations reported in this paper , 1 Monte Carlo step is equivalent to 1 minute . Since this model has several enhancements over the previous model [19] , there are a different number of parameters , which necessitates recalibration of all the parameters . Therefore , some parameters take on different values .
The canonical benchmark for validating models of tumor-induced angiogenesis is the rabbit cornea assay [39] , [40] . In this in vivo experimental model , tumor implants are placed in a corneal pocket approximately 1–2 mm from the limbus . New vessel growth is measured with an ocular micrometer at 10× , which has a measurement error of ±0 . 1 mm or 100 µm . Initially , growth is linear and sprout extension speeds are estimated at a rate of 0 . 5 mm/day , or 20 . 8±4 . 2 µm/hr . Sprouts then progress at average speeds estimated to be between 0 . 25–0 . 50 mm/day , or 10 . 4–20 . 8±4 . 2 µm/hr . More recent measurements of sprout extension speeds during angiogenesis are reported in Kearney et al . [26] . In this study , embryonic stem cells containing an enhanced green fluorescent protein are differentiated in vitro to form primitive vessels . Day 8 cell cultures are imaged within an ∼160 µm2 area at 1 minute intervals for 10 hours and show sprouting angiogenesis over this period . The average extension speed for newly formed sprouts is 14 µm/hr and ranges from 5 to 27 µm/hr . For cell survival , growth factor is present and is qualitatively characterized as providing a diffuse , or shallow , gradient . No quantitative data pertaining to growth factor gradients or the effect of chemotaxis during vessel growth are reported [26] . We use the above experimental models and reported extension speeds as a close approximation to our model of in vivo angiogenesis for quantitative comparison and validation . We simulate new sprout formation originating from a parent vessel in the presence of a diffusible VEGF field , which creates a shallow VEGF gradient . We measure average extension speeds over a 14 hour period in a domain 100 µm by 160 µm . As was done in Kearney et al . [26] , we calculate average sprout velocities as total sprout tip displacement in time and measure this displacement as the distance from the base of the new sprout to the sprout tip . Figure 1A shows average sprout extension speed over time for our simulated sprouts . Reported speeds are an average of at least 10 independent simulations using the same initial VEGF profile and parameter set as given in Table 1 . Error bars represent the standard error from the mean . The average extension speeds of our simulated sprouts are within the ranges of average sprout speeds measured by both Kearney et al . [26] and Gimbrone et al . [39] . Table 2 summarizes various morphological measurements for the simulated sprouts . It shows that the average velocity , thickness , and cell size of the simulated sprouts compare favorably to relevant experimental measurements . Sprout velocity is given at 10 hours for direct comparison to [26] and averaged over 14 hours . Sprout thicknesses and cell size are within normal physiological ranges . There are many different cell shapes and sizes and vessel morphologies , however , that can be obtained in vivo and in vitro given different environmental factors ( VEGF profile , ECM topology and stiffness , inhibitory factors , other cell types , etc . ) . In this manuscript , we investigate several of these dependencies and as we discuss below specific model parameters can be tuned to reproduce different cellular interactions and environments . Figure 1A indicates that average sprout extension speed changes as a function of time . Within the first two hours , speeds average ∼30 µm/hr and the new sprout consists of only 1–2 endothelial cells . At two hours , sprouts contain an average of 3 cells , and at 4 hours , there are a total of 5–6 cells . Over time , as more cells are added to the developing sprout , cell-cell adhesion and cellular adhesion to the extracelluar matrix slow the sprout extension speed . The inset in Figure 1A shows the geometry of the computational domain and simulated sprout development at 7 . 8 hours . As shown , simulated sprouts are approximately one cell diameter wide , which compares quantitatively well to reported VEGF induced vessel diameters [41] , [42] . Here and in all simulation snapshots , tip cells are identified with a ‘T’ . In moving multicellular clusters , rear retraction is a collective process that involves many cells simultaneously [16] . A natural result of the cell-based model is that cells exhibit rear retraction , which refers to the ability of a cell to release its trailing adhesive bonds with the extracellular matrix during migration . Collective migration , another characteristic dynamic observed during sprout growth , is also evident during the simulations ( see videos ) . The VEGF concentration profile in picograms ( pg ) at 7 . 8 hours is given in Figure 1B . Higher concentrations of VEGF are encountered as the cells approach the tumor . However , because cell uptake of VEGF is small compared to the amount of available VEGF , it is difficult to discern the heterogeneities in the VEGF profile from this figure . Figure 1C is the VEGF gradient profile ( pg ) at 7 . 8 hours and is a better indicator of the changes in local VEGF concentration . This image shows larger gradients in the proximity of the tip cell and along the leading edges of the new sprout . On average , simulated sprouts migrate 160 µm and reach the domain boundary in approximately 15 . 6 hours , before any cells in the sprout complete their cell cycle and proliferate . We do not expect to see proliferation in the new sprout because the simulation duration is less than the 18 hour cell cycle and the cell cycle clock is set to zero for newly recruited cells to simulate the very onset of angiogenesis . In our simulations , sprout extension is facilitated by cell recruitment from the parent vessel . Between 15 and 20 cells are typically recruited , which agrees with the number of cells we estimate would be available for recruitment based on parent vessel cell proliferation reported by Kearney et al . [26] . In those experiments [26] , proliferation in the parent vessel was measured for day 8 sprouts , which likely has cells at various stages in their cell cycles . Proliferation in the new sprout is another mechanism for sprout extension . Thus , we consider the possibility that cells recruited from the parent vessel may be in different stages of their cell cycles by initializing the cell cycle clock of each recruited cell at randomly generated times . We observe no differences in extension speeds , sprout morphology , or the number of cells recruited as a result of the assumption we make for cell cycle initialization ( or random ) . This suggests that , in the model , stalk cell proliferation and cell recruitment from the parent vessel are complementary mechanisms for sprout extension . By adjusting key model parameters , we are able to simulate various morphogenic phenomena . For example , by increasing the chemotactic sensitivity of cells in the sprout stalk and decreasing the parameter controlling cellular adhesion to the matrix , , we are able to capture stalk cell migration and translocation along the side of a developing sprout ( Video S1 ) . This phenomena , where stalk cells weaken their adhesive bonds to the extracellular matrix and instead use cell-cell adhesion to facilitate rapid migration , frequently occurs in embryogenesis ( personal communication with C . Little ) and is described as preferential migration to stretched cells [43] . Compare Video S1 with Figure 1 ( f ) in Szabo et al . 2007 [43] . Figure S1 shows the morphology for one particular set of parameter values corresponding to weaker cell-cell and cell-matrix adhesion and stronger chemotaxis . In this simulation , cells elongate to approximately 40 µm in length , fewer cells are recruited from the parent vessel , and the average extension speed at 14 hours slows to 6 . 8 µm/hr . The length scale is consistent with experimental measurements of endothelial cell elongation [23] , [44] . Figure 5 from Oakley et al . 1997 shows images from experiments using human fibroblasts stained for actin ( e ) and tubulin ( f ) on micro-machined grooved substratum [45] . These experiments demonstrate that cells alter their shape , orientation , and polarity to align with the direction of the grooves ( double-headed arrow ) , exhibiting topographic , or contact , guidance . Figure S2 is a simulation designed to mimic these experiments by isolating the cellular response to topographical guidance on similarly patterned substratum . In this simulation , there is no chemotaxis and no cell-cell contact; cells respond only to topographical cues in the extracellular matrix . Simulated cells alter their shape and orient in the direction of the matrix fibers . Figure S2 bears a striking resemblance to the cell shapes observed in [45] . We are also able to simulate interstitial invasion/migration by a single cell by turning off proliferation and cell recruitment but leaving all other parameters unchanged ( Video S2 ) . This simulation is especially relevant in the context of fibroblast recruitment during wound healing and tumor cell invasion ( e . g . , glioblastoma , the most malignant form of brain cancer [46] ) , where understanding cell-matrix interactions and directed motility are critical mechanisms for highly motile or invasive cell phenotypes . We design a set of numerical experiments allowing us to observe the onset of angiogenesis in extravascular environments of varying matrix fiber density . We consider matrix fiber densities given as a fraction of the total interstitial area , . As a measure of matrix orientation equivalency , the total fiber orientation in both the x and the y direction is calculated as we increased the matrix density . The total x and total y fiber orientation do not vary with changes in total matrix density . Besides varying the matrix density , all other parameters are held fixed . All simulations last the same duration corresponding to approximately 14 hours . The average rate at which the sprout grows and migrates , or its average extension speed , is calculated as the total tip cell displacement in time . Average extension speeds in microns per hour ( µm/hr ) versus matrix fiber density ( ) are graphed in Figure 2A at various times ( 2 , 5 , 10 , 14 hours ) during sprout development . We identify qualitative measures to describe and differentiate between various capillary sprout morphologies , such as the thickness of the sprout , its tortuosity , and whether sprout branching or anastomosis occur . Following Kearney et al . , we define a sprout branch as one or more cells that extend , or bud , from the primary sprout body at least 10 µm [26] . We report capillary sprout thickness and the incidence of branching versus the fraction of matrix fibers present in the stroma in Figure 2B . Figure 2 demonstrates that the density of the matrix impacts the average rate at which a capillary sprout extends and the resulting sprout morphology . At very low ratios ( ) , the matrix fibers are sparse , disconnected filaments ( Figure 3A ) . In a study of vasculogenesis using endothelial cells plated on varying densities of collagen or fibronectin , cell attachment , spreading , and tube formation are maximal on dishes of intermediate density , reported to be 100–500 ng/cm2 [47] . Whereas , at matrix densities below 100 ng/cm2 , cells detach from the substrate and lose their viability [47] . Our model predicts a coincident interruption of normal angiogenesis and loss of sprout viability at very low matrix fiber densities ( <0 . 10 ) . Moreover , experimental data shows that matrices with lower fibril density transfer more strain to the cell [14] . We capture the morphological consequences of this relationship by inferring strain rate effects on morphology through changes in matrix density . A simulation of sprout development on a low fiber density matrix can be seen in Figure 3A and shows severe cell elongation at . Compare these cells with those shown in the inset of Figure 1A , which is an identical simulation except for an increase in the ECM density ( ) . This higher density matrix has an effect similar to that of transferring less strain to the cells and consequently the cells are rounder . Additionally , because there are more focal adhesion sites in this denser matrix , cells are able to maintain their cell-cell contacts and develop as a cohesive body . We do not report migration speeds for or because sprouts show developmental defects , that is , cells are severely elongated , detach from each other , do not grow , or do not form a cohesive sprout body . For , the fiber network is highly inhomogeneous , and we know that lower matrix densities transfer a larger amounts of strain to the cells . As a result we see an increase in cell spreading and a thickening of the new sprout as compared to those morphologies seen for ( compare Figures 3A and 3B ) . These values of correspond to the same fraction of collagen present in subcutaneous tissue ( ) and some skeletal muscle ( ) [28] . Figure 2B quantifies the incidence of branching for sprouts developing in different matrix densities . Remarkably , we see a distinct range of densities , 0 . 20–0 . 30 , where new buds develop from the main sprout body and branches begin to form ( see arrow in Figure 3B ) . This observation suggests that a high degree of fiber heterogeneity , which is related to mechanical mechanisms , such as ECM tension transfer to cells , may promote branching . This observation is consistent with reports that collagen matrix tension induces directional sprouting in endothelial cells [15] . Figure 3C shows sprout development on a matrix where . Morphologies that could be interpreted as anastomosis ( loop formation ) are evident and are only seen at this density . Figure 2A shows ( i ) a clear range of matrix density that encourages sprout migration and results in faster average speeds and ( ii ) density ranges that present a physical barrier to migration and inhibit sprout growth and results in slower extension speeds . The peak in the graph at indicates that sprout extension speeds are fastest at intermediate densities between and suggests an optimal matrix density for promoting angiogenesis . For comparison , this range of matrix density is near the physiological fraction of collagen fibers found in the cornea [28] . A possible mechanistic explanation for the existence of a peak extension velocity is that the mechanical properties of the ECM around provide contact guidance cues that complement or are aligned with chemotactic gradients . Referring again to Figure 2A , we see that peak migration speeds are prominent at 2 hours , but are still evident , although to a lesser extent , at 10 and 14 hours . Thus , these results do not depend on time . Our finding that maximum migration speeds depend on matrix density is supported by empirical measurements of endothelial cell migration speeds on various fibronectin concentrations ( 0 . 5 , 1 , 5 , 20 , 40 µg/cm2 ) demonstrating peak migration speeds at intermediate concentrations ( 5 µg/cm2 ) [48] . As matrix density increases , the network of connected fibers is extensive . Higher fiber density translates into greater matrix homogeneity and a loss of strong guidance cues from fiber heterogeneity . Chemotaxis then plays a stronger role in sprout guidance thereby producing linear sprouts ( Figure 3D ) . Consequently , we do not observe any branching at densities above . At a fiber density of , less tension is transferred to the cells . Cells experiencing less tension are rounder . Wider and slower sprouts form at this matrix density ( Figure 3E ) . Above , very high matrix densities actually establish a physical barrier to migration and we see a corresponding reduction in sprout extension speed due to increased focal adhesion contacts . Figure 3F shows complete inhibition of angiogenesis at as cell adherence to matrix fibers dominates chemotactic incentives . Looking at Figures 2B and 3 , average sprout thickness is within the observed physiological range of 1–3 cells wide , but does show a dependency on matrix density . For very low densities ( A ) , the cells form a very thin , unstable sprout ( <1 cell wide ) . For intermediate densities ( B–D and Figure 1A ) , sprouts are stable and approximately 10–15 µm wide ( 1–2 cells ) . As matrix density increases ( E ) , sprouts thicken and are on average 20–25 µm wide ( 2–3 cells ) . As Figure 3F shows , at very high densities , sprouts are unable to form . The results presented in this section were very recently confirmed by experiments performed independently and unbeknownst to us by Prof . Sarah Heilshorn and Amir Shamloo in the Materials Science and Engineering Department at Stanford University ( personal communication , manuscript in preparation ) . In Heilshorn and Shamloo's experiments , sprouting formation from dermal microvascular endothelial cells is studied in different collagen concentrations ( 0 . 3 , 0 . 7 , 1 . 2 , 1 . 9 , and 2 . 7 mg/mL ) in a microfluidic device ( for details on their microfluidic device see [49] ) . The cells are subjected to equilibrium VEGF concentration gradients of 50 ng/mL/mm ( with minimum and maximum VEGF concentrations of 100 and 150 ng/mL at the boundaries of the cell culture chamber ) and are incubated for 2–4 days . No sprout formation occurs at 0 . 3 mg/mL . At low collagen concentrations ( 0 . 7 mg/mL ) , some tracks of cells can be seen to form unstable sprouting structures and sprouts are less than 10 µm wide ( compare to Figure 3A ) . Stable sprouting can be seen at a collagen concentrations of 1 . 2 and 1 . 9 mg/mL and sprout are 8±2–18±4 µm thick ( compare to Figure 3B–D and inset in Figure 1A ) . In addition , branching of sprouts is only observed at a collagen concentration of 1 . 2 mg/mL confirming our finding that branching occurs only in a specific matrix density range . At final collagen concentration of 2 . 7 mg/mL , sprouts are 45±15 µm thick or do not grow at all ( compare to Figure 3E , F ) . Our model accurately predicts both the qualitative and quantitative relationships between matrix density and sprout thickness and occurrence of branching confirmed by experiment . Based on our earlier observations , the density of the ECM affects capillary sprout migration speeds . As matrix density is increased , a connected fibrous network develops which could be a mechanism for differences in observed average speeds . We hypothesized that peak extension speeds occur when the mechanical properties of the ECM provide contact guidance cues that are aligned with the chemotactic gradients . To examine the effects of matrix fiber alignment on average rates of capillary sprout elongation , we devise another set of numerical experiments . If matrix fiber alignment plays a prominent role in sprout migration , we would expect more rapid rates of sprout elongation when matrix fibers are aligned parallel to VEGF gradients than when fibers are aligned perpendicular to the gradient . We look at three specific cases: matrix fibers aligned perpendicular to VEGF gradients , matrix fibers aligned parallel to the VEGF gradient , and a combination of horizontal and vertical fibers only . We compare these test cases with the baseline simulations of sprout development on matrices of random fiber orientation . We distinguish and refer to these three cases by the angle that is formed between the fiber axis and the axis of the VEGF gradient , that is , 0° denotes a matrix with fibers aligned parallel to the gradient and 90° identifies a matrix of fibers perpendicular to the VEGF gradient . These numerical experiments represent a simplified replica of the matrix fiber restructuring and fiber alignment that occurs as a result of the tractional forces exerted by endothelial cells during migration [11] , [15] . All matrices have the same matrix fiber density . As matrix fiber density increases , both the number of focal adhesion binding sites available in the ECM and the connectivity of the fiber network increase . As a measure of connectivity , we consider the network connected if there exists a continuous path along matrix fibers from the parent vessel to the source of chemoattractant . As the density of matrix fibers increases , there will be a density that guarantees network connectedness . This threshold density is known as the percolation threshold . Our model fiber networks are constructed by randomly placing fibers at randomly selected but discrete orientations: 0° , ±30° , ±45° , ±60° , and 90° . Consequently , our fiber network most closely approximates a triangular lattice . We estimate that the percolation threshold in our fiber networks occurs between . Recall that we define matrix density , , as the fraction of total tissue space occupied by collagen fibers . This can be interpreted as the probability that a matrix fiber occupies , that is , a bond exists between , two neighboring lattice sites . The bond percolation thresholds depend on lattice geometry and is 0 . 35 for a two-dimensional triangular lattice [50] . The matrix percolation threshold observed in our random matrices corresponds to the bond percolation threshold for a two-dimensional triangular lattice . Interestingly , this percolation threshold is coincident with the density at which our model predicts maximum sprout extension rates . This finding suggests that capillary sprout extension is positively related to the connectedness of the network . We believe that this is because , at the percolation threshold , “tracks” of matrix form , and , near this matrix density , the adhesive and chemotactic energies are well balanced . These factors , in combination , provide strong contact guidance cues to and facilitate the motility of the developing sprout . Referring again to Figure 2A , as the density of the matrix increases above the percolation threshold , sprout extension slows . Thus , even though a connected fiber network is also present at higher densities , higher matrix density is also commensurate with a greater number of focal adhesion binding sites , which impedes cell , and therefore sprout , mobility . Figure 4A–C reports the average extension speed of new sprouts forming on these restructured matrices for respectively . The baseline for comparison is the average extension speed for sprouts formed on matrices with random fiber alignment and is plotted as a solid black line in each plot . At , there are fewer focal adhesion sites in the ECM and the matrix fibers do not form a well connected network . Consequently , at this density , matrix fiber alignment does not have a strong effect on sprout extension speeds . At and , sprouts achieve statistically significant higher average extension speeds when the fibers are aligned parallel to the VEGF gradient ( 0° ) than when fibers are aligned perpendicular to the chemogradient ( 90° ) . The slowest speeds occur on matrices with fibers aligned perpendicular to the VEGF gradient . Interestingly , sprout extension speeds on a matrix composed of randomly oriented fibers are almost as fast as those observed on matrices aligned parallel to the gradient ( 0° ) . The reason for this is clear if we consider the vector describing net contact guidance cues due to fiber orientation . For strictly 0° or 90° matrices , the net contact guidance cues are in the 0° and 90° directions respectively . For matrices composed of fibers aligned randomly in both 0° and ±90° , the net cue is at a ±45° angle . This explains why 0° matrices facilitate the fastest extension speeds and 90° matrices the slowest . For matrices with completely random fiber orientations , the resultant contact guidance cue is at a ±11° angle . This is calculated by vector summation of the discrete fiber orientations at 0° , ±30° , ±45° , ±60° , 90°: . Thus , , and . Since the contact guidance cue for random matrices is approximately aligned with the VEGF gradient , this accounts for our observation that the corresponding extension speeds are close to those speeds recorded on 0° matrices . In these computer generated matrices , the fibers are oriented at discrete angles and thus have a net orientation . Biologically , we are not limited to these discrete angles . Depending on the tissue type , fibers may already be aligned , for instance in muscle , or the tissue may be isotropic and lack any structural orientation . Compared to , the effect of matrix fiber alignment is greatest at . This is because at , the fiber network is well connected and provides adequate focal adhesion sites , but still retains sufficient anisotropy such that strong guidance cues are transferred through fiber orientation . At higher densities ( ) , even though there are ample focal adhesion binding sites , the matrix is more homogeneous , matrix “tracks” become less evident , and strong migratory cues from matrix anisotropies are lost . Consequently , the effect of matrix alignment on average extension speed decreases . These results support our hypothesis that when mechanical or contact guidance cues from the ECM are aligned with the direction of cell chemotaxis , these mechanisms for motility cooperate and promote sprout extension . In light of the above results , we construct patterned matrix topographies to look at the effect of unambiguous contact guidance cues on cell shape , orientation , and sprout morphology . In these numerical experiments , instead of distributing fiber bundles , we engineer matrix cord patterns that vary in width and orientation . As a baseline , we augment a matrix of randomly distributed fibers with horizontal cords 7 . 2 µm thick ( Figure 5A ) . Figure 5B–E shows sprout development on matrix cords 7 . 2 µm thick aligned horizontally , horizontal cords 2 . 2 µm thick , vertical cords 2 . 2 µm thick , and crosshatched cords . Horizontal cords are aligned parallel to the VEGF gradient ( 0° ) ; vertical cords are perpendicular to the gradient ( 90° ) ; crosshatched cords form a ±45° angle with the gradient . Except for the topography of the ECM , all other model parameters are unchanged . We find a strong correspondence between fiber alignment and cell shape and orientation . We define cell orientation as the axis of elongation . In Figure 5A , the density of ambient fibers is great enough to form a well connected mesh and facilitate migration , whereas the higher density matrix cords present a physical barrier that requires more energy to overcome . The anisotropy of the fiber mesh promotes variable cell shapes with no obvious cell orientation . In contrast , in the absence of an ambient fiber mesh , cells quickly adhere to the matrix cords ( Figure 5B ) . Cells orient and elongate in the direction of the horizontal cords . Figure 5C shows the result of reducing cord thickness roughly 1/2 cell diameter from 7 . 2 to 2 . 2 µm . Cells dramatically elongate and orient in the direction of the VEGF gradient . Compare these two cases to Figure 5A and notice that thinner more linear sprouts develop when strong and unambiguous contact guidance cues are aligned in the direction of chemotaxis . Next we examine the effects of matrix cords aligned perpendicular to the gradient . The results are shown in Figure 5D . In this case , although the sprout migrates toward higher concentrations of VEGF , cells elongate and clearly orient in the direction of the matrix cords , perpendicular to the gradient . Figure 5E depicts sprout formation on crosshatched matrix topography . Again , cells orient in the direction of the matrix cords , here at ±45° angles with respect to the gradient . The resulting morphology is a sprout approximately 2 cell diameters thick , notably thicker than the sprouts that develop with strong contact guidance cues aligned in the direction of chemotaxis ( Figure 5B , C ) . Fiber orientation also modulates cell recruitment . When cells elongate and orient in the direction of the VEGF gradient , fewer cells are recruited from the parent vessel and sprout extension is largely due to cell elongation . Compare Figure 5: ( A ) with no obvious cell orientation 15 cells are recruited , ( B ) 11 cells are recruited when cells are oriented in the direction of the VEGF gradient , ( C ) only 3 cells are recruited when cells dramatically elongate , ( D ) 19 cells are needed when cell orientation is perpendicular to the chemoattractant gradient , and ( E ) 19 cells are recruited when cells orient at ±45° with respect to the gradient . These results demonstrate the important role of contact guidance and tissue structure in determining cell shape and orientation . During angiogenesis , endothelial cells not only realign matrix fibers , but they also secrete matrix degrading proteases that break down extracellular matrix proteins and facilitate sprout migration through the stroma [20] . To study the effect of matrix degradation on sprout development , we implement matrix degradation by allowing the tip cell to degrade ∼ ( 0 . 55 µm ) 2 of matrix each minute . We choose this rate of degradation as a rough approximation based on numerical studies of tip cell collagen proteolysis [51] . This rate of degradation is , however , dependent on the availability of ECM , that is , a cell will degrade matrix only if matrix is present . Average sprout extension speeds are recorded and compared with the average extension speeds without matrix degradation for different matrix densities . Figure 6 graphically represents average extension rate pairs for sprouts forming with and without matrix fiber degradation at and shows that the effect of matrix degradation depends on matrix density . At and , matrix degradation results in approximately a 37% increase in average sprout extension speeds at 14 hours . As matrix fibers are degraded , fewer cell-matrix adhesion sites are bound and therefore cellular attachment is reduced resulting in increased motility . At a matrix density of , tip cell matrix degradation only seems to have a significant influence on extension speed at earlier times ( 0–5 hours ) . This observation suggests that the increase in motility due to a loss of bound focal adhesion sites is limited . On the other hand , Figure 6 also shows that for , tip cell matrix degradation has the greatest effect at later times , after 10 hours , indicating that at higher densities , a more significant reduction in matrix density is needed before the cluster of cells can realize any noticeable change in sprout progression . Taken together , these results offer some insight into why velocity curves peak at intermediate matrix densities . On more sparse matrices , , matrix degradation actually slows sprout extension . While this may seem counterintuitive , at lower densities , further reducing fiber density reduces the effectiveness of the ECM to provide a cellular support system that is necessary for normal sprout migration and formation . Thus , depending on the density of the matrix , matrix degradation may result in faster or slower extension speeds , thereby providing pro- and anti-angiogenic effects respectively . Referring to Figure 3F , at the initial endothelial cell is unable to penetrate the stroma and angiogenesis is completely inhibited . In otherwise identical simulations , however , when the tip cell actively degrades the matrix fibers , the tip cell carves out a path through the ECM and a new vessel sprout develops ( Figures 7A , B ) . This result is empirically validated by very recent experiments from Davis et al . showing that human endothelial cells in extracellular collagen matrices degrade a path through the ECM [52] . This path is referred to as a vascular guidance tunnel and can be seen in Figure 7B . The effect of degradation is to decrease the density of the ECM and this decrease is entirely localized to the area under and immediately surrounding the sprout body ( Figures 7B , D ) . However , when we vary ECM density systematically as in our numerical experiments ( Figure 2A ) , the reduction is a uniform reduction . Thus , when comparing extensions speeds associated with changes in ECM density from these two different mechanisms ( one from degradation that is highly spatially heterogeneous and the other a uniform reduction in ECM density ) , one must instead calculate the density of the ECM under the sprout and compare sprout velocities at this density . When this is considered , the extension speeds measured when cells degrade the matrix are in agreement with those measured at the corresponding lower ECM density . This subtlety , however , illuminates an important distinction - that sprout development and progression is not independent of the mechanism for matrix reduction ( spatially uniform versus localized ) . Because the velocity curve steepens above , it is quite expected that at these higher densities , any reduction in matrix density will have a significant effect on sprout velocity . Thus , since degradation is very spatially localized mechanism for matrix reduction , the effect of degradation is even more pronounced at higher densities , which is seen in Figure 6 . In our model , without degradation we observe no branching at matrix fiber densities above . Figures 7C , D show the progress of sprout development at 14 hours with ECM degradation at ( also see Video S3 ) . A new sprout branches from the primary sprout body , an event that emerges only as a result of featured cellular and molecular level dynamics; no rule specifically incorporating branching is imposed . Tip cell degradation reduces ECM density and sets up very high local anisotropies in the matrix fiber structure ( Figure 7D ) , thereby providing strong contact guidance cues to the developing sprout . Figure 7E shows the VEGF gradient profile in picograms ( pg ) that was generated during the simulation shown in Figure 7C . This image shows stronger VEGF gradients develop along the leading edges of sprout . Of interest is that even though there is a strong chemotactic incentive at the branching bifurcation , the lack of ECM prohibits sprout progression . These results lend additional support to our hypothesis that high matrix heterogeneities created by tip cell degradation may be a mechanism for sprout branching . To ascertain the variability and sensitivity of our results to the choice of parameters , we vary one parameter at a time , holding fixed all other Table 1 parameters , and record our observations . A summary and explanation of the effects of parameter perturbation can be found in Table 3 . Decreasing is equivalent to increasing the strength of the bond between endothelial cells . For values of below 10 , cell shapes are grossly contorted and unrealistic . For , cells elongate to increase their cell-cell contact area . As increases , cell-cell adhesion weakens . For , cells move to reduce their surface area contact with each other and are generally rounder in shape . For , cell-cell adhesion becomes too weak relative to the chemotactic energy acting on the cell , and consequently , the tip cell migrates away from the main sprout . Similarly , lower values of correspond to stronger cell-matrix adhesion energies . For , cells are unnaturally distorted in an effort to increase the contact area between the matrix fibers and the cell membrane . At , a relatively strong cell-matrix adhesion bond , sprout morphologies are noticeably thicker and more tortuous . Intermediate values ( ) provide a good balance between contact guidance and release of focal adhesion bonds . Sprout morphologies and extension speeds are relatively insensitive to parameter variability within this range . Above , contact guidance is weak . In this case , chemotaxis is the dominant mechanism governing sprout guidance and , consequently , more linear sprouts develop . An extraordinarily large value , , is equivalent to complete inhibition of cell-matrix adhesion , for example by blocking integrin receptors . Thus , at , endothelial cells do not adhere to matrix fibers at all and are unable to migrate , even in the presence of chemotatic incentives . The results are insensitive to the binding energies between matrix fibers , , and between interstitial fluid molecules , , because they are each collectively identified by the same ID and are therefore always like neighbors . Insensitivity is indicated with an “I” in Table 1 . In addition , the results do not depend on the compressibility properties of the matrix fibers or interstitial fluid , , since the total mass of these ECM components is conserved . We vary between 0 . 3 and 3 . Decreasing makes it easier for the cells to deviate from their target volume . Therefore , at , the cells grow to a larger size overall , and consequently , fewer cells are recruited from the parent vessel . However , average extension speeds are not affected . That sprouts are able to maintain their average velocity with fewer recruited cells highlights cell growth as another mechanism for sprout extension . On the other hand , increasing produces smaller cells , and in this case , more cells are recruited . At , the tip cell migrates away from the main body of the sprout . This is because of the chemotactic sensitivity differential between the tip cell and the stalk cells . Here the relative pressure on a cell to maintain its target volume ( ) is greater than the chemotactic energy of the stalk cells , but not greater than the chemotactic incentives for the tip cell . Thus the tip cell detaches . Figure 8 shows how the average extension speed of a sprout varies with increasing chemotactic sensitivity , . Average speeds are calculated at 14 hours . Above , the physical integrity of the endothelial cells is compromised and cells dissociate due to the relatively strong chemotactic stimulus . Below , chemotactic gradients provide insufficient migratory cues relative to the adhesion energies and the growth constraint , and consequently , the initial cell does not migrate into the stroma . At intermediate values ( ) , sprouts migrate faster with increasing , but sprout morphologies are unaffected . To determine the effect of changing the probability that energetically unfavorable events occur , we vary the parameter , where is the Boltzmann constant and is the effective temperature that corresponds to the amplitude of cell membrane fluctuations . Increasing decreases the probability that an update adding to system energy will be accepted . Increasing effects faster average sprout extension speeds , but no noticeable changes in cell shape , the number of cells recruited , or sprout morphology . This parameter becomes insensitive as it decreases because it is moderated by the continuity constraint . For example , as the probability to accept a change that increases system energy goes up ( decreasing ) , we would expect cells to break up easily , but in this case , the continuity constraint circumvents this effect .
Increased understanding leading to the ability to control angiogenesis in vivo has serious clinical implications . Angiogenesis is a crucial event to many physiological processes . Embryonic development and endometrium vascularization , arteriogenesis resulting from ischemia and vessel occlusion , wound healing and tissue repair are all homeostatic processes that require new vessel growth for normal function . Angiogenesis can also lead to pathological conditions . Tumor angiogenesis , proliferative diabetic retinopathy and macular degeneration , psoriasis and rheumatoid arthritis occur when angiogenesis is unhalted [54] . On the other hand , insufficient vessel growth can lead to heart attack , stroke , and impaired ulcer and wound healing . Existing angiogenic therapies can be broadly categorized as those that ( 1 ) target growth factors or growth factor cell receptors that stimulate vessel growth , ( 2 ) block cell invasion into the stroma , and ( 3 ) directly induce endothelial cell apoptosis . Because of its established prominence in both homeostatic and aberrant angiogenesis , VEGF and its receptors are prime therapeutic targets . There is overwhelming experimental evidence that in order to form functional vessels , the various VEGF isoforms must be precisely regulated and that the blockage of even a single growth factor might limit tumor-induced vascular growth [20] , [23] , [55] . The most promising approaches to anti-angiogenesis therapies are those based on blocking VEGF or VEGF receptors [56] . VEGF neutralizing antibodies , soluble VEGF receptors , and receptor tyrosine kinase inhibitors are examples of therapies currently being utilized or that are undergoing clinical trials [57] . One problem associated with targeting growth factors as therapeutic agents is that they are often constitutively expressed in vivo and can be proteolytically released . Thus tight control is , in practice , hard to maintain . For example , it is known that connective tissue , which contains some of the same fibrous proteins that are found in the ECM , can significantly inhibit cell migration and prevent the formation of sprouts [20] . The ECM and cell-matrix associations also provide promising possibilities for angiotherapy , but have only more recently received attention as targets and are in less advanced stages of clinical development . Consequently , modeling and simulation have the potential to contribute to and propel further advancement . Current therapeutic interventions aimed at cell-matrix interactions during angiogenesis focus on tissue inhibitors of metalloproteinases ( TIMPs ) and on integrin-mediated cellular adhesion [54] . Blocking proteolysis is intended to inhibit cellular migration into the stroma and to prohibit MMP-dependent release and activation of ECM sequestered angiogenic factors . Our results indicate that regulating the cellular production of matrix degrading proteases can shift sprout velocity curves for the purpose of promoting or inhibiting angiogenesis . We show that at low matrix densities ( ) , matrix degradation has anti-angiogenic effects , whereas above , degradation facilitates sprout progression . Using our model , we regulate cell-matrix binding affinity ( ) and control the number of focal adhesion binding sites available in the ECM ( density modulation ) to test the efficacy of integrin specific anti-angiogenic therapies . The integrin receptor is significantly upregulated in angiogenic vessels compared to mature vessels [54] , making this receptor one logical therapeutic choice . Setting is equivalent to blocking integrin receptors . Our simulations show that decreasing the binding affinity of integrin receptors prevents endothelial cells from adhering to matrix fibers and cells are unable to migrate even in the presence of chemotactic incentives . We also show that cellular motility is inhibited at high matrix densities . This is due to the greater number of focal adhesion binding sites available . Simulations suggest that regulating the affinity or number of cell-matrix focal adhesion sites either biochemically or mechanically produces anti-angiogenic effects . In these collective studies , we use the model to isolate and examine variations in fiber density and structure , and proteolytic matrix degradation as independent mechanisms that control vascular morphogenesis . Nonetheless , the integrin , protease , and growth factors systems are highly connected and provide regulatory feedback for each other [54] . Thus , there is still a need for more in depth investigations of the relationship between extracellular stimuli and cellular function . In particular , studies focusing on intracellular signaling and cross-talk between the integrin and growth factor receptors are of key importance . In addition , experimental models are needed to measure critical biochemical activity , for example , matrix proteolysis during angiogenesis , and to verify the predictions made herein regarding the pro- and anti-angiogenic effects of manipulating the ECM .
|
A cell migrating in the extracellular matrix environment has to pull on the matrix fibers to move . When the matrix is too dense , the cell secretes enzymes to degrade the matrix proteins in order to get through . And when the matrix is too sparse , the cell produces matrix proteins to locally increase the “foothold” . How cells interact with the extracellular matrix is important in many processes from wound healing to cancer invasion . We use a computational model to investigate the topography of the matrix on cell migration and coordination in the context of tumor induced new blood vessel growth . The model shows that the density of the matrix fibers can have a strong effect on the extension speed and the morphology of a new blood vessel . Further results show that matrix degradation by the cells can enhance vessel sprout extension at high matrix density , but impede sprout extension at low matrix density . These results can potentially point to new targets for pro- and anti-angiogenesis therapies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"cardiovascular",
"disorders/vascular",
"biology",
"developmental",
"biology/morphogenesis",
"and",
"cell",
"biology",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"cell",
"biology/cell",
"growth",
"and",
"division",
"mathematics",
"cell",
"biology/cell",
"adhesion",
"computational",
"biology/systems",
"biology"
] |
2009
|
Topography of Extracellular Matrix Mediates Vascular Morphogenesis and Migration Speeds in Angiogenesis
|
All of our current knowledge of African trypanosome metabolism is based on results from trypanosomes grown in culture or in rodents . Drugs against sleeping sickness must however treat trypanosomes in humans . We here compare the transcriptomes of Trypanosoma brucei rhodesiense from the blood and cerebrospinal fluid of human patients with those of trypanosomes from culture and rodents . The data were aligned and analysed using new user-friendly applications designed for Kinetoplastid RNA-Seq data . The transcriptomes of trypanosomes from human blood and cerebrospinal fluid did not predict major metabolic differences that might affect drug susceptibility . Usefully , there were relatively few differences between the transcriptomes of trypanosomes from patients and those of similar trypanosomes grown in rats . Transcriptomes of monomorphic laboratory-adapted parasites grown in in vitro culture closely resembled those of the human parasites , but some differences were seen . In poly ( A ) -selected mRNA transcriptomes , mRNAs encoding some protein kinases and RNA-binding proteins were under-represented relative to mRNA that had not been poly ( A ) selected; further investigation revealed that the selection tends to result in loss of longer mRNAs .
Trypanosoma brucei subspecies and related parasites infect humans , cattle , camels , and horses , causing substantial economic losses throughout the tropics [1 , 2] . Human sleeping sickness in East Africa is caused by Trypanosoma brucei rhodesiense , a zoonotic parasite which differs from Trypanosoma brucei brucei ( which is found in cattle ) only by the acquisition of a single gene enabling survival in human serum [3] . Trypanosoma brucei gambiense causes a more chronic human disease in West Africa . After an initial phase in which the parasites are restricted to the blood and tissue fluids , T . gambiense and T . rhodesiense penetrate the central nervous system ( CNS ) . T . rhodesiense disease is usually fatal , whereas some T . gambiense-infected people are asymptomatic [4] . The organisms completely evade adaptive humoral immunity because they show antigenic variation , repeatedly changing their surface coat of variant surface glycoprotein ( VSG ) . As a consequence , disease control has to rely on chemotherapy of detected cases , combined with insecticides and traps to control the tsetse fly vector . There are , however , very few drugs available to treat African trypanosomiasis , they are all unacceptably toxic , and resistance is arising [5] . Moreover , within the CNS , trypanosomes are sensitive only to drugs that cross the blood-brain barrier , limiting therapeutic options for the late stage of the disease . All of our knowledge of the biochemistry and molecular biology of T . brucei depends on laboratory models , and this includes the early phases of drug development . Targeted approaches rely on biochemical knowledge gained from culture alone for target selection; in the phenotypic approach , compounds are initially screened using cultured trypanosomes . Promising leads are then tested in rodent models . Within rodents , as in other mammals , T . brucei spread throughout the blood and tissue fluids and invade the brain . Most trypanosomes within the rat brain parenchyma appear degraded , although cells of normal appearance are seen in the pia mater and cerebrospinal fluid ( CSF ) [6] . Many trypanosomes are also found in the adipose tissue of mice; in this case , transcriptome analysis suggested metabolic differences from blood trypanosomes [7] . It is therefore possible that differences between trypanosomes at different sites could contribute to treatment failure . In natural T . brucei infections , the trypanosomes are pleomorphic . Proliferating forms have long slender morphology , and obtain ATP through glycolysis . The parasites produce a soluble signal known as "Stumpy Inducing Factor" ( SIF ) , whose identity is still unknown [8 , 9] . At high density , when the SIF concentration reaches a critical threshold , the trypanosomes arrest in G1 , and acquire a more stumpy shape [10–12] . Stumpy forms have increased expression of some mitochondrial proteins; markers that are absent ( or much less expressed ) in bloodstream forms include ESAG9 [13] , PAD1 [14] and the protein phosphatase PIP39 [15] . The higher expression of mitochondrial protein mRNAs means that stumpy forms are pre-adapted to differentiate into the procyclic form which multiplies in the tsetse midgut , since procyclic forms rely on mitochondrial energy metabolism . Procyclic forms lack VSG , instead having a surface coat of procyclin proteins containing Glu-Pro ( EP ) or Gly-Pro-Glu-Glu-Thr ( GPEET ) repeats . Relative to long slender forms , stumpy forms have slightly increased procyclin mRNA expression [16 , 17] , although the protein is not made . The density at which long slender trypanosomes cease to proliferate , and become stumpy , differs according to the environment [8] . In rodents , maximal densities of fully pleomorphic parasites are 1–5 x 108/ml [16–18] for the initial parasitaemia , with differentiation initiating above 5 x 107/ml . In contrast , differentiation-competent trypanosomes in liquid culture arrest at 1–2 x 106/ml [8] ) . In experimentally infected cattle [19] or Mastomys natalensis rats ( but not Swiss mice ) [18] , parasitaemias are lower during chronic infection than in the initial wave . Humans who present for sleeping sickness diagnosis , have usually been infected for some time , and rarely show T . rhodesiense parasitaemias above 106/ml , although a parasitaemia of 108/ml was recently recorded in a Polish tourist [20] . The reasons for the low densities during chronic infection are unknown . The infection may be suppressed via innate immunity and inflammatory responses; there may be metabolic constraints; or SIF may accumulate more readily . In cattle , the infectivity of the parasites for tsetse seems to be relatively unaffected by the parasitaemia level [19] . Perhaps SIF levels in cattle are high despite low parasitaemias , or the trypanosomes are more receptive to it than in the initial wave; or alternatively there may be tissues in which parasite densities are substantially higher than in the blood . After multiple passages in rodents or culture , African trypanosomes lose the ability to make stumpy forms , becoming monomorphic . It is these forms that are used for most biochemical and molecular biology experiments . In the work described in this paper , we set out to characterize T . rhodesiense in human patients . We asked two questions: To answer these questions , one should ideally compare the proteomes of different parasites growing in different environments . Messenger RNA levels do not predict protein levels reliably , because trypanosomes have strong regulation of translation [21–23] , and protein degradation rates are presumably also important . However , due to the very limited amount of material available , and the small numbers of parasites in comparison with host cells , proteome characterization of trypanosomes from patients is not feasible . We therefore instead analyzed transcriptomes . The results suggest that cultured trypanosomes are in most respects a satisfactory model for parasites in humans . The gene expression profiles also indicated that parasites in human CSF are , if anything , growing more actively than those in human blood .
Samples of blood and CSF were obtained from patients presenting for diagnosis at the clinic in Lwala hospital , Kaberamaido district , which is located in the T . b . rhodesiense focus of North Eastern Uganda . We tested a variety of methods for RNA preparation using mixtures of trypanosomes with blood from the Heidelberg University blood bank . All gave acceptable yields of intact RNA , with the best results being obtained by suspension of buffy coat trypanosomes in denaturing solutions such as Trizol . In contrast , use of such methods with field blood samples resulted in exceptionally low yields of RNA ( a few nanograms ) , and the preparations were much too degraded to allow RNASeq library preparation . This was true even if the purified RNA was initially resuspended in a solution containing RNase inhibitors . We succeeded in obtaining sufficient RNA for sequencing from blood only when the samples were placed directly into PAXgene tubes [24] . For CSF , some RNAs were purified using Trizol and some using the PAXgene tubes . Samples from blood with the highest trypanosome densities , and from CSF with highest ratios of trypanosomes to leukocytes , were chosen for RNA preparation . For CSF , human rRNA was depleted; for the blood mRNA samples , both rRNA and haemoglobin mRNA were depleted . Subsequently cDNA libraries were prepared and sequenced . To simplify the alignment and counting , an easy-to use pipeline was created; this can be downloaded from [25] . Within this pipeline , the sequences were first trimmed to exclude the 60 most abundant sequences; these include not only the adapters , but also the most over-represented rRNAs . Removal of these over-represented sequences greatly simplified and sped up the subsequent alignment . After alignment and read counting , libraries from samples giving adequate numbers of trypanosome reads were re-sequenced to increase the read depth . All of the resulting datasets are available at Array Express . For CSF , parasitaemias varied from 4–66 x 104/ml , and between 1% and 10% of reads were trypanosome-specific . These numbers roughly correlated with the ratio of trypanosomes to white blood cells , and suggested that a CSF white blood cell contained about 10 times more mRNA than the trypanosomes ( Table 1 and S1 Table , sheet 1 ) . PAXgene sampling lyses the parasites . This meant that all blood parasitaemias had to be estimated using the diagnostic thin films , using rat samples for calibration . Bloodstream parasitaemias were 100–1000 times higher than those in CSF , and the percentage of reads aligning to the T . brucei TREU927 genome varied between 6% and 77% ( Table 1 and S1 Table , sheet 1 ) , but these reads included a substantial ( and variable ) proportion that corresponded to rRNA . For comparison , when poly ( A ) + RNA from Leishmania brazilienesis mouse skin lesions was sequenced , about 1% of the reads mapped to the Leishmania genome [26] . In this paper , we discuss the trypanosome transcriptomes . Results for the human mRNAs will be analysed separately . To enable direct comparison of the human results with an experimental sample obtained using exactly the same methods , we infected 8 immunocompetent rats with two T . rhodesiense isolates that we had obtained 1–2 years earlier from patients attending the same clinic as the current ones [17] . These trypanosomes had undergone 2 mouse passages prior to infection; their genomes are similar to each other [17] . Transcriptomes from single rats had been obtained from these lines previously ( RBC1 and RBC4 in S1 Table ) , but those parasites were morphologically uncharacterized and parasitaemias were unknown [17] . This time , blood for RNA preparation was taken at parasitaemias ranging from 5 x 107–2 x108 ( Fig 1A and S1 Table sheet 1; samples "RBD" ) . RNA was prepared from all samples and treated exactly as for human blood . Sufficient reads for analysis were obtained from six samples . After infection of immunocompetent mice with EATRO1125 cells , PAD1 mRNA became detectable when parasitaemias attained about 2 x 108/ml , and stumpy forms were present when this density had persisted for 3 days [27] . To see whether the newly-isolated trypanosomes behaved similarly in rats , thin blood films from four rats were stained for PAD1 and for DNA , and the distance between the nucleus and kinetoplast was measured . We did not detect any stumpy forms: the cells were longer than stumpy forms and no PAD1 was detected ( Fig 1B ) . We also searched three of the relevant raw sequence files for short sequences specific to PAD1 [27] , but could not find any matches . This suggests that in trypanosomes from this region , the PAD gene family has diverged too far to allow PAD1 identification from sequence alone . Unexpectedly , there was no correlation between cell density and parasite length ( Fig 1B ) . In the immunocompetent rats the immune response was presumably contributing to parasitaemia control . Sample RBD3 , which had the shortest parasites among the three tested samples , had relatively high levels of PAD gene family mRNA ( S1 Table sheet 2 ) , although we were unable to detect PAD1 protein or PAD1-specific sequence . This sample unfortunately yielded too few trypanosome-derived reads for statistically valid transcriptome analysis . The low read count was presumably partly caused by the low parasitaemia , but since it is known that stumpy forms have low mRNA content [28] , this might also have contributed . In addition to the samples described above , we incorporated three previously published datasets for blood trypanosomes from rats infected with either culture-adapted or fresh T . rhodesiense , and several datasets from mouse blood . Three of the mouse transcriptomes were new data for well-characterised long slender trypanosomes of a pleomorphic strain ( samples "MBA" ) . Data for mouse adipose tissue ( Mad ) were also included . Finally , a variety of published results from cultured parasites ( Cult ) were included; these were almost uniformly from cells at low enough densities to be in log phase . Details of all datasets studied are in Table 2 and further information is in S1 Table , sheet 2 . Raw sequence data that had been obtained from other labs were re-analysed using our own pipeline in order to ensure that parameters for alignment and read counting were identical . Most of the transcriptomes had been prepared using mRNA that had been purified either by poly ( A ) selection ( giving "poly ( A ) +" RNA ) , or by depletion of rRNA ( giving "ribo-minus" RNA ) . Both selections involve hybridization to oligonucleotides coupled to magnetic beads . For poly ( A ) selection the RNA is bound to oligo d ( T ) in high salt , washed with lower salt buffer , then eluted with water . For rRNA depletion , a set of oligonucleotides complementary to rRNA is used , with moderate salt conditions; the RNA is allowed to bind , then the supernatant is taken for further analysis . To align the transcriptomes , we wrote various user-friendly scripts . First , there is a python script that aligns the reads while allowing for the peculiar nature of kinetoplastid genomes [25] . Since mRNA annotation is incomplete and many mRNAs have numerous different processing sites , the script counts only the reads that align to open reading frames . Trypanosome genomes have many repeated genes , so to account for this , the application is set to allow each read to align up to 20 times . In the subsequent analysis , over-counting of the repeated genes is avoided by considering only a set of "unique" open reading frames in which only one representative of each sequence was present . ( This list is adapted from [29] ) . As a consequence , the RNA abundances for each gene , as estimated by reads per million reads ( RPM ) , if normalized to open reading frame length , should approximate to the level of mRNA . To analyse the data statistically we used another custom application , DEseqU1 , which runs in RStudio and uses DEseq2 [30] for significance estimation . The application yields principal component analysis , which shows which transcriptomes are closely related . In addition , it allows analysis according to gene functions and cell cycle regulation [31] . We assigned gene functions using a combination of the annotations in TritrypDB , and manual annotations based on publications; all are listed in S1 Table , sheet 3 . Both the unique gene list and the assigned gene functions can be changed by editing the relevant text files . Finally , heat maps were generated using another RStudio script , ClusterViewer . rmd . This is included as S1 app . The included folder enables readers to examine the data in this paper for themselves , either by looking at all genes , or by examining specific functional categories . ClusterViewer can be adapted easily for other datasets by changing the input file and a few lines of the script , as described in the instructions . The unique gene list does not include variant surface glycoproteins ( VSGs ) . To find VSGs expressed in human patients , we took the two largest datasets and assembled all mRNAs as contigs . Next , we searched for the 13nt sequence that is shared by the 3'-untranslated regions . The pipeline used to do this is at https://github . com/klprint/IdentifyVSGs and the assembled VSGs are in the supplement . All read counts are presented in S1 Table , sheet 3 . Normalized values ( reads per million reads ) are in S1 Table , sheet 4 . Only datasets with at least 3x105 reads aligning to the unique gene set were analysed ( S1 Table , sheet 1 ) ; this resulted in exclusion of two samples for each set of field isolates . The principal component analysis in Fig 2 shows how the different transcriptomes are related . It covers 63% of the total variance , with 44% of the variation on the x-axis . Strikingly , mRNAs prepared in similar ways mostly clustered together irrespective of source ( Fig 2 ) . The only exception was one poly ( A ) + culture dataset ( CultE ) , which was quite similar to the ribo-minus culture transcriptomes; the reason for this is unknown and it will not be considered further . Among the poly ( A ) + mRNAs , transcriptomes from four rats infected with recent Ugandan isolates [17] showed considerable variation; RBC1 and RBC2 were extremely similar to those of from a culture-adapted Ugandan strain ( RBA ) and from long slender EATRO1125 strain trypanosomes in mice ( MBA ) , whereas two samples that had relatively high expression of stumpy-form marker mRNAs ( RBC3 and RBC4 ) [17] were somewhat apart from the main cluster . Transcriptomes that had been generated using RNA that was both poly ( A ) selected and rRNA depleted clustered separately from the others . As reported previously , within this set there were differences between cells from adipose tissue , blood and culture [7] . However , the results from cultured cells [32] were surprisingly different from others despite similar growth conditions ( S1 Table Sheet 2 ) . This discrepancy is presumably due to technical differences , so these datasets were not included in subsequent comparisons . Before comparing the human samples with the others , we looked at the effects of poly ( A ) selection in more detail ( S1A Fig ) . We were surprised to see that , according to both cluster ( Fig 3 , S2 Table sheet 3 ) and enrichment analyses ( S2 Table sheet 2 ) , mRNAs encoding protein kinases ( Fig 3A ) and RNA-binding proteins ( Fig 3B ) were selectively lost after poly ( A ) selection . Readers can analyse this themselves using the cluster viewer which is in the Supplement . By Northern blotting , we confirmed this result for two RNA binding protein mRNAs , ZC3H32 and ZC3H8 ( S2A Fig ) . Loss of these particular functional sets might have been meaningful—perhaps they have very short poly ( A ) tails . On the other hand , many of the most affected protein kinase and RNA-binding protein mRNAs are rather long , either because of long open reading frames , or long 3'-untranslated regions ( 3'-UTRs ) ( S3 Fig ) . We therefore wondered whether poly ( A ) selection was resulting in the loss of long mRNAs . A comparison using all available datasets ( Fig 4A , S2 Table ) confirmed this suspicion , with particularly strong effects above about 4kb ( 212 on the graph ) . Oddly , when we divided the datasets according to how the parasites had been grown , we found a clear length effect for RNA from cultured parasites ( Fig 4B ) but not RNA from rat blood ( S2B Fig ) . However the poly ( A ) + rat blood samples were biologically much more diverse and less well characterized than the cultures , so inter-sample variation might conceal a length effect . We therefore decided to follow up the result for cultures . There are two obvious technical reasons why long mRNAs might get lost during poly ( A ) selection . One is degradation . Indeed , it has previously been demonstrated that poly ( A ) selection can result in preferential loss of sequence towards the mRNA 5'-end [33] . We counted only reads from open reading frames , so any mRNAs that were broken in the 3'-UTR would fail to be counted . Somewhat unexpectedly , there was no correlation between poly ( A ) +/ribo-minus ratios and the annotated 3'-UTR length ( S2C Fig ) . This conclusion must however be regarded with caution because 3'-UTR lengths in the database are sometimes too short . Further investigation suggested that it was indeed mRNA length that was important , rather than the functional class of the encoded protein . For mRNAs encoding both cytoskeletal proteins and protein kinases , the correlation between abundance and length was greater for the total mRNA length than it was for the open reading frame alone ( Fig 4C; S2D–S2F Fig ) . Moreover , scrutiny of published RNASeq read density maps ( TritrypDB ) for several of the outliers among protein kinase mRNAs suggested that the annotated 3'-UTR lengths were incorrect . Similar length-abundance correlations were seen for mRNAs encoding RNA-binding proteins ( Fig 4E ) , cell cycle proteins ( S2G Fig ) and translation factors ( S2H Fig ) . There was less correlation for transporters ( S2I Fig ) and none for the relatively short mRNAs encoding ribosomal proteins ( Fig 4C ) . To investigate the effect of mRNA length directly , we experimentally changed the lengths of two open reading frames ( Tb927 . 4 . 1500 and Tb927 . 8 . 1050 ) , by integration of a yellow fluorescent protein ( YFP ) open reading frame at various positions relative to the endogenous start codon . This generated progressively shorter mRNAs ( Fig 5A and 5B ) . The YFP mRNAs were measured by Northern blotting ( Fig 5C and 5E , and S4 Fig ) and by reverse transcription followed by real time PCR ( qRT-PCR ) . The mRNA from the upstream puromycin resistance cassette ( PAC ) ( Fig 5A ) was used as a loading control . For both genes , the mRNA sizes were as expected ( Fig 5C and 5E ) , but for Tb927 . 8 . 1050 there was also a shorter mRNA species ( Fig 5F ) , which was roughly 20% of the total irrespective of length . Up to a length of 8 kb , results from blots and qRT-PCR were comparable ( Fig 5D and 5F ) . The only exception was for the shortest mRNAs , which had a truncated GFP ORF; for these , results from qPCR were anomalous . The abundance of total mRNA from the Tb927 . 4 . 1500 reporters decreased with increasing length ( Fig 5C and 5D ) . To find out whether this might be due to differing half-lives , we measured mRNA abundance by qRT-PCR 30min after inhibition of splicing and transcription . Reassuringly , the half-life of the full-length fusion mRNA was similar to that of the unmodified version . However , the half-life increased with progressive truncations ( Fig 5B ) . Thus for the Tb927 . 4 . 1500 locus , the increase in mRNA abundance with decreasing length can probably largely be attributed to increased mRNA stability . In contrast , for Tb927 . 8 . 1050 no reproducible length effect was seen on abundance ( Fig 5E and 5F ) and this was also the case for preliminary half-life measurements . As in other organisms [34] , codon optimality can affect trypanosome mRNA half-lives ( M . Carrington , Cambridge University , personal communication ) . This does not explain the difference between the two loci: for both , deletions of the open reading frame by GFP integration resulted in progressive increases in codon optimality ( Fig 5B ) . Perhaps there are differences in codon distribution: clusters of non-optimal codons might have a bigger effect on translation than non-optimal codons that are spread uniformly throughout the sequence , and the positions of non-optimal codons relative to the start codon are known to be important [34] . Contrary to our previous conclusions from modeling [23 , 35] , results for these two genes did not yield any evidence for an effect of mRNA length that was independent of the half-life . More accurate half-life measurements would be needed to confirm this . In the whole transcriptome analysis 60% of the mRNAs that were significantly depleted after poly ( A ) selection were longer than 4kb . Indeed , for the YFP fusion mRNAs , poly ( A ) selection caused loss of reporter mRNAs longer than 5 kb , and there was also some loss of the 4 . 5 kb Tb927 . 8 . 1050 locus mRNA ( Fig 5D , 5F and 5G ) . We concluded that poly ( A ) selection can cause loss of mRNAs longer than 4 kb , but also that losses are variable . Perhaps other sequence characteristics also contribute to the mRNA yield . In subsequent comparisons , we considered only the datasets derived from rRNA-depleted RNA . The principal component analysis for these samples ( Fig 6A ) suggested that from the trypanosomes' point of view , there are few differences between rat and human blood; the total number of mRNAs with significantly different abundance , 125 , was very low ( S2 Table ) and probably not far from random variation . The CSF transcriptomes were separated from those for blood , but it was notable that our only samples from the CSF and blood of a single patient ( C71 and B71 ) were relatively similar . Notably , the CSF parasite transcriptomes more closely resembled those of log-phase cultured cells ( Fig 6A ) than did those of blood parasites . Differences between the various samples could be due to differences in available nutrients or immune responses , but most obviously , from the presence of stumpy forms , since densities in blood were much higher than in CSF . Cluster analysis of the samples from humans and rats only ( S5 Fig , S3 Table ) distinguished two groups . One group included the CSF samples and two human blood samples ( HB71 and HB73 ) , and the other group included the remaining blood samples . The former group showed higher expression of mRNAs encoding cytoskeletal proteins , several translation factors , tRNA charging enzymes , RNA degradation pathway proteins , and some protein kinases ( clusters 3 and 16 , S3 Table ) . At the same time , it showed lower expression of mRNAs encoding numerous mitochondrial proteins ( clusters 12 and 18 ) . Comparison of the human blood and CSF parasite transcriptomes ( S2 Table , sheet 1 ) revealed 320 mRNAs that were lower in blood , and these were again significantly enriched in mRNAs encoding cytoskeletal proteins ( S2 Table , sheet 2 ) . 830 mRNAs were higher in blood , with enrichment for mitochondrial electron transport and amino acid transport ( S2 Table , sheet 2 ) . Cell-cycle-regulated genes that were over-expressed in blood mainly show peak expression in G1 , whereas many of those that were more abundant in CSF peak in S-phase ( Fig 6B ) [36] . All of these results suggested that the CSF parasite population included more actively multiplying parasites than the bloodstream populations . Expression of mitochondrial proteins suggested that that the bloodstream populations included some parasites that were beginning to differentiate to stumpy forms . To examine the link between gene expression and cell density , we looked at a few examples . Results for the translation initiation factor EIF4A , the stumpy-inducing phosphatase PIP39 and a cytochrome oxidase subunit ( Fig 6C–6E ) , as well as various other regulated mRNAs ( S5 Fig ) revealed no simple relationship between expression and cell density . For the rat blood samples , there was also no correlation with attaining the plateau of parasitaemia . This was consistent with our previous morphological analysis of the rat samples ( Fig 1B ) . Finally , we looked at the differences between cultured and human parasites . We focus here mainly on the CSF parasites: differences between culture and blood were more difficult to interpret due to the likely presence of growth-arrested parasites in the bloodstream ( see above ) . The results suggested that there are indeed some differences which must be considered when using cultured parasites as a model . Genes that were more highly expressed in the human CSF samples included those encoding four membrane proteins; but these were mainly from multi-gene families , which can vary between strains . The products of CSF up-regulated mRNAs were also enriched for ribosomal proteins ( S2 Table sheet 2 ) . More interestingly , the increased mRNAs encoded nine protein kinases , and three potential cyclins , two of which ( CYC10 and CYC11 ) were also increased in human blood ( S2 Table sheet 1 ) . The mRNAs in rat blood and cultured trypanosomes were previously compared in a ribosomal profiling study [22] . There was no significant correlation between those results and ours . Nevertheless , in that study too , the blood parasites had higher levels of mRNAs encoding CYC10 and CYC11 , various protein kinases , protein phosphatases , and RNA-binding proteins ( S2 Fig Sheet 1 ) . Cultured trypanosomes had higher mRNA levels than CSF trypanosomes for mRNAs encoding half of the Sm complex and some other splicing factors; various RNA-binding proteins including ZFP2 and ZFP3; 26 cytoskeletal proteins; 51 mitochondrial proteins , 26 proteins involved in vesicular transport , numerous translation initiation factors , and the whole of the core proteasome . The differences for blood trypanosomes were to some extent similar , but in this case the cultures also expressed more translation initiation factor mRNA . These results might mean that the cultured cells are multiplying faster than the cells in the patients . Since all of the ribo- culture datasets were from monomorphic trypanosomes , we also compared poly ( A ) + mRNAs from pleomorphic and monomorphic cultures . This revealed that cultured pleomorphic EATRO1125 had lower expression of numerous RNA-binding proteins , and some RNA decay and cytoskeletal proteins ( S2 Table Sheet 1 ) , than cultured monomorphic Lister 427 . Cell cycle analysis ( Fig 7A ) suggests that this reflects more active division of the Lister 427 cultures . Unsurprisingly , this indicates that pleomorphic cultured cells ought to resemble parasites in patients more closely than monomorphic cells do . The zinc finger protein ZC3H11 is a positive regulator of mRNAs encoding protein refolding chaperone complexes [37] , and is required for the survival of procyclic forms after heat shock . After a brief heat shock , ZC3H11 mRNA is unaffected but its translation is strongly induced [38] . Remarkably , in all of our new ribo-minus transcriptomes , ZC3H11 was among the ten most abundant mRNAs , being 4–6 times more abundant than in cultured trypanosomes . Comparison of all the datasets also revealed strain differences: ZC3H11 mRNA levels were lower in culture-adapted parasites than in the newly-isolated ones ( Fig 7B ) . This is definitely a difference in regulation: ZC3H11 is a single-copy gene in the new rat blood trypanosomes as well as in Lister 427 [17] . All of the human patients from our study had fever so it may be that the ZC3H11 mRNA is stabilized by prolonged elevated temperatures . Paradoxically , though , many chaperone mRNAs were significantly lower in the CSF parasites than in culture , and some known ZC3H11 target mRNAs [37] showed a weak inverse correlation with ZC3H11 mRNA ( Fig 7C–7E ) . Further comparison revealed over 200 mRNAs with expression that was either negatively or positively correlated with ZC3H11 mRNA in both ribo-minus and poly ( A ) + datasets ( S2 Table Sheet 4 ) . Products of negatively correlated mRNAs included cytoskeletal proteins , protein kinases and phosphatases , while for positively correlated mRNAs there were some mitochondrial and ribosomal proteins . The meaning of these results is unclear , since the level of ZC3H11 protein does not correlate with the mRNA [39] . Contrary to the results shown here , in the Jensen ribosome foot-printing study , ZC3H11 mRNA was lower in blood than in culture parasites—but the blood parasites yielded 10 times more ribosome footprints [22] . ZC3H11 is also phosphorylated , which might affect its activity [37 , 39] . This is , to our knowledge , the first study that compares transcriptomes of parasites from several different labs , with different strains , growth conditions , and RNA preparation methods . We discovered that each of these affects the transcriptome . It was already known that poly ( A ) selection and rRNA depletion affect RNA-Seq-derived trypanosome transcriptomes [40] , and such effects have been comprehensively demonstrated for mRNAs for other species , including humans [41 , 42] . From analysis of the data , combined with reporter experiments , we concluded that technical factors , such as trapping of RNA in the matrix , strongly contribute to depletion of long mRNAs . The reason that the differences are concentrated within mRNAs encoding particular functional protein classes may be that these classes have a disproportionate number of long mRNAs; in the case of both protein kinases and RNA-binding proteins , this is because their 3'-UTRs are longer than average ( S3 Fig ) . mRNAs encoding ribosomal proteins are , in contrast , unusually short ( S3 Fig ) . Since the 3'-UTR annotations are not all correct , the extent to which other factors contribute is not certain . Some mRNAs may not be retained on the oligo d ( T ) matrix because they have very short poly ( A ) tails . Although poly ( A ) tails usually protect from degradation and promote translation , well-expressed Opisthokont mRNAs tend to have short tails [43] . We compared mRNA half-lives [35] and ribosome densities [22 , 23] , but for these characteristics we found no significant difference between the trypanosome mRNAs that were enriched or depleted by poly ( A ) selection . When we started this study , we expected that of the available laboratory models , trypanosomes growing in rodent blood would have transcriptomes that most closely resembled those of pleomorphic trypanosomes growing in humans . Our results confirmed this expectation , but also revealed some intriguing differences between trypanosomes growing in different environments . Several hundred mRNAs were significantly different between cultured and human-grown trypanosomes . The functions of the proteins encoded by those mRNAs suggested that the cultured parasites might be multiplying faster than parasites in blood , and that the blood parasites were affected by environmental stresses . Although some of the differences in gene expression might have been due to the growth environments , others can probably be attributed to culture adaptation of the parasites . All of the data from cultures were for Lister 427 parasites , which were probably originally isolated from a cow in Tanganyika ( now Tanzania ) ( see http://tryps . rockefeller . edu/DocumentsGlobal/lineage_Lister427 . pdf ) . The cells have been serially passaged for many years , and in culture since about 1985 . To understand the effects of culture adaptation it will be necessary to follow parasite genomes and transcriptomes during that process . There was no systematic correlation between human parasitaemias and expression of mRNAs that are increased in stumpy forms . In rats , in the samples analysed , there was also no correlation between parasitaemia and parasite morphology . Stumpy forms of T . gambiense were originally reported in the descending phase of human parasitaemia [44]; perhaps some of the patients were also in that phase . Moreover , some tissues may harbour higher trypanosome densities than are present in the blood , and thus accumulate stumpy induction factor . The resulting stumpy parasites might subsequently escape into the circulation . Previous rodent studies did not support this idea [6 , 7 , 45 , 46] . However , recent results do suggest that tissue and blood parasitaemias may be different . T . b . gambiense were found in the skin of asymptomatic humans lacking detectable blood parasitaemia [47] , and relatively high proportions of stumpy-form ( PAD1-positive ) T . brucei were detected in mouse skin and adipose tissues [47 , 48] . An important motivation for our study was to find out whether CSF trypanosomes are significantly less metabolically active than those in blood , and thus less susceptible to drug treatment . Overall , if transcriptomes can be taken as a guide to enzyme expression , the results did not provide evidence for systemic metabolic differences between blood and CSF trypanosomes . If anything , the CSF parasites are likely to be more metabolically active than those in the blood , and thus more susceptible to any drug that targets parasite metabolism or multiplication .
For the human studies , ethical approval of protocols was obtained from the Ministry of Health and Uganda National Council of Science and Technology ( Ethical approval No . HS 729 ) , Uganda , and the ethics committee of University of Heidelberg , Germany . All patients recruited into this study received written and verbal information explaining the purpose of the study and they gave informed consent . The ethical consent forms were written in English and translated into the local languages . For the children and adolescent participants ( below 18 years ) , parents or guardians gave informed consent on their behalf . Animal experiments in this work were carried out in accordance with the local ethical approval requirements of the University of Edinburgh and the UK Home Office Animal ( Scientific Procedures ) Act ( 1986 ) under license number 60/4373 , or in Makerere University with approval of the College of Veterinary Medicine Animal Resources and Biosecurity research and ethics committee , with approval number SBLS/REC/16/137b . Samples were collected as described previously [17] during routine sleeping sickness diagnosis at Lwala hospital in the Kaberamaido district of North-Eastern Uganda . In order to confirm that all the cases were T . b . rhodesiense infections , PCR was carried out on the SRA gene as described in [49] . Up until sample 60 , both blood and CSF samples were centrifuged to obtain a cell pellet for CSF , and a buffy coat for the infected blood . These were either resuspended directly into Trizol , then frozen in liquid nitrogen , or the cells were stored in liquid nitrogen for later RNA extraction . For samples 61 onwards , both CSF and blood were placed directly into PAXgene tubes . For both PAXgene and Trizol , RNA was prepared according to the manufacturer's instructions . CSF cell counts were made directly from undiluted samples . For blood , cell counts were estimated at the clinic using thin smears stained with Giemsa . To convert these values to parasitaemias , we used blood from infected rats , counting parasites in diluted samples in a haemocytometer , and counting parasites from the same samples on dried smears . The results for human parasitaemias are therefore only approximations . Human RNA samples were initially checked for their integrity on the Agilent Bioanalyzer 2100 ( Agilent RNA Nano 6000 kit , 5067–1511 ) . The human blood samples showed considerable degradation . All samples ( blood and spinal fluid ) were prepared for sequencing using the Illumina TruSeq Total Stranded RNA preparation kit ( Illumina , RS-122-2301 ) . Between 75–750 ng total RNA was used as input material . rRNA depletion was performed on the samples dependent on their origin; those from blood were depleted with RiboGlobin ( Illumina ) , those from the spinal fluid with Ribo-Gold ( H/M/R ) ( Illumina ) . Since these kits are optimised for depletion of mammalian rRNA , most trypanosome rRNA remained in the sample . Due to the degradation of the samples , the binding time for depletion was increased to 5 minutes , and the subsequent fragmentation time was decreased from the normal 8 minutes to 3 minutes . PCR cycles were decreased from the recommended 15 to 13 cycles for the human samples . All human samples were processed with the same batches of Paxgene tubes and reagents for RNA handling and library preparation . Different batches were used for the rat samples . The finished libraries were equimolar pooled and sequenced with the Illumina NextSeq500 System , at the EMBL Genomics Core Facility , where 75 Single-end reads were generated ( Illumina , FC-404-2005 ) . The raw data are available at Array express under accession numbers E-MTAB-5293 and E-MTAB-5294 ( human ) and E-MTAB-6125 ( rat ) . The rat samples were sequenced 1–2 years after the human samples . Trypanosomes were purified on DEAE cellulose , then RNA was isolated using RNeasy column purification ( Qiagen ) with on-column DNase treatment according the the manufacturer's instructions . Total RNA samples were subjected to oligo ( dT ) -selection and paired-end sequencing at the Beijing Genomics Institute . RNAseq datasets were retrieved as FASTQ files . In case of paired-end data sets ( MB_A , MB_B , MAd ) , only one end per sample was analysed to ensure comparable results with single-end data . All sets were processed as follows: First , the overall read quality was investigated using FastQC [50] . Hereby , overrepresented sequences were identified , making up more than 0 . 1% of all reads in a set . Since this overrepresentation is not expected in a standard RNAseq experiment , and in our experience these sequences have often rRNA origin , they were removed using Cutadapt [51] . The resulting cleaned reads were then aligned to the T . brucei TREU927 genome ( release 9 . 0 ) using bowtie 2 [52] with a maximum mismatch count of one and each read was allowed to align to the genome up to 20 times . This helps in making sure that each read originating in a multi-gene family , aligns in each member , which is necessary for later subsetting for a unique gene list [29] . The alignment was used for read counting , and utilizing a custom script which is based on Samtools [53] to count the number of aligned reads within each annotated coding sequence . The whole process was automated using a python based pipeline [25] . For statistical analysis we used a unique gene list which holds single representatives for each multi-gene family , and unique genes ( total about 7000 genes ) [29] . All analysed data sets were combined in one read count table . Genes for which no data could be retrieved in some samples ( NA-values ) , were removed , and the rest was analysed using the R package DESeq2 [31 , 54 , 55] . The DESeq2 experimental design included only the affiliation of each sample to the original data set . The significance level alpha was set to 0 . 01 . Heat maps for overall comparison and co-regulation studies were generated using the rlog function of DESeq2 . The rlog function transformed the read counts and normalized the data to the sequencing depth and also shrank the effect size of genes with low read counts to prevent overestimation . The rlog transformed counts were then given to the pheatmap [56] package which was instructed to generate previously mentioned numbers of kmeans-clusters of all unique genes , to deduce the euclidean hierarchical clustering of the kmeans-clusters and the samples and to plot the final heat map . Each gene was annotated using a manually curated list ( see Supplementary Tables ) . Class enrichment within clusters was done using Fisher's exact test . Occurrence of each gene class within the studied cluster and within the unique gene list ( all genes were removed which had incomplete data in the read table ) was identified and for each class a two dimensional contingency table was generated . Fisher's exact test p-value for overrepresentation was calculated and corrected using the Benjamaini-Hochberg method for multiple testing [57] . PAC-YFP cassettes were integrated into the genome after cloning of suitable fragments into the plasmid p2675 [58] to direct homologous recombination [59] . All oligonucleotides used are listed in S4 Table . Clones were checked by Northern blotting . Total RNA was made using the RNAeasy Midi kit ( Qiagen ) or peqGold Trifast ( Paqlab ) . Poly ( A ) + RNA was selected using the Qiagen Oligotex mRNA kit . After denaturing formaldehyde gel electrophoresis , the RNA was subject to limited depurination ( 0 . 25M HCl , 15 min ) to ensure efficient transfer of longer mRNAs . Northern blots were hybridised with radioactive probes covering the whole YFP or PAC ORFs . For quantitative PCR ( RT-qPCR ) reverse transcription was done using random hexamer primers , with Superscript IV at 50°C , 15 minutes and the qPCR was done using LightCycler 480 SYBR Green I Master mix ( Roche ) or Luna Universal qPCR Master Mix ( NEB ) using LightCycler 480 II , Roche . Melting curves were checked using 95°C 10 s , 4 . 8°C/s; 65°C 1 min . , 2 . 5°C/s; 95°C 0 . 11°C/s . For the qPCR slightly different procedures were used . The protocol for LC480 master mix was; denaturation 95°C 1min . , 4 . 8°C/s; 45 amplification cycles of 95°C 20 s , 4 . 8°C/s , hybridization 60°C 20 s , 2 . 5°C/s , elongation 72°C 7 s , 4 . 8°C/s , 45 cycle . For Luna-Master Mix hybridization was for 30 s , and we used 40 cycles . Signals or measurements for YFP were normalized to those from PAC , to allow for differences in input RNA and for possible copy number variation . Then , one of the shortest mRNAs was used as a standard to calculate relative mRNA amounts . To estimate mRNA half-lives , we inhibited mRNA processing and transcription using sinefungin and Actinomycin D , and RNA was isolated 30 minutes later [40] . RNA from cells with and without inhibition was quantified by RT-qPCR .
|
African trypanosomes cause sleeping sickness in humans and various diseases of domestic and wild animals . Until now , all of our current knowledge of African trypanosome metabolism is based on results from trypanosomes grown in in vitro culture or in rodents . Drugs against sleeping sickness must however treat trypanosomes in humans . We here examine the way in which genes are expressed in human sleeping sickness trypanosomes from the blood and cerebrospinal fluid of human patients , and compare our results with those from trypanosomes growing in culture and rodents . Gene expression profiles of trypanosomes from human blood and cerebrospinal fluid were quite similar and there was no evidence for differences that might affect drug susceptibility . The RNAs in laboratory-adapted parasites grown in in vitro culture also quite closely resembled those in parasites from humans . The results showed that technical differences in the way RNA is made can have strong effects on measured gene expression profiles .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Methods"
] |
[
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"messenger",
"rna",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"trypanosoma",
"brucei",
"protozoans",
"genome",
"analysis",
"molecular",
"biology",
"techniques",
"rna",
"sequencing",
"research",
"and",
"analysis",
"methods",
"genomics",
"proteins",
"molecular",
"biology",
"biochemistry",
"rna",
"cytoskeletal",
"proteins",
"trypanosoma",
"eukaryota",
"blood",
"anatomy",
"nucleic",
"acids",
"physiology",
"transcriptome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"organisms"
] |
2018
|
Transcriptomes of Trypanosoma brucei rhodesiense from sleeping sickness patients, rodents and culture: Effects of strain, growth conditions and RNA preparation methods
|
Dengue viruses ( DENV ) are the causative agents of dengue , the world’s most prevalent arthropod-borne disease with around 40% of the world’s population at risk of infection annually . Wolbachia pipientis , an obligate intracellular bacterium , is being developed as a biocontrol strategy against dengue because it limits replication of the virus in the mosquito . The Wolbachia strain wMel , which has been introduced into the mosquito vector , Aedes aegypti , has been shown to invade and spread to near fixation in field releases . Standard measures of Wolbachia’s efficacy for blocking virus replication focus on the detection and quantification of virus in mosquito tissues . Examining the saliva provides a more accurate measure of transmission potential and can reveal the extrinsic incubation period ( EIP ) , that is , the time it takes virus to arrive in the saliva following the consumption of DENV viremic blood . EIP is a key determinant of a mosquito’s ability to transmit DENVs , as the earlier the virus appears in the saliva the more opportunities the mosquito will have to infect humans on subsequent bites . We used a non-destructive assay to repeatedly quantify DENV in saliva from wMel-infected and Wolbachia-free wild-type control mosquitoes following the consumption of a DENV-infected blood meal . We show that wMel lengthens the EIP , reduces the frequency at which the virus is expectorated and decreases the dengue copy number in mosquito saliva as compared to wild-type mosquitoes . These observations can at least be partially explained by an overall reduction in saliva produced by wMel mosquitoes . More generally , we found that the concentration of DENV in a blood meal is a determinant of the length of EIP , saliva virus titer and mosquito survival . The saliva-based traits reported here offer more disease-relevant measures of Wolbachia’s effects on the vector and the virus . The lengthening of EIP highlights another means , in addition to the reduction of infection frequencies and DENV titers in mosquitoes , by which Wolbachia should operate to reduce DENV transmission in the field .
Dengue fever is caused by an RNA virus belonging to the genus Flavivirus and is primarily vectored by the mosquito Aedes aegypti . Dengue viruses ( DENVs ) causes a spectrum of symptoms ranging from a mild fever to the life-threatening dengue shock syndrome [1–4] and collectively has become the most prevalent arthropod-borne viruses affecting humans today [5 , 6] . The geographic range of dengue is increasing largely due to human population growth and urbanization , especially in tropical and subtropical regions [2] . Approximately 390 million people from over 100 countries are estimated to contract dengue annually [7] . The suboptimal efficacy of a tetravalent dengue vaccine in recent phase IIb and phase III trials , and increasing insecticide resistance in mosquito populations has highlighted the urgent need to develop other alternative strategies to lessen the burden of this disease [8–10] . The use of the obligate endosymbiont Wolbachia pipientis has become a promising novel strategy to control dengue [9] . Wolbachia is a maternally inherited intracellular bacterium that is naturally found in a wide range of arthropod species including ~40% of all insect species [11] . Wolbachia is best known for its ability to induce diverse reproductive abnormalities in its hosts that result in its spread through invertebrate host populations [12] . A . aegypti does not carry Wolbachia naturally but has been stably transinfected with the bacterium [13–15] . In both semi-field cage experiments and more recently in field sites in Cairns , Australia , the wMel strain of Wolbachia has successfully invaded natural populations of A . aegypti , risen to near-fixation within a few months of release and remained established in those sites unaided [15 , 16] . In laboratory studies , Wolbachia infection in Aedes and Anopheles mosquitoes has been shown to interfere with replication of a broad range of pathogens including viruses , filarial nematodes , bacteria and malaria parasites [17–20] . However , there are exceptions to this Wolbachia–mediated antiviral property . For example , contrary to most systems , Wolbachia infection in a Culex species enhanced rather than inhibited West Nile virus infection [21] . The mechanism ( s ) underlying Wolbachia’s antiviral properties in the mosquito A . aegypti are only partially understood . It has been shown that Wolbachia primes the innate immune system of the symbiont [17 , 22] , competes for host resources critical for viruses [23] and manipulates the host viral defense pathways such as the microRNA pathway [24] . In laboratory-reared strains of wMel infected mosquitoes captured from field release regions , this antiviral activity , as measured by reduced infectivity of mosquitoes and reduced viral titers in tissues , remained strong even one year after field deployment [25] . This evidence bodes well for the long-term stability of the Wolbachia-based biocontrol effect against DENV . What remains is to test the ability of Wolbachia-infected mosquitoes to reduce transmission of human disease in a dengue endemic region . Such trials are currently underway in Vietnam and Indonesia [9] . Vectorial capacity , that is a quantitative measure of the efficiency of a vector-borne disease’s transmission , is determined by several factors including vector density , probability of a vector biting a human , vector competence , extrinsic incubation period ( EIP ) , and longevity [26 , 27] . EIP is the viral incubation period or delay between when a mosquito imbibes a dengue-infected blood meal and when the mosquito is capable of transmitting the virus to another individual . During the EIP , the virus must infect and escape the midgut , disseminate to the mosquito hemocoel and finally infect the salivary glands where it may be secreted into the saliva . EIP is a critical factor epidemiologically , as the earlier a pathogen is secreted through the saliva , the more humans the vector has the potential to infect over its lifespan [27] . This is particularly the case for A . aegypti as it tends to acquire multiple blood meals during a single gonotrophic cycle every 1–2 days [28] . Without a small animal model , the best proxy for infectiousness of the mosquito is the detection of the virus in mosquito saliva [29] . Measuring the transmission potential of mosquitoes is technically difficult due largely to the uncertainty of when a mosquito may feed and the small volume that mosquito salivate . Using a forced salivation method on pooled samples , wMel-infected mosquitoes were shown to be less likely to expectorate saliva-containing DENV compared to wild-type ( WT ) mosquitoes [15] at 14 days post-infection ( DPI ) . Here we use a non-destructive method to repeatedly sample pools of DENV-infected mosquitoes to assess the effect of wMel on the EIP of mosquitoes recaptured from the Wolbachia release site in Cairns , Australia . Our work demonstrates that wMel induces a delay in virus arrival in mosquito saliva . In addition , we show that wMel reduces the frequency that DENV can be detected in the saliva as well as its titer . Lastly , this study reveals that higher titers of DENV in the blood meal lead to shorter EIP , higher saliva DENV titers and reduced survival of the mosquitoes .
Two populations of mosquitoes were used in this study . WT mosquito ( not infected with Wolbachia ) eggs routinely collected from ovitraps outside the Wolbachia release zone in Cairns , Australia . wMel-infected ( wMel . F ) mosquito eggs were collected from Wolbachia release zone in mid 2012 . A . aegypti species identification was based on specific morphological characteristics . All mosquitoes were screened for Wolbachia infection using qPCR [16] . To retain genetic diversity all WT mosquitoes were used within 4 generations of the field . To prevent drift between the two lines and maintain genetic diversity , 20% of the males in the wMel . F line were replaced each generation with WT males . After hatching , larvae were reared at a standard density of 150 individuals per 3 L of distilled water in 30 x 40 x 8cm plastic trays and fed fish food ( Tetramin Tropical Tablets , Tetra , Melle , Germany ) until pupation . Pupae were transferred to 30 x 30 x 30cm cages to allow adult emergence at a density of approximately 400 individuals per cage . All mosquitoes were maintained in a controlled environment insectary at 25°C , ~70% relative humidity , with 12:12h light:dark cycle . Adults were allowed to feed on a 10% sucrose diet ad libitum . A dengue virus serotype 3 ( DENV-3 ) strain , which was originally isolated from a patient during the 2008/2009 outbreak in Cairns [30] was used in this study . It had been passaged five times in Aedes albopictus C6/36 cells to generate sufficiently high titer for infection . Virus was propagated , harvested and stored in single-use aliquots as previously described [31] . Virus stocks were titrated using plaque assays to a titer of 2 x 107 plaque forming units ( PFU ) /ml . This strain was selected for this study because it is capable of achieving a high titer to maximize the chance of infecting the mosquitoes following ingestion . The strain was also responsible for causing one of the largest outbreaks in recent history ( >900 cases ) in far north Queensland . Only 5% of circulating strains in the region have led to more than 100 cases of disease in a single outbreak with the next largest after the 2008/2009 outbreak causing >400 cases in 1997/1998 [30] . For infection , five to eight day old female A . aegypti were deprived of sucrose for approximately 18 h and then allowed to feed on a blood meal consisting of defibrinated sheep blood mixed with an equal volume of 2 x 107 PFU/ml to a final DENV concentration of 1 x 107 PFU/ml through a piece of desalted porcine intestine stretched over a water-jacketed membrane feeding apparatus for three hours at 37°C . Blood-engorged mosquitoes were sorted the following day under CO2 and placed in 250mL cups ( Sarstedt , Germany ) in groups of ten . The entire experiment was repeated with 1 x 106 PFU/ml of DENV . To ensure that virus in sheep blood remains infectious over the duration of the feed , aliquots of the mixture of DENV-3 with sheep blood were taken at zero and four hours in a separate experiment also carried out at 37°C . Live infectious virus was serially diluted and titered in the wells of a 96-well microtiter plate seeded with confluent monolayers of C6/36 cells . Plates were incubated for 10 days and fixed with PBS/acetone . Virus infection was identified in the fixed cell monolayers using a cell culture enzyme immunoassay [32] . The flavivirus-reactive monoclonal antibody , 4G2 ( TropBio , Townsville , Australia ) , was used as the primary antibody [30] . The titer as measure in tissue culture infective dose ( TCID50 ) of the DENV-3 remains stable over a four-hour period ( S1 Table ) . Saliva was collected from mosquitoes in groups of ten using a method that exploits the fact that mosquitoes expectorate virus when they sugar feed [33] . Blood-engorged mosquitoes were briefly anaesthetized under CO2 and placed in each 44mm x 55mm ( diameter x length ) 250ml polypropylene cups ( Sarstedt , Germany ) covered by a piece of 100% polyester curtain lace ( Spotlight Pty Ltd , Australia ) . Saliva was collected in 10 . 8mm x 46mm ( diameter x length ) 2ml polypropylene screw-cap tubes ( Sarstedt , Germany ) . The cap of a 2ml polypropylene screw-cap tube was attached to the bottom of the inside of the 250ml cup containing the mosquito using a small piece of adhesive plasticine ( Bostik , Thomastown , Vic , Australia ) . Two hundred μl of 10% sucrose solution was then pipetted into the cap; this was the only source of food and fluid for the mosquito . It was expected that mosquitoes would expectorate into the sucrose during feeding . Mosquito survival was recorded before each collection . During the collection , the mosquito was anaesthetized under CO2 and the pre-labelled body of the 2ml tube was carefully screwed onto the cap containing the 10% sucrose solution . The sealed tube was then removed from the cup , a new cap was affixed to the bottom of the 250ml cup , and new sucrose solution was pipetted into the cap . This method eliminates the requirement to pipette small volumes of expectorate and sucrose solution , which would likely to result in a loss of material . Tubes were then stored at -80°C . Collections were made every day from four to 14 days post-infection ( DPI ) and all samples were visually screened for mosquito body parts and blood spots from regurgitation or feces at collection . None were found . Pilot experiments were carried out to determine that DENV RNA would maintain its integrity in the 10% sucrose by spiking caps containing 200μl of 10% sucrose solutions with 10μl of serial dilutions of stock DENV , ranging from 107 to 103 copies of DENV . The spiked sucrose solutions were then held at 25°C , ~70% relative humidity and collected at day 0 , 1 , 2 and 5 . DENV remained detectable at least 5 days post-inoculation with no sign of decrease in RNA copy number ( S1 Fig ) . To determine how sucrose positive results correlated with other measures of DENV positivity we examined salivary glands and saliva via forced feeding . First DENV fed mosquitoes were provided food dye ( Queen fine food Pty . Ltd . Australia ) containing 10% sucrose solution in the standard cap assay four days after blood feeding . After a further 24 hours mosquitoes were visually inspected for food dye in their crops . Those that had fed on the sugar solution were then forced to salivate into a capillary tube [34 , 35] containing sterile RPMI1640 medium ( Invitrogen ) . This medium was used instead of the standard mineral oil because the oil was found to interfere with the efficacy of the downstream RNA extraction . With this modification salivation could not be confirmed visually by the formation of bubbles as is possible with oil . The midgut , head and salivary gland were dissected from the individual after salivation and prepared for DENV quantification . As a negative control , a tube cap containing sucrose was maintained at the bottom of the cup . The body of the tube was tightly screwed to the cap to prevent infected mosquitoes from salivating into the sucrose . All were negative for DENV . To determine the likelihood of occurrence of false positive in the sucrose feeding salivation assay due to contamination of mosquito fecal material , we carried out saliva collections with a modification to the mosquito housing . A false floor made of mesh ( opening 1 . 44 mm2 ) was laid over the bottom of the container preventing direct contact from the mosquito and the sucrose collection cap but allowing fecal matter to pass through . Mosquitoes were orally infected with live virus ( 4 X 106 genomic copies/mL ) and were given sucrose on the top of the housings covered in mesh using cotton wools . A total of 45 WT and 50 wMel . F dengue fed mosquitoes were held in these containers from 1 to 15 DPI . Sucrose cups were collected and checked for presence of DENV every two days . All housings were scored for visual evidence of fecal contamination below the mesh and sucrose caps that contained visible blood excluded from further analysis . Samples were thawed at room temperature and 200μl of Lysis Buffer containing 5 . 6 μg of Carrier RNA from a PureLink Pro 96 Viral RNA/DNA Kit ( Life Technologies , Carlsbad , CA , USA ) was added to each tube . Approximately 80μl of the sucrose solution evaporated during a day period between collections , resulting in an increase in viscosity of the remaining solution . We ensured that the sucrose solution was mixed well with the Lysis Buffer containing the Carrier RNA by shaking the tubes in a mini Bead Beater ( Biospec products , Bartlesville , Ok , USA ) for 1 min . The use of the Carrier RNA was essential as it both binds to viral RNA to increase its affinity for the silica matrix and reduces any viral RNA degradation from nucleases that may be present in the sample . Viral RNA was then extracted using the PureLink Pro 96 Viral RNA/DNA Kit according to manufacturer’s instructions . The EIP for each replicate was defined as the first time point at which DENV became detectable in saliva . A two-step approach was used to synthesize cDNA of DENV RNA and subsequent quantification using qPCR as previously published [31] . Viral titer was expressed as dengue virus copy number per part using absolute quantification . A 107-bp fragment from the 3’ UTR region of the DENV ( that is , in the same region that the primers amplify ) was amplified and cloned into the pGEM-T vector system ( Promega , Madison , WI ) . The plasmid was transformed into Escherichia coli , extracted using phenol-chloroform , and linearized by restriction enzyme digest . The copy number of the linearized plasmid was measured using the NanoDrop spectrophotometer . A standard curve of 106 , 105 , 104 , 103 , 102 , 50 and 10 of DENV fragment copies was constructed from a serial dilution . The limit of detection was set at 10 copies for this study , as it is the last dilution of the standard curve that amplified at least 95% of the time in 28 replicates . The concentrations of DENV in the samples were extrapolated from the standard curve and expressed as concentration per part by back calculating to the initial concentration of RNA [36] . Saliva volume was measured at five different ages ( 5 , 11 , 17 , 23 and 30 days ) spanning the lifetime of the mosquitoes to assess whether failure to detect DENV may be the result of low saliva production . Mosquitoes were starved for approximately 18 hours prior to undergoing forced salivation into mineral oil . The diameter of the saliva droplets was measured using an ocular micrometer at 40 X magnification . The volumes of the droplets were calculated using the sphere formula as previously published [34 , 37] . To determine if Wolbachia infection and/or the age of the mosquitoes affect feeding frequency , feeding rates were monitored in WT and wMel . F mosquitoes ( ~35 individuals each ) post DENV feed ( as above ) using food dye colored sucrose . At 5 , 11 17 and 24 DPI mosquitoes were randomly allocated 10% sucrose dyed either with 5% blue or pink food coloring ( Queen fine food Pty . Ltd . Australia ) . After 24 hours , mosquitoes were visualized under a dissecting microscope and then given the alternate colored sucrose solution . Visual inspections were then carried out every 24 hours . Once an individual mosquito was scored for the number of days required for the second dye to arrive in the abdomen it was removed from the assay . EIP was analyzed using a general linear model with an identity link function and normally distributed errors with Wolbachia infection status , DENV titer of bloodmeal and their interactions as factors . Saliva volume and feeding frequency data were analyzed using a general linear model with an identity link function and normally distributed errors . Mosquito age and Wolbachia infection status were tested as factors . The number of days infective and saliva DENV titer were analyzed using Mann-Whitney U tests due to deviation from normality . Saliva DENV titer was analyzed using a generalized linear model with a logit link function and a normal error distribution . Survival data were analyzed using Kaplan-Meier analysis and log-rank statistics . Statistical analysis was performed with the software Statistica 8 . 0 ( Statsoft , Inc . USA ) .
There was very high agreement and significant correlation between sucrose feeding salivation assay positive for DENV and both DENV infection of the salivary gland and forced salivation results ( Table 1 ) . In only one case was a sucrose feeding salivation positive result not recapitulated by the DENV infection in the salivary gland . The midgut and head tissues from this same mosquito were positive , however , which may indicate a possible failure of the RNA extraction for the salivary gland tissue . The sucrose feeding assay identified a greater number of positive samples than forced salivation ( 10 vs 6 ) . In three out of four the cases the salivary gland , head and midgut of these mosquitoes were also DENV positive . These differences are not surprising given that the forced feeding assay is a poor baseline control for salivation for several reasons . First , without visual confirmation of salivation due to the requirement of using medium for collection purposes not all of the mosquitoes will have participated in forced salivation . From our saliva volume assay with forced salivation in mineral oil ( n = 547 ) we estimated that mosquitoes fail to expectorate 13 . 2% of the time . Second , forced expectoration can produce incredibly small volumes , from 0 . 11 to 23 . 63 nl . Despite flushing the capillary tubes with the medium the entire contents is cannot be completely collected . Third , because the mosquitoes were forced to expectorate after the 24 hr sucrose collection window there may be real differences in whether they are secreting virus and hence positivity of samples . The design of the fecal contamination assay was effective as feces were present below the mesh in 100% of the housings . The rate of DENV positivity in sucrose due to strict fecal contamination ranged from 0% to 4 . 8% but averaged 0 . 88 and 0 . 25% for WT and wMel , respectively , across all time points ( S2 Fig ) . Two different titers of DENV blood meal were used to orally infect the mosquitoes: a high titer blood meal of 107 PFU/ml and a lower concentration of 106 PFU/ml . The EIP for each replicate was defined as the first time point at which DENV became detectable in the sucrose feeding solution . Overall there was an significant main effect of Wolbachia infection ( Fig 1 ) with wMel . F experiencing a lengthier EIP compared to Wolbachia-free WT mosquitoes ( df = 1 , F = 11 . 6 , P<0 . 01 ) . There was also a significant difference in EIP due to blood meal titer ( df = 1 , F = 10 . 3 , P<0 . 01 ) . There was no significant Wolbachia by blood meal titer interaction effect ( df = 1 , F = 0 . 84 , P = 0 . 36 ) on EIP . When mosquitoes were infected with the high titer blood meal , the mean EIP of wMel . F was 6 . 0 ± 0 . 58 DPI as compared to 4 . 9 ± 0 . 21 DPI in WT mosquitoes . When mosquitoes were infected with a 106 PFU/ml DENV blood meal , the mean EIP of wMel . F was 7 . 8 ± 0 . 77 DPI compared to 5 . 9 ± 0 . 29 DPI in WT mosquitoes ( Fig 1 ) . This difference suggests that the wMel slows the arrival of virus in the saliva , particularly at a lower orally infected virus concentration . A biological replicate ( a pool of mosquitoes in a cup ) was considered infective for a particular time point if DENV was detected in the sucrose solution on that day . Overall , wMel . F exhibited fewer infective days compared to WT mosquitoes ( Fig 2 ) . When mosquitoes were infected with a 107 PFU/ml DENV blood meal , the presence of wMel reduced the median number of days infective from 6 days to 1 day ( Z = 4 . 47 , P<0 . 0001 ) . The same differential was seen when mosquitoes were infected with 106 PFU/ml DENV blood meal ( Z = 3 . 48 , P<0 . 001 ) . The presence of wMel also reduced the amount of DENV in the mosquito saliva . After a high DENV titer blood meal ( Fig 3A ) , the median copy number of DENV in wMel . F mosquito saliva was 2633 copies across all timepoints measured , a 4 . 9 fold decrease as compared to the median copy number of 12826 in WT mosquitoes ( Z = 3 . 51 , P<0 . 001 ) . It is also noted that no DENV was detected in saliva from wMel mosquitoes after 11 DPI . This is likely a stochastic result due to low infection frequencies ( unique to wMel ) and declining population sizes due to age associated mortality . When mosquitoes were infected with a low titer blood meal ( Fig 3B ) , the presence of wMel lowered ( 2 . 6 fold ) the median copy number of DENV in mosquito saliva from 198 copies to 76 copies ( Z = 2 . 08 , P<0 . 05 ) . The amount of virus in the mosquitoes’ saliva was positively correlated with the titer of DENV used to infect the mosquitoes . WT mosquitoes expectorated more DENV when infected with a high DENV titer blood meal as compared to a low titer ( Z = 12 . 3 , P<0 . 0001 ) . The same effect of DENV titer in the blood meal on saliva DENV titer was observed in wMel . F mosquitoes ( Z = 3 . 78 , P<0 . 001 ) . The DENV copy number varied hugely by almost 4 logs from 507 to 2458333 copies in WT mosquitoes fed with a 107 PFU/ml DENV blood meal . While there are trends in titer across DPI overall the effect of day was not significant ( df = 1 , F = 0 . 95 , P = 0 . 33 ) . The presence of wMel infection lengthened the lifespan of mosquitoes as compared to WT following DENV infection . This effect was significant for mosquitoes fed both high ( P<0 . 001 , Fig 4A ) and low ( P<0 . 0001 , Fig 4B ) DENV titer blood meals . The titer of DENV in the blood meal was inversely correlated with mosquito survival . WT mosquitoes infected with 107 PFU/ml of DENV died more quickly than those infected with the lower titer of 106 PFU/ml ( P<0 . 001 ) . This suggests that DENV infection is costly to mosquitoes and that Wolbachia is providing some protection to the host . Using each cup of 10 mosquitoes as a biological replicate , we assessed the dynamics of infection of the mosquitoes over time . After a high DENV titer blood meal ( Fig 5A ) , WT mosquitoes quickly became infective and peaked at 88 . 9% infective ( 16/18 ) at 6 DPI . In contrast , wMel . F mosquitoes remained largely non-infective with a maximum of 31 . 6% ( 6/19 ) of the replicates expectorating detectable DENV at 5 DPI . A similar trend was seen when mosquitoes were fed a low titer blood meal ( Fig 5B ) . The proportion of replicates infective declines as the mosquito age even when most of the replicates are infective at earlier timepoints . To determine if the dynamics in saliva DENV titer and infectivity can be explained by mosquito saliva production we compared the saliva volume of the two mosquito populations over time in their natural DENV uninfected state ( Fig 6 ) . wMel . F mosquitoes produced less saliva as compared to WT mosquitoes ( df = 1 , F = 55 . 3 , P<0 . 0001 ) . The age of the mosquitoes was also a determinant of saliva volume ( df = 1 , F = 17 . 5 , P<0 . 0001 ) . In WT mosquitoes , saliva volume increased as mosquitoes aged until 17 days of age and then declined . In wMel . F mosquitoes saliva volume remained low throughout the lifespan of the mosquitoes . The saliva volume also varied hugely between individuals especially on day 17 ( ranged 0 . 11–23 . 63nL ) in WT mosquitoes . The dynamics of expectorated saliva volume mirror that of DENV titer and the dynamics of the infective mosquitoes , all peaking between 5 and 11 DPI . Together , the results suggest that wMel . F mosquitoes expectorate less DENV , which can at least be partly explained by producing less saliva and older mosquitoes are less likely to expectorate detectable DENV as their saliva volume deceases with age . To determine if Wolbachia infection and/or age of the mosquitoes can affect the sucrose feeding frequency , we compared the feeding frequency of the two mosquito populations over time in their DENV infected state ( Fig 7 ) . wMel . F mosquitoes fed more frequently ( shorter intervals ) as compared to WT mosquitoes ( df = 1 , F = 4 . 8 , P<0 . 05 ) . Mosquitoes also fed more often as they aged ( DPI effect: df = 1 , F = 8 . 0 , P<0 . 01 ) . There was no effect of food dye ( df = 1 , F = 0 . 074 , P = 0 . 79 ) .
One year after wMel-infected A . aegypti field deployment in Cairns , Australia , there is no sign of attenuation in its ability to limit viral replication in the body of the mosquito and reduce viral dissemination to the head of the mosquitoes [25] . Using mosquitoes captured from the same field sites we show that the Wolbachia-mediated blocking effect is translatable to saliva-based measures of vector competence . Our findings show that the presence of wMel not only reduces the proportion of mosquitoes with transmission potential but also delays their EIP , thus further reducing the capacity of the mosquitoes to transmit dengue . Additionally , for the first time we present data that shows that wMel significantly reduces the frequency of mosquitoes that expectorate DENV and lowers virus titer in the saliva . We also offer a potential mechanism demonstrating that wMel infected mosquitoes produce little saliva . Ideally , the saliva of individual mosquitoes instead of pools would be assayed for the presence of DENV in order to better understand variation at the individual level , for infection status and saliva DENV titer . However wMel reduces viral replication so effectively only a small portion of the mosquitoes display dissemination to the head tissue compared to WT mosquitoes , 6 vs 62% , respectively , as reported by previous studies with DENV-2 [15 , 25] . A subset of mosquitoes exhibiting dissemination may then eventually acquire transmission capability [38] . Rearing and producing infections in sufficient numbers of individuals to overcome the strength of blocking is simply intractable . This is especially the case for low-titer blood meals of 106 PFU/ml as on average only ~5% of replicates were infective at any time point . Another limitation of this study is the reliance on qPCR to detect and quantify virus . qPCR is a sensitive and efficient method to quantify DENV , however it does not differentiate infectious from non-infectious virus [36] . In contrast , plaque assays , which require much larger volumes of starting material , quantify infectious particles . While there is a strong correlation between the two measures , the estimated RNA copy number is usually 2–5 logs higher than the number of infectious units due to the presence of immature and/or defective virus [39 , 40] . It is therefore possible that our estimates of EIP are skewed towards earlier timepoints that could lead to an over-estimation of the mosquitoes’ transmission potential . Indeed , when using a plaque assay laboratory strains of wMel mosquitoes infected with DENV-2 92-T strain were found to completely block dengue transmission as none of the pooled ( 0/36 ) mosquito saliva had detectable infectious dengue virus at 14 DPI [15] . However , the relatively poor sensitivity of the plaque assay in combination with the destructive nature of the mosquito saliva collection process made it impossible to repeatedly assay mosquito saliva from the same individual for infectious virus . It is also clear that the DENV-2 92-T strain is much less infectious in mosquitoes as compared to DENV-3 isolate used in this study . In WT mosquitoes an average blood feeding event produces ~60 percent of mosquitoes with disseminated infections [31] whereas in pilot studies our DENV-3 isolate achieved an infection rate near 100% . Future studies should examine the generality of DENV blocking and effects on transmission parameters for diverse viral genotypes including representatives of the other three serotypes . Our estimate of mean EIP for this dengue strain in these mosquitoes ranges from 4–10 days , which is shorter than a previously estimated mean of 15 days ( 5 to 33 days at a 95% confidence ) at 25°C for dengue viruses in general [41] . These findings from our sucrose collection assay are confirmed by our validation assay and other studies have argued that EIP of DENV is shorter than is commonly perceived . For example , DENV antigen was detected in the salivary gland in more than a third of mosquitoes examined as early as 4 DPI [42] , and a small fraction of mosquitoes were found to have naturally leaky abdominal midguts that could facilitate rapid systemic infection [43] . Furthermore , our measure of EIP is in line with estimates of the total incubation period ( EIP + Intrinsic incubation period ) of 9–11 days obtained from patient records from the 2008/2009 outbreak during which this dengue isolate was collected [30] . After infection , DENV titer is highly dynamic within the insect [39] . A previous study examining the kinetics of DENV in the whole mosquito showed that midgut titer increases , peaks and eventually declines as mosquitoes age [42] . DENV dynamics in mosquito saliva is rarely studied due to its intractability . In one study using plaque assays on forced salivation on individual mosquitoes , it was found that the proportion of mosquitoes expectorating DENV decreased in older ( 21 days ) mosquitoes as compared to young ( 6 days ) mosquitoes despite 100% body infections . This suggests that old mosquitoes do not salivate detectable DENV [44] , a finding that concurs with measures for our WT strain where infections decline in saliva after 8 DPI . The effect of wMel on saliva production and feeding frequency , however , was more surprising given that the strain is considered to have few fitness effects on the host . More specifically , wMel has no effect on the fecundity and egg viability of the mosquitoes but reduces the mean lifespan of the mosquitoes by approximately 10% [15] . Another strain of Wolbachia wMelPop , in A . aegypti radically reduces the mean lifespan by more than 40% and causes late acting blood feeding defects previously showed reductions in saliva production in mosquitoes at 26 and 35 days of age but not at 5 days [37] . Our study shows similar results with no effect at 5 days but differences beginning at 11 days . These data suggest that , even in the absence of other virulent effects , Wolbachia infection in mosquitoes may affect saliva production and mosquito behavior such as feeding frequency . Using direct mosquito feeding on dengue patients , the infectious dose or the titer for each of the four DENV serotypes required to orally infect mosquitoes ranges from 6 . 3 to 7 . 5 log10 RNA copies/mL of patient plasma , although viremia in some patients may exceed 10 log10 [45] . We found that the high viral infectious dose is associated with shorter EIP and higher DENV titer in the mosquito saliva . This suggests that the transmission potential of the mosquito may be directly influenced by human host viremia levels . We also show the ability of wMel to offer protection against DENV , slowing the time to death and increasing survival . Whether this lengthening of lifespan can potentially enhance the vectorial capacity of wMel mosquitoes requires further study . Future experiments should also be focused on the effect of patients with high viremia on EIP , especially titers that would not be achievable in the laboratory using cell culture methods . The relationship between saliva volume and DENV titer has not been studied in A . aegypti . In the case of Plasmodium falciparum , saliva volume is positively correlated with the number of parasites in the salivary gland [46] . We argue that , in general , the dynamics of infective mosquitoes and DENV titer in saliva can at least be partly explained by the dynamics of saliva volume . The fact that older mosquitoes produce less saliva suggests that there is a window during their lifetime when infected mosquitoes may expectorate virus during a bite . This also points out possible shortfalls in formulas used to calculate vectorial capacity of mosquito-borne disease that do not take into consideration the epidemiological variation in virus expectoration with mosquito age . These findings are particularly important where the likely efficacy of Wolbachia in reducing DENV transmission is being modeled in advance of field trials . Environmental factors such as temperature and larval nutritional status affect the length of EIP [47] . Recently , the diurnal temperature range ( DTR ) , which reflects the degree of daily fluctuating temperature , was found to significantly change the outcome of infection and survival of the mosquitoes , and the EIP of DENV as compared to a constant temperature [48] . Similar fluctuating temperature at different mean baselines was also found to affect Plasmodium development and dissemination in Anopheles that demonstrated the generality of the effect of temperature on parasite transmission potential [49] . Wolbachia density , which is shown to be a determinant of viral blocking [50] , is affected by temperature too [51] . Varying constant temperatures was found to alter the extent of parasite blocking in somatically Wolbachia transinfected Anopheles mosquitoes [52] . This raises the importance of not confining the study of vector-pathogen interactions to a constant rearing temperature of 25°C . As large-scale field releases of Wolbachia infected mosquitoes are currently underway in sites each with its unique baseline temperature and diurnal temperature range , it is paramount to understand how the tripartite interactions between mosquitoes , dengue virus and Wolbachia may change in terms of temperature regimes . Gut microbiota of the mosquitoes was known to influence the outcome of vector by pathogens [53 , 54] . How Wolbachia may impact the mosquito microbiota and thus influence the outcome of vector-borne pathogens also need to be investigated . In conclusion , we found that the wMel infection lengthens the EIP of DENV , reduces the frequency that the virus is expectorated and decreases the amount of DENV RNA copy number in saliva as compared to wild-type mosquitoes . A reduction in saliva production in wMel mosquitoes can at least partially explain the above observations . These saliva-based traits offer more disease relevant measures of the symbiont’s effects on virus than using measures such as infection and dissemination . The shift in EIP , in particular , indicates an additional means by which Wolbachia could modulate virus transmission . The opposing effects of wMel prolonging mosquito survival post DENV infection needs further investigation .
|
Dengue is endemic in more than 100 countries and is transmitted by the mosquito Aedes aegypti . The use of the symbiotic bacterium Wolbachia has become a potential biocontrol approach against dengue virus for two reasons . First , Wolbachia spreads rapidly through populations by manipulating host reproduction to its advantage . Second , Wolbachia limits viral replication in the mosquito by competing with the virus for essential host resources . Following field release in Cairns , Australia in 2011 , the wMel strain of Wolbachia has successfully invaded wild mosquito populations , infecting nearly all individuals . To test whether limited dengue replication in wMel mosquitoes translates to a reduction in dengue transmission potential , we used a non-destructive assay to repeatedly quantify dengue virus in mosquito saliva . We found that wMel significantly delayed the time it took for mosquito saliva to become infectious , reduced the frequency of dengue virus that was expectorated by mosquitoes and lowered the virus titer in mosquito saliva . We also showed that wMel infection suppresses saliva production in mosquitoes that may , in part , explain our findings . The saliva-based nature of the work provides a more accurate assessment of Wolbachia’s ability to limit disease transmission and suggests that Wolbachia may have positive impacts on transmission not only by reducing the number of infectious mosquitoes in a population but also delaying the arrival of virus in the saliva .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Wolbachia Reduces the Transmission Potential of Dengue-Infected Aedes aegypti
|
Fungi are a large group of eukaryotes found in nearly all ecosystems . More than 250 fungal genomes have already been sequenced , greatly improving our understanding of fungal evolution , physiology , and development . However , for the Pezizomycetes , an early-diverging lineage of filamentous ascomycetes , there is so far only one genome available , namely that of the black truffle , Tuber melanosporum , a mycorrhizal species with unusual subterranean fruiting bodies . To help close the sequence gap among basal filamentous ascomycetes , and to allow conclusions about the evolution of fungal development , we sequenced the genome and assayed transcriptomes during development of Pyronema confluens , a saprobic Pezizomycete with a typical apothecium as fruiting body . With a size of 50 Mb and ∼13 , 400 protein-coding genes , the genome is more characteristic of higher filamentous ascomycetes than the large , repeat-rich truffle genome; however , some typical features are different in the P . confluens lineage , e . g . the genomic environment of the mating type genes that is conserved in higher filamentous ascomycetes , but only partly conserved in P . confluens . On the other hand , P . confluens has a full complement of fungal photoreceptors , and expression studies indicate that light perception might be similar to distantly related ascomycetes and , thus , represent a basic feature of filamentous ascomycetes . Analysis of spliced RNA-seq sequence reads allowed the detection of natural antisense transcripts for 281 genes . The P . confluens genome contains an unusually high number of predicted orphan genes , many of which are upregulated during sexual development , consistent with the idea of rapid evolution of sex-associated genes . Comparative transcriptomics identified the transcription factor gene pro44 that is upregulated during development in P . confluens and the Sordariomycete Sordaria macrospora . The P . confluens pro44 gene ( PCON_06721 ) was used to complement the S . macrospora pro44 deletion mutant , showing functional conservation of this developmental regulator .
Fungi ( Eumycota ) are a group of eukaryotes that are present in almost all habitats; therefore they do not only play a great role in nature , but also influence human life in many ways [1] . About 100 , 000 fungal species have been described , but it is estimated that the actual number might exceed 1 . 5 million [2] . The largest group among the Eumycota is the Ascomycota ( or ascomycetes ) , which comprise the Saccharomycotina , Taphrinomycotina , and Pezizomycotina . The former groups contain many unicellular species ( yeasts ) or species that develop only few hyphae or develop hyphae only under certain conditions ( dimorphic fungi ) , whereas the Pezizomycotina are generally filamentous fungi capable of producing highly differentiated multicellular structures , the most complex of which are fruiting bodies for the protection and dispersal of sexual spores [3] , [4] . The most basal groups of Pezizomycotina are the Pezizomycetes and the Orbiliomycetes that form open fruiting bodies called apothecia with exposed meiosporangia ( asci ) . Phylogenetically derived groups ( e . g . Sordariomycetes , Eurotiomycetes and Dothideomycetes ) mostly differentiate closed fruiting bodies where the asci develop within and protected by mycelial structures [3] , [5]–[8] . In the last decade , genomes of many filamentous ascomycetes have been sequenced and are invaluable for the analysis of the evolution of species as well as for understanding physiological and morphological properties of fungi . In fact , fungi are among the groups of eukaryotes with the highest number of sequenced genomes to date ( http://www . ncbi . nlm . nih . gov/genome/browse/ ) , largely because they include many model organisms , species of medical , agricultural or biotechnological importance . In addition , they usually have compact genomes with short introns and relatively few repetitive regions or non-coding DNA compared to plants and animals , thus making genomic analysis less complex . However , while there are at least ten genome sequences available for each of the more derived groups ( Sordariomycetes , Leotiomycetes , Eurotiomycetes and Dothideomycetes ) , only one Orbiliomycete and one Pezizomycete genome have been sequenced , namely those of a nematode-trapping fungus , Arthrobotrys oligospora , and the black truffle , Tuber melanosporum , respectively [9] , [10] . A . oligospora ( teleomorph Orbilia auricolor [11] ) belongs to a group of nematode-trapping soil fungi that comprises only a few known species , which are mostly analyzed for their ability to develop specialized trapping structures , while fruiting body formation is not well studied in this group . The 40 Mb genome of A . oligospora encodes ∼11 , 500 protein-coding genes , similar to the size and coding capacity of other ascomycete genomes [10] . In contrast , the 125 Mb genome of T . melanosporum is much larger than those of other sequenced ascomycetes , but contains fewer protein-coding genes . This genome expansion is mostly due to a large number of transposable elements that make up 58% of the truffle genome [9] . Truffles are symbiotic fungi that form mycorrhizal interactions with plant roots; and it has been noted that a biotrophic life-style , either as symbiont or pathogen , is often correlated with an increase in genome size , e . g . caused by repetitive sequences , in many fungi [12] . Furthermore , truffles have a highly specialized fruiting body that is adapted to growth within the soil , in contrast to fruiting bodies of almost all other filamentous fungi , which develop above ground . Thus , even though the truffle genome is of great interest for both ecological and economic reasons , it is difficult to distinguish between features that are ancestral with respect to the filamentous ascomycete lineage , specifically with respect to fruiting body formation , versus features that are adaptations to the truffle-specific life style , i . e . adaptations to mycorrhizal symbiosis or to below-ground fruiting body development . Consequently , the genome sequence of another member of the Pezizomycetes with fruiting body development that is more typical of filamentous ascomycetes will be of great value for evolutionary and morphogenetic analyses . To fill this gap , we sequenced and analyzed the genome and development-dependent transcriptomes of the Pezizomycete Pyronema confluens . P . confluens was established as a model organism for the analysis of cell biology and fruiting body development in filamentous ascomycetes during the first half of the 20th century . It was instrumental in the elucidation of the dikaryotic phase during sexual development of higher ascomycetes [13]–[17] . In the last decade , P . confluens was used in comparative studies of gene expression during sexual development of ascomycetes [6] , [18] , [19] . It is a soil-living saprobe found in forests in temperate climates . In nature , its fruiting bodies ( apothecia ) usually appear on the ground after forest fires [20] . Under laboratory conditions , P . confluens has a short life cycle where typical apothecia containing eight-spored asci are formed within six days ( Figure 1 ) . This is rare among members of the Pezizomycetes , many of which do not easily reproduce sexually in the laboratory . Also , P . confluens is homothallic ( self-fertile ) , therefore no crossing partner of different mating type is needed for the fungus to complete its sexual cycle [17] . In addition , P . confluens can also be used to study the effects of light on fruiting body formation , because in contrast to many other filamentous ascomycetes , fruiting body development in this fungus is strictly light-dependent [6] , [21] . A previous analysis based on sequence data from 15 proteins showed that the P . confluens lineage is positioned at the base of the filamentous ascomycetes in a phylogenetic tree [6] . Phylogenomic analysis based on the genome data from this study and including sequences from T . melanosporum and A . oligospora confirms this basal position with the Pezizomycetes as sister group to the Orbiliomycetes ( Figure 2 ) . In two previous small pilot studies , P . confluens was used for comparative expression analyses to identify genes with evolutionary conserved expression patterns during fruiting body development in ascomycetes [6] , [19] . These studies already indicated that gene expression patterns during development might be conserved even over large evolutionary distances . Here , we sequenced the genome and development-dependent transcriptomes of P . confluens with the following objectives: ( i ) To work towards closing the sequence gap among basal filamentous ascomycetes , and , thus , to learn more about the evolution of fungal genomes . ( ii ) To use the genome and transcriptome data to study the biology of a basal filamentous ascomycete in comparison with more derived species , with a focus on sexual development .
The genome of the P . confluens strain CBS100304 was sequenced with a combination of Roche/454 and Illumina/Solexa sequencing similar to what was described previously for Sordaria macrospora [22] . A summary of the sequence reads that were used for the P . confluens assembly is given in Table S1 . The final assembly consists of 1 , 588 scaffolds ( 1 , 898 contigs ) with a total size of 50 Mb , a scaffold N50 of 135 kb and a GC content of 47 . 8% ( Table 1 ) . To estimate the genome size independently of the assembly , k-mer analyses based on the Illumina/Solexa reads were performed using an algorithm described for the potato genome [23] . The analysis resulted in one clear peak , as one would expect for haploid genome ( Figure S1 ) . Based on the analysis for different k-mer lengths ( 31 and 41 ) , a total genome size of ∼50 . 1 Mb was predicted which is close to the total length of the assembly . For transcriptomics , we performed RNA-seq for three different conditions: sexual development ( sex ) , long-term culturing in the dark ( DD ) , and a mixture of different vegetative tissues ( vegmix ) , in two biological replicates per condition ( Table S1 ) . For each condition , RNA from different time points was pooled to represent a high number of genes that are expressed during the corresponding condition ( see Materials and Methods for details ) . RNA for sex samples was extracted from mycelia grown in minimal medium in surface culture in constant light . Only under these conditions is P . confluens able to develop fruiting bodies , whereas growth in darkness , submerged , or in complete medium prevents sexual development . We used RNA from 3d , 4d , and 5d old mycelia to cover the initial stages of sexual development up to the development of young fruiting bodies ( Figure 1 ) . The DD samples comprised RNAs from mycelia grown in minimal medium in submerged culture in constant darkness , which prevents fruiting body formation . The vegmix samples also contained only RNAs from mycelia that could not develop fruiting bodies , but from a mixture of growth conditions different from the DD condition . We argue that the use of different mycelia sharing the common denominator of “no fruiting bodies” would allow us to focus on genes that are differentially expressed during fruiting body morphogenesis by comparing the three different conditions . Thus , genes that are differentially regulated in the comparisons sex/DD and sex/vegmix , but not DD/vegmix are candidates for genes that are regulated in a sexual development-dependent manner . RNAs from different growth conditions were also used to allow a high read coverage of as many genes as possible for annotation purposes as described previously for Sordaria macrospora [24] . Therefore , gene model predictions were performed ab initio as well as evidence-based on the RNA-seq data ( details in Materials and Methods ) . The output from different gene prediction pipelines was merged using MAKER [25] . Gene model predictions were scanned for consistency , and ∼10% of the predicted gene annotations were corrected manually to improve the exon/intron structure . To address the question whether the assembly and annotation cover the complete gene space of P . confluens , we performed a BLASTP search with a eukaryotic core gene set as previously described [26] . All of the 248 single-copy core genes were present in the P . confluens predicted peptides , suggesting that the assembly covers the complete core gene space . Untranslated regions ( UTRs ) were also modeled by the gene prediction pipeline , and were refined based on the manually curated gene set using custom-made Perl scripts as described previously [24] . For the current assembly , 13 , 369 protein-coding genes are predicted with an average CDS length of 1 , 093 and an average transcript length of 1 , 483 bases ( Table 1 , Table S2 ) . The median length of 5′ and 3′ UTRs are 156 and 200 bases , respectively , similar to findings in T . melanosporum , Aspergillus oryzae and S . macrospora [24] , [27] , [28] . Furthermore , we predicted 605 tRNAs , and assembled an rDNA unit comprising the 18S , 5 . 8S and 28S rRNA genes as well as the internal transcribed spacers ( ITS ) 1 and 2 . Based on this annotation , the majority of the RNA-seq reads map to exonic regions as expected ( Figure S2 ) . Spliced sequence reads identified in RNA-seq mapping results can not only be used to improve the exon/intron structures of predicted genes , but also to address the question of natural antisense transcripts ( NATs ) , because the consensus sequences at the 5′ and 3′ ends of introns allow strand determination even in non-strand-specific RNA-seq data . NATs can play a role in the regulation of gene expression , and were found to be pervasive in metazoans [29] , [30] . To identify putative NATs in P . confluens , we extracted predicted splice sites in antisense orientation to annotated genes from the mapping results obtained with Tophat [31] . Antisense splice sites covered by at least five spliced sequence reads , and with a coverage of more than 10% of the average coverage of the predicted sense-transcript were checked manually to remove splice sites that were most likely due to annotation errors , within repeat-rich regions , or close to sequence gaps . This yielded 376 antisense splice sites in 281 genes ( Table S3 ) , indicating that natural antisense transcripts are present in P . confluens . The number of genes with NATs is most likely underestimated , because we set stringent criteria , and non-spliced antisense transcripts could not be discovered by this analysis . In T . melansporum , only 33 NATs were identified from RNA-seq data by gene modeling , but this low number most likely also is an underestimate due to the method used [27] . Few studies have addressed the presence of NATs at a genome-wide scale in filamentous fungi , and the number of NATs that were identified in the ascomycetes Aspergillus flavus , Aspergillus niger , and Magnaporthe grisea [32]–[34] , and in the basidiomycetes Ustilago maydis , Coprinopsis cinerea , and Schizophyllum commune [35]–[37] is in a range similar to our findings in P . confluens . Thus , NATs appear to be present across all groups of filamentous fungi , even though they do not seem to be as pervasive as in metazoans . When compared with the genome of its closest sequenced relative , T . melanosporum [9] , the P . confluens genome is much smaller ( 50 Mb versus 125 Mb ) , but contains nearly twice the number of protein-coding genes ( 13 , 369 versus 7 , 496 ) . The large size of the truffle genome is mostly due to an expansion of transposons and other repeated elements , with nearly 58% ( ∼71 Mb ) of the genome consisting of repeats larger than 200 bp [9] . Repeat analyses based on similarity to known repeat classes as well as de novo repeat finding showed that in P . confluens , transposable elements longer than 200 bp constitute only 12% ( ∼6 Mb ) of the genome , with low complexity regions and simple repeats constituting an additional 0 . 21 and 1 . 03% , respectively ( Table S4 ) . Very few of the repeats show a high percentage of sequence identity with the repeat consensus sequences , indicating a high degree of divergence among repeats ( Figure S3 ) . This is different from findings in T . melanosporum , Fusarium oxysporum , and Pyrenophora tritici-repentis [9] , [38] , [39] , but similar to Nectria haematococca [40] . One reason for this finding might be that long perfect repeats ( longer than the Roche/454 read lengths of 300–400 bp ) were lost in the assembly process; however , this would not apply to shorter perfect repeats . Also , the k-mer analysis showed a single main peak corresponding to a genome size of 50 Mb ( Figure S1 ) , and did not show any major extra peaks , indicating that most of the genome sequence is represented in the assembly . Thus , the presence of repeats with dissimilar sequences suggests that the repeats are evolutionary rather old , and that genome defense processes are active in P . confluens . A search for genes that might be involved in chromatin modification and other silencing processes showed that P . confluens comprises gene sets similar to those in other fungi where genome defense processes were identified . Interestingly , there is even a slight expansion in some families of putative RNA interference genes ( Table S5 ) . This suggests that in its current state , the P . confluens genome is well-protected against repeat spreading , and that the high repeat content of the T . melanosporum genome is not characteristic for all members of the Pezizales . Among the genes that might be involved in genome defense is PCON_06255 , a homolog of the N . crassa rid gene . In N . crassa , the cytosine DNA methyltransferase RID is essential for repeat induced point mutation ( RIP ) [41] . The RIP process introduces C∶G to T∶A mutations in duplicated sequences of more than 400 bp and at least 80% sequence identity during the sexual cycle , and thus is a means to control the spread of duplicated sequences including transposons [42]–[44] . The rid homolog PCON_06255 is slightly upregulated during sexual development ( Table S2 ) , consistent with a role for the gene product under these conditions; however , an analysis of the genomic DNA for the presence of regions that might have been subjected to RIP yielded only a small fraction ( 0 . 46% ) of the genome ( Table S6 ) . Thus , RIP does not seem to play a major role in genome defense in P . confluens . Overall , with respect to genome size , gene number , and transposon content , the P . confluens genome is more similar to those of many higher filamentous ascomycetes than to the truffle genome , indicating that these features in truffle might be a consequence of the specialized life style . It has been shown previously that the number of introns per gene varies greatly between different fungal lineages , with filamentous ascomycetes harboring one to two introns per kilobase [45] . However , at the time of the study , no Pezizomycete genome sequences were available , therefore we performed an analysis of intron content in P . confluens and T . melanosporum in comparison with nine other filamentous fungi representing major fungal lineages ( Figure 3 ) . The comparison included 747 genes for which orthologs could be identified in all analyzed fungi . About 7% of the analyzed P . confluens genes do not contain introns , similar to chytrids and zygomycetes , and less than in most higher ascomycetes where 10–12% of the investigated genes are intron-free . This is also much less than in hemiascomycete yeasts , e . g . S . cerevisiae , where only a minority of genes contains introns [45] . However , it is almost twice the number found in basidiomycetes , where less than 4% of the analyzed genes do not contain introns . The average number of introns per kb is 1 . 83 in P . confluens , which is larger than that of the higher filamentous ascomcyetes with the exception of Aspergillus nidulans , but still in the same range as found in the previous study [45] . Interestingly , T . melanosporum has 2 . 49 introns per kb , the highest value of all investigated ascomycetes . These data might indicate that within filamentous ascomcyetes there is a tendency towards net intron loss that is more pronounced in the evolutionary derived lineages than in the basal Pezizomycetes . However , the exceptionally high intron content of T . melanosporum might also be life style-specific or connected to the high repeat content in this fungus . The availability of transcriptome data for P . confluens also allowed us to compare overall expression levels across the genome . In a previous study with metazoans , genes could be grouped in two classes with high and low expression levels , respectively , independent of species , tissue type , or type of experiment [46] . In fungi , this has been addressed only in the filamentous ascomycete Sordaria macrospora where the situation is different , because there were up to three expression peaks depending on the conditions analyzed [24] . Interestingly , the situation in P . confluens is more similar to metazoans , with two main peaks that represent high and low expression in all three conditions investigated ( for details see Text S1 and Figure S4 ) . The closest relative of P . confluens with a sequenced genome is T . melanosporum , therefore , we used the MUMmer package [47] to determine regions of sequence similarity and possible synteny , i . e . the order of genes within the genome , between the two species . However , even though both species belong to the order Pezizales , there is little sequence similarity at nucleic acid level ( data not shown ) , therefore we used the PROmer algorithm from the MUMmer package to compare the in silico-translated genomic sequences ( Figure 4 ) . Even at amino acid level , only ∼11% of the P . confluens genome align with the T . melanosporum genome , compared to more than 66% in the highly syntenic genomes of S . macrospora and N . crassa ( Figure 4A and B ) . A dot plot analysis of the PROmer results also indicated a low degree of overall synteny between P . confluens and T . melanosporum ( Figure S5A ) . In a second analyis of synteny , we identified orthologs for all P . confluens genes in the predicted proteomes of ten filamentous fungi by reciprocal BLAST analysis [48] , and used the positions of orthologous proteins on sequenced chromosomes or contigs to determine synteny ( Figure S5B ) . A dot plot for this comparison of P . confluens with T . melanosporum also shows that there is little overall synteny , in contrast to the comparison of the highly syntenic genomes of N . crassa and S . macrospora . The low degree of synteny between P . confluens and T . melanosporum might be explained by their large evolutionary distance . Estimation of divergence times of the Pyronema and Tuber lineages using r8s [49] placed their most recent common ancestor at least 260 Mya ( million years ago ) , nearly twice the time of the estimated divergence of Tuber from its sister groups within the Tuberaceae ( ∼156 Mya , [50] ) . Nevertheless , an analysis of the number of syntenic gene pairs or gene triplets showed that although the number of syntenic pairs and triplets in P . confluens versus T . melanosporum is still lower than in N . crassa versus S . macrospora , it is much higher than in comparisons of P . confluens with other ascomycetes ( Figure 4C ) . This indicates that many regions of microsynteny exist between P . confluens and T . melanosporum whereas overall chromosomal synteny was lost . In contrast , in comparisons of P . confluens with basidiomycete , zygomycete , or chytrid species , we found few syntenic gene pairs or triplets . In filamentous ascomycetes , the master regulators of sexual reproduction are the various genes that reside at the mating type ( MAT ) loci [51] . They encode transcription factors that regulate the sexual cycle . Heterothallic ascomycetes have a bipolar mating type system , with isolates possessing one of two non-allelic versions ( idiomorphs ) of a single MAT locus , termed MAT1-1 and MAT1-2 [52] . The MAT1-1-1 and MAT1-2-1 genes encode transcription factors with a conserved alpha domain and high-mobility group ( HGM ) -domain , respectively [51] . Conversely , homothallic ascomycetes carry both MAT loci within a single genome , with the two loci either fused together , located within close proximity , or on separate chromosomes [53]–[57] . BLASTP searches with MAT1-1-1 proteins of different filamentous ascomycetes revealed the presence of a MAT1-1-1 gene ( PCON_07491 , scaffold 329 ) encoding a putative transcription factor with an alpha domain that is most similar to the MAT1-1-1 protein of T . melanosporum [58] ( Table S7 ) . In the majority of Sordariomycetes , two other genes are also located in the MAT1-1 locus: the MAT1-1-2 gene encoding a protein with a PPF domain harboring the three invariant residues proline ( P ) , proline and phenylalanine ( F ) , and the MAT1-1-3 gene encoding a protein with a high-mobility-group ( HMG ) domain as a DNA-binding motif [55] , [59] . Homologs of these two MAT1-1-specific mating type genes could not be identified in the genome of P . confluens . A BLAST search with MAT1-2-1 HMG domain mating type proteins identified the ORF PCON_08389 ( scaffold 381 ) encoding a HMG domain protein as a putative MAT1-2-1 homolog . The encoded protein displayed the highest degree of identity to the mating type protein MAT1-2-1 of Gibberella indica ( Table S7 ) . The genes APN2 , encoding a putative DNA lyase , and SLA2 , encoding a cytoskeleton assembly control factor , have been reported to be adjacent to MAT loci in many filamentous ascomycetes [55] , [59]–[62] . The MAT1-1 locus of P . confluens is flanked by genes encoding proteins of unknown function ( Figure 5 , Table S7 ) . A homolog of APN2 ( PCON_08385 ) is located 10 kb upstream of the MAT1-2 locus . A SLA2 homolog ( PCON_02178 ) is also present in the P . confluens genome but neither on scaffold 329 ( MAT1-1 ) nor on scaffold 381 ( MAT1-2 ) . Recently the genes flanking the mating type locus of the Pezizomycete T . melanosporum have been identified as GSTUMT00001088001 and GSTUMT00001092001 at the left flank and at the right flank , respectively [58] . Only a homolog of GSTUMT00001088001 is conserved in P . confluens ( PCON_01243 , scaffold 1068 ) but is not located adjacent to the mating type genes . Interestingly , PCON_08391 at the right flank of MAT1-2 and PCON_07490 at the left flank of MAT1-1 encode proteins with a high degree of similarity ( 59 . 2% identity in 1005 amino acids overlap ) , and for both proteins the closest homolog in T . melanosporum is GSTUMT0008232001 ( Figure 5 , Table S7 ) . Aspergilli also contain only one copy of this gene . Therefore , one might hypothesize that this gene was duplicated in P . confluens during a recombination event that led to the presence of both mating type loci in one genome and therefore to homothallism . Indeed the phylogenomic tree in phylomeDB shows that PCON_08391 and PCON_07490 are species-specific paralogs . In summary , the genome of the homothallic P . confluens has two putative MAT loci , typical for homothallic filamentous ascomycetes . The MAT1-1 and MAT1-2 loci encode an alpha domain and an HMG domain transcription factor , respectively . The mating type loci are not fused and not in close proximity , similar to the situation in several Eurotiomycetes [63] . Only the MAT1-2 locus is flanked by the conserved APN2 gene , whereas both MAT loci are flanked by a pair of paralogous genes not found in this location in other ascomycetes ( Figure 5 ) . In this respect , P . confluens is more similar to derived filamentous ascomycetes than to its closest sequenced relative , T . melanosporum , where the MAT loci are not flanked by either APN2 or SLA2 . Thus , the P . confluens MAT loci might reflect a putative intermediate state between the relatively conserved genomic arrangement of mating type loci found in higher filamentous ascomycetes and the specific arrangements found in T . melanosporum . We also searched for homologs to other proteins involved in sexual development and signaling , e . g . pheromone and pheromone receptor genes as well as genes involved in pheromone processing and downstream signaling . With the exception of pheromone genes , which are weakly conserved in ascomycetes , conserved genes for sexual development were found in P . confluens , too ( for details see Text S2 and Table S7 ) . A quantitative analysis of gene expression across the P . confluens genome was performed based on RNA-seq ( Table S2 ) . Of the 13 , 369 annotated genes , only 58 were not expressed ( i . e . have no RNA-seq reads mapped to their exon sequences ) in any of the conditions tested . Our analysis was focused on development-dependent gene expression with one tested condition allowing sexual development ( sex ) , and two conditions that only allow the formation of vegetative mycelium ( DD and vegmix ) . Genes that are regulated mostly by developmental factors should be differentially expressed in the comparisons sex/DD and sex/vegmix , but not DD/vegmix . Of the predicted 13 , 369 protein-coding genes , 5 , 565 ( 41% ) are regulated differentially in at least one of the three comparisons with thresholds >2 or <0 . 5 , and 3 , 616 genes ( 27% ) are differentially regulated in at least one comparison with thresholds >4 and <0 . 25 . With the less stringent thresholds , 506 ( 4% ) genes are downregulated and 1 , 804 ( 13% ) genes are upregulated in both sex/DD and sex/vegmix , but not differentially regulated in DD/vegmix; with the more stringent thresholds , the numbers are 229 ( 2% ) and 1 , 460 ( 11% ) genes . Studies in animals and plants have shown that genes associated with sexual reproduction evolve more rapidly than genes with other functions [64] , [65] . This can be observed not only for single genes , but across genes that are transcriptionally expressed in organs involved in sexual reproduction [66] . In fungi , few studies have addressed this question so far . A genome comparison of Candida and related yeast species showed that meiotic genes undergo rapid evolution [67] , and similar findings were made in studies of mating type and pheromone signaling genes in filamentous ascomycetes [68] , [69] . A recent EST analysis of Neurospora intermedia and comparison with other Neurospora species indicated that sex-associated genes , i . e . those genes that are preferentially expressed during sexual development , are rapidly evolving in fungi , too [70] . Here , we approached this question from a different angle by analyzing gene expression levels for P . confluens genes with different degrees of evolutionary conservation to find out if genes with different lineage-specificities are preferentially expressed under any of the conditions that we investigated ( Figure 6 , Table S8 ) . First , we extended our orthology analysis as described in the previous section to include the predicted proteomes of 14 fungi from the major fungal groups ( chytrids , zygomycetes , ascomycetes , basidiomycetes ) , adding two ascomycetous yeasts and an additional chytrid to the previous dataset ( Table S8 ) . For downstream analysis , only genes without hits ( orphan genes ) or with clear reciprocal BLAST hits ( orthologs ) were used . Genes that are members of gene families with more than one paralog where clear orthologs could not be determined in the analyzed genomes were excluded from the analysis . This left 6 , 706 P . confluens genes in the final analysis that were sorted in six lineage-specific groups ( a–f ) ranging from P . confluens orphan genes ( the largest group with 5 , 737 genes ) to genes that are conserved in all analyzed fungal genomes ( Figure 6 , Table S8 ) . For these genes , we analyzed derived peptide lengths , and expression in the three conditions sex , DD , and vegmix . Median peptide lengths were shortest in the orphan genes ( Figure 6A ) . This is consistent with observations from a broad range of organisms where conserved proteins are on average longer than poorly conserved proteins [71] . However , at least some of the short peptides may be artifacts of annotation problems , because no homology-based information to aid annotation was available for these genes . But while some of these genes might be due to spurious annotation , more than 40% of the orphan genes are differentially expressed in at least one condition , which might indicate some functionality ( Figure 6B ) . Interestingly , more than 20% of the orphan genes are upregulated during sexual development ( in sex/DD and sex/vegmix ) , while less than 2% are downregulated . This percentage of differentially expressed genes as well as genes upregulated during sexual development is much higher than in all the other groups where less than 5% of genes are upregulated during sexual development . Furthermore , the percentage of up- and downregulated genes is not much different in the other groups . These trends were also observed when this analysis was performed with lineage-specificity groups obtained from phylogenomics analysis ( Figure S6 , Tables S8 and S9 ) . This analysis allowed the differentiation between P . confluens orphan genes , Pezizales-specific genes , and genes that are specific to Pezizales and Orbiliales; and an increase in peptide lengths as well as a decrease in the percentage of genes that are upregulated during sexual development is correlated with decreasing lineage specificity ( Figure S6 ) . The expression trends can also be seen when analyzing overall expression levels as measured by RPKM ( reads per kilobase per million counted reads ) values ( Figure 6C , Figure S6C ) . Overall median expression is significantly lower for orphan genes than for the other groups , with a strong increase in Pezizales-specific genes and further slight increase in genes specific to filamentous ascomycetes . A general trend for more conserved genes to have higher expression has also been observed in other organisms [72] , [73] . When looking at RPKM values in the three conditions tested , there are no significant differences between conditions within the lineage-specific groups with the exception of the orphan genes . In this group , the median expression is significantly higher in the sexual development condition . Thus , while overall expression is lower for orphan genes , this group comprises much more genes with specific expression during sexual development than the more conserved groups . There are several hypotheses to explain this finding . One would be the above-mentioned rapid evolution of sex-associated genes leading to apparent orphan genes in its most extreme form . This should be observed especially in species where no sequence information is available for close relatives , as is the case with P . confluens . Increased evolutionary rates have generally been observed in genes with higher lineage-specificity , independent of putative function , in an analysis of seven ascomycete genomes [74] . However , there are other mechanisms that may lead to the presence of orphan genes . One is gene loss in all but one ( observed ) species , although this is unlikely to occur on a larger scale , i . e . for thousands of genes in a single species . Another is the de novo gene birth from previously non-coding sequences , a process that in recent years was acknowledged as probably being more common than previously thought [75] , [76] . One might speculate that the high number of sex-associated orphan or less conserved genes in P . confluens indicates that sexual development allows the “testing” of novel gene-inventions . This might be feasible in filamentous fungi where sexual development is usually not the only means of propagation , and therefore novel genes that are deleterious for sexual reproduction under some circumstances might be retained by purely vegetative propagation until more compatible conditions occur . Another reason could be a more general trend for less conserved genes to be involved in group- or species-specific processes as was found in an analysis of gene expression during different stages of vegetative growth and conidiation in N . crassa [77] . Further analyses of more species and transcriptomes from different conditions will be necessary to address these questions . It has been shown in many fungi that light can cause developmental changes . In several ascomycetes , illumination promotes vegetative reproduction via conidiospores , while sexual reproduction is observed in darkness [78] . In contrast to this , it was noted already at the beginning of the last century that the formation of apothecia in P . confluens is light-dependent [13] , [21] , and this was confirmed by our studies . Both constant illumination ( LL ) as well as a 12 h photoperiod promote fruiting body formation , whereas in constant darkness ( DD ) , P . confluens is sterile ( Figure S7 , Figure 7 ) . This complete light-dependency of fruiting body development is uncommon in ascomycetes . It was discussed that perithecia formation in Trichoderma reseei might be activated by light [79] , but it was shown later on that sexual development in this fungus can also be observed in darkness [80] . However , light-dependent fruiting body development was observed in several basidiomycetes , e . g . Schizophyllum commune and Coprinopsis cinerea , and in both species , blue light constitutes the effective part of the visible spectrum [81]–[83] . We found that this is also the case for P . confluens where blue ( 400–500 nm ) but not green or red light allows fruiting body formation ( Figure 7A ) . This confirms an early study from the 1920s that found wavelengths of 400–550 nm to promote sexual development in this fungus [21] . Despite the fact that no phenotypic responses to wavelengths other than blue have been observed in P . confluens yet , its genome encodes putative photoreceptors that cover a range of wavelengths ( Table 2 ) , and most of these genes are expressed ( Figure 7 , Table S10 ) . This includes homologs of N . crassa WC-1 and WC-2 ( PCON_03119 and PCON_05086 ) , transcription factor/photoreceptor proteins that are part of the white collar complex and mediate all blue-light responses [84]–[88]; however , no homolog was found for the VVD protein that functions in light adaptation in N . crassa and T . reseei [80] , [89] , [90] . Two putative phytochromes ( PCON_06747 and PCON_08526 ) are present in the P . confluens genome , the first of which is orthologous to FphA that was shown to be the photoreceptor mediating repression of sexual development by red light in A . nidulans [91] . P . confluens also encodes an ortholog of the N . crassa opsin-related protein ( PCON_01637 ) that lacks a conserved lysine residue for chromophore binding , but has no homolog of the rhodopsin NOP-1 , a putative green-light receptor [92]–[94] ( Table 2 ) . BLAST searches in the T . melanosporum genome also failed to identify a NOP-1 homolog , suggesting that the Pezizales might lack a gene for this type of photoreceptor . A gene encoding a putative cryptochrome is present in P . confluens ( PCON_04132 , Table 2 ) , but was the only putative photoreceptor gene for which no expression could be detected under the conditions tested . All other putative photoreceptor genes are expressed , and induced by white light in long-term illumination experiments ( 4 d LL versus 4 d DD , Figure 7B ) , but only moderately or not induced by short-term light pulses ( from 5 to 60 min , Figure 7C ) . Interestingly , we also observed some induction with green light for several of the genes , which is surprising because P . confluens lacks a rhodopsin-type receptor that was hypothesized to mediate green-light responses in fungi [92] . However , the green-light filter we used has a slight transmission wavelength overlap with the blue light filter ( Figure S8 ) , therefore at least part of the green-light responses might be mediated by residual blue light . Nevertheless , P . confluens seems to have some green-light sensitivity , because the orp gene is light-induced in the short- and long-term illumination experiments , and this effect is stronger with green light than with blue light ( Figure 7 , B and C ) . We also looked for homologs to genes that act downstream of photoreceptors in light signal transduction in other fungi . Interestingly , P . confluens contains a frequency ( frq ) homolog ( PCON_09365 ) , and thus is the most distant relative of N . crassa in which this gene is found . The frq gene encodes the main regulator of circadian rhythmicity in N . crassa , and is a direct target of the white collar complex , but so far frq homologs were only found in Sordariomycetes , Dothideomycetes , and Leotiomycetes [95] , [96] . The identification of a frq homolog in P . confluens suggests that frq was present in the ancestor of filamentous ascomycetes and was lost several times during evolution , because the T . melanosporum genome does not contain a frq homolog ( data not shown ) , and no homolog has been detected in the Eurotiomycetes [95] , [96] . Similar to N . crassa [84] , [97] , the P . confluens frq is strongly induced by short light pulses , and this reaction is mainly mediated by blue light ( Figure 7C ) . frq was also upregulated in the long-term illumination experiments ( Figure 7B ) . In N . crassa , an antisense transcript of frq is also upregulated by light , and is involved in light-dependent resetting of the circadian clock [98] . Analysis of antisense splice sites did not show any NATs for the P . confluens frq , although we cannot exclude the possibility of non-spliced antisense transcripts with this analysis . However , a splice site analysis of RNA-seq reads for frq indicated that there might be an alternatively spliced intron in the sense direction overlapping the predicted start codon of the open reading frame ( Figure S9 ) . In N . crassa , an intron overlapping the frq start codon is alternatively spliced resulting in two different forms of the FRQ protein [99] , [100] . To determine whether a similar mechanism might occur in P . confluens , we performed RT-PCR analysis of a region covering the predicted alternatively spliced intron . Interestingly , there are indeed two different transcript , and alternative splicing of the intron is light-dependent with an increased ratio of spliced versus non-spliced transcript in the light ( Figure S9 ) . In the N . crassa frq , alternative splicing of the AUG-covering intron is temperature-dependent [99] , [100] , therefore it seems that similar principles might be at work in these two distantly related fungi , but with different input signals . We also identified an ortholog for the GATA-type transcription factor NsdD/SUB-1/PRO44 ( PCON_06721 ) that was shown to mediate late light responses in N . crassa [101] and is essential for sexual development in A . nidulans , Aspergillus fumigatus , N . crassa and S . macrospora [102]–[105] . The developmental function of this protein appears to be conserved ( see later section ) , and expression analyses indicate that light responses , and therefore regulatory activities , might also be similar to those in higher fungi , because the P . confluens pro44 is strongly light induced after long- and short-term illumination ( Figure 7 ) . We also searched for homologs to carotenoid biosynthesis genes that are known to be light-induced in other fungi [78] , [106] , [107] . P . confluens encodes homologs to the three enzymes AL-1 , AL-2 , and AL-3 from N . crassa [108]–[111] , and the corresponding al genes ( PCON_03421 , PCON_03432 , and PCON_05718 ) are strongly light-induced under short- and long-term illumination ( Figure 7B , C ) . Under long-term illumination , blue light has an even stronger effect than white light on al gene expression . It has been shown in a previous biochemical analysis that the orange-pinkish pigments that characterize P . confluens cultures grown in white or blue light ( Figure 7A ) are carotenoids [112] . Thus , assuming a function of the al genes in P . confluens similar to that in N . crassa , it seems likely that these carotenoids are synthesized by the products of the al genes . Analysis of cultures grown on complete medium that does not support fruiting body formation even in the light indicates that pigment synthesis is independent of fruiting body formation , because light-grown cultures are pigmented even in the absence of apothecia formation ( Figure S7 ) . In summary , our data suggest that light-sensing and signal transduction in P . confluens might be comparable to mechanisms in the distantly related species N . crassa , and might thus be conserved to a large degree in filamentous ascomycetes . However , output from the light-signaling pathways might be somewhat different , because fruiting body development is strictly light-dependent in P . confluens , but not in most other ascomycetes . Blue light has the strongest effect on both morphological as well as gene expression phenotypes , but our data hint at sensitivity to other wavelengths , too , especially in the green part of the visible spectrum . Similar findings were made in S . macrospora , where phenotypic changes were observed in response to green light , and in N . crassa , where gene expression was found to be modified in mutants of the putative green-light receptor NOP-1 [22] , [113] . We searched for conserved protein domains in the predicted proteins from P . confluens and seven other filamentous fungi to identify protein families that are expanded in P . confluens ( Table S11 ) . Among the expanded gene families are two that encode mostly small , extracellular proteins , namely CBM_14 and Defensin_2 domain proteins . The CBM_14 domain ( Chitin binding Peritrophin-A domain ) is mainly found in metazoa , and in fungi so far has been described only in the Avr4 protein from Cladosporium fulvum and , with one gene per genome , in the genomic sequences from several Aspergilli [114] , [115] . BLASTP analysis in GenBank revealed that there are also some predicted CBM_14 proteins in other Eurotiales ( Figure 8A ) , but not in other fungal groups . In contrast , there are 13 proteins with CBM_14 domain in P . confluens ( Figure 8A ) . All of these are predicted as extracellular and have a putative cleavable N-terminal signal peptide for co-translational insertion into the ER ( data not shown ) . They are mostly 80–140 amino acids long with the exception of PCON_04108 ( 351 amino acids ) , and contain no other recognized domains besides CBM_14 . Some of the corresponding genes are clustered within the same genomic region , indicating that the genes might have arisen through duplications at certain gene loci: PCON_09939 , PCON_09940 , PCON_09946 , and PCON_09947 lie within 20 kb of scaffold 486 , and PCON_05983 and PCON_05987 lie within 9 kb on scaffold 228 . The RNA-seq data show that more than half of the CBM_14 domain proteins are upregulated during sexual development . Furthermore , the overall expression levels of these proteins vary greatly ranging from no sequence reads in certain conditions to >13 , 000 reads ( normalized to kb of mRNA , Figure S10 ) . For example , the clustered genes PCON_09939 to PCON_09947 are preferentially or only expressed during sexual development , while others are more strongly expressed under non-sexual conditions . To address the expression of some CBM_14 domain genes in more detail and distinguish between regulation by sexual development , light , and growth conditions ( surface versus submerged ) , we performed qRT-PCR for four genes under the conditions LL , DD , LLK and DDK ( light and darkness in surface culture and submerged culture , Figure 8B ) . These combinations distinguish if a gene is differentially regulated during sexual development ( i . e . in LL/DD and in LL/LLK ) , or regulated by light ( i . e . in LL/DD and LLK/DDK ) or regulated by surface versus submerged growth ( i . e . in LL/LLK and DD/DDK ) . PCON_04108 and PCON_09947 are upregulated during sexual development , but not consistently regulated by light or surface culture ( PCON_09947 is slightly upregulated by light , but the extent of regulation is far lower than the development-dependent regulation ) . PCON_09794 is downregulated during sexual development , but not regulated by the other two stimuli , whereas PCON_06720 is downregulated during sexual development and ( to a lesser extent ) downregulated in the light and in surface cultures . Overall , the four genes have distinct expression patterns , which might indicate that they are functional in P . confluens . The Avr4 protein from the phytopathogenic C . fulvum was shown to bind chitin and protect it from hydrolysis by plant chitinases [115] . P . confluens is non-pathogenic , but one might speculate that secreted CBM_14 domain proteins might protect the fungus from microbial attacks in its soil habitat . A second expanded gene family encoding small , secreted proteins is the Defensin_2 family ( Table S11 ) . Defensin_2 domain proteins are mostly known from arthropods where they are part of the immune system and act against bacteria [116] , [117] . In fungi , only one Defensin_2 domain protein has been described in detail , namely Plectasin from Pseudoplectania nigrella , a member of the Pezizales [118] , [119] . In P . confluens , the family comprises six members ( PCON_01606 to PCON_01611 ) , all of which are between 92–96 amino acids long , including predicted signal peptides of 15–23 amino acids , and are encoded by a cluster of genes within 10 kb on scaffold 1117 ( Figure S11A and B ) . Two of the genes ( PCON_01607 and PCON_01611 ) are pseudogenes , and interestingly the expression of the two pseudogenes is much lower than that of the other four genes under all conditions investigated by RNA-seq ( Figure S11C ) . Both pseudogenes have one of the functional Defensin genes as closest homolog ( Figure S11D ) , and one might speculate that they have arisen from gene duplications within the Defensin gene cluster , but lost ( most of ) their expression and function which was retained by their closest homologs . Interestingly , all genes including the pseudogenes are downregulated during sexual development ( Figure S11C ) . However , a comparison of intergenic regions showed that these are not conserved , in contrast to the coding regions . On the one hand , this might indicate that the regulatory sequences responsible for development-specific regulation are too small or non-conserved to be detected in these comparisons . Another explanation could be that regulation is achieved through chromatin organization of the complete gene cluster , similar to what was described for secondary metabolism gene clusters in fungi [120] , [121] . In addition to the Defensin_2 domain proteins in P . confluens and P . nigrella , searches in other sequenced fungal genomes identified Defensin_2 domain proteins only in Eurotiomycetes ( Figure S8D ) . This phylogenetic distribution is similar to that of the CBM_14 domain proteins . One might speculate that both classes of small , secreted proteins arose from horizontal transfer events from insects into fungi; alternatively , this could be a case of gene loss or rapid evolution in the other ascomycete groups . Horizontal gene transfer has been acknowledged as an important mechanism in fungal evolution only in recent years , and the availability of genome sequences has made in-depth analyses possible [122] . A transfer event of carotenoid biosynthesis genes from fungi into insects has been shown already [123] , therefore it seems possible that a transfer in the reverse direction might also occur . Interestingly , both the Defensin_2 as well as the CBM_14 domain proteins might be involved in defense mechanisms against microorganisms in arthropods . One might hypothesize that the corresponding genes were acquired by fungi through horizontal gene transfer , and were retained because they offer a selective advantage in the microorganism-rich soil habitat . A third gene family that is expanded in P . confluens compared to T . melanosporum ( but not to other ascomycetes ) comprises genes with HET ( heterokaryon incompatibility protein ) domains . While there are only two HET-domain containing protein in T . melanosporum [124] , there are 15 in P . confluens ( Table S11 ) . However , there are 11–101 predicted HET domain proteins in various species of higher filamentous ascomycetes , therefore the number in T . melanosporum might have been reduced by selective gene loss . In the Sordariomycetes N . crassa and Podospora anserina , HET domain proteins were shown to mediate heterokaryon incompatibility ( HI ) [125] , [126] . Homologs to many known HI proteins can be found in P . confluens ( Table S12 ) ; however , none of the P . confluens HET-domain proteins contains additional WD repeat , NACHT , leucine- or glycine-rich repeat domains that are found in the HET domain in HI proteins from N . crassa and P . anserina [127] . Therefore , if HI is present in P . confluens , it is unlikely to be mediated by the same proteins that regulate HI in Sordariomycetes . A number of protein families are have fewer members or are missing in P . confluens compared to more derived ascomycete groups ( Table S11 ) . Most prominent among these are gene families involved in secondary metabolism ( see next section ) , transporter protein families , and several gene families involved in carbohydrate metabolism . The low number of genes for some transporter families might be connected to the limited capabilities for the production ( and presumably export ) of secondary metabolites; whereas the reduced number of genes for certain enzymes involved carbohydrate metabolism might either be a niche-specific adaptation or indicate that the expansion of carbohydrate-activating enzymes occured only in more derived ascomycete groups . Overall , gene family contraction in P . melanosporum is much less prominent than in T . melanosporum [9] . In contrast to the gene families described in the previous section , genes encoding enzymes for the biosynthesis of polyketides or non-ribosomal peptides , typical secondary metabolites of filamentous ascomycetes , are underrepresented in the P . confluens genome ( Figure S12A ) . There are seven putative non-ribosomal peptide synthase ( NRPS ) genes , and one polyketide synthase ( PKS ) gene , much fewer than in the genomes of higher filamentous ascomycetes [128]–[130] . The predicted NRPS protein PCON_02859 has the typical domain structure of siderophore NRPSs and is part of a cluster of genes homologs of which are involved in siderophore biosynthesis in other fungi [131] , [132] ( Figure S12B ) . A second putative NRPS gene ( PCON_07777 ) is not clustered and does not have homology to NRPSs with known function . The remaining five NRPS genes all have a domain structure that is typical for alpha-aminoadipate reductase ( AAR ) -type NRPSs ( Figure S12A ) , and ( with the exception of PCON_04030 ) all have high sequence similarity to aminoadipate semialdehyde dehydrogenase , an enzyme of lysine biosynthesis that is conserved in fungi [129] . Most fungi have only one AAR-type NRPS [129] , therefore the high number of corresponding genes in P . confluens is somewhat unusual . Possible explanations may be selective amplification of this specific gene family or loss of most other NRPS genes with exception of AAA-type NRPS genes . However , at least PCON_04030 might have a function other than lysine biosynthesis , because the gene is located adjacent to the single PKS gene ( PCON_04029 ) in a gene cluster that also contains other genes encoding enzymes that might be involved in the biosynthesis of secondary metabolites ( Figure S12C ) . The genes in this cluster might be involved in the production of a hybrid polyketide/non-ribosomal peptide . The existence of gene clusters encoding separate PKS and NRPS proteins that act in a common biosynthetic pathway was demonstrated , for example , in A . nidulans , where such a cluster is responsible for the production of Emericellamide [133] . The single predicted PKS PCON_04029 is a type I PKS . In filamentous ascomycetes , there is usually one type III PKS encoded in the genome [130] , but type III PKSs are missing in P . confluens . A low number of PKS and NRPS genes was also found in T . melanosporum , and therefore seems to be typical for lower filamentous ascomycetes rather than a result of the truffle-specific life-style [9] . In summary , our analysis shows that the presence and clustering of NRPS and PKS genes is already established in P . confluens . In combination with the fact that no PKS genes and only the single AAR-type NRPS gene were found in Taphrinomycotina , Saccharomycotina , and zygomycete genomes [130] , this suggests that the evolution and expansion of PKS and NRPS gene families began in a common ancestor of filamentous ascomycetes , whereas the evolution of type III PKS genes might be a later event that occurred in higher filamentous ascomycetes . However , at present it cannot be excluded that the low number of putative PKS and NRPS genes is an adaptation to specific ecological niches in both T . melanosporum and P . confluens [134] , [135]; more Pezizomycete genome sequences will be needed to resolve this question . The number of putative transcription factor genes ( excluding general transcription factors that regulate RNA polymerase ) in filamentous fungi varies from 182 in N . crassa to more than 600 to 800 in several Fusarium species [39] , [136]–[138] . In truffle , 201 transcription factor genes were predicted [139] , while our survey of the P . confluens genome indicated 177 putative transcription factor genes ( Table 3 , Table S13 ) . Similar to other filamentous ascomycetes , the largest group comprises putative Zn2-Cys6 binuclear cluster ( Zn cluster ) proteins; thus , the regulatory capacity of P . confluens appears to be similar to that of other filamentous fungi . 54 of the putative transcription factor genes are differentially expressed in at least one of the comparisons that were investigated ( Figure S13 ) . Eight genes are strongly upregulated during sexual development , and among these is PCON_02619 , the gene encoding the ortholog of STE12 , a transcription factor that was shown to be involved in sexual development in yeast and filamentous ascomycetes [140]–[143] . Expression of PCON_02619 and five additional transcription factor genes was characterized in more detail by qRT-PCR ( Figure S14 ) . Development-dependent expression was confirmed for those genes that were predicted to be differentially regulated during sexual development by the RNA-seq analysis , showing that our sampling strategy is indeed suitable for identifying developmentally regulated genes . The STE12 ortholog PCON_02619 was confirmed as one of the most strongly upregulated transcription factors during sexual development . Interestingly , the corresponding SteA gene in truffle is downregulated in fruiting bodies [139] , suggesting a functional diversification of this conserved transcription factor among Pezizales . Comparison of gene expression patterns can serve to identify core genes that are involved in biological processes , because conservation of expression is a strong indicator for functional significance [144] , [145] . In previous studies , we have already demonstrated that development-dependent expression of several genes is conserved in P . confluens and other , more derived filamentous ascomycetes , and that conservation of gene expression can be used as a criterion to identify genes that play a role during sexual development in fungi [6] , [18] , [19] . Here , we compared the RNA-seq results from P . confluens with published data from different developmental stages of S . macrospora [24] ( Table S14 ) . A cluster analysis of RPKM values for all orthologous gene pairs showed that overall expression patterns in total vegetative or sexual mycelia from S . macrospora are more similar to those from total mycelia from P . confluens than to expression patterns from isolated young fruiting bodies ( protoperithecia ) of S . macrospora ( Figure 9 ) . Similar results were obtained in a cluster analysis of expression ratios where comparisons of S . macrospora protoperithecia with total mycelia cluster separately from other comparisons ( Figure S15 ) . This indicates that similar tissues/organs in different species might have more similar expression patterns than different tissues/organs from the same species; in other words , tissue/organ-specific gene expression might be conserved across fungi . While more comparative studies of specific fungal organs or cell types are needed to confirm this , this finding is similar to results from organ-specific gene expression analysis in mammals [66] . Among the genes that were upregulated during sexual development in both species are the predicted transcription factor gene PCON_06721 and its S . macrospora ortholog pro44 . The S . macrospora gene was shown previously to be essential for fruiting body development , similar to the corresponding homologous genes in A . nidulans , N . crassa , and A . fumigatus [102]–[105] . In N . crassa , the homologous sub-1 gene was also shown to be light-regulated at the level of transcription , and to be involved in light-regulation of downstream genes [101] . PCON_06721 ( P . confluens pro44 ) is also light-induced ( see section about light-dependent regulation and Figure 7 ) , and further qRT-PCR analysis of PCON_06721 showed that it is upregulated by both light and sexual development , indicating that PCON_06721 might be involved in sexual development and development-independent light reactions ( Figure 9B ) . PCON_06721 and its homologs in other fungi encode GATA-type transcription factors . Whereas the C-terminal GATA domain is highly conserved in all homologous proteins , the N-terminal part of the protein is only weakly conserved ( Figure S16 ) . To address the question whether the developmental function of PCON_06721 might be conserved despite limited sequence conservation , we transformed an S . macrospora Δpro44 strain with a construct expressing the P . confluens PCON_06721 gene ( Figure 9C ) . The S . macrospora Δpro44 is sterile and forms protoperithecia , but no mature perithecia . Transformation with PCON_06721 restored the fertility of the deletion mutant , demonstrating that the gene from the basal filamentous ascomycete P . confluens is functional in the more derived Pezizomycete S . macrospora . This indicates that this transcription factor gene is one of the core regulators of sexual development across filamentous ascomycetes . Here , we have analyzed the genome and development-dependent transcriptomes of P . confluens . This is the second Pezizomycete genome to be sequenced , but the first of a Pezizomycete with a more “typical” saprobic lifestyle and apothecium when compared to the more specialized life style of the truffle T . melanosporum . Overall genomic synteny with T . melanosporum is low , but regions of microsynteny between P . confluens and truffle are more numerous than between P . confluens and other ascomycetes , indicating that the two Pezizomycetes are more closely related to each other than to other ascomycete groups; however , the level of synteny still suggests a wide evolutionary range within the Pezizomycetes . This is consistent with phylogenetic analyses based on rDNA sequences that placed Pyronema and Tuber in subgroups C and B , respectively , of the Pezizomycetes [146] , [147] . The P . confluens genome has a number of characteristics that are similar to higher filamentous ascomycetes , and distinct from T . melanosporum , namely its size of 50 Mb , gene content of ∼13 , 400 protein-coding genes , and low repeat content . However , several typical features of higher filamentous ascomycetes are different in P . confluens , allowing conclusions about the evolution of these features in fungi . For example , the mating type genes are conserved , but in contrast to higher filamentous ascomycetes , their genomic environment is not . Also , clustered genes for secondary metabolites exist , but in much lower numbers than in other species . On the other hand , P . confluens has a full complement of fungal photoreceptors , and expression studies indicate that light-sensing and signaling might be similar to more derived species and therefore represent basic features of filamentous ascomycetes . Several families encoding predicted small secreted proteins are expanded in P . confluens and present in only few other fungal groups , making it possible that they were acquired by horizontal gene transfer . By analyzing spliced RNA-seq reads in antisense direction to annotated genes , we were able to deduce the presence of natural antisense transcripts in P . confluens; and this principle might be of interest for non-strand-specific RNA-seq experiments in other organisms . Interestingly , among the P . confluens orphan genes , a disproportionally high number is upregulated during sexual development , consistent with a hypothesis of rapid evolution of sex-associated genes . Comparative transcriptome analysis with S . macrospora identified the transcription factor gene PCON_06721 , the ortholog of S . macrospora pro44 , as upregulated during sexual development in both species , and complementation of an S . macrospora deletion mutant with the P . confluens gene demonstrated the conserved function of this regulator of sexual development . In summary , the P . confluens genome helps to close a sequence gap at the base of the filamentous ascomycetes , and the genome and transcriptome data are valuable resources for the analysis of fungal evolution and sexual development .
The sequenced strain is Pyronema confluens CBS100304 , obtained from the CBS ( Centraalbureau voor Schimmelcultures , Utrecht , NL ) . The strain was grown on minimal medium as previously described [6] or on cornmeal medium [148] . Sordaria macrospora strains used in this study are the wild type ( FGSC 10222 ) and a pro44 deletion mutant from the strain collection of the Department of General and Molecular Botany at the Ruhr-Universität Bochum ( Nowrousian , Teichert and Kück , unpublished ) . S . macrospora was grown on cornmeal medium as previously described [148] . For standard cultures , white light with a spectral range from 400 to 700 nm ( Osram L36W/840; 1 . 57 lx at culture level ) was used . For wavelength-dependent development and gene expression analyses , LEE filters ( Andover , UK ) with different transmission characteristics were used ( light intensity at the level of the cultures given in lux , for transmission data see Figure S8 ) : far red ( LEE Filter 787 marius red; 0 . 02 lx ) , red ( LEE Filter 106 primary red; 0 . 54 lx ) , green ( LEE Filter 139 primary green; 0 . 60 lx ) and blue ( LEE Filter 363 medium blue; 0 . 14 lx ) . For light induction experiments , samples were harvested under far-red light ( Philips PF712E darkroom safe light ) after 4 d continuous light ( LL ) , or continuous darkness ( DD ) , or DD and 5–60 min of light induction with the respective wavelengths . Genomic DNA from P . confluens was prepared from mycelium grown for 3 days in minimal medium . Mycelium was frozen in liquid nitrogen , pulverized , and incubated in equal volumes of lysis buffer ( 0 . 6 M , 10 mM EDTA , 100 mM Tris-HCl pH 8 . 0 , 1% SDS ) and phenol/chloroform ( 1∶1 ) at room temperature for 10 min with constant shaking . After centrifugation , the supernatant was again treated with an equal volume phenol/chloroform ( 1∶1 ) , and this step was repeated until the supernatant was clear . It was then treated with RNase A , and afterwards again with phenol/chloroform . Genomic DNA was precipitated with sodium acetate ( pH 7 . 0 ) and ethanol . Roche/454 sequencing was performed with 20 µg genomic DNA at Eurofins MWG GmbH ( Ebersberg , Germany ) on a GS FLX system . Illumina/Solexa paired-end sequencing was performed with 5 µg genomic DNA at GATC Biotech ( Konstanz , Germany ) on a HiSeq 2000 . An overview of obtained sequence reads is given in Table S1 . The 454 raw data were extracted from sff files and converted to fasta files using sff_extract . py ( Jose Blanca and Bastien Chevreux , http://bioinf . comav . upv . es/sff_extract/index . html ) . 454 and Illumina raw data were trimmed with custom-made Perl scripts to remove reads with undetermined bases ( “N” ) and for trimming of low quality bases ( phred score <10 ) from the 3′ end as described ( available at http://c4-1-8 . serverhosting . rub . de/public/software . html ) [24] , [104] . 454 reads were assembled with the Celera assembler [149] . The trimmed 454 and Illumina reads as well as the 454-based Celera assembly were used for an assembly with Velvet 1 . 1 . 04 [150] with the following parameters for velveth ( k ) and velvetg ( all others ) : k 41 , exp_cov 100 , cov_cutoff 2 , long_mult_cutoff 0 , ins_length 300 . Overlapping Velvet scaffolds were merged further using CAP3 [151] . The rDNA unit ( scaffold 1635 ) was assembled separately from 454 reads . BLAST searches with rDNAs from S . macrospora and several publicly available Pezizales rDNAs against the 454 reads and the Celera assembly were used to obtain sequences with homology to rDNA . Theses reads were assembled with CAP3 [151] to obtain an rDNA unit that contains the 18S , 5 . 8S , and 28S rRNA genes as well as the internal transcribed spacers 1 and 2 . k-mer frequencies were analyzed based on the Illumina reads with an algorithm described for potato genome [23] . The algorithm was used to write a custom Perl program . Based on the fastq data of the Illumina reads , k-mers of 31 and 41 bases were analyzed . For RNA extraction , P . confluens mycelia were grown as described in liquid medium ( minimal medium or cornmeal medium ) either in darkness ( and harvested in dark red light ) or in light , as surface cultures or submerged ( shaking cultures ) [6] . For the analysis of effects of different wavelengths on fruiting body formation and gene expression , LEE filters ( Andover , UK ) were used as described above . RNA was prepared with the RNeasy lipid tissue mini kit ( Qiagen , Hilden , Germany ) as described [6] . Reverse transcription and qRT-PCR were performed as described previously [18] , [152] , oligonucleotide primers are given in Table S15 . For RNA-seq analysis , 50 µg total RNA from several growth conditions were pooled to generate the sex , DD , and vegmix samples . RNA for sample sex was extracted from mycelia grown in minimal medium in surface culture in constant light . Only under these conditions is P . confluens able to develop fruiting bodies . Equal amounts of RNAs from mycelia grown for 3 , 4 , and 5 d were pooled to represent a high number of genes that are expressed during fruiting body development . The DD samples comprised RNAs from mycelia grown for 3 , 4 , and 5 d in minimal medium in submerged culture in constant darkness , which prevents fruiting body formation . The vegmix samples also contained only RNAs from mycelia that could not develop fruiting bodies , but from a mixture of growth conditions different from the DD samples: for the vegmix samples , we used mycelia grown for 3 d in complete medium ( cornmeal medium ) in surface culture in constant darkness , mycelia grown for 3 d submerged in minimal medium in constant light , and mycelia grown for 3 d submerged in minimal medium in constant darkness . Two independent biological replicates of each condition ( sex , DD , and vegmix ) were used for sequencing . cDNA and library preparation for RNA-seq as well as Illumina/Solexa paired-end sequencing was performed at GATC Biotech ( Konstanz , Germany ) . Indexed cDNA libraries for multiplexing were prepared with the TrueSeq RNA sample preparation kit ( Illumina , San Diego , CA , USA ) . One library was prepared for each independent replicate for each of the three conditions ( sex , DD , and veg ) , and the resulting six libraries were pooled and sequenced in one lane on a HiSeq 2000 . An overview of obtained sequence reads is given in Table S1 . RNA-seq reads were assembled with Trinity [153] , and assembled transcripts were mapped to the genome sequence with PASA [154] . The longest full length ORFs identified by PASA were used to train AUGUSTUS and SNAP , then gene models were predicted independently with AUGUSTUS , SNAP , and GeneMark-ES [155]–[158] . The resulting annotation from each of the prediction programs was used together with the RNA-seq data as input to MAKER , a program that integrates the different sources of gene evidence [25] . Detailed parameters that were used for the gene predictions are available at https://github . com/hyphaltip/fungi-gene-prediction-params/ . Initial automated gene predictions were checked for consistency ( e . g . presence of start/stop codons ) and manually curated in about 10% of all cases . UTR predictions were refined/improved based on the RNA-seq data using custom-made Perl scripts as described previously [24] . For each of the predicted proteins , the protein with the highest sequence identity in GenBank ( nr ) was determined using BLASTP [48] , and putative localizations of the predicted proteins were determined with WoLF PSORT [159] ( Table S2 ) . Searches for conserved protein domains ( Pfam-A domains , http://pfam . sanger . ac . uk/ ) were performed with the HMMER 3 . 0 program hmmsearch [160] , [161] . A chi-square test to determine which domains are over- or underrepresented in P . confluens was performed in R . tRNAs were predicted using a combination of Infernal 1 . 0 , tRNAscan-SE , TFAM 1 . 0 , and ARAGORN [162]–[165] . Analysis of transposable elements and other repeats was performed with RepeatMasker ( A . F . A . Smit , R . Hubley , P . Green; www . repeatmasker . org ) based the RepbaseUpdate library [166] and a library of de novo-identified P . confluens repeat consensus sequences that was generated by RepeatModeler ( A . F . A . Smit , R . Hubley; www . repeatmasker . org/RepeatModeler . html ) . First , the P . confluens genome sequence was analyzed using the RepBase Update library and species-specification “fungi” . In a second step , repeats were identified de novo from the P . confluens genome using RepeatModeler , and the RepeatMasker analysis was repeated with the P . confluens-specific repeat library generated by RepeatModeler . The results of both RepeatMasker runs were combined using custom-made Perl scripts to remove redundancy and only keep non-overlapping repeat regions . Histograms of percent divergence , percent insertions , and percent deletions compared to the repeat consensus sequences were generated based on the output information from RepeatMasker . Composite RIP indices were calculated with Perl script RIP_index_calculation . pl ( https://github . com/hyphaltip/fungaltools/blob/master/scripts/RIP_index_calculation . pl ) on DNA sequences of 500 bp in sliding windows ( window step size 100 bp ) based on a method used in [167] . Briefly , a RIP product index ( TpA/ApT ) and RIP substrate index ( CpA+TpG/ApC+GpT ) [168] , [169] are calculated . Sequences that have been subjected to RIP have a product index of at least 1 . 1 and a substrate index of less than 0 . 9 , while sequences that have not been subjected to RIP have a product index of less than 0 . 8 and substrate index of at least 1 . 1 . The composite RIP index is calculated by subtracting the substrate index from the product index; positive values imply that the DNA has been subjected to RIP [167] . Predicted splice sites from the junctions . bed output of Tophat [31] were analyzed with respect to strand based on the intron consensus sequences ( 5′ GT or GC , 3′ AG ) and overlap with annotated protein-coding genes using custom-made Perl scripts based on BioPerl modules [170] . Splice sites in antisense direction to annotated genes were further filtered to include only sites covered by at least five spliced sequence reads , and with a coverage of more than 10% of the average coverage of the predicted sense-transcript . Remaining putative antisense splice sites were checked manually to remove splice sites that were most likely due to annotation errors or within repeat-rich regions or close to sequence gaps . Regions of sequence similarity were determined with the PROmer algorithm from the MUMmer package version 3 . 23 [47] . The resulting files were used as input to mummerplot , and percent identity plots and dot plots of PROmer results were visualized with gnuplot ( www . gnuplot . info ) based on the mummerplot output files . An orthology-based analysis of synteny was performed by determining orthologs for all P . confluens genes in the predicted proteomes of ten filamentous fungi by reciprocal BLAST analysis [48] , and using custom-made Perl scripts based on BioPerl modules [170] to determine the positions of orthologous proteins on sequenced chromosomes or contigs . The predicted proteomes of P . confluens and the following 17 other fungal species were used for the reconstruction of the phylome: Agaricus bisporus [171] , Arthrobotrys oligospora [10] , Blumeria graminis [172] , Coccidioides immitis [173] , Emericella nidulans [174] , Gibberella zeae [138] , Laccaria bicolor [175] , Mycosphaerella graminicola [176] , Neosartorya fischeri [177] , Neurospora crassa [178] , Phaeosphaeria nodorum [179] , Saccharomyces cerevisiae [180] , Schizosaccharomyces pombe [181] , Sclerotinia sclerotiorum [182] , Sordaria macrospora [22] , Tuber melanosporum [9] , Yarrowia lipolytica [183] . 13 of the proteomes belonged to other Pezizomycotina species , S . cerevisiae and Y . lipolytica represented the Saccharomycotina and S . pombe , L . bicolor and A . bisporus served as outgroups . The phylome , meaning the complete collection of phylogenetic trees for each gene in a genome , was reconstructed in an automated process that mimics a manual phylogenetic tree reconstruction process [184]: homology search , multiple sequence alignment and phylogenetic reconstruction . For each protein encoded in the P . confluens genome we performed a Smith-Waterman search against the proteome database formed by the genomes listed above . Results were then filtered using an e-value threshold of 1e-05 and a continuous overlapping region between the query and the result of 0 . 5 . A maximum of 150 sequences were taken . Multiple sequence alignments were then reconstructed using three different programs: MUSCLE v3 . 8 [185] , MAFFT v6 . 712b [186] , and DIALIGN-TX [187] . Alignments were reconstructed in forward and reverse ( i . e using the Head or Tail approach [188] ) . The resulting alignments were then combined using M-COFFEE [189] . A trimming step was performed using trimAl v1 . 3 [190] ( consistency-score cutoff 0 . 1667 , gap-score cutoff 0 . 9 ) . Trees were reconstructed using PhyML [191] . First the best fitting model was selected by reconstructing neighbor joining trees as implemented in BioNJ [192] using seven different models ( JTT , LG , WAG , Blosum62 , MtREV , VT and Dayhoff ) . The two best models in terms of likelihood were used to reconstruct maximum-likelihood trees . Four rate categories were used and invariant positions were inferred from the data . Branch support was computed using an aLRT ( approximate likelihood ratio test ) based on a chi-square distribution . Resulting trees and alignments are stored in phylomeDB [184] ( http://phylomedb . org ) , with the phylomeID 203 . Orthologs between P . confluens and the other species included in the phylome were based on the phylogenetic trees reconstructed in the phylome ( Table S9 ) . A species-overlap algorithm , as implemented in ETE v2 [193] was used to infer orthology and paralogy relationships . The algorithm traverses the tree and at each node it calls speciation or duplication depending on whether there are common species at both sides of the node . Expanded protein families were detected based on the trees reconstructed in the phylome . For each tree , we used ETE v2 [193] to find nodes that exclusively contained P . confluens sequences . Only those nodes with more than five sequences were considered as expansions . Overlapping expansions were fused when they shared more than 20% of their members . Expansions were then annotated using a BLAST search against UniProt . The species tree was build using a concatenation method . 426 single-copy , widespread genes were selected . The concatenated alignment was further trimmed using trimAl [190] ( gap-score cutoff 0 . 5 and conservation 0 . 5 ) . The final alignment contained 277 , 192 positions . The tree was reconstructed using PhyML [191] . LG model [194] was selected and a 4-categories GAMMA distribution was used . A bootstrap of 100 repetitions was also reconstructed . Additionally a species tree based on the super-tree reconstruction program DupTree [195] was reconstructed . All the trees reconstructed in the phylome were used as input ( 6 , 949 trees ) . Both species trees showed the same topology . r8s [49] was used to estimate the divergence between P . confluens and T . melanosporum based on the species tree inferred from the concatenated alignment . Two analyses were run using different estimates of the divergence between Schizosaccharomyces pombe and the remaining acomycetes as calibration point . For each analysis the smoothing parameter was estimated using cross-validation . In the first analysis , the divergence between S . pombe and the remaning ascomycetes was put at 723 . 86 Mya [196] ( www . timetree . org ) , this resulted in a divergence time between T . melanosporum and P . confluens of 260 . 38 Mya . If a more ancestral divergence point is selected as a calibration point ( 1147 . 78 Mya , [197] ) , then the divergence time between the two species of interest is 413 . 30 Mya . RNA-seq reads were cleaned with custom-made Perl scripts as described [24] and mapped to the P . confluens genome sequence using Tophat [31] . Custom-made Perl scripts using modules from the BioPerl toolkit [170] were used to determine the number of reads that mapped to each annotated protein-coding gene based on the SAM files with the mapping information ( output from Tophat ) , and quantitative analysis was done with two different methods ( “classical” and with LOX [198] ) as described previously [24] . We also used the two Bioconductor packages DESeq [199] and baySeq [200] in the R computing environment ( version 2 . 12 . 1 ) ; however , similar to previous analyses , the number of differentially expressed genes was >10% , and under these conditions the statistical models upon which these methods are based are no longer valid [24] , therefore the results were not used for further analyses ( data not shown ) . Results from both LOX and “classical” analyses agreed well with qRT-PCR results for selected genes ( see e . g . Figure S14 , compare qRT-PCR results for LL/DD with RNA-seq results for sex/DD ) . LOX calculates expression ratios and Bayesian credible intervals and p-values for differential expression . The classical analysis consists of the calculation of expression ratios , standard deviation , and coefficient of variance from read counts normalized to the total number of read counts for the sample , and genes were sorted into five groups ( 0–4 ) according to the following criteria: genes in group 4 have ratios of ≤0 . 25 or ≥4 in all independent biological replicates , genes in group 3 have a mean ratio of ≤0 . 25 or ≥4 and a coefficient of variance <0 . 5 , genes in group 2 have ratios of ≤0 . 5 or ≥2 in all independent biological replicates , genes in group 1 have a mean ratio of ≤0 . 5 or ≥2 and a coefficient of variance <0 . 5 , and group 0 contains all other genes ( with the exception of genes for which no ratios could be calculated due to a lack of read coverage , these were not included in the analysis ) . To classify genes as differentially expressed , a consensus was determined for each gene based on the results from both the classical and LOX analysis; a gene was labeled as up-regulated ( 1 ) , down-regulated ( −1 ) or not differentially expressed ( 0 ) under the conditions that were compared , when the following criteria were met: ( a ) ( “normal” analysis ) expression ratios from both classical and LOX analysis had to be >2 and <0 . 5 , LOX Bayesian probability for differential expression = 1 , and the gene had to be in groups 1–4 in the classical analysis; ( b ) ( stringent analysis ) thresholds for expression ratios were set to >4 and <0 . 25 . For the analysis of reads that mapped to different genomic regions ( e . g . , exons , introns , intergenic regions ) , reads were counted based on the SAM files with the mapping information using custom-made Perl scripts as described previously [24] . To determine the distribution of expression frequencies ( Figure S4 ) , the coverage for locus tags of protein-coding gene was determined as the average coverage for the bases of the predicted mRNA ( normalized to coverage per kilobase per million counted bases in the sample ) . Curve fitting and clustering of the data by expectation-maximization was performed on the log2-transformed RNA-seq data using the R package mclust [201] . Plasmid pRSnat_06721_OE for complementation of an S . macrospora pro44 deletion mutant was generated by homologous recombination in yeast as described [102] . It contains the PCON_06721 ORF under control of the Aspergillus nidulans gpd promoter and trpC terminator in vector pRSnat , which contains a nourseothricin resistance cassette for selection in S . macrospora [202] . The S . macrospora Δpro44 strain was transformed with pRSnat_06721_OE as described [104] , [203] . Multiple alignments were created in CLUSTALX [204] and trimmed with Jalview [205] , and the same alignment was used for analysis by distance-matrix ( DM ) or maximum parsimony ( MP ) . Phylogenetic analyses were made with PAUP version 4 . 0b10 for Windows ( D . L . Swofford , distributed by Sinauer Associates , copyright 2001 Smithsonian Institution ) . DM and MP analyses were performed using 10 , 000 bootstrap replicates . Consensus trees were graphically displayed with TREEVIEW [206] . The sequence and annotation data are available under the accession numbers HF935090–HF936677 ( annotated scaffolds at the European Nucleotide Archive ENA , http://www . ebi . ac . uk/ena/data/view/HF935090-HF936677 ) , CATG01000001–CATG01001898 ( primary , non-annotated contigs from which scaffolds were assembled ) , and BioProject acc . PRJNA65573 . The sequence reads that were used for the assembly of the P . confluens genome were submitted to the NCBI sequence read archive ( accession number SRA059523 ) . The RNA-seq reads and derived expression ratios were submitted to the GEO database ( accession number GSE41631 ) .
|
Fungi are a morphologically and physiologically diverse group of organisms with huge impacts on nearly all ecosystems . In recent years , genomes of many fungal species have been sequenced and have greatly improved our understanding of fungal biology . Ascomycetes are the largest fungal group with the highest number of sequenced genomes; however , for the Pezizales , an early-diverging lineage of filamentous ascomycetes , only one genome has been sequence to date , namely that of the black truffle . While truffles are among the most valuable edible fungi , they have a specialized life style as plant symbionts producing belowground fruiting bodies; thus it is difficult to draw conclusions about basal ascomycetes from one truffle genome alone . Therefore , we have sequenced the genome and several transcriptomes of the basal ascomycete Pyronema confluens , which has a saprobic life style typical of many ascomycetes . Comparisons with other fungal genomes showed that P . confluens has two conserved mating type genes , but that the genomic environment of the mating type genes is different from that of higher ascomycetes . We also found that a high number of orphan genes , i . e . genes without homologs in other fungi , are upregulated during sexual development . This is consistent with rapid evolution of sex-associated genes .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
The Genome and Development-Dependent Transcriptomes of Pyronema confluens: A Window into Fungal Evolution
|
MicroRNAs belonging to the miR-34 family have been proposed as critical modulators of the p53 pathway and potential tumor suppressors in human cancers . To formally test these hypotheses , we have generated mice carrying targeted deletion of all three members of this microRNA family . We show that complete inactivation of miR-34 function is compatible with normal development in mice . Surprisingly , p53 function appears to be intact in miR-34–deficient cells and tissues . Although loss of miR-34 expression leads to a slight increase in cellular proliferation in vitro , it does not impair p53-induced cell cycle arrest or apoptosis . Furthermore , in contrast to p53-deficient mice , miR-34–deficient animals do not display increased susceptibility to spontaneous , irradiation-induced , or c-Myc–initiated tumorigenesis . We also show that expression of members of the miR-34 family is particularly high in the testes , lungs , and brains of mice and that it is largely p53-independent in these tissues . These findings indicate that miR-34 plays a redundant function in the p53 pathway and suggest additional p53-independent functions for this family of miRNAs .
The tumor-suppressor protein p53 is a master regulator of the stress response and provides a key barrier to cellular transformation and tumorigenesis [1] . Upon oncogene activation , DNA damage , and other forms of cellular stress , p53 accumulates in the nucleus where it induces or represses the transcription of a myriad of genes . Ultimately , p53 activation results in cell cycle arrest , apoptosis , or senescence , depending on the cellular context and the type of stimulus [2] . Although transcription-independent mechanisms have been reported [3] , p53 mainly acts as a transcription factor for a large array of downstream effectors [4] , including the proapoptotic proteins Puma , Noxa , and Bax , as well as the cell cycle inhibitor , p21 [5]–[11] . The essential tumor-suppressive function of p53 is further highlighted by the observation that this pathway is inactivated in the vast majority of human cancers [1] , [12] . Several groups have recently suggested that miRNAs are also components of the p53 pathway . In particular , three highly related miRNAs—miR-34a , miR-34b , and miR-34c ( Figure 1A ) —are directly induced upon p53 activation in multiple cell types and have been proposed to modulate p53 function [13]–[20] . The precursors of these miRNAs are transcribed from two distinct loci: the miR-34a locus on chromosome 1p36 and the miR-34b∼c locus on chromosome 11q23 . Canonical p53-binding sites are located in the promoter regions of both miR-34a and miR-34b∼c , and these miRNAs are bona fide direct transcriptional targets of p53 [13] , [17] , [18] . Ectopic expression of members of the miR-34 family is sufficient to induce cell cycle arrest or apoptosis , depending on the cellular context [14] , [17]–[21] . Furthermore , loss-of-function studies using miR-34 antagonists have provided some evidence that this miRNA family is required for p53 function [13] , [18] , [22]–[24] . Many of the predicted miR-34 target genes encode for proteins that are involved in cell cycle regulation , apoptosis , and growth factor signaling . These include Cyclin E2 , cMyc , MET , BCL-2 , SIRT1 , and members of the E2F family of transcription factors [13] , [17] , [17] , [23] , [25] . Consistent with a possible tumor-suppressor role , loss of expression of members of the miR-34 family has been reported in human cancers . Hemizygous deletion of the chromosomal region containing the miR-34a locus has been described in neuroblastomas and pancreatic cancer cell lines [14] , [21] . Similarly , loss of 11q23 , containing the miR-34b∼c locus , has been reported in prostate cancers [26] . Epigenetic silencing of miR-34 members has also been reported in human cancers . Promoter hyper-methylation of miR-34a is observed in non-small-cell lung cancers and melanomas [27] , [28] , and silencing of miR-34a and miR-34b∼c has been described in human epithelial ovarian cancers [29] . Although these observations point towards an important role for miR-34 members as critical downstream effectors of p53 and potential tumor suppressors , these hypotheses have not been formally tested using miR-34-deficient animals and cells . One notable exception is a recent elegant paper by Choi and colleagues demonstrating that miR-34-deficient MEFs are more susceptible to reprogramming [30] . However , the consequences of miR-34 loss on p53 function were not examined in detail . Here we report the generation of mice carrying targeted deletion of all three members of the miR-34 family and systematically investigate the impact of miR-34 loss on the p53 pathway . We show that complete genetic inactivation of miR-34 does not detectably impair the p53 response in a variety of in vivo and in vitro assays . These findings highlight likely redundancies among p53's downstream effectors , show that the miR-34 family is largely dispensable for p53 function in vivo , and suggest possible p53-independent functions .
To investigate the biological functions of miR-34 , we first examined the expression of this family of miRNAs under basal conditions and in response to p53 activation in vivo . Under basal conditions , miR-34a and miR-34b∼c expression is particularly intense in the testis , brain , and lung of adult mice ( Figure 1B–1D ) . MiR-34b∼c expression seems largely restricted to these three tissues , while miR-34a is detectable , albeit at lower levels , also in a variety of other organs ( Figure 1B–1D ) . Consistent with previous reports indicating that miR-34a expression is under the direct control of p53 [13] , [17] , [18] , we detected reduced levels of this miRNA in a subset of p53-deficient tissues ( heart , small and large intestine , liver and kidney ) , but the levels of both miR-34a and miR-34b∼c remained high in the brains , testes and lungs ( Figure 1B–1D ) of p53−/− mice , a finding that suggests that p53-independent mechanisms determine basal miR-34 transcription in these tissues . These results were obtained using two independent techniques: quantitative real time polymerase chain reaction ( qPCR ) and Northern blotting . The specificity and sensitivity of these assays were validated using miR-34-deficient mice as controls ( Figure 1B–1D and Figure 2D ) . Exposure to ionizing radiation , which leads to p53 stabilization and transcriptional activation , resulted in substantial miR-34a induction in the spleen , thymus , small and large intestine of wild-type mice , but not in the other tissues examined ( Figure S1 ) . We also observed modest but significant miR-34c induction in the thymus , small and large intestine of irradiated mice , but not in the other tissues examined . To investigate the physiologic functions of the miR-34 family and to determine the extent to which its induction is required for p53 function , we generated mice carrying targeted deletion of both miR-34a and miR-34b∼c loci ( Figure 2A–2C ) . To allow temporally and spatially restricted deletion , we also generated a conditional miR-34a KO allele ( miR-34afl , Figure 2A ) . Northern blot and qPCR analysis confirmed the loss of expression of the respective miRNAs in homozygous mutant animals ( Figure 2D ) . Importantly , homozygous deletion of miR-34a did not lead to compensatory up-regulation of miR-34b∼c , and vice versa ( Figure 2D and data not shown ) . MiR-34a−/− and miR-34b∼c−/− single KO mice were viable and fertile and were obtained at the expected Mendelian frequency ( Figure 2E , 2F ) . The sequence similarity between the three miR-34 family members ( Figure 1A ) , which share the same “seed” , suggests that they may be functionally redundant . To examine the consequences of complete loss of miR-34 function , we crossed miR-34a−/− and miR-34b∼c−/− mice to generate compound mutant animals carrying homozygous deletion of all three family members ( miR-34TKO/TKO ) . Complete loss of miR-34 expression in miR-34TKO/TKO animals was confirmed by Northern blot and qPCR ( Figure 2D ) . MiR-34TKO/TKO mice of both sexes were obtained at approximately the expected Mendelian frequency ( Figure 2G ) , did not display obvious macroscopic defects ( Figure S2 ) , and were fertile ( data not shown ) . A full histological examination ( Figure S3 ) , complete blood cell count ( Figure S4 ) , and serum chemistry analysis ( Figure S5 ) did not detect any statistically significant defects in adult miR-34TKO/TKO mice of both sexes . An analysis of the major myeloid and lymphoid populations of the bone marrow , spleen and thymus also did not reveal any statistically significant difference between wild-type and miR-34TKO/TKO mice ( Figure S6 ) . Next , we sought to determine whether loss of miR-34 expression affects the p53 response in vitro . We focused on the three best-characterized p53-dependent processes: replicative senescence , response to DNA damage , and response to oncogene activation [31]–[35] . The ability to proliferate indefinitely is one of the hallmarks of cancer cells [36] and also one of the most striking consequences of p53 inactivation at the cellular level [35] . To investigate the role of miR-34 in replicative senescence , mouse primary fibroblasts ( MEFs ) derived from wild-type , p53−/− , and miR-34TKO/TKO embryos were serially passaged . Although we detected a remarkable induction of miR-34a and miR-34c expression in late-passage wild-type MEFs compared to early-passage MEFs ( Figure 3A ) , miR-34-deficient MEFs became senescent with a kinetic identical to wild-type MEFs ( Figure 3B ) . This is in stark contrast with p53-deficient MEFs , which as expected proliferated indefinitely ( Figure 3B ) . The only significant difference we observed was a slight but reproducible increase in the proliferation rate of early passage miR-34-deficient fibroblasts compared to controls ( Figure 3B , 3C ) . We next examined the role of miR-34 in the response to the DNA damaging agent doxorubicin . As previously reported [37] , doxorubicin treatment leads to stabilization of p53 ( Figure 3D ) and up-regulation of its downstream targets p21 ( Cdkn1a ) , Mdm2 , Puma and Noxa ( Figure 3D–3F ) . Expression of members of the miR-34 family was similarly upregulated in response to p53 stabilization ( Figure 3G ) . Although as predicted , p53-null cells failed to arrest in G1 in response to doxorubicin treatment , the response of miR-34TKO/TKO MEFs was indistinguishable from that of wild-type cells ( Figure 3H–3I ) . Consistent with these results , doxorubicin treatment caused similar activation of p53 and of its downstream targets in wild-type and miR-34TKO/TKO MEFs ( Figure 3E and 3F ) . The experiments described above were performed on asynchronously growing early-passage MEFs and as such may not be sensitive enough to detect a modest effect of miR-34 loss on the S-phase checkpoint . To measure cell cycle progression more accurately , we first synchronized MEFs by serum starvation and then released the cells in complete medium containing colcemid , a mitotic spindle inhibitor . With this experimental design , upon release in complete medium , cells synchronously proceed from G1 through S phase and then accumulate at the M ( 4N ) phase . This assay provides a more sensitive way to determine the ability of cells to transition through the S-phase and allows detection of subtle defects in the DNA damage-induced S-phase checkpoint . Although a reproducibly larger fraction of miR-34TKO/TKO cells was able to transition through the S phase after ionizing radiation compared to wild-type MEFs ( Figure 3J ) , we observed a similar difference in non-irradiated MEFs ( Figure 3J ) . The most logical interpretation of these results is that miR-34-deficient MEFs , rather than being more resistant to irradiation-induced cell cycle arrest , possess a slightly faster basal proliferation or more rapid re-entry into the cell cycle following serum starvation . This interpretation is also consistent with the faster proliferation rate displayed by miR-34-deficient MEFs ( Figure 3B , 3C ) and with the observation by Lal and colleagues that miR-34a is involved in modulating the cellular response to growth factors [38] . We also examined the consequences of miR-34 loss in MEFs on the expression of a subset of its previously reported direct targets [17] , [20] , [23] , [25] . We detected modest upregulation of cMyc , E2f3 , Met and Sirt1 in miR-34-deficient MEFs , while Bcl2 was expressed at similar levels in wild-type and mutant cells ( Figure 3K ) . The upregulation of Myc and E2f3 might contribute to the increased proliferation rate we have observed in miR-34 deficient MEFs . Having established that miR-34 is not required for cell cycle arrest in response to genotoxic stress in MEFs , we next sought to determine whether this miRNA family might contribute to p53-induced apoptosis . Thymocytes respond to ionizing radiations by rapidly undergoing apoptosis , an effect that is dependent on p53 [39] . We therefore examined the effects of DNA damage on thymocytes from wild-type , p53−/− , and miR-34TKO/TKO mice . As expected , p53−/− thymocytes were almost entirely resistant to irradiation-induced apoptosis; however , wild-type and miR-34-deficient cells were equally sensitive to DNA damage-induced apoptosis , as judged by dose-response and time-course experiments ( Figure 4A , 4B ) . To exclude the possibility that tissue culture conditions may have masked a physiologic role of miR-34 in modulating the p53 response , we next examined the consequences of p53 activation in miR-34-deficient tissues directly in vivo . Age- and sex-matched wild-type , miR-34TKO/TKO and p53−/− mice were exposed to 10 Gy of ionizing radiation and euthanized 6 hours later . Ionizing radiation induced similar activation of the p53 pathway and of its downstream effectors in wild-type and miR-34TKO/TKO mice ( Figure 4C ) . Analogous to what we observed in thymocytes in vitro , the apoptotic response was equally dramatic in wild-type and in miR-34-deficient mice , while it was virtually absent in p53−/− animals ( Figure 4D–4G ) . Based on these results we conclude that miR-34 function is not required for p53-induced cell-cycle arrest and apoptosis in response to genotoxic stresses . The p53 pathway provides a crucial barrier against the neoplastic transformation of primary cells [40] . Supra-physiologic proliferative stimuli , such as those caused by sustained oncogene activation , lead to transcriptional activation of p19Arf , which in turn results in stabilization and activation of p53 , and consequently apoptosis or cell cycle arrest [41] . For example , ectopic expression of a constitutively active K-Ras ( K-RasV12 ) in wild-type MEFs leads to oncogene-induced senescence , but the concomitant inactivation of p53 is sufficient to allow full cellular transformation [31] . To test whether miR-34 plays a role in this context , we ectopically expressed oncogenic K-Ras in wild-type , miR-34TKO/TKO , and p53−/− MEFs . As shown in Figure 5A , complete loss of miR-34 function was not sufficient to allow primary MEFs to be transformed by K-RasV12 alone , while p53-deficient MEFs were readily transformed in the same assay . However , when MEFs were co-transduced with oncogenic K-Ras and E1A , which binds to and inhibits the retinoblastoma protein ( pRb ) [42] , we observed a slight increase in the number of foci formed in miR-34TKO/TKO MEFs compared to wild-type cells ( Figure 5A , 5B ) . These results show that while miR-34 alone is not required for p53-mediated tumor suppression in MEFs , its loss might cooperate with inactivation of the Rb pathway in promoting cellular transformation . To extend our analysis to an in vivo setting , we next examined whether miR-34 inactivation is sufficient to accelerate spontaneous and oncogene-induced transformation in mice . P53-deficient mice exhibit a high incidence of spontaneous tumors , in particular lymphomas and sarcomas [43]–[45] , and p53 inactivation greatly accelerates tumor formation in a variety of mouse models of human cancer [46]–[51] . To determine whether loss of miR-34 expression leads to increased spontaneous tumorigenesis , we aged a cohort of 14 miR-34TKO/TKO and 12 wild-type mice . The animals were monitored for at least 12 months ( wild-type = 359 days; miR-34TKO/TKO = 359 days ) and up to 17 . 3 months ( wild-type = 521 days; miR-34TKO/TKO = 521 days ) . All wild-type and miR-34TKO/TKO mice appeared healthy and miR-34TKO/TKO mice did not show a reduction in life span compared to wild-type controls ( Figure S7 ) . For comparison , the median survival of p53−/− mice has been reported to be 4 . 5 months and by 10 months of age all p53−/− mice have died or developed tumors [45] . In addition , ∼40% of p53+/− mice develop tumors by 16 months of age [45] . Thus , although a longer follow-up of miR-34TKO/TKO mice may be needed to uncover very subtle defects in tumor suppression , we conclude that loss of miR-34 expression does not lead to a substantial increase in spontaneous tumorigenesis . We next sought to determine whether loss of miR-34 might accelerate tumor formation in response to genotoxic stress . P53−/− mice irradiated shortly after birth display accelerated tumorigenesis compared to non-irradiated littermates [52] . We therefore exposed a cohort of 14 miR-34TKO/TKO and 11 wild-type mice to 1 Gy of ionizing radiation soon after birth and monitored them for 42–60 weeks . Both wild-type and miR-34-deficient mice appeared healthy throughout the follow-up period ( Figure S7 ) , in striking contrast with the ∼15 weeks reported median tumor-free survival of irradiated p53−/− mice [52] . Although it will be important to follow a larger cohort of animals over a more prolonged period , these results suggest that miR-34 does not provide a potent barrier to tumorigenesis in response to genotoxic stress in vivo . Finally , we sought to determine whether genetic ablation of miR-34 could contribute to tumor formation in cooperation with a defined oncogenic lesion . For these experiments , we chose the Eμ-Myc model of B cell lymphomas [53] . A crucial tumor-suppressive role for p53 is well established in this mouse model and inactivation of the p53 pathway results in greatly accelerated lymphomagenesis [46] , [47] , [54] . However , even in this context complete loss of miR-34 expression was not sufficient to accelerate tumor formation . The incidence and latency of B cell lymphomas was virtually identical in Eμ-Myc;miR-34TKO/TKO and Eμ-Myc;miR-34+/+ mice ( Figure 5C ) and the resulting tumors displayed similar histopathological features and extent of spontaneous apoptosis ( Figure 5D–5E ) .
We have reported the generation of mice carrying targeted deletion of miR-34a , miR-34b and miR-34c , and we have investigated the consequences of loss of miR-34 expression on p53-dependent responses in vitro and in vivo . Our results show that complete loss of miR-34 expression is compatible with normal development and that the p53 pathway is apparently intact in miR-34-deficient mice . Our observation that inactivation of miR-34 does not impair p53-mediated responses in vitro and in vivo is particularly relevant because a key role for miR-34 in the p53 pathway had been previously proposed by a number of independent groups . The results presented in this paper do not necessarily conflict with previous experiments showing that ectopic expression of miR-34 can induce many of the most characteristic consequences of p53 activation; here we have tested whether miR-34 is necessary for p53 function and not whether it is sufficient . More difficult , however , is to reconcile our findings with previous reports of impaired p53-function in cells treated with miR-34 antagonists . Because previous work has relied on the use of miRNA antagonists to inhibit miR-34 function , it is possible that some of the previous observations reflected miR-34-independent off-target effects . It is also possible that other miRNAs sharing sequence similarities with miR-34 may compensate for miR-34 loss in the knock-out animals . In particular , members of the miR-449 family ( miR-449a , b and c ) have the same “seed” sequence as miR-34 , and miR-34 antagonists could in principle impair their function as well . A conclusive test for this hypothesis will require the generation of compound miR-34 and miR-449 mutant animals , but several lines of evidence suggest that this explanation is not particularly likely . First , in the tissues and cells used in our experiments , the expression of miR-449 members is much lower compared to miR-34a and miR-34c , as judged by multiple independent methods including qPCR , Northern blotting and high throughput sequencing ( Figure S8 and data not shown ) . A notable exception is represented by the testis , in which expression of miR-449a is particularly elevated ( Figure S8 ) . In addition , miR-449 expression is not substantially increased in miR-34-null mice , and activation of the p53 pathway does not lead to significant upregulation of miR-449 ( Figure S8 ) . We would like to emphasize that our results do not necessarily indicate that members of the mIR-34 family are not components of the p53 pathway . Given the essential tumor-suppressive function exerted by p53 , it is perhaps not surprising that multiple and partially redundant effector arms are recruited in response to its activation . It is plausible that the simultaneous inactivation of multiple effector arms is required to measurably impair p53 function . Consistent with this model is our observation that while loss of miR-34 expression alone does not allow the transformation of primary cells by oncogenic K-Ras , it slightly increases the efficiency of transformation when combined with inactivation of the Rb pathway by E1A ( Figure 5A , 5B ) . In this context , it will be important to systematically probe the extent of functional cooperation between this family of miRNAs and other , previously characterized p53 effectors . We also wish to point out that in this manuscript we have investigated the best-characterized functions of p53 ( cell cycle arrest , apoptosis and tumor suppression ) and it remains possible that miR-34 participates in other p53-dependent processes . For example , p53 has been proposed to modulate autophagy [55] and stem cell quiescence [56] , [57] and we cannot exclude that miR-34 plays an important role in these contexts . Future studies using the miR-34-deficient animals we have generated will be needed to test these possibilities . With respect to the potential tumor suppressive role of miR-34 , our experiments indicate that loss of miR-34 expression does not lead to an obvious increase in tumor incidence in mice and does not cooperate with Myc in the context of B cell lymphomagenesis . However , the tumor suppressive function of miR-34 might be restricted to specific tissues and loss of miR-34 might cooperate with specific oncogenic lesions . In humans , for example , loss of miR-34 expression has been reported in a large fraction of primary melanomas , prostatic adenocarcinomas and small cell lung cancers [27] , [28] , among others . Introducing the miR-34-null alleles we have generated into mouse models of these types of human cancers will be important to fully explore the tumor suppressive potential of this family of miRNAs . An additional issue raised by the results presented in this manuscript relates to possible p53-independent functions of miR-34 . We show that under basal conditions the expression of both miR-34 loci is particularly elevated in the testes and , to a lesser extent , in the brains and lungs of mice . Importantly , in these three tissues , miR-34 expression is almost entirely p53-independent ( Figure 1B–1D and [58] ) , a finding that suggests that additional transcription factors control the expression of this family of miRNAs in the absence of genotoxic or oncogenic stresses . A role for miR-34c in spermatogenesis and in controlling the first zygotic cleavage has been recently proposed [58] , [59] . Although our observation that single KO and miR-34TKO/TKO mice produce viable offspring argues against an essential role for miR-34 in these processes , members of the related miR-449 family , that are particularly highly expressed in the testis ( Figure S8 ) , could partially compensate for miR-34 loss in this context . Recent reports have also implicated miR-34 in neuronal development and behavior [60] , [61] and a role for miR-34c in learning and memory [62] , as well as in stress-induced anxiety [63] , has been reported . In addition , inactivation of miR-34 expression has been recently shown to lead to accelerated neurodegeneration and ageing in Drosophila melanogaster [64] . A detailed behavioral and neuroanatomical analysis , as well as a careful characterization of the long-term consequences of miR-34-loss will be essential to confirm and extend these hypotheses in mice . In conclusion , we have reported the generation and characterization of miR-34-deficient mice with a particular focus on the consequences of miR-34 loss on the p53 pathway . The genetically engineered mouse models described in this study will be essential to further investigate the physiologic functions and the tumor suppressive potential of this important miRNA family .
The “recombineering” method [65] was used to modify a BAC clone ( RP-23-410P10 ) containing the miR-34a locus to generate the miR-34a conditional knockout allele . A frt-Neo-frt-loxP cassette was first inserted ∼480 bp downstream of the pre-miR-34a sequence . Gap-repair was used to retrieve a 9 . 6 kbp fragment containing the frt-Neo-frt-loxP cassette , ∼4 kb of 3′ homology arm , and ∼3 . 7 kb 5′ homology arm , and including the pre-miR-34a sequence . The fragment was cloned into the targeting plasmid pKS-DTA , and a second loxP site was introduced into a unique KpnI site located ∼500 bp upstream of the pre-miR-34a sequence . The final targeting construct was linearized with NotI and electroporated into V6 . 5 murine embryonic stem cells ( ESC ) . Following selection with G418 , ESC colonies were isolated and screened by Southern blotting using DNA probes mapping outside the targeted region . Two targeted clones were expanded and injected into C57BL/6 blastocysts to generate chimeric mice . High contribution chimeras were subsequently crossed to Actin-flpe transgenic mice [66] to excise the frt-Neo-frt cassette and generate the miR-34a conditional knockout allele ( miR-34afl ) or crossed to CAG-Cre mice [67] to excise the entire region flanked by the loxP sites and obtain the constitutive miR-34a KO allele ( miR-34aΔ ) . Lastly , miR-34a+/fl and miR-34a+/− were intercrossed to obtain miR-34afl/fl and miR-34a−/− animals . To generate mice carrying deletion of the miR-34b∼c bicistronic cluster , we used recombineering to replace a 1 . 3 kbp DNA region in BAC RP-23-281F13 containing pre-miR-34b and pre-miR-34c with a frt-Neo-frt cassette . A 8 . 4 kbp DNA fragment containing the frt-Neo-frt cassette , the 3 . 7 kbp 5′ homology arm , and 2 . 8 kbp of 3′ homology arm was retrieved from the engineered BAC and cloned into pKS-DTA . The resulting targeting vector was linearized by NotI and electroporated into V6 . 5 ESCs . Upon selection , two independent clones were injected into C57BL/6 blastocysts . High contribution chimeras were crossed to Actin-flpe transgenic mice for germline transmission of the targeted allele and to delete the Neo cassette resulting in the miR-34b∼cΔ allele . The miR-34b∼c+/− mice were intercrossed to obtain miR-34b∼c−/− animals . The Eμ-Myc mice were generated and described by Adams and colleagues [53] and the p53−/− mice were generated by Jacks and colleagues [44] . Genotyping protocols are provided in Text S1 . All animal studies and procedures were approved by the MSKCC Institutional Animal Care and Use Committee . Mice were maintained in a mixed 129SvJae and C57BL/6 background . The miR-34a<floxed> mice and the miR-34b∼c−/− mice are available to the research community through The Jackson Laboratory ( JAX Stock Numbers 018545 and 018546 ) . Primary MEF lines were generated from E13 . 5 embryos using standard protocols . miR-34TKO/TKO embryos were obtained by intercrossing miR-34 mutant mice . Wild-type MEFs were generated in parallel . p53−/− embryos were obtained by intercrossing p53+/− mice . Genotyping protocols are provided in Text S1 . MiR-34 wild-type and miR-34TKO/TKO MEF lines were also verified by qPCR . RNA extraction was performed by homogenizing tissues and cells in TRIzol reagent ( Invitrogen ) according to manufacturer's instructions . For Northern blotting , 15 µg of each RNA sample was loaded into a 15% Urea-PAGE gel and blotted onto a Hybond-N+ nylon membrane ( GE Healthcare ) . The blots were then hybridized with 32P-labeled probes specific for miR-34a , miR-34c , and U6 . qPCR was performed using primers and probes by Applied Biosystems according to manufacturer's instructions . Sno-135 was used for normalization . Passage 2 or 3 primary MEFs were used for all experiments and cultured at 37°C ( 5% CO2 ) in DME-HG with 10% FBS ( complete medium ) or 0 . 1% FBS ( starvation medium ) supplemented with L-glutamine , penicillin , streptomycin , and β-mercaptoethanol . For BrdU cell cycle analysis , wild-type , miR-34TKO/TKO , and p53−/− MEFs were plated in complete medium at 70% confluence , treated with varying doses of doxorubicin for 16 hours or treated at different time points , and pulsed with 10 µM BrdU for one hour . The BD Pharmingen APC-BrdU kit was used to process harvested samples and used according to manufacturer's protocol . For the irradiation experiments , 150 , 000 wild-type , miR-34TKO/TKO and p53−/− MEFs were seeded into each well of a 6-well culture plate and starved for 72 hours . MEF lines were then trypsinized and resuspended in complete medium and either irradiated ( 20 Gy , Cs-137 irradiator , Shepherd Mark-I ) or left untreated . Cells were replated into complete medium containing 500 ng/ml colcemid at 70% confluence and harvested 24 h later . Samples were processed as mentioned above and stained with 7-AAD . Flow cytometry was performed using FACSCalibur ( BD Biosciences ) , and data were analyzed using FlowJo software ( TreeStar ) . Wild-type and miR-34TKO/TKO MEFs were seeded into a 6-well plate ( 40 , 000 cells/well ) and counted every day for the growth curves . The standard 3T3 protocol was followed to determine the cumulative population doublings of wild-type , miR-34TKO/TKO , and p53−/− MEFs . Briefly , 3×105 cells were seeded in a 6 cm2 dish and counted and passaged every three days . Thymocytes were isolated from sex-matched , age-matched wild-type , miR-34TKO/TKO , and p53−/− mice and seeded at a density of 1×106 cells/ml in MEF medium . Thymocytes were then treated with various doses of irradiation ( 2 , 4 , 6 , 8 , and 10 Gy , Cs-137 irradiator , Shepherd Mark-I ) or left untreated . For the time course experiments , thymocytes were treated with 5 Gy of irradiation and harvested 4 , 8 and 24 h after treatment . Samples were stained with AnnexinV and propidium iodide ( Roche ) according to manufacturer's protocol . Flow cytometry was performed using FACSCalibur ( BD Biosciences ) and data were analyzed using FlowJo software ( TreeStar ) . Phoenix cells ( Orbigen ) were transfected using FUGENE 6 ( Promega ) with retroviral constructs of K-RasV12 alone or together with E1A according to manufacturer's instructions . Wild-type , miR-34TKO/TKO , p53−/− MEFs were seeded at 70% confluence and infected with virus . Plates were fixed with methanol and stained with crystal violet two weeks after infection . Foci were quantified using ImageJ . Cells were lysed in RIPA buffer containing protease inhibitors . Proteins ( 25 µg ) were separated on a NuPAGE Bis-Tris gel ( Invitrogen ) , and transferred onto a PVDF membrane ( Millipore ) . Blocking was performed with 5% milk in TBST . Primary antibodies used were anti-p21 ( 1: 1000 , Santa Cruz , F-5 ) , anti-Mdm2 ( 1∶1000 , Abcam , 2A10 ) , anti-Met ( 1∶1000 , Millipore , 07-283 ) , anti-Bcl2 ( 1∶500 , Cell Signaling , #2876S ) , anti-E2f3 ( 1∶500 , Millipore , PG37 ) , anti-Sirt1 ( 1∶1000 , Cell Signaling #2028 ) , anti-cMyc ( 1∶1000 , Cell Signaling , D84C12 ) , and anti-α-Tubulin ( Sigma , DM1A ) . The anti-p53 antibody ( 1∶300 ) was a kind gift of Kristian Helin ( BRIC , Denmark ) . Secondary antibodies were obtained from Cell Signaling . ECL reagents were obtained from GE Healthcare . Western blot bands were quantified using ImageJ . Mice were irradiated with 10 Gy and sacrificed 6 hours after . PFA-fixed , paraffin-embedded sections were deparaffinized in xylene , and rehydrated . The samples were stained with Cleaved Caspase-3 antibody ( Cell Signaling , #9664 ) overnight , according to Cell Signaling protocol . The samples were also counterstained with 0 . 1% alcoholic Eosin Y solution ( Sigma-Aldrich ) or 30% hematoxylin . The sections were then dehydrated and mounted in Permount ( Fisher Scientific ) . Sample pictures were quantified using ImageJ .
|
MicroRNAs ( miRNAs ) are small , non-coding RNAs that broadly regulate gene expression . MicroRNA deregulation is a common feature of human cancers , and numerous miRNAs have oncogenic or tumor suppressive properties . Members of the miR-34 family ( miR-34a , miR-34b , and miR-34c ) have been widely speculated to be important tumor suppressors and mediators of p53 function . Despite the growing body of evidence supporting this hypothesis , previous studies on miR-34 have been done in vitro or using non-physiologic expression levels of miR-34 . Here , we probe the tumor suppressive functions of the miR-34 family in vivo by generating mice carrying targeted deletion of the entire miR-34 family . Our results show that the miR-34 family is not required for tumor suppression in vivo , and they suggest p53-independent functions for this family of miRNAs . Importantly , the mice generated from this study provide a tool for the scientific community to further investigate the physiologic functions of the miR-34 family .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"models",
"cancer",
"genetics",
"model",
"organisms",
"genetics",
"biology",
"mouse",
"genetics",
"and",
"genomics",
"gene",
"function"
] |
2012
|
Intact p53-Dependent Responses in miR-34–Deficient Mice
|
A form of dwarfism known as Meier-Gorlin syndrome ( MGS ) is caused by recessive mutations in one of six different genes ( ORC1 , ORC4 , ORC6 , CDC6 , CDT1 , and MCM5 ) . These genes encode components of the pre-replication complex , which assembles at origins of replication prior to S phase . Also , variants in two additional replication initiation genes have joined the list of causative mutations for MGS ( Geminin and CDC45 ) . The identity of the causative MGS genetic variants strongly suggests that some aspect of replication is amiss in MGS patients; however , little evidence has been obtained regarding what aspect of chromosome replication is faulty . Since the site of one of the missense mutations in the human ORC4 alleles is conserved between humans and yeast , we sought to determine in what way this single amino acid change affects the process of chromosome replication , by introducing the comparable mutation into yeast ( orc4Y232C ) . We find that yeast cells with the orc4Y232C allele have a prolonged S-phase , due to compromised replication initiation at the ribosomal DNA ( rDNA ) locus located on chromosome XII . The inability to initiate replication at the rDNA locus results in chromosome breakage and a severely reduced rDNA copy number in the survivors , presumably helping to ensure complete replication of chromosome XII . Although reducing rDNA copy number may help ensure complete chromosome replication , orc4Y232C cells struggle to meet the high demand for ribosomal RNA synthesis . This finding provides additional evidence linking two essential cellular pathways—DNA replication and ribosome biogenesis .
The faithful and timely duplication of a cell’s genome is required every round of division . During eukaryotic S phase , DNA replication initiates at multiple sites along each chromosome called origins of replication . Eukaryotic replication initiation has been best characterized in the budding yeast Saccharomyces cerevisiae , where chromosomal origins were first identified by their ability to maintain recombinant plasmids after transformation into yeast [1] . The majority of these Autonomous Replication Sequences or ARS elements correspond to the ~300 chromosomal origins of replication that are scattered across the genome and share a core consensus sequence called the ACS ( ARS consensus sequence ) [2] . Subsequent biochemical and genetic work in yeast identified many of the essential genes for replication initiation [3–5] . Features that define origins in higher eukaryotes differ significantly from yeast ARSs , but the proteins that carry out origin recognition and initiation are strikingly conserved in sequence and structure across eukaryotes ( S1 Fig ) [6–9] . In budding yeast , a six-membered protein complex called the Origin Recognition Complex ( Orc1-6 ) binds ARSs throughout the cell cycle [10] . To become competent ( or licensed ) for initiation , additional proteins are recruited by ORC during the M and G1 phases [11] . The first licensing factor to bind is Cdc6 , which facilitates the recruitment of the Mcm2-7 helicase component through an interaction with Cdt1 [9 , 12–16] . Collectively , this protein complex is known as the Pre-Replication Complex ( Pre-RC ) and , once assembled on an origin , licenses it to initiate DNA replication or “fire” in the subsequent S phase . During the onset of S phase CDK- and DDK-dependent phosphorylation events complete assembly of the replisomes , including two helicase complexes , allowing replication to proceed bi-directionally from the origin of replication [17–19] . As chromosome replication is essential for cell division , there has been a tacit assumption that mutations that impair the function of proteins involved in DNA replication would be incompatible with metazoan life . Yet , researchers reported in 2011 that amino acid substitutions in proteins involved in the initiation of DNA replication , proteins first identified in S . cerevisiae , are responsible for a form of proportionate dwarfism called Meier-Gorlin syndrome ( MGS ) [20 , 21] . Individuals with MGS have phenotypes that include short stature , small external ears and missing or underdeveloped kneecaps [22 , 23]—phenotypes not obviously associated with chromosome replication defects . The specific genetic variants found in patients with MGS include homozygous or compound heterozygous alterations in six different Pre-RC genes ( ORC1 , ORC4 , ORC6 , CDT1 , CDC6 , and MCM5 ) [20 , 21 , 24] . Recent work has identified de novo autosomal dominant mutations in Geminin ( encoded by GMNN ) , an inhibitor of DNA replication that is unique to higher eukaryotes [25] . Additionally , biallelic mutations in CDC45 , which is required for both origin initiation and elongation during S phase , have been found to be causative for some cases of MGS [26] . Considering the known roles of these proteins in origin initiation , a reasonable hypothesis is that these mutations are adversely affecting DNA replication and thus reducing cell proliferation so that individuals harboring these variants are uncommonly small . Consistent with this hypothesis , previous work using Epstein-Barr virus replication as an assay has found that immortalized fibroblasts and cultured lymphoblastoid cells derived from MGS patients are diminished in their ability to initiate replication [27 , 28]; however , defects in replication initiation did not always correlate with slowed S phase in these cells [27] . Additionally , MGS mutations have been shown to affect aspects of cell biology other than DNA replication , such as centrosome duplication and cilia formation [27 , 29] . The defective cilia formation phenotype observed in MGS cells is thought to contribute to some of the developmental abnormalities associated with this condition [27] . Although the proteins linked with MGS have been studied extensively in yeast and other eukaryotes , it is not clear how MGS mutations might affect chromosome replication to give rise to the phenotypes observed in humans . Therefore , understanding how MGS mutations affect chromosome replication may shed light on how they contribute to the phenotypes in humans . In this study , we have replaced the genomic copy of the budding yeast ORC4 with a mutated version ( orc4Y232C ) bearing a tyrosine-to-cysteine change that is orthologous to the Tyr174Cys mutation reported in human patients ( S2A and S2B Fig ) [21] . We find that yeast cells bearing this orc4Y232C allele have a longer cell cycle time that is mostly accounted for by a lengthened S phase . Additionally , we find that in orc4Y232C cells more than 85% of the earliest firing origins are unaltered in their time and/or efficiency; however , the origins present in each copy of the ribosomal DNA ( rDNA ) are severely compromised in their ability to fire and the number of copies of the rDNA repeat drops from ~150 to as few as 10 . Previous work with ORC1 and ORC2 temperature sensitive mutants also revealed shrinkage of the rDNA; however , that work did not provide a model for how the copy number reduction occurred and proposed a checkpoint control of genome-wide replication initiation as an explanation for loss of rDNA repeats [30] . Our findings with the orc4Y232C allele reveal that the mechanism for rDNA copy number loss is chromosome XII breakage as a consequence of the “random replication gap” problem and that insufficient replication initiation outside of the rDNA locus is not likely the cause for rDNA shrinkage . Furthermore , we show that the reduction in rDNA copy number , by restricting rRNA synthesis , constrains the translational capacity , possibly explaining the slow growth observed in orc4Y232C cells . While it remains to be seen whether these phenotypes are also common to Meier-Gorlin patient cells , our characterization of the orc4Y232C allele in S . cerevisiae highlights an unsuspected pathway linking replication dysfunction and growth control .
A missense mutation in ORC4 has been shown to be causative for some instances of MGS in humans [21] . This specific mutation results in an amino acid substitution ( Tyrosine to Cysteine ) at position 174 of the human Orc4 protein and occurs in a region with homology to the AAA+ ( ATPases Associated with diverse cellular Activities ) related domain of the S . cerevisiae Orc4p [21] . The initial work investigating the equivalent MGS mutation ( orc4Y232C ) in S . cerevisiae revealed a slow-growth phenotype [21] . In that experiment , the investigators constructed a strain that had the orc4Y232C allele on a plasmid rescuing the inviable chromosomal deletion of ORC4 . Because the orc4Y232C allele was not in its native location , it was not clear to what extent the slow growth was due to plasmid loss or poor expression from the plasmid versus other , more wide-spread defects in chromosome replication or segregation . To explore in more detail the consequences of the orc4Y232C mutation , we replaced the wild type yeast ORC4 allele with the MGS equivalent allele orc4Y232C at its native chromosomal locus . Cells with the chromosomal orc4Y232C allele grow more slowly than wild type ( population doubling time of 2 . 7 hr . vs . 2 . 4 hr; Fig 1A ) . The more marked difference in growth rates reported for the strains analyzed by Guernsey et al . [21] is probably due to loss of the plasmid bearing the sole source of Orc4p in their experiments . To explore the slow growth phenotype , we performed flow cytometry on cells synchronously proceeding through the cell cycle after having been arrested at START by treatment with alpha-factor ( Fig 1B ) . Initiation of S phase appears to be slightly delayed in the mutant but after replicating much of its genomic DNA the mutant shows a more pronounced delay in reaching a full G2 DNA content and progressing back into G1 . We reasoned that a delay in S phase entry could result if the orc4Y232C mutation altered early origin firing . To examine this possibility we used an assay that specifically examines early origin activation across the genome [31] . This assay uses microarray hybridization to measure the levels of single stranded DNA exposed at replication forks . Both the peak position and peak amplitude of ssDNA formed at genomic loci are informative . While we do not fully understand the molecular processes that give rise to peaks of different amplitudes—e . g . , number of cells that have activated a particular origin vs . amount of ssDNA revealed at different forks—the results from different replicates of the experiment are highly reproducible . We find that origins that are known to fire early and are efficient produce the peaks of greatest magnitude , while later firing and less efficient origins produce smaller or no peaks in this assay [31] . We carried out the ssDNA assay on the wild type ORC4 and orc4Y232C mutant and a representative comparison of the two is shown in Fig 1C ( to view comparisons of all chromosomes see S3 Fig ) . We observed differences in origin usage in ORC4 compared to orc4Y232C and could discern three classes of origins . Some origins had comparable ssDNA peaks in both strains ( cf . origins at 420 kb , 490 kb , 890 kb ) . Some origins that were active in ORC4 showed significant reduction in ssDNA accumulation in orc4Y232C ( cf . 200 kb , 390 kb , 830 kb ) . Perhaps most surprisingly , the converse was also true: some origins that are late or inefficiently initiated in ORC4 cells had significant ssDNA peaks in the orc4Y232C mutant , indicating earlier or more efficient initiation than in wild type ( cf . 15 kb , 170 kb , 670 kb ) . To assess quantitatively the magnitude in change to the early replication program of origin firing , we quantified the area under the peaks observed in the ssDNA assays for ORC4 and orc4Y232 . Peak areas in ORC4 vs . orc4Y232C were highly discordant ( Fig 1D; R2 = ~0 . 14 ) . These results highlight substantial differences in the early origin firing of origins in orc4Y232 cells: of the 213 origins represented in the scatter plot ( Fig 1D ) , 31 are significantly depressed or delayed in activity in the orc4Y232C mutant ( x-axis: outlined in orange ) while another 24 origins are now detected as firing early ( y-axis: outlined in orange ) . However , overall , the number of early firing origins was similar between wild type and orc4Y232 cells . While the ssDNA assay limits us to monitoring only early-firing origins we anticipate that even if no more origin activation occurred ( as is the case for cells with a deletion of CLB5 , the major S phase cyclin ) , the completion of S phase would only be delayed by an additional 15 minutes [32] . The flow cytometry data of a synchronous culture of orc4Y232C cells showed more than a 15 minute delay in completion of S phase and cell division ( Fig 1B ) . For example , by 85 minutes wild type cells had divided and were undergoing a second round of DNA synthesis . The mutant cells do not show a comparable DNA content profile until 120 minutes after release . We hypothesized that this 35 minute delay could be due to cells encountering problems in late S phase and that the gradual shift to 2C DNA content results from slow progress in completing genomic DNA replication . Interestingly , patient-derived lymphoblastoid cell lines harboring an ORC4-MGS mutation also exhibit a delayed S phase transit [27] . An alternative hypothesis for the gradual shift to 2C DNA content that is observed in orc4Y252C cells may be that they have completed genome replication on time and are continuing to replicate mitochondrial DNA but are delaying the G2/M transition for other reasons . However , the observation that cells with the orc4Y252C mutation are unable to form colonies on plates with 200 mM hydroxyurea ( HU; Fig 1E ) , a drug that inhibits ribonucleotide reductase and thereby slows the progress of replication forks , supports the hypothesis that the cell cycle delay stems from a chromosome replication defect . When comparing the flow cytometry profiles of the orc4Y232C cells relative to their wild type ORC4 control cells , we noticed a consistent shift to an approximately 10% lower DNA content for the orc4Y232C cells ( Fig 2A ) regardless of their cell cycle phase . Considering that a haploid yeast genome is ~13 Mb in size , this difference in peak location means that ~1 . 3 Mb of sequence is missing from the mutant . Since the ORC4 and orc4Y232C strains analyzed in this experiment are haploid , the difference in DNA content cannot be ascribed to chromosome aneuploidy . As an alternative possibility , we investigated whether loss of repetitive sequences could explain the shift in the mutant’s flow cytometry histogram . Candidate repetitive DNA species included mitochondrial DNA ( mtDNA; ~85 kb , ~50 copies per cell ) , the native 2-micron plasmid ( 6 . 3 kb , ~50 copies per cells ) , and ribosomal DNA ( rDNA; 9 . 1 kb , 100–200 copies per cell ) [33–35] . The 2-micron plasmid and mitochondrial genomes are autonomous elements and copy number variation can occur through failure of replication and/or segregation [33 , 36] . In contrast , the rDNA locus is a tandemly repeated array located on chromosome XII and copy number changes occur through homologous inter- or intra-chromosome recombination [37 , 38] . To determine which , if any , of these repeated sequences might account for the missing genomic DNA content of the orc4Y232C strain we performed quantitative Southern blotting of the mutant and wild type strains . We found that the mutant cells retained most ( ~88% ) of their mtDNA ( whose replication is ORC-independent ) , but retained only ~21% of the 2-micron plasmid ( a selfish DNA element that is dependent on ORC for maintenance [39] ) and ~23% of their rDNA copies relative to the wild type cells ( Fig 2B ) . The small decrease in mtDNA cannot account for the shift observed in the flow cytometry profiles; however , together the loss of 2-micron and rDNA repeats is sufficient to account for the missing DNA content . Reduction of the 2-micron plasmid would have no significant impact on the cell’s health as cells completely cured of their 2-micron plasmids actually enjoy a slight selective growth advantage [40] . However , the reduction in the rDNA locus could have a more significant impact on cell physiology . The rDNA locus comprises over half of the physical length of chromosome XII and accounts for ~10% of the nuclear genome ( Fig 2C ) [41] . Each 9 . 1 kb repeat contains the template for making ribosomal RNAs ( rRNAs ) , the main structural component of ribosomes [42] . Based on the quantitative Southern blot analysis , we estimate that the rDNA copy number was reduced from ~150 copies in wild type cells to ~30 copies in the mutant . CHEF gel analysis of whole yeast chromosomes confirmed the expected size reduction of chromosome XII in the mutant ( Fig 2D ) . Thus , the deficit of ~120 copies of rDNA in the mutant would account for ~1 . 1 Mb of the missing sequence , with the ~40 fewer copies of 2-micron plasmid ( ~252 kb ) accounting for the remainder of the missing ~1 . 3 Mb . Since the shrinkage of chromosome XII in orc4Y232C cells was such a striking phenomenon , we tested six additional isolates of the mutant to determine if this size reduction was a consistent phenotype of the orc4Y232C mutation . Immediately after selection for loss of the ORC4 allele ( ~20 generations ) , all six isolates had a smaller chromosome XII than wild type , with an rDNA locus of ~30–100 repeats ( Fig 2E ) . To determine the stability of the rDNA locus in the different orc4Y232C isolates over time we monitored the size of their chromosome XII after long-term growth of the strains . We found that after ~100 generations , all six populations had arrived at the same rDNA copy number of ~30 repeats ( S4 Fig ) , suggesting that the orc4Y232C mutation imposes strong selection for cells that have reduced their rDNA copy number . To determine the dynamics of rDNA repeat loss we analyzed samples collected during the ~100 generations of growth for one isolate ( e ) ( Fig 2F ) . Rather than a steady decrease in rDNA copy number during propagation of the culture , we observed a decreasing abundance of the original chromosome XII size with concomitant increased abundance of the final size , with no obvious populations with intermediate sizes . These results support the hypothesis that a sub-population of cells with the final rDNA copy number already existed in the initial culture or appeared shortly thereafter , and had a selective advantage over cells with the longer rDNA . The descendants of the decreased-rDNA variants eventually took over the population , approaching fixation by ~80 generations of growth ( Fig 2F; indicated by the lane marked with an asterisk ) . Next , we asked if the selection for a loss of rDNA repeats was a phenotype specific to the orc4Y232C allele . To test this hypothesis , we introduced a different MGS-like mutation into budding yeast . Some cases of MGS have been reported to be caused by mutations in CDC45 , a core component of the CMG complex , the replicative helicase that travels with the replisome during S phase [26 , 43] . A specific missense mutation ( P463L ) occurs at an evolutionary conserved position of the human Cdc45 protein [26]; therefore , we introduced the equivalent MGS mutations into budding yeast ( cdc45P542L ) . We analyzed the chromosome XII size of five cdc45P542L isolates immediately after selection for loss of the CDC45 allele and found that all the isolates had shortened chromosome XII ( S5 Fig ) . Although we observe a contraction of the rDNA locus in cdc45P542L yeast cells , it is important to note that shortening of this locus in the cdc45P542L and orc4Y232C mutants analyzed may occur through different mechanisms . However , since both Cdc45p and Orc4p are required for DNA replication , we sought to determine if defects during DNA replication may be responsible for this unusual phenotype . Because each rDNA repeat contains a potential origin of replication ( rDNA ARS ) [44] , we asked if compromised replication initiation at the rDNA ARS could be responsible for the delay in completion of S phase and ultimately in the reduction of rDNA copy number . Replication of the rDNA locus is a bit unusual: 1 ) Each repeat contains a potential origin but only a subset of them—usually those rDNA ARSs downstream of transcriptionally active repeats—serve as origins in any given S phase [45] . 2 ) Because of the high transcriptional activity , replication is almost entirely unidirectional—enforced by the replication fork barrier ( RFB ) that blocks forks from entering the 3’ end of the 35S transcription unit in a direction that opposes transcription [44] . Compared to bidirectional replication elsewhere in the genome , the unidirectional replication in the rDNA would require twice the number of initiation events for the same territory to be replicated in the same amount of time . 3 ) The rDNA locus completes its replication late in S phase [46] . We reasoned that if origin initiation at the rDNA locus were less efficient in orc4Y232C cells it could explain the delay in completion of S phase and the reduction in rDNA copy number because cells that had suffered a reduction in rDNA length would complete replication more quickly and their descendants would take over the culture . To test the efficiency of rDNA origin firing , we carried out 2D gel electrophoresis of genomic DNA from cells in logarithmic growth [47] . By focusing on the NheI fragment that contains the rDNA ARS at its center we can detect the fraction of rDNA repeats that gives rise to bubble intermediates ( active origin ) relative to the fraction of repeats in which the origin region is passively replicated by a fork moving through ( inactive origin ) ( Fig 3A ) . The wild type cells showed the expected frequency of repeats with an active origin ( roughly 1 in 5 repeats has an active origin; Fig 3B ) . In contrast , orc4Y232C cells had a greatly reduced “bubble arc” relative to the intensities of the “Y arc” , the RFB pause site and the non-replicating “1N” spot . After adjusting for copy number differences in rDNA repeats between the two strains we conclude that origin activation within the rDNA array is a very rare event in the orc4Y232C mutant cells . What might be the cause for the reduced origin activity of rDNA origins in the mutant cells ? One possibility is that the mutant allele results in a less stable protein , therefore limiting the amount of Orc4Y232C protein available to form Pre-RCs on rDNA origins . Such is indeed the case in orc2-1 mutant cells , where protein destabilization leads to reduced pre-RC formation; simply increasing the dosage of the mutant protein by providing an additional copy of the mutant gene on a plasmid alleviates the mutant phenotype [48] . Accordingly , we asked whether an additional copy of the orc4Y232C allele on a centromere plasmid would provide any level of rescue of the rDNA locus contraction . An orc4Y232C clone with ~30 rDNA repeats was transformed with plasmids containing either the ORC4 or orc4Y232C allele and chromosome XII size was measured after ~30 generations of growth . The plasmid copy of ORC4 resulted in a significant increase in the rDNA locus while the same plasmid with the orc4Y232C allele provided no rescue ( S6 Fig ) . These results suggest that it may not be protein levels that are contributing to reduced rDNA origin firing , but rather , the rDNA origin is less efficiently activated in cells with the mutant ORC complex . To test whether inefficient rDNA origin firing was responsible for contraction of the rDNA locus we reasoned that a compromised rDNA ARS would exacerbate the phenotype . A naturally occurring rDNA variant , obtained from the Robert Mortimer vineyard strain RM11-1a , was introduced into the laboratory background BY4741 [49] . This variant , which has a T to C transition in a highly-conserved residue of the rDNA ACS ( S7 Fig ) , is known to reduce but not eliminate rARS activity [49] . Cells with this weakened rDNA variant in an otherwise wild type background show no change in growth rate ( compare Fig 1A and S8 Fig ) . For convenience , we shall hereafter refer to the lab ( BY4741 ) version of the rDNA as rDNABY and the vineyard strain ( RM11-1a ) variant rDNA as rDNARM . Plating assays of the wild type and orc4Y232C strains are shown in Fig 4A . While growth of the orc4Y232C rDNABY strain was comparable to that of ORC4 strains with either rDNA version at 30°C and 37°C , growth of the orc4Y232C rDNARM strain was reduced at 30°C and was undetectable at 37°C . Furthermore , the slow growth phenotype of the orc4Y232C rDNARM strain is recapitulated in liquid medium at 30°C ( S8 Fig ) : relative to the wild type strain , the growth rate of orc4Y232C rDNARM is reduced by 54 minutes as compared to 18 minutes for orc4Y232C rDNABY . We analyzed the size of the rDNA locus by measuring the BamHI genomic fragment from chromosome XII that contains the intact rDNA locus plus adjacent single copy sequences ( Fig 4B ) by CHEF gel electrophoresis and found that the rDNA locus had undergone an even greater reduction in the orc4Y232C rDNARM strain compared to the orc4Y232C rDNABY strain ( Fig 4C ) —from 30 copies to ~10 copies . An rDNA locus of this size ( ~91 kb ) is at the upper limit of a replicon that could reasonably be expected to be replicated by a fork established at the nearest upstream origin in the flanking unique sequences [50 , 51] . 2D gel analysis ( Fig 4D ) confirmed that there is no detectable origin initiation within the rDNA locus itself . Unlike the cells containing the BY version of the rDNA ARS , all the initial orc4Y232C isolates that contained the RM version of the rDNA ARS showed this drastic reduction in rDNA content and did not require an extended period of growth to achieve the steady-state reduction in rDNA copy number ( Fig 4E ) . Cells released from an alpha-factor arrest into a synchronous S phase showed a similar late-S/G1 delay but an additional and exaggerated delay in entry into S phase ( ~40 minutes in orc4Y232C rDNARM compared to ~20 minutes in ORC4 rDNARM; Fig 4F; compare with Fig 1B ) . If the prolonged G1 to S delay in orc4Y232C rDNARM cells were caused by insufficient origin activation across the genome , we would anticipate the total number of early active origins to be lower in this strain . To test this hypothesis , we carried out the same ssDNA assay as before on both ORC4 rDNARM and orc4Y232C rDNARM cells . The full set of comparisons for all the chromosomes between ORC4 rDNARM and orc4Y232C rDNARM cells are shown in S9 Fig and quantification of the area under peaks and pairwise comparisons between the strains are shown in S10 Fig . Similar to our previous observation , we found differences in the activation of early origins between ORC4 rDNARM and orc4Y232C rDNARM cells; however , the total number of active origins was similar between the two . Additionally , pairwise comparisons of the two ORC4 or two orc4Y232C strains with different rDNAs showed extremely good concordance , indicating that the rDNA ARS genotype does not affect early genome wide replication initiation dynamics . These observations suggest that the prolonged G1 to S transition observed in the cell cycle analysis was not due simply to inadequate genome-wide origin activation in orc4Y232C rDNARM cells but suggest that some aspect of defective rDNA replication contributes to both the entry into and completion of S phase . Why are orc4Y232C rDNARM cells slower to complete S phase and enter cell division ? Are they experiencing difficulty in replicating their chromosome XII due to reduced origin activity at the rDNA locus , and if so , what might the consequences be ? In the process of measuring rDNA copy number in the different isolates of orc4Y232C rDNARM we noticed an additional minor band ( Fig 4E; indicated by an asterisk ) that was present only in the orc4Y232C rDNARM samples and that migrated faster than the predominant chromosome XII band . We ruled out the possibility that the presence of the additional band was due to cross hybridization of the probe we were using because we did not observe its presence in the lane loaded with DNA from ORC4 rDNARM cells . Therefore , we hypothesized that the additional band might be a result of breakage of chromosome XII , specifically at the rDNA locus . As previously mentioned , we estimated that the rDNA copy number of orc4Y232C rDNARM cells is ~10 copies , which should result in a reduction in the overall size of chromosome XII to ~1 . 13 Mb ( Fig 4G ) . If chromosome XII were breaking at the rDNA locus , the two resulting sizes should be ~450 kb ( left of the rDNA locus ) and ~590 kb ( right of the rDNA locus ) and we should be able to detect these two different products using probes specific to different locations along the chromosome ( Fig 4G ) . When hybridizing the Southern blot with a probe specific to the right of the rDNA locus ( R-rDNA ) , in addition to the full-length chromosome XII , we also detected a band that migrated with a similar mobility to chromosomes V and VIII ( ~580 kb ) in the lane loaded with orc4Y232C rDNARM DNA ( Fig 4H ) . However , when probing the same Southern blot for a sequence to the left of the rDNA locus ( L-rDNA ) , the minor band previously detected was not seen; instead , a new minor band appeared at a size of ~450 kb . Based on the locations of the probes used and the different sizes of minor bands detected , we believe that chromosome XII is breaking at the rDNA locus in a small population of orc4Y232C rDNARM cells . To determine if chromosome breakage is a general feature observed at other chromosomes in orc4Y232C rDNARM cells , we next probed the same Southern blot for a sequence on the second largest yeast chromosome ( chromosome IV ) and observed only a single band in both lanes ( Fig 4H ) , indicating that chromosome breakage is specific to the rDNA locus in orc4Y232C rDNARM cells . Lastly , to determine if the chromosome breakage observed in orc4Y232C rDNARM cells was a transient or recurrent event , we assayed for chromosome breakage again after ~60 generations of growth and observed only the major band corresponding to the intact chromosome XII ( Fig 4H ) . This result suggests that the locus specific breakage in orc4Y232C rDNARM cells occurs during a very narrow window , shortly after loss of the ORC4 allele . The combination of the orc4Y232C mutation and the rDNARM locus reduces the growth rate of cells further compared to that of the orc4Y232C rDNABY strain . While reducing the size of the rDNA array could permit completion of genome replication , it may come at a cost—namely , in the ability to make enough ribosomes to support robust growth . Just what is the lower limit of rDNA repeat number a yeast cell can tolerate and still support a normal ribosome population ? When the rDNA copy number in yeast is artificially reduced to ~40 copies there are no obvious negative effects on growth rate and the cells increase the density of Pol I RNA polymerases per rDNA repeat to produce levels of rRNA similar to cells with ~150 copies of rDNA [52] . However , only a finite number of polymerases can be loaded onto a single rDNA repeat before space becomes limited . Based on the estimates of Pol I density and polymerization rates , yeast cells with ~30 copies of rDNA fall just short of being capable of producing levels of rRNA needed to support a normal growth rate [52] . By these calculations , orc4Y232C cells with only ~10 copies of rDNARM could be further compromised in their ability to make rRNA . Therefore , we next tested our hypothesis that in addition to replication defects , the slow growth and prolonged G1 phase observed in orc4Y232C strains is due to an inability to meet the demand for ribosome production . To test whether orc4Y232C cells with a reduced rDNA copy number also have decreased levels of rRNA , we separated the total nucleic acid content from cells by gel electrophoresis ( S11 Fig ) and then independently transferred to hybridization membrane the two different parts of the gel containing either genomic DNA ( Southern blot ) or rRNA ( northern blot ) . We found that both orc4Y232C mutants with a reduced rDNA copy number showed a reduction in 25S rRNA levels compared to their respective wild type counterparts ( Fig 5A ) . However , orc4Y232C rDNARM cells were more severely affected , being able to make only approximately half as much 25S rRNA as wild type cells . If the drastic reduction of rRNA in orc4Y232C rDNARM cells was compromising ribosome production , we proposed that ribosomal proteins would also be reduced in orc4Y232C rDNARM cells . Since the per-cell fluorescence output of GFP is directly proportional to its concentration in living cells [53] , we reasoned that we could determine the relative abundance of ribosomal proteins per cell by GFP labeling a ribosomal protein and then measuring the cells’ relative fluorescence using flow cytometry . Ribosomal protein levels are tightly regulated , with excess proteins being targeted for rapid destruction [54] , so the abundance of ribosomal proteins is a good proxy for the abundance of ribosomes . We constructed strains of the wild type ORC4 and orc4Y232C mutant in the rDNABY and rDNARM backgrounds harboring a GFP-tagged version of the single-copy ribosomal protein Rpl10 ( Rpl10-GFP ) and measured their relative fluorescence during exponential growth by flow cytometry . The histogram of orc4Y232C rDNARM cells is shifted to the left ( a decrease of nearly one log ) compared to wild type ( Fig 5B ) . To determine if the shift were statistically significant , we recorded the fluorescence value for each of the 20 , 000 events recorded for each sample and performed a Wilcoxon Rank-Sum Test . The decrease in fluorescence we observed in orc4Y232C rDNARM cells is statistically significant ( p-value < 2 . 2e-16 ) . We therefore conclude that there are fewer Rpl10-GFP molecules per cell in the orc4Y232C rDNARM strain compared to wild type . The same decrease in fluorescence was not observed in orc4Y232C rDNABY cells ( S12 Fig ) , consistent with the relatively modest reduction of 25S rRNA measured in this strain ( Fig 5A ) . To investigate whether the reduction of ribosomal components in orc4Y232C rDNARM cells might negatively impact their growth when translation is slightly restricted , we tested their ability to grow in the presence of a low concentration of the translation inhibiting drug cycloheximide that is tolerated by wild type ORC4 cells . We observed that orc4Y232C rDNARM cells could form colonies on the plate without cycloheximide; however , their growth was severely inhibited on the plate containing the drug ( Fig 5C ) . This observation supports the notion that the slower growth of orc4Y232C rDNARM cells and their delay in G1 is due to a lower translation capacity , rooted in their inability to make sufficient numbers of ribosomes .
In an attempt to uncover the link between potential chromosome replication defects and the phenotypes that characterize Meier-Gorlin syndrome we sought to identify the cellular and molecular defects conferred by the ORC4-MGS mutation ( orc4Y232C ) using yeast as a model system . We find that haploid yeast cells harboring orc4Y232C at the endogenous chromosomal locus grow slowly , with altered S phase kinetics . Cells have reduced DNA content resulting from the loss of most copies of the nuclear 2-micron plasmid and from shrinkage of the chromosomal rDNA locus from the normal ~150 repeats to 30 repeats . Introduction of a single nucleotide variant in the rDNA origin ( rDNARM ) reduces its function as an origin in ORC4 cells and results in a further shrinkage of the rDNA locus to ~10 copies in orc4Y232C cells . Replication initiation within the rDNA locus is greatly diminished in both rDNA types; in orc4Y232C cells with rDNARM , replication of the rDNA locus is thus mostly dependent on a single fork initiated at one of the adjacent , upstream origins in the unique flanking sequences . We propose that this situation is responsible for the S/G2 delay and for the growth advantage enjoyed by cells that have a shortened rDNA locus . The presence of the rDNARM origin exacerbates all phenotypes associated with the orc4Y232C mutant: slower growth , further reduction in rDNA origin firing , shorter rDNA locus , chromosome XII breakage specifically within the rDNA , and a more rapid selective sweep of the short rDNA variants through the cell population . We also find that the rDNARM strain with the orc4Y232C mutation experiences a drastic decrease of ribosomal components , which may account for the additional ~30 minute G1 delay before cells enter S phase . Across the yeast genome , replication initiation is also altered by the orc4Y232C mutation: some unique chromosomal origins share with the rDNA origin a reduction or delay in their activation but others are advanced in their replication . While we have only examined the origins that fire early in S phase , it seems unlikely , given the density of unique origins and the proportion that are affected by the orc4Y232C allele , that there would be other stretches of the genome that would require a single replication fork to travel ~90 kb—the distance required in the truncated rDNARM locus . Thus , even though replication dynamics are altered genome-wide in the orc4Y232C mutant , we conclude that difficulties in rDNA replication , resulting in a loss of rDNA repeats and ultimately a ribosome deficiency , are at the heart of the phenotypes of yeast harboring this allele . Our work demonstrates a strong selection for reducing rDNA copy number when ORC function is compromised . A link between impaired ORC function and variation in rDNA copy number has been noted in previous studies . Ide et al . showed that temperature sensitive mutations in two other ORC complex proteins , Orc1 and Orc2 , result in the shrinkage of the rDNA locus when cells are grown at the restrictive , or semi-restrictive temperatures [30] . In addition , they demonstrated that the shortened rDNA locus was responsible for suppressing the temperature sensitivity of these mutations . Finally , they demonstrated that replication difficulties in the rDNA triggered a Rad53 checkpoint response and that rDNA reduction attenuates this checkpoint response . Based on these observations , the authors proposed that the rDNA locus plays an important role in monitoring when origin initiation across the genome is compromised . The specific molecular event that was responsible for activation of the checkpoint response was not addressed in that study , nor was it clear why rDNA shrinkage would attenuate the checkpoint response . Our work provides new insights into the observations made by Ide et al . and suggests an alternative interpretation of their results . One limitation of the Ide et al . study is that although their model assumes reduced origin firing across the genome in the orc mutants , origin activity was only examined for one chromosomal origin ( ARS1 ) other than rARS . In our broader assessment of origin activity , we find that ORC4 and orc4Y232C cells have a very similar number of early active origins , arguing that there is unlikely to be replication defects genome-wide . Rather , it appears that the rDNA itself is the “weak link” suffering from replication gaps . Our results suggest that as in our orc4Y232C cells , chromosome breakage at the rDNA locus is likely responsible for triggering the Rad53 response observed by Ide et al . ; we presume that as in the orc4Y232C strain , the orc1 and orc2 cells with reduced rDNA would circumvent the rDNA replication gap problem and would therefore attenuate the checkpoint signal . Finally , although Ide , et al . demonstrated shrinkage of the rDNA locus partially suppressed the temperature sensitive mutations in Orc1 and Orc2 they did not explore how this rDNA shrinkage might affect rRNA synthesis . We find that having fewer rDNA repeats may allow for complete replication of this locus in orc4Y232C cells , but at a cost . Losing too many rDNA repeats , as in the case of orc4Y232C rDNARM cells , limits the amount of rRNA that can be transcribed . Therefore , orc4Y232C cells must walk a tightrope , as it were—too many copies of rDNA and they suffer chromosome XII instability , but too few copies and they are unable to meet the demand for ribosomal RNAs . A second study expanded on the idea that the ~150 rDNA origins compete for limiting replication initiation factors with the ~300 unique origins across the yeast genome [55] . The authors of that study discovered that this competition is regulated oppositely by two histone deacetylases—Sir2 and Rpd3 . When initiation in the rDNA is increased , replication at some unique genomic origins is reduced , and vice versa . In their experiments , ORC was not the limiting factor . Instead , they showed that they could increase initiation at unique origins in a sir2Δ strain by overexpressing three of the initiation factors known to be in limiting supply ( Sld7 , Sld3 and Cdc45 ) . We initially entertained the possibility that reduction of rDNA copy number in the orc4Y232C mutant strain might restore a more favorable replication initiation balance between the rDNA and unique origins . However , three observations make this explanation unlikely . First , we did not find an overall reduction in unique chromosomal origin firing in the orc4Y232C mutant—the number of early origins that failed to fire was similar to the number of new origins that appeared in the early firing class . Second , if competition between rDNA origins and unique origins were causing the growth defect in orc4Y232C , reducing the efficiency of the rDNABY origin by replacing the locus with rDNARM should have improved growth of the orc4Y232C strain—instead , growth was further restricted and the only locus to suffer from failed replication initiation was the rDNA . Third , introducing a second copy of orc4Y232C on a centromere plasmid did not produce any rescue in the size of the rDNA locus—suggesting that the Orc4Y232C protein was not in limiting supply . Together , these observations led us to hypothesize that the orc4Y232C mutation has altered the protein’s function . What aspect of ORC function has been altered by the orc4Y232C mutation ? One possibility is that this single amino acid substitution has changed the DNA sequence recognized by the ORC complex . Origins in budding yeast share a similar core sequence called the ACS ( ARS Consensus Sequence , an approximately 17 bp AT-rich sequence necessary but not sufficient for origin function ) and variation in this sequence has been shown to impact origin usage [2] . Analysis of the different groups of origins did not reveal any simple pattern ( s ) of polymorphism within the ACS that distinguished wild type specific origins ( i . e . , origins that showed reduced activity in the mutant ) from mutant specific origins ( i . e . , origins that had ssDNA peaks in the mutant but not in wild type ) , although we cannot rule out the possibility that such sequence differences do exist . A second possibility is that there is some aspect of chromatin structure or nuclear architecture that is influencing origin choice in the orc4Y232C mutant . One class of origins whose activity was influenced by the orc4Y232C allele was those near centromeres: 11 of the 31 wild type specific origins were located within 10 kb of a centromere ( S13 Fig ) . To determine if this apparently skewed distribution of wild type specific origins was significantly different than would be expected to occur by chance , we performed a permutation test . We randomly labeled 31 of the of the 213 origins used to generate the scatter plots ( Fig 1D and S10 Fig ) as wild type specific and asked whether at least 11 of those origins were within 10 kb of a centromere . In 10 , 000 trials of this test , we found no occurrence of 11 or more affected origins being within 10 kb of a centromere ( p < 10−4 ) . Relevant to this discussion is the observation that we and others have made that indicates that centromeres promote early firing time of origins in their vicinity [56–58] . The early firing of centromere proximal origins is thought to occur as a consequence of the kinetochore protein Ctf19 recruiting the S-phase kinase , DDK , to phosphorylate components of the pre-RC for replication initiation [57] . Whether the ORC complex with the orc4Y232C variant is deficient in this interaction is unknown . Lastly , nucleosome occupancy or transcription factor binding around the origin may be preventing ORC containing Orc4Y232C protein from binding to some origins . Moving forward , it will be important to determine if the centromere itself , chromatin state , or specific proteins that make up the kinetochore are influencing these changes . ORC is known to play important roles in cellular processes other than DNA replication; however , its role maintaining a healthy ribosome population has yet to be characterized . Here we show that properly functioning ORC is necessary for maintaining adequate ribosomal RNA levels through its role in rDNA replication . But when the function of Orc4 is compromised , leading to a loss of rDNA repeats and the capacity to make rRNA , how might a ribosome deficiency cause slow growth in cells ? The obvious answer would be that cells cannot meet the demand for protein production—but how might certain processes such as translation and protein degradation be affected ? Does the translation of certain housekeeping mRNAs take priority over non-essential ones or do cells manage to deal with a ribosome deficit by speeding up the rate of translation of all mRNAs equally ? Ribosome profiling experiments to identify and quantify the mRNAs that are being actively translated by ribosomes will prove helpful in addressing these questions [59] . Additionally , if cells are struggling to keep up with the protein synthesis demands , it is possible that they modulate their protein turnover rates in response . Coupling mass spectrometry based proteomics with metabolic pulse-labeling or cycloheximide treatment of cells will help shed light on the rate of protein turnover and the general state of the proteome in cells with a ribosomal RNA deficiency [60] . Is our proposed model for loss of rDNA repeats relevant to higher eukaryotes ? Just as in yeast , the human rDNA is present as tandemly repeated arrays; however , there are some notable differences . In yeast the rDNA locus is located on a single chromosome with two transcription units separately producing the 5S rRNA transcript and 35S rRNA precursor , which is processed into the 18S , 5 . 8S and 25S rRNAs [42] . In humans , rDNA clusters are located on six different chromosomes [61] . A locus near the end of the long arm of Chromosome 1 contains 50–200 copies of a 2 . 2 kb repeat that produces the 5S rRNA [61] . Loci at the ends of the short arms of Chromosomes 13 , 14 , 15 , 21 , and 22 contain 43 kb repeats coding for a 47S transcript that gets processed into the 5 . 8S , 18S and 28S rRNAs [61] . The number of repeats on each chromosome varies and individuals show a wide distribution of copy numbers at these five loci , ranging from between 10 to more than 100 repeats per locus [62] . While most chromosomal origins in humans are thought not to be defined by primary DNA sequence , replication initiation events in the 43 kb rDNA repeats are confined to the non-transcribed spacer [63] . What might be the consequences of impaired replication at the rDNA locus in higher eukaryotes ? Bloom’s syndrome ( BLM ) is a rare autosomal recessive disorder characterized by short stature , immunodeficiency , and predisposition for cancer [64] . Individuals with BLM have mutations in the gene BLM , which encodes a member of the RecQ family of DNA helicases that acts during DNA replication [64] . Cells derived from individuals with BLM exhibit a high frequency of sister chromatid exchange and genomic instability [64] . Particularly , BLM cells exhibit elevated levels of instability at the rDNA locus [65]; however , the exact cause for this instability is unknown . Deletion of the BLM orthologue SGS1 in budding yeast also results in increased instability at the rDNA locus [66] . Additionally , cancers in both human and mouse have also been shown to exhibit elevated levels of instability at the rDNA locus [67 , 68] . Stultz et al . found an increase in rDNA rearrangements in the majority of tumor samples they analyzed from lung and colorectal cancers , leading them to the conclusion that the rDNA locus is a recombination hotspot in some cancers [67] . Lastly , mice deficient for MCM2 ( a component of the Pre-RC ) develop lymphomas and exhibit elevated levels of DSBs at the 45S rDNA repeats in their genomes [68] . What could be the consequence ( s ) of increased instability at the rDNA locus in cancer ? Recently , Xu et al . analyzed rDNA copy number in both human and mouse cancer genomes and contrary to their initial prediction , they found that rDNA copy number is reduced in the cancer state [69] . One explanation the authors propose for this unexpected result is that having fewer rDNA repeats may allow for more efficient DNA replication and thus greater cell proliferation in cancer . Given the fact that the rDNA locus is highly sensitive to replication stress in higher eukaryotes and its copy number can change rapidly during the disease state , a similar problem with rDNA replication—due to reduced origin initiation—could lead to a reduction in rDNA copy number and possibly limit the amount of rRNA and thus reduce ribosome levels in higher eukaryotes . Meier-Gorlin syndrome phenotypes include a number of skeletal defects that have similarities to other , better-understood syndromes , all associated with deficiencies in ribosome biogenesis [70] . For example , Treacher Collins syndrome ( TCS ) can be caused by one of three different genetic mutations that are likely to affect ribosome production [71 , 72] . One of the TCS mutations is in TCOF1 , a gene that encodes a protein called Treacle that is involved in rDNA transcription through its interaction with upstream binding factor ( UBF ) [73] . Treacle also interacts physically with human Nop56p , a component of the rRNA modifying box C/D small nucleolor ribonucleoprotein complex [74] . The other two genes , POLR1C and POLR1D , are subunits of RNA Polymerases I and III [72] . Work with mouse models of this syndrome suggest that ribosome deficiencies are reducing the migration of neural crest cells that are essential for proper craniofacial development [75] . Whether derived from neural crest cells or mesenchymal cells , chondrocytes also have a high demand for ribosomes as they divide , enlarge and produce collagen and other bone-matrix proteins [70] . In addition to TCS , Postaxial Acrofacial Dystosis ( POADS ) , Diamond Blackfan Anemia ( DBA ) , Roberts Syndrome ( RBS ) , Schwachman-Diamond Syndrome ( SDS ) , and Cartilage-Hair Hyploplasia ( CHH ) are distinct ribosomopathies with a common set of skeletal malformations [70] . Their genetic mutations affect ribosome production in different ways—a mutation that reduces uracil biosynthesis ( POADS ) [76 , 77] , mutations in individual ribosomal protein genes ( DBA ) ( reviewed in[78 , 79] ) , a mutation in a gene for cohesion ( ESCO2 ) that in yeast ( ECO1 ) reduces 18S and 28S production ( RBS ) [80 , 81] , a mutation in SBDS ( yeast SDO1 ) that reduces maturation of the 60S ribosomal subunit ( SDS ) [82–84] , and a mutation in RMPR , the RNA component of RNase MRP , a snoRNA involved in rRNA processing ( CHH ) [85 , 86] . A ribosomopathy that is associated with a deficiency of rDNA has not yet been identified; however , one might speculate on the phenotypes that may be manifested with such a disorder [87] . Further work will be necessary to determine whether the phenotypes in orc4Y232C yeast are consistent with cells derived from individuals with MGS . Additionally , determining if the slow growth phenotype observed in orc4Y232C yeast is due primarily to defects in chromosome replication or a ribosome deficiency will prove necessary in elucidating how mutations in proteins necessary for origin initiation may inadvertently affect cellular processes in addition to DNA replication .
A complete list of yeast strains and plasmids can be found in the S1 Table . BY4741 was used as wild type ( ORC4 ) for this study and all strains were derived from this background . The rDNA locus from the Robert Mortimer vineyard strain RM11 was introduced into the BY4741 background by standard backcrossing ( 10 times ) to create ORC4 rDNARM [49] . Subsequently , the MGS-like variant orc4Y323C was introduced into either ORC4 rDNABY or ORC4 rDNARM by two-step gene replacement [88] . A plasmid containing URA3 and the orc4Y232C allele was integrated at the ORC4 locus; correct integrants were confirmed by PCR and Southern analysis . We selected for loss of the integrated sequences through homologous recombination by selecting against the URA3 gene on plates containing 5-FOA . To screen for clones that had lost the wild type ORC4 allele and had kept the orc4Y232C allele we performed PCR using an allele specific oligonucleotide as one of the PCR primers . The MGS-like allele cdc45P542L was introduced into BY4741 using CRISPR-Cas9 following the steps described by Laughery et al [89] . For all experiments , yeast cultures were grown at 30°C in synthetic complete medium supplemented with 2% glucose unless stated otherwise . Protein structure visualization was performed using UCSF Chimera [90] . Cell cycle progression was examined using flow cytometry . Early log phase cells ( OD660 ~0 . 25 ) were arrested in G1 by the addition of alpha factor at a final concentration of 3 μM . When >90% of the population was un-budded ( time equivalent to ~1 . 5 population doublings ) , cells were synchronously released into S-phase by the addition of Pronase ( Calbiochem ) at a final concentration of 0 . 3 mg/ml . After the release into S phase , cells were harvested at 5-minute intervals , mixed with sodium azide at a final concentration of 0 . 1% and fixed with 70% ethanol . Cells were prepared for flow cytometry as previously described [32] . Flow cytometry was performed on a BD FACSCanto II and data were analyzed using FlowJo software . Stationary phase cells were embedded in agarose plugs and prepared using standard procedures [91] . CHEF gel analysis of whole yeast chromosomes was performed using a BioRad CHEF-DR II Pulsed Field Electrophoresis System . Whole chromosomes were resolved in 0 . 8% LE agarose gels with a switch time ramped from 300–900 seconds at 100 volts for 68 hours in 0 . 5X TBE at 14°C . BamHI digested chromosomal DNA fragments were resolved in 1 . 0% LE agarose gels with a switch time ramped from 47–175 seconds at 165 volts for 62 hours in 0 . 5X TBE at 14°C . Gels were stained with ethidium bromide and photographed . Southern blotting and hybridization were performed using standard procedures . Mid-log phase cells were harvested and embedded in 0 . 5% low melt agarose ( SeaPlaque ) in 50 mM EDTA and prepared as previously described [92] with slight modifications ( see http://fangman-brewer . genetics . washington . edu/plug . html ) . DNA was subsequently digested in-gel using NheI ( NEB ) . 2D gel electrophoresis was used to visualize the relative abundance of replication intermediates and was performed as described by [47] . Gels were blotted and hybridized with a probe specific to the rDNA origin of replication . DNA was harvested from stationary phase cells via the “Smash-and-grab” DNA isolation protocol [93] and subsequently digested with EcoRV ( NEB ) . Digested DNA was separated by electrophoresis in a 0 . 7% Agarose LE ( GeneMate ) gel and then blotted following standard Southern blotting protocols . Blots were then sequentially hybridized with probes specific to ACT1 ( single copy control ) , mtDNA , 2-micron plasmid , and then rDNA . The hybridization signal was analyzed using the BioRad Personal Molecular Imager and Quantity One software . Hybridization signals of repetitive sequences were first normalized to ACT1 and then relative to wild type . Our initial replacement of ORC4 with the mutant orc4Y232C allele was performed using a “pop-in/pop-out” strategy ( [88]; see Yeast strains and plasmids section above ) . To track long-term rDNA copy number changes following replacement of wild type with mutant orc4Y232C , we picked fresh “pop-out” candidates and inoculated liquid cultures for growth to saturation ( ~30 generations ) . After confirming the loss of ORC4 by allele specific PCR , plugs for CHEF gel analysis were made for the orc4Y323C positive isolates ( a-f; Fig 2D ) . From the initial overnight cultures , a 1/100 dilution was made into 5 ml of fresh medium and allowed to grow to saturation ( ~7 generations ) . Growing of cells to saturation and then diluting back into fresh medium was performed for a total of 10 times which accounted for a total of ~100 generations of growth . In each round , cells were harvested when the culture reached saturation and embedded in agarose plugs for CHEF gel analysis . The relative abundance of 25S ribosomal RNA was measured by quantitative hybrid Southern/northern blotting . The total nucleic acid content of cells was extracted using a version of the “Smash-and-grab” DNA isolation protocol [93] with the modification that cell walls were enzymatically disrupted using Zymolyase - 20T ( Amsbio ) instead of glass beads . Nucleic acids were resuspended in TE pH 8 and then separated by electrophoresis in a 1 . 5% LE agarose gel with ethidium bromide ( 0 . 3 μg/ml ) . After the ribosomal RNA was separated away from genomic DNA , the gel was photographed and then cut to separate the two portions containing genomic DNA and rRNA . Subsequently , the two different portions of the gel were blotted following standard Southern ( genomic DNA ) and northern ( rRNA ) blotting protocols . The Southern blot was probed for ACT1 as a single copy control and the northern blot was probed for 25S rRNA sequence . Because the amount of rRNA on the blot could be in excess of the probe , hybridization of the northern blot was limited to 2 hours to ensure that the hybridization signal was proportional to the amount of target sequence . The hybridization signals were analyzed using a Bio-Rad Personal Molecular Imager and Quantity One software . The 25S rRNA hybridization signal was first normalized to ACT1 and then relative to wild type . The GFP fluorescence of living cells was measured using flow cytometry . Strains were grown overnight in synthetic complete medium at 30°C . Fresh cultures were made by diluting these overnight cultures back to a starting OD660 of ~0 . 05 . When cultures reached an OD660 of ~0 . 6 , cells were diluted and sonicated , then analyzed directly using a BD Accuri C6 flow cytometer . Flow cytometry data was exported and analyzed using FloJo software . A detailed protocol for this assay has previously been published [94] . Cells growing in log phase ( OD660 ~0 . 25 ) were arrested in G1 by the addition of alpha factor and then synchronously released into S phase in the presence of 200 mM HU . Samples were collected every 15 minutes after release into S phase and cells were embedded in agarose plugs and then spheroplasted . The ssDNA from either S phase or G1 control samples were then differentially labeled with either Cy5- or Cy3-dUTP by in-gel random-primed labeling using exo- Klenow polymerase ( NEB ) without denaturation of template . The differently labeled DNAs were then collected and co-hybridized to Agilent G4493A yeast 4x44K ChIP to chip DNA microarrays according to the manufacturer’s recommendations . The data from scanned microarrays was extracted using Agilent’s Feature Extraction software . The ssDNA microarray data are available at NCBI GEO under accession no . GSE104671 . Areas under ssDNA peaks were assessed from Loess-smoothed microarray data ( coordinates spaced 500 bp apart ) using a custom Python script and a reference list of origins and their locations from OriDB [95] ( http://cerevisiae . oridb . org/ ) . Since most of the genome is double-stranded and therefore not a template for ssDNA labeling , the mean genome-wide signal was used as a “threshold” value for each sample . The “ssDNA peak area” for each origin was then calculated as a sum of S/G1 value at the origin’s location and the sequentially-added S/G1 values from adjacent data points until three “below threshold" values were reached on each side of the origin . The script does not distinguish overlapping origin peaks and therefore overlapping early firing origins in close proximity were manually curated and excluded from the analysis . This Python script is available as a supplemental file ( S1 File ) .
|
The origin recognition complex ( ORC ) is essential for licensing replication origins during M/G1 for their firing in the subsequent S phase . Individuals with a rare form of dwarfism called Meier-Gorlin syndrome ( MGS ) have mutations in proteins required for origin activation , including various subunits of ORC . To better understand the molecular and cellular consequences of these mutations , we introduced an equivalent MGS mutation in ORC4 into yeast . We find that origin activity in the ribosomal DNA ( rDNA ) repeats is severely compromised in yeast that harbor the MGS allele . Consequently , cells that have reduced their rDNA copy number from ~150 to fewer than 30 copies overtake the culture . Although the loss of rDNA repeats helps ensure the complete replication of chromosome XII during S phase , cells with fewer rDNA repeats struggle to meet the high demand for ribosomal RNA .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"molecular",
"probe",
"techniques",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"fungi",
"model",
"organisms",
"dna",
"replication",
"experimental",
"organism",
"systems",
"dna",
"molecular",
"biology",
"techniques",
"synthesis",
"phase",
"cellular",
"structures",
"and",
"organelles",
"gel",
"electrophoresis",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"electrophoretic",
"techniques",
"electrophoretic",
"blotting",
"molecular",
"biology",
"genetic",
"loci",
"ribosomes",
"yeast",
"biochemistry",
"rna",
"southern",
"blot",
"eukaryota",
"ribosomal",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"non-coding",
"rna",
"organisms"
] |
2017
|
Defective replication initiation results in locus specific chromosome breakage and a ribosomal RNA deficiency in yeast
|
Topoisomerase II is a major component of mitotic chromosomes but its role in the assembly and structural maintenance of chromosomes is rather controversial , as different chromosomal phenotypes have been observed in various organisms and in different studies on the same organism . In contrast to vertebrates that harbor two partially redundant Topo II isoforms , Drosophila and yeasts have a single Topo II enzyme . In addition , fly chromosomes , unlike those of yeast , are morphologically comparable to vertebrate chromosomes . Thus , Drosophila is a highly suitable system to address the role of Topo II in the assembly and structural maintenance of chromosomes . Here we show that modulation of Top2 function in living flies by means of mutant alleles of different strength and in vivo RNAi results in multiple cytological phenotypes . In weak Top2 mutants , meiotic chromosomes of males exhibit strong morphological abnormalities and dramatic segregation defects , while mitotic chromosomes of larval brain cells are not affected . In mutants of moderate strength , mitotic chromosome organization is normal , but anaphases display frequent chromatin bridges that result in chromosome breaks and rearrangements involving specific regions of the Y chromosome and 3L heterochromatin . Severe Top2 depletion resulted in many aneuploid and polyploid mitotic metaphases with poorly condensed heterochromatin and broken chromosomes . Finally , in the almost complete absence of Top2 , mitosis in larval brains was virtually suppressed and in the rare mitotic figures observed chromosome morphology was disrupted . These results indicate that different residual levels of Top2 in mutant cells can result in different chromosomal phenotypes , and that the effect of a strong Top2 depletion can mask the effects of milder Top2 reductions . Thus , our results suggest that the previously observed discrepancies in the chromosomal phenotypes elicited by Topo II downregulation in vertebrates might depend on slight differences in Topo II concentration and/or activity .
Type II topoisomerases are large ATP-dependent homodimeric enzymes that transiently cleave double stranded DNA , pass a second DNA double helix through the break , and then reseal the break [1] , [2] . In this way , Topo II enzymes solve a variety of topological problems that normally arise in double stranded DNA during processes such as replication , transcription , recombination and sister chromatid segregation [1] , [2] . Topo II enzymes are structurally and functionally conserved , and the genomes of all eukaryotes harbor at least one Topo II enzyme . Vertebrates have two Topo II isoforms , alpha and beta [3]; these enzymes have identical catalytic activities but distinct localization patterns during mitosis . The beta isoform is primarily cytoplasmic , while most of Topo II alpha is concentrated in mitotic chromosomes [4] . In contrast , yeast and Drosophila have a single Topoisomerase II ( Top2 ) gene . Notably , both the Drosophila Top2 and each of the human Topo II genes can rescue the phenotype of yeast Top2 mutants , highlighting the strong functional conservation of type II topoisomerases [5]–[7] . Topo II alpha is a major component of vertebrate mitotic chromosomes [8]–[10] . In vivo studies have shown that Topo II alpha has a dynamic behavior and that chromosome-associated Topo II alpha is rapidly exchanged with the cytoplasmic pool [4] , [11] , [12] . In fixed mitotic chromosomes , Topo II alpha exhibits a discontinuous localization pattern with Topo II alpha alternating with cohesin along chromatid axes [13] , [14] . There is also evidence that in some systems Topo II alpha accumulates at centromeres in prometaphase and metaphase , suggesting a role of this enzyme in the regulation of centromere structure and/or cohesion [4] , [12] , [15]–[17] . Studies in yeast have shown that Top2 is not required for completion of DNA synthesis but plays essential roles in mitotic chromosome condensation and sister chromatid segregation . Failure to decatenate sister chromatids results in anaphase chromatin bridges that cause chromosome breakage during anaphase or cytokinesis [18]–[22] . Loss of Topo II activity does not affect S phase progression and disrupts sister chromatid separation also in vertebrate cells [2] , [23]–[25] . However , the role of Topo II in vertebrate chromosome structure is rather controversial , possibly due to species-specific differences in chromosome organization and/or the different methods used to inhibit Topo II function ( chemical inhibitors , immunodepletion , mutations or RNAi ) . For example , treatment of Indian muntjac cells with the Topo II inhibitor ICRF-193 caused frequent failures in sister chromatid individualization , the process by which duplicated DNA is resolved into two distinct chromatids [26] . In contrast , Topo II inhibition with ICRF-193 did not affect sister chromatid resolution in both Chinese hamster ovary cells ( CHO ) and baby hamster kidney ( BHK ) cells [27] . Moreover , Topo II alpha inhibition in different vertebrate systems resulted in a variety of chromosome morphologies ranging from relatively mild effects on the axial compaction of chromosomes to severe defects in chromosome condensation [14] , [24] , [28]–[32] . Although Topo II beta is not normally able to compensate for Topo II alpha loss , overexpression of Topo II beta in human cells can correct the defects caused by Topo II alpha depletion [33] . In addition , it has been shown that DT40 avian cells and human cells depleted of both Topo II alpha and Topo II beta display chromosomal defects more severe than those observed in cells lacking Topo II alpha alone [17] , [24] , [25] , [32] . These results suggest that the two Topo II isoforms are partially redundant . Thus , studies of cells depleted of both Topo II alpha and Topo II beta are particularly relevant to define the role of Topo II in the maintenance of proper chromosomal architecture . An analysis of DT40 avian cells conditionally depleted of both Topo II isoforms showed that they exhibit extensive anaphase chromatin bridges , defective cytokinesis and polyploid cells [24] . Similar defects were observed in human cells lacking both Topo II alpha and Topo II beta [17] , [25] , [32] . In addition , cytological analysis showed that the latter cells exhibit severe defects in chromosome structure ranging from chromosome entangling to disrupted chromosome morphology [25] . Another controversial issue is the existence of a decatenation checkpoint triggered by loss of Topo II activity . The existence of such a checkpoint was suggested by studies in human cells showing that catalytic inhibitors of Topo II such as ICRF-187 or ICFR-193 are able to induce a caffeine-sensitive G2 delay that is dependent on ATR and BRCA1 , but apparently independent of the DNA damage checkpoint [34]–[36] . However , a decatenation checkpoint is not present in both S . cerevisiae and S . pombe , in which Top 2 mutations cause minimal cell cycle delays [2] , . Recent RNAi-based studies have also shown that depletion of Topo II alpha alone or both Topo II alpha and II beta does not trigger a decatenation checkpoint in vertebrate cells [24] , [25] . In addition , no G2 delay was observed in Topo II-depleted vertebrate cells that were also treated with ICRF-193 [24] , [38] . In contrast , expression of certain mutant forms of Top2 resulted in a G2 arrest in budding yeast cells [22] , [37] . Collectively , these results indicate that loss of Topo II , and thus DNA catenation per se , is not able to induce a cell cycle delay and that a G2 checkpoint is instead activated by specific DNA lesions caused by catalytically inactive forms of Topo II . Several studies have addressed the role of Top2 in Drosophila . Early work showed that injection of anti-Top2 antibodies or Top2 inhibitors into live Drosophila embryos result in strong defects in chromosome condensation and sister chromatid segregation at anaphase [39] . Two RNAi-based studies on S2 tissue culture cells showed that Top2-depleted fixed cells exhibit defects in longitudinal compaction of chromosomes , defective sister chromatid segregation and extensive anaphase bridges [40] , [41] . Another study performed on live S2 cells , in which Top2 activity was reduced by either RNAi or chemical inhibition , did not detect defects in chromosome condensation and suggested that Top2 is required for centromere resolution and to prevent incorrect microtubule-kinetochore attachment [42] . Remarkably , downregulation of Top2 did not affect the mitotic index , indicating that Top2 deficiency does not activate cell cycle checkpoints in S2 cells [40] , [42] . Other studies suggested that Drosophila Top2 is involved in homolog pairing in cell cultures [43] , modulation of insulator function [44] , and regulation of polytene chromosome structure [45] . Surprisingly , the role of Top2 in the maintenance of mitotic and meiotic chromosome structure in living flies has never been investigated . In a previous study we identified viable mutants in the solofuso ( suo ) gene . suo1 and suo2 mutant spermatocytes exhibit severely defective ana-telophases with extensive chromatin bridges; these telophases give rise to achromosomal secondary spermatocytes that are able to assemble bipolar spindles and divide in the complete absence of chromosomes [46] . Here we show that suo1 and suo2 are weak mutant alleles of the Top2 gene and describe the isolation and characterization of a stronger Top2 mutation . We show that modulation of Top2 function by means of these mutant alleles and in vivo RNAi results in multiple cytological phenotypes including site-specific chromosome aberrations , heterochromatin undercondensation , polyploidy , and complete disruption of chromosome morphology accompanied by a cell cycle arrest . These phenotypes recapitulate most of the phenotypes observed in vertebrate cells and indicate that Drosophila chromosomes are exquisitely sensitive to the residual level of Top2 in the cell .
In a previous study we identified and characterized suo1 and suo2 , two viable but sterile mutants that exhibit many chromatin bridges and lagging chromosomes in male meiosis [46] . suo was originally mapped to the polytene interval 37C-37F5 uncovered by Df ( 2L ) VA17 [46] . Using the Exelis kit of deficiencies , we restricted this interval to the polytene band 37E1 , uncovered by Df ( 2L ) Exel9043 , which contains four genes: CG10237 , RanGAP , Hs2st and Topoisomerase II ( Top2 ) . Complementation tests revealed that suo mutations are not allelic to either CG10237 or RanGAP . However , sequencing of Hs2st and Top2 in suo mutants did not reveal alterations in the protein coding exons of these genes with respect to those of another stock of the “Zuker collection” from which the suo mutants have been originally isolated ( see Materials and Methods for detailed explanation ) . We thus screened a collection of 1 , 500 lethal mutations for those that fail to complement suo1 ( see Materials and Methods ) . We isolated a lethal mutation that in combination with suo1 is semi-lethal and elicits the same meiotic phenotype observed in suo mutants ( see below ) ; this mutation was initially named suo3 . DNA analysis of suo3 mutants revealed a G-A transition at nucleotide 1 , 040 of the Top2 coding sequence , corresponding to the 5′ splicing site of the second intron of the gene . This substitution results in a premature stop codon that would lead to a truncated form of the protein . These results indicate that suo3 is allelic to Top2 . Thus , we renamed suo1 , suo2 and suo3 as Top2suo1 , Top2suo2 and Top2suo3 , respectively . We note that we analyzed only the protein coding sequences of the Top2suo1 and Top2suo2 mutants but not the introns the UTRs or the 5′ regulatory sequences . Thus , the molecular lesions resulting in these mutations remain to be determined . As previously reported , both Top2suo1 and Top2suo2 homozygous flies are viable but sterile in both sexes [46] . Top2suo3 homozygous individuals die during embryonic development , but this early lethal phenotype is due to a second site mutation , as Top2suo3/Df ( 2L ) Exel9043 ( henceforth Top2suo3/Df ) animals survive until an early third instar larval stage . Top2suo3/Df larvae are devoid of imaginal discs like strong mitotic mutants [47] and exhibit small brains compared to controls at the same stage . Top2suo2/Df individuals are viable but male and female sterile; Top2suo1/Df and Top2suo1/Top2suo3 mutant flies are semi-lethal , with a few escapers . Together these results allow the ordering of our Top2 mutations in an allelic series with Top2suo3>Top2suo1>Top2suo2 . To assess the amount of the Top2 protein in different mutant combinations , we performed a Western blot analysis on larval extracts using a polyclonal antibody directed against Drosophila Top2 [48] . In wild type extracts , this antibody recognized a single band of the expected size ( 170 kDa ) . The intensity of this band was reduced to approximately 80% of the control level in extracts from Top2suo1 homozygous brains , showed further reduction in extracts from both Top2suo1/Df and Top2suo1/Top2suo3 larvae , and was virtually undetectable in Top2suo3/Df extracts ( Figure 1A , B ) . Immunostaining with an anti-Top2 antibody [49] showed that Top2 is enriched in interphase nuclei and mitotic chromosomes . In particularly favorable preparations we could observe a discontinuous Top2 distribution on chromosomes with alternating stained and unstained regions and no specific centromeric enrichments ( Figure 1C–E ) . This staining pattern is consistent with that observed in vertebrate chromosomes in which the Topo II negative regions are more enriched in condensins compared with the Topo II positive regions [13] , [14] . An alternating condensin/Top2 immunostaining pattern has been previously observed also in Drosophila S2 cells , although it was less sharply defined than in vertebrate chromosomes [42] , [50] . In Top2suo1/Df and Top2suo1/Top2suo3 brains immunostained for Top2 , both mitotic chromosomes and interphase nuclei showed a weaker fluorescence compared with wild type ( Figure 1F and G ) ; the nuclei of Top2suo3/Df did not show any Top2 signal . These results agree with the genetic data and indicate that Top2suo1 and Top2suo2 are hypomorph alleles , while Top2suo3 is a very strong and potentially null allele . The availability of flies expressing different levels of Top2 allowed us to examine the roles of this protein in a physiological context . We focused on the phenotypes elicited by Top2 mutants in three different Drosophila tissues: testes , brain and salivary glands . During prophase I , Drosophila spermatocytes organize their chromatin in 3 main distinct clusters that localize at the periphery of the nucleus . Each of these clusters corresponds to one of the major Drosophila bivalents ( X-Y , 2-2 and 3-3 ) ; the small fourth chromosome bivalent sometimes is separated from the three major chromatin clumps but more often associates with the X-Y bivalent . Before meiosis , chromosomes become progressively individualized within each chromatin territory but bivalents remain well separated from each other until metaphase I , when they congress at the center of the cell [51]–[54] . Previous work has shown that chromatin clumps morphology is affected by mutations in the condensin II coding genes and in genes that specify proteins required for homologous paring of the achiasmatic male chromosomes [55]–[58] . We thus analyzed chromatin organization within the prophase nuclei of Top2suo1/Df primary spermatocytes . To compare mutants and wild type spermatocytes , we stained cells for DNA ( with DAPI ) , tubulin and the centriole marker DSpd-2 [59] , and used centriole length as an additional criterion for stage identification ( Figure 2 ) . Top2 spermatocytes at stages S2 and S3 were not very different from those of wild type ( see ref [52] for stage description ) . However , at stage S4 they showed a higher number of distinct chromatin masses than wild type controls . This increase in chromatin masses was more evident in S5 spermatocytes; Top2suo1/Df spermatocyte nuclei displayed an average number of 5 . 6 chromatin masses per cell ( n = 60 ) while control cells showed an average of 3 . 4 masses/nucleus ( n = 60 ) . Notably , in Top2 mutant spermatocytes at stages S4 and S5 chromatin clumps of similar size were often closely apposed , suggesting that homologs were unpaired but remained in the vicinity of one another ( Figure 2 ) . In the subsequent stages of Top2/Df spermatocyte growth , the number of chromatin clumps progressively reduced . At prometaphase ( stage M1 ) , mutant spermatocytes displayed 3 compact clumps like wild type cells ( Figure 2 , and ref [46] ) ; at metaphase ( stage M3 ) they showed a central chromatin mass like wild type controls [46] . Collectively , these results indicate that during the early stages of spermatocyte growth Top2 is required for homolog co-mingling in a single nuclear territory . However , at later stages of spermatocyte development the homologs appear to pair through a Top2-independent mechanism . Previous studies characterized the meiotic phenotype of mutant males homozygous or hemizygous for Top2suo1 or Top2suo2 [46] . Here we examined the meiotic phenotype of flies bearing the Top2suo3 allele after staining for tubulin , DSpd-2 and DNA . In Top2suo3/Df testes , no meiotic divisions were found due a severe defect in germ cell proliferation . The meiotic phenotype of Top2suo1/Top2suo3 males was similar to that of Top2suo1/Df males , with all ana-telophase I figures characterized by highly defective chromosome segregation and chromatin bridges ( Figure 3A , B ) . As previously described , in approximately one half of these aberrant ana-telophases , all chromosomes segregated to a single pole , giving rise to secondary spermatocytes that divided in the complete absence of chromosomes . Secondary spermatocytes that received some chromosomes also displayed extensive defects in chromosome segregation with frequent chromatin bridges ( Figure 3A; see also ref [46] ) . These results demonstrate that Top2 is required for chromosome segregation during both meiotic divisions of Drosophila males , highlighting a crucial role of the protein not only in sister chromatid separation but also in bivalent segregation . Because the methanol-based fixation technique used to preserve spindle structure results in poorly defined chromosomes , we fixed larval testes in 45% acetic acid to maintain proper chromosome morphology ( see Materials and Methods ) . The chromosomes of all metaphase and early anaphase I figures of Top2suo1/Df and Top2suo1/Top2suo3 spermatocytes displayed very severe structural defects . In metaphase figures , bivalents appeared as chromatin masses in which individual sister chromatids were no longer discernible ( Figure 3C ) . Interestingly , approximately 60% of these aberrant metaphase figures ( n = 18 ) showed one or more long chromatin protrusions that were not observed in wild type spermatocytes ( Figure 3C ) . Similar protrusions have been observed in Top2-depleted meiotic cells of Drosophila females; these protrusions emanated from the chromatin mass of metaphase I figures and contained the centromeres at their tips , suggesting that they were generated by the pulling forces exerted by the spindle ( Hughes and Hawley; cosubmitted ) . In anaphase-like figures , individual chromosomes were also no longer recognizable , and the two presumptive daughter cells were almost invariably connected by a number of entangled and irregularly condensed chromatin threads ( Figure 3C ) . These entangled threads resulted in long ana-telophase chromatin bridges suggesting that the kinetochores of the entangled chromosomes are pulled away by the spindle microtubules ( Figure 3B and 3C ) . The different appearance of metaphase and anaphase chromosomes is a likely consequence of the spindle pulling forces , which would stretch and partially disentangle the metaphase chromosomes . However , in approximately 50% of the cases the chromosome tangles generated by loss of Top2 function were not resolved and gave rise to telophase figures in which all chromosomes remained in one of the daughter cells ( Figure 3A ) . To address the controversial role of Topo II in the maintenance of proper mitotic chromosome architecture , we first analyzed Top2suo1/Top2suo1 , Top2suo1/Df and Top2suo1/Top2suo3 brains fixed without hypotonic and colchicine pre-treatments . In contrast with the spermatocyte chromosomes , the chromosomes of Top2suo1/Top2suo1 mutant brains were morphologically normal . Chromosome morphology was substantially normal also in Top2suo1/Df and Top2suo1/Top2suo3 mutant brains; we observed only a few cells ( ∼5% ) showing a mild undercondensation of the heterochromatic regions of the major autosomes but the appearance of euchromatic arms was always normal ( Figure 4 ) . To determine whether the chromosomal phenotypes observed in spermatocytes and brain cells were due to different levels of Top2 we prepared testis extracts from Top2suo1/Top2suo1 mutants . Western blotting analysis showed these extracts exhibit a ∼50 reduction in the Top2 abundance . This reduction is comparable to that observed for Top2suo1/Df brains ( ∼44%; Figure 1A ) in which the chromosome morphology is essentially normal . This suggests that spermatocyte and brain chromosomes require different levels of Top2 activity to achieve a regular structure , and that spermatocyte chromosomes are particularly sensitive to Top2 depletion . In Top2suo1/Top2suo1 mutant brains fixed without colchicine and hypotonic pre-treatments , the mitotic index ( MI , the proportion of cells engaged in mitosis; see Materials and Methods ) and the frequency of anaphases were comparable to those of wild type brains ( Table 1 ) . Top2suo1/Df brains also showed a normal MI but displayed a significant increase in the frequency of anaphases with respect to both control and Top2suo1/Top2suo1 brains . This finding is likely to reflect a delay in progression through anaphase of mutant cells . In addition , 37% of the anaphases observed in Top2suo1/Df brains displayed chromatin bridges and/or acentric fragments that failed to correctly segregate ( Figure 4 A ) . However , we never observed entire lagging chromosomes with unseparated sister chromatids . Thus , in Top2suo1/Df brains centromere resolution does not appear to be affected as occurs in Top2-depleted S2 cells [42] . Top2 mutant brains fixed without colchicine and hypotonic pre-treatments also displayed several chromosome aberrations ( CABs ) . To define the type and frequency of CABs elicited by the different Top2 mutant alleles we incubated dissected brains in saline with 10−5 M colchicine for 1 hour and treated them with hypotonic solution before fixation . Colchicine arrests mitotic cells in metaphase and hypotonic treatment results in chromosome spreading facilitating CAB scoring . In Top2suo2/Top2suo2 brains the CAB frequency was not significantly higher than that of wild type controls , which exhibit 0 . 008 CABs per cell ( Table 2 ) . Top2suo1/Top2suo1 mutant brains showed approximately 0 . 04 CABs/cell , while brains of both Top2suo1/Df and Top2suo1/Top2suo3 larvae displayed CAB frequencies ranging from 0 . 25 to 0 . 53 per cell ( Figure 4B; Table 2 ) . An analysis of the CABs observed in Top2suo1/Df and Top2suo1/Top2suo3 brain metaphases reveled a striking specificity . In mutant males , most CABs were isochromatid breaks that preferentially involved the entirely heterochromatic Y chromosome ( 48 . 3% of isochromatid breaks ) and the third chromosome heterochromatin ( 46 . 1% ) . In addition , mutant males displayed 18–19% chromosome exchanges , most of which ( ∼98% ) were dicentrics or translocations generated by breaks in the Y chromosome and 3L heterochromatin ( Figure 4B ) . A similar CAB pattern was observed in XXY females in which 76% of the aberrations involved the Y chromosome , the third chromosome heterochromatin or both ( Table 2 ) . Examination of selected rearrangements displaying particularly clear heterochromatin banding revealed that the Y and third chromosome breakpoints are non-randomly distributed; the third chromosome breakpoints were clustered in region h47 of 3L heterochromatin , while the Y breakpoints were localized in several regions with a preference for regions h4-h5 and h19-h22 ( Figure 4B and C; see ref [60] for a cytological map of Drosophila heterochromatin ) . In Top2suo1/Df and Top2suo1/Top2suo3 XX females most of the CABs were isochromatid breaks in region h47 of 3L heterochromatin . In contrast with males , these females did not display chromosome exchanges ( Table 2 ) . However , should these exchanges occur between the third chromosomes and involve the heterochromatic regions , they would be hard to detect as they would appear as intact chromosomes 3 . Importantly , in 65% of the mutant metaphases with a broken Y chromosome the acentric fragment was either missing or present as an additional element to a normal chromosome complement ( the Y is easily recognized for its brightly fluorescent bands; see Figure 4 ) . Similarly , 35% of the cells with a broken third chromosome showed only the centric portion of the chromosome . We also observed many cells ( 20% ) with a normal chromosome complement and an extra acentric fragment of the size and the appearance of an entire 3L euchromatic arm . We thus assume that this acentric fragment is the exact complement of the centric element broken within 3L heterochromatin . Cells carrying either an isocromatid break lacking the acentric fragment or a normal chromosome complement plus an extra acentric fragment are likely to be generated during the anaphase of the previous cell cycle by the breakage of bridges between entangled chromatids . The bridges observed in Top2 mutant anaphases are likely to involve sister rather than nonsister chromatids because the analysis of 100 anaphase figures revealed that the bridges , or the acentric fragments generated by their rupture , never involve a banded Y chromatid and a uniformly stained autosomal or XL chromatid ( Figure 4A and S1 ) . Thus , although we cannot formally exclude that some of the evenly stained bridges/fragments are generated by entanglements between nonsister chromatids , we assume most bridges observed in Top2 mutant anaphases are due to aberrant associations between the sister chromatids of either the Y or the third chromosome ( Figures 4 and 5 ) . As shown in Figure 4B ( panels 1–6 ) and depicted in Figure 5 , the acentric fragment resulting from the rupture of a bridge is subject to three different fates: it could be lost , co-segregate with its intact sister chromatid , or co-segregate with its complementary centric fragment . The hypothesis that incomplete isochromatid breaks are generated during the anaphase of the previous cell cycle is supported by the analysis of 103 chromosome exchanges ( selected on the basis of their cytological quality ) involving the Y and 3L heterochromatin . Indeed , as shown in Figures 4 and 5 , the majority ( 82% ) of these rearrangements involve at least one element that can only be generated by the rupture of an anaphase bridge: either a centric fragment devoid of its acentric companion ( panels 1 and 4 of Figures 4B and 5 ) or an acentric fragment accompanied by its intact homologue ( panels 2 and 5 of Figures 4B and 5 ) . Interactions between these elements and between them and the two complementary fragments of a broken chromosome ( which are also likely to be generated during anaphase; see panels 3 and 6 of Figures 4B and 5 ) would result in the 9 types of chromosome rearrangements shown in figure 4B and depicted in Figure 5 . In Figure 5 , we depicted only the outcomes of single breaks in the Y and/or in one of the two third chromosomes . We did not consider the rearrangements generated by breaks in both chromosomes 3 or by two breaks in the Y chromosome . However , these complex CABs are rather rare and represent only 3% of the rearrangements involving the Y and/or the third chromosome; some examples of these complex rearrangements are shown in figure S2 . An analysis of the frequencies of the various types of rearrangements permits us to pinpoint the anaphase events that led to their formation . As shown in Figure 5 , the rearrangements that include centric Y and third chromosome elements with no complementary acentric fragments are 41 and 52 , respectively; those that contain intact Ys or third chromosomes plus extra fragments are 30 and 20 , respectively; and those including centric Y and third chromosome elements plus the respective complementary fragments are 32 and 31 , respectively . Given that the cells contain two chromosomes 3 and a single Y , this analysis establishes that the Y is more frequently involved in anaphase breaks than the third chromosome , consistent with the fact that the Y is breakable in multiple sites and the third chromosome in a single site . In addition , it appears that the most frequent anaphase event that will subsequently generate chromosome exchanges is the transmission to one of the two daughter cells of a centric element not accompanied by its complementary fragment . These findings are consistent with the observed frequencies of isochromatid breaks ( Table 2 ) . However , the relative frequencies of complete isochromatid breaks ( centric element plus complementary fragment; Table 2 ) are slightly higher than those derived from the analysis of the exchanges . The simplest explanation for this discrepancy is that the anaphase events leading to the chromosomal configurations depicted in Figure 4 are not completely independent , so that the frequencies of double events is not the one expected based on independency of single events . In summary , our findings strongly suggest that most CABs observed in Top2suo1/Df and Top2suo1/Top2suo3 mutants derive from site-specific breaks generated during the anaphase of the previous cell cycle . While Top2suo1/Df and Top2suo1/Top2suo3 mutant cells appear to progress almost normally through mitosis despite the presence of CABs , Top2suo3/Df brains displayed a dramatic drop in the MI . In brains stained only with DAPI we were unable to observe clear mitotic figures . We thus analyzed Top2suo3/Df brains stained with DAPI and immunostained for both tubulin and the mitotic H3 phospho-histone , which marks mitotic chromatin [61] . In 25 brains examined , we observed only 10 metaphase-like figures ( a single wild type brain usually contains 50–100 mitotic figures ) . The chromosomes of these rare metaphases appeared as chromatin masses in which individual chromosomes and sister chromatids were no longer recognizable ( Figure 6 ) . However , despite these severe chromosomal defects , all metaphases showed bipolar spindles with relatively normal microtubule densities . Growing evidence indicates that Drosophila somatic cells require kinetochore-driven MT growth for correct bipolar spindle formation [62] , [63] . Thus , the observation that Top2suo3/Df cells can assemble a normal bipolar spindle suggests that Top2-depleted chromosomes/kinetochores retain the ability to drive MT nucleation . The extremely low MI observed in Top2suo3/Df brain preparations might be the consequence of a checkpoint that prevents cells to progress through the cell cycle and enter mitosis . To ask whether the interphase arrest of Top2-deficient cells is caused by checkpoint activation we examined mei-41; Top2suo3/Df and Top2suo3/Df; tefu/tefu double mutants . mei-41 and tefu are the Drosophila orthologues of ATR and ATM , respectively; the kinases encoded by these genes are involved in the signaling pathway of the S and G2-M DNA damage checkpoints [64] , [65] , and ATR is thought to mediate the DNA decatenation checkpoint [35] . We found that neither double mutant displays a MI higher than that seen in Top2suo3/Df mutants ( in all cases we examined more than 20 brains and observed less than one mitotic cell per brain ) . Thus , neither ATR nor ATM loss was able to rescue the block in mitotic progression in Top2suo3/Df brains . To obtain further insight into the type of DNA lesions caused by Top2 deficiency we examined brains from mei-41; Top2suo1/Top2suo3 and Top2suo1/Top2suo3; tefu/tefu larvae . mei-41; Top2suo1/Top2suo3 brains showed a substantial increase in the frequency of CABs compared to either single mutant ( Table 2 ) . In doubly mutant males , the relative frequencies ( with respect to the total number of CABs ) of isochromatid deletions involving the Y chromosome or the third chromosome heterochromatin and the relative frequencies of Y-3 exchanges were comparable to those found in Top2suo1/Top2suo3 mutants . This suggests that downregulation of mei-41 ( ATR ) function might either increase the frequency of anaphase entanglements generated by Top2 deficiency or inhibit repair of the site-specific chromosome breaks they produce . Previous studies have shown that mutations in tefu ( ATM ) cause both telomeric fusions ( TFs ) and CABs [66] . Top2suo1/Top2suo3; tefu/tefu brains displayed frequencies of TFs and CABs that appear to be the sum of those observed in Top2suo1/Top2suo3 and tefu/tefu mutants . Thus , we conclude that tefu ( ATM ) and Top2 function in different pathways that mediate maintenance of chromosome integrity . The Top2 mutants analyzed here showed different mitotic phenotypes . Top2suo1/Df and Top2suo1/Top2suo3 brain cells showed a normal MI and frequent CABs that preferentially involve the Y and the 3L heterochromatin . In contrast , Top2suo3/Df brains showed a drastically reduced MI and rare metaphases with collapsed and/or shattered chromosomes . Top2suo1/Df and Top2suo1/Top2suo3 brains contain ∼60% less Top2 than wild type brains , and Top2 is undetectable in Top2suo3/Df brains ( Figure 1A ) . This suggests that larval brains with intermediate Top2 levels would produce intermediate phenotypes that would provide additional information on Top2 function . Thus , we analyzed the phenotype produced by RNAi in flies bearing an inducible UAS-Top2 RNAi construct . Animals bearing this construct and an Actin-GAL4 driver died as third instar larvae; the brains of these larvae displayed a Top2 content that was barely detectable in Western blots but definitely higher than that of Top2suo3/Df larvae ( Figure S3 ) . Consistent with this finding , the mitotic phenotype of Top2 RNAi brains was milder than that observed in Top2suo3/Df brains . The frequency of mitotic figures in brains stained for DNA , mitotic phosphohistone and tubulin was still very low ( an average of 7 mitotic figures per brain; 23 brains examined ) but basic chromosome morphology was maintained . The analysis of chromosome preparations from colchicine-treated and acetic acid-fixed brains revealed a phenotype that was not previously observed in Top2suo1/Df or Top2suo3/Df mutants . Most metaphases were either polyploid ( 31% ) or hyperploid ( 27% ) and showed extensive chromosome breakage ( Figure 7A and B ) . These metaphases showed many free autosomal arms generated either by breakage or drastic undercondensation of centric heterochromatin ( Figure 7A ) . We estimated that the frequency of these free arms was approximately three-fold higher than the frequency of breaks in euchromatin . This finding , together with the fact that the Y chromosome was invariably broken in multiple fragments or rearranged , suggests that in Top2 RNAi cells most CABs involve heterochromatin . The presence of hyperploid and polyploid metaphases in Top2-depleted brains suggests that these aberrant cells experienced nondisjunction events or complete cell division failures in the previous cell cycles . We also observed several prometaphase figures in which the homologous chromosomes were as tightly paired as in control cells ( Figure 7A ) , suggesting that Top2 activity is not essential for somatic pairing in larval brain mitotic cells . Examination of salivary glands revealed a role of Top2 in the control of polytene chromosome structure . In polytene preparations from Top2suo1/Df , Top2suo1/Top2suo3 and Top2suo3/Df males we consistently observed a specific alteration in the morphology of the X chromosome , which appeared bloated and often detached from the chromocenter ( Figure 8 and Figure S4 ) . In Top2suo1/Df and Top2suo1/Top2suo3 males , this deformed X chromosome conserved a banded appearance but its bands were less sharp than those of either the autosomes of the same nucleus or an X chromosome from wild type males . In general , the X chromosome condensation phenotype observed in Top2suo1/Df and Top2suo1/Top2suo3 mutants was rather variable but clearly visible in all polytene nuclei . A more severe defect was consistently observed in Top2suo3/Df and Top2 RNAi males , which showed a specific bloating and shortening of the X chromosome accompanied by a partial or total loss of its typical banding pattern ( Figure 8 and Figure S4 ) . In contrast , Top2suo1/Df , Top2suo1/Top2suo3 and Top2suo3/Df females displayed morphologically normal polytene chromosomes . These results indicate that Top2 is required for proper chromatin organization of the dosage-compensated X chromosome of males . We note that the defect in X chromosome organization and condensation observed in Top2suo3/Df and Top2 RNAi males is quite different from that caused by Cap-H2 ( condensin ) overexpression; an excess of Cap-H2 caused a tremendous axial compaction of all arms of polytene chromosomes [67] , which was never observed in the X chromosome of Top2 mutants . Defects in the X chromosome condensation were previously observed in male polytene nuclei of Imitation switch ( ISWI ) , Su ( var ) 2-5 ( HP1 ) and Su ( var ) 3-7 mutants [68]–[70] . HP1 and Su ( var ) 3-7 are interacting proteins with multiple roles in Drosophila chromatin regulation [71] , [72]; ISWI is the ATPase subunit of several chromatin remodeling complexes including CHRAC , ACF and NURF [73] . Dosage compensation and the architecture of compensated chromatin depend on the Male Specific Lethal ( MSL ) complex , which includes Msl1 , Msl2 , Msl3 , Mle , the Mof acetyltransferase and the roX1 and roX2 noncoding RNAs ( see refs [74] and [75] for review ) . Previous studies have shown that blocking H4k16 acetylation completely rescues the X chromosome condensation defects caused by ISWI , Su ( var ) 2-5 and Su ( var ) 3-7 mutations [70] , [76] , suggesting that histone acetylation modulates chromatin compaction of the male X chromosome . Based on these results , we investigated the relationships between Top2 and the dosage compensation system . Previous work has shown that the X chromosome of Top2 mutants binds the Msl1 and Mle proteins [45] . Our immunostaining experiments showed that the bloated X chromosome of Top2 mutants also binds the dosage compensation factor Mof ( Figure 9 ) . In addition , we found that the bloated appearance of the X chromosome of Top2 mutants is completely rescued in mof; Top2 double mutants ( Figure S5 ) . Although mutations in ISWI and Top2 both affect X chromosome structure in male polytene nuclei , they do not result in identical phenotypes . A direct comparison between the polytene chromosomes of the mutants revealed that the X chromosome of ISWI mutants is shorter and more compact than those of either wild type or Top2 mutant larvae ( Figure 8 ) . Because previous studies have suggested that Top2 and ISWI are both contained in the CHRAC remodeling complex [77] , we asked whether the male X chromosome deformation caused by Top2 mutations is a consequence of a failure to recruit ISWI at polytene chromosomes . Immunostaining of salivary glands from Top2suo1/Top2suo3 males showed that an anti-ISWI antibody decorates the X chromosome , although with a diffuse pattern reflecting the altered architecture of the chromosome ( Figure 9 ) . However , immunostaining of salivary glands from mof; Top2suo1/Top2suo3 doubly mutant males showed that the X with rescued morphology recruits a normal amount of ISWI despite the absence of Top2 ( Figure S5 ) . Thus , Top2 requirement for compensated chromatin modeling appears to be independent of ISWI localization .
We have shown that mutations in Top2 affect chromatin organization within the primary spermatocyte nuclei . During prophase I , wild type spermatocytes exhibit 3 main distinct chromatin clusters , which correspond to the major Drosophila bivalents ( X-Y , 2-2 and 3-3 ) ; the small fourth chromosome bivalent is either separated from these chromatin masses or associated with the X-Y bivalent [51]–[54] . We found that at early growth stages ( S1 and S2 ) the chromatin distribution within the spermatocyte nuclei of Top2suo1/Df males was not substantially different from wild type , suggesting that spermatogonial divisions are not severely affected . However , at stages S4 and S5 mutant spermatocytes displayed approximately twice as many chromatin masses as their wild type counterparts . In addition , in most mutant nuclei masses of similar size were closely apposed , suggesting a separation of the homologs within each chromatin territory . In the subsequent stages of spermatocyte growth , the number of chromatin masses in mutant nuclei progressively decreased , so that at prometaphase they displayed 3 compact chromatin clumps like their wild type counterparts . Chromosome behavior during spermatocyte growth and male meiosis has been investigated in previous studies , which revealed a complex pairing mechanism [53]–[58] , [78] . The X and the Y pair through their rDNA regions , while no specific euchromatic or heterochromatic pairing sites have been identified for the major autosomes , which are thus likely to exploit a homology-based pairing mechanism [53] , [54] , [78] . Tagging of allelic chromosome sites using the GFP-Lac repressor/lacO system or fluorescent in situ hybridization ( FISH ) showed that the major autosomes are tightly paired during the S1 and S2 stages . However , pairing is suddenly lost at the S2/S3 transition; the chromosomes remain then unpaired throughout the rest of meiosis but are included in a common nuclear territory until they condense prior to meiotic division [53] , [54] , [78] . One open problem is the mechanism underlying homolog co-mingling within the territories . Such co-mingling is unlikely to be the result of a canonical meiotic pairing , as the homologs remain uncondensed throughout prophase . It has been thus postulated that during early prophase the homologs might be held together in a single territory by chromatin entanglements [53] , [54] . Our results are consistent with this idea and lead us to hypothesize that Topo II plays an active role in generating the entanglements that mediate homolog association . However , it is equally possible that Topo II is required for some kind of chromatin modifications that are important for homolog conjunction within the territories . The chromatin organization defects within the prophase nuclei of Top2suo1/Df spermatocytes are very different from those previously observed in Cap-H2 and Cap-D3 mutants . In these mutants the chromatin remains diffuse within the spermatocyte nuclei from stage S4 through S6 , indicating that the Cap-H2 and Cap-D3 condensin II subunits are required for the formation of the intranuclear territories that comprise the homologous chromosomes [58] . The chromatin organization defect in Top2suo1/Df spermatocytes it is also different from that caused by mutations in genes mediating achiasmate homolog pairing in Drosophila males ( teflon , MNM and SNM ) . Spermatocytes of mutants in these genes display diffuse and slightly expanded chromatin territories during stages S4–S6; at prometaphase they show up to eight distinct chromatin clumps corresponding to unpaired univalents [54]–[57] . Collectively , the available results suggest that condensins ( Cap-H2 and Cap-D3 ) , the proteins required for homolog conjunction ( teflon , MNM and SNM ) and Top2 play distinct roles in chromatin organization during spermatocyte growth . As previously suggested , condensins are essential for territory formation and appear to function in opposition to homolog conjunction [58] . Top2 , Teflon , MNM and SNM are all required for proper territory formation and homolog pairing . Top2 is primarily required for homolog conjunction and correct territory organization during stages S4–S6 of spermatocyte growth , whereas Teflon , MNM and SNM are primarily required for meiotic chromosome pairing during prometaphase and metaphase . However , Top2 might have a redundant role in metaphase chromosome pairing that would be masked by the activity of Teflon , MNM and SNM , which would be able to mediate homolog pairing even when Top2 is reduced . The finding that Top2 mediates an aspect of homologous chromosome pairing in males is intriguing , as this enzyme ensures proper biorientation of achiasmatic homologs in females ( Hughes and Hawley; cosubmitted ) . Thus , despite the profound differences between Drosophila male and female meiosis , both types of meiotic divisions share a common Top2-depedent mechanism to facilitate achiasmate chromosome pairing . Previous studies have shown that loss of Topo II function results in species-specific meiotic defects . In top2 mutant cells of S . cerevisiae , premeiotic DNA synthesis , recombination and chromosome condensation are not affected but cells arrest at metaphase I and do not undergo the first meiotic division . However , top2 rad50 double mutants , in which recombination and synaptonemal complex formation are suppressed , perform the first meiotic division but not the second [79] , [80] . A similar meiotic phenotype has been observed in Top2 mutant cells of S . pombe , which exhibit only a mild defect in the final steps of meiotic chromosome condensation and arrest at metaphase . This arrest is relieved by mutations in rec7 that strongly reduce recombination [81] . Thus , in both budding and fission yeast , Topo II has little or no role in chromosome condensation but its activity is required for segregation of recombinant chromosomes at meiosis I and , at least in S . cerevisiae , for sister chromatid separation at meiosis II . Studies on meiotic cells from mouse and Chinese hamster injected with Topo II inhibitors did not reveal gross defects in chromosome condensation at the doses used in the experiments . However , the inhibitors induced a substantial meiotic delay and resulted in anaphase bridges and lagging chromosomes at both the first and the second meiotic anaphase [82]–[85] . It has been also suggested that the defect in homolog separation at meiosis I was due to a primary defect in chiasmata resolution [82] . In contrast , studies on mouse pachytene spermatocyte cultured in vitro showed that treatments with the Topo II inhibitors ICRF-193 and teniposide cause drastic defects in chromosome condensation . In teniposide-treated spermatocytes , both chromatin condensation and sister chromatid individualization were strongly affected . The effects of ICRF-193 were milder and some chromosomes managed to condense reaching a diplotene-like configuration [86] . We showed that weak Drosophila Top2 mutants ( Top2suo1/Top2suo1 and Top2suo2/Df ) with virtually no defects in brain cell mitoses exhibit strong defects in chromosome segregation during both meiotic divisions of males ( this report and ref [46] ) . In addition , we have shown that in Top2suo1/Df and Top2suo1/Top2suo3 testes all meiotic divisions exhibit severe defects in chromosome structure and segregation . In most cells , the chromosomes formed amorphous metaphase I masses where the sister chromatids were no longer discernible . In addition , these chromatin masses often emanated protrusions that are likely to correspond to stretched pericentric regions . Despite the strong defect in chromosome structure , Drosophila spermatocytes did not arrest at metaphase like yeast cells [79] or mouse spermatocytes treated with teniposide [86] . This finding is consistent with the limited delay caused by the spindle checkpoint in Drosophila male meiosis and by the inability of this checkpoint to prevent spermatid formation and differentiation [87] , [88] . Given that Drosophila male meiosis is achiasmatic , our observations strongly suggest that the aberrant meiotic chromosome segregation observed in Top2 mutant males is the consequence of a primary defect in chromatin folding within the chromosomes . As shown in Figure 3 , this defect is particularly evident when the chromosomes are pulled away by the meiotic spindle . In the anaphase-like figures of Top2 mutants , the chromatids are not individualized and the spindle poles are connected by an irregular network of chromatin threads . This suggests that the chromatin fibers of both the homologous and the heterologous chromosomes as well as those of the sister chromatids remain trapped by multiple entanglements , which prevent correct chromosome segregation during both meiotic divisions . The observation that Top2 is required for homolog conjunction during early meiotic prophase and then for correct chromosome segregation during anaphase is intriguing . We favor the hypothesis that Top2 has two independent activities , one required for catenation of the homologs within each chromosome territory and one required for proper chromatin folding within the metaphase chromosomes . However we cannot exclude the possibility that Top 2 activity during early prophase results in aberrant chromatin configurations that interfere with the chromatin folding processes leading to proper chromosome assembly . Frequent anaphase bridges have been also observed in spermatocytes from mutants in the Cap-H2 and Cap-D3 genes . However , the morphology of the meiotic chromosomes of these mutants is not severely disrupted as occurs in Top2 mutants . Judging from the published micrographs , the chromatids of Cap-H2 and Cap-D3 mutants are clearly individualized and their appearance is not very different from that of their wild type counterparts [58] . Our data do not provide an explanation of why the meiotic chromosomes of Drosophila males are much more sensitive to Top2 depletion than brain chromosomes . We can only envisage that the two types of chromosomes contain different proteins and/or different numbers of high affinity Top2 binding sites . If these binding sites were more frequent in meiotic than in mitotic chromosomes , then a ∼50% reduction of the Top2 protein would be sufficient to disrupt meiotic chromosome organization in males , but would have only a limited effect on mitotic chromosomes of larval brains . A meiotic phenotype reminiscent of that seen in Top2 mutant males has been observed after RNAi-mediated depletion of Top2 in females ( Hughes and Hawley; cosubmitted ) . In wild type female meiosis , the heterochromatic regions of the homologous chromosomes remain connected during prometaphase I by chromatin threads that ensure proper biorientation of achiasmatic homologues; these homologous connections are then resolved at later stages of meiosis allowing chromosome segregation . In Top2- depleted oocytes , heterochromatic regions of chromosomes usually fail to separate during prometaphase and metaphase I , and are often stretched into long protrusions with centromeres at their tips ( Hughes and Hawley and references therein ) . These findings indicate that Top2 is required for resolution of the DNA entanglements that normally connect homologous heterochromatic regions during female meiosis and suggest that the pulling forces exerted by the spindle generate chromatin protrusions . However , the meiotic phenotypes elicited by Top2 depletion in males and females are similar but not identical . While in female meiosis only the heterochromatic regions appear to be affected , in male meiosis both euchromatin and heterochromatin are affected . A higher sensitivity of heterochromatin to Top2 depletion is consistent with our observations on Top2suo1/Df and Top2suo1/Top2suo3 mitotic cells , which exhibit chromosome breaks that preferentially involve the 3L and the Y heterochromatin . We have shown that relatively weak mutant combinations of Top2 alleles ( Top2suo1/Df and Top2suo1/Top2suo3 ) only exhibit chromosome aberrations ( CABs ) , most of which involve specific regions of the Y and third chromosome heterochromatin . Severe RNAi-mediated Top2 depletion results in extensive chromosome breakage involving all chromosome regions with a preference for heterochromatin . Previous studies with pharmacological inhibitors of Topo II have also shown that treatments with these drugs cause CABs , but it is currently unclear to which extent these drugs directly induce DNA lesions , cause DNA damage via Topo II inhibition , or affect DNA stability through other off target effects [89] . However , CABs have been also observed in Top2 mutants of budding and fission yeasts [18] , [20] , [21] , [90] and in vertebrate cells depleted of Topo II by RNAi [17] , [24] , [25] , [32] , [91] . In both yeast and vertebrate systems , most CABs induced by Topo II deficiency are thought do be produced by breakage of the anaphase chromatin bridges generated by failure to decatenate sister chromatids [17] , [25] , [32] , [90] , [91] . We have shown that a relatively modest reduction of the Top2 level results in many isochromatid breaks and chromosome exchanges ( translocations and dicentric chromosomes ) that primarily involve 4 regions of the entirely heterochromatic Y chromosome ( regions h1-2; h4-6 , h19-21 and h24-25 ) and a specific region of the 3L heterochromatin ( region h47 ) . To the best of our knowledge , previous studies did not detect site-specific chromosome aberrations after inhibition of Topo II function . What is then the mechanism underlying the chromosome damage specificity in weak Top2 Drosophila mutants ? Two observations help answering this question . First , in mutant brain cells not treated with colchicine , 37% of the anaphases displayed chromatin bridges or lagging chromosome fragments generated by severing of the bridges . Second , most CABs observed in colchicine-treated cells were “incomplete” chromosome type aberrations ( i . e involving both sister chromatids ) . Namely , they consisted in broken centric chromosomes not accompanied by a corresponding acentric fragment , in normal chromosome complements with an additional acentric fragment , in Y-3 translocations lacking the reciprocal element , or Y-3 dicentric chromosomes lacking the acentric fragment . As illustrated in Figure 5 , these aberrations are likely to be the consequence of chromosome breaks generated during the anaphase of the previous cell cycle . We propose that these breaks preferentially occur in chromosomal sites whose stability is particularly dependent on Top2-mediated DNA decatenation . In Top2 deficient cells , these sites would not be properly untangled and would break when the sister chromatids are pulled apart by the mitotic spindle . It has been shown that a prominent Top2 cleavage target is the 359 bp Drosophila satellite DNA , which is mainly found in the centric heterochromatin of the X chromosome [92] . We examined the extant maps of satellite DNA and transposable element distribution along Drosophila heterochromatin [93]–[95] but did not find any sequence that uniquely maps to the Top2-sensitive regions . Thus these regions might share an as yet unidentified DNA or might correspond to junctions between different DNA sequences ( e . g . satellite-satellite; satellite transposon , or transposon-transposon ) . Top2 RNAi brain cells contain very small amounts of Top2 and exhibit only a few divisions , most of which are hyperploid or polyploid . The few scorable diploid figures almost invariably displayed incomplete aberrations involving the Y or third chromosome heterochromatin , often accompanied by breaks in other chromosomes . We could not assess the presence of incomplete aberrations in polyploid metaphases , most of which displayed many apparent breaks of the centric heterochromatin of the major autosomes . These discontinuities in chromosome structure could be either due to drastic failures of heterochromatin condensation or to real breaks generated by the rupture of chromatin bridges during anaphase . Our results do not permit us to discriminate between these possibilities , but we favor the first . Our observations on different Top2 mutant combinations and Top2 RNAi cells revealed different and apparently contradictory effects on cell cycle progression . In Top2suo1/Df brains that exhibit a ∼60% reduction in the wild type Top2 level , the MI was comparable to that of wild type controls , but mutant brains displayed an increase in the frequency of anaphases . These data are consistent with previous studies on Drosophila S2 cells showing that RNAi-mediated depletion of Top2 does not affect the MI and causes only a small increase in the anaphase frequency [40] , [42] . The MI was not substantially affected also in DT40 and human cells depleted of both Topo II alpha and Topo II beta [24] , [25] , [32] . In contrast , in Top2 RNAi brains and Topsuo3/Df brains the MI was reduced by one and two orders of magnitude , respectively . Top2 RNAi brains also displayed many aneuploid and polyploid cells . Polyploidy has been also observed in chicken and human cells lacking Topo II activity , and was attributed either to defects in cytokinesis or to a reentry into interphase following a mitotic arrest ( restitution ) [24] , [25] . The low MI and the extensive chromosome damage in RNAi brains did not allow us to reliably pinpoint the mechanism of polyploid cells formation . Polyploidy in Drosophila brains can be generated by either restitution or cytokinesis failure ( see for example refs [47] and [63] ) . It is thus possible that the polyploid cells of Top2 RNAi brains were generated through both mechanisms . The observations on weak Top2 mutants ( Top2suo1/Df and Top2suo1/Top2suo3; this study ) and Top2 RNAi S2 cells [40] , [42] strongly suggest that Drosophila does not have a decatenation checkpoint that arrests cell cycle in response to loss of Top2 function ( see Introduction ) . This conclusion agrees with recent data indicating that Topo II depletion and the resulting excess of DNA catenation does not trigger a G2 arrest in vertebrate cells [24] , [25] , [38] , [96] . However , a catenation-independent but Topo II-dependent checkpoint is activated by interruptions of the decatenation process caused by catalytically inactive forms of Topo II [22] , [37] , [96] . We found that in Top2suo3/Df and Top2 RNAi brains the MI is drastically reduced , indicating that cells are blocked in interphase . We also showed that this block is not relieved by mutations in either mei-41 ( ATR ) or tefu ( ATM ) . Because these kinases are involved in the signaling pathways that mediate most cell cycle checkpoints [64] , [65] , and because ATR has been previously implicated in the decatenation checkpoint [35] , we believe that the interphase block observed in the nearly complete absence of Top2 is not due to the activation of a checkpoint . We instead believe that this block could depend on the failure to remove supercoils during DNA replication , which would cause extensive DNA damage and make the cell unable to sustain cell cycle progression . We first reported that mutations in Top2 cause a specific alteration of the X chromosome morphology in male polytene chromosomes ( Bonaccorsi et al . , 50th Drosophila Research Conference; abstract 350b , 2009 ) . This observation was confirmed and extended by Hohl and coworkers [45] , who showed that in polytene nuclei of Top2 mutants the X chromosome retains the ability to recruit the MSL dosage compensation complex . In agreement with this study , we found that the bloated X chromosomes of Top2 mutants are decorated by anti-Mle , anti-Msl3 and anti-Mof antibodies . However , we have been unable to assess whether the staining intensity is the same as that of a normally condensed wild type X . Thus , it is quite possible that a reduction in Top2 expression partially affects the association of the MSL complex with the male X chromosome as recently suggested [97] . Regardless of the role of Top2 in recruitment and/or stabilization of the MSL complex , the observation that loss or inhibition of Top2 activity specifically disrupts the X chromosome morphology in males is fully consistent with the ChIP/Mass Spec experiments indicating that Top2 is the major MSL interactor [98] and with the co-IP assays showing that Top2 interacts with MSL through its Mle component [97] . Previous studies have shown that the X chromosome of polytene nuclei from Top2 mutants is decorated by antibodies against histone H4 acetylated at lysine 16 ( H4K16ac ) . This post-translational modification is mediated by the Mof histone acetyltransferase , whose association with the X chromosome depends on Mle; a mutation in mle or blocking H4k16 acetylation rescues the X chromosome condensation defects caused by mutations in ISWI [76] . We found that the X chromosome phenotype elicited by mutations in Top2 is rescued in mof; Top2 double mutants , and that both the bloated X of Top2 mutants and the reconstituted X of mof; Top2 double mutants normally recruit the ISWI protein . These results suggest that loss of Top2 does not affect condensation of the dosage compensated chromatin by inhibiting ISWI recruitment . However , in the absence of Mof-mediated H4k16 acetylation the chromatin compaction functions of Top2 and ISWI are both dispensable . The genetic interaction between Top2 and mof is consistent with the previously shown physical and functional interactions between topoisomerase II and histone deacetylases ( HDACs ) [2] , and with the synergistic cytotoxic effects caused by simultaneous inhibition of HDAC and Topo II [99] , [100] . We have shown that meiotic chromosomes are extremely sensitive to Top2 depletion and exhibit drastic defects in chromosome morphology even in weak Top2 mutants . Top2 downregulation in brain mitotic cells by either mutations or in vivo RNAi produced different cytological phenotypes . Moderate Top2 depletion ( Top2suo1/Df ) did not affect chromosome structure , and produced site-specific chromosome aberrations generated by the rupture of anaphase bridges . Severe Top2 depletion ( Top2 RNAi ) strongly reduced the MI , and induced heterochromatin undercondensation , extensive chromosome breakage , aneuploidy and polyploidy . Finally , complete ( or nearly complete ) Top2 deficiency ( Top2suo3/Df ) caused an interphase block and disrupted chromatid individualization in the rare diving cells . These phenotypes indicate that Drosophila chromosomes are exquisitely sensitive to the residual level of Top2 in the cell . In addition , they recapitulate most , if not all , phenotypes previously observed in vertebrate cells exposed to Topo II inhibitors or RNAi against Topo II ( see above ) . Thus , our results suggest that the previously observed discrepancies in vertebrate chromosome phenotypes elicited by Topo II downregulation might depend on the type of chromosomes examined ( e . g . mitotic vs meiotic ) , slight differences in Topo II activity , or both .
Top2suo1 and Top2suo2 mutant alleles were previously isolated by a cytological screen of a collection of male sterile mutants induced by EMS in C . Zuker laboratory [46] , [101] . Top2suo3 was isolated from a collection of about 1 , 500 lines carrying lethal mutations on chromosome 2 , arisen in the Zucker collection of viable mutants [101] . All mutations were kept in stock over the second chromosome balancer CyOTbA , bearing the Tubby1 ( Tb1 ) dominant transgene [102] . Df ( 2L ) Exel9043 was obtained from the Bloomington Drosophila Stock Center . mei-4129D [103] was kept in stock over an FM7-GFP balancer . The mei-4129D; Top2suo1/Top2suo3 and mei-4129D; Top2suo3/Df ( 2L ) Exel9043 double mutants were obtained by crossing mei-4129D/FM7-GFP; Top2suo3/CyOTbA females to FM7-GFP/Y; Top2suo1/CyOTbA and to FM7-GFP/Y; Df ( 2L ) Exel9043/CyOTbA males , respectively . Male larvae carrying both mutations were identified based on their non-GFP non-Tb phenotype . tefu3 ( or atm3 , a gift of S . D . Campbell; ref [104] ) was kept in stock over the TM6C balancer carrying the Stubble ( Sb ) and Tb dominant markers . The Top2suo1/Top2suo3; tefu3/tefu3 double mutant was obtained by crossing Top2suo1/CyOGFP; tefu3/TM6C females to Top2suo3/CyOGFP; tefu3/TM6C males . Doubly mutant larvae were identified on the basis of their non-GFP , non-Tb phenotype . The mof1 mutant stock was kindly provided by J . Lucchesi . To construct mof; Top2 double mutants we crossed mof1/FM7-GFP; Top2suo1/CyOTbA females to FMT-GFP/Y; Top2suo3/CyOTbA males; non-GFP and non-Tb male larvae were then selected for cytological examination of polytene chromosomes . Iswi mutant larvae were generated by crossing y w; Iswi2 sp; +/T ( 2;3 ) B3 CyO , TM6B females to Iswi1 Bc/SM5 , Cy sp males and recognized for the Bc non-Tb phenotype ( Iswi mutant stocks are a gift of D . Corona; see ref [76] ) . For in vivo RNAi-experiments , flies carrying a Top2 RNAi construct ( line 4570; VDRC collection ) were crossed to males carrying the Actin-Gal4 driver . The Oregon R laboratory strain was used as wild type control . All the stocks were maintained at 25°C on a standard medium . For markers , balancers and special chromosomes details see FlyBase ( http://www . flybase . org ) . Extract preparation and Western blotting were performed according to ref . [105] using a rabbit antiserum directed against aa 534–950 of Top2 ( gift of P . Fisher; see ref [48] ) diluted 1∶2 , 500 . The anti-alpha tubulin ( SIGMA ) and the anti-lamin ( Dm0 , Hybridoma Bank ) antibodies used for loading control were diluted 1∶20 , 000 and 1∶2 , 500 , respectively . Bands were detected with Chemi Doc XRS+ ( BIORAD laboratories ) , and densitometric analysis was performed using the Image Lab 4 . 0 . 1 software . To analyze the morphology and the integrity of metaphase chromosomes , brains from third-instar larvae were dissected in saline ( NaCl 0 . 7% ) . After incubation for 1 h with colchicine ( 10−5 M in saline ) , brains were treated for 8 min with hypotonic solution ( 0 . 5% Na Citrate ) , squashed in 45% acetic acid under a 20×20 mm coverslip , and immediately frozen in liquid nitrogen . To analyze anaphases and assess mitotic parameters , larval brains were disssected in saline , directly squashed without colchicine and hypothonic pretreatment and immediately frozen . The mitotic index ( MI ) was calculated by determining the average number of mitotic figures per optic field as described previously [47] . To visualize meiotic chromosome morphology ( Figure 3C ) , pupal testes isolated in saline were squashed in 45% acetic acid and frozen in liquid nitrogen . For polytene chromosome analysis , salivary glands were dissected in saline , squashed in 45% acetic acid and then frozen in liquid nitrogen . After removal of the coverslip , slides were air dried and mounted in Vectashield H-200 ( Vector Laboratories ) containing the DNA dye DAPI . To analyze nuclear organization of primary spermatocytes larval and pupal testes were fixed according to ref [52] . To analyze mitosis , brains from third instar larvae were dissected and fixed according to ref [106] . After several rinses in phosphate-buffered saline ( PBS ) slides were incubated overnight at 4°C with a monoclonal anti-alpha tubulin antibody ( Sigma Aldrich ) diluted 1∶1 , 000 in PBS and either a rabbit anti-DSpd-2 ( 1∶5 , 000; ref [59] ) or a rabbit anti-phosphorylated histone H3 ( 1∶1000 Millipore ) , also diluted in PBS . Primary antibodies were detected by 1 hour incubation at room temperature with FITC-conjugated anti-mouse IgG ( 1∶20; Jackson Laboratories ) and CY3-conjugated anti-rabbit IgG ( 1∶300; Invitrogen ) , both diluted in PBS . For anti-Top2 immunostaining , larval brains fixed according to ref [107] were incubated overnight with a rabbit anti-Top2 antibody ( gift of Donna J . Ardnt-Jovin; ref [49] ) diluted 1∶100 in PBS . After rinsing in PBS , slides were incubated for 1 hour at room temperature with a CY3-conjugated anti-rabbit IgG ( 1∶300 ) . For polytene chromosome immunostaining , salivary glands were dissected in saline , fixed as described in ref . [108] and incubated overnight with either of the following polyclonal antibodies diluted in PBS: rabbit anti-ISWI ( 1: 100; gift of D . Corona ) and rabbit anti-Mof , ( 1: 100; gift of J . Lucchesi ) . CY3-conjugated anti-rabbit IgG ( 1: 300; Invitrogen ) used as secondary antibodies . Immunostained preparations were mounted in Vectashield H-200 ( Vector Laboratories ) with DAPI . All cytological preparations were examined with a Zeiss Axioplan fluorescence microscope , equipped with a cooled charged–coupled device ( CCD camera; Photometrics CoolSnap HQ ) . Grayscale images were collected separately , pseudocolored and merged .
|
Type II topoisomerases ( Topo II ) are enzymes that disentangle DNA molecules during essential cellular processes such as DNA replication , chromosome condensation and mitotic cell division . Topo II is a major component of mitotic chromosomes and it is a well known target for cancer chemotherapy . Topo II inhibitors block the Topo II enzymatic activity leading to extensive DNA damage , which ultimately kills the cancer cell . Thus , investigating the role of Topo II in the assembly and structural maintenance of chromosomes is not only relevant to understand chromosome biology but might also have a translational impact on cancer therapy . Here we used Drosophila as model system to analyze the effect of Topo II depletion on chromosome stability . We show that the chromosomal phenotypes of mutant flies vary with the amount of residual Topo II , ranging from site-specific chromosome breaks , variations in chromosome number ( aneuploidy and poliploidy ) and dramatic defects in chromosome morphology . The chromosomal phenotypes observed in flies recapitulate all phenotypes seen in Topo II-depleted vertebrate chromosomes , reconciling the phenotypic discrepancies reported in previous studies . In addition , our finding that the Topo II dependent phenotypes vary with the residual amount of the enzyme provides useful information on the possible outcome of cancer therapy with Topo II inhibitors .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"cell",
"biology",
"biology",
"and",
"life",
"sciences"
] |
2014
|
The Analysis of Mutant Alleles of Different Strength Reveals Multiple Functions of Topoisomerase 2 in Regulation of Drosophila Chromosome Structure
|
In higher eukaryotes , replication program specification in different cell types remains to be fully understood . We show for seven human cell lines that about half of the genome is divided in domains that display a characteristic U-shaped replication timing profile with early initiation zones at borders and late replication at centers . Significant overlap is observed between U-domains of different cell lines and also with germline replication domains exhibiting a N-shaped nucleotide compositional skew . From the demonstration that the average fork polarity is directly reflected by both the compositional skew and the derivative of the replication timing profile , we argue that the fact that this derivative displays a N-shape in U-domains sustains the existence of large-scale gradients of replication fork polarity in somatic and germline cells . Analysis of chromatin interaction ( Hi-C ) and chromatin marker data reveals that U-domains correspond to high-order chromatin structural units . We discuss possible models for replication origin activation within U/N-domains . The compartmentalization of the genome into replication U/N-domains provides new insights on the organization of the replication program in the human genome .
Comprehensive knowledge of genetic inheritance at different development stages relies on elucidating the mechanisms that regulate the DNA spatio-temporal replication program and its possible conservation during evolution [1] . In multi-cellular organisms , there is no clear consensus sequence where initiation may occur [2] , [3] . Instead epigenetic mechanisms may take part in the spatial and temporal control of replication initiation in higher eukaryotes in relation with gene expression [4]–[9] . For many years , understanding the determinants that specify replication origins has been hampered by the small number ( approximately 30 ) of well-established replication origins in the human genome and more generally in mammalian genomes [1] , [7] , [10] . Recently , nascent DNA strands synthesized at origins were purified by various methods [11]–[14] to map a few hundreds putative origins in 1% of the human genome . For unclear reasons , the concordance between the different studies is very low ( from to ) [12]–[15] . In a completely different approach to map replication origins , previous in silico analyses of the nucleotide compositional skew of the human genome showed that the sign of abruptly changed from to when crossing known replication initiation sites . This allowed us to predict putative origins at more than a thousand sites of sign inversion ( -jumps ) along the human genome [16] , [17] . Further analyses of patterns identified 663 megabase-sized N-domains whose skew profile displays a N-like shape ( Fig . 1A ) , with two abrupt -jumps bordering a DNA segment whose skew linearly decreases between the two jumps [16]–[21] . Skew N-domains have a mean length of Mb and cover 29 . 2% of the human genome . The initiation zones predicted at N-domains borders would be specified by an open chromatin structure favorable to early replication initiation and permissive to transcription [21] , [22] . The determination of HeLa replication timing profile [23] and the analysis of available timing profiles in several human cell lines [24]–[26] allowed us to confirm that significant numbers of N-domains borders harbor early initiation zones active in germline as well as in somatic cell types [18] , [27] . Recent studies have shown that replication induces different mutation rates on the leading and lagging replicating strands [27] . This asymmetry of rates acting during evolution has generated the skew upward jumps that result from inversion of replication fork polarity at N-domain extremities . The skew profile along N-domains would result from superimposed effects of transcription and of replication [19] , [20] , [28]–[31] . Accordingly , the linear decrease of the skew ( Fig . 1A ) may reflect a decrease in the proportion of replicating forks propagating from the left ( 5′ ) to the right ( 3′ ) N-domain extremity . This organization of replication in a large proportion of the genome contrasts with the previously proposed segmentation of mammalian chromosomes in regions replicated either by multiple synchronous origins with equal proportion of forks coming from both directions ( 0 . 2–2 . 0 Mb Constant Timing Regions ) or by unidirectional replication forks ( 0 . 1–0 . 6 Mb Transition Timing Regions ) [25] , [32]–[34] . Here , to determine the existence of a new type of replication domains presenting gradients of replication fork polarity , we establish ( i ) that the replication fork polarity and the compositional skew are proportional to each other , ( ii ) that the replication fork polarity can be directly extracted from the derivative of the replication timing profile . Taking advantage of replication timing profiles in several human cell types [23] , , we show that the derivative of the replication timing profile of N-domains is shaped as a N . The corresponding U-shape of the replication timing profile is not specific to the germline but is generally observed in all replication timing profiles examined , thus establishing these “U-domains” as a new type of replication domains , consistent with the recent experimental observation of multiple replication initiations in most Transition Timing Regions in several human cell lines [35] . As observed with the early initiation zones bordering N-domain extremities , those specific to the U-domains are significantly enriched in open chromatin markers as well as insulator-binding proteins CTCF [36] , [37] and are prone to gene activity . Analysis of recent Hi-C data [38] reveals that U-domains correspond to self-interacting structural chromatin units . These data make a compelling case that the “islands” of open chromatin observed at U-domains borders are at the heart of a compartmentalization of chromosomes into chromatin units of independent replication and of coordinated gene transcription .
To establish the existence of replication domains associated with replication fork polarity gradients , we first demonstrate the relations between replication fork polarity , nucleotide compositional skew and derivative of the replication timing profile . Under appropriate hypotheses , the skew resulting from mutational asymmetries associated with replication is proportional to the fork polarity at position on the sequence ( Material and Methods ) : ( 1 ) where ( resp . ) is the proportion of forks replicating in the ( resp . ) direction on the Watson strand . The linear decrease of in N-domains from positive ( end ) to negative ( end ) values thus likely reflects a linear decrease of the replication fork polarity with a change of sign in the middle of the N-domains . This result strongly supports the interpretation of N-domains ( Fig . 1A–C ) as the signature of a higher-order organization of replication origins in germline cells . The replication fork polarity can also be directly deduced from replication timing data under the central hypotheses that the replication fork velocity is constant and that replication is bidirectional from each origin . Note that recent DNA combing experiments in HeLa cells have shown that replication fork velocity does not significantly vary during S phase which strongly supports the former hypothesis [35] . We demonstrate that the replication fork polarity is the product of the derivative of the mean replication timing ( MRT ) and the replication fork velocity ( Material and Methods ) : ( 2 ) The fork polarity should therefore provide a direct link between the skew and the derivative of the replication timing profile in germline cells . To test this relationship , we used a substitute for germline MRT , the replication timing profiles of seven somatic cell lines ( one embryonic stem cell , three lymphoblastoid , a fibroblast , an erythroid and HeLa cell lines ) ( Material and Methods ) . We first correlated the skew with , in the BG02 embryonic stem cells , over the 22 human autosomes ( Fig . 1D ) . The significant correlations observed in intergenic ( , ) , genic ( , ) and genic ( , ) regions are representative of the correlations observed in the other 6 cell lines ( Table 1 ) . These correlations are as strong as those obtained between the profiles in different cell lines ( Supplementary Table S1 ) , as well as those previously reported between the replication timing data themselves [26] , [34] , [39] . The correlations between and are even stronger when focusing on the 663 skew N-domains ( Table 1 ) . The correlations obtained in intergenic regions ( ) are recovered to a large extent in genic regions ( ) where the transcription-associated skew was hypothesized to superimpose to the replication-associated skew [18]–[20] . Further evidence of this link between and was obtained when averaging , for the different cell lines , the profiles inside the 663 skew N-domains after rescaling their length to unity ( Fig . 1E ) . These mean profiles are shaped as a N , suggesting that some properties of the germline replication program associated with the pattern of replication fork polarity are shared by somatic cells . According to Equations ( 1 ) and ( 2 ) , the integration of the skew is expected to generate a profile rather similar to the replication timing profile . In segments of linearly changing skew , the integrated function is thus expected to show a parabolic profile . The integrated function when estimated by the cumulative skew ( Fig . 1B ) along N-domains of a 11 . 4 Mb long fragment of human chromosome 10 , indeed displays a U-shaped ( parabolic ) profile likely corresponding the replication timing profile in the germline . Remarkably , the 6 N-domains effectively correspond to successive genome regions where the MRT in the BG02 embryonic stem cells is U-shaped ( Fig . 1C ) . The 7 putative initiation zones ( to ) corresponding to upward -jumps ( Fig . 1A ) , co-locate ( up to the kb resolution ) with MRT local extrema which supports that they are highly active in BG02 . These initiation zones can present cell specificity as exemplified by the putative replication origin which is inactive ( or late ) in both the K562 erythroid and GM06990 lymphoblastoid cell lines ( Fig . 1C ) resulting in domain “consolidation” [40] . Two neighboring U-domains ( and ) in BG02 merged into a larger U-domain in the K562 and GM06990 cell lines . Note that the other 3 N-domains ( , , and ) are replication timing U-domains common to BG02 , K562 and GM06990 . To detect U-domains in replication timing profiles at genome scale , we developed a wavelet-based method ( Material and Methods , and Supplementary Text S1 ) which allowed us to identify in the 7 human cell lines from 664 ( TL010 ) up to 1534 ( BG02 ) U-domains of mean size ranging from 0 . 966 Mb ( HeLa R2 ) up to 1 . 62 Mb ( TL010 ) and covering from 39 . 6% ( TL010 ) to 61 . 9% ( BG02 ) of the genome ( Table 2 ) . For each cell line , the average MRT profile of U-domains has an expected parabolic shape ( Fig . 2A ) representative of individual U-domains ( Fig . 2C and Supplementary Figs . S1A–S9A ) . Inside the U-domains , the derivative is N-shaped ( Fig . 2D and Supplementary Figs . S1B–S9B ) like the skew profile inside N-domains ( Supplementary Figs . S1F–S9F ) . When rescaling the size of each U-domains to unity for a given cell line , these profiles superimpose onto a common N-shaped curve well approximated by the average profile ( Fig . 2B ) . To determine the amounts of U-domains conserved in different cell types , we computed for each cell type pair the mutual covering of the corresponding sets of U-domains ( two U-domains are shared by two different cell lines if each domain covers more than of the other domain ( Table 3 ) ) . Taking as reference the matching obtained for the two BJ ( 68 . 6% and 74 . 3% ) and HeLa ( 51 . 8% and 54 . 6% ) cell replicates , the matchings between the other cell lines were statistically significant and comparable ( from 40% to 65% for the mutual covering of lymphoblastoid cell lines ) . The number of U-domain shared by cell type pairs were all significantly larger than the number expected by chance ( , Supplementary Table S2 ) . For example BG02 shares 197 and 189 U-domains with K562 and GM06990 respectively , when only and are expected by chance ( Supplementary Table S3 ) . This corresponds to a significant proportion ( ) of the U-domains of the individual cell lines ( Table 3 ) , as compared to the matchings ( ) expected by chance ( Supplementary Table S4 ) . A significant percentage of N-domains correspond to U-domains ( e . g . from 12 . 5% in BJ R1 up to 23 . 7% in BG02 ) . This explains that when representing the MRT profile of BG02 instead of the skew , along the set of N-domains ordered according to their size , we can recognize the edges of many N-domains ( Supplementary Figs . S1D–S9D ) . The same observation can be made when comparing the profiles ( Supplementary Figs . S1E–S9E ) to the corresponding skew profiles ( Supplementary Fig . S1F ) . Note that the N-domains match only of the U-domains of various cell lines due to the very stringent N-domain selection criteria [19] , [20] that yielded only 663 N-domains ( 29 . 2% of the genome ) as compared to much larger U-domain numbers ( of the genome; Table 2 ) . Replication timing U-domains are robustly observed in all cell lines , covering of the human genome . For each cell type , about half U-domains are shared by at least another cell line , namely BG02 ( 38 . 4% ) , K562 ( 61% ) , GM06990 ( 59 . 2% ) , BJ R1 ( 51 . 6% ) , HeLa R1 ( 44 . 7% ) . This is also true for the skew N-domains ( 50 . 2% ) that likely correspond to replication timing U-domains in the germline . However about half of the genome that is covered by U-domains corresponds to regions of high replication timing plasticity where replication domains may ( i ) reorganize according to the so-called “consolidation” scenario ( merging of two U-domains into a larger one ) ( Fig . 1C ) , ( ii ) experience some boundary shift and ( iii ) emerge in a late replicating region as previously observed in the mouse genome during differentiation [40] . Genome-wide investigation of chromatin architecture has revealed that , at large scales ( from 100 kb to 1 Mb ) , regions enriched in open chromatin fibers correlate with regions of high gene density [41] . Moreover there is a growing body of evidence that transcription factors are regulators of origin activation ( reviewed in Kohzaki and Murakami 2005 ) . We ask whether the remarkable genome organization observed around N-domain borders [19] is maintained around replication timing U-domain borders and to what extent it is mediated by a particular chromatin structure favorable to early replication origin specification [22] . When mapping DNase I sensitivity data ( Material and Methods ) [42] on the U-domains , we observed that the mean coverage is maximal at U-domain extremities and decreases significantly from the extremities to the center that is rather insensitive to DNase I cleavage ( Fig . 3A and Supplementary Fig . S10 ) . This decrease , from values significantly higher than the genome-wide average value , extends over 150 kb , whatever the size of the replication timing U-domain ( Supplementary Fig . S11A–C ) suggesting that , for all examined cell lines , early replicating U-domains borders are at the center of kb wide open chromatin regions . We observed a significant anti-correlation between DNase I cleavage sensitivity data and replication timing data in BG02 ( DNase H1-hESC: , ) , K562 ( , ) and GM06990 ( , ) cell lines as well as in the other four cell lines ( data not shown; note that this was still observed when controlling for the GC content ) . This is further supported by open over input chromatin ratio data obtained from human lymphoblastoid cells [41] . We observed that the regions presenting an open/input ratio also decreased significantly ( 3-fold ) from U-domain borders to centers ( Fig . 3B ) . Cytosine DNA methylation is a mediator of gene silencing in repressed heterochromatic regions , while in potentially active open chromatin regions , DNA is essentially unmethylated [43] . DNA methylation is continuously distributed over mammalian chromosomes with the notable exception of CpG islands ( CGIs ) and in turn of certain CpG rich promoters and transcription start sites ( TSSs ) . Along the observation that the hypomethylation level of CGIs extends to about 1 kb in flanking regions , we used 1 kb-enlarged CGI coverage as an hypomethylation marker ( Material and Methods ) [22] . When averaging over the U-domains detected in BG02 , we robustly observed a maximum of CGI coverage at U-domain borders as the signature of hypomethylation and a decrease over a characteristic distance of kb ( Fig . 3C ) , similar to what we found for DNase I sensitivity coverage ( Fig . 3A ) . This contrasts with the GC-content profile that strongly depends on the U-domain size and decreases very slowly toward the U-domain center without exhibiting any characteristic scale ( Supplementary Fig . S11D–F ) . These observations are consistent with the hypothesis that early replication origins at U-domain borders are associated with CGIs that are possibly protected from methylation by colocalization with replication origins [44] . Open chromatin markers have been associated with genes . For example 16% of all DNase I hypersensitive sites ( HS ) are in the first exon or at the TSS of a gene and 42% are found inside a gene [45] . Also , more than 90% of broadly expressed housekeeping genes have a CpG-rich promoter [46] . Remarkably , the mean profiles of Pol II binding Chip-Seq tag density ( Material and Methods ) along U-domains detected in BG02 , K562 and GM06990 cell lines strongly decay over kb away from U-domain borders ( Fig . 3D ) . This indicates that , whatever the cell line , the open chromatin regions around replication U-domains are prone to transcription whereas U-domain central regions appear , on average , transcriptionally silent . Importantly , we have reproduced the analyses of open chromatin markers near U-domain borders that do not match with a N-domain border ( at 100 kb resolution ) and confirmed that the results reported in Fig . 3 apply to the initiation zones at U-domains borders of every cell line ( Supplementary Fig . S12 ) . It is widely recognized that the 3D chromatin tertiary structure provides some understanding to the experimental observation of the so-called replicon and replication foci [2] , [47] . In particular , replicon size , which is dictated by the spacing between active origins , correlates with the length of chromatin loops [8] , [47] , [48] . The chromosome conformation capture technique [38] has provided access to long-range chromatin interactions as a footprint of the different levels of chromatin folding in relation with gene activity and the functional state of the cell . From a comparative analysis of replication timing data and Hi-C data correlation matrix in the human genome , some dichotomic picture has been proposed where early and late replicating loci occur in separated compartments of open and closed chromatin respectively [34] , [38] . Here , instead of considering the partitioning of the chromosomes derived from all intrachromosomal interactions of each locus ( using a principal component of the principal component analysis of the Hi-C data over each chromosome ) , we focused on interactions between loci separated by short genomic distances ( Mb ) over which the contact probabilities are the highest [38] . First , we performed this zoom in the Hi-C contact matrix in the K562 cell line at the 100 kb resolution ( Material and Methods ) for the 11 . 4 Mb fragment of human chromosome 10 which contains four U-domains in K562 ( Fig . 1; , , and ) . We found that these four U-domains remarkably correspond to four matrix square-blocks of enriched interactions ( Fig . 4A ) . Hence , we recover that early replicating zones that border a U-domain ( e . g . and separated by 3 . 9 Mb ) , have a high contact probability as the signature of 3D spatial proximity . However , we also observe a high contact probability of the two early replicating borders with the late replicating U-domain center and interactions appear sparse for loci in separate U-domains ( e . g . and separated by 3 . 6 Mb ) . Further examination of the average behavior of intrachromosomal contact probability as a function of genomic distance for the complete genome corroborates these observations . We found that the mean number of interactions between two 100 kb loci of the same U-domain decays when increasing their distance as observed genome-wide ( Fig . 4B ) . Importantly , the mean number of pairwise interactions is significantly higher inside the U-domains than genome-wide and this seems to depend on the U-domain length . In particular , we found that the smaller the domain , the higher the mean number of interactions which is probably a signature of a more open chromatin structure . When comparing the contact probability between two loci inside a U-domain or lying in neighboring U-domains ( Fig . 4C ) , we observed that the latter is higher than the former for distances smaller than the characteristic size ( kb ) of the open chromatin structure at U-domain borders ( Fig . 3 ) . Above this characteristic distance , the tendency is reversed and the ratio increases up to 2 for distances Mb ( Fig . 4C ) . These data suggest that the segmentation of the genome into replication timing U-domains corresponds to some spatial compartmentalization into self-interacting structural chromatin units insulated by two boundaries of open , accessible , actively transcribed chromatin . This conclusion is strengthened by the observation that U-domain borders are significantly enriched in the insulator binding protein CTCF ( Fig . 5 ) , that is known to be involved in chromatin loop formation conditioning communication between transcriptional regulatory elements [36] , [37] , [49] , [50] . Quantitatively similar results were obtained for the lymphoblastoid GM06990 cell line for which both replication timing and Hi-C data were available ( Supplementary Fig . S13 ) . The mapping of open chromatin marks along U-domains revealed that they are bordered by early replication initiation zones likely specified by a kb wide region of accessible , open chromatin permissive to transcription . Such a strong gradient of open chromatin environment was not observed around a large fraction of the 283 replication origins identified in ENCODE regions [12]; only 29% overlap a DNase I hypersensitivity site and half of them do not present open chromatin marks and are not associated with active transcription [22] . Furthermore , the typical inter-origin distance in human cells is 50–100 kb [12] , [4] , a much smaller value than the mean U-domain size ( 1–1 . 5 Mb ) . These data can be reconciled in a model [51] , [52] where replication origins fire independently and their properties ( intrinsic firing time probability , efficiency ) are specified by the chromatin state: efficient early replicating origins in euchromatic regions ( U-domains borders ) and late replicating or less efficient origins in heterochromatic regions ( U-domains centers ) . A more dynamical model can also be proposed in which replication first initiates at U-domain borders followed by a chromatin gradient-mediated succession of secondary origin activations . These origins may be remotely activated by the approach of a center-oriented fork that may stimulate initiation due to changes in DNA supercoiling in front of the fork or to association of chromatin remodelers or origin triggering factors with replication fork proteins [35] . This “domino” model could explain why replication progresses from U-domain borders much faster ( 3–5 times ) than the known speed of single fork [8] , [35] , [48] . Indeed the U-shape of the replication timing profile indicates that the replication wave accelerates ( effective velocity equals the inverse of the replication timing derivative , Equation ( 2 ) ) as the signature of an increasing origin firing frequency during the S-phase [53] . It will be essential to determine to what extent the chromatin state influences fork progression and origins activations and whether outside of U-domains , the genome replicates according to a similar or completely different scenario .
We use the formalism of Markov processes to prove that replication-associated asymmetries between the substitution rates of the two DNA strands induce , in the limit of small asymmetries , a nucleotide compositional skew proportional to the replication fork polarity ( the average direction of a locus' replication ) . Models of DNA composition evolution are usually written in the form of an autonomous and homogeneous system of first-order differential equations [54]: ( 3 ) where is the vector which represents the state of the system , i . e . for , is the frequency of at time , and for , is the substitution rate of . A general and well-known property of a Markov process like Equation ( 3 ) is that tends exponentially towards the equilibrium value , defined as . The evolution on the complementary strand is given by the same equation but for and , defines the frequency vector on the complementary strand , is the substitution rate matrix on the complementary strand , and denotes the complementary base of . Under no-strand-bias conditions [55] , the same substitution rates affect the two strands , i . e . leading to the so-called parity rule of type 2 ( PR2 ) : and [56]–[59] . Departure from this symmetry condition can thus be quantified by decomposing into symmetric and antisymmetric parts , the latter accounting for the establishment of a nucleotide compositional strand asymmetry during evolution . According to our previous studies of the skew in mammalian genomes [16]–[20] , [29]–[31] , we can reasonably suppose that replication and transcription are the main mechanisms responsible for deviations in PR2 . If we concentrate on the effect of replication on DNA composition , we may consider intergenic regions only: then the substitution rate matrix can be written as ( 4 ) where is a substitution rate matrix satisfying the no-strand bias conditions ( ) , is the substitution rate matrix associated with replication and ( resp . ) the proportion of forks replicating the region of interest in the ( resp . ) direction . can be easily decomposed into a symmetric part: ( 5 ) and an antisymmetric part: ( 6 ) which turns out to be proportional to the fork polarity: ( 7 ) Under the assumption that is significantly smaller than , namely ( 8 ) we can use perturbation theory to solve Equation ( 3 ) and to show that if the compositional skews: ( 9 ) are initially null ( ) , then the total skew will be proportional to the fork polarity at all times up to terms of order ( Equation ( 8 ) ) : ( 10 ) where is a function that depends only on and . Using the mean nucleotide substitution rate matrix computed in the intergenic regions on each side ( 300 kb windows ) of the -upward jumps [27] , the coefficients of were found to be much smaller than those of with ( Supplementary Text S1 ) . Thus , according to Equation ( 10 ) , the observed linear decrease of the skew in N-domains from positive ( end ) to negative ( end ) values likely reflects the progressive linear decrease of the replication fork polarity with a change of sign in the middle of the skew N-domains . These results provide strong support to the interpretation of skew N-domains ( Fig . 1A ) as independent replication units in germline cells . As previously pointed out in [52] , the derivative of the replication timing profile does not provide a direct estimator of the replication fork velocity as it also depends on the fork polarity . Here , we demonstrate that the replication fork polarity can be directly deduced from replication timing data under the central hypothesis that the replication fork speed is constant and that replication is bidirectional from each origin . For a given cell cycle , let be the number of activated origins , their positions along the genome and their initiation times . Then the configuration ( where and when the origins of replication fire during the S-phase ) completely specifies the spatio-temporal replication program ( Fig . 6 ) [51] , [52] . If we denote the event “the fork coming form meets the fork coming from ” whose space-time coordinates are: ( 11 ) then the replication timing and fork orientation at spatial position are given by ( Fig . 6 ) : ( 12 ) We clearly see that since then the fork orientation is equal to times the derivative of the replication timing: ( 13 ) Under the hypothesis of constant fork velocity , this relationship holds in whole generality in each cell cycle and at every locus without any specific asumption on the distribution of initiation events . By definition , the replication fork polarity is the population average over cell cycles of the fork orientation: . Hence , when averaging over cell cycles , Equation ( 13 ) yields: ( 14 ) where we have used the fact that the spatial derivative commutes with the population average and that by definition . The replication fork polarity therefore provides a direct link between the skew and the derivative of the MRT ( Equations ( 10 ) and ( 14 ) ) in germline cells . Sequence and annotation data were retrieved from the Genome Browsers of the University of California Santa Cruz ( UCSC ) [60] . Analyses were performed using the human genome assembly of March 2006 ( NCBI36 or hg18 ) . As human gene coordinates , we used the UCSC Known Genes table . When several genes presenting the same orientation overlapped , they were merged into one gene whose coordinates corresponded to the union of all the overlapping gene coordinates , resulting in 23818 distinct genes . We used CpG islands ( CGIs ) annotation provided in UCSC table “cpgIslandExt” . The coordinates of the 678 human replication N-domains for assembly NCBI35/hg17 were obtained from the authors [19] and mapped using LiftOver to hg18 coordinates; we kept only the 663 N-domains that had the same size after conversion . We determined the mean replication timing profiles along the complete human genome using Repli-Seq data [23] , [26] ( Supplementary Text S1 , and Supplementary Fig . S14 ) . For embryonic stem cell line ( BG02 ) , three lymphoblastoid cell lines ( GM06990 , H0287 , TL010 ) , a fibroblast cell line ( BJ , replicates R1 and R2 ) , and erythroid K562 cell line , Repli-Seq tags for 6 FACS fractions were downloaded from the NCBI SRA website ( Studies accession: SPR0013933 ) [26] . For the HeLa cell line we computed the mean replication timing ( MRT ) instead of computing the S50 ( median replication timing ) as in [23] . We developed a segmentation method of the MRT profile into U-domains based on the continuous wavelet transform . This method amounts to perform objective ( U- ) pattern recognition in 1D signals where the U-motif is picked out from the background signal variations ( Supplementary Text S1 , and Supplementary Fig . S15 ) . For the analysis of correlations , we reported the Pearson's product moment correlation coefficient and the associated P-value for no association ( ) . All statistical computations were performed using the R software ( http://www . r-project . org/ ) . We used the DNaseI sensitivity measured genome-wide [42] . Data corresponding to Release 3 ( Jan 2010 ) of the ENCODE UW DNaseI HS track , were downloaded from the UCSC FTP site: ftp://hgdownload . cse . ucsc . edu/goldenPath/hg18/encodeDCC/wgEncodeUwDnaseSeq/ . We plotted the coverage by DNase Hypersentive Sites ( DHSs ) identified as signal peaks at a false discovery rate threshold of 0 . 5% within hypersensitive zones delineated using the HotSpot algorithm ( “wgEncodeUwDnaseSeqPeaks” tables ) . When several replicates were available , data were merged . We used ChIP-seq data using antibody for Pol II and CTCF from Release 3 ( Mar 2010 ) of the ENCODE Open Chromatin track [11] , [61] . Data were downloaded from the UCSC FTP site: ftp://hgdownload . cse . ucsc . edu/goldenPath/hg18/encodeDCC/wgEncodeChromatinMap . We plotted coverage by regions of enriched signal in ChIP experiments , called based on signals created using F-Seq [62] ( “wgEncodeUtaChIPseqPeaks” tables ) . Significant regions were determined at an approximately 95% sensitivity level . We always used the most recent version of data . We used the spatial proximity maps of the human genome generated using Hi-C method [38] . We downloaded 100 kb resolution maps for GM06990 and K562 cell lines from the GEO web site ( GSE18199_binned_heatmaps ) : http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE18199 . Open over input chromatin ratio data from human lymphobastoid cells were obtained from the authors [41] . Coordinates of N-domains and U-domains in the investigated 7 cell lines can be downloaded from: http://perso . ens-lyon . fr/benjamin . audit/ReplicationDomainsPLoSComputBiol2012/ .
|
DNA replication in human cells requires the parallel progression along the genome of thousands of replication machineries . Comprehensive knowledge of genetic inheritance at different development stages relies on elucidating the mechanisms that regulate the location and progression of these machineries throughout the duration of the DNA synthetic phase of the cell cycle . Here , we determine in multiple human cell types the existence of a new type of megabase-sized replication domains across which the average orientation of the replication machinery changes in a linear manner . These domains are revealed in 7 somatic cell types by a U-shaped pattern in the replication timing profiles as well as by N-shaped patterns in the DNA compositional asymmetry profile reflecting the existence of a replication-associated mutational asymmetry in the germline . These domains therefore correspond to a robust mode of replication across cell types and during evolution . Using genome-wide data on the frequency of interaction of distant chromatin segments in two cell lines , we find that these U/N-replication domains remarkably correspond to self-interacting folding units of the chromatin fiber .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"physics",
"genomics",
"nucleic",
"acids",
"genetics",
"epigenetics",
"biology",
"computational",
"biology",
"dna",
"biophysics",
"genetics",
"and",
"genomics"
] |
2012
|
Replication Fork Polarity Gradients Revealed by Megabase-Sized U-Shaped Replication Timing Domains in Human Cell Lines
|
Multiple endocrine neoplasia type 2B ( MEN2B ) is a highly aggressive thyroid cancer syndrome . Since almost all sporadic cases are caused by the same nucleotide substitution in the RET proto-oncogene , the calculated disease incidence is 100–200 times greater than would be expected based on the genome average mutation frequency . In order to determine whether this increased incidence is due to an elevated mutation rate at this position ( true mutation hot spot ) or a selective advantage conferred on mutated spermatogonial stem cells , we studied the spatial distribution of the mutation in 14 human testes . In donors aged 36–68 , mutations were clustered with small regions of each testis having mutation frequencies several orders of magnitude greater than the rest of the testis . In donors aged 19–23 mutations were almost non-existent , demonstrating that clusters in middle-aged donors grew during adulthood . Computational analysis showed that germline selection is the only plausible explanation . Testes of men aged 75–80 were heterogeneous with some like middle-aged and others like younger testes . Incorporating data on age-dependent death of spermatogonial stem cells explains the results from all age groups . Germline selection also explains MEN2B's male mutation bias and paternal age effect . Our discovery focuses attention on MEN2B as a model for understanding the genetic and biochemical basis of germline selection . Since RET function in mouse spermatogonial stem cells has been extensively studied , we are able to suggest that the MEN2B mutation provides a selective advantage by altering the PI3K/AKT and SFK signaling pathways . Mutations that are preferred in the germline but reduce the fitness of offspring increase the population's mutational load . Our approach is useful for studying other disease mutations with similar characteristics and could uncover additional germline selection pathways or identify true mutation hot spots .
Multiple endocrine neoplasia ( MEN ) type 2 is characterized by thyroid cancer , variable penetrance of tumors or hyperplasia in other endocrine organs and mutations in RET , the receptor tyrosine kinase proto-oncogene “rearranged during transfection” [1] , [2] . The disease is transmitted in an autosomal dominant fashion . MEN2 has two subtypes: MEN2A ( OMIM 171400 ) accounts for ∼90–95% of cases including a less penetrant sub-form ( familial medullary thyroid carcinoma , FMTC [2] , [3] ) , while MEN2B ( OMIM 162300 ) makes up the remaining ∼5–10% . MEN2B is characterized by a number of interesting genetic features . ( 1 ) Half of all new cases result from sporadic mutations , the vast majority ( >95% ) of which arise in the male germline [4] , [5] . ( 2 ) The average age of the males who transmit a new mutation to their children is greater than that of the average age of all fathers ( paternal age effect [4] ) . ( 3 ) The overwhelming majority ( 95% , [6] ) of new MEN2B mutations occur at the same nucleotide site ( c . 2943T>C ) resulting in the same amino acid substitution ( M918T ) [7]–[9] . ( 4 ) Given that most new cases are caused by mutations at this one site , the incidence of the disease implies that the c . 2943T>C nucleotide substitution frequency is several hundred fold greater than the genome average mutation frequency estimated from evolutionary sequence comparisons [10]–[15] and direct disease incidence data [16] , [17] ( see Text S1 for the detailed calculation ) . One possible explanation for the elevated frequency and paternal age effect is that the c . 2943T nucleotide site in RET is unusually susceptible to undergoing the T>C transition mutation compared to a T elsewhere in the genome ( hot spot model ) . An alternative possibility ( germline selection model ) is that the c . 2943T>C mutation is not unusually susceptible to mutation but , as a result of the biochemical consequences of the MEN2B amino acid substitution , the mutated self-renewing Ap spermatogonial stem cell ( SrAp ) is provided with a proliferative advantage . ( The designations Ap , and Ad which we discuss later in this manuscript , refer to the cytological staining properties of the pre-meiotic A-pale and A-dark spermatogonia , respectively; reviewed in [18] ) . There is considerable evidence that the two common Apert syndrome FGFR2 mutations confer a germline advantage on human SrAp ( reviewed in [19] , [20] ) . Apert syndrome also shares many of the same interesting genetic features as MEN2B . Unfortunately , the biochemical role normally played by FGFR2 in mammalian SrAp function is virtually unknown [21]–[23] . One advantage , then , to asking whether a germline selective advantage is responsible for the elevated MEN2B mutation frequency and paternal age effect is that wild type RET's role has been extensively studied in mouse testis and is known to be required for the self-renewal of mouse spermatogonial stem cells ( SSC , reviewed in [24] , [25] ) . Our observation that the MEN2B mutation does in fact confer a germline selective advantage allows a unique insight into the molecular pathways involved in positive germline selection in human males .
We follow our recently developed approach [26] , [27] by measuring the spatial distribution of the MEN2B c . 2943T>C mutation in fourteen testes from normal men . We then quantitatively test whether or not these distributions are consistent with the hot spot model that predicts a uniform distribution of SrAp with new mutations or the selection model that predicts these mutant cells will be clustered . Both models assume that the germ cells that undergo mutation are uniformly distributed in the testis ( for details supporting this assumption see Text S2 ) . Each testis was cut into 6 slices and each slice into 32 pieces of approximately equal size . The amount of DNA in each piece was quantitated and the frequency of mutant MEN2B molecules was established for each piece using a highly sensitive modification of allele-specific PCR called PAP [28] that gave a false positive rate of 4 . 7×10−7 ( based on analysis of ∼2 . 7×108 control genomes ) . For each testis piece we estimate the mutation frequency per million genomes ( pmg ) . In Figure 1 , this frequency is represented by a heat map with colors ranging from light gray to dark brown . Dataset S1 contains mutation frequency estimates for each testis piece . In Table 1 , we use several statistics to summarize this data . For each testis , we consider the average mutation frequency of all the pieces ( Av ) . Many testes have individual pieces with frequencies that are very different from the average . For each testis , we also identify the piece with the maximum mutation frequency ( Mx ) . In order to normalize for the varying average frequencies among different testes , we consider the ratio of Mx to Av in each testis ( Mx/Av ) . In addition , we consider the fraction of testis pieces with mutation frequencies less than 50 pmg ( F<50 ) ; in Figure 1 these pieces are colored light or dark gray . The youngest age group is made up of three individuals 19 , 21 , and 23 years of age . For this age group , the Av ranges from 1 to 15 pmg . The Mx ranges from 13 to 65 pmg . Figure 1 shows that all the pieces' mutation frequencies are colored light gray ( <25 pmg ) , dark gray ( 25 to 50 pmg ) , or pink ( 50 to 500 pmg ) . The few pieces colored pink in this age group are in the low end of the pink range , since the one with the greatest frequency is only 65 pmg . For each testis the F<50 ranges from 95% to 100% . Six individuals , aged 36 to 68 years , comprise the middle-aged group . For these testes , the Av ranges from 19 to 1 , 188 pmg . In contrast to the youngest age group , each testis has a small number of pieces with mutation frequencies that are several orders of magnitude greater than the remaining pieces . The Mx ranges from 643 to 48 , 884 pmg . These high frequency pieces are more darkly colored in Figure 1 , and are often clustered together in the same slice or in adjacent slices . The sample with the lowest Av ( #59089 ) also has the lowest Mx , and the sample with the highest Av ( #854-2 ) also has the highest Mx . The Mx/Av ratio ranges from 34 to 139 . The F<50 fraction is still high , ranging from 76% to 99% . Both the Av and the Mx are greater for the middle-aged group than the youngest age group . However , within the middle-aged group there is no obvious correlation between frequency and age . Indeed the testis with both the lowest Av and Mx is from a 45 year old ( #59089 ) , and the testis with both the highest Av and Mx is from a 54 year old ( #854-2 ) . So the extreme frequencies come from individuals with ages in the middle of the group , and with ages that are close to each other . The oldest age group containing five individuals aged 75 to 80 years is heterogeneous . Two of the individuals ( #64302 and #60954 ) have frequency values typical of the middle-aged group: the Av are 75 and 203 pmg , the Mx are 4 , 372 and 6 , 673 pmg , the Mx/Av are 58 and 33 , and the F<50 are 98% and 84% . The remaining three individuals ( #60955 , #57650 , and #60507 ) have much lower frequency values typical of the youngest age group: the Av are 10 pmg or less , the Mx are 56 pmg or less , and the F<50 are 99% and 100% . The three low frequency old samples will be further discussed later in the Results section . For discussion purposes , we define a testis as having “substantial” mutation clusters if Mx is greater than 500 pmg: this group includes all of the middle-aged samples and the first two from the oldest age group , while excluding all of the youngest samples and the last three from the oldest age group . Previously , we developed a model based on what is known about human germ-line development and maturation to quantitatively test whether the mutation distribution in a testis is consistent with a hot spot model [26] , [27] . Here we briefly review the model , apply it to the c . 2943T nucleotide site in the RET gene , and discuss a new variant to the model . The computer programs to simulate all the models discussed here and elsewhere in the paper can be found in Protocol S1 . The hot spot model has two phases that we call the growth-phase and the adult-phase . The growth-phase models the testis from zygote formation to puberty . During this phase , divisions of the male germ-line cells are symmetric and self-renewing , and the number of such cells increases exponentially . Similar to a Luria and Delbruck “mutation jackpot” in bacteria [29] , a mutation arising early in this phase will be shared by more descendent germ-line cells than will later mutations . The primordial germ cells migrate to the site of gonad formation and form the seminiferous cords early in fetal development [30] , [31] and since germ cells are expected to remain physically close to their ancestors once the chords are formed , further cell divisions of early mutations can result in mutation clusters . There are approximately 30 growth-phase generations [32] . The germ-cells originating during the growth-phase eventually form the adult SrAp . These cells cycle throughout a man's life providing many opportunities for new mutations . The adult-phase portion of the model considers the testis after puberty . During this phase , the SrAp divide asymmetrically to produce a daughter SrAp ( self-renewal ) and another daughter cell whose descendants , after a few additional divisions , will produce sperm . In an adult male , the SrAp divide every 16 days [33] , and therefore from an individual's age we can estimate the number of adult phase generations that his SrAp cells have experienced . In our model any new mutation in the adult phase can produce only one mutant SrAp self-renewing cell lineage . The model has only one free parameter: the mutation rate per cell division . For each testis , the data is the mutation frequencies of the 192 testis pieces . In order to test the hot spot model using the maximum likelihood approach one would need to calculate the probability , as a function of the model parameters , of the mutation frequencies for all 192 testes pieces . Unfortunately , none of the models we consider are amenable to such calculations . One could estimate this probability function by counting the number of computer simulations of the model that exactly match all 192 frequencies . However , the probability of exactly matching all 192 frequencies is so low that this approach is not feasible . The goodness-of-fit strategy that we pursue instead is we test whether there are values of the model parameters such that computer simulations of the model can approximately match the three summary statistics Av , Mx/Av , and F<50 simultaneously . These statistics attempt to summarize both the mutation frequency and the clustering observed in the testes . Say , for example , a model predicts a more uniform distribution of frequencies than was observed so that simulations which approximate the observed Av statistic also feature much lower than observed Mx/Av ratios . Since this model fails to capture both the mutation frequency and the clustering observed in the testis , we would reject such a model . Alternatively , suppose another model approximately matches the three summary statistics simultaneously . Since this model reproduces both the mutation frequency and the clustering observed in the testis , we would declare such a model consistent with the data . Let us consider the example of testis #374-1 from a 62 year old . The observed Av is 68 pmg ( Table 1 ) . In simulations , we vary the mutation rate per cell division until we find the value of this model parameter such that the simulated Av best matches the observed Av . We simulate the model using this parameter value until we have one million simulations where the simulated Av is within 5% of the observed Av , and then we compare the other statistics for these simulations ( Table 2 ) to the actual data . For the observed data the Mx/Av ratio is 85 , while in 95% of simulations the ratio is between 2 . 1 and 4 . 3 . Indeed , in one million simulations this ratio is always less than was observed in the data . Similarly the observed Mx is 5 , 784 pmg , while in 95% of simulations the Mx is between 144 and 288 pmg . Since we only consider those simulations such that the simulated Av is within 5% of the observed Av , the results for the two statistics Mx and Mx/Av closely correspond . Since we find the ratio Mx/Av more intuitive , we will only consider it subsequently . Likewise , for the data the F<50 statistic is 90% , while in 95% of simulations this fraction is between 25% and 35% . In one million simulations this fraction is always less than was observed in the data . Thus we are able to strongly reject the hot spot model with p-value less than 10−6 . In Table 2 , we see the same conclusion holds for the remaining seven testes with substantial mutation clusters . Note that for testis #59089 in 95% of simulations the F<50 statistic is between 99% and 100% because the Av ( 19 pmg , Table 1 ) is less than 50 pmg . In the hot spot model , a mutation early in the growth phase could produce a mutation cluster . In order to match the observed Av , however , the inferred value of the mutation rate per cell division model parameter is low enough such that mutations early in the growth phase are rare . Since the SrAp divide every 16 days , in a 62 year old male there have been approximately 500 times more adult phase cell divisions than growth phase divisions [27] and mutations in the adult phase do not produce mutation clusters . Consequently , in simulations of the hot spot model that match the observed Av the distribution of mutations is more uniform than observed . Furthermore , even if one does not agree with the modeling details , the youngest age group has markedly lower Av and Mx statistics than the middle-aged group ( Table 1 ) . Therefore , the increase in the mutation frequencies and the growth of the mutation clusters occurs in the adult , not during development . Finally , we previously examined the distribution of a likely neutral mutation in testis samples using the same approach [27] . We assayed a C to G mutation in the intron of the CAV1 gene on chromosome 7 . This presumably neutral mutation was studied in testes 374-1 and 374-2 ( 62 years of age ) and involved the same DNA samples we used for the MEN2B analysis . The summary statistics are identical for both testes ( Av = 3 , Mx = 20 , Mx/Av = 6 . 67 and F<50 = 100% ) and similar to the MEN2B data from much younger donors . Simulations showed that the relatively uniform distribution of mutations was consistent with the hot spot model . Based on work in the mouse [34] and human [35] , we also consider a variant to the hot spot model where the SrAp in the adult phase , independent of whether or not they have acquired the disease mutation , may divide symmetrically . As in the original model , each SrAp in the adult phase divides every 16 days , but now there are three possible types of divisions ( the probabilities of these types sum to one ) . This variant introduces a second model parameter q . With probability 1-2q , the SrAp cells divide asymmetrically as in the original hot spot model . However , now with probability q , the SrAp cells divide symmetrically producing two SrAp cells: both daughter SrAp cells share any accumulated mutations and since these cells remain physically near each other , multiple symmetric divisions would produce a mutation cluster . Also with probability q , in order to keep the number of SrAp cells approximately constant [36] , [37] , an SrAp cell can produce two differentiated daughter cells ( B spermatogonia ) that both go on to make sperm thus eliminating one SrAp cell lineage . For a given mutation rate per cell division the Mx/Av and F<50 statistics increase with the value of the symmetric parameter q , therefore to make the test as conservative as possible we only consider the case where q equals the maximum possible value 0 . 5 ( so one-half of the divisions produce two SrAp cells and one-half produce two B spermatogonia ) . As in the test of the original hot spot model , we simulate the model with the mutation rate per cell division that best matches the observed Av until we have one million simulations with Av within 5% of the observed Av . We again consider sample #374-1 . For the data the Mx/Av ratio is 87 , while in 95% of simulations this ratio is between 5 . 6 and 12 . 9 . For the data F<50 is 90% , while in 95% of simulations this fraction is between 55% and 70% . The symmetric variant to the hot spot model increases these statistics greater than the level achieved by the original hot spot model , but not as high as is observed for the data . Since in one million simulations both the Mx/Av ratio and the F<50 fraction were always less than was observed in the data , this variant is also strongly rejected with p-value less than 10−6 . As shown in Table 2 , the same conclusion holds for the other testes with substantial mutation clusters . Previously , in order to explain the mutation clustering for the Apert syndrome mutations , we had proposed a role for selection [26] , [27] . The selection model is based on the original hot spot model , and adds a selection parameter p: at each adult phase generation , a mutated SrAp divides symmetrically with probability p and asymmetrically with probability 1-p ( after a symmetric division , each daughter SrAp reverts to asymmetric divisions until the next rare symmetric division ) . A similar model was proposed by Crow [38] . Unlike the symmetric hot spot model considered above , non-mutated SrAp cells in the adult phase can only divide asymmetrically . Since the SrAp daughter of an SrAp cell is expected to remain near its progenitor , these rare symmetric divisions can cause mutation clusters to form and grow locally over time . The motivation for the selection model is that in model organisms it has been shown that stem cells can switch from asymmetric to symmetric divisions and back again , and that such behavior can depend on factors intrinsic and extrinsic to the stem cells [39] , [40] . Consider again sample #374-1 for the MEN2B mutation . The selection model has two free parameters: the mutation rate per cell division and the selection parameter p . With these two free parameters , we can now try to match both the Av and the Mx . As before , we only consider those simulations such that the simulated Av is within 5% of the observed Av . For the data the Mx/Av ratio is 85 , and in 95% of simulations this ratio is between 26 and 92 . For the data the F<50 fraction is 90% , and in 95% of simulations this fraction is between 86% and 93% . Therefore the selection model is consistent with the data for this testis . The inferred selection parameter p is 0 . 0084 , so if the mutated SrAp cells divide symmetrically approximately 1% of the time , this is sufficient to form mutation clusters similar to what is observed in the testes . Moreover , if we now take the inferred mutation rate per cell division ( 4 . 4×10−11 ) and set the selection parameter p equal to zero , then simulations of the model produce mutation frequencies similar to the already established genome averages [10]–[17] , implying that the mutation rate per cell division is not elevated at this nucleotide [26] , [27] . The selection model can explain the paternal age effect since the mutation clusters will grow as the man ages and the male mutation bias since this growth is only in the male germline . The selection model can also explain the elevated mutation frequencies and the clustering observed for all the other testes with substantial mutation clusters ( results not shown ) . However , this model predicts that the samples in the oldest age group will have the highest mutation frequencies and the most intense mutation clusters , and thus cannot explain the low mutation frequencies and lack of mutation clusters observed in three of the testes from this age group ( see next heading ) . We have also considered a “combined” model which merges the symmetric variant to the hot spot model with the original selection model: all SrAp randomly divide asymmetrically , divide symmetrically or divide to produce two differentiated daughter cells , but the mutant SrAp are more likely than the wild type SrAp to divide symmetrically . However , we did not pursue this combined model further since it introduces an additional model parameter without improving the fit of the selection model . Our results for the oldest individuals were surprising in that three ( #60955 , #57650 , and #60507 ) of the five samples had unexpectedly low levels of the c . 2943T>C MEN2B mutation similar to young testes ( Table 1 ) . One trivial explanation for such low levels of mutation was germ cell degradation in these three older samples and that this data should be discarded . To examine this question we looked at the distribution of a different mutation . We used the version of our assay originally designed for the Apert syndrome c . 755C>G mutation [26] , [27] on the same 14 testes we studied for the MEN2B mutation ( plus one 21 year old sample , #63205 , which we had not studied for MEN2B ) . Figure S1 shows the Apert syndrome mutation distribution for all the testes and Dataset S2 contains the mutation frequency estimates for every piece . Table 3 summarizes the mutation frequency statistics for each testis and Figure 2 shows the mutation distribution for the five testes in the oldest age group . The results showed substantial Apert mutation clusters in all five older testes including those with the fewest MEN2B mutations . Therefore general germ cell degradation in the three testes cannot explain the heterogeneity in the MEN2B data . Another observation , which will play a part in the subsequent modeling , is that for the middle-aged group of testes the median Av is ∼4-fold higher and the median Mx is ∼3-fold higher for the Apert mutation compared to the MEN2B mutation ( see Table 1 and Table 3 ) . To try and explain the MEN2B data on the oldest age group , we concluded that the only acceptable model modification was to incorporate age-dependent cell death . Researchers have shown that the number of SrAp cells decreases as men grow old [36]: from the ages of 31–40 to 61–70 there is a slight decrease from 162 cells per mm2 of seminiferous tubule cross section to 120 per mm2 , but from the ages of 61–70 to ages 81–90 there is a much more rapid decrease from 120 mm2 to 57 per mm2 . There is a similar pattern of decrease for the A-dark spermatogonia ( Ad ) . Believed to act as “reserve” stem cells , the Ad remain quiescent until the number of SrAp cells is sufficiently diminished to activate the Ad to replace the SrAp [41] . Since the Ad have not been cycling as frequently as the SrAp until this point , they are less likely to have acquired any mutations , and thus the pool of SrAp cells is replenished with a fresh supply of primarily non-mutated cells . We have incorporated cell death into the selection model by assuming that all SrAp , whether or not they are mutated , die at the same rate . The details of this new model can be found in Text S3 . The selection model incorporating cell death can explain all of the testes data for both MEN2B and Apert syndrome . For those testes with substantial MEN2B mutation clusters , as before , we varied the mutation rate per cell division and the selection parameter to try to match both the Av and the Mx , and we only considered those simulations such that the simulated Av was within 5% of the observed Av . Table 4 shows that this model is consistent with these testes . For those testes without substantial mutation clusters , we did not fit each testis separately ( many low values of the model parameters would suffice ) but rather for a given age and set of parameter values we simulated the model many times to see how often the simulations were typical of a young donor and how often they were typical of a middle-aged donor ( see Text S3 for details ) . For MEN2B , we found that when we set the selection parameter at the low end of the range in Table 4 then most simulations of an older individual were typical of a young donor . However , when we increased the selection parameter to the median value in Table 4 then most simulations of an older individual were typical of a middle-aged donor . Thus a relatively slight variation in the selection parameter between individuals can explain the heterogeneity in the older donors for MEN2B . As for the Apert syndrome mutation , the Av and Mx values in Tables 1 and 3 are greater for the Apert mutation than the MEN2B mutation , leading to slightly higher inferred values for the selection parameter for the Apert mutation ( see Table S1 ) . When we increased the selection parameter to the median value for the Apert mutation in Table S1 then almost all of the simulations of an older individual were typical of a middle-aged donor in agreement with Table 3 . The slight increase in the value of the selection parameter for Apert syndrome compared to MEN2B can explain the difference in the oldest age groups for these two mutations . Finally for MEN2B and the youngest donors , even using the greatest values of both the mutation rate per cell division and the selection parameter from Table 4 , almost all simulations of a 23 year old are typical of a young donor in agreement with Table 1 . For these parameter values , the probability of a substantial mutation cluster developing in a 23 year old is very small due to the relatively low number of adult phase generations .
The MEN2B mutation causes aggressive malignant tumors in the thyroid gland during early childhood and later-appearing tumors in the adrenal glands ( pheochromocytoma ) as well as other kinds of abnormal growths [2] , [44] but testis cancer has not been reported as a feature of men with the disease ( [2] updated 5/4/2010 ) . A study of the most common type of human testis cancers ( seminomas ) and rare spermatocytic seminomas both failed to find tumors carrying the MEN2B mutation [45] , [46] . Using a mouse model of MEN2B [47] testicular cancer was not observed in either homozygotes ( RetMEN2B/MEN2B ) or heterozygotes RetMEN2B/+ . Men with MEN2B disease ( carrying the mutation in all their cells ) can father children with MEN2B [4] , and RetMEN2B/MEN2B and RetMEN2B/+ mice also show normal sperm production [47] . These results indicate that the functional properties of the MEN2B protein are consistent with normal spermatogenesis and spermiogenesis but do not contribute to germ cell tumor formation . On the other hand hyperactivation of RET in Ret+/+ mice by overexpressing glial derived neurotrophic factor ( GDNF ) , the major ligand of RET in the testis , results in germ cell tumors that are seminoma-like and led to disruption of normal spermatogenesis [48] . We conclude that the activating effect of the MEN2B mutation on RET function in SrAp cells is minimal compared to RET function in the mouse germline when continually activated by GDNF . A sufficient selective advantage relative to wild type SrAp is necessary to allow cluster formation yet the achievement of this advantage by MEN2B SrAp cells still allows the production of the differentiating daughter spermatogonia needed for normal sperm production and transmission to the next generation . To understand how MEN2B RET alters the normal function of SrAp cells we must consider RET's normal biochemical properties ( reviewed in [44] , [49] , [50] ) . RET is a receptor tyrosine kinase activated after forming a complex with both GDNF and a member of the GDNF-family α co-receptors anchored to the cell surface ( GFRα1 ) . Complex formation results in RET dimerization and induction of RET's tyrosine kinase activity resulting in trans-autophosphorylation of critical tyrosines in each RET monomer's intracellular domain . Interactions between these phosphorylated tyrosine docking sites and adapter or signaling proteins initiates a variety of downstream signaling pathways . Wild type RET functions in a wide variety of cells and tissues including the central and peripheral nervous system , during the development of the kidney and in a variety of other organs ( including testis ) . Whether cell proliferation , cell survival , differentiation or a myriad of other cell processes are stimulated or inhibited by RET activation depends on which RET protein isoforms are expressed , the specific cell types involved , their developmental stage and the expression patterns of many additional proteins that function in downstream signaling pathways including Ras/MAPK , SFK and PI3K/AKT among others ( reviewed in [44] , [49] , [50] ) . Since published work on RET's direct role in the adult human testis is extremely limited we look to studies of the mouse's RET protein for help in understanding how a new MEN2B mutation in a wild type testis might lead to a germline selective advantage of the newly mutated cell in humans . RET signaling is critical for the continuing self-renewal of spermatogonial stem cells in the mouse ( SSC ) and thus spermatogenesis ( reviewed in [23] , [24] ) . Self-renewal requires balancing the number of SSC divisions that lead to more SSCs against SSC divisions that produce precursors of differentiating spermatogonia so that both the number of stem cells and the amount of sperm production is sustained throughout life [23] , [25] , [51]–[53] . We suggest that pathways involved in maintaining this balance are subtly modified by the M918T mutation . Using mouse knockout and overexpression models , experiments on Gdnf , Gfrα1 or Ret have shown that all three genes are critical for SSC self-renewal . Other experiments using testis cell cultures grown in serum-free chemically defined media showed that GDNF alone could promote SSC self-renewal for long periods of time [54] implicating RET as being critical for this process . Transgenic mice carrying a phenylalanine mutation ( Y1062F ) at one of RET's critical tyrosines [55] lose all germ cells within three weeks suggesting that Y1062 function influences this process . Studies in vitro using mouse undifferentiated spermatogonial cell cultures grown in serum-free media showed that the PI3K ( phosphatidylinositol-3 kinase ) /AKT and SFK ( SRC family kinases ) pathways play an important role in SSC self-renewal [23] , [25] , [51] . It has also been proposed that RAS activation is involved in SSC self-renewal [53] . The biochemical consequences of the M918T mutation on RET function have been studied extensively in a variety of tissues and cell types but not the germline ( reviewed in [44] , [49] , [50] , [56] ) . The M918T protein can be activated as a monomer even before being inserted into the cell membrane unlike wild type RET , which normally requires dimerization through ligand binding at the cell surface . The mutant protein also alters its own pattern of tyrosine autophosphorylation in the intracellular domain of RET . This can result in weakened signaling for some downstream pathways and/or activation of pathways not normally signaled by wild type RET . Notably , the M918T mutant protein also shows constitutively high levels of tyrosine phosphorylation especially at Y1062 , the docking site that influences SSC self-renewal in mice . Proteins that normally bind to Y1062 in wild type RET might be expected to signal the PI3K/AKT , and/or RAS/MAPK downstream pathways at unusually high levels in the testis . Constitutive phosphorylation of the RET binding site ( Y981 ) for SRC ( a SFK member ) would also be expected to affect SRC signaling . Based on the role of RET in mouse SSC self-renewal we propose that the M918T mutation modifies the signal transduced from GDNF activated RET downstream through the SFK , PI3K/AKT and possibly RAS pathways leading to the production of mutation clusters . In the human context self-renewal can be achieved by balancing asymmetric and the two types of symmetric SrAp cell divisions as described by our models . The details of how the SFK , PI3K/AKT and possibly RAS pathways enable normal SrAp self-renewal in the germline and how the MEN2B mutation alters these processes to confer a selective advantage to mutated SrAp cells are yet to be determined . Apert syndrome ( OMIM 101200 ) is characterized by premature closing of the sutures between the bones of the skull ( craniosynostosis ) due to gain of function mutations in the receptor tyrosine kinase fibroblast growth factor receptor 2 gene ( FGFR2 ) , one of four such receptors ( FGFR1-4 ) . The disease manifestations of MEN2B and Apert syndrome are very different yet , like MEN2B mutations , new Apert syndrome germline mutations also arise at an unexpectedly high frequency ( 100–1 , 000 times that expected ) at a limited number of nucleotide sites ( c . 755C>G or c . 758C>G ) , virtually always occur in the male parent and exhibit a paternal age effect ( reviewed in [19] , [38] , [42] , [43] , [57] ) . Both Apert syndrome mutations are distributed in testes as clusters rather than uniformly ( [26] , [27] and this paper ) and , together with other results [22] , [38] , [58]–[61] , support the idea that the unexpectedly high frequency of the two Apert nucleotide substitution mutations and paternal age effect results from a selective advantage acquired by mutated SrAp cells over wild type cells . Normal FGFR2 activation follows ligand binding to a subset of the 18 human fibroblast growth factors [62] , dimerization and transphosphorylation . FGFR2 downstream signaling pathways can influence cell proliferation , cell survival , differentiation and a myriad of other cell functions in many different cell and tissue types ( reviewed in [63]–[67] ) . Unfortunately , compared to RET , virtually nothing is known about the downstream signaling pathways stimulated by activated FGFR2 in mouse undifferentiated spermatogonial cultures or other germ cells [21]–[23] although the addition of one of the fibroblast growth factors ( basic bFGF/FGF2 ) , also a ligand of FGFR2 , is required for the self-renewal of mouse SSC in cell culture ( reviewed in [23] ) . The two Apert syndrome mutations increase FGFR2's FGF ligand binding affinity and alter ligand specificity . Consequently , FGFR2 has an unusually long residence in the cell membrane during which time interactions with other proteins could produce abnormally persistent downstream activation signals [63]–[66] , [68]–[72] . Whether mouse or human SSC self-renewal also involves FGFR2 signaling through the PI3K/AKT and SFK pathways or some other pathway is not known although activated FGFR2 can signal through the Ras/MAPK and PI3K/AKT ( and possibly SFK [73] ) pathways in non-germline cells and tissues . More information about the normal function of RET and FGFR2 in spermatogonia are needed to provide an explanation for how both these disease mutations might perturb the signaling landscape to elicit a positive germline selective advantage . Our discovery that MEN2B seems to provide a spermatogonial selective advantage will make it possible to leverage the information on the function of RET with regard to what may be learned about FGFR2 and vice versa and to find out the similarities and differences between the pathways that lead to a germline selective advantage in each case . In a theoretical analysis twenty years ago Hastings [74] , [75] studied the consequences of germline selection on the mutational load of a population . The focus of Hastings' work was primarily concerned with very rare recessive alleles already existing in the population but primarily found only in heterozygotes . Hastings suggested that germline genetic events leading to loss of heterozygosity ( e . g . gene conversion , mitotic crossing over ) in an individual heterozygous for the rare allele could produce premeiotic germ cells homozygous for this allele . If homozygosity for this allele was selectively disadvantageous in both the germline and at the level of an individual the potential mutational load of this allele on the population would be significantly reduced . Hastings also suggested that rare recessive but selectively advantageous alleles could increase in frequency in the population by the same reasoning . Finally , he recognized the possibility that some mutations may have a germline selective advantage but also a selective disadvantage for individuals in the population . This would create what he called a “mitotic drive” system and increase the mutational load of the population . The MEN2B and Apert mutations seem to be realizations of this idea . In both examples the mutation incidence and the magnitude of the paternal age effect are markedly greater than would be expected simply by an increase in the mutation rate per cell division . There are additional candidate de novo disease mutations at other loci that might also provide a germline selective advantage ( reviewed in [19] , [20] , [38] , [42] , [43] ) to spermatogonia . Of course the ability of any such mutation to increase the human mutational load depends upon not significantly interfering with the state of differentiation of the stem cell so as to permit mutant sperm formation and transmission to the next generation . As we learn more about the germline signaling pathways involved and the effect of specific mutations on those pathways it may be possible to identify other mutations that might either provide a selective advantage and be transmitted to the next generation or those that will not be transmitted to the next generation in spite of their germline selective advantage because they interrupt a fundamental aspect of spermatogonial differentiation . The testis dissection method can be useful in studying these questions in animals .
The study was approved by the Institutional Review Board of the University of Southern California . It involved anonymous organ donors and was certified as exempt: 45 CFR 46 . 101 ( b ) ( 4 ) . Testes were obtained from the National Disease Research Interchange ( NDRI , Philadelphia , PA ) . No donors were accepted if they had been treated with drugs known to interfere with normal spermatogenesis . All samples were frozen within 10–12 h after death . Details of the dissections , DNA isolation and quantitation of the amount of DNA in each testis piece have been published [26] , [27] . We modified a highly specific amplification assay that uses dideoxy-terminated PCR primers in a reaction buffer with added pyrophosphate ( pyrophosphorolysis-activated PCR , or PAP [28] ) . In general our assay is almost identical to that used in our earlier work [26] , [27] , [61] . Each MEN2B amplification reaction contained 20 mM HEPES ( pH 7 . 0 ) , 30 mM KCl , 50 µM Na4PPi , 2 mM MgCl2 , 80 µM of each dNTP and 160–320 nM of each primer . The MEN2B specific primer sequences were 5′TGCGTGGTGTAGATATGATCAAAAAGGGATTCAATTGCCGdd3′ ( Biosearch ) and 5′ TCCATCTTCTCTTTAGGGTCGGATTCCAGTTAAATGGACdd 3′ ( IDT ) . Each reaction also included 2 µM Rox , 0 . 2 X Syber Green I , 0 . 04 unit/uL TMA31FS DNA polymerase ( Roche Molecular Systems ) , and DNA containing 25 , 000 genomes from the testis piece . [Research samples of Tma31FS DNA polymerase may be obtained from Dr . Thomas W . Myers , thomas . myers@roche . com , Director , Program in Core Research , Roche Molecular Systems , Inc . , 4300 Hacienda Drive , Pleasanton , CA 94588] . PCR was carried out in 384 well plates using either a Roche LightCycler 480 or a Applied Biosystems 7900 . The cycling conditions ( Roche LightCycler 480 ) were: initial denaturation 1 min , 94°C and 130 cycles of 6 s , 94°C and 1 min , 73°C . Initial denaturation using the Applied Biosystems 7900 was 1 min , 94°C followed by 130 cycles of 6 s , 94°C and 1 min 15 s , 74 . 4°C . In this paper the false positive rate in the MEN2B assay was 4 . 7×10−7 based on an analysis of 269 million wild type genomes . We initially estimated the MEN2B mutation frequency using ten reactions ( 25 , 000 genomes per reaction ) from every testis piece . If less than 5/10 reactions were positive we took that number as an estimate of the mutation frequency ( after Poisson correction ) . If 5 or more reactions were positive we repeated the experiment using diluted samples of the piece until fewer than 50% were positive . The presence of a single mutant molecule in any reaction was detected by examining the kinetics of fluorescence increase as a function of cycle number using quantitative PCR and evaluation of the PCR product melting profile . Sample PAP data has already been published as a supporting figure in an earlier publication [27] . In every experiment 20 negative controls each contained 25 , 000 human blood genomes from unaffected individuals ( Promega ) . Twenty positive controls each contained 25 , 000 control blood genomes and an average of 0 . 5 or 1 genome of MEN2B DNA with the c . 2943T>C mutation ( kindly provided by Dr . Robert Hofstra ) . The estimate of the total testis mutation frequency ( mutations per million genomes ) was the average of the frequencies of the pieces weighted by the number of genomes in those pieces . The computer code and instructions to simulate all of the models can be found in Protocol S1 .
|
Multiple endocrine neoplasia type 2B ( MEN2B ) is a highly aggressive thyroid cancer syndrome . MEN2B offspring with unaffected parents almost always received a new mutation from the father . Moreover , this mutation is almost always at the same nucleotide in the RET proto-oncogene . Thus MEN2B's incidence should equal the average single nucleotide mutation frequency , but the observed incidence is 100–200 times greater . One explanation is that the mutation rate at the causal nucleotide is significantly elevated above the genome average . Another is that human testis stem cells acquiring this mutation have a selective advantage over non-mutated ones and this advantage increases the mutation's frequency in the testis . Computational analysis of our testis dissection and mutation assay data rejects the hot spot but not the selective advantage explanation . Because the normal RET gene is known to be critical for mouse testis stem cell function , we now have an important insight into what biochemical pathways are altered by the MEN2B mutation to provide this selective advantage in humans . Germline selection explains the unexpectedly high incidence of MEN2B , why the mutation's origin is almost always in the father , and why the probability a child is born with this disease increases with the father's age .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"biology",
"computational",
"biology",
"evolutionary",
"biology",
"population",
"biology",
"genetics",
"and",
"genomics"
] |
2012
|
Positive Selection for New Disease Mutations in the Human Germline: Evidence from the Heritable Cancer Syndrome Multiple Endocrine Neoplasia Type 2B
|
The relative proportion of additive and non-additive variation for complex traits is important in evolutionary biology , medicine , and agriculture . We address a long-standing controversy and paradox about the contribution of non-additive genetic variation , namely that knowledge about biological pathways and gene networks imply that epistasis is important . Yet empirical data across a range of traits and species imply that most genetic variance is additive . We evaluate the evidence from empirical studies of genetic variance components and find that additive variance typically accounts for over half , and often close to 100% , of the total genetic variance . We present new theoretical results , based upon the distribution of allele frequencies under neutral and other population genetic models , that show why this is the case even if there are non-additive effects at the level of gene action . We conclude that interactions at the level of genes are not likely to generate much interaction at the level of variance .
In view of the apparent conflict between the observations of high proportions of additive genetic variance ( often half or more of the phenotypic variance , and even more of the total genetic variance ) and the recent reports of epistasis at quantitative trait loci ( QTL ) [8] , we consider explanations beyond that of simple sampling errors and bias of estimates . We focus particularly on the role that the distribution of gene frequencies may play in the relation between the genetic model and the observed genetic variance components . Genetic variance components depend on the mean value of each genotype and the allele frequencies at the genes affecting the trait [3] , [4] , [17] . Unfortunately the allele frequencies at most genes affecting complex traits are not known , but the distribution of allele frequencies can be predicted under a range of assumptions . This distribution depends on the magnitude of the evolutionary forces that create and maintain variance , including mutation , selection , drift and migration . As the effects on fitness of genes at many of the loci influencing most quantitative traits are likely to be small , we can invoke theory for neutral alleles to serve as a reference point . An important such reference is the frequency distribution under a balance between mutation and random genetic drift due to finite population size in the absence of selection . If mutations are rare , the distribution of the frequency ( p ) of the mutant allele is f ( p ) ∝1/p , i . e . approximately L-shaped [2] , [35] , [36] , with the high frequency at the tail being due to mutations arising recently . The allele which increases the value of the trait may be the mutant or ancestral allele , so its frequency has a U-shaped distribution f ( p ) ∝1/p+1/ ( 1−p ) = 1/[p ( 1−p ) ] . As we shall use it often , we define the ‘U’ distribution explicitly by this formula . For loci at which the mutants are generally deleterious , the frequency distribution will tend to be more concentrated near p = 0 or 1 than for this neutral reference point . As another simple reference point we use the uniform distribution , f ( p ) ∝1 , 1/ ( 2N ) ≤ p ≤ 1−1/ ( 2N ) , with N the population size . This approximates the steady state distribution of a neutral mutant gene which has been segregating for a very long time [2] , and also has much more density at intermediate gene frequencies than the ‘U’ distribution . Our third reference point is at p = 0 . 5 , as in populations derived from inbred crosses , and is the extreme case of central tendency of gene frequency . These analyses assume a gene frequency distribution which is relevant to no selection . For a more limited range of examples we consider the impact of selection on the partition of variance . We consider a limited range of genetic models , some simple classical ones and others based on published models of metabolic pathways or results of QTL mapping experiments . Uniform: f ( p ) = 1 , assuming N is sufficiently large that the discreteness of the distribution and any non-uniformity as p approaches 1 or 0 can be ignored , i . e . integrated over 0 to 1 . This and the ‘U’ gene frequency distributions are , for simplicity , assumed to be continuous . Neutral mutation model ( ‘U’ ) : f ( p ) ∝1/[p ( 1−p ) ] . To standardise the distribution , with population size N assumed to be large , note that Thus , where K∼ln ( 2N ) . Genetic variance components are obtained by integration of expressions for the variance as a function of p for a specific model of the gene frequency distribution . For multiple locus models the distribution of all loci is assumed to be identical and there is no linkage disequilibrium . We focus on the contribution of additive genetic variance ( VA ) to genotypic variance ( VG ) .
Many general points are illustrated by two simple examples , the single locus model with dominance and the two locus model with AA interaction , so we consider these in more detail . For the single locus model with genotypic values for CC , Cc and cc of +a , d and −a , respectively , VA = 2p ( 1−p ) [a+d ( 1−2p ) ]2 and VD = 4p2 ( 1−p ) 2d2 . For d = a , i . e . complete dominance of C , VA = 8p ( 1−p ) 3a2 and VD = 4p2 ( 1−p ) 2a2 and thus: at p = 0 . 5 , VA = ( 2/3 ) VG; if the dominant allele is rare ( i . e . p → 0 ) , VG → 8p and VA/VG → 1 , and if it is common , VG → 4p2 and VA/VG → 0 . Note , however , that VG and VA are much higher when the dominant allele is at low frequency , e . g . 0 . 1 , than are VG and VD when the recessive is at low frequency , e . g . p = 0 . 9 . Even for an overdominant locus ( a = 0 ) , all genetic variance becomes additive at extreme gene frequencies . Considering now expectations ( E ) over the frequency distributions , let η2 = E ( VA ) /E ( VG ) , an equivalent to narrow sense heritability if VE = 0 . For the ‘U’ distribution , η2 = 1−d2/ ( 3a2+2d2 ) and for the uniform distribution , η2 = 1−2d2/ ( 5a2+3d2 ) . Hence , for a completely dominant locus , η2 = 0 . 8 and η2 = 0 . 75 respectively; whereas VA/VG = 0 . 67 for p = 0 . 5 . In summary , the fraction of the genetic variance that is additive genetic decreases as the proportion of genes at extreme frequencies decreases ( Table 2 ) . The genotypic values ( see Theory section ) for the simple AA model for double homozygotes BBCC and bbcc are +2a and for bbCC and BBcc are 0 , and all single or double heterozygotes are intermediate ( +a ) . With linkage equilibrium , VA/VG = 1−HpHq/[Hp+Hq−3HpHq] , where the heterozygosities are Hp and Hq at loci B and C . Thus VA/VG → 1 if either locus is at extreme frequency ( i . e . p or q → 0 or 1 ) , and equals 0 when p = q = 0 . 5 . If p = q , for gene frequencies 0 . 1 , 0 . 2 , 0 . 3 and 0 . 4 , VA/VG = 0 . 88 , 0 . 69 , 0 . 43 and 0 . 14 . For the uniform distribution η2 = 2/3 , and for the ‘U’ distribution , the variances are a function of the population size , because more extreme frequencies are possible at larger population sizes . Thus η2 = ( 2−4/K ) / ( 2−3/K ) , where K = ln ( 2N ) , so η2 → 1 for large K . Any residue is VAA . These two examples , the single locus and A × A model , illustrate what turns out to be the fundamental point in considering the impact of the gene frequency distribution . When an allele ( say C ) is rare , so most individuals have genotype Cc or cc , the allelic substitution or average effect of C vs . c accounts for essentially all the differences found in genotypic values; or in other words the linear regression of genotypic value on number of C genes accounts for the genotypic differences ( see [3] , p 117 ) . Hence almost all VG is accounted for by VA . With the ‘U’ distribution , most genes have one rare allele and so most variance is additive . Further examples ( Table 2 ) illustrate this point , including the duplicate factor and complementary models where there is substantial dominance and epistasis . These models show mostly VA for the ‘U’ distribution for a few loci but the proportion of the variance which is additive genetic declines as the number increases . With many loci , however , such extreme models do not explain the covariance of sibs ( i . e . any heritability ) or the approximate linearity of inbreeding depression with inbreeding coefficient , F , found in experiments [3] , [4] , [40] , [41] , [42] , or the linearity in response to artificial selection [43] . We also analysed a well-studied systems biology model of flux in metabolic pathways [38] , [39] , [44] and found again that the expected proportion of VG that is accounted for by VA is large ( Table 3 ) . A number of QTL analyses using crosses between populations ( some inbred , some selected ) have been published in which particular pairs ( or more ) of loci have been identified to have substantial epistatic effects [8] . We consider examples of the more extreme cases of epistasis found , obtaining variance components by numerical integration . Results are shown in Table 4 , for examples from [8] deliberately chosen as extreme . Even so , the proportion of the genetic variance that is additive is high with the ‘U’ distribution , except in the dominance × dominance example . Further , as these examples were selected by Carlborg and Haley and us as cases of extreme epistasis , it is not unreasonable to assume that the real epistatic effects are smaller than their estimates . A test of the hypothesis that the lack of non-additive variance observed in populations of humans or animals is because gene frequencies near 0 . 5 are much less common than those more extreme , not because non-additive effects are absent , is to compare variance components among populations with different gene frequency profiles . For crops such as maize and for laboratory animals , estimates can be got both from outbreds and from populations with gene frequencies of one-half derived from crosses of inbred lines . There are a limited number of possible contrasts and linkage confounds comparisons of variation in F2 and later inter se generations , however , so it is difficult to partition variation between single locus and epistatic components ( e . g . [17] ch . 7 ) . The most extensive data are on yield traits in maize . The magnitudes of heritability and of dominance relative to additive variance estimated for different kinds of populations in a substantial number of studies ( including 24 on F2 and 27 on open-pollinated , i . e . outbreds ) have been summarised [59] . Average estimates of h2 were 0 . 19 for open-pollinated populations , 0 . 23 for synthetics from recombination of many lines , 0 . 24 for F2 populations , 0 . 13 for variety crosses and 0 . 14 for composites . Estimates of VA/VG ( from tabulated values of VD/VA [59] ) were 0 . 57 , 0 . 55 , 0 . 50 , 0 . 42 and 0 . 43 , respectively , which are inconclusive but indicate relatively more dominance variance at frequencies of 0 . 5 . Analyses of the magnitude of epistasis at the level of effects , rather than variance , do not provide consistent patterns . For example , in two recent analyses of substantial data sets of F2 populations of maize , one found substantial epistasis [60] and the other almost none [61] . In an analysis of a range of traits in recombinant inbred lines , F2 and triple test crosses [62] in Arabidopsis thaliana , there was substantial additive genetic and dominance variance for all traits , with most estimates of VD/VA in the range 0 . 3 to 0 . 5 , essentially no significant additive × additive epistatic effects , but several cases of epistasis involving dominance [63] . Although there does appear to be more dominance variance in populations with gene frequencies of one-half than with dispersed frequencies , from these results we cannot reject or accept the hypothesis that there is relatively much more epistatic variance in such populations . One explanation is indeed that there is not a vast amount of epistatic variance in populations at whatever frequency , although another is that maize has unusually small amounts of epistasis . Many additive QTL were identified in an analysis of a line derived from the F2 of highly divergent high and low oil content lines from the long term Illinois maize selection experiment , but with almost no evidence of epistasis or indeed dominance effects [64] . In contrast , an F2 of divergent lines of long-term selected poultry and an F2 from inbred lines of mice showed evidence of highly epistatic QTL effects for body weight [65] , [66] . We do not claim to understand these different results , but as has been pointed out [67] , [68] , QTL with significant epistatic interaction effects might not represent the majority of QTL with small effects contributing to gene networks . We have summarised empirical evidence for the existence of non-additive genetic variation across a range of species , including that presented here from twin data in humans , and shown that most genetic variance appears to be additive genetic . There are two primary explanations , first that there is indeed little real dominant or epistatic gene action , or second that it is mainly because allele frequencies are distributed towards extreme values , as for example in the neutral mutation model . Complete or partial dominance of genes is common , at least for those of large effect; and epistatic gene action has been reported in some QTL experiments [8] , [69] . Detailed analyses in Drosophila melanogaster , using molecular and genetic tools available for it , identify substantial amounts of epistasis , including behavioural traits [70] and abdominal bristle number [71] , yet most genetic variation in segregating populations for bristle number appears to be additive ( as noted above ) . But many QTL studies of epistatic gene action suffer from a high degree of multiple testing , increasingly so the more loci and orders of interaction are included , such that they may be exaggerating the amount of epistasis reported . On the assumption that many of the effects are indeed real , we have turned our attention to the second explanation . The theoretical models we have investigated predict high proportions of additive genetic variance even in the presence of non-additive gene action , basically because most alleles are likely to be at extreme frequencies . If the spectrum of allele frequencies is independent of which are the dominant or epistatic alleles , VA/VG is large for almost any pattern of dominance and epistasis because VA/VG is low only at allele frequencies where VG is low , and so contributes little to the total VG . The distribution of allele frequencies is expected to be independent of which are the dominant or epistatic alleles for neutral polymorphisms; but under natural selection the favourable allele is expected to be common and lead to high or low VA/VG depending on whether it is dominant ( low VA ) or recessive ( high VA ) . The equivalent case for epistasis is that all genotype combinations except one is favourable ( low VA ) vs . only one genotype combination is favourable ( high VA ) . If genetic variation in traits associated with fitness is due almost entirely to low frequency , deleterious recessive genes which are unresponsive to natural selection , these traits would show low VA/VG . However , neither the empirical evidence nor the theory supports this expectation . There seems to be substantial additive genetic variance for fitness associated traits [21] and fitness itself [30] , [31] , [72] . Although heritabilities for such traits may be low , they show high additive genetic coefficient of variation ( evolvability ) [29] , and the correlation of repeat records is typically little higher than the heritability ( e . g . , litter size in pigs ) , indicating that VA/VG is one-half or more . In agreement with this , when the life history of deleterious , recessive mutants was modelled , VA/VG was found to be 0 . 44 ( Table 6 ) , basically because rare recessives contribute so little variance , albeit most is VD , in non-inbred populations . We believe we have a plausible gene frequency model to explain the minimal amounts of non-additive genetic and particularly epistatic variance . What consequences do our findings have ? For animal and plant breeding , maintaining emphasis on utilising additive variation by straightforward selection remains the best strategy . For gene mapping , our results imply that VA is important so we should be able to detect and identify alleles with a significant gene substitution effect within a population . Such variants have been reported from genome-wide association studies in human population [9] , [10] , [11] , [12] , [13] . Although there may well be large non-additive gene effects , the power to detect gene-gene interactions in outbred populations is a function of the proportion of variance they explain , so it will be difficult to detect such interactions unless the effects are large and the genes have intermediate frequency . Thus we expect that the success in replicating reported epistatic effects will be even lower than it is for additive or dominance effects , both because multi-locus interactions will be estimated less accurately than main effects and because they explain a lower proportion of the variance . Finally , if epistatic effects are real , gene substitution effects may vary widely between populations which differ in allele frequency , so that significant effects in one population may not replicate in others .
|
Genetic variation in quantitative or complex traits can be partitioned into many components due to additive , dominance , and interaction effects of genes . The most important is the additive genetic variance because it determines most of the correlation of relatives and the opportunities for genetic change by natural or artificial selection . From reviews of the literature and presentation of a summary analysis of human twin data , we show that a high proportion , typically over half , of the total genetic variance is additive . This is surprising as there are many potential interactions of gene effects within and between loci , some revealed in recent QTL analyses . We demonstrate that under the standard model of neutral mutation , which leads to a U-shaped distribution of gene frequencies with most near 0 or 1 , a high proportion of additive variance would be expected regardless of the amount of dominance or epistasis at the individual loci . We also show that the model is compatible with observations in populations undergoing selection and results of QTL analyses on F2 populations .
|
[
"Abstract",
"Model",
"Results/Discussion"
] |
[
"genetics",
"and",
"genomics/complex",
"traits",
"evolutionary",
"biology",
"genetics",
"and",
"genomics",
"genetics",
"and",
"genomics/medical",
"genetics",
"genetics",
"and",
"genomics/population",
"genetics"
] |
2008
|
Data and Theory Point to Mainly Additive Genetic Variance for Complex Traits
|
In addition to their biological function , protein complexes reduce the exposure of the constituent proteins to the risk of undesired oligomerization by reducing the concentration of the free monomeric state . We interpret this reduced risk as a stabilization of the functional state of the protein . We estimate that protein-protein interactions can account for of additional stabilization; a substantial contribution to intrinsic stability . We hypothesize that proteins in the interaction network act as evolutionary capacitors which allows their binding partners to explore regions of the sequence space which correspond to less stable proteins . In the interaction network of baker's yeast , we find that statistically proteins that receive higher energetic benefits from the interaction network are more likely to misfold . A simplified fitness landscape wherein the fitness of an organism is inversely proportional to the total concentration of unfolded proteins provides an evolutionary justification for the proposed trends . We conclude by outlining clear biophysical experiments to test our predictions .
The toxicity due to protein misfolding and aggregation has a considerable effect on the viability of living organisms [1]– . Consequently , cells are under strong selection pressure to evolve thermodynamically stable [6] and aggregation-free protein sequences [7] . The internal region of stable proteins has a tightly packed core of hydrophobic residues . A mutation in the core may disrupt the entire protein structure . Consequently , the core residues are strongly conserved [8] , [9] . In contrast , mutations on the surface contribute weakly to the thermodynamic stability of proteins [10] yet surfaces show significant level of conservation [11] owing to protein-protein interactions . Recent high throughput experiments have established that proteins interact with each other on a genome-wide scale [12] . Such ‘small world’ networks are thought to facilitate biological signaling and ensure that cells remain robust even after a random failure of some of its components [13] . It is thought that evolutionarily , multi-protein complexes are favored over larger size of individual proteins [14] since large proteins are difficult to fold and expensive to synthesize while small interacting proteins can fold independently and then efficiently assemble into large complexes . Individual interaction between proteins can give rise to cooperativity and allostery which results in a finer control over the functional task the protein complex performs . Protein-protein interactions ( PPI ) are also thought to prevent protein aggregation [15] , [16] . Lastly , many proteins can perform promiscuous function in that they can partake in multiple protein complexes . Interestingly , proteins in higher organisms are involved in more interactions and form larger protein complexes compared to more primitive life forms [17] . Here , we hypothesize an additional biophysical advantage for protein-protein interactions . Proteins bound to their interaction partners effectively present a lower monomer concentration inside the cell . Since free monomers are susceptible to misfolding/unfolding and toxic oligomerization , interacting proteins may face a reduced risk towards the same . This reduced risk can be interpreted as interaction-induced stabilization — stabilization due to the protein-protein interaction network — of an otherwise monomeric protein ( see Fig . 1 for a cartoon ) . We propose that by giving proteins an additional stability , each protein in the interaction network acts as an evolutionary capacitor [18] , [19] in the evolution of its binding partners: proteins are allowed to explore the less stable regions ( regions of low intrinsic stability ) of the sequence space as long as they are stabilized by their interaction partners . Inversely , unstable proteins are expected to receive significant additional stability from the interaction network . Below we outline the empirical evidence for our hypothesis and suggest clear biophysical and evolutionary experiments to test it further .
Fig . 2 shows the histogram of the estimated interaction-induced stability for cytoplasmic yeast proteins for whom abundance , interaction , and localization data is available ( see Methods for the details of the calculations ) . Note that the average PPI induced stability is and can be as high as . This stabilization is dependent not only on the number of interaction partners of a given protein or the strengths of those interactions but also on the relative abundances of the interaction partners . In fact , the interaction-induced stability of a protein correlates strongly with the relative concentration of its binding partners ( Spearman . This suggests a plausible mechanism of stabilization of a protein without changing its sequence viz . via adjusting the expression levels of its interaction partners ( see Discussion below ) . The estimated values are of the same order of magnitude as the inherent stabilities of proteins , ( ) [9] . Given that random mutations are more likely to destabilize proteins [6] , we expect protein-protein interactions to act as secondary mechanisms to stabilize proteins and to interfere with the evolution of protein stability . To explore the evolutionary consequences of the interaction-induced stability , we investigate a simplified fitness model of a toy proteome consisting of 15 proteins ( see Methods , Text S1 , and Table S1 ) . Briefly , the fitness of the cell depends only on the total concentration of unfolded proteins in it [20] . During the course of evolution , each protein acquires random mutations that change either a ) its inherent stability or b ) the dissociation constant of its interaction with a randomly selected interaction partner . Even though protein abundance and protein-protein interactions evolve at the same time scale as protein stability , the former are dictated largely by the biological function of the involved proteins . Incorporating the fitness effects of changes in expression levels and interaction partners in our simple model is non-trivial . Thus , in order to specifically probe the relation between stability and interactions , we do not allow proteins to change their abundance and interaction partners . In the model , the concentration of unfolded proteins and thus the fitness of the proteome depends on the total stability of individual proteins . While random mutations are more likely to make proteins unstable , protein-protein interactions increase the total stability . In the canonical ensemble description of the evolution of fitness [21] , the inverse effective population size ( ) , the evolutionary temperature quantifies the importance of genetic drift . The effective population size modulates the competition between destabilizing random mutations and stabilizing protein-protein interactions . We find that at higher effective populations , proteins are inherently stable and only the least stable proteins ( small ) receive high stabilization from the interaction network ( high ) . At low effective population , due to genetic drift , proteins are inherently destabilized and protein-protein interactions serve as the primary determinant of the effective stability of proteins . Fig . 3 shows the dependence of average inherent stability ( ) , average interaction-induced stability ( ) , and average total stability ( ) with effective population size . Interestingly , the total stability ( ) of proteins remains relatively insensitive to changes in population size . We observe that the correlation coefficient between the inherent stability and the interaction-induced stability itself varies with the effective population size . Even though its magnitude decreases , interaction-induced stability becomes more and more correlated with inherent stability as population size increases ( See Fig . 4 ) . In real life organisms , interaction-induced stability acts on a need basis for proteins and serve as a secondary stabilization mechanism . In the drift-dominated regime , which is unlikely to be realized in real life organisms ( except probably in parasitic microbes with low population sizes ) , interaction-induced stability becomes the dominant player in the evolution of total stability of proteins [17] . We next examine if this prediction from the toy model holds for real organisms . Proteome-wide information about the inherent stability of proteins is currently unavailable . Previously , in silico estimates of protein aggregation propensity have been used as proxy for protein stability [22] , [23] . We use the TANGO [24] algorithm to estimate protein aggregation propensity . It is known that TANGO aggregation propensity correlates strongly and negatively with protein stability [24] . TANGO has been verified extensively with experiments on peptide aggregation [24] and has been previously used to study the evolutionary aspects of protein-protein interactions [22] , [25] . Similar analysis for Aggrescan [26] can be found in Text S1 and Table S3 . We find that the aggregation propensity is correlated positively with the interaction-induced stability ( Spearman ) . As expected [2] , the aggregation propensity is negatively correlated with protein abundance ( Spearman ) . The correlation between and does not depend on this underlying dependence and persists even after controlling for total abundance ( partial Spearman ) ( See Table S2 ) . This result suggests in the proteome of baker's yeast , protein stability correlates negatively with interaction-induced stability . The fitness cost of protein aggregation is directly proportional to the amount of aggregate [20] . Thus , the selection forces that make protein sequences aggregation-free act more strongly on highly expressed proteins [1] , [2] , [22] . Our hypothesis suggests that the proteins that are bound to their interaction partners present a lower concentration of the free monomeric state in vivo ( low ) and automatically lower the misfolding/aggregation induced fitness cost , even if highly abundant ( high ) . The selection forces to evolve an aggregation-free sequence may be weaker for such proteins . Consequently , the aggregation propensity should be principally correlated with the free monomer concentration rather than the total abundance . Indeed , we observe that the estimated monomer concentration and the aggregation propensity are correlated negatively ( Spearman ) . Importantly , this correlation is not an artifact of the underlying correlation between the aggregation propensity and total abundance ( partial Spearman ) . At the same time , the partial correlation coefficient between the aggregation propensity and the total protein abundance controlling for the estimated monomer concentration is minimal ( partial Spearman ) . In short , the total free monomer concentration of a protein ( rather than , its total abundance ) might be a better variable to relate to evolutionary and biophysical constraints on the protein . We have thus far shown that a protein's interaction partners can significantly stabilize its folded state and this stabilization interferes with the evolution of the inherent stability of the protein . We now explore the reverse viz . the evolutionary consequences of the ability of each protein to impart stability to its interaction partners . The concept of evolutionary capacitor has been previously introduced for the heat shock protein HSP90 [18] , [19] , which is also a molecular chaperone and a highly connected hub in the PPI network ( 70 interaction partners in the current analysis ) . An elevated concentration of HSP90 buffers the potentially unstable variation in proteins , which may allow proteins to sample a wider region of the sequence space , which may often lead to functional diversification [27] . Similar to HSP90 , each protein in the interaction network has some ability to stabilize its interaction partners to a certain extent . Consequently , we study the evolutionary capacitance of individual proteins in the context of the interaction network by estimating the effect of protein knockout on ppi-induced stability in silico . Proteins with higher evolutionary capacitance are defined as those with the higher cumulative destabilizing effect on the proteome . We write , ( 1 ) For each protein , the sum in Eq . 1 is carried out over all proteins that are destabilized due to its knockout . Here , we assume that the potential of a given protein knockout to generate multiple phenotypes depends on the loss of stability of its interaction partners caused by its knockout . We hypothesize that , similar to unstable proteins requiring HSP90 to fold , the interaction partners of proteins with high capacitance should be unstable . In fact , the capacitance of a protein and the mean aggregation propensity of its interaction partners are strongly correlated ( Spearman ) . The capacitance is significantly correlated with even after controlling for the abundance of the protein ( partial spearman ) and the number of its interaction partners ( partial spearman ) . This suggests that a protein needs to be present in sufficient quantity and should interact with a large number of proteins in order to effectively act as a capacitor . We have presented evidence that all proteins can act as an evolutionary capacitor , albeit with variable effectiveness , for their interaction partners . Traditionally , evolutionary capacitors are understood to be chaperones that buffer phenotypic variations by helping misolding-prone proteins fold in a proper structure [19] . Not surprisingly , when we carried out functional term enrichment analysis using gene ontology [28] , we found that approximately half of the top 20 capacitors have ‘chaperone’ in their name . The top 20 are also over represented in the chaperone-like molecular function of protein binding and unfolded protein binding ( ) and the biological process of protein folding ( ) . These findings validate our definition of capacitors that were previously identified as chaperones . Interestingly , some of the predicted capacitors do not currently have a protein folding-related functional annotation . These need more experimental investigation ( see supplementary File S1 for the list ) . This suggests that previously identified evolutionary capacitor HSP90 may in fact only be one among the broader set of evolutionary capacitors . Every protein in the interaction network is an evolutionary capacitor for its interaction partners and evolutionary capacitor is a quantitative distinction rather than a qualitative one .
In cellular homeostasis , the total concentration of any protein can be written as the sum of its free folded monomer concentration , a fraction comprising of insoluble oligomers and unfolded peptide , and as part of all protein complexes containing ( See Fig . 5 ) . In our computational model , for simplicity and owing to the nature of the large scale data [34] , we restrict protein complexes to dimers [35] , thus for all proteins that interact with , ( 2 ) Conservation of mass implies , ( 3 ) The concentration of each dimer satisfies the law of mass action , ( 4 ) We can write the balance between the three states of the protein , ( See Fig . 1 ) , as two equilibrium equations ( 5 ) ( 6 ) Note that comprises of a collection of biologically unusable states of the protein viz . the misfolded/unfolded and the oligomerized state any of which may convert to/interact with the folded monomeric state . Consequently , the first equilibrium is a collection of thermodynamic equilibriums . The equilibrium constant will thus depend not only on the temperature but also on and . If among the unfolded , misfolded , and the oligomerized states the former dominates the population comprising then , where is the thermodynamic stability of the free monomeric state . Similarly , is given by , ( 7 ) and depends not only on the dissociation constants but also the free concentrations of the interacting partners of protein and on the topology of the interaction network in the organism . Here too , we assume that a ) only the folded monomeric forms of proteins interact with each other and b ) there is no appreciable interaction between the collective unfolded state of protein and any state of any other protein . We have also neglected the role of chaperones in actively reducing the concentration of the unfolded/misfolded/aggregated state by turning it over to the folded state . In fact , some of the chaperones are included in of our mass action equilibrium model and prevent unfolding by sequestering the folded state ( see below and the discussion section ) . By combining mass conservation ( Eq . 3 ) with Eq . 5 and Eq . 6 , ( 8 ) In the above development , we have made a crucial assumption that only . Note that in the absence of interactions , . We identify as the additional decrease in the insoluble fraction due to protein-protein interactions . We define the interaction-induced stability as , ( 9 ) We downloaded the latest set of interacting proteins in baker's yeast from the BIOGRID database [36] . To filter for non-reproducible interactions and experimental artifacts , we retained only those interactions that were confirmed in two or more separate experiments . For the sake of simplicity , we only considered cytoplasmic proteins [37] with known concentrations [38] . This lead to proteins connected by interactions . The in vivo stability of a protein is a combination of its thermodynamic stability , resistance to aggregation or oligomerization , and resistance to degradation [39] . Note that the interaction-induced stability of a protein depends on the stability of its interaction partners ( see Eq . 6 , Eq . 7 , and Eq . 9 ) . Unfortunately , the exact dependence of the in vivo protein stability on its sequence is unclear and there exist no reliable data or sequence dependent computational estimates for the thermodynamic stability of proteins . Moreover , , and thus ( Eq . 6 , Eq . 7 , and Eq . 9 ) , can be estimated even in the absence of the knowledge of . In our estimates of , we assume that is given simply byHere , is obtained by solving the mass action equations [35] iteratively ( see below ) . This is equivalent to assuming that all the proteins are equally and highly stable ( for all proteins ) . The thus calculated serves as the upper limit of interaction-induced stability . In the supplementary materials ( Text S1 , Fig . S1 , Fig . S2 , and Tables S4 and S5 ) , we show that different assignments of the equilibrium constants including a simple model of protein stability [40]–[42] do not change the qualitative nature of our observations . The dissociation constants for protein-protein interactions follow a lognormal distribution with a mean nM [35] . The majority of interactions between proteins are neither too weak nor unnecessarily strong . Common sense dictates that it does not make sense to decrease the dissociation constant between two proteins beyond the point where the abundance limiting protein spends all of its time in the bound state . Motivated by these evolutionary arguments to minimize unnecessary protein production and to avoid unnecessarily strong interactions , Maslov and Ispolatov [35] devised a recipe to assign dissociation constants to individual protein-protein interactions . viz . for interacting proteins and , the dissociation constant . We also explore a few other assignment rules for dissociation constants ( see supplementary Text S1 , Fig . S3 , and Table S6 ) . We solve for free concentrations iteratively [35] . We start by setting for all proteins and iteratively calculate from ( 10 ) till two consecutive estimates of fall within of each other for all proteins . As noted above , the toxic effects of misfolding and aggregation may be the chief determinant of protein sequence evolution [2] , [4] , [5] . The dosage dependent fitness effect of misfolded proteins [20] motivates us to introduce a simple biophysical model for fitness of the proteome ( See Eq . 11 ) , ( 11 ) is the scaling factor . Potentially , can be estimated from fitness experiments by introducing measured quantities of unfolded protein in the cell [20] . We explore the evolution of a hypothetical proteome to investigate the interplay between protein stability and protein-protein interactions . We believe that protein abundances and the topology of the interaction network are largely dictated by biological function . It is non-trivial to incorporate the fitness effect of changes in gene expression level and the network topology in our simplified model . Thus , to specifically probe the relation between stability and interactions , we concentrate on the effect of toxic gain of function due to misfolding and aggregation on cellular fitness and not include changes in gene expression levels and network topology . In this aspect , our model is in the same spirit as previously proposed models [6] , [41]–[48] . The effect of random mutations on average destabilizes proteins and the dynamics of the evolution of thermodynamic stability of proteins can be modeled as a random walk with negative average velocity [6] . We consider the thermodynamic stability as a proxy for the in vivo stability of proteins . We construct the cytoplasm of a hypothetical organism with 15 proteins . The number of proteins is low due to computational restrictions . The proteome is evolved by sampling the dissociation constants from the lognormal distribution while introducing random mutations in proteins that change their stability . At each generation , the fitness is evaluated and the progeny is accepted at a certain evolutionary temperature ( defined as the inverse of the effective population size , ) [21] . We run a total of generations for each evolutionary temperature and analyze the organism in the latter half of the evolutionary run ( details of the model and a brief description of the population genetics terminology is in supplementary Text S1 ) . The notion of protein stability relevant to this study is the propensity of a protein to avoid structural transformations that may render it unemployable for biological function . For example , for a small and highly soluble protein , this stability corresponds to the thermodynamic stability of the native state while for a large multi domain protein , it may correspond to the thermodynamic stability of one of its domains against the partially unfolded state . In short , thermodynamic stability of the folded state with respect to the unfolded , partially folded state , and the misfolded state all contribute to the in vivo stability of proteins [39] . Though there is a lack of proteome-wide estimates of thermodynamic stability of proteins , the aggregation propensity can be estimated from the sequence [24] , [26] and is known to be correlated with protein stability [24] . In our correlation analysis , we use the estimated aggregation propensity as a proxy for in vivo protein stability and explore the relationship between interaction-induced stability and protein stability . The aggregation propensity was estimated for the same proteins used in the mass action calculation to estimate . We tested the TANGO [24] and Aggrescan [26] to estimate the aggregation propensity of proteins . Previously , TANGO has been used [22] , [23] , [49] to understand the relation between protein abundance and instability . We show results for TANGO in the main text . Aggrescan results ( supplementary Text S1 and Table S3 ) are quite similar .
|
The folded form of proteins is only marginally stable in vivo and constantly faces the risk of aggregation , unfolding/misfolding , and other aberrant interactions . For most proteins , the folded form is also the functionally relevant one and forces of natural selection strongly modulate its stability . In vivo , proteins interact with each other on a genome-wide scale . Usually , the interaction of a protein and its binding partners requires both the proteins to be in the folded form and as a result , the interactions tend to shift the population of a protein towards the folded form . Consequently , protein-protein interactions interfere with the evolution of protein stability . Here , we present empirical evidence and theoretical justification for proteins' ability to stabilize the folded form of their interaction partners and allow them to explore the region of the sequence space that corresponds to proteins with less stable structure . We argue that the ‘evolutionary capacitance’ – previously thought to be a property of the chaperone HSP90 , a special class of proteins – is a property of all proteins , albeit to a different degree .
|
[
"Abstract",
"Introduction",
"Results",
"Methods"
] |
[
"evolutionary",
"modeling",
"biology",
"computational",
"biology"
] |
2013
|
Evolutionary Capacitance and Control of Protein Stability in Protein-Protein Interaction Networks
|
Gram-negative bacterial pathogens deliver a variety of virulence proteins through the type III secretion system ( T3SS ) directly into the host cytoplasm . These type III secreted effectors ( T3SEs ) play an essential role in bacterial infection , mainly by targeting host immunity . However , the molecular basis of their functionalities remains largely enigmatic . Here , we show that the Pseudomonas syringae T3SE HopZ1a , a member of the widely distributed YopJ effector family , directly interacts with jasmonate ZIM-domain ( JAZ ) proteins through the conserved Jas domain in plant hosts . JAZs are transcription repressors of jasmonate ( JA ) -responsive genes and major components of the jasmonate receptor complex . Upon interaction , JAZs can be acetylated by HopZ1a through a putative acetyltransferase activity . Importantly , P . syringae producing the wild-type , but not a catalytic mutant of HopZ1a , promotes the degradation of HopZ1-interacting JAZs and activates JA signaling during bacterial infection . Furthermore , HopZ1a could partially rescue the virulence defect of a P . syringae mutant that lacks the production of coronatine , a JA-mimicking phytotoxin produced by a few P . syringae strains . These results highlight a novel example by which a bacterial effector directly manipulates the core regulators of phytohormone signaling to facilitate infection . The targeting of JAZ repressors by both coronatine toxin and HopZ1 effector suggests that the JA receptor complex is potentially a major hub of host targets for bacterial pathogens .
A prevailing concept for plant-pathogen interactions highlights the continuing battles between the activation of plant immune responses upon pathogen perception and the subversion of host immunity by virulence factors produced by successful pathogens . One branch of the plant immunity system is based on the recognition of pathogen- or microbe-associated molecular patterns ( PAMP/MAMPs ) , which leads to a signal transduction cascade , and eventually PAMP-triggered immunity ( PTI ) [1] . PTI , broadly referred as basal defense in plants , restricts the growth of the vast majority of potential pathogens encountered by plants in the surrounding environment [2] , [3] . However , successful pathogens produce virulence factors to effectively suppress PTI . For example , Gram-negative bacterial pathogens , such as Pseudomonas syringae , inject type III-secreted effectors ( T3SEs ) into the host cell to actively inhibit PTI [4] , [5] . As a counter-attack strategy , plants have evolved nucleotide-binding leucine-rich repeat ( NB-LRR ) proteins to perceive specific T3SEs , directly or indirectly , and elicit effector-triggered immunity ( ETI ) , which is often associated with localized programmed cell death at the infection sites [2] , [3] , [6] . Recent studies suggest that many P . syringae T3SEs suppress PTI and/or ETI by targeting important components of plant immunity [5] , [7] , [8] . Although the virulence targets of a few T3SEs have been characterized , the molecular mechanisms by which the majority of T3SEs subvert host resistance or facilitate nutrient acquisition remain elusive . HopZ1 is a P . syringae T3SE that belongs to the widely distributed YopJ family of cysteine proteases/acetyltransferases produced by both plant and animal bacterial pathogens [9] . The YopJ-like T3SEs share a conserved catalytic core , consisting of three key amino acid residues ( histidine , glutamic acid , and cysteine ) , which is identical to that of clan-CE ( C55-family ) cysteine proteases [10] . However , several members of the YopJ effector family have been shown to possess acetyltransferase activity . YopJ and VopA modify their target proteins ( mitogen-associated protein kinases and Ikkα/β ) in animal hosts and the acetylation of these host targets blocks their phosphorylation and the subsequent defense signal transduction [11] , [12] . PopP2 produced by the plant pathogen Ralstonia solanacerum has an autoacetylation activity , which is essential for its recognition in resistant plants; however , whether PopP2 can modify its target proteins in the host remains unknown [13] . Two functional HopZ1 alleles , HopZ1a and HopZ1b , have been identified in P . syringae [9] . HopZ1b is produced by P . syringae pv . glycinea ( Pgy ) strains , which are the causal agents of bacterial blight disease on soybean ( Glycine max ) [9] . HopZ1bPgyBR1 ( HopZ1b in Pgy strain BR1; hereafter HopZ1b ) promotes P . syringae multiplication in soybean; whereas the closely-related HopZ1aPsyA2 ( HopZ1a in P . syringae pv . syringae strain A2; hereafter HopZ1a ) triggers an HR in soybean cultivar Williams 82 and Arabidopsis thaliana accession Columbia-0 ( Col-0 , wild-type ) [14] . HopZ1 mutants with the catalytic cysteine residues ( C216 in HopZ1a or C212 in HopZ1b ) substituted by alanines lose the virulence function or the HR-triggering activity , indicating that the functions of HopZ1 alleles require their enzymatic activities [9] , [14] . In addition , HopZ1 has a potential N-terminal myristoylation site ( Gly2 ) which directs the proteins to the plasma membrane [14] , [15] . This myristoylation site of HopZ1a contributes to its avirulence function in both soybean and Arabidopsis [14] , [15] . However , it is not clear whether this myristoylation site is important for the virulence function of HopZ1a . HopZ1 exhibited weak cysteine protease activities in vitro [9] . Recent studies showed that HopZ1a also possessed an acetyltransferase activity and could use tubulin as a substrate in vitro . Modification of tubulin is associated with the disruption of microtubule cytoskeleton , which may contribute to bacterial pathogenesis [16] . To identify potential host targets of HopZ1 , we conducted yeast two-hybrid screening using a cDNA library of the natural host soybean and identified several HopZ1-interacting proteins ( ZINPs ) . ZINP1 ( 2-hydroxyisoflavanone dehydratase , GmHID1 ) is a key enzyme in the soybean isoflavone biosynthetic pathway and a positive regulator of soybean basal defense . HopZ1 induces the degradation of GmHID1 , and hence a decreased isoflavone production in soybean , resulting in increased plant susceptibility to bacterial infection [17] . HopZ1 also enhances bacterial infection in Arabidopsis , which does not have a putative ortholog of GmHID1 . To understand the mechanisms underlying the virulence function of HopZ1a in Arabidopsis , we characterized another family of ZINPs , which were identified as jasmonate ZIM-domain ( JAZ ) proteins . JAZs are key transcriptional repressors of the jasmonate ( JA ) signaling pathway and major components of the JA receptor complex [18] , [19] , [20] . JA plays an important role in regulating plant responses to biotic and abiotic stresses . Some P . syringae strains produce the JA-mimicking phytotoxin coronatine , which efficiently activates JA signaling to facilitate bacterial entry into plant apoplastic space and suppress defense [21] , [22] , [23] . Therefore , HopZ1a may also target the JAZ proteins to promote bacterial infection . Consistent to this hypothesis , HopZ1a was previously reported to induce the expression of the JA/ethylene marker gene AtPDF1 . 2 in Arabidopsis , indicating that it could activate JA/ethylene signaling [24] . Here , we report that HopZ1a directly interacts with JAZ proteins of soybean and Arabidopsis . We show that HopZ1a induces the degradation of AtJAZ1 , and promotes JA-responsive gene expression during P . syringae infection . Furthermore , HopZ1a functionally complements the growth deficiency of a P . syringae pv . tomato mutant that does not produce coronatine . All these activities depend on the intact catalytic core of HopZ1a , which acetylates JAZ proteins in vitro . Taken together , our results suggest that HopZ1a facilitates bacterial infection by manipulating the JA signaling pathway in Arabidopsis .
Using yeast two-hybrid screens , we identified the HopZ1a-interacting proteins ( ZINPs ) from a soybean cDNA library [17] . Among them , ZINP3 ( Gm7g04630 ) was interesting because it shows significant homology to the Jasmonate ZIM-domain ( JAZ ) proteins . We designated ZINP3 as GmJAZ1 because it is most similar ( 51% similarity in full-length amino acid sequences and 62% similarity in the ZIM and Jas domains ) to AtJAZ1 in Arabidopsis . GmJAZ1 was then further pursued as a direct target of HopZ1a . We first confirmed the physical interaction between HopZ1a and GmJAZ1 by in vitro pull-down using recombinant GST-HopZ1a and GmJAZ1-HA proteins over-expressed in E . coli . GST-HopZ1a or GST ( empty vector ) was purified from whole cell lysate using glutathione resins and then incubated with an equal amount of whole cell lysate of E . coli expressing GmJAZ1-HA . GST-HopZ1a-bound resins , but not GST-bound resins , provided enrichment of GmJAZ1-HA ( Fig . 1A ) , suggesting that HopZ1a interacted with GmJAZ1 in vitro . The catalytic mutant HopZ1a ( C216A ) also interacted with GmJAZ1 , similar to wild-type HopZ1a ( Fig . 1A ) . We next examined the sub-cellular localization of GmJAZ1 to determine whether it co-localizes with HopZ1a in plant cells . GmJAZ1-YFP was expressed in Nicotiana benthamiana using Agrobacterium-mediated transient expression . Yellow fluorescence was examined in the pavement cells of the infiltrated leaves at 48 hours post inoculation ( hpi ) using confocal microscopy . Fluorescence was detected both on the plasma membrane and in the nucleus ( Fig . S1 ) . Previous studies reported that HopZ1a mainly locates on the plasma membrane with a sub-pool of HopZ1a in the nucleus [17] . These results suggest that GmJAZ1 and HopZ1a could co-localize in plant cells . To further confirm that HopZ1a indeed enters the nucleus , we performed nuclear fractionation of N . benthamiana cells expressing HopZ1a ( C216A ) . The catalytic mutant HopZ1a ( C216A ) was used in this experiment , because the expression of the functional HopZ1a triggers cell death in N . benthamiana [9] , [14] . Consistent with the previous confocal microscopy data [17] , we detected the presence of HopZ1a ( C216A ) from both cytosolic and nuclear fractions ( Fig . S2 ) . These data confirmed that HopZ1a and GmJAZ1 co-localize in N . benthamiana cells . We further used the bimolecular fluorescence complementation ( BiFC ) assay to determine the interaction between HopZ1a and GmJAZ1 in planta . HopZ1a ( C216A ) and GmJAZ1 were fused to the nonfluorescent N-terminal domain of YFP ( 1–155 aa , nYFP ) and the C-terminal domain of YFP ( 156–239 aa , cYFP ) , respectively , at their C-termini . When the fusion genes were co-expressed in N . benthamiana , fluorescence was detected on the plasma membrane and in the nucleus ( Fig . 1B ) , consistent with the subcellular localization of GmJAZ1 and HopZ1a . Taken together , these experiments demonstrate the interaction of HopZ1a and GmJAZ1 in vitro and in planta . We have previously observed HopZ1-mediated degradation of another HopZ1-interacting protein GmHID1 when GmHID1 and HopZ1 were transiently co-expressed in N . benthamiana [17] . Therefore , we examined whether HopZ1a can also induce the degradation of GmJAZ1 . GmJAZ1-FLAG and HopZ1a-HA were co-expressed in N . benthamiana , and the abundance of GmJAZ1 was determined at 20 hpi before the onset of visible cell death symptoms , which usually starts at 30 hpi . We chose 20 hpi because the expression level of GmJAZ1 was too low for protein analysis at earlier time points . A significant reduction of GmJAZ1 protein level was observed in N . benthamiana leaves co-expressing wild-type HopZ1a-HA , compared to leaves expressing the HopZ1a catalytic mutant or infiltrated with Agrobacterium carrying the empty vector ( Fig . 2A ) . These results suggest that HopZ1a induces the degradation of GmJAZ1 in plant cells and the degradation requires the enzymatic activity of HopZ1a . Incubation of GmJAZ1 and HopZ1a proteins purified from E . coli did not lead to observable changes in the abundance of GmJAZ1 ( Fig . 1A ) . We suspected that a plant factor ( s ) might be required for this process and therefore performed a semi-in vitro degradation assay by incubating proteins extracted from N . benthamiana tissues expressing GmJAZ1 or HopZ1a individually . Total proteins extracted from leaves expressing GmJAZ1 or HopZ1a were mixed and incubated at 4°C for six hours before the abundance of GmJAZ1 was examined using western blots . Again , a significant decrease in GmJAZ1 protein level was observed in the presence of wild-type HopZ1a , but not the catalytic mutant HopZ1a ( C216A ) ( Fig . 2B ) . These data suggest that HopZ1a induces GmJAZ1 degradation in plant cells . To exclude the possibility that the reduced GmJAZ1 protein levels might have been resulted from cell death triggered by wild-type HopZ1a in N . benthamiana , we performed two control experiments . Firstly , we co-expressed the green fluorescence protein ( GFP ) with HopZ1a-HA or HopZ1a ( C216A ) -HA in N . benthamiana . The GFP protein levels remained unchanged in the presence of either wild-type or the catalytic mutant of HopZ1a ( Fig . S3A ) . Secondly , we performed the semi-in vitro degradation assay of GmJAZ1 using AvrRpt2 , which also elicits cell death in N . benthamiana [25] . Incubation with plant protein extracts expressing AvrRpt2 did not change the abundance of GmJAZ1 ( Fig . S3B ) . This suggests that the reduced abundance of GmJAZ1 was not a result of HopZ1a-induced cell death in N . benthamiana . Because GmJAZ1 is an ortholog of Arabidopsis JAZ proteins ( AtJAZs ) , we examined whether HopZ1a also targets AtJAZs . Arabidopsis produces twelve JAZ orthologs ( Fig . S4 ) . Among them , seven were tested for their interactions with HopZ1a using in vitro pull-down . Our data showed that AtJAZ2 , AtJAZ5 , AtJAZ6 , AtJAZ8 and AtJAZ12 interacted with HopZ1a in vitro ( Fig . 3A ) . Although AtJAZ1 shares the highest sequence similarity with GmJAZ1 ( Fig . S4 ) , the interaction of AtJAZ1 with HopZ1a could not be determined because we were unable to express AtJAZ1 in E . coli at a level suitable for the pull-down assay . We next confirmed the interaction between HopZ1a and AtJAZ6 in planta using BiFC . Similar to HopZ1a-GmJAZ1 interaction , yellow fluorescence was observed from plasma membrane and nucleus in cells co-expressing HopZ1a ( C216A ) -nYFP and AtJAZ6-cYFP ( Fig . 3B ) . AtJAZ6 by itself was mainly located in the nucleus , but could also be detected in cytosol ( Fig . S2 ) . These data suggest that HopZ1a and AtJAZ6 co-localize and interact in plant cells . Several effectors from the YopJ family , including HopZ1a , have been shown to possess acetyltransferase activities . To determine whether JAZs are substrates of HopZ1a , we performed in vitro enzymatic assay using C14-labeled acetyl-CoA . Recombinant HIS-SUMO-HopZ1a or HIS-SUMO-HopZ1a ( C216A ) proteins were expressed in E . coli and purified using nickel column . The HIS-SUMO tag was then removed by ubiquitin like protease 1 ( ULP1 ) . Tag-free HopZ1a or HopZ1a ( C216A ) proteins were incubated with purified HIS-GmJAZ1or MBP-AtJAZ6-HIS proteins in the presence of the cofactor inositol hexakisphosphate ( IP6 ) , and the acetylation of HopZ1a , GmJAZ1 and AtJAZ6 was detected by autoradiography as previously described [26] . Our experiments showed that both GmJAZ1 ( Fig . 4A ) and AtJAZ6 ( Fig . 4B ) were acetylated by wild-type HopZ1a , which also exhibited autoacetylation . The acetylation of GmJAZ1 appeared to be weaker in the autoradiograph compared to that of AtJAZ6 . This is due to the low expression level of GmJAZ1 in E . coli , which only allowed us to use a much lower amount ( 1 µg ) , compared to AtJAZ6 ( 10 µg ) in the reactions . Nonetheless , we consistently detected the acetylated form of GmJAZ1 when it was incubated with HopZ1a , but not HopZ1a ( C216A ) , suggesting that GmJAZ1 and AtJAZ6 are both substrates of HopZ1a . We sometimes could observe a background level of acetylation in tagged AtJAZ6 ( MBP-AtJAZ6-HIS ) when it was incubated with HopZ1a ( C216A ) . Although this background acetylation was very weak compared to the acetylation of MBP-AtJAZ6-HIS , we decided to use the tag-free AtJAZ6 proteins to further confirm its acetylation by HopZ1a . Again , we observed strong acetylation of AtJAZ6 by HopZ1a , but not by HopZ1a ( C216A ) using only 5 µg of AtJAZ6 in the reaction ( Fig . S5 ) . These results demonstrate that GmJAZ1 and AtJAZ6 are acetylation substrates of HopZ1a . JAZ proteins share three conserved domains: the C-terminal Jas motif [27] , the ZIM domain in the central region [28] , and a weakly conserved N-terminal region [19] . Because the conserved Jas domain is essential for the instability of JAZs in response to JA and the JA-mimicking phytotoxin coronatine [18] , [19] , we examined the impact of the Jas domain in the interaction between JAZs and HopZ1a . We constructed the mutant AtJAZ6ΔJas by deleting ten highly conserved amino acids ( from seine191 to lysine200 ) within the Jas domain . In vitro pull-down assay showed that AtJAZ6ΔJas did not bind HopZ1a ( Fig . 4C ) . Furthermore , AtJAZ6ΔJas was not acetylated by HopZ1a in vitro ( Fig . 4D ) or degraded by HopZ1a when these two proteins were co-expressed in N . benthamiana ( Fig . 4E ) . These results demonstrate that HopZ1a-induced JAZ degradation requires direct interaction of HopZ1a with AtJAZ6 , which is mediated by the Jas domain . Although we observed the degradation of GmJAZ1 and AtJAZ6 when they were co-expressed with HopZ1a in N . benthamiana , it is important to examine whether HopZ1a can promote JAZ degradation during bacterial infection . For this purpose , we inoculated transgenic Arabidopsis plants expressing 35S-HA-AtJAZ1 with P . syringae producing HopZ1a or HopZ1a ( C216A ) . The Arabidopsis pathogen Pseudomonas syringae pv . tomato strain DC3000 ( PtoDC3000 ) is well-known to induce AtJAZ degradation through the production of coronatine , which acts as a JA mimic [22] . The mutant PtoDC3118 is deficient in coronatine production and therefore no longer degrades JAZs [29] . Importantly , PtoDC3118 expressing HopZ1a from its native promoter also significantly reduced the abundance of AtJAZ1 at 6 hpi ( Fig . 5A ) . The level of AtJAZ1 remained unchanged in tissues infiltrated with PtoDC3118 carrying the empty vector or expressing the catalytic mutant HopZ1a ( C216A ) . These data strongly suggest that HopZ1a can induce AtJAZ1 degradation during bacterial infection . Because HopZ1a elicits HR in Arabidopsis ecotype Col-0 , we performed two experiments to exclude the possibility that HopZ1a-triggered AtJAZ1 degradation was a result of plant cell death . First , we examined whether another effector AvrRpt2 could induce AtJAZ1 degradation . Although AvrRpt2 also triggers HR in Arabidopsis Col-0 , the abundance of AtJAZ1 was unchanged when the HA-AtJAZ1-expressing plants were inoculated with PtoDC3118 expressing AvrRpt2 ( Fig . 5A ) . Next , we generated the transgenic Arabidopsis line expressing 35S-HA-AtJAZ1 in the zar1-1 mutant background , which is abrogated in HopZ1a-triggered HR [30] . Again , the AtJAZ1 protein level was significantly reduced by HopZ1a ( Fig . 5B ) , confirming that HopZ1a delivered by P . syringae leads to AtJAZ1 degradation in a cell death independent manner . A major regulatory mechanism of JAZs in the presence of JA or coronatine is through COI-dependent ubiquitin-proteasome degradation . COI1 is an F-box protein that determines the substrate specificity of a Skp/Cullin/F-box ( SCF ) E3 ubiquitin ligase-SCFCOI1 [31] . Together with JAZ , COI1 is also a critical component of the JA co-receptor complex [19] , [20] , [22] . We examined whether COI1 is required for HopZ1a-induced JAZ degradation using the transgenic Arabidopsis line expressing 35S-HA-AtJAZ1 in the coi1-30 ( SALK_035548 ) mutant background . As expected , PtoDC3000 , which induces JAZ degradation through coronatine production , was unable to reduce the abundance of AtJAZ1 in the absent of COI1 . Interestingly , PtoDC3118 expressing HopZ1a also no longer induced the degradation of AtJAZ1 in the coi1 mutant plants ( Fig . 5C ) . These data suggest that , similar to coronatine- and JA-mediated AtJAZ degradation , COI1 is required for the degradation of AtJAZ1 by HopZ1a . In Arabidopsis , JAZ proteins are repressors of JA transcription factors ( e . g . AtMYC2 ) that are involved in the expression of JA-responsive genes [32] , [33] , [34] . Since HopZ1a induces the degradation of AtJAZ1 , we examined whether it could induce the expression of JA-responsive genes during bacterial infection . Real-time RT-PCR was carried out to determine the transcript levels of JA-responsive genes in Arabidopsis . Five-week old zar1-1 plants were inoculated with PtoDC3118 expressing HopZ1a or HopZ1a ( C216A ) at OD600 = 0 . 2 ( approximately 2×108 cfu/mL ) . The transcript levels of two early JA-responsive genes , AtJAZ9 and AtJAZ10 [34] , were analyzed at 6 hpi . Both genes were induced approximately ten fold in plants infected by PtoDC3118 ( HopZ1a ) , whereas their expression was not changed in tissues infected by PtoDC3118 expressing HopZ1a ( C216A ) ( Fig . 6A ) . The level of gene induction by HopZ1a was lower than that by coronatine , as shown by the approximately 40-fold induction of AtJAZ9 and AtJAZ10 in plants infected with PtoDC3000 . This is consistent with the partial vs . complete degradation of AtJAZ1 by HopZ1a or coronatine during bacterial infection . Nonetheless , these experiments suggest that bacterium-delivered HopZ1a can activate JA signaling . Recent findings showed that coronatine can suppress salicylic acid ( SA ) accumulation , probably as a consequence of the activation of JA signaling [35] . Because SA-associated defense confers resistance against biotrophic and hemibiotrophic pathogens , reduced SA accumulation would lead to defense suppression . In particular , coronatine is able to repress the expression of the SA synthetic enzyme isochorismate synthase gene 1 ( AtICS1 ) in Arabidopsis . We then examined the impact of HopZ1a on the expression of AtICS1 . Arabidopsis zar1-1 plants were inoculated with PtoDC3000 or PtoDC3118 carrying empty vector , HopZ1a , or HopZ1a ( C216A ) at OD600 = 0 . 2 ( approximately 2×108 cfu/mL ) . Consistent with the prior findings , the transcript abundance of AtICS1 was reduced in plants infected with PtoDC3000 when compared to PtoDC3118 carrying the empty vector or HopZ1a ( C216A ) ( Fig . 6B ) . The expression of AtICS1 was also reduced in plants inoculated with PtoDC3118 expressing HopZ1a , to a similar level as that in leaves inoculated with PtoDC3000 . These data confirmed that , like coronatine , HopZ1a activates JA signaling and represses SA accumulation during bacterial infection . Coronatine facilitates the infection of PtoDC3000 by activating JA signaling in Arabidopsis [21] . The coronatine-deficient mutant PtoDC3118 exhibits a significant reduction in bacterial population especially when the plants are infected by dipping inoculation [36] . Since HopZ1a also activates JA signaling , we examined whether HopZ1a could complement the growth deficiency of PtoDC3118 . The Arabidopsis zar1-1 mutant plants were dipping-inoculated by PtoDC3000 or PtoDC3118 carrying the empty vector , HopZ1a , or HopZ1a ( C216A ) . Three days post infection ( dpi ) , the bacterial populations of PtoDC3118 carrying the empty vector or expressing HopZ1a ( C216A ) were approximately 200 fold lower than that of PtoDC3000 ( Fig . 7A ) . Importantly , PtoDC3118 expressing wild-type HopZ1a multiplied to a significantly higher level ( about 10 fold ) than PtoDC3118 or PtoDC3118 expressing HopZ1a ( C216A ) ( Fig . 7A ) . Although the population of PtoDC3118 ( HopZ1a ) is lower than that of PtoDC3000 , this partial complementation of the growth deficiency of PtoDC3118 is consistent with the partial degradation of AtJAZ1 ( Fig . 5A ) and the lower levels of JA-responsive gene induction ( Fig . 6A ) by PtoDC3118 ( HopZ1a ) compared to PtoDC3000 . To further confirm that the function of HopZ1a is specifically related to its ability to activate the JA pathway , we introduced HopZ1a into PtoDC3000 and performed the same bacterial growth assay . HopZ1a was previously shown to enhance the infection of PtoDC3000 [30] . However , despite numerous trials , we did not observe any enhancement of HopZ1a on in planta multiplication of PtoDC3000 . In fact , we consistently detected a decrease in the population of PtoDC3000 ( HopZ1a ) compared to PtoDC3000 carrying the empty vector ( Fig . 7A ) . Thus , HopZ1a can partially substitute for coronatine to promote bacterial infection . Because COI1 is required for HopZ1a-induced degradation of AtJAZ1 , we then examined whether COI1 is also required for the virulence activity of HopZ1a in Arabidopsis . For this purpose , we generated coi1-1 , zar1-1 double mutant Arabidopsis plants , which were inoculated by PtoDC3118 carrying the empty vector , HopZ1a or HopZ1a ( C216A ) . The bacterial populations of these three strains remained the same ( Fig . 7B ) , confirming that the virulence activity of HopZ1a requires COI1 . To further confirm that HopZ1a was unable to activate JA signaling in the mutant plants , we also determined the transcript levels of the JA-responsive genes AtJAZ9 and AtJAZ10 , as well as the SA-biosynthetic gene AtICS1 after bacterial inoculation . Similar to PtoDC3000 , PtoDC3118 expressing HopZ1a was also unable to induce the expression of AtJAZ9 and AtJAZ10 or suppress the expression of AtICS1 ( Fig . S6 ) . These data suggest that both the phytotoxin coronatine and the effector HopZ1a activate JA signaling in a COI1-dependant manner . HopZ1a contains a potential N-terminal myristoylation site ( Gly2 ) which may direct the association of HopZ1a with plasma membrane in plant cells [14] , [15] . We therefore examined whether the Gly2 residue was required for the virulence function of HopZ1a . The bacterial population of PtoDC3118 carrying HopZ1a ( G2A ) was similar to that of PtoD3118 carrying wild-type HopZ1a in Arabidopsis zar1-1 plants ( Fig . S7 ) . This result demonstrates that HopZ1a ( G2A ) retained the virulence activity to promote PtoD3118 multiplication in Arabidopsis , indicating that the potential myristoylation site is not required for the virulence function of HopZ1a . Taken together , our experiments showed that HopZ1a enhanced P . syringae infection in Arabidopsis in a manner similar to coronatine , which is a potent activator of JA signaling during bacterial infection .
T3SEs manipulate a variety of cellular processes in eukaryotic hosts for the benefit of pathogen infection . Emerging data suggest that bacterial pathogens have evolved various effectors to manipulate the signaling of JA and SA , which are important plant hormones that regulate defense response [37] . SA-dependent defense plays a major role in plant immunity against biotrophic and hemibiotrophic pathogens , such as Hyaloperonospora arabidopsidis and P . syringae [35] , [37] . The P . syringae effector HopI1 directly targets Hsp70 in choloroplasts to suppress SA accumulation and thereby facilitate bacterial infection [38] . In addition , the Xanthomonas campestris effector XopD was also shown to repress SA signaling during bacterial infection of tomato [39] , [40] , [41] . Here , we report that the P . syringae effector HopZ1a represses SA accumulation , probably as an indirect effect of the activation of JA signaling . In this study , we report that HopZ1 directly targets JAZs , the key negative regulators of JA signaling [18] , [19] . Because JA signaling pathway is antagonistic to SA-dependent defense , activating JA signaling would be an attractive bacterial strategy to suppress host defense and facilitate pathogenesis of these pathogens . Indeed , recent studies have shown a remarkable example in which the P . syringae phytotoxin coronatine structurally mimics the active form of JA and targets the JAZ repressors for degradation to efficiently activate JA signaling [20] , [22] . However , it has been rather puzzling that only a few P . syringae strains produce coronatine [42] . A previous study showed that a T3SE , AvrB , was also able to promote JA signaling , apparently through an indirect mechanism via the activation of MPK4 [43] , [44] . We show here that the effector HopZ1a directly interacts with JAZs and at least some JAZs can be used by HopZ1a as substrates for acetylation . Importantly , HopZ1 mediates the degradation of AtJAZ1 in Arabidopsis and promotes bacterial infection in a COI1-dependent manner . This new finding raises the exciting possibility that JAZ repressors ( hence the JA receptor complex ) may be a common hub of host targets for diverse bacterial virulence factors . Consistent with this notion , oomycete pathogens were also found to produce effectors that interact with AtJAZ3 [45] . These pieces of evidence suggest that the JA receptor complex might be the Achilles' heel in plant defense system during the arms race with microbial pathogens . HopZ1a enhances the in planta multiplication of PtoDC3118 , but not that of PtoDC3000 , in Arabidopsis . A weak growth enhancement of PtoDC3000 by HopZ1a was reported previously [30] . We were unable to replicate the published data , probably due to differences in experimental conditions . Our experiments , including the JA-responsive gene expression , JAZ protein degradation and bacterial in planta multiplication , consistently suggest that HopZ1a activates the JA signaling pathway in a manner similar to coronatine . However , HopZ1a only partially complements the function of coronatine . This could be because HopZ1a is not as potent as coronatine in inducing the degradation of JAZs . Coronatine has dual functions during the pre-entry and post-entry stages of bacterial infection [23] , whereas the type III secretion genes are generally believed to be expressed after the bacteria have entered the apoplast [46] . Although it remains to be determined whether HopZ1a could promote stomata opening at the pre-entry stage , it is possible that HopZ1a mainly substitutes coronatine function inside the plant tissues . In addition , proteins might not be as stable as metabolites in planta , which may also explain the partial complementation of HopZ1a for the virulence deficiency of the coronatine mutant PtoDC3118 . As transcription regulators , JAZs are believed to function in the nucleus . However , HopZ1a was previously shown to mainly locate on the plasma membrane and this localization was mediated by a potential myristoylation site in the N-terminus [15] . Our protein-protein interaction and localization analyses showed that HopZ1a is also located in the nucleus and it interacts with JAZs both in the nucleus and on the cytosol/plasma membrane . Importantly , the mutant HopZ1a ( G2A ) , which is abolished for its localization on the plasma membrane , was still able to promote PtoDC3118 infection . These data demonstrate that the membrane localization of HopZ1a is only important for host recognition [14] , [15] , but not required for virulence activity . This is consistent with the primary localization of JAZs in the nucleus as transcription repressors . YopJ-like T3SEs produced by plant pathogens appear to have various enzymatic activities . AvrXv4 of Xanthomonas campestris was reported to be a SUMO protease [47] . AvrBsT , also from Xanthomonas , exhibited a weak cysteine protease activity in vitro [48] . PopP2 in R . solanacearum has autoacetylation and trans-acetylation activities in vitro , but it does not seem to acetylate its host target proteins [13] . Recently , HopZ1a was demonstrated to acetylate tubulin [16] . Our experiments showed that GmJAZ1 and AtJAZ6 are also substrates of HopZ1a . Importantly , we found that HopZ1a induces the degradation of AtJAZ1 during bacterial infection . In the presence of the active form of JA or coronatine , the F-box protein COI1 interacts with JAZs via the C-terminus Jas motif and recruits JAZs to the 26S proteasomes for degradation [18] , [19] , [22] . The fact that HopZ1a-mediated AtJAZ1 degradation is dependent on COI1 suggests that this degradation could also be dependent on the 26S proteasomes as a consequence of JAZ modification by HopZ1a . Post-translational modifications , including acetylation , have been shown to induce or repress proteasomal degradation . For example , in mammalian cells , the acetyltransferase ARD1 acetylates Hypoxia-inducible factor 1α ( HIF-1α ) , which promotes its ubiquitination and proteasomal degradation [49] . Further investigations are needed to determine how HopZ1a-mediated acetylation of JAZs could facilitate COI1-dependent degradation of JAZ repressors .
Pseudomonas syringae , Agrobacterium tumefaciens and Escherichia coli strains were grown as described [50] . Bacteria strains and plasmids used in this study are summarized in Table S1 . To construct plasmids for bimolecular fluorescence complementation ( BiFC ) assay , full-length cDNA of GmJAZ1 or AtJAZ6 and hopZ1a ( C216A ) were cloned into the vectors pSPYCE and pSPYNE [51] , respectively . To examine the subcellular localization of GmJAZ1 , full-length cDNA was cloned into the vector pEG101[52] . The recombinant plasmids were introduced into Agrobacterium tumefaciens strain C58C1 ( pCH32 ) , which were then used to infiltrate 3–4 week old Nicotiana benthamiana plants using a protocol described previously [14] . Functional fluorophore were visualized in the infiltrated leaves using a Leica SP2 Laser Scanning Confocal Microscope ( Leica Microsystems ) at 48 hpi for subcellular localization and BiFC . DAPI was used to stain the nucleus in plant cells [53] , [54] . To construct GST-fusion plasmids , the full-length hopZ1a gene was inserted into the vector pGEX4T-2 ( GE Healthcare Life Science ) . GmJAZ1-HA gene was cloned into the vector pET14b ( Navogen ) , which has 6×His in the N-terminus . The AtJAZ genes were cloned into the plasmid vector pET-Mal with maltose binding protein ( MBP ) in the N-terminus and 8×His in the C-terminus [55] . In vitro pull-down assays were carried out using GST pull-down protein∶protein interaction kit ( Pierce ) according to the manufacturer's instruction . Briefly , GST or GST-HopZ1a was expressed in E . coli strain BL21 ( DE3 ) . Soluble proteins were incubated with 50 µL glutathione agarose beads ( Invitrogen ) for one hour at 4°C . The beads were washed ( 20 mM Tris-HCl ( PH 7 . 5 ) , 150 mM KCl , 0 . 1 mM EDTA and 0 . 05% Triton X-100 ) five times and then incubated with equal amount of bacterial lysates containing JAZ proteins at 4°C for overnight . The beads were washed five times again , and the presence of the JAZ proteins on the beads was detected by western blots using anti-HA or anti-His antibodies conjugated with horseradish peroxidase ( HRP ) ) ( Santa Cruz Biotechnology Inc . ) . Pseudomonas syringae expressing the hopZ1a-HA genes was induced in M63 minimal medium containing 1% fructose at room temperature overnight [50] . HopZ1 expression was detected by western blots using the anti-HA antibody . For JAZ degradation assay in N . benthamiana , hopZ1a-HA or 3×FLAG-hopZ1a were co-expressed with GmJAZ1-FLAG or AtJAZ6-YFP-HA using Agrobacterium-mediated transient expression as previously described [9] , [56] . Leaf disks were collected at 20 hpi , and then grounded in 2×Laemmli buffer . The abundances of GmJAZ1 and AtJAZ6 were analyzed by western blots . For the semi-in vitro protein degradation assay , GmJAZ1-FLAG and 3×FLAG-HopZ1a were over-expressed individually in N . benthamiana using Agrobacterium-mediated transient expression . Total proteins were extracted from the infected leaf tissues at 20 hpi using an extraction buffer containing 200 mM NaCl , 50 mM Tris ( pH 7 . 6 ) , 10% Glycerol , 0 . 1% NP-40 , protease inhibitor cocktail ( Roche ) , 10 mM DTT , 1 mM PMSF . Protein extracts were mixed in equal volume for six hours at 4°C with gentle agitation before the abundance of GmJAZ1 was examined by western blots . For the in vivo JAZ degradation assay , six-week-old 35S-HA-AtJAZ1 transgenic Arabidopsis plants were hand infiltrated with bacterial suspensions of PtoDC3000 or PtoDC3118 carrying the empty vector ( EV ) , expressing HopZ1a , HopZ1a ( C216A ) or AvrRpt2 at OD600 = 0 . 2 ( approximately 2×108 cfu/mL ) . Leaves infiltrated with water were used as a negative control . Six hours post inoculation , total proteins were extracted from four leaf discs ( 0 . 5 cm2 ) in 100 µL of 2×SDS sample buffer . The lysates were incubated at 95°C for 10 minutes followed by centrifugation at 14 , 000 rpm for 5 minutes . The abundance of AtJAZ1 was then analyzed by western blots . Homozygous coi1-30 mutant plants were selected on 1/2×Murashige & Skoog ( MS ) medium supplemented with 50 µM JA . Seedlings that were insensitive to JA treatment , i . e . without inhibited root growth symptoms , were transplanted in soil and infected with P . syringae after six weeks . HIS-GmJAZ1 , HIS-SUMO-HopZ1a , HIS-SUMO-HopZ1a ( C216A ) , HIS-SUMO-AtJAZ6 , MBP-AtJAZ6-HIS , and MBP-AtJAZ6ΔJas-HIS were over-expressed in the E . coli strain BL21 ( DE3 ) and then purified using nickel resins . HIS-SUMO-HopZ1a and HIS-SUMO-AtJAZ6 proteins were then cleaved by ULP1 protease , producing protein mixtures with both HIS-SUMO and either tag-free HopZ1a or AtJAZ6 . The protein mixtures were incubated with nickel resin again and the tag-free HopZ1a or AtJAZ6 proteins were collected from the flow through . In a standard acetylation assay , 2 µg purified HopZ1a or HopZ1a ( C216A ) was incubated with 10 µg MBP-AtJAZ6 , 5 µg AtJAZ6 or 1 µg GmJAZ1 at 30°C for one hour in 25 µL of reaction buffer containing 50 mM HEPES ( pH 8 . 0 ) , 10% glycerol , 1 mM DTT , 1 mM PMSF , 10 mM sodium butyrate , 1 µL [14C]-acetyl-CoA ( 55 µci/µmol , ) and 100 nM IP6 . The reaction mixtures were then subjected to SDS-PAGE and acetylated proteins were detected by autoradiography as previously described [11] , [26] , [57] after exposure at −80°C for five days . After autoradiograph , the protein gels were removed from the filter paper and stained with Coomassie blue as a loading control . The transcript abundances of AtJAZ9 , AtJAZ10 or AtICS1 in Arabidopsis leaf tissues were analyzed by real-time RT-PCR using SYBR Green Supermix ( BioRad Laboratories ) and an CFX96 Real-Time PCR Detection System ( BioRad Laboratories ) . Total RNA was isolated from three independent biological replicates using Trizol , and DNA was removed using DNase I ( Fermentas ) . Reverse transcription was performed using M-MLV Reverse Transcriptase ( Promega ) with 1 µg of total RNA in a 25 µL reaction . The cDNAs were then used as templates for real-time PCR using gene-specific primers , which are listed below . AtActin was used as internal standard when compared the expression of AtJAZ9 and AtJAZ10 in different treatment . AtUBQ5 was the internal control used for the normalization of AtICS1 expression . AtActin: 5′-GGTGTCATGGTTGGTATGGGTC-3′ and 5′-CCTCTGTGAGTAGAACTGGGTGC-3′ AtJAZ9: 5′-ATGAGGTTAACGATGATGCTG-3′ and 5′-CTTAGCCTCCTGGAAATCTG-3′ AtJAZ10: 5′-GTAGTTTCCGAGATATTCAAGGTG-3′ and 5′-GAACCGAACGAGATTTAGCC-3′ AtUBQ5:5′-GACGCTTCATCTCGTCC-3′ and 5′-GTAAACGTAGGTGAGTCC-3′ AtICS1: 5′-GGCAGGGAGACTTACG-3′ and 5′-AGGTCCCGCATACATT-3′ Arabidopsis plants were planted as previous described [30] , [36] . The leaves of five-week old plants were dipped into the bacterial suspensions at an OD600 = 0 . 2 ( approximately 2×108 cfu/mL ) for 15 seconds . The inoculated plants were then transferred to a growth chamber ( 22°C and 16/8 light/dark regime , 90% humidity ) . Bacterial populations were determined as colony forming units ( cfu ) per cm2 using a previously described procedure [50] . Statistical analyses were performed using JMP 8 . 0 ( SAS Institute Inc . ) .
|
Many Gram-negative bacterial pathogens rely on the type III secretion system , which is a specialized protein secretion apparatus , to inject virulence proteins , called effectors , into the host cells . The type III secreted effectors ( T3SEs ) directly target host substrates in order to promote bacterial colonization and disease development . Therefore , the identification and characterization of the direct host targets of T3SEs provides important insights into virulence strategies employed by bacterial pathogens to cause diseases . Here , we report that the plant pathogen Pseudomonas syringae T3SE HopZ1a physically interacts with and modifies the jasmonate ZIM-domain ( JAZ ) proteins in plant hosts . JAZ proteins are components of the receptor complex of the plant hormone jasmonates ( JA ) and key transcription repressors regulating JA-responsive genes . HopZ1a belongs to the widely distributed YopJ ( for Yersinia Outer Protein J ) family of T3SEs with a potential acetyltransferase activity . P . syringae producing HopZ1a , but not the catalytic mutant , leads to the degradation of AtJAZ1 during infection . As a result , HopZ1a activates JA signaling and promotes bacterial multiplication in Arabidopsis . This work provides the first example of a bacterial effector that subverts host immunity by directly targeting the receptor complex of a defense-associated hormone in plants .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Bacterial Effector Activates Jasmonate Signaling by Directly Targeting JAZ Transcriptional Repressors
|
Oseltamivir is relied upon worldwide as the drug of choice for the treatment of human influenza infection . Surveillance for oseltamivir resistance is routinely performed to ensure the ongoing efficacy of oseltamivir against circulating viruses . Since the emergence of the pandemic 2009 A ( H1N1 ) influenza virus ( A ( H1N1 ) pdm09 ) , the proportion of A ( H1N1 ) pdm09 viruses that are oseltamivir resistant ( OR ) has generally been low . However , a cluster of OR A ( H1N1 ) pdm09 viruses , encoding the neuraminidase ( NA ) H275Y oseltamivir resistance mutation , was detected in Australia in 2011 amongst community patients that had not been treated with oseltamivir . Here we combine a competitive mixtures ferret model of influenza infection with a mathematical model to assess the fitness , both within and between hosts , of recent OR A ( H1N1 ) pdm09 viruses . In conjunction with data from in vitro analyses of NA expression and activity we demonstrate that contemporary A ( H1N1 ) pdm09 viruses are now more capable of acquiring H275Y without compromising their fitness , than earlier A ( H1N1 ) pdm09 viruses circulating in 2009 . Furthermore , using reverse engineered viruses we demonstrate that a pair of permissive secondary NA mutations , V241I and N369K , confers robust fitness on recent H275Y A ( H1N1 ) pdm09 viruses , which correlated with enhanced surface expression and enzymatic activity of the A ( H1N1 ) pdm09 NA protein . These permissive mutations first emerged in 2010 and are now present in almost all circulating A ( H1N1 ) pdm09 viruses . Our findings suggest that recent A ( H1N1 ) pdm09 viruses are now more permissive to the acquisition of H275Y than earlier A ( H1N1 ) pdm09 viruses , increasing the risk that OR A ( H1N1 ) pdm09 will emerge and spread worldwide .
The influenza NA inhibitor antiviral drug oseltamivir is a key element of public health defences against influenza , and was used during the early stages of the A ( H1N1 ) pdm09 influenza pandemic to lessen the burden of disease in infected patients [1] , [2] . Resistance to oseltamivir most commonly results from mutations in the NA protein . The most common oseltamivir resistance ( OR ) mutation detected in A/H1N1 viruses is the NA H275Y mutation . Prior to 2007 , the incidence of OR influenza viruses was generally low ( <1% ) [3]–[7] . In vitro and in vivo virological studies demonstrated that OR seasonal A ( H1N1 ) viruses had attenuated viral replication kinetics in cell culture , mice and ferrets [8]–[10] , and therefore were considered to pose only a minimal threat to public health [8] . However in 2008 , OR ( H275Y ) seasonal A ( H1N1 ) viruses emerged and spread globally within 12 months , in the absence of oseltamivir selection pressure [11]–[14] , clearly demonstrating that the fitness of H275Y seasonal A ( H1N1 ) viruses was no longer compromised by the resistance mutation . Subsequent investigations revealed the presence of several “permissive” mutations ( R222Q , V234M , and possibly D354G ) in the NA of 2008–2009 seasonal A ( H1N1 ) viruses that enabled the acquisition of H275Y without compromising viral fitness [15]–[17] . In 2009 , the OR seasonal A ( H1N1 ) virus was replaced by the oseltamivir-sensitive ( OS ) ( NA 275H ) A ( H1N1 ) pdm09 virus . Since its emergence , there has been a concern that the same NA H275Y mutation may also become fixed within circulating A ( H1N1 ) pdm09 viruses . Since 2009 , virological surveillance has reported that the proportion of OR A ( H1N1 ) pdm09 viruses encoding the NA H275Y mutation has remained around 1% globally , and for the first two years following its emergence only limited sporadic transmissions of H275Y A ( H1N1 ) pdm09 viruses were reported between individuals in closed or near-contact settings [18]–[22] . However , in the United States [23] , United Kingdom [24] and Australia [25] during 2011 there was a notable increase in the detection of OR A ( H1N1 ) pdm09 viruses amongst community patients who had not received oseltamivir treatment . The largest cluster of cases occurred in 2011 around the city of Newcastle , within the Hunter New England region of Australia ( subsequently referred to as HNE2011 ) , where 15% of the A ( H1N1 ) pdm09 viruses collected between May and September 2011 were OR , including a peak frequency of 24% in July [26] . Genetic analysis revealed that these viruses were virtually identical , suggesting emergence from a single source [26] . Epidemiological investigations revealed that this OR virus had spread in the near absence of oseltamivir treatment , prompting concern that these A ( H1N1 ) pdm09 viruses may have obtained the capability to acquire the NA H275Y mutation without compromising viral fitness , much as seasonal A ( H1N1 ) viruses had done previously . Experiments with early A ( H1N1 ) pdm09 viruses ( from 2009 ) demonstrated that introduction of the H275Y mutation decreased total NA activity , largely by decreasing NA expression levels [15] . Therefore enhancing the total NA activity of viruses containing H275Y is likely to be a key factor required for the efficient replication and transmission of OR A ( H1N1 ) pdm09 viruses . Previously , we used a computational analysis to predict that two NA mutations present in a large number of A ( H1N1 ) pdm09 viruses sampled during 2010–2011 ( V241I and N369K ) and a third NA mutation which , although absent from the majority of A ( H1N1 ) pdm09 viruses , was present in viruses from the HNE2011 cluster ( N386S ) , could potentially offset the deleterious effect of H275Y upon the A ( H1N1 ) pdm09 NA [26] . Subsequent analysis of more recent A ( H1N1 ) pdm09 NA sequences submitted to GISAID and GenBank since mid 2012 revealed that V241I and N369K are now present in virtually all ( >99% ) of globally circulating A ( H1N1 ) pdm09 viruses , whereas the N386S NA mutation has not been maintained in contemporary A ( H1N1 ) pdm09 viruses ( Figure S1 ) . Here we combine our ferret competitive mixtures model [27] and a series of in vitro assays with quantitative modelling to assess the relative fitness of A ( H1N1 ) pdm09 viruses bearing these potentially permissive mutations ( PPMs ) and demonstrate that two of these mutations do indeed influence the fitness of OR A ( H1N1 ) pdm09 influenza viruses .
All experiments involving ferrets were conducted with the approval of the CSL Limited/Pfizer Animal Ethics Committee ( permit numbers 791 and 801 ) . All procedures were conducted according to the guidelines of the National Health and Medical Research Council as described in the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes [28] . All of the viruses described were isolated at the WHO Collaborating Centre for Reference and Research on Influenza , Melbourne , Australia , through routine virus sampling as part of the WHO Global Influenza Surveillance and Response System ( GISRS ) . The first ferret experiment used a “natural” ( non-reverse genetics derived ) virus pair . One of the viruses used was obtained from the cluster of OR cases that occurred during the HNE2011 outbreak [25] , [26]: A/Newcastle/17/2011 ( NA 275Y ) ( New17 OR ) ( for GISAID accession numbers see Table S1 ) . The fitness of this virus was compared against an OS virus which was obtained from the same location and time period: A/Newcastle/163/2011 ( NA 275H ) ( New163 OS ) . Full genome sequence analysis revealed a high degree of amino acid similarity between this pair of viruses ( New17 OR:New163 OS = 99 . 5% ) and no amino acid differences that have previously been associated with virulence , receptor binding , antiviral resistance or other aspects of viral function , other than the OR NA H275Y mutation . For experiments using reverse-engineered virus pairs , denoted with the prefix rg ( such as rgNew17 ) , RNA was extracted and cloned from two OR A ( H1N1 ) pdm09 viruses: the New17 OR virus from the HNE2011 outbreak , and A/Perth/261/2009 ( Perth261 OR ) , an OR A ( H1N1 ) pdm09 virus isolated from a hospitalised patient undergoing oseltamivir treatment during the early period of the 2009 pandemic . The NA segment of the early A ( H1N1 ) pdm09 virus A/California/7/2009 ( Cal7 OS ) was also used for an in vitro assessment of NA expression and enzymatic activity . All growth of viruses in vitro was performed in Madin-Darby canine kidney ( MDCK ) cells ( ATCC , #CCL-34 ) , using maintenance media [Dulbecco's modified Eagle's medium ( DMEM ) ( Sigma-Aldrich ) , supplemented with 2 mM L-glutamine ( SAFC Biosciences ) , 1 M HEPES ( SAFC Biosciences ) , 1% nonessential amino acids ( NE ) ( SAFC Biosciences ) , 0 . 05% sodium bicarbonate ( SAFC Biosciences ) , 2% penicillin-streptomycin ( Sigma-Aldrich ) and 4 μg/ml TPCK trypsin ( Sigma-Aldrich ) ] . Following initial isolation , viruses were subjected to two rounds of plaque purification to achieve a homogenous population , as previously described [27] . Thereafter viruses were amplified once at 35°C for 48 h , and the supernatant was frozen in aliquots and stored at −80°C for use in ferret infection experiments . The infectious titre of each stock was determined by titration on MDCK monolayers and the 50% tissue culture infectious dose ( TCID50 ) /ml was calculated according to the method of Reed and Muench [29] . RNA was extracted from cell culture supernatants as previously described [30] . Each influenza genome segment was amplified by RT-PCR and subjected to DNA sequencing as described by Hurt et al . [27] . DNA sequences obtained were translated and the deduced amino acid sequences aligned using Clone Manager v9 . 11 ( Scientific & Educational Software ) . All eight genome segments of the New17 OR and Perth261 OR viruses were amplified by RT-PCR and incorporated into the pHW2000 reverse genetics ( rg ) virus rescue plasmid [31] ( kindly provided by Dr . Richard Webby , St Jude Children's Research Hospital , Memphis , USA ) . In order to investigate the effect of the NA PPMs upon viral fitness , plasmids encoding the New17 OR and Perth261 OR NA genome segments were subjected to site-directed mutagenesis using the GeneArt Site-Directed Mutagenesis System ( Life Technologies ) and relevant primer pairs ( sequences available upon request ) . Following sequencing to confirm that the correct mutation had been inserted and that no other changes had been acquired , recombinant viruses were generated by transfection of all eight plasmids into a co-culture of human embryonic kidney-293T and MDCK cells , as previously described [31] . The correct NA sequence was confirmed by DNA sequencing for all generated viruses . All amino acid positions are described relative to the first methionine of the N1 NA . The NA protein surface expression and total activity of each NA variant were determined using 293T cells transfected with plasmids encoding the NA variants . The assays were performed as described in [32] , with the following modifications: the IRES-mCherry was replaced with an IRES-GFP , the HA tag was replaced with a V5 tag , and the antibody used was the anti-V5 AF647-conjugated antibody ( Invitrogen 45–1098 ) at a 1∶200 dilution . NA activity results were compared using a two-tailed t test . In order to assess the impact of each mutation upon viral replication efficiency , each virus was subjected to single-step and multi-step replication cycle experiments . Single-step replication experiments involved the infection of MDCK cells at a high multiplicity of infection ( MOI ) of 1 TCID50/cell , followed by sampling of the cell-culture supernatant every 2 h for 12 h , while multi-step replication experiments involved the infection of cells at a low MOI of 0 . 01 TCID50/cell , followed by sampling of the cell-culture supernatant every 12 h until 60 h . Viruses were also assayed for in vitro replication in the presence and absence of 250 nM oseltamivir carboxylate ( kindly provided by Hoffmann-La Roche , Switzerland ) as described by Yen et al . [33] . Briefly , confluent monolayers pre-treated for 2 h with either PBS or PBS containing 250 nM oseltamivir were infected at a MOI of 0 . 001 TCID50/cell with each virus for 30 min . Immediately thereafter excess virus particles were removed [34] . Cell monolayers were washed once with 0 . 9% aqueous NaCl ( pH 2 . 2 ) to remove unbound virus particles , followed by two washes with PBS to adjust the pH back to pH 7 . 2 . Cells were then grown in the presence or absence of 250 nM oseltamivir in maintenance media for 48 h , after which the cell culture supernatant was sampled and frozen at −80°C prior to subsequent titration on MDCK cells . An in vitro NA inhibition assay was performed as previously described by Hurt et al . [35] , to determine the concentration of oseltamivir required to inhibit 50% of the NA activity ( IC50 ) of each virus using a logistic curve fit program ( Robosage , Glaxo-SmithKline , UK ) . The relative fitness of virus pairs was assessed using a ferret competitive mixtures model previously described by our laboratory [27] . In preparation for inoculation into ferrets , each pair of viruses was diluted to 1×105 TCID50/ml and used to inoculate ferrets either as pure populations ( Virus A:Virus B , 100:0% , 0%:100% ) or as a series of deliberately prepared mixtures ( Virus A:Virus B , 80%:20% , 50%:50% , 20%:80% ) based on their infectivity titre [27] . Each of the pure population or mixtures was considered an experimental group . Two groups of ferrets were used to assess the 50%:50% mixture . Each experimental group comprised three ferrets; one ferret served as an artificially infected donor ferret and two ferrets served as sequentially naturally infected recipient ferrets ( the 1st recipient was infected by the donor and the 2nd recipient was infected by the 1st recipient ) . On day 0 of each experiment , donor ferrets were anesthetised intramuscularly with 20 mg/ml Ilium Xylazil-20 ( Troy-Laboratories , Australia ) and intranasally inoculated with 0 . 5 ml PBS containing 5×104 TCID50 of virus . Once inoculated , each donor ferret was housed separately in a high efficiency particulate air ( HEPA ) filtered cage . After 24 h , 1st recipient ferrets were co-housed with the donor ferrets to allow virus transmission . Beginning on day 1 post-infection ( pi ) of the donor ferret , all ferrets were nasal washed daily , as described [27] . Nasal washes collected from 1st recipient ferrets were analysed immediately following collection for influenza A virus using a rapid point-of-care test ( Directigen EZ Flu A+B , Becton Dickinson and Company ) . Once influenza infection was confirmed in a 1st recipient ferret it was transferred to a clean cage housing a 2nd recipient ferret to allow virus transmission to the 2nd recipient ferret . All ferrets were euthanized on days 10 or 11 pi . All ferrets used in these experiments were 6–12 months old , of mixed gender , and all received food and water ad libitum . Prior to infection all ferrets were confirmed as seronegative to currently circulating human influenza viruses using a haemagglutination inhibition assay . To determine the infectious virus titre , samples were titrated on MDCK cells and incubated at 37°C in 5% ( v/v ) CO2 as described by Hurt et al . [27] with the modification that , after 4 days of incubation the presence of haemagglutinating virus in each well was assessed using turkey red blood cells , and the virus titre calculated according to the method of Reed and Muench [29] . A TaqMan one-step quantitative RT-PCR assay capable of detecting the influenza matrix gene segment [36] was used to detect influenza viral RNA in ferret nasal washes . This assay was performed using a SensiFast Probe Lo-ROX One-Step kit ( Bioline ) and an Applied Biosystems 7500 Fast Real Time PCR System ( Life Technologies ) . Cycle threshold ( Ct ) values for each sample were compared to those obtained for a set of RNA transcripts encoding the A/California/7/2009 A ( H1N1 ) pdm09 matrix genome segment , which were included as a control in each assay . These RNA transcripts were generated using the pGEM-A/Cal/7/2009 matrix plasmid kindly provided by Heidi Peck ( WHO Collaborating Centre for Reference and Research on Influenza , Melbourne , Australia ) . The relative proportion of each virus in ferret nasal washes was determined using pyrosequencing allele quantitation analysis . Viral RNA was extracted from nasal washes as described above and subjected to RT-PCR using a MyTaq One-Step RT-PCR kit ( Bioline ) and the primer pairs shown in Table S2 . Each RT-PCR assay produced double-stranded DNA fragments which encompassed the NA H275Y , V241I or N369K mutation sites . Single-stranded biotinylated DNA was purified from each RT-PCR product and subjected to pyrosequencing analysis using the relevant sequencing primer ( Table S2 ) and a PyroMark ID pyrosequencer ( Biotage ) as described by Deng et al . [36] . In Deng et al . [36] , we showed that the difference between the pyrosequencing estimate and the known proportion of H275 vs . Y275 in pure viruses and defined mixtures was between 0 . 5 and 10 . 5% . Following validation of the assays used in this study we determined a similar accuracy range using purified plasmids ( at a concentration range of 10−2 to 10−6 DNA copies ) for each of the NA H275Y , V241I or N369K pyrosequencing assays . Therefore assay of a pure viral population may indicate the presence of a minor ( <10% ) proportion of the alternative viral population . Within the competitive-mixtures model , fitness differences between competing strains arise in two contexts: within-host replication kinetics and host-to-host transmission . The overall fitness of one strain compared to another arises as a combination of these two factors . Here we developed a within-host mathematical model of virus kinetics to provide a quantitative estimate of the relative within-host viral replication fitness of the strains used in the competitive mixtures experiments . We assessed the transmission fitness of strains using our previously developed model [27] , [37] . Briefly , we modelled the within-host replication of the two influenza strains using the classic Target cell – Infectious cell – Virus ( TIV ) paradigm [38]–[40] . Free infectious virus infects healthy ‘target’ ( epithelial ) cells , which following a latent phase , become infectious , releasing progeny virus particles that subsequently infect further target cells . Free virus is removed from the system due to natural decay and ( time-independent ) immune responses . Our particular model extends the standard TIV paradigm through inclusion of two co-infecting strains that compete for the same target cell reservoir ( see also [41] ) . Furthermore , by modelling both infectious and total ( infectious and non-infectious ) viral matter [42] , our model may be fitted to both TCID50 and RNA based assay data , providing more precise estimates of relevant within-host parameters . The experiments investigated mutations within the NA gene , and a primary function of NA is to aid in the release of budded viruses from the surface of infected cells [43] , [44] . We therefore assumed that observed differences in within-host viral kinetics between strains arose through a difference in the production rate of infectious virus from infected cells . The ratio of this estimated production rate by strain served as a measure of the relative within-host replication fitness of the two strains . Further details of the mathematical model and how it was fitted to the available data are presented in Supplementary Text S1 . Using the within-host and transmission models we calculated the relative within-host and relative transmission fitness values along with 95% confidence intervals ( 95% CI ) for the virus pair used in each ferret experiment . Relative fitness values of >1 indicated an advantage for virus B over A in each comparison pair , provided that the 95% CI did not cross 1 . All available A ( H1N1 ) pdm09 NA protein sequences derived from viruses that infected human hosts were downloaded from the Global Initiative on Sharing All Influenza Data website ( http://www . gisaid . org ) ( Supplementary Text S2 ) and the influenza virus resource at the National Centre for Biotechnology Information [45] . After removal of duplicate sequences for unique viral strains , 14234 sequences were aligned using MAFFT [http://www . ncbi . nlm . nih . gov/pubmed/18372315] . To generate the figure that illustrates the evolution timeline of the NA V241I and N369K mutations , the NA sequence from A/California/07/2009 was used as reference and the percentages of occurrences for each of the mutations were calculated on a monthly basis since April 2009 . To keep the figure legible , only mutations found in 100% of all circulating viruses in any of the months since April 2009 were retained . The drug resistance H275Y and the N386S mutations were also kept for reference .
An in vitro NA inhibition assay was used to assess the oseltamivir susceptibility of the HNE2011 OR and OS viruses . As expected , New17 OR ( which encoded NA 275Y ) had an approximately 115-fold higher IC50 compared to New163 OS ( Table S3 ) . Cell culture experiments further confirmed that oseltamivir had no effect on the replication of the New17 OR virus whereas it impaired the replication of the New163 OS virus ( Figure S2 ) . Given the spread of HNE2011 H275Y OR viruses , we hypothesised that they may have had equivalent or superior fitness to similar OS strains . To test this hypothesis we used the ferret competitive mixtures model to assess the relative fitness of an HNE2011 OR virus ( New17 OR ) compared with an OS virus from the same location and time ( New163 OS ) . Following viral inoculation of ferrets , pure populations of the New17 OR and New163 OS viruses replicated to equivalent titres ( in groups of three ferrets , mean [±SD] peak titres for the two viruses were 5 . 3±0 . 9 and 5 . 0±0 . 9 log10TCID50/ml respectively ) with a mean [±SD] duration of shedding of 6 . 0±0 . 0 vs . 5 . 7±1 . 2 days respectively ( Figure 1A , S3 ) . Furthermore there were no significant differences in mean weight loss , mean temperature increases or other clinical signs amongst the groups of ferrets inoculated with pure populations ( data not shown ) . Pyrosequencing analysis showed the maintenance of pure populations ( i . e . values were >90% and within the expected variability of the assay – see materials and methods ) ( Figure 1B ) . In the four groups of ferrets inoculated with virus mixtures , a pure population of OR virus ( >90% ) was observed by the end of the infection in the 2nd recipient of the 80∶20 group and the 1st recipient of one of the 50∶50 groups ( Figure 1B ) . Mixtures persisted until the end of the infection in the other ferrets . Mathematical analysis of the data found that , while there was no significant difference in transmissibility , the OR virus had enhanced within-host viral replication fitness ( relative within-host fitness value [95% CI] = 1 . 07 [1 . 02; 2 . 59] ) relative to the OS virus ( Table 1 ) . The transmission event between the 1st and 2nd recipient of one of the 50∶50 groups showed an unusual transfer from >90% OR to <5% OR virus . Sequence analysis did not reveal any new mutations within the NA genes of this virus pair , although a HA P154S amino acid change was detected in the virus mixture within the nasal washes of the 2nd recipient ferret in this group . This change was not present in any of the nasal washes from the donor or 1st recipient ferrets . An in vitro NA expression system was used to investigate the effect of the NA mutations V241I , N369K and S386N on NA surface expression and enzymatic activity . Reversal of the V241I and N369K PPMs individually from the New17 OR NA protein ( New17 I241V OR and New17 K369N OR ) resulted in approximately 40% and 35% reductions in NA surface expression and enzymatic activity respectively ( Figure 2A ) . However , reversal of both mutations reduced NA expression and activity by approximately 60% ( Figure 2A ) . In contrast , removal of the N386S mutation from the HNE2011 NA ( New17 S386N OR ) enhanced the activity and surface expression by approximately 50% ( Figure 2A ) , demonstrating that this mutation was unlikely to be having a beneficial effect upon the fitness of the New17 OR virus . Hence , subsequent investigations were restricted to the NA V241I and N369K mutations . Removal of H275Y from the New17 OR NA protein enhanced its surface expression and enzymatic activity by approximately 50% ( Figure 2A ) , demonstrating the detrimental effect of the H275Y mutation upon surface expression and enzymatic activity . To further evaluate the impact of both H275Y and the PPMs V241I and N369K in vitro , we introduced the mutations into an early A ( H1N1 ) pdm09 OS virus from 2009 . While the incorporation of H275Y into the Cal7 OS NA protein reduced its surface expression and activity by 50% , addition of the V241I and N369K mutations partially offset these losses , by approximately 40% and 20% respectively ( Figure 2B ) . Addition of both V241I and N369K together offset these losses by approximately 70% ( Figure 2B ) . We then determined the effect of removing both V241I and N369K from a HNE2011 OR virus on in vivo within-host and transmission fitness . Initial comparison of in vitro replication kinetics at low and high MOI showed that replication of the rgNew17 I241V , K369N OR virus was delayed for the first 4-6 h pi compared to the rgNew17 OR virus ( Figure S4A , B ) . Following inoculation in ferrets , there were no significant differences in morbidity between the groups of ferrets inoculated with pure populations of the rgNew17 OR and rgNew17 I241V , K369N OR viruses ( data not shown ) , and both viruses replicated to similar titres ( 4 . 9±0 . 4 and 5 . 5±0 . 3 log10TCID50/ml respectively ) and were shed for an equivalent duration ( 5 . 3±0 . 6 days for both viruses ) ( Figure 3A , S5 ) . Pyrosequencing analysis revealed that a pure virus population was maintained in each of these groups ( Figure 3B ) . In the four groups of ferrets inoculated with virus mixtures , a pure population of rgNew17 OR virus , that was maintained upon subsequent transmission to further recipient ferrets , was observed by the end of the infection in the 1st recipient of both of the 50∶50 groups and the 20∶80 group , as well as in the donor of the 80∶20 group ( Figure 3B ) . Modelling indicated that the rgNew17 OR virus exhibited significantly superior transmission ( relative fitness value [95% CI] = 2 . 69 [1 . 00; 7 . 24] ) and within-host viral replication fitness ( relative fitness value [95% CI] = 1 . 82 [1 . 35; 7 . 46] ) , compared to the rgNew17 I241V , K369N OR virus ( Table 1 ) , demonstrating that removal of NA 241I and 369K impaired the ability of the rgNew17 I241V , K369N OR virus to out-compete the rgNew17 OR virus both during replication within hosts and upon transmission between them . To investigate the individual influence of the NA V241I and N369K mutations upon the fitness of recent OR A ( H1N1 ) pdm09 viruses , rgNew17 OR was compared with isogenic rg viruses from which either the NA V241I ( rgNew17 I241V OR ) or N369K ( rgNew17 K369N OR ) mutations had been removed . The in vitro replication kinetics of rgNew17 OR , rgNew17 I241V OR and rgNew17 K369N OR viruses along with the “natural” New17 OR virus was first determined in MDCK cells at a low and high MOI . All viruses replicated efficiently in MDCK cells with most of the New17 OR recombinant viruses following a similar growth pattern to that of the “natural” New17 OR virus from which they were derived ( Figure S4A , B ) . However , rgNew17 K369N OR showed delayed replication for the first 6–8 h pi at a high MOI ( Figure S4A ) , suggesting that removal of NA 369K had a somewhat detrimental effect upon virus growth in vitro . Analysis of the rgNew17 OR vs . rgNew17 I241V OR virus pair in ferrets showed that pure populations of the two viruses replicated to 4 . 0±0 . 5 , and 5 . 2±0 . 4 log10TCID50/ml respectively , and were shed for 5 . 3±1 . 2 and 6 . 0±1 . 0 days respectively ( Figure S6A , C ) . In addition , there were no significant differences in morbidity between the groups of ferrets inoculated with the pure populations ( data not shown ) . Pyrosequencing analysis revealed that a pure virus population had been maintained in each of these groups ( Figure S6B ) . In the four groups of ferrets inoculated with virus mixtures , a pure population of rgNew17 OR virus , that was maintained upon subsequent transmission to further recipient ferrets , was observed by the end of the infection in the 1st recipient of one of the 50∶50 groups and the 80∶20 group . Modelling revealed that the rgNew17 OR virus exhibited significantly superior transmission ( relative transmission fitness value [95% CI] = 2 . 22 [1 . 24; 3 . 97] ) and within-host viral replication fitness ( relative within-host fitness value [95% CI] = 3 . 96 [1 . 17; 6 . 83] ) compared to the rgNew17 I241V OR virus ( Table 1 ) , demonstrating that the V241I NA mutation is important in contemporary A ( H1N1 ) pdm09 OR viruses for both within-host and transmission fitness . To assess the effect of the K369N mutation , ferrets were inoculated with the rgNew17 OR and rgNew17 K369N OR viruses . Pure populations of the rgNew17 OR and rgNew17 K369N OR viruses replicated to 5 . 0±0 . 5 and 4 . 1±0 . 6 log10TCID50/ml respectively , and both viruses were shed for a similar duration of 5 . 0±1 . 2 and 5 . 3±0 . 6 days respectively ( Figure S7A , C ) . Furthermore , there were no significant differences in morbidity between the groups of ferrets inoculated with the pure populations ( data not shown ) . Pyrosequencing analysis revealed that a pure virus population had been maintained in each of these groups ( Figure S7B ) . In the four groups of ferrets inoculated with virus mixtures , a pure population of rgNew17 OR virus was observed by the end of the infection in the 1st recipient of both of the 50∶50 groups and the 80∶20 group , as well as in the 2nd recipient of the 20∶80 group . Virus transmitted to subsequent recipient ferrets in these groups persisted as a pure population of rgNew17 OR virus , with the exception of a 2nd recipient ferret in one of the 50∶50 groups in which rgNew17 OR virus accounted for 87% of the total virus present in the nasal wash on the final day of the experiment . Modelling found that , while the rgNew17 OR virus exhibited significantly superior within-host viral replication fitness compared to the rgNew17 K369N OR virus ( relative within-host fitness value [95% CI] = 3 . 14 [1 . 17; 5 . 30] ) , there was no statistical evidence for a transmission fitness difference between the viruses ( Table 1 ) . In vitro assays demonstrated that introduction of the NA V241I and N369K mutations into an earlier A ( H1N1 ) pdm09 OR NA protein , that did not naturally contain these mutations , resulted in enhanced enzymatic activity and NA expression ( Figure 2 ) . Therefore the effect of introducing these PPMs into an early A ( H1N1 ) pdm09 OR virus ( Perth261 OR ) was assessed in ferrets . Rg viruses were generated encoding the complete Perth261 OR genome without any changes ( rgPerth261 OR ) and with the NA V241I and N369K mutations ( rgPerth261 V241I , N369K OR ) . In vitro replication kinetics showed that both rg viruses grew more rapidly and to higher virus titres than the “natural” parent ( Perth261 OR ) virus , at both a low and high MOI ( Figure S4C , D ) . Following inoculation into ferrets , there were no significant differences in morbidity between the groups of ferrets inoculated with pure populations of rgPerth261 OR and rgPerth261 V241I , N369K OR ( data not shown ) , and both viruses replicated to similar titres of 5 . 8±1 . 0 and 5 . 4±1 . 2 log10TCID50/ml respectively , and were shed for a similar duration of 5 . 3±0 . 6 and 5 . 7±0 . 6 days respectively ( Figure 4A , S8 ) . Pyrosequencing analysis revealed that a pure virus population was maintained in each of these groups ( Figure 4B ) . In the four groups of ferrets inoculated with virus mixtures , a pure population of rgPerth261 V241I , N369K OR virus , that was maintained upon subsequent transmission to further recipient ferrets , was observed by the end of the infection in the donor in the 20∶80 group , and in the 1st recipient in the 80∶20 group , and both of the 50∶50 groups ( Figure 4B ) . Modelling showed that while the rgPerth261 V241I , N369K OR virus exhibited superior within-host viral replication fitness ( relative within-host fitness value [95% CI] = 1 . 86 [1 . 37; 7 . 24] ) compared to the rgPerth261 OR virus , there was no significant difference in transmission fitness between the viruses ( Table 1 ) . Hence in the case of this experiment , the lack of any rgPerth261 OR virus in the 2nd recipient ferrets in each of the groups inoculated with virus mixtures , demonstrates how a virus with superior within-host replication fitness but with equivalent transmission fitness , may replicate more efficiently within successive infected hosts , such that its relative proportion increases prior to each transmission event , eventually dominating the virus mixture .
The PPMs NA V241I and N369K are now present in >99% of circulating A ( H1N1 ) pdm09 viruses ( Figure S1 ) . The experiments and accompanying mathematical analysis presented in this study demonstrate that these mutations enable A ( H1N1 ) pdm09 viruses to maintain robust viral fitness when they acquire the NA H275Y oseltamivir resistance mutation . The computational analysis reported previously [26] , in conjunction with the in vitro experiments presented here , demonstrate that NA V241I and N369K are indeed permissive mutations that act by enhancing both the surface expression and total activity of H275Y A ( H1N1 ) pdm09 NA proteins , similar to the effect that the R222Q and V234M mutations had on the NA of H275Y seasonal A ( H1N1 ) viruses [15] . The data obtained from the competitive-mixture ferret experiments performed in this study were subjected to mathematical modelling to determine the relative within-host and transmission fitness of each virus pair . In this way it was possible to investigate whether the presence or absence of the NA V241I and/or the N369K permissive mutations enabled the OR viruses to replicate more efficiently within ferrets , and/or be more efficiently transmitted between ferrets . In using the within-host mathematical model to calculate the relative within-host viral replication fitness ( summarised in Table 1 ) we made the assumption that the observed strain-dependence in viral kinetics arose due to differing infectious virus production rates between strains ( see Supplementary Text S1 for mathematical details ) . While there are other plausible biological explanations for the observed within-host viral kinetics , a careful consideration of these alternatives ( see Supplementary Text S1 for details ) suggests they are not reconcilable with the picture of outgrowth that we see across multiple host-to-host transmission events in many of the competitive-mixtures experiments performed in this work . Previous studies in ferrets , guinea pigs and mice revealed broadly equivalent [46]–[50] , or lower fitness [51] of early H275Y OR A ( H1N1 ) pdm09 viruses compared to genetically similar OS ( NA 275H ) strains . In most of these studies , the viruses used did not contain the PPMs V241I and N369K which were shown here to improve the fitness of the OR viruses . However a recent study by Abed et al . [50] showed that the introduction of the V241I and N369K PPMs into an early H275Y A ( H1N1 ) pdm09 virus resulted in higher virus titres in ferret nasal washes . Abed et al . [50] also noted that introduction of a T289M NA mutation ( which was identified as a PPM by computational analyses , but has not yet been detected in circulating strains ) into an early H275Y A ( H1N1 ) pdm09 virus resulted in greater weight loss , enhanced mortality and higher lung viral titres in mice . Most recently in 2013 two other NA mutations ( N44S and N200S ) have become almost universally observed in A ( H1N1 ) pdm09 viruses , whilst at the same time an NA V106I mutation , which was rapidly acquired at the beginning of the A ( H1N1 ) pdm09 pandemic , has been lost ( Figure S1 ) . Whether these more recently acquired NA changes have an impact upon the fitness of OR A ( H1N1 ) pdm09 viruses remains to be investigated . Given the apparent robust fitness of the H275Y HNE2011 viruses in this study , the obvious question is why have they not yet re-emerged ? One explanation is that a high level of circulating A ( H1N1 ) pdm09 viruses may be required for a A ( H1N1 ) pdm09 OR virus to become established and spread . The Australian HNE2011 virus cluster emerged [25] , [26] during a season when A ( H1N1 ) pdm09 viruses accounted for almost 40% of all influenza A and B viruses detected globally but , in 2012 and 2013 , the proportion of A ( H1N1 ) pdm09 viruses circulating has been considerably lower ( 9% and 25% respectively ) [52] . In the most recent 2013/14 Northern Hemisphere influenza season , a cluster of A ( H1N1 ) pdm09 H275Y OR viruses that contained both the V241I and N369K PPMs plus an additional N386K NA mutation , was detected in Sapporo , Japan [53] , during a period of the season where A ( H1N1 ) pdm09 viruses contributed approximately 50% of the circulating influenza strains [54] . Although the majority of Japanese A ( H1N1 ) pdm09 OR viruses are currently localised to the Sapporo prefecture ( as of Feb 3 , 2014 , the frequency of resistance in Sapporo was 88% [15/17] compared with 7% [22/298] for the whole of Japan [55] ) , there have been reports of genetically similar A ( H1N1 ) pdm09 OR viruses also being detected in China during the same time period [56] . Such clusters need to be closely monitored to determine if spread of OR viruses is occurring into other regions . Apart from NA PPMs , it may be that other properties , such as antigenic novelty , are also necessary for an OR virus to spread widely . In 2007–2008 , the H275Y NA mutation became fixed in a new seasonal A ( H1N1 ) antigenic variant ( A/Brisbane/59/2007-like ) , suggesting that the antigenic novelty of the OR virus assisted its prolific spread [57] . The results of this study show that A ( H1N1 ) pdm09 viruses have now acquired permissive NA mutations which allow them to retain viral fitness when the H275Y NA mutation is present , raising the possibility of rapid global spread of an OR A ( H1N1 ) pdm09 virus if H275Y were to arise in an antigenically drifted virus . A ( H1N1 ) pdm09 viruses have now been circulating in humans for over four years , but are yet to undergo a significant antigenic change ( as evidenced by the continued inclusion of A/California/7/2009 in the human seasonal influenza vaccine since 2009 ) . As the H1 component of the vaccine has been updated , on average , every 2 . 8 years ( range 1 to 8 years ) , and the H3 component every 1 . 8 years ( range 1 to 4 years ) since 1980 , it is reasonable to anticipate that A ( H1N1 ) pdm09 viruses will undergo antigenic change in the near future . At present oseltamivir remains the primary drug of choice for the treatment of human influenza infection worldwide , although the newly licensed neuraminidase inhibitor laninamivir has recently become widely used in Japan [58] . Until laninamivir becomes more widely available , oseltamivir will continue to remain the most accessible option for the prevention and treatment of influenza . Given the data presented here and recent reports of community-wide spread of oseltamivir resistant virus in the absence of drug selection pressure [23]–[25] , there is an urgent need to reassess the almost exclusive reliance upon oseltamivir both for the treatment of human influenza infection and as the primary component of antiviral drug stockpiles for use during influenza pandemics . Alternatives include the other widely available influenza NA inhibitor drug , zanamivir , against which resistant viruses are rarely detected [59] , and laninamivir , which is likely to become licensed and more widely available in coming years . It is notable that the majority of OR viruses ( including those containing the NA H275Y mutation ) retain sensitivity to zanamivir and laninamivir . Furthermore , future research efforts should investigate new antiviral drugs including those that target viral components other than the NA , which may be suitable for use alone or in combination with the current NA inhibitors [60] . Here we demonstrate that contemporary A ( H1N1 ) pdm09 viruses have acquired NA mutations which permit the acquisition of NA H275Y without compromising viral fitness . These mutations , which are now present in virtually all circulating A ( H1N1 ) pdm09 viruses , enhance the surface expression and enzymatic activity of the A ( H1N1 ) pdm09 H275Y NA protein in vitro and result in enhanced viral fitness in vivo . Hence , the risk that H275Y A ( H1N1 ) pdm09 viruses will spread globally , in a similar manner to OR seasonal A ( H1N1 ) viruses in 2007–2008 , now appears greater than at any time since the A ( H1N1 ) pdm09 lineage emerged in 2009 .
|
Antimicrobial resistance is an increasing problem for the treatment of infectious diseases . In 2007–2008 human seasonal A ( H1N1 ) influenza viruses rapidly acquired resistance to the most commonly used anti-influenza drug oseltamivir , via a H275Y amino acid mutation within the neuraminidase ( NA ) protein . In 2009 the oseltamivir sensitive A ( H1N1 ) pdm09 virus ( encoding NA 275H ) emerged in the human population , rapidly replacing the oseltamivir resistant seasonal A ( H1N1 ) virus . However , there is increasing concern that currently circulating A ( H1N1 ) pdm09 viruses may similarly acquire oseltamivir resistance ( via the NA H275Y mutation ) and become widespread . Here we demonstrate that two novel amino acid changes present in virtually all recent A ( H1N1 ) pdm09 viruses ( NA V241I and N369K ) enable the acquisition of the NA H275Y oseltamivir resistance mutation without compromising viral fitness . As such recent A ( H1N1 ) pdm09 viruses are now one step closer to acquiring widespread oseltamivir resistance .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"antimicrobials",
"animal",
"models",
"of",
"infection",
"organismal",
"evolution",
"infectious",
"diseases",
"microbial",
"mutation",
"medicine",
"and",
"health",
"sciences",
"emerging",
"infectious",
"diseases",
"viral",
"persistence",
"and",
"latency",
"microbial",
"evolution",
"virology",
"emerging",
"viral",
"diseases",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"co-infections",
"microbiology",
"antivirals",
"evolutionary",
"biology",
"viral",
"evolution"
] |
2014
|
Estimating the Fitness Advantage Conferred by Permissive Neuraminidase Mutations in Recent Oseltamivir-Resistant A(H1N1)pdm09 Influenza Viruses
|
Clostridium botulinum neurotoxin ( BoNT ) causes flaccid paralysis by disabling synaptic exocytosis . Intoxication requires the tri-modular protein to undergo conformational changes in response to pH and redox gradients across endosomes , leading to the formation of a protein-conducting channel . The ∼50 kDa light chain ( LC ) protease is translocated into the cytosol by the ∼100 kDa heavy chain ( HC ) , which consists of two modules: the N-terminal translocation domain ( TD ) and the C-terminal Receptor Binding Domain ( RBD ) . Here we exploited the BoNT modular design to identify the minimal requirements for channel activity and LC translocation in neurons . Using the combined detection of substrate proteolysis and single-channel currents , we showed that a di-modular protein consisting only of LC and TD was sufficient to translocate active protease into the cytosol of target cells . The RBD is dispensable for cell entry , channel activity , or LC translocation; however , it determined a pH threshold for channel formation . These findings indicate that , in addition to its individual functions , each module acts as a chaperone for the others , working in concert to achieve productive intoxication .
Botulinum neurotoxin ( BoNT ) inhibits synaptic exocytosis in peripheral cholinergic synapses causing botulism , a severe disease characterized by descending flaccid paralysis . Clostridium botulinum strains express seven BoNT isoforms known as serotypes A to G [1] . Each BoNT isoform is synthesized as a single polypeptide chain with a molecular mass of ∼150 kDa . The precursor protein is cleaved either by clostridial or host cell proteases into two polypeptide chains linked by a disulfide bridge [1]–[3] . Structurally , the mature toxin consists of three modules [1] , [4]–[6]: a 50 kDa light chain ( LC ) Zn2+-metalloprotease , and the 100 kDa heavy chain ( HC ) which encompasses the N-terminal ∼50 kDa translocation domain ( TD ) , also denoted HN and the C-terminal ∼50 kDa receptor-binding domain ( RBD ) also termed HC . This tri-modular architecture has a physiological counterpart . The RBD determines the cellular specificity mediated by the high affinity interaction with a surface protein receptor , SV2 for BoNT/A [7] , [8] and BoNT/E [9] , and synaptotagmins I and II for BoNT/B and BoNT/G [10] , and a ganglioside ( GT1B ) co-receptor [7] , [8] , [10] , [11] . Then , BoNTs enter sensitive cells via receptor-mediated endocytosis [1] , [12] , [13] . Exposure of the BoNT-receptor complex to the acidic milieu of endosomes [12]–[16] induces a conformational change leading to the insertion of the HC into the endosomal bilayer membrane , thereby forming transmembrane channels [17]–[20] . Previous studies have provided compelling evidence for the retrieval of a folded LC protease capable of proteolyzing its SNARE ( soluble NSF attachment protein receptor ) substrates , which are essential for synaptic vesicle fusion and neurotransmitter release [1] , [2] , [21] , [22] , only after productive translocation across lipid bilayer membranes and release from the channel [23] . These results indicate that the HC of BoNT/A acts as both a channel and a transmembrane chaperone for the LC to ensure a translocation competent conformation during LC transit [19] , [23] . They also support the view that the TD module is the conduit for the passage of the LC module from the interior of the acidic endosome into the cytosol allowing access to the intracellular SNARE substrates [1] , [13] , [19] , [23] . Thus , BoNT represents a fascinating example of molecular partnership: the HC chaperone activity driven by a pH gradient across the endosome prevents aggregation of the LC in the acidic vesicle interior , maintains the LC unfolded conformation during translocation , and releases the LC after it refolds at the neutral cytosolic pH . In the process , the HC channel is occluded by the LC during transit and open after completion of translocation and release of cargo [23]–[25] . Here , we investigated the minimal domain requirements for BoNT channel activity and for productive LC translocation . We characterized the isolated TD and the TD-disulfide linked to the catalytic LC using a translocation assay with single molecule resolution on excised membrane patches of neuroblastoma Neuro 2A cells [24] , [25] . The system allowed us to dissect the minimal domain requirements and pH constraints for channel activity and for LC translocation . We utilized a cell-based neurotoxicity assay to assess the impact of the single molecule studies at the cellular level . The RBD was determined to be dispensable for both channel activity and LC translocation under mild acidic conditions ( pH∼6 ) . In contrast , the RBD restricted insertion of the TD into the membrane until localized to an acidic endosomal compartment where low pH ( pH∼5 ) induced channel insertion concurrent with partial unfolding of the LC cargo , thereby triggering productive LC translocation and ultimate release into the cytosolic compartment .
Unless otherwise specified , all chemicals were purchased from Sigma-Aldrich . Purified native BoNT/A holotoxin and HC were from Metabiologics . E . coli strain DH5α cells were used in the cloning procedures and E . coli strains BL21 . DE3 . pLysS and BL21 . DE3 ( RIL ) were used for protein over expression . DNA corresponding to the BoNT/A catalytic and translocation domains was obtained from two separate starting vectors: BoNT/A LC ( unpublished results; a wild-type BoNT/A catalytic domain sequence ) and TD-SDmut [26]; a wild-type BoNT/A TD amino acid sequence with silent mutations to disrupt an internal Shine Dalgarno site ) . The catalytic domain was amplified with primers that introduced a 5′ NdeI site ( GGAATTCCATATGCCATTTGTTAATAAAC ) and 3′ Sac1 site ( GGCGAGCTCGCTTATTGTATCCTTTATCTAATG ) and ligated into a TOPO vector ( Invitrogen , Carlsbad , CA ) . An internal SacI site was then removed by QUIK-Change ( Stratagene , La Jolla , CA ) mutagenesis using the primer CTCTGGCACACGAACTGATCCACGCTGGTC and its reverse complement . The BoNT/A TD was amplified with primers that introduced a 5′ SacI site ( GCCGAGCTCTGAACGATCTGTGTATCAAAGTTAATAATTGGG ) and 3′ XhoI site ( CCGCTCGAGGTTCTTAATATATTCAGTAAATGTAG ) and ligated into a TOPO vector . The catalytic and translocation domains were excised from the TOPO vectors using NdeI/SacI and SacI/XhoI , respectively and ligated into a pET24b vector that had been digested with NdeI/XhoI . The use of the engineered SacI site results in the insertion of an Arg between K447 ( cat ) and A448 ( trans ) , a feature that was included in the design to improve the efficiency of trypsin activation . The BoNT/A TD was expressed and purified from the TD-SDmut construct as described previously [26] with the following modification . BoNT/A TD was eluted from the Ni-affinity column in the presence of 0 . 2% dodecyl maltoside ( DDM ) and concentrated to 0 . 375 mg/ml in 0 . 2% DDM , 20 mM Tris , 150 mM NaCl , pH 8 . 0 . The large dilutions used for single molecule studies insure that the effect of detergent on cell membranes is negligible . No detergent was required for LC-TD purification . BoNT/A LC-TD was expressed in E . coli BL21 . DE3 ( RIL ) cells . A 10 ml aliquot was added to one L of Terrific Broth and induced at an OD = 600 nm using 1 mM isopropyl-β-D-thiogalactopyranoside ( IPTG ) . Cells were harvested after overnight expression at 18°C and lysed using a French Press . The protein was purified on a Ni-NTA column followed by a Fast Flow Q Sepharose column and S200 gel filtration column . Di-chain BoNT/A LC-TD was generated by cleavage with trypsin: BoNT/A LC-TD ( 0 . 5 mg/ml ) was incubated with 1 µg/ml trypsin overnight at 22±2°C . Thereafter , trypsin was inactivated with 0 . 25 mg/ml trypsin soybean inhibitor for 15 min at 22±2°C . The purity of all BoNT/A proteins was determined qualitatively with SDS-PAGE analysis ( Figure S1 ) . Recombinant SNAP-25 [27] , [28] was incubated with 60 ng of BoNT/A holotoxin or LC-TD in 13 . 2 mM Hepes ( pH 7 . 4 ) , 20 mM dithiothreitol , and 1 µM Zn ( CH3COOH ) 2 for 30 min at 37°C . SDS-PAGE ( 12% ) was used to visualize cleavage of SNAP-25 by BoNT/A proteins [29] ( Figure S2 ) . Excised patches from Neuro 2A cells in the inside-out configuration were used as described [24] , [25] . Current recordings were obtained under voltage clamp conditions . Records were acquired at a sampling frequency of 20 kHz and filtered online to 2 kHz with Gaussian filter . All experiments were conducted at 22±2°C . To emulate endosomal conditions the trans compartment ( bath ) solution contained ( in mM ) NaCl 200 , NaMOPS [3- ( N-morpholino ) propanesulfonic acid] 5 , ( pH 7 . 0 with HCl ) , tris- ( 2-carboxyethyl ) phosphine ( TCEP ) 0 . 25 , ZnCl2 1 , and the cis compartment ( pipet ) solution contained ( in mM ) NaCl 200 , NaMES [2- ( N-morpholino ) ethanesulfonic acid] 5 , ( pH 5 . 3 or pH 6 . 0 with HCl ) . When the cis compartment was tested with pH 7 buffer , the trans compartment solution set to pH 7 . 0 was used . The osmolarity of both solutions was determined to be ∼390 mOsm . ZnCl2 was used to block endogenous channel activity specific to Neuro 2A cells [30] , [31] . BoNT protein reconstitution and channel insertion was achieved by supplementing 2 . 5–5 µg/ml BoNT holotoxin , HC , LC-TD , TD , to the pipet solution , which was set to pH 5 . 3 , pH 6 . 0 or pH 7 . 0 . Analysis was performed on single bursts of each experimental record . Only single bursts were analyzed due to the random duration of quiescent periods . A single burst is defined as a set of openings and closings lasting ≥50 ms bounded by quiescent periods of ≥50 ms before and after . Single-channel conductance ( γ ) was calculated from Gaussian fits to current amplitude histograms . The total number of opening events ( N ) analyzed was 165 , 936 . Time course of single-channel conductance change for each experiment was calculated from γ of each record , where t = 0 s corresponds to onset of channel activity , and average time course was constructed from the set of individual experiments for a single condition . The half-time for completion of single-channel conductance change event ( t½ ) was calculated from Sigmoidal fit to the average time course . The voltage-dependence of channel opening was calculated from measurements of the fraction of time that the channel is open ( Po ) as a function of voltage by integration of γ histograms where γ is 60≤γ≤75 pS . Statistical values represent means±SEM , unless otherwise indicated . n denotes the number of different experiments . Cleavage of endogenous SNAP-25 within Neuro 2A cells exposed to BoNT/A and truncation proteins was investigated as described [32]; Neuro-2A cells were seeded at a density of ∼120 , 000 cells per well in a 12-well tissue culture plate in DMEM culture medium . After incubation for 24 h , the media were removed and replaced with serum-free media containing 5 . 0 µg of BoNT/A holotoxin , LC-TD or TD . After incubation for ∼48 h , the cells were harvested by removing the media , adding 160 µl of 1× NuPAGE SDS sample buffer ( Invitrogen ) , and boiling for 10 min . SDS-PAGE and Western blotting were conducted using standard protocols . Proteins within the whole-cell extract samples were separated by SDS-PAGE on a 12% Bis-Tris NuPAGE gel in MOPS/SDS running buffer ( Invitrogen ) before transfer to a 0 . 2 µm nitrocellulose membrane for 120 min at 30°C [29] . After blocking in 2% skim milk/H2O for 20 min at room temperature , the membrane was washed three times for 5 min at room temperature with TBST [25 mM Tris ( pH 7 . 4 ) , 137 mM NaCl , 2 . 7 mM KCl , and 0 . 1% Tween 20] . Primary antibody , anti-SNAP-25 mouse monoclonal IgG1 ( 200 µg/ml; Santa Cruz Biotechnology , Santa Cruz , CA ) diluted 1∶1 , 000 into 2% skim milk/H2O , was added , and the blot was incubated for 20 min at room temperature followed by four 5-min washes with TBST at room temperature . Next , secondary antibody , goat anti-mouse HRP-conjugated ( 10 µg/ml; Pierce , Rockford , IL ) diluted 1∶500 into 2% skim milk/H2O , was added , and the blot was incubated for 1 h at room temperature followed by washing for 120 min at room temperature . Bands were visualized with 4 ml of SuperSignal West Dura Chemiluminescent Substrate ( Pierce ) and analyzed with an X-OMAT 2000A processor ( Kodak ) . Western blot analysis was quantified with the use of Image J software ( NIH ) .
Earlier studies were suggestive of channel forming activity by the N-terminal half of the HC ( HN ) [17] . We generated several truncation constructs of BoNT/A holotoxin , purified them and examined their channel activity; these include the HC , HN − the N-terminal half of the HC called the translocation domain ( TD ) [26] , and the LC and TD linked by a disulfide bridge ( LC-TD ) [33] ( Figure 1 ) . The LC-TD was expressed as a single polypeptide chain of ∼100 kDa with a disulfide bridge; the functionally relevant di-chain protein was generated by trypsin cleavage of the linker between the disulfide bridge cysteine residues . Trypsin nicking did not disrupt the disulfide crosslink , as demonstrated by SDS PAGE analysis ( Figure S1 ) , or affect the enzymatic activity of the LC ( Figure S2 ) . Channel formation was monitored on excised membrane patches from neuroblastoma cells under conditions that recapitulate the pH and redox gradients across endosomes: the cis compartment , defined as the compartment containing BoNT/A proteins , was held at pH 5 . 3 and the trans compartment was maintained at pH 7 . 0 and supplemented with the membrane nonpermeable reductant tris- ( 2-carboxyethyl ) phosphine ( TCEP ) . The LC-TD ( Figure 2C ) , and TD ( Figure 2B ) exhibited channel activity with characteristics indistinguishable from those of BoNT/A HC ( Figure 2A ) : channel activity occurred in bursts of fast transitions between closed and open states interspersed between periods of no channel activity; discrete channel openings displayed a distinctive single-channel conductance ( γ ) ∼65 pS ( in symmetric 0 . 2 M NaCl ) , with a high probability to reside in the open state ( Po ) , as illustrated by the records depicted in the corresponding bottom traces displayed at higher time resolution . The LC-TD and TD display a voltage dependence similar to the unoccluded BoNT channel which resulted after completion of LC translocation [24] , [25]; V1/2 , the voltage at which Po = 0 . 5 , −67 . 2±2 . 9 mV for holotoxin , −59 . 0±9 . 1 mV for LC-TD , and −64 . 0±4 . 2 mV for TD . Together , these results demonstrate that the channel activity of BoNT is confined to the TD . We previously demonstrated that channel activity for BoNT HC depends on the presence of a pH gradient ( ΔpH ) across the membrane [23]–[25] . The HC is a di-modular protein consisting of the TD and the RBD ( Figure 1 ) . Given that the TD forms channels similar to those of the HC ( Figure 2 ) , the question arises as to whether the RBD modulates the TD channel activity , particularly with regards to the observed dependence on ΔpH . As shown in Figure 3A , the TD channel activity was practically equivalent when the cis compartment solution was adjusted to pH 6 or pH 7 , thereby reducing ( top and middle records ) or eliminating ( bottom record ) the ΔpH . Channel activity retained the hallmark features of the channel measured for the isolated HC [23] , [34] and for holotoxin after completion of productive translocation [24] , [25] irrespective of ΔpH , namely , the bursting pattern of channel activity with an invariant γ at ∼65 pS ( Figure 3A ) . The voltage dependence of the Po was modulated by attenuating or eliminating ΔpH: The V1/2 was right shifted to ∼−47 mV when the cis compartment was maintained at pH 6 . 0 for TD ( Figure 3B ) . The interactions between the TD and the RBD therefore modulate the pH threshold for membrane insertion and channel formation; in the absence of the RBD the TD readily forms channels at neutral pH and in the absence of a ΔpH . Translocation of BoNT/A LC by the BoNT/A channel can be monitored in real time and at the single molecule level in excised membrane patches from neuroblastoma cells [24] , [25] . LC translocation through the HC channel requires conditions which emulate those prevalent across endosomes . Translocation was observed as a time-dependent increase in Na+ conductance through the HC channel recorded at −100 mV ( Figure 4A , top panel ) . Initial channel activity exhibited small , discrete events with γ∼14 pS ( Figure 4A , top , left panel ) . Progressively , γ underwent a continuous increase until reaching a stable value of 68 pS ( Figure 4A , top , right panel ) , a constant conductance monitored for the duration of the experiment ( Figure 4C , black circle ) . This steady-state γ was also the characteristic conductance of isolated HC recorded under identical conditions [24] , [25] , [34] . The half-time for completion of such an event ( t½ ) , estimated from the transition to high conductance , was ∼150 s ( Figure 4C ) . We interpret the intermediate conductance states as reporters of discrete transient steps during the translocation of the LC across the membrane; a schematic representation is depicted under the records shown in Figure 4A , top [24] , [25] . During protease translocation , the protein-conducting channel progressively conducts more Na+ around the polypeptide chain before entering an exclusively ion-conductive state . For BoNT holotoxin , channel formation and LC translocation are dependent on ΔpH; no channels were detected when the internal solution containing BoNT was held at pH 6 ( Figure 4A , middle trace ) or pH 7 ( Figure 4A , bottom trace ) ( Figure 4C , black triangle and black square ) , in agreement with our previous findings [23] . In contrast , LC-TD translocation ( Figure 4B , top panel ) proceeded even under a modest ΔpH ( 6 on the cis- and 7 on the trans- compartments ) ( middle trace ) . Remarkably , when the excised membrane patches were bathed in symmetric neutral pH solutions , the LC-TD formed HC-like channels; low conductance intermediate states were not detected ( Figure 4B , lower panel , and 4C , blue ) . Circular dichroism analysis of the LC at pH 7 indicates a high α-helical content incompatible with a translocation competent conformation [23] . These results imply that the LC remains folded at pH 7 and therefore cannot go through the ∼15 Å diameter of the TD channel [23] , [35] , thereby allowing expression of the TD channel activity unperturbed by cargo . The findings shown in Figure 4 imply that , while BoNT/A holotoxin readily enters neuronal cells via receptor-mediated endocytosis , the LC-TD devoid of the RBD would insert and be trapped in the plasma membrane unable to access its cytosolic substrate SNARE protein . To test this notion , we utilized a cell-based assay that monitors the amount of intact versus cleaved endogenous SNAP-25 protein within Neuro-2A cells [32] . This assay is highly reliable when given a 48 hr exposure to the BoNT proteins . In the absence of BoNT/A holotoxin or in the presence of isolated TD [26] , SNAP-25 remained intact ( Figure 5A , lanes 1 and 8 ) , whereas in the presence of BoNT/A holotoxin , a lower-molecular-weight proteolysis product of ∼24 KDa was detected ( Figure 5A , lane 2 ) . Incubation of cells with LC-TD resulted in proteolysis of SNAP-25 ( Figures 5A and 5B , lanes 3 and 4 ) . The extent of proteolysis attained by LC-TD was comparable to that produced by holotoxin , albeit it required higher protein concentration consistent with a lower efficacy . A single-chain BoNT/A LC-TD unexposed to trypsin ( lane 5 ) or the LC-TD that had been trypsin nicked and the disulfide bridge reduced prior to the assay ( lane 6 ) did not cleave SNAP-25 . Remarkably , cleavage of SNAP-25 by BoNT/A LC-TD does not occur when cells are preincubated with 2 µM bafilomycin , an inhibitor of the vesicular proton pump and , therefore , of endosomal acidification [36]–[38] ( lane 7 ) . These results indicate that BoNT/A LC-TD enters neurons without the aid of the RBD . Translocation requires nicking of the LC cargo from the TD carrier and does not arise from leaky cells that uptake LC via a non-specific mechanism . The implication is , therefore , that the inserted LC-TD uses the constitutive endocytic pathway to enter cells , is subsequently processed and undergoes translocation of LC with the consequent cleavage of the intracellular substrate . Together , the results shown in Figures 4 and 5 provide compelling evidence that the RBD is not necessary for channel activity or LC translocation , and that the LC-TD is a di-modular BoNT endowed with the ability to deliver folded and active LC protease into the cytosol of target cells .
The tri-modular holotoxin encompassing the LC protease , the TD and the RBD enters neurons via receptor mediated endocytosis determined by the RBD module , and achieves the cytosolic co-localization of the LC protease with its substrate by the chaperone activity of a TD protein-conducting channel . Here we demonstrate the BoNT channel-forming entity is confined to the TD ( Figure 2 ) which , at variance with the holotoxin and the di-modular HC , displays channel activity irrespective of a transmembrane ΔpH ( Figure 3 ) . We further show that for the di-modular HC the pH threshold for membrane insertion and channel formation is modulated by the interactions between these two modules and/or by the RBD interaction with the SV2 receptor ( Figure 4A and 4C ) . Previous studies indicate that channel formation occurs concomitantly with protein translocation; therefore we investigated the minimal domain requirements for productive LC translocation . We established that the di-modular LC-TD is sufficient for productive translocation of active protease ( Figure 4 and Figure 5 ) given that the RBD responsible for cell binding and internalization was unnecessary for BoNT/A LC translocation ( Figure 4B ) . These results are similar to those obtained for the translocation of diphtheria toxin catalytic domain by its translocation domain [39] . The absence of the RBD confers to the LC-TD a wider pH range for translocation activity ( Figure 4 ) . Having identified this minimal entity , the next set of questions is what this tells us about the BoNT tri-modular design in the context of cellular toxicity . The following inferences can be derived from our analysis . The RBD determines targeting of BoNT to the peripheral nervous system and insures efficient intoxication by at least three mechanisms . In foodborne botulism , the RBD binds receptors on the mucosal surface of gut epithelial cells independent of associated BoNT complex proteins [13] , [40] . The holotoxin then undergoes receptor mediated endocytosis and transcytosis with subsequent delivery to the basolateral side of the epithelial cell [40]–[42] . Once released into the circulation , BoNT reaches cholinergic nerve terminals and a second round of cell entry occurs . BoNTs enter sensitive neurons via receptor-mediated endocytosis determined by its high affinity interaction with a surface protein receptor and a ganglioside co-receptor [1] , [7] , [8] , [10]–[13] . During cell binding and intracellular traffic , the RBD restricts the TD from membrane insertion until its residence within the acidic interior environment of endosomes . Without the RBD , the TD readily inserts into the plasma membrane of neuronal cells and forms channels ( Figure 3 ) . For the LC-TD , the LC remains tethered to its TD carrier on the cell surface by the disulfide linkage , and folded in the extracellular neutral pH environment . Accordingly , the RBD serves to chaperone the LC and TD , insuring that partial unfolding of the LC is concomitant with TD channel formation thereby promoting productive LC translocation . A similar example of a toxin RBD-receptor interaction regulating the pH threshold required for pore formation has been observed for anthrax toxin [43] and likely reflects intoxication requirements shared between the two toxins . The lack of structural similarity between BoNT and anthrax toxin [44] , however , suggests that the mechanisms by which this regulation occurs will be unique . A LC-RBD di-modular protein has not been characterized thus far . We conjecture that such an entity would bind to peripheral neurons and undergo receptor-mediated endocytosis; however , in the absence of concurrent cargo translocation , the LC protease would be irreversibly inactivated by permanent exposure to the acidic endosome interior . At neutral pH , why does the LC-TD translocate from the plasma membrane in cell-based neurotoxicity assays but not in the single molecule studies ? To understand this apparent discrepancy , we first consider the properties of the individual BoNT modules in the context of neuronal cells . The isolated LC , a globular protein with a thermolysin-like fold at neutral pH [4] , [45] , does not enter neurons to cleave its cytosolic substrate [1] , [23] . Multiple studies have demonstrated that , upon exposure to acidic pH , the LC partially unfolds and loses catalytic activity , whereas it remains folded and enzymatically active at neutral pH [23] , [46] , [47] . The LC requires the TD for access to the cytosol , but the conductance of TD channels suggests a pore size of ≤15 Å in diameter [23] , [25] , [34] , [35] . Therefore , the LC must enter an acidic environment to translocate into the cytosol . In the absence of RBD , the TD inserts into the membrane of Neuro 2A cells and would be open at the negative membrane potentials prevalent at the cell membrane ( Figure 4B ) . Thus , the TD channel would dissipate the electrochemical gradients across the plasma membrane , potentially disrupting normal cellular function and activating both the constitutive and regulated endocytosis pathways to recover cellular homeostasis . By usurping the endocytotic machinery and entering acidic endosomes , translocation of the tethered LC could then proceed [23]–[25] . Completion of translocation and release of the LC requires the reduction of the disulfide linkage [23]–[25] . Consistent with this notion is the finding that cells exposed to either isolated LC or single chain LC-TD preserve intact their endogenous SNAP-25 content ( Figure 5 ) . Furthermore , and in accordance with our findings , is the documented ability of LC-TD to induce paralysis in mice at relatively high protein concentrations [33] . The fact that BoNT/A LC-TD intoxicates neurons in the absence of the neuronal targeting RBD implies that LC-TD may be developed into a molecular device with a broad spectrum of cellular specificity . An initial evaluation of the translocation activity of LC-TD in a number of non-neuronal cell lines , which thus far have included CHO ( derived from Chinese hamster ovary cells ) and Vero ( derived from monkey kidney epithelial cells ) cells , show that these cells do not display translocation activity . The difference between these cultured epithelial cells and neuronal cells is likely to be more complex than just the absence of a toxin receptor and co-receptor , as the endo/exocytotic machinery and the membrane lipid composition may vary in substantial and deterministic ways . This issue constitutes a new line of investigation which may provide unsuspected insights into pathways or components involved in neuronal trafficking or recycling of neurotransmitter vesicles . The findings presented here highlight the molecular design of BoNT , a modular machine in which each component functions individually yet their tight and concerted interplay implies that each domain serves as a chaperone for the others . The RBD insures that TD channel formation occurs concurrently with LC unfolding to a translocation competent conformation ( Figure 4B and 4C ) . At the positive membrane potential prevailing across endosomes the TD channel would be closed until it is gated by the LC to initiate its translocation into the cytosol [34] . The TD itself protects the LC within the acidic milieu of the endosome interior , chaperones the LC to the cytosol , and releases the LC in an enzymatically active conformation to access its cytosolic substrate SNARE proteins [23] . This modular design of BoNT and similar modular toxins [44] is therefore emerging as a tool for biomolecule delivery to predetermined target cells [38] , [48]–[54] . Model cargo proteins have been tethered to enzymatically inactive BoNT and demonstrated to translocate and function within the neuronal cell [51] . These studies have focused on inactivation or removal of the enzymatic domain and tethering of additional cargo proteins . By contrast , replacement of the neuronal targeting RBD with one that recognizes a distinct , unique cellular surface protein ( for reviews see [2] , [44] , [52]–[54] ) could transform BoNT TD into a widespread delivery system for a diverse array of cargo proteins to the target tissue of choice , provided cargo proteins reversibly unfold and refold at the beginning and the end of translocation thereby retaining a tight association with the protein-conducting channel throughout the process .
|
Botulinum neurotoxin , widely acknowledged to be the most potent toxin known , is a modular nanomachine and a marvel of protein design . This neurotoxin exploits a modular design to achieve its potent toxicity , which relies on one of its modules—the heavy chain channel—to operate as a specific protein translocating transmembrane chaperone for another of its component modules—the light chain protease . Our study shows that a di-modular protein consisting only of protease and translocation domains is sufficient to translocate active cargo into the cytosol of target cells . The receptor binding domain is dispensable for channel activity or LC translocation; however , it regulates the pH threshold of channel insertion into the membrane . The botulinum neurotoxin modular design embodies a tool for biomolecule delivery to predetermined target cells .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"biophysics/protein",
"folding",
"cell",
"biology/cell",
"signaling",
"biotechnology/protein",
"chemistry",
"and",
"proteomics",
"biochemistry/protein",
"folding",
"neurological",
"disorders",
"cell",
"biology/membranes",
"and",
"sorting",
"physiology/cell",
"signaling",
"neuroscience/neuronal",
"signaling",
"mechanisms",
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"biophysics/membrane",
"proteins",
"and",
"energy",
"transduction",
"biophysics",
"biochemistry/membrane",
"proteins",
"and",
"energy",
"transduction"
] |
2008
|
Botulinum Neurotoxin Devoid of Receptor Binding Domain Translocates Active Protease
|
Race , specifically African ancestry , and obesity are important risk factors for uterine fibroids , and likely interact to provide the right conditions for fibroid growth . However , existing studies largely focus on the main-effects rather than their interaction . Here , we firstly provide evidence for interaction between categories of body mass index ( BMI ) and reported-race in relation to uterine fibroids . We then investigate whether the association between inferred local European ancestry and fibroid risk is modified by BMI in African American ( AA ) women in the Vanderbilt University Medical Center bio-repository ( BioVU ) ( 539 cases and 794 controls ) and the Coronary Artery Risk Development in Young Adults study ( CARDIA , 264 cases and 173 controls ) . We used multiple logistic regression to evaluate interactions between local European ancestry and BMI in relation to fibroid risk , then performed fixed effects meta-analysis . Statistical significance threshold for local-ancestry and BMI interactions was empirically estimated with 10 , 000 permutations ( p-value = 1 . 18x10-4 ) . Admixture mapping detected an association between European ancestry and fibroid risk which was modified by BMI ( continuous-interaction p-value = 3 . 75x10-5 ) around ADTRP ( chromosome 6p24 ) ; the strongest association was found in the obese category ( ancestry odds ratio ( AOR ) = 0 . 51 , p-value = 2 . 23x10-5 ) . Evaluation of interaction between genotyped/imputed variants and BMI in this targeted region suggested race-specific interaction , present in AAs only; strongest evidence was found for insertion/deletion variant ( 6:11946435 ) , again in the obese category ( OR = 1 . 66 , p-value = 1 . 72x10-6 ) . We found nominal evidence for interaction between local ancestry and BMI at a previously reported region in chromosome 2q31-32 , which includes COL5A2 , and TFPI , an immediate downstream target of ADTRP . Interactions between BMI and SNPs ( single nucleotide polymorphisms ) found in this region in AA women were also detected in an independent European American population of 1 , 195 cases and 1 , 164 controls . Findings from our study provide an example of how modifiable and non-modifiable factors may interact to influence fibroid risk and suggest a biological role for BMI in fibroid etiology .
Uterine leiomyomata , also referred to as fibroids , are benign growths arising from myometrial smooth muscle cells . As the most common pelvic tumor in women , prevalence of fibroids ranges from 20 to 77% [1–5] , and accounts for $9 . 4-$34 billion dollars annually in healthcare costs [5–7] . As the leading indication for hysterectomy ( 33% ) in US women of reproductive age , it represents a significant source of burden for women and the healthcare system [8] . There are large disparities in fibroid risk across racial and ethnic populations [8] . Compared with European American ( EA ) women , African American ( AA ) women are two to three times more likely to be diagnosed with fibroids [2;3] , which are also larger in size and greater in number [2;3;9] . AAs also have an approximately 10 year earlier onset of fibroids [4] , are more likely to have hysterectomy and seven-fold more likely to have myomectomies for the treatment of fibroids [10] . The role of genetic predisposition in this disparity is supported by two admixture mapping studies of AAs which demonstrated that greater proportion of European ancestry was inversely associated with fibroids in AA women [11;12] . Obesity is associated with higher fibroid risk with most studies reporting a positive but non-linear relationship with categories of body mass index ( BMI ) , an association that may be mediated by elevated bioavailable estrogen and/or testosterone associated with obesity [13–18] . The role of endogenous sex hormones in the etiology of fibroids is widely accepted , with factors related to higher cumulative exposure showing increased risk , such as greater age at menopause , after which fibroid risk decreases , and earlier age at menarche [19] . Interestingly , the Uterine Fibroid Study ( UFS ) showed positive associations between categories of BMI and fibroids , irrespective of fibroid number or size in black women , but not in white women [18] . Non-modifiable factors such as race/ethnicity and genetics , and modifiable risk factors such as obesity likely interact together to provide the right conditions for fibroid growth . However , most existing studies relating to fibroids have largely focused on these factors individually rather than their interaction . Recognizing this crucial gap in the literature , the primary goal of this study was to evaluate interactions between local genetic ancestry across the genome and BMI in relation to fibroid presence in AA women from the Vanderbilt University Medical Center ( VUMC ) Synthetic Derivative ( SD ) electronic medical record ( EMR ) database and bio-repository ( BioVU ) , and the Coronary Artery Risk Development in Young Adults ( CARDIA ) study .
Comparing women with data available on fibroid status and genotype information in BioVU and CARDIA , cases were more likely to be obese ( 59 . 2% and 58 . 7% in BioVU and CARDIA , respectively ) than controls ( 50 . 7% and 47 . 3% in BioVU and CARDIA , respectively ) ( Table 1 ) . Average age at fibroid diagnosis was 40 . 6 in the BioVU and 40 . 0 in the CARDIA; age distribution of controls was similar to those of cases ( Table 1 ) . On average , cases had lower proportion of average European ancestry across the genome than controls ( Table 1 ) . Similar trends of lower European ancestry in cases versus controls were observed when median European ancestry was visualized by strata of BMI category ( Fig 1a–1d ) . Distribution of characteristics of cases and controls in the larger SD and CARDIA datasets ( including women with and without genotype data ) were similar to those described above ( Table 1 ) . AA race was positively associated with fibroids in both the SD and CARDIA , as expected ( Table 1 ) . AA women in BioVU with genetic data available were comparable to AA women in the larger SD , with the exception of age , where average age was more balanced across cases and controls in BioVU than in the larger SD ( S1 Table ) . In the larger SD and CARDIA datasets , we evaluated the association between reported race and fibroid risk stratified by BMI category ( Table 2 ) . In both datasets , EA women were less likely to have fibroids than AA women across all strata of BMI ( Table 2 ) . However , in the SD , this inverse association strengthened monotonically with increasing strata of BMI until the last strata . The magnitude of the inverse association between race and fibroid presence was smallest in the normal-weight ( <25 kg/m2 ) strata ( Odds Ratio ( OR ) = 0 . 53 , 95% Confidence Interval [CI] 0 . 43 , 0 . 64 ) , and largest in the obese ( 30–35 kg/m2 ) strata ( OR = 0 . 36 , 95% CI 0 . 30 , 0 . 43 ) . Formal tests for interaction using the likelihood ratio test showed evidence for effect modification between categories of BMI and reported race ( P = 0 . 01 ) . Visual evaluation of the interaction in the SD by strata of race , with odds ratios representing odds of fibroids across categories of BMI showed a steeper slope for increased odds in African Americans than for European Americans , the source of interaction ( S1 Fig and S2 Table ) . With far fewer cases and controls per strata compared to the SD , this trend was not similar in CARDIA . We then evaluated the association between genetically inferred global ancestry and uterine fibroids in the subset of African American women in the SD with genetic data available ( BioVU ) and in CARDIA . Every 10% increase European ancestry was associated with 0 . 88 ( 95% CI: 0 . 80 , 0 . 97 ) and 0 . 86 ( 95% CI: 0 . 71 , 1 . 03 ) decreased odds in uterine fibroids in BioVU and CARDIA , respectively ( S3 Table ) . European ancestry was inversely associated with uterine fibroids across each strata of BMI category , however , there was no significant trend for interaction notable for global ancestry ( S3 Table ) . We then evaluated interactions between BMI ( continuous ) and local European ancestry across the genome ( inferred through genetic data ) in relation to fibroids in BioVU AA and CARDIA AA women ( Fig 2a ) and then performed a fixed-effects meta-analysis ( Fig 3 ) . Local ancestry estimates in the chromosome 6p24 region showed strongest evidence of interaction with BMI in relation to fibroids ( Fig 3 and S2 Fig ) . Although no interaction surpassed the canonical genome-wide significance threshold of 5x10-8 , the strongest signal in the chromosome 6p24 region passed empirically estimated statistical significance threshold through 10 , 000 permutations ( p-threshold = 1 . 18x10-4 ) . Interaction ORs ( BMI x local ancestry estimate ) from meta-analysis at marker rs6457825 was 0 . 95 ( P = 3 . 75x10-5; P for heterogeneity [Het-P] = 0 . 91 ) ( Table 3 ) . The association between local European ancestry at this marker and fibroids decreased monotonically with increasing categories of BMI ( Table 3 ) ; the strongest admixture mapping signal was observed in the obese ( BMI >30 kg/m2 ) category ( Fig 4 ) . Each unit increase in European ancestry at this marker was associated with 0 . 51 ( P = 2 . 23x10-5; Het-P = 0 . 50 ) reduced odds of fibroids in the obese category ( Table 3 ) . Investigation of single nucleotide polymorphism ( SNP ) and BMI interactions ( as a continuous term ) using imputed/genotyped variants in the chromosome 6p24 region in AA specific datasets showed 1 , 735 SNPs had p-value less than 0 . 05 ( Fig 1b ) , with the most-statistically significant SNPs hovering around the ADTRP gene ( Fig 5a ) . Interaction analyses reflecting test-for-trend across categories of BMI , and stratified analyses suggested an insertion/deletion ( indel ) variant was associated with fibroid risk and the association differed by BMI strata ( Table 4 ) . Compared to the reference allele ( CTT ) , each additive unit of the effect allele ( C ) for variant chr6:11946435 , located in the Androgen Dependent TFPI Regulating Protein ( ADTRP ) gene was positively associated with fibroid presence in the obese category ( Meta-analysis OR = 1 . 66; P = 1 . 72x10-6; Het-P = 0 . 83 ) ( Table 4 and Supplemental S3 Fig ) . Consistent with the direction of association with local ancestry estimates in this region , the allele frequency for the effect allele ( here , also the allele associated with greater odds for fibroids ) is higher in the African populations ( 73% ) than in the European populations ( 17% ) from the 1000 Genomes reference panel . Conditioning this association on local ancestry at marker rs6457825 attenuated the OR in the obese strata from 1 . 66 to 1 . 46 ( P = 7 . 70x10-4 ) . Of the 1 , 735 BMI-SNP interactions with p-value less than 0 . 05 in the AA specific meta-analyses , 1 , 213 SNPs were available and eligible for analysis in an independent imaging-confirmed replication dataset of European American ( EA ) women ( 1 , 195 cases and 1 , 164 controls ) from BioVU . Approximately 7 . 8% of these SNPs ( 94 SNPs ) had interaction p-values less than 0 . 05 in the BioVU EA set ( Fig 1b ) . Peaks present in AA-specific interaction analyses were attenuated when BioVU EA interaction estimates were meta-analyzed together with AA-specific estimates ( Fig 5a and 5b ) , as the magnitude of interaction estimates were also attenuated in BioVU EA for the top hits in this region ( Supplemental S5 Table ) . The second strongest signal in the genetically inferred local ancestry by BMI ( continuous ) interaction analyses in AA women was found in the chromosome 2q31-32 region at marker rs12999125 ( Meta-analysis OR = 1 . 04; P = 2 . 29x10-4; Het-P = 0 . 77 ) ( Table 5 , S4 , S5 and S6 Figs ) . Further investigation of this signal in analyses stratified by BMI category showed that European ancestry was inversely associated with fibroid risk in the normal weight category ( Meta-analysis OR = 0 . 55; P = 5 . 79x10-2; Het-P = 0 . 92 ) and that this association trended from inverse to positive across increasing categories of BMI ( Table 5 ) . Investigation of SNP and BMI interactions ( as a continuous term ) using imputed and genotyped variants in the chromosome 2q31-32 region in the AA specific datasets revealed 2 , 025 SNPs that had p-value less than 0 . 05 , of which only 1 , 253 were available and eligible in the BioVU EA replication set . Approximately 15% of these ( 183 SNPs ) also had interaction p-values less than 0 . 05 in the BioVU EA set ( Fig 1b ) and meta-analysis of AA and EA specific sets strengthened ( Fig 6b and Supplemental S5 Table ) the relatively weak signals observed in the AA-specific meta-analysis ( Fig 6a ) . Interaction analyses as a continuous term , as a test-for-trend across categories of BMI , and stratified analyses showed SNP rs71430182 had the most consistent evidence across all three datasets as well as across all three methods of interaction assessed; strongest evidence was obtained with the BMI ( continuous ) x SNP interaction method ( p = 7 . 15x10-5 ) ( Table 6 ) . Consistent with evidence in the local-ancestry-BMI interaction analyses , effect allele G was inversely associated with fibroids in the normal-weight ( BMI < 25 kg/m2 ) category and the inverse association approached the null across increasing categories of BMI ( Table 6 ) . The effect allele is found in greater frequency in the European reference population ( CEU EAF: 0 . 86 ) compared with their African counterpart ( YRI EAF: 0 . 61 ) . Conditioning the SNP-BMI interaction on local ancestry at marker rs71430182 for the AA specific analyses did not attenuate the effect estimate . For example , meta-analysis OR for AA specific analyses ( BMI categories coded as 0 , 1 , 2 x SNP ) decreased slightly from 1 . 37 to 1 . 36 when adjusted for rs71430182 .
The genetic basis of uterine fibroid risk and racial disparity in uterine fibroids are not well understood . Additionally , understanding of how modifiable risk factors such as obesity interact with non-modifiable risk factors such as race and genetic ancestry to influence uterine fibroid etiology is minimal . In this study , we first showed evidence of effect modification between BMI and reported race using the SD . Compared with AA women , EA women were less likely to have fibroids and effect sizes were stronger in the heavier BMI categories , with the strongest association found in the obese ( BMI 30–35 kg/m2 category ) . Then using local ancestry estimates inferred from GWAS data in AA women , we show that BMI modifies the association between local European ancestry and fibroid risk in AA women from the BioVU and CARDIA study in two different genetic regions: chromosome 6p24 and chromosome 2q31-32 . Further evaluation for evidence of interaction between SNPs around the chromosome 6p24 region and BMI suggested the signals were race-specific ( only present in AA women , but not in EA women ) , which could be due to allelic heterogeneity and/or reduced detection power due to differences in allele frequency for SNPs in this region . However , in chromosome 2q31-32 , although we detected suggestive evidence of interaction between BMI and local ancestry followed by suggestive evidence of interaction between BMI and SNPs in the region in AA women , incorporating EA women strengthened evidence for interaction between BMI and SNPs suggesting a common mechanism of action for this region across these two populations in relation to fibroids . ADTRP is an androgen dependent gene [20] that regulates the expression of Tissue Factor Pathway Inhibitor ( TFPI ) [21;22] , the gene product of which is a natural antagonist of Tissue Factor ( TF ) . TF , a trans-membrane glycoprotein , is the main trigger of the blood coagulation cascade [23] . Non-membrane bound isoforms of TF have been shown to trigger angiogenesis and anti-apoptotic activity [24] . TF and TFPI proteins are found in the myometrial tissue [25] . With regard to obesity , TF concentration in the blood is higher in obese individuals , and has been shown to be reversible with weight reduction [26] . Additionally , considering androgen dependency of ADTRP mRNA expression [20] , our finding of the strongest association in obese women is paralleled with the evidence that obese women have higher average circulating levels of testosterone compared with non-obese women [26] . Furthermore , our results extend etiological evidence to findings from a recent study that showed a positive association between higher free-testosterone levels and incident uterine fibroids , but an inverse association with recurrent fibroids [27] . The mechanistic relationships between obesity , androgens , ADTRP , TF and TFPI have not been studied in relation to uterine fibroids , and the relationships we detect here require additional functional studies . The second BMI-local ancestry interaction that we noted in chromosome 2q31-q32 region , although only marginally significant , harbors several genes with relevance to fibroids including COL3A1 and COL5A2 as the closest genes ( S5 Fig ) . One of the widely hypothesized models for fibroid formation has been related to abnormal tissue repair , disordered healing and altered extracellular matrix formation in parallel with keloid formation . Over expression of COL3A1 mRNA [28–30] , a key component in extracellular matrices , along with over-expression of COL5A2 mRNA and higher presence of irregularly aligned collagen fibrils have been noted in fibroid-tissue compared with normal myometrial tissue [30] . Furthermore , both these genes have been found to be highly expressed in transformed fibroblasts in the Genotype-Tissue Expression ( GTEx ) Project database . Studies have shown abnormal over-expression of basic fibroblast growth factor ( bFGF ) mRNA [31] and bFGF ligand-receptor [32] in fibroid tissues compared with matched normal-myometrial tissue . In addition to COL3A1 and COL5A2 , it is noteworthy that TFPI also lies within this admixture mapping peak and is within 1 . 75 Mb of the top chromosome 2 signal ( S5 Fig ) . The observation that the strongest signal in this analysis ( ADTRP ) is a potent activator of the TFPI gene , which is the second strongest signal in the analysis further strengthens the hypothesis that obesity and testosterone related pathways may modify fibroid risk through these genes . Further adding evidence for TPFI as a causal candidate gene for uterine fibroids , a study comparing gene expression between uterine fibroid tissue and normal myometrial tissue reported 3 . 9 times lower expression of the TFPI gene in fibroid tissues , on average [33] . We took several measures at the design and analysis stages of this study to ensure internal and external validity of study findings . First , we designed a novel approach for gene discovery at various levels of ancestry variables to find evidence: reported-race , global-ancestry and local ancestry . We used two independent data sources to provide interval validation of study results , where possible . We implemented several steps to reduce misclassification of fibroid status , an important and difficult consideration for uterine fibroids research . We limited the choice of our data sources to individuals for who fibroid status was confirmed through imaging modalities . Almost all cases in CARDIA were confirmed though transvaginal ultrasound ( TVUS ) and a few through hysterectomy reports . Importantly , all controls were also confirmed for absence of fibroids through TVUS in CARDIA . An even more rigorous algorithm was applied for the SD/BioVU , where cases were confirmed to have uterine fibroids through imaging reports and controls were required to have two or more imaging reports free of uterine fibroids . To minimize possibility of reverse causality , we considered an average BMI measure for the SD/BioVU participants that was reflective of their adult life up until the time of fibroid diagnosis ( BMI points during pregnancy were excluded ) . For CARDIA , a more traditional cohort design , we took BMI measures at the time of the ultrasound visit , or the preceding visit if absent . Despite these strengths , a few considerations are worth reflecting on while interpreting study results . Our statistical models included , age , principal components , BMI , local ancestry and the interaction term between BMI and local ancestry . We were unable to adjust for other potential confounding factors , such as smoking , parity , age at menarche and oral contraceptive use that may have influenced estimates for the interaction term . However , we considered the implications of this potential limitation thoroughly while designing the study . Simulation studies have shown bias to be a major issue when there is interaction between confounder and gene of interest , but that the association between confounder and outcome would have to be large , as would the interaction between confounder and genetic factor [34] . With the exception of race , we are not aware of other risk factors that have extremely large effect estimates on uterine fibroids . However , due to our inability to adjust for additional effects , albeit likely small effects , we are not able to completely rule out residual confounding . With regard to comparability of cases and controls between the SD and BioVU ( smaller subset of the SD ) , fibroid cases and controls tended to have similar distributions across the SD and BioVU with regard to BMI but not with regard to age . Fibroid controls in BioVU tended to be a slightly older subset of SD controls . Therefore , the BioVU subset is not fully representative of the larger SD . However , this does not disrupt the internal validity of the global and local ancestry results . Instead , by having a slightly older set of controls , potential misclassification of would-be cases as controls is further minimized in the BioVU subset . We did not find a statistically significant interaction between global ancestry and BMI in either BioVU or CARDIA . It is interesting that , although not significant , the direction of interaction across increasing categories of BMI for global ancestry is opposite than that noted for reported-race in the SD and the top hit for local ancestry in BioVU and CARDIA . Global ancestry provides an average of local ancestry associations , and in light of local ancestry results where we report interactions going in two different directions , it is likely that the genetic architecture of uterine fibroids is complicated at the least . In comparing results from imputed regions at the top two loci , it is notable that adjustment for local ancestry attenuated the odds ratio at the chromosome 6p24 region , but negligibly for the chromosome 2q31-32 region . Intuitively , the correlations between local ancestry and genotypes are higher for the chromosome 6p24 region which also had the greater ancestral allele frequency difference ( 0 . 17 , and 0 . 73 , in EUR and AFR respectively ) than for the top SNP in the chromosome 2q31-32f region ( 0 . 86 , and 0 . 61 in EUR and AFR , respectively ) . Therefore , adjustment for local ancestry at the top SNP had the largest impact on chromosome 6p24 than in chromosome 2q31-32 region . Several findings from our study are in agreement , at least in part , with previously published findings . The UFS study showed a positive association between BMI and fibroids in AA women , but not in EA women , suggesting effect modification , even though a formal test was not conducted [18] . We also present our effect modification results as evaluations between BMI and fibroids by strata of race ( EA and AA ) ( shown in S2 Table for SD , S6 Table for CARDIA ) . Compared with the normal-weight category , uterine fibroid ORs for overweight ( BMI 25–30 kg/m2 ) and obese ( BMI 30–35 kg/m2 ) EA women were lower than for AA women ( S2 Table , S1 Fig ) . However , contrary to the UFS , effect estimates for the highest BMI category ( BMI >35 kg/m2 ) were similar in EA and AA women . The effect modification was only apparent in the SD , which was considerably larger than the CARDIA sample size ( S6 Table ) . The lack of consistency between the SD and CARDIA may reflect the difference in the types of fibroid cases between the two sets . The SD likely has greater number and proportion of symptomatic fibroid cases seeking care than the CARDIA , where most cases were incidental discoveries through TVUS examinations . In addition to the vast difference in sample sizes between the two studies , the populations are selected in a different manner , where one is based on EMRs , while the other is based on self-selection , people willing to participate in a cohort study . Furthermore , differences in cultural attitudes towards care seeking behaviors , especially for sensitive topics such as uterine fibroids may have further influenced effect estimates when comparing interaction between reported-race across the SD and CARDIA . It is intriguing though that interaction estimates across BioVU and CARDIA for genetically inferred ancestry , whether global or local ancestry , are similar across the two studies . In the same vein of thought , this is likely reflective of using a more objective approach to estimate ancestry within a group of women ( African American women , regardless of the source population BioVU or CARDIA ) , instead of comparing women between reported-races . To our knowledge , this is the first admixture mapping study to directly evaluate effect modification of the association between genetic ancestry and fibroids by BMI . Two additional admixture mapping studies have been previously conducted for uterine fibroids in AA women [11;12] . Similar to the two previous reports from the Black Women’s Health Study ( BWHS ) and the UFS , mean differences in average genetic European ancestry between cases and controls in BioVU and CARDIA ranged from 1 . 5% to 2 . 7% excess European ancestry in controls than in fibroid cases . Using BWHS , Wise and colleagues reported suggestive and statistically significant peaks in chromosomes 2 , 4 and 10 , with varying directions of risk associated with higher local European ancestry proportion [11] . Using the UFS , Zhang and colleagues reported suggestive associations between local ancestry and fibroids in chromosome 1q42 . 2 and 2q32 . 2 , with one locus suggesting a positive association and the other suggesting a negative association between local European ancestry and fibroids [12] . As the third admixture mapping study relating to fibroids , our study corroborates suggestive evidence presented by the previous two studies for the chromosome 2q32 . 2 region and highlights the importance of further investigating this region . The nearest markers reported for this region are within 1Mb in proximity ( rs256552 reported by Zhang et al . ) and within 2 . 5 Mb ( rs6710083 reported by Wise et al . ) of the signal observed in this study ( S6 Fig ) . Even though both these studies report excess European ancestry in uterine fibroid cases in chromosome 2q32 compared with the genome-wide average in cases , which is not in complete agreement with our observation , this is the only replicated region identified thus far with regards to admixture mapping . More notably , Zhang et al . show an increasing Z-score trend across increasing categories of BMI , where the effect of local European ancestry would again increase fibroid risk . Although the reason for discrepancy in the direction of association between this study and others is not completely clear , differences in study designs may provide a reasonable explanation . Previous admixture mapping studies used case-only approaches , which compute Z-scores at local ancestry in relation to global average , whereas we used a case-control design that allowed us to adjust for potential confounders in statistical models . Additionally , the primary focus of our study was different from previous studies; we focused on the interaction between BMI and local ancestry in relation to fibroids , while previous studies focused on association between local ancestry and fibroids . Availability of comparable results from previously published studies , evaluating local ancestry and BMI interaction in a case-control setting adjusted for global ancestry may facilitate future comparisons and may provide more insight into observed differences . As the first admixture mapping study that formally evaluated interactions between local ancestry and BMI in relation to fibroid risk , we provide statistically significant evidence for interaction in the ADTRP gene and suggestive evidence for its immediate down-stream target , TFPI , in two independent samples of AA women from the BioVU and the CARDIA study . Our study further highlights the power and flexibility of admixture mapping to identify risk loci for uterine fibroids that are modified by obesity among AA women , a population that is likely at highest risk for fibroids . Further confirmation of these findings and further characterization of the mechanisms involved may suggest therapeutic approaches for this high risk population .
Participants for this investigation were derived from the SD EMR database , located at VUMC , Nashville , TN , and from the CARDIA study . The Institutional Review Board at VUMC approved this study . Both the SD database and CARDIA study have been described in detail in previous publications [35;36] . Briefly , the SD consists of de-identified clinical data obtained from patients attending all clinics associated with the Vanderbilt University Medical Center hospital system . Clinical data are abstracted from multiple sources including diagnostic and procedure codes , basic demographics , discharge summaries , progress notes , health history , multi-disciplinary assessments , laboratory values , imaging reports , medication orders , and pathology reports . Women of age 18 years or older , with at least one diagnostic or procedural code for pelvic imaging in the SD , were considered to be eligible for case-control selection in this investigation [37] . CARDIA is a prospective cohort study that recruited 5 , 115 EA and AA participants ( 54 . 5% female ) between 18–30 years of age at baseline from the years 1985–86 in four clinical centers ( Birmingham , AL; Chicago IL; Minneapolis , MN; and Oakland , CA ) in the US . Participant characteristics were collected and various health outcomes were measured at baseline and during follow-up visits at years 2 , 5 , 10 , 15 , 20 , 25 and 30 [35] . Additionally , at year 16 , the CARDIA Women’s Study ( CWS ) , performed ancillary to the existing study , administered TVUS examinations to non-pregnant women who had attended the year 15 exam and had at least one intact ovary by self-report . Women with TVUS examination at year 16 or with additional information on fibroid presence in previous visits were the target population of this investigation . Analyses were first conducted using the SD , then replicated using CARDIA samples , followed by a meta-analysis where applicable . Cases and controls in the SD were selected using an algorithm described in detail previously [37] . Women who are at least 18 years of age with EMR data in the SD with at least one procedural code for imaging with ultrasound , magnetic resonance imaging , or computed tomography were eligible for selection . Women with at least one diagnostic/procedure code for fibroids , defined by the International Classification of Diseases 9 ( ICD-9 ) or by the Current Procedural Terminology ( CPT ) criteria , were considered cases . To be considered as controls , women needed to have pelvic imaging codes in at least two different time points with no ICD-9 , or CPT code indicative of fibroid presence , and no mention of fibroids related key words or hysterectomy related keywords in imaging reports , operative reports , pathology summaries , or in the Problem List in patient file . We identified 1 , 314 AA cases , 2 , 697 AA controls , 2 , 699 EA cases and 10 , 161 EA controls in the SD with or without genotyping data available . In the CWS , for women who received a TVUS examination , trained staff recorded information on the number of fibroids , and largest fibroid dimension from three perpendicular planes . For this investigation , women with one or more fibroids of any size were considered cases and women with no fibroids as controls . Additionally , several women who were eligible for CWS but had indicated fibroids as a reason for hysterectomy from baseline till year 15 were considered cases . In CWS , there were 402 AA cases ( 88 self-reported hysterectomy ) , 148 AA controls , 209 EA cases ( 15 self-reported hysterectomy ) and 300 EA controls , with or without genotyping data available . For participants in the SD , age at first diagnosis of fibroids or first pelvic imaging with fibroids was noted for cases and age at last pelvic imaging without mention of fibroids was noted for controls . BMI was computed as the average BMI starting at age 18 and up until the time of fibroid diagnosis for cases and up until the last pelvic imaging without fibroids for controls . BMI during pregnancy was excluded from average calculations . The BMI variable computed from the SD thus reflects the mean non-pregnant BMI in cases and controls over time . Race was coded as AA or EA based on third-party identification in the SD EMR . For women who were part of the CWS , age was recorded and BMI was directly measured for cases and controls during the study visit at year 16 . For the small proportion of cases that were identified by self-reported reason for hysterectomy , age and BMI were determined using CARDIA visits that were closest to the time of hysterectomy . Race was recorded as AA or EA based on self-report at baseline in CARDIA and validated at the Year 2 examination . Fibroid cases and controls identified as AA or EA in the SD with DNA available and proper consent to use DNA for research ( BioVU ) were genotyped in the Affymetrix Axiom Biobank Genotyping Array and the Affymetrix Axiom World Array 3 ( Affymetrix Inc . , Santa Clara , CA , USA ) . CARDIA cases and controls were genotyped with the Affymetrix Genome-Wide Human single nucleotide polymorphism ( SNP ) Array 6 . 0 ( Affymetrix Inc . , Santa Clara , CA , USA ) . Standard QC measures were taken for genotyping data from AA women for both datasets using PLINK [38] . Individuals with low genotyping rates ( <95% ) were removed from consideration followed by SNPs with low genotyping quality ( <95% ) . Individuals with inconsistency in reported versus genetically determined sex were excluded , followed by individuals with first degree or higher relatedness identified by identity-by-descent sharing from a random selection of approximately 100 , 000 autosomal SNPs . SNPs with minor allele frequencies less than 1% and SNPs that deviated from the Hardy-Weinberg equilibrium at p <10−6 threshold ( in controls ) were excluded . Upon QC completion , limiting to individuals with complete information on key covariates , there were 1 , 233 AA women ( 539 cases and 694 controls ) and 410 AA women ( 264 cases and 146 controls ) available for primary analysis in the BioVU and CARDIA , respectively . Additionally , data were available from 2 , 359 EA women from the BioVU ( 1 , 195 cases and 1 , 164 controls ) after QC which were used for replication of interaction estimates for select markers from candidate regions , described below in the Statistical analysis section . We first identified consensus autosomal SNPs present in both the BioVU and CARDIA post-QC datasets . For consensus SNPs , we further limited inclusion to SNPs with allele-frequency differences ( delta ) >0 . 2 between the 1000 Genomes African and European populations . A total of 20 , 000 SNPs were identified , which were then used for local and global ancestry estimation . Local ancestry estimation was performed using Local Ancestry in admixed Populations Ancestry ( LAMP-ANC ) with proxy ancestral allele frequency inputs for SNPs from Europeans and Africans to infer local ancestry across the genome [39] . We used allele frequency estimates from the African and European populations nested in the Phase 3 1000 Genomes reference panels as ancestral allele frequencies for Africans and Europeans for the 20 , 000 consensus markers described above [40] . The following settings and parameters were assumed: LD-pruning was set to r-squared value of 0 . 1 , recombination rate was set to 1x10-8 , time since admixture was assumed to be seven generations and the proportion of admixture estimates was set to 0 . 2 for European ancestry . For the resulting output from LAMP-ANC , local ancestry was then coded as 0 , 1 or 2 European ancestry calls for each marker . Global ancestry ( average European ancestry per individual ) was then calculated by summing the local ancestry calls across the genome and dividing by the total number of markers used in ancestry estimation . Local and global ancestry estimates were inferred separately for BioVU and CARDIA . Comparison of the top principal component from EigenSoft [41] with global ancestry estimates showed strong correlations , 98% in BioVU and 99% in CARDIA ( S7a & S7b Fig ) . We used multiple logistic regression to evaluate the association between reported race and fibroid risk in EA and AA women by strata of BMI categories ( BMI < 25 kg/m2 , 25–30 kg/m2 , 30–35 kg/m2 , and > 35 kg/m2 ) in the SD and CARDIA datasets , separately . Interaction between BMI and race was evaluated by performing the likelihood ratio test in StataIC , version 12 ( StataCorp , College Station , TX , USA ) , obtained by comparing the following reduced and full models . In the reduced model uterine fibroids was modeled against race ( 0 for African American and 1 for European ) , and k-1 BMI category indicator variables ( BMI 25–30 kg/m2 , 30–35 kg/m2 , and > 35 kg/m2 , with BMI <25/m2 serving as the reference variable ) , adjusted for age . The full model included variables in the reduced model and also included interaction terms between race ( 0 , and 1 ) and the three BMI indicator categories to provide 3-degrees of freedom for the likelihood ratio test . Estimates for the association between race and uterine fibroids for individuals in each BMI category was obtained from the full model . Equivalently , the association between categories and BMI and uterine fibroids for individuals in each race category were also obtained from the full model . A flow chart detailing primary study populations by analysis type and availability of GWAS data is shown in Fig 2a . To evaluate the potential impact of including cases based on self-reported hysterectomy due to fibroids with TVUS confirmed cases , we performed sensitivity analyses by excluding TVUS non-confirmed cases . As interaction estimates and trends were similar for both types of analyses ( S7 Table ) we opted to include fibroid cases ascertained by hysterectomy status in the remaining analyses . For primary analyses , in the smaller subset of BioVU AA and CARDIA AA women for whom local ancestry estimates were inferred , we performed interaction analyses between BMI ( continuous ) and local ancestry estimates ( additive model: 0 , 1 or 2 copies of European ancestry ) across the genome in a multiple logistic regression framework while adjusting for age , and first 10 principal components ( Eigensoft ) , using PLINK software [38] . All analyses were performed separately for each of the two datasets . Resulting effect estimates were then aggregated using inverse variance-weighted fixed effects meta-analysis in METAL [42] . Threshold for statistical significance for admixture mapping interaction analyses was estimated using 10 , 000 min-p permutation tests . Briefly , case-control status was randomly shuffled in each iteration to break the association between markers/ancestry loci and outcome of interest in BioVU and CARDIA separately . Additionally , case-control status and the first principal component were paired and shuffled together in order to maintain the correlation between global ancestry and case-control status . Beta and standard errors resulting from each iteration were then aggregated using fixed-effects meta-analysis . The smallest interaction p-value across all markers from the meta-analysis in each iteration was stored and repeated for 10 , 000 iterations to yield 10 , 000 min-p test statistics . The p-value at the 5th percentile of the rank-ordered statistics across the 10 , 000 test statistics constitutes the empirically estimated threshold for statistical significance ( P = 1 . 18x10-4 ) . Statistically significant and suggestive ( P = 5x10-4 ) signals were further evaluated for the association between local European ancestry and fibroid presence by strata of BMI category ( BMI < 25 kg/m2 , 25–30 kg/m2 , and >30 kg/m2 ) while adjusting for age and first 10 principal components . Then to search for interactions between BMI and SNPs in relation to fibroids , we tested for interactions in the candidate regions using genotyped and imputed SNPs in the BioVU AA and CARDIA AA datasets ( Fig 2b ) . We imputed 5–10 mega-base regions around the BMI-and-ancestry interaction peaks , using the Phase 3 1000 Genomes cosmopolitan reference panel using IMPUTE2 [43] . Genotyped and imputed variants with minor allele frequencies greater than 5% in these regions were then allowed to interact with BMI in the following ways: 1 ) BMI ( continuous ) x SNP for maximal detection power with Probabel [44] , 2 ) BMI category ( 0 , 1 , or 2 , as a continuous term ) x SNP as a test for trend of association for SNP across meaningful BMI categories with Probabel , 3 ) standardized BMI ( per standard deviation ) x SNP , and 4 ) by strata of BMI for ease of interpretation with SNPTESTv2 [45] . Dataset specific analyses were then meta-analyzed using METAL . We estimated p-value threshold ( p = 6 . 9x10-6 ) here by first estimating the effective number of independent tests with simpleM in each dataset ( 7 , 215 and 6 , 885 for BIOVU and CARDIA , respectively ) , and then dividing 0 . 05 by the estimate that was the more conservative of the two . For BMI-stratified analyses , where appropriate , conditional analysis adjusting for local ancestry marker with the strongest signal was conducted to evaluate whether the stratum specific association between SNP and fibroids was independent of local ancestry . Finally , the subset of markers that interacted with BMI ( continuous term ) at a p-value threshold less than 0 . 05 in meta-analysis of AA specific datasets were tested for interaction in an independent dataset of European American women ( BioVU EA ) who were confirmed as fibroid cases ( N = 1 , 195 ) and controls ( N = 1 , 164 ) with pelvic imaging codes . Same tests of interaction were performed as with the AA specific datasets , although BMI ( continuous ) x SNP interaction was used as the primary test for statistical evidence for replication ( p < 0 . 05 ) in BioVU EA .
|
Although it is postulated that obesity and non-modifiable risk factors such as race or genetic ancestry may interact to jointly influence uterine fibroid growth , most existing studies have not evaluated their interaction . In this study we exhibit evidence for interaction across several tiers of investigation . We first show that the association between reported/third-party identified-race ( African American ( AA ) and European American ( EA ) ) and fibroid risk is modified by body mass index ( BMI ) categories . We then reveal evidence for interaction between two genetically-inferred local European ancestry regions ( top two regions: chromosome 6p24 , ADTRP; chromosome 2q31-32 , TFPI , COL5A2 ) along the genome and BMI in relation to fibroids in two independent AA populations . Of intrigue , genes in these top two regions are mechanistically related , where ADTRP gene-product is an immediate upstream regulator for TFPI . Then at the genotype level , we show that interaction between genotyped/imputed variants and BMI is race-specific for chromosome 6p24 region , present in AAs but not in EAs , whereas , trans-ethnic , common across AAs and EAs for the 2q31-32 region , as replicated in an independent population of EA women . Our multi-tiered investigation supports evidence for interaction between reported race , genetic ancestry and BMI in relation to fibroid risk .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"body",
"weight",
"medicine",
"and",
"health",
"sciences",
"african",
"americans",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"ethnicities",
"physiological",
"parameters",
"mathematics",
"statistics",
"(mathematics)",
"chromosome",
"mapping",
"obesity",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"reproductive",
"system",
"procedures",
"gene",
"mapping",
"chromosome",
"biology",
"mathematical",
"and",
"statistical",
"techniques",
"hysterectomy",
"molecular",
"biology",
"people",
"and",
"places",
"cell",
"biology",
"surgical",
"excision",
"physiology",
"meta-analysis",
"body",
"mass",
"index",
"genetics",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"physical",
"sciences",
"human",
"genetics",
"statistical",
"methods",
"chromosomes"
] |
2017
|
African genetic ancestry interacts with body mass index to modify risk for uterine fibroids
|
Feathers have complex forms and are an excellent model to study the development and evolution of morphologies . Existing chicken feather mutants are especially useful for identifying genetic determinants of feather formation . This study focused on the gene F , underlying the frizzle feather trait that has a characteristic curled feather rachis and barbs in domestic chickens . Our developmental biology studies identified defects in feather medulla formation , and physical studies revealed that the frizzle feather curls in a stepwise manner . The frizzle gene is transmitted in an autosomal incomplete dominant mode . A whole-genome linkage scan of five pedigrees with 2678 SNPs revealed association of the frizzle locus with a keratin gene-enriched region within the linkage group E22C19W28_E50C23 . Sequence analyses of the keratin gene cluster identified a 69 bp in-frame deletion in a conserved region of KRT75 , an α-keratin gene . Retroviral-mediated expression of the mutated F cDNA in the wild-type rectrix qualitatively changed the bending of the rachis with some features of frizzle feathers including irregular kinks , severe bending near their distal ends , and substantially higher variations among samples in comparison to normal feathers . These results confirmed KRT75 as the F gene . This study demonstrates the potential of our approach for identifying genetic determinants of feather forms .
Birds have evolved many unique and interesting features , allowing them to adapt and radiate into various ecological niches . They display a great degree of diversity in feathers and other body parts . Domesticated birds exhibit an even greater diversity in phenotypes than their wild ancestors , thus providing an excellent opportunity to explore the genetic basis underlying variation in morphology , physiology , and behavior . As Darwin noted , the domestic chicken displays a remarkable level of phenotypic diversity [1] and it is the most phenotypically variable bird , especially in terms of feather form [2] . However , the genetic and developmental basis of this diversity is unclear . Understanding the genetic basis of plumage variability in the chicken would provide insight into how evolutionary diversification in morphological traits could occur rapidly during adaptive radiations or under strong sexual selection . The development of a feather has to be coordinated by an enormous number of molecular and cellular machineries [3]–[17] . The feather is the most complex keratinized structure of the vertebrate integument and has vital importance for physiological and functional requirements . The complex organization of feathers allows for a variety of potential morphological changes to occur . Modifications of the feather include deterrence of feather development , changes in feather structure , inhibition of feather molting , and alterations of feather growth rates [18] . The structure of feathers includes the rachis ( feather backbone ) , ramus ( branches ) and barbules ( branches off the ramus , which enable them to form an organ capable of moving air to provide flight ) ( Figure 1A ) . In the chicken , embryonic downy feathers are radially symmetric and fluffy and have a very short rachis or none at all . The branches in downy feathers only include the ramus and barbules ( Figure 1A ) . Most adult chicken feathers are bilaterally symmetric and include a rachis , ramus , and barbules ( Figure 1A ) . The rachis and ramus are composed of two layers: the outer , thin cortex ( composed of solidly compacted squamous cells ) and the inner , thick medulla ( composed of empty polyhedral pith cells ) [19] . Various feather types are essential characteristics of domestic chicken breeds . Although the molecular and cellular basis of feather development has been well characterized [3]–[6] , little is known about the genes influencing feather growth , pigment pattern , length , distribution , and structure . The existence of a reference genome sequence has placed the chicken as an important model organism for understanding genome evolution , population genetics , and the genetic basis of phenotypic traits [20]–[33] . Owing to the close relationships within Class Aves , the molecular and genetic understanding of phenotypic variations discovered in chickens are likely to be applicable to wild bird species . Therefore , chicken genetic and genomic studies can provide information for studying development and evolution of avian species [2] , [34]–[42] . Frizzle feathers have been described in domesticated birds and established as varietal characteristics in domestic chickens [18] , [43] . The contour feathers of the frizzle chicken all curl outward and upward . Due to an altered feather rachis structure and morphology , they cannot lie flat against the body . Usually , rectrices and remiges are less affected by the mutation but have an irregular appearance . Other prominent modifications , such as thickening of the barbs and barbules , alteration of the hooklets and other structural abnormalities have also been observed [44] . The frizzle mutation has been reported to occur in a single autosomal gene denoted F that shows incomplete dominant inheritance [44] , [45] . In order to dissect the genetic mechanism underlying frizzle feathers , we conducted a whole genome linkage scan and mapped the causative genetic mutation to the linkage group E22C19W28_E50C23 . By analyzing the candidate genes in the associated interval , we found that the F mutation is caused by a deletion in a conserved region of an α-keratin . The causative effect of the KRT75-MT was confirmed by a retrovirus-mediated misexpression of the wild-type or mutated K75 protein in the feather follicle during regeneration in chickens with normal plumage . Interestingly , mutations in KRT75 have also been identified in mammals , causing structural abnormalities in hair in humans [46] , [47] and mice [48] . This implies a fundamental role for K75 in building the architecture of skin appendages .
The adult frizzle chicken shows a distinct disorientation of feathers ( Figure 1B ) . Upon hatching , the first-generation radially symmetric feathers of frizzle chicks do not show curves ( Figure 1B ) . The frizzle phenotype starts to appear when the first-generation feathers are replaced with second-generation bilateral feathers which have a rachis . At this stage , both body and wing flight feathers twist toward a dorsal orientation . Normally , feathers are bent along the dorsal to ventral orientation . However , in frizzle chickens , the feathers are bent along the ventral to dorsal orientation ( Figure 1C ) . The pennaceous vane on both dorsal and ventral sides of frizzle feathers show normal branching when compared to white leghorn controls ( Figure 1C ) . Their rachis backbones are determined by our computer-aided analyses ( Figure 1D ) . The definitions of “s” and “θ” are shown in the schematic drawing of a feather in Figure 1E . We define the backbone of the rachis by a curve at equal distance to the two edges of the rachis , because rachis has a width . The accumulated distance from the most proximal end of the rachis along the backbone is defined as “s” . We then , arbitrarily draw a straight line as the coordinate , and define the angle between the local tangent line ( the best straight-line approximation to the rachis backbone at that point ) and the straight line as “θ” . Thus the bending of the rachis is represented by the function θ ( s ) , which quantitatively reflect the changing curvature of different feathers ( Figure 1E ) . Features of θ ( s ) quantitatively reflect some subtle differences beyond a visual inspection of the images . For example , the heterozygous chicken feather shows not only kinks along its length but also curves over the entire rachis in comparison to the homozygous chicken whose feathers have dramatically increased curvature and less kinking ( Figure 1F ) . These features may be correlated to the growth of the feather in response to the expression of a different allele . Feather sections from the mature , top region show that the rachis of frizzle feathers has a smaller medulla compared to the normal leghorn controls ( Figure 2A ) . The medulla is localized in the inner , ventral region of the rachis and is composed of empty polyhedral pith cells [19] . These observations suggest that the frizzle phenotype is caused by a defect in the ventral part of the rachis . We examined rachidial morphogenesis at different regeneration time points by plucking body feathers and allowing them to regenerate for 10 days or 30 days . After regeneration , feather follicles were dissected and paraffin sections were prepared . Three different levels of cross-section of a 30-day regenerated sample , from mature ( Figure 2D and S1A , level III ) to immature ( Figure 2D and Figure S1A , level I ) , were studied by H&E staining ( Figure 2B ) . These results indicate that a clear defect in the ventral region of the rachis may be responsible for the altered medulla formation . PCNA ( Proliferating Cell Nuclear Antigen ) staining showed that the cell proliferating zone in the frizzle rachis in level I is much narrower than that of the controls ( Figure 2C , upper panel ) . The TUNEL staining showed there is no cell death in the PCNA positive cell proliferating zone ( Figure 2C , lower panel ) . Detailed PCNA and TUNEL staining ( Figure S1B–S1E ) , summarized in Figure 2D and Figure S1F , show there are no differences in cell proliferation and programmed cell death between normal and frizzle feathers at level II and level III . Our cell proliferation data suggests that the cell proliferation zone at an immature level ( level I ) of the frizzle rachis is narrow compared to that found in a normal rachis , perhaps contributing to smaller medulla formation in the frizzle rachis . Finally , we compared the feathers of homozygous frizzle chickens with white leghorn controls at embryonic day 12 . The base of the feather filament appeared normal . However , the tip of each frizzle feather filament appears to be randomly twisted in both the body and wing feathers ( Figure 2E , upper two panels ) . To examine possible differences between frizzle and normal feather branching morphogenesis , we used whole mount in situ hybridization with a probe targeting SHH which is expressed in marginal plate cell [49] . The frizzle feathers showed the same expression of SHH as controls ( Figure 2E , lower panel ) . This suggests that embryonic frizzle feather branching occurred normally even though the tip of frizzle feathers were randomly twisted . In order to locate the gene underlying the frizzle trait , a genome scan was conducted on progeny of crosses between the same heterozygous frizzle rooster , PF1 , and five different wild-type feathered hens . A total of 2678 SNPs were genotyped in 45 birds and linkage analysis of the genotyping data identified two SNPs , rs16687483 and rs16687610 within the linkage group E22C19W28_E50C23 that yielded a LOD score of 7 . 34 and 6 . 5 , respectively . Haplotype sharing of SNPs between family members identified a shared haplotype extending from rs14689023 to rs16687610 in 21 of 22 frizzle birds . A possible recombination event in the region between rs16687483 and rs16687610 was evident in the frizzle female Y61F ( Figure 3 ) . A cluster of keratin genes was found within the genomic interval to which the frizzle locus was mapped by the above analysis . Mutations in keratins are obvious candidates for altered feather phenotypes [50] , [51] . Keratins purified from the frizzle feather showed a slightly altered amino acid content , produced distinct X-ray diffraction patterns , and exhibited quantitative banding changes on SDS-PAGE gels [52] , [53] . To identify possible causative variants , we PCR-amplified and sequenced partial gene regions of the 14 keratin candidate genes ( Table S1 ) and found only one significant variation in a coding sequence ( GenBank accession number JQ013796 ) , namely , a deletion covering the junction of exon 5 and intron 5 in the KRT75 gene ( chrE22C19W28_E50C23:658 , 389–658 , 472 ) ( Figure S2 , Table S2 ) . This deletion mutation showed complete segregation with the frizzle phenotype in all the frizzle offspring within the F1 generation of the experimental crosses ( Figure S3 and Figure S4 ) . Frizzle chickens sampled from different populations in Taiwan with the distinctive homozygous and heterozygous feather phenotypes demonstrated two mutant alleles and a single mutant allele , respectively ( Figure 4A ) . The deletion was not observed in other breeds of normal chickens . Other variants discovered by sequencing genomic DNA from the frizzle chicken were also found in non-frizzle chickens except for one nonsynonymous SNP ( Table S2 ) . The effects of variants in other genes were not subjected to functional studies . We isolated RNA from the feather follicles 2-weeks after plucking of normal and F/F chickens and surveyed the expression of KRT75 in the feather follicles . We confirmed that KRT75 is expressed in feather follicles of both normal and F/F chickens ( Figure S5 ) . Sequence analysis of the coding sequence of KRT75 cDNA showed that the loss of the authentic splice site at the exon5/intron 5 junction activates a ‘cryptic’ splice site in exon 5 ( Figure 4B ) , resulting in a 69-bp in-frame deletion within the coding region ( CDS positions 934–1 , 002 ) . The cryptic splicing site in exon 5 contains 6-bp ( 5′-GTGAAG-3′ ) that resembled those at the authentic splice site . The mutated K75 thus contains a deletion of 23-amino acids within a conserved region ( Figure 4B and Figure S6 ) . The deletion covers the entire part of link L2 and some parts of the coiled-coil segments of 2A and 2B in K75 ( Figure S7 ) [54] . The length of link L2 is highly conserved in all keratin proteins and required for changing the azimuth of the coiled-coil over a short distance axially to reorient the apolar residues in coiled-coil segment 2A appropriately in terms of energetic stability [55] . Therefore , the loss of link L2 might significantly disrupt the structure over the coiled-coil segments 2A and 2B , potentially preventing the proper dimerization of keratin , consistent with a dominant-negative mode of action . To locate the KRT75 transcripts in embryonic and adult feathers , we generated a KRT75 full-length antisense RNA probe . Section in situ hybridization showed that KRT75 is expressed in barb ridges but restricted to the region destined to become the ramus , at embryonic day 13 ( E13 ) ( Figure 5A ) . In the normal regenerating adult feather , we found that KRT75 was expressed in both the rachis and the ramus ( Figure 5B ) . To ensure the specificity of our KRT75 probe , we made probes from the 3′untranslated region ( UTR ) which show the same expression pattern as our probe to the coding region ( data not shown ) . The regenerating frizzle feathers show the same pattern of KRT75 expression as those in normal controls ( Figure 5C , compared to 5B ) , suggesting that KRT75 mRNA is expressed in the normal regions in the frizzle mutant chicken feathers and the phenotype must result from a dysfunction of the protein . We then compared the expression of KRT75 and feather keratin at mature regions of the wildtype body feather ( Figure 5D ) . We found that KRT75 and feather keratin were co-expressed in the rachis and ramus . KRT75 was expressed in the ventral part , whereas feather keratin was expressed in the dorsal part of the rachis and ramus . To study the expression pattern differences between KRT75 and feather keratin , we examined the expression of these two keratins in the rachis and ramus from the immature to the mature stage of normal wing feathers ( Figure 5E–5H ) . KRT75 is expressed in the ventral part but feather keratin is expressed in the dorsal part of the immature rachis . Ventral regions of the feather formed a medulla that expressed KRT75 , whereas feather keratin was expressed in the dorsal region ( Figure 5E , 5F ) . In the ramus , KRT75 and feather keratin were expressed in a similar pattern as in the rachis ( Figure 5G , 5H ) . In summary , in both the rachis and ramus , KRT75 and feather keratin were expressed in a complementary pattern ( Figure 5F , 5H ) . K75 was present in the medulla but feather keratin was not . These data confirm that although body and wing feathers have differences in symmetry and size , they show similar expression patterns of KRT75 . To further examine the alteration of K75 protein expression in the frizzle rachis , we performed double immunostaining using antibodies to both K75 and feather keratin . Figure 5I shows staining of a section at level II , adjacent to the section shown in Figure 2B . In the normal rachis , feather keratin protein ( red ) is expressed in the dorsal part and in regions surrounding the medulla , while K75 protein ( green ) is expressed in the ventral rachis as well as at lower density in the medulla . The protein expression pattern is the same as the mRNA expression pattern ( Figure 5D ) . In the frizzle rachis , K75 protein is only expressed in the narrower ventral region but the feather keratin domain expanded to cover the medulla which is reduced in size . The perturbed keratin organization in the frizzled rachis suggests that the frizzled phenotype may be caused by the KRT75 mutation . To test the function of KRT75 in feather development , we constructed RCAS-KRT75-WT and RCAS-KRT75-MT viruses to misexpress the normal and mutant forms in embryonic and adult chickens . RCAS-KRT75-WT virus did not produce feather malformations in chicken embryos; however , some keratin-like depositions were found ( N = 10/10 ) ( compare middle and left panels in Figure 6A ) . Whole mount in situ hybridization of KRT75 confirmed the ectopic expression of KRT75 in the feather filaments ( insert in Figure 6A , middle panel ) . In comparison , misexpression of KRT75-MT generated feathers with curved tips , mimicking the frizzle phenotype ( N = 8/20 ) ( Figure 6A , right panel ) . We further characterized feather phenotypes produced as a result of KRT75 misexpression by H&E staining , PCNA staining , AMV-3C2 staining ( for RCAS virus detection ) , KRT75 section in situ hybridization and TUNEL assay ( for detection of apoptosis ) in serial paraffin sections at different levels of the filament ( from proximal - level I , to distal - level IV , as shown in Figure S8 ) . Figure 6B–6F shows sections at level III , which are close to the feather filament tip . We did not detect significant alterations based on H&E ( Figure 6B ) and PCNA staining ( Figure 6C ) among the treated samples and controls . The treated samples displayed strong AMV-3C2 staining ( Figure 6D , arrows ) and ectopic KRT75 expression ( Figure 6E , arrows ) . The detailed H&E and TUNEL staining at different levels are shown in Figure S8B–S8D′ . In normal development , programmed cell death appeared in the peripheral epidermis at level I , II and III but eventually apoptosis occurred in all distal tip cells ( level IV ) ( Figure S8B′ ) . Cell death was detected infrequently in the proximal to middle region of the epidermis shown in Figure S8B′ ( red arrow ) . However , both treated samples induced ectopic cell apoptosis but KRT75-MT misexpression induced significantly increased TUNEL positive cells ( Figure 6F , Figure S8C′ and S8D′ ) . We conclude that the misexpressed mutant form of KRT75 induces significant ectopic cell apoptosis , which may be responsible for the randomly curved feather morphology in the RCAS-KRT75-MT infected feathers . To verify that the adult frizzle phenotype is due to the identified KRT75 mutant , we misexpressed KRT75-WT or KRT75-MT by injecting the RCAS virus into adult feather follicles after plucking . Misexpressing KRT75-WT produced twisted feathers ( N = 5/12 ) ( Figure 6G ) . The control feathers involving only plucking or injecting RCAS-GFP did not show the twisted phenotype ( N = 0/20 ) . Cross sections of the twisted feathers showed the asymmetrical distribution of ectopically expressed KRT75 in the ramogenic zone of the feather follicle ( Figure 6H ) . Misexpression of the mutant form of KRT75 produced the curved feathers but the curvature only existed at the tip of the feather ( N = 6/10 ) ( Figure 6I ) . Control feathers on the right wing did not show any unusual curvature ( N = 0/10 ) ( Figure 6I ) . Since misexpressing the mutant form of KRT75 in a normal feather follicle that contains numerous normal KRT75 transcripts only affects the distal feather tip , we presume that its effect may be masked by high levels of endogenous wild type transcripts and limited to the softest part of the rachis at the tip . Images of flight feathers sampled from two wings of the same chicken in our experiments are shown in Figure 7A . Even though visual inspection of images of the control and KRT75-MT transfected feathers only reveal subtle differences , computer-aided analyses showed that ectopic expression of mutant K75 substantially changed the way the feathers bend along their rachis . Under normal circumstances , the natural bending of feathers from either side of a wildtype chicken would be expected to display reflective or mirror image symmetry to that of the opposite wing ( Figure S9 ) . Instead of the wild-type gentle inward bend , the end of the infected feather was twisted abruptly away from the body ( Figure 7B ) , as a consequence of the viral KRT75-MT misexpression during the feather growth of the left wing . While θ ( s ) of three control feathers generally converge on the length-normalized coordinate , these curves of θ ( s ) determined from the KRT75-MT transfected feathers are rather diverse . They exhibit anomalous bending and kinky structures that are qualitatively different from those of the controls . Our analyses also revealed that the over-expression of KRT75-WT resulted in twisted feathers and increased the curvature of the feather in a smooth manner ( Figure S10 ) , suggesting that excessive K75 may affect the physical properties of the feather . Ectopic mouse K75 ( K6hf ) was reported to co-localize with K8 , K17 and K18 in cultured PtK2 rat kangaroo kidney epithelial cells [56] . To explore the role of KRT75-MT in disrupting the intermediate filament structure in a dominant fashion , we transfected PtK2 cells with RCAS expressing either KRT75-WT or KRT75-MT . K18 ( red ) is present in the cytosplasm in a network configuration and K75 ( green ) is weakly positive in control PtK2 cells ( Figure S8E ) . After transfection with wild type KRT 75 , KRT 75 is expressed strongly . The keratin network is still maintained ( Figure . S8F ) . When the mutant K75 form is expressed , both K18 and K75 accumulate around the nucleus ( Figure S8G , white arrows ) . This is similar to what was found for mouse K75 [48] . Our data suggests that avian KRT75-MT can act in a dominant negative fashion to disrupt the keratin filament network .
KRT75 is a member of the type II epithelial α-keratin gene family [58]–[60] . The feather mainly consists of two types of keratin proteins: α- and β-keratins . An obligate heteropolymer is formed by two types of α-keratin , an acidic type and a basic/neutral type , and culminates into the 8–10 nm-thick intermediate filament [61] , [62] . The polymerization partner of K75 is unclear but it may be K17 in mammals [48] . In contrast to α-keratin , a fibrous protein rich in alpha helices , β-keratin is rich in stacked β-pleated sheets . β-keratins are only found in reptiles and birds , whereas α-keratins exist in all vertebrates [50] . K75 is not a hard feather keratin per se . Although the feather mainly consists of feather-specific β-keratins , cellular and biochemical studies have shown that α-keratin plays an important role in the early formation of rachides , barbs , and barbules [51] . The molecular mechanisms for accumulating α-keratin in down feathers and regenerating feathers are still largely unknown . It has been proposed that during the development of regenerating feathers , the α-keratin in the initial tonofilaments of sheath cells is replaced by feather-specific β-keratin [51] , [63] , [64] . Ultrastructural studies indicate that bundles of keratin filaments of 8–12 nm in diameter ( α-pattern ) are initially formed in differentiating barb/barbule cells and later replaced by 3–4 nm-thick filaments ( β-pattern ) [51] , [65] . Our studies however , revealed that the α-keratin and β-keratin actually accumulated in different parts of the rachis and ramus . We found that KRT75 is expressed in the ventral part that is destined to become the medulla , whereas β-keratin is expressed in the dorsal part of the rachis and ramus that is destined to become the cortex . Biochemical studies have indicated that feather α-keratins are mainly acidic , while basic α-keratins are thought to be rare in feathers [51] . Cytokeratins have been proposed to have a role in the formation of an initial and temporary scaffold for the deposition of immense quantities of compact feather keratins [51] . The identification of KRT75 , which encodes a type II cytokeratin ( basic ) , as a major determinant of normal feather structure suggests that basic α-keratins are also critical for feather formation . In the cytoplasm of rachis sheath cells , a higher quantity of α-keratin bundles is initially deposited . This observation may explain why the rachis is more severely affected than the barb and barbule in KRT75 mutant chickens . During feather filament development , keratinocytes eventually die , either leaving space or leaving a keratinized structure . Before those events occur , localized proliferation and apoptosis of keratinocytes either add or remove cells in different places , thus shaping the feather , including the rachis [66] . Our data show that apoptosis is expanded in frizzle compared to control adult feathers in the immature feather region . In contrast , the proliferation zone is decreased in frizzle chicken feathers compared to controls . Our functional studies on embryonic chicken feathers show that apoptosis is increased within the inner epithelium of embryonic feathers expressing ectopic KRT75 compared to controls and is increased even further in embryonic feathers expressing the mutant form of KRT75 . Currently , we do not know whether the mechanism is through a classic mechanical role or through an alternate pathway as was seen for KRT17 in mammalian hair follicles [67] . Our results show that KRT75 unquestionably plays a significant role in the normal development of feathers . However , the cellular mechanisms underlying the frizzle phenotype have not been specifically probed because the role of α-keratin in feather development is largely unexplored . The mutation could potentially affect feather formation in many ways such as altering the mechanical properties of the feather , weakening the initiation of keratin formation , causing abnormal scaffolding for feather keratin deposition , impairing α- and β-pattern replacement , or perturbing β-keratin polymerization . Our findings support the importance of α-keratin in feather formation . Our results also demonstrate the power of using mutants of domestic chickens as a genetic model to unravel biological functions that are difficult to reveal by traditional biochemical and cellular studies . The action of the F gene is localized in the feather follicle and is not a consequence of a metabolic disorder [68] . However , the F gene may also have other pleiotropic effects that cause physiological abnormalities . Frizzle plumage may cause the acceleration of basal metabolism due to the loss of body heat , leading to alterations in organ size ( e . g . , enlargement of the heart , spleen , gizzard , and alimentary canal as well as lack of hypodermal fat deposits ) and numerous physiological anomalies ( e . g . , higher food intake , oxygen consumption , heart rate , volume of circulating blood as well as delayed sexual maturity or decreased fertility ) [18] , [68]–[71] . An autosomal recessive modifier gene mf , which restricts the effect of F , has also been found in some chicken breeds [72] , [73] . Pathogen free stocks are required for RCAS mediated gene misexpression . The available SPAFAS chickens are of White Leghorn . Landauer [72] indicated White Leghorns chickens are likely enriched for the recessive modifier of the frizzle phenotype . This may help explain the less impressive phenotypes we observed with misexpression of a virally-derived frizzle protein , although whether this is really the case remains an open question . In mammals , the α-keratin K75 ( Keratin 75 or cytokeratin 75 , formerly known as K6hf or hfK6 ) is a hair follicle-specific epithelial keratin [58] . K75 plays an essential role in hair and nail formation . The KRT75 gene is specifically expressed in the companion layer [the cellular layer that lies between the outer ( ORS ) and inner ( IRS ) root sheaths] , the upper germinative matrix region of the hair follicle ( where mitosis takes place and hair keratins are produced ) , and the medulla of the hair shaft [74]–[78] . Mutations in KRT75 have been associated with the hair disorderpseudofolliculitis barbae ( PFB ) . Pseudofolliculitis barbae is a common human hair disorder characterized by a pustular foreign body inflammatory reaction that is induced by ingrown hairs of the facial and submental ( barbea ) regions after regular shaving [79] , [80] . This abnormal hair orientation phenotype is somewhat similar to the frizzle feather we report here . Interestingly , an unusual Ala12Thr polymorphism in the 1A alpha-helical segment of K75 has been associated with PFB by examination of a three-generation Caucasian family as well as 100 individuals affected with PFB and 100 unaffected controls . Modeling and transfection studies led the investigators to conclude that the Ala12Thr substitution is disruptive at later stages of filament assembly and could represent one of the genetic factors leading to this complex phenotype . This abnormal hair orientation phenotype is somewhat similar to the frizzle feather we report here . Besides , another hair-follicle-specific epithelial keratin is also known to be associated with the autosomal dominant wooly hair syndrome [81] , [82] . A total of 54 functional keratin genes in the human genome can be divided into 28 type I genes and 26 type II genes [83] , [84] . KRT75 is located within the type II keratin gene cluster on chromosome 12 of humans and chromosome 15 of mice [85] . Twenty out of the 26 type II keratin genes are epithelial keratins and six encode hair keratins [80] . KRT75 is tightly linked with epithelial keratin genes KRT6A/B/C and hair keratin genes KRT81–86 in humans [83] . Mutations in human hair-follicle specific epithelial type II keratins are known to cause structural defects of differing severity in hair , nail , and skin [86] , but the regulation of these keratins during proliferation and differentiation is yet to be elucidated . The chicken and zebra finch genomes contain only 28 and 27 α-keratin genes respectively compared to 41 genes in the anole genome [87] . β-keratin duplications occurred more frequently in birds than in reptiles [87] and they may have replaced some important roles in the formation of hard appendages in birds , thus the remaining α-keratins in the birds' genomes should play irreplaceable roles in the formation of epithelial and epidermal appendages of birds . The frizzle feather might represent a phenotype that could also be caused by mutations in genes involved in rachis structure but other than KRT75 . Thus , this group of appendage structural mutants can be considered as a sub-category of ectodermal dysplasia . Interestingly , we observed that the rachis and barbs of feathers from homozygous frizzle chickens were easily broken during handling , and they were also easier to pluck . However , no abnormalities in nails were noted ( unpublished observation ) . Further studies will provide clues into the architectural principles controlling how various skin appendages are built . It also calls for more molecular investigation into the role of this gene cluster in the evolution and development of the feather . Another interesting phenomenon we report is the use of the cryptic splice site in this mutant . A cryptic splice site is a suppressed splice site that is recognized but usually not utilized by the splicing machinery until a mutation activates it , either by strengthening the cryptic splice site or disrupting an authentic splice site [88] , [89] . Disruption of authentic splice sites crucial for identification of the 5′ or 3′ splice sites frequently result in complete exon skipping or in activating of the use of cryptic splice sites [90] . For genes with many introns , it is thought that up to 50% of mutations that cause disease actually affect splicing , either through the activation of cryptic splice sites , exon skipping , or disruption of alternative splicing [91]–[94] . In conclusion , we show that a single KRT75 allele is the major determinant of frizzle feathers in chickens . It is most interesting to compare the phenotypes caused by a mutation in KRT75: pseudofollolliculitis barbae in human and frizzle feather in chicken . The phenotypes appear to be very different , but indeed fundamentally similar in that both exhibit appendage architecture defects . In chicken , the defect appears to be exaggerated due to the elaborate morphogenesis of feathers . Thus the ability to identify a gene that contributes to feather morphology highlights the potential contribution of chicken genetics to the understanding of feather variations . It also illustrates how the progress in chicken genomics provides a new approach to dissect basic biological questions [28] , in this case , the molecular determinants of feather forms . Finally , this body of work highlights the importance of the cluster of α-keratin genes and invites further exploration into their role in normal and aberrant vertebrate developmental processes and provides an impetus for analyzing the relationship between α- and β-keratin during feather evolution .
Animal care and experiments were conducted according to the guidelines established by the USC Institutional Animal Care and Use Committee . For the linkage and association mapping , DNA from blood or feathers was opportunistically obtained as a byproduct of National Poultry Improvement Plan testing . The first pedigree was derived from a bantam White-tailed Japanese/Silver Penciled Plymouth Rock frizzle heterozygote rooster and the second from a bantam Red Cochin frizzle heterozygote rooster . These males were crossed with 9 normal feathered bantam hens of the following breeds: Araucana , Barred Plymouth Rock , and White Leghorn . Eight to eighteen offspring were evaluated from each female . The first cross produced 33 frizzles and 29 wild-type and the second 19 frizzles and 27 wild-type for a total of 108 chicks . Frizzle embryos used in embryonic studies were generated by mating a homozygous frizzle rooster with homozygous frizzle hems . Embryos were collected at different embryonic stages . Feathers for image analysis were from sex-matched white Plymouth Rock siblings that were both heterozygous and homozygous for the F gene . Chicks were hatched and raised in the USC animal facility . For the functional study , pathogen free fertilized eggs were purchased from SPAFAS , Preston , CT . Some of these eggs hatched and the chickens were used for functional studies on adult feather follicles . Additional frizzle chickens were opportunistically sampled from farms in Wanhua and Tamsui , Taiwan . Feathers were approximated as two-dimensional objects , which defined a primary plane for our imaging . Typical images of the feathers were taken from the dorsal side . We identified the fringes of the rachis by edge-finding algorithms we previously developed and defined the backbone of the rachis as a curve equidistant to the two fringes . The curve was smoothed by a length scale of 3 mm to suppress noises due to image errors . We parameterized this backbone by s , the accumulated distance from the proximal end that is defined as the origin ( s = 0 ) . The total length of the rachis means the accumulated distance from the origin to the distal end . We described the bending of the rachis by a function θ ( s ) , in which θ is an angle ( with an arbitrary reference ) representing the tangent of the backbone at the location s . The generated or gene misexpressed feathers were fixed in 4% paraformaldehyde at 4°C overnight followed by procedures described by Jiang et al . ( 1998 ) for immunohistochemistry and 7 µm paraffin sections were prepared [95] . PCNA and AMV-3C2 antibodies are from Chemicon ( CBL407 ) and Hybridoma Bank respectively . Double fluorescent immunostaining was done using K75 antibody ( ab76486; Abcam , MA ) and feather keratin antibody from Dr . Roger Sawyer . Section were imaged with a Zeiss 510 confocal microscope ( University of Southern California Liver Center ) . DAPI was used to visualize the nuclei . TUNEL assay was preformed according to the protocol provided by Millipore ( catalog number S7101 ) . We performed PCR for the full length chicken KRT75 by using sense primer ( 5′-ATGTCTCGCCAGTCCACCG-3′ ) and antisense primer ( 5′-TTAGCTCCTGTAACTTCTCC-3′ . The PCR product was inserted into the p-drive plasmid ( Qiagen ) . Antisense probe was made to detect the KRT75 mRNA by section or whole mount in situ hybridization . Non-radioactive in situ hybridization was performed according to procedures described by Chuong et al ( 1996 ) [96] . SHH antisense probe was generated as previously described [49] . Feather keratin antisense probe was used to detect feather keratin B [97] . In order to locate the gene underlying the frizzle trait , a genome scan was conducted on progeny of crosses between the same heterozygous frizzle rooster , PF1 , and five different wild-type feathered hens . A total of 2678 SNPs were genotyped using the Illumina Goldengate assay . The average genotype call rate obtained for the 45 birds in the study was 99 . 37% ( range 98 . 32–99 . 74% ) providing approximately 2661 genotypes per bird . The genotype data was screened for Mendelian incompatibilities using PEDCHECK , while MERLIN was used to assess the data for occurrence of double recombination events over short genetic distances , which are most likely due to genotyping error . MLINK of the FASTLINK package was used to perform two-point linkage analysis . An autosomal dominant mode of inheritance with complete penetrance and a mutant allele frequency of 0 . 001 was used in the analysis . Chicken genomic DNA was isolated from blood using the Blood & Cell Culture DNA Mini Kit ( Qiagen , Hilden , Germany ) . For mutation analysis of the keratin candidate genes , we got 49 PCR amplicons of the selected candidates amplified from chicken genomic DNA ( Table S1 ) . Primers to cover some intronic and exonic regions of 14 keratin genes were designed using the CLC Bio 6 . 0 ( Aarhus , Denmark ) . All amplicons were sequenced directly after treatment with exonuclease I and shrimp alkaline phosphatase by standard methods . Each amplicon was sequenced using BigDye terminator sequencing kits and standard protocols ( Applied Biosystems , Santa Clara , CA ) . A significant mutation was detected in only one amplicon amplified by a pair of primers ( 5′-CCATGGACAACAACCGCAAC-3′ and 5′-TTTCCTTCCTTCCTTCCAATCCT-3′ ) . Feather follicle samples were collected from a homozygous frizzle chicken 2 weeks after plucking and then were immersed in RNALater ( Ambion , Austin , TX ) and stored at −20°C . After thawing , total RNA was isolated by homogenization and extraction using the RNeasy Tissue Midi kit ( Qiagen , Hilden , Germany ) . Each RT-PCR reaction was carried out with 1 ug of total RNA . Primers were designed for the KRT75 gene ( 5′-TTTCTTCTTTCCCTCCCACT-3′ and 5′- GTTCTGCTTCCCCTGATTAT-3′ ) . KRT75 cDNA PCR products were cloned into the pCR8/GW/TOPO Gateway entry vector ( Invitrogen , Carlsbad , CA ) and sequenced . An LR recombination reaction was performed to transfer the cDNAs to a Gateway compatible RCASBP-Y DV vector [98] . KRT75 wildtype ( KRT75-WT ) and KRT75 mutant forms ( KRT75-MT ) were cloned to RCAS by the Gateway system ( Invitrogen , Carlsbad , CA ) . Virus was made according to Jiang et al . , 1998 [95] and concentrated by ultra-centrifugation . Around 30 contour feathers from the middle back of the body were plucked and then 10 days or 30 days were allowed to pass . At the collection points , regenerated feathers were directly plucked or dissected and the whole single follicles were prepared for sectioning . For embryonic studies , RCAS-KRT75-WT or RCAS-KRT75-MT virus was injected into the amniotic cavity of E3 chicken embryos . Samples were collected at E13 . RCAS-GFP was injected into different embryos as a control . For adult feathers , about 100 µl of virus was injected into the empty follicles after plucking the primary flight feathers in the left wing . The feathers on the right wing were collected at same time as controls . Feather follicles from a different chicken injected with RCAS-GFP were used as an alternative control . Feather morphogenesis was observed after 1–2 months of regeneration . PtK2 cells ( American Tissue Culture Collection , MD ) were transfected with plasmid ( RCAS ) encoding KRT75-WT or KRT75-MT by lipofection ( Lipofectamine , Invitrogen , CA ) . After 48 hours samples were fixed in 4% paraformaldehyde and stained with antibodies to K18 ( AV40206; Sigma , MO ) and K75 . DAPI was used to visualize the nuclei . Images were obtained using a Zeiss 510 confocal microscope . The sequence of F allele of KRT75 has been submitted to GenBank with the accession number JQ013796 .
|
With the availability of a sequenced chicken genome , the reservoir of variant plumage genes found in domestic chickens can provide insight into the molecular mechanisms underlying the diversity of feather forms . In this paper , we identify the molecular basis of the distinctive frizzle ( F ) feather phenotype that is caused by a single autosomal incomplete dominant gene in which heterozygous individuals show less severe phenotypes than homozygous individuals . Feathers in frizzle chickens curve backward . We used computer-assisted analysis to establish that the rachis of the frizzle feather was irregularly kinked and more severely bent than normal . Moreover , microscopic evaluation of regenerating feathers found reduced proliferating cells that give rise to the frizzle rachis . Analysis of a pedigree of frizzle chickens showed that the phenotype is linked to two single-nucleotide polymorphisms in a cluster of keratin genes within the linkage group E22C19W28_E50C23 . Sequencing of the gene cluster identified a 69-base pair in-frame deletion of the protein coding sequence of the α-keratin-75 gene . Forced expression of the mutated gene in normal chickens produced a twisted rachis . Although chicken feathers are primarily composed of beta-keratins , our findings indicate that alpha-keratins have an important role in establishing the structure of feathers .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"physiology",
"genetics",
"biology",
"anatomy",
"and",
"physiology",
"genetics",
"and",
"genomics"
] |
2012
|
The Chicken Frizzle Feather Is Due to an α-Keratin (KRT75) Mutation That Causes a Defective Rachis
|
RNA-protein interaction plays important roles in post-transcriptional regulation . Recent advancements in cross-linking and immunoprecipitation followed by sequencing ( CLIP-seq ) technologies make it possible to detect the binding peaks of a given RNA binding protein ( RBP ) at transcriptome scale . However , it is still challenging to predict the functional consequences of RBP binding peaks . In this study , we propose the Protein-RNA Association Strength ( PRAS ) , which integrates the intensities and positions of the binding peaks of RBPs for functional mRNA targets prediction . We illustrate the superiority of PRAS over existing approaches on predicting the functional targets of two related but divergent CELF ( CUGBP , ELAV-like factor ) RBPs in mouse brain and muscle . We also demonstrate the potential of PRAS for wide adoption by applying it to the enhanced CLIP-seq ( eCLIP ) datasets of 37 RNA decay related RBPs in two human cell lines . PRAS can be utilized to investigate any RBPs with available CLIP-seq peaks . PRAS is freely available at http://ouyanglab . jax . org/pras/ .
RNA-binding proteins ( RBPs ) are essential in many post-transcriptional regulatory processes , such as alternative splicing , stability , localization and editing [1] . For example , RBP Quaking plays important roles in pre-mRNA splicing and mRNA export [2]; RBP HuR is an mRNA stability and splicing regulator [3]; RBP Ataxin-2 promotes mRNA stability and protein expression [4] . RBPs achieve their functions via binding to RNAs; therefore , it is of vital importance to study RNA-protein interaction . Cross-linking and immunoprecipitation followed by sequencing ( CLIP-seq ) approaches have been widely used to detect the binding peaks of RBPs at the transcriptome scale [5–9] . Thus , the examination of CLIP-seq peaks informs us of the functional targets of RBPs . Existing computational approaches for analyzing CLIP-seq data focus on detecting RBP binding peaks [10–19] or differential RBP binding peaks between two different conditions [11 , 14 , 15 , 20] . Computational methods for predicting the functional consequence of RBP binding peaks are less well-established [21–23] . Some studies suggest that the binding preferences of RBPs are associated with their specific functions . For example , HuR binding preferentially occurs close to the 3’ splicing site , which is consistent with its known function on alternative splicing [3]; Ataxin-2 , an mRNA stability regulator , has a tendency to bind close to the polyadenylation site [4] . A recent study revealed that RBP TDP-43 regulates poly ( A ) site usage in a position-dependent way [22] . In this paper , we develop a new approach named Protein-RNA Association Strength ( PRAS ) , which incorporates the intensity and positional information of CLIP-seq peaks to quantitate the association between an RBP and its targets . We apply PRAS to study two CUGBP ELAV-like family proteins , CELF4 and CELF1 with both CLIP and perturbation RNA-seq data available . CELF4 ( also known as Brunol4 ) is expressed as an mRNA regulator in the central nervous system across species [24 , 25] . The deficiency of CELF4 is associated with a complex neurobehavioral disorder including seizures and autism-like features in human [26 , 27] and in mice [28] . iCLIP studies revealed that CELF4 preferentially binds , almost exclusively in 3’ untranslated regions ( UTRs ) , to mRNAs encoding many important neurological functions , [29] . CELF1 is implicated in myotonic dystrophy [30] . CELF1 is highly expressed in early embryonic stages and are then down-regulated dramatically in skeletal muscle and the heart during development [31 , 32] . CELF1 has been reported to promote transcript deadenylation and the abnormal up-regulation of its protein level could contribute to the myotonic dystrophy pathology [33 , 34] . A more refined understanding of the functional targets of CELF RBPs is essential for understanding the impact of CELF in development and diseases , and may provide clues as to the mechanisms by which CELF impacts mRNA function . In addition , to demonstrate the robustness of PRAS , we examined its performance of detecting the functional targets in a large-scale collection of eCLIP data of RBPs in the integrated encyclopedia of DNA elements in the human genome ( ENCODE ) . By applying PRAS to the eCLIP peaks of the RNA decay regulators , we demonstrate that PRAS outperforms other existing methods and also provide deeper understanding in the post-transcriptional regulation of these RBPs .
The basis of PRAS is to score a potential functional target of an RBP based on both the intensities and positions of its binding sites . Our pipeline of calculating PRAS is shown in Fig 1 . First , given a CLIP-seq dataset , the significant cross-linking sites that are within a small interval of each other ( default: 20 nt ) are merged as RBP binding peaks . If the called binding peaks are provided , we will use them directly . Second , if a reference position is provided by the user based on known knowledge of the function of the RBP , PRAS will use it directly; if no reference position is given , PRAS will set it based on the RBP’s binding preference , e . g . , the distal end of the 3’ UTR of the transcript ( aka polyadenylation site ) . Finally , each transcript is scored as the sum of the intensities of the binding peaks weighted by the distances between the mid points of the binding peaks and the preselected reference position . All mRNAs are then ranked by the PRAS scores and can be tested for associations with functions . As described in Fig 1 , the PRAS score is based on the weighted sum of the intensities of the binding given detected CLIP-seq peaks . In the study that analyzed the interaction between DNA and proteins with ChIP-seq datasets , the exponential decay function was used to characterize the decreasing effects of a transcription factor binding peak on its targets with increasing distances [35] . Therefore , we here construct the score to describe the regulatory effect of an RBP on its targets in a similar way . Specifically , we define the PRAS score for an mRNA as: S=∑irie−di/d0 , ( 1 ) where ri is the intensity ( CLIP-seq read counts ) of the ith peak cluster of the RBP , di is the distance ( number of nucleotides ) between the reference position and the ith peak cluster , and d0 is a constant . For both CELF4 and CELF1 in mouse , we set the reference position as the distal 3’ UTR and the constant d0 = 1000 nt . Note that d0 = 1000 nt is the default setting , but not a hard-set option in PRAS . For the RNA decay regulators in human , we set the constant d0 = 500 nt . The details of d0 estimation for RBPs in mouse and human are described in the Results and Discussions sections . PRAS is implemented in Python ( version 2 . 7 . 14 or above ) and R ( version 3 . 3 . 2 or above ) scripts and has minimum requirements for the inputs . To reformat the annotation file , PRAS takes use of gtfToGenePred , a toolkit from the UCSC Genome Browser [36] . PRAS also uses BEDTools [37] to efficiently obtain the overlapping between the binding sites and the annotation regions . The annotation file should be the Gene Transfer Format ( GTF ) format and the peak file ( no special requirement for the peak caller ) should be the Browser Extensible Data ( BED ) format as the required input files , which are both the standard file formats . Details of usage can be found on the instruction page of our website: http://ouyanglab . jax . org/pras/ .
CELF4 is expressed in excitatory neurons of the adult mouse brain , from which iCLIP data are available [25 , 28 , 29] . We collected the significant cross-linking sites detected by iCount ( http://icount . fri . uni-lj . si ) with false discovery rate ( FDR ) less than or equal to 0 . 05 . We conducted a metagene analysis involving all 9 , 193 mRNAs that are bound by CELF4 and noted an enrichment of iCLIP reads at the distal ( 3’ end ) versus proximal ( 5’ end ) 3’ UTR ( S1 Fig ) . This preference suggests a potentially functional role of CELF4 binding close to the polyadenylation site . We calculated the PRAS scores for CELF4 binding mRNAs with the polyadenylation site as the reference position , which gives the binding sites closer to the polyadenylation site higher weights . We estimated the decay parameter d0 in Eq ( 1 ) based on the strength of the peak intensity decay shown in S1 Fig . In detail , we defined the weighting formula as w=e−d/d0 according to Eq ( 1 ) . The highest average peak density , 0 . 843 , appears at 63 nt to the 3’ end of 3’ UTR and the average peak density at 1000 nt upstream to the 3’ end of the 3’ UTR is 0 . 285 ( S1 Fig ) . We calculated w as the ratio between the average peak intensity at the 1000 nt upstream to the 3’ UTR and that of the 3' end of the 3’ UTR , which is 0 . 285/0 . 843 = 0 . 339 . By plugging d = 937 nt ( which is 1000 nt– 63 nt ) and w = 0 . 339 into the weighting formula , we obtained the estimation of 866 nt for d0 , which is approximately the default of 1000 nt . For comparison , we applied the expressRNA procedure of Rot et al . [22] , which sums the number of reads in CLIP peaks within 200 nt upstream and downstream flanking the polyadenylation sites ( Fig 2 ) . We also applied the procedure in Wang et al . [34] , which calculated the score as the number of significant CLIP peaks per kilobase ( noted as PPK; Fig 2 ) . Each of the three measurements ranks CELF4 binding mRNAs from high to low scores . We then evaluated the performance of PRAS , expressRNA , and PPK on a list of known functional targets previously validated by qPCR in wild-type and Celf4 null mouse brain , totaling 23 mRNAs [29] ( see details in S1 Text ) . To investigate the ability of the three measurements to identify CELF4 functional targets , we performed receiver operating characteristic ( ROC ) analysis . We extracted the log fold change ( LFC ) of the qPCR values in Celf4 null mouse brain over wild-type . The mRNAs with positive and negative LFCs were labelled as CELF4-degraded and CELF4-stabilized genes , respectively . The area under the curve ( AUC ) of the ROC curve was used to measure the prediction performance of the methods . We found that PRAS perfectly distinguished the PCR-validated CELF4-degraded and CELF4-stabilized genes ( AUC = 1 ) , outperforming expressRNA ( AUC = 0 . 867 ) and PPK ( AUC = 0 . 7 ) ( Fig 3A ) . This result suggests that given CLIP peaks , PRAS has greater ability to capture the functional targets of CELF4 compared to expressRNA and PPK . In addition , we examined the quantitative relationship between the PRAS scores and the qPCR LFCs of these known targets . A negative Pearson’s correlation coefficient ( -0 . 60 ) was obtained , suggesting that the more negative qPCR LFC a target has , the larger the PRAS score is ( Fig 3B ) . The advantage of PRAS over expressRNA and PPK can be attributed to two factors . First , PRAS utilizes the binding bias of CELF4 towards the distal 3’ UTRs of its validated targets ( Fig 3C ) . expressRNA partially utilizes this bias by considering the 200 nt flanking region around the polyadenylation site , whereas PPK does not consider the binding bias . Second , unlike expressRNA which only considers a fixed flanking region , PRAS considers all binding peaks , which decreases loss of important RBP binding sites . The analysis of the validated targets of CELF4 suggests the importance of binding near the polyadenylation sites as a potential factor on how it regulates gene expression . By applying different decay parameter d0 to PRAS , we found that PRAS obtained equally good performance over a reasonable range of d0s ( S2A and S2B Fig ) . A d0 that falls out of certain range will decrease the performance of PRAS ( S2A and S2B Fig ) , because a too small d0 can filter out the majority of iCLIP signals and a too large d0 approximates the uniform weighting . The stable performance of PRAS with d0 chosen around 1000nt shows the robustness of PRAS ( S2 Fig ) . To assess the ability of PRAS to detect RBP functional targets in the entire transcriptome , we extracted the top 500 genes ranked by permutation test p-values in the differential expression test between the wild-type and Celf4 null mouse brain based on existing microarray datasets [29] . We calculated the LFC for gene expression in Celf4 null over wild-type mouse brain . The mRNAs have lower abundance ( LFC < 0 ) in Celf4 null genotype are more likely to be CELF4-stabilized targets , while the mRNAs with higher abundance ( LFC > 0 ) in Celf4 null brain were more likely to be CELF4-degraded targets . We sought to assess the ability of PRAS on capturing CELF4-stabilized vs . CELF4-degraded targets . Specifically , we first set a sequence of cutoffs as the quantiles ( from 0 . 05 to 0 . 95 with step size as 0 . 05 ) of the distribution of the absolute value of the expression LFCs . Second , for each cutoff , we extracted a subset of genes whose absolute expression LFC is larger or equal to the cutoff . Finally , for each subset of potential CELF4 targets , we calculated the Spearman’s correlation coefficient between the expression LFCs and the PRAS scores , in which the magnitude and sign of the correlation reflect the association between the two . For comparison , we also applied the same correlation analysis to expressRNA and PPK ranking scores . Line-charts of the Spearman’s correlation coefficient of the three methods are shown in Fig 4A . We observed that the more stringent the expression LFC cutoff for the gene subset was set , the stronger the negative correlation between the PRAS score and the expression LFC was obtained , which suggests that PRAS is more powerful in capturing more reliable CELF4-stabilized targets . In addition , the expressRNA score is less correlated with the expression LFC , and the direction of the correlation between the PPK score and the expression LFC flips at different cutoffs . The results suggest that PRAS has greater ability to select the regulated mRNA targets compared to expressRNA and PPK . We also extracted the top 500 genes ranked by their adjusted p-values in the differential expression test between the wild-type and Celf1 over-expression in mouse muscle based on published RNA-seq datasets [34] . In this dataset , mRNAs that have higher abundance upon Celf1 over-expression ( LFC > 0 ) are more likely to be CELF1-stabilized targets while those that have lower abundance upon Celf1 over-expression ( LFC < 0 ) were more likely to be CELF1-degraded targets . We evaluated the performance of the three aforementioned methods using the same analysis as with CELF4 . We used the 3’ end of the 3’ UTR as the reference site in PRAS to rank the mRNA targets based on the reported binding preference of CELF1 [34] . PRAS has a stronger negative correlation with the expression LFC compared to expressRNA and PPK for each subset of the potential CELF1 targets ( Fig 4B ) . These results suggest that PRAS is more powerful in capturing the reliable CELF1-degraded targets , consistent with the main regulatory function of CELF1 [34] . Next , we used different reference sites in PRAS for scoring functional targets of CELF4 and CELF1 in order to examine the effect of the reference site selection . We scored the targets of CELF4 using the 5’ end of the 3’UTR as the reference site in PRAS ( PRAS 5’ ) and did a similar correlation analysis as above . We observed that the PRAS 5’ score is also negatively correlated with the expression LFC and the magnitude of correlation improves with increasingly stringent cutoffs ( S3A Fig ) . However , the magnitude of the correlation is not as high as that of PRAS with the 3’ end of 3’ UTR as the reference site ( PRAS 3’ ) ( Fig 4A ) . We also similarly analyzed the targets of CELF1 using PRAS 5’ . Again , the PRAS 3’ has stronger negative correlation with the expression LFC than PRAS 5’ for the more reliable CELF1 targets ( S3B Fig ) . The results indicate that known biological knowledge can aid in reference site selection in PRAS for identifying the functional targets of the CELF proteins . The results also suggest that both the CELF4 and CELF1 proteins may regulate mRNAs via the distal 3’ UTRs while having opposite effects on their targets . Indeed , this is plausible because CELF proteins play various roles in both co-transcriptional and post-transcriptional RNA regulation , as well as translation inhibition in different cellular contexts [38–40] . To examine the difference of taking the raw or the normalized read density of the CLIP peaks as the input of PRAS , we then used the Celf4 null iCLIP-seq as the negative control for the wild-type CELF4 iCLIP to score the functional targets of CELF4 with the 3’ end of the 3’UTR as the reference site . Specifically , we replaced the iCLIP-seq read counts ri in Eq ( 1 ) by the enrichment ratio ri×log2 ( rici ) as suggested by Van Nostrand et al . [41] , where ci is the Celf4 null iCLIP-seq read counts of the ith peak cluster . We noted the PRAS score using the raw read intensity and the enrichment ratio of peaks as PRAS-raw and PRAS-norm , respectively . By applying the correlation analysis as above , we found that PRAS-norm has achieved stronger negative correlation with the expression LFC than PRAS-raw ( S4 Fig ) . This improvement of performance indicates the important role of the negative control in reducing the noise , which is consistent with the results in [42] . Even though PRAS-raw cannot achieve as good performance as PRAS-norm , the difference in the performance between them is small ( S4 Fig ) , which indicates that PRAS can handle the situation where the negative control of CLIP-seq is not available , such as the CELF1 data in our study . To further compare the functional relevance of the targets identified by PRAS , expressRNA and PPK , we performed gene ontology ( GO ) analysis on the top 500 mRNA targets of CELF4 ranked by each score ( Fig 5A–5C ) , which is similar to the analysis shown in Wagnon et al . [29] . There is much greater enrichment ( 5 to 40 orders based on p-values ) of the categories related to suspected CELF4 function in the targets identified by PRAS than those identified by expressRNA and PPK . For example , in the class of “Biological Process” , most of the top 10 significant categories for PRAS top-ranked targets are related to neuron or synaptic functions and ion transport , consistent with prior studies on CELF4 [29] . These results suggest that PRAS captures CELF4 functional targets more precisely than the other methods being compared . To demonstrate that PRAS has the potential for wide adoption , we further applied PRAS to the eCLIP data [42] in two human cell lines , K562 and HepG2 , from the ENCODE consortium [43] . Specifically , we selected the RBPs that are related to the RNA decay function [41] because this function can be clearly quantified at gene level in the differential expression ( DE ) analysis between the RBP knockdown and the wild-type RNA-seq samples . We collected the DE analysis results by DESeq [44] from ENCODE and obtained 37 distinct RBPs , which include 28 and 32 RBPs in HepG2 and K562 cell line , respectively . We then applied PRAS to the eCLIP data using the enrichment ratio over the control sample described above as the peak intensities . In the parameter settings in PRAS , we selected the reference site for each RBP from 4 candidates: transcription start site , translation initiation site , translation termination site , and transcription end site , based on eCLIP peak intensity distribution along the transcript . S5 Fig presents four example RBPs assigned with 4 different reference sites . To simplify the analysis , we applied d0 = 500 nt to all the selected RBPs according to the distribution ( S6 Fig ) of the estimated decay parameters as described previously . This general selection of d0 may not achieve the best performance of PRAS but is likely to be comparable with the best d0 selection as discussed in the CELF4 data . After obtaining the PRAS scores , we did the correlation analysis of the DE ( adjusted p-value < = 0 . 05 ) genes for each RBP . We found PRAS scores achieved significantly stronger correlation with the LFC in gene expression in comparison to expressRNA and PPK , with p-value equal to 3 . 8e-9 and 4 . 4e-4 , respectively ( Fig 6A ) . We then separated the RBPs by their reference site usage and found that the translation termination site and the transcription end site , both of which are related to the 3’ UTR , constitute the majority of the RNA decay regulators’ reference sites ( Fig 6B ) . It suggests the essential association between the 3’ UTR of transcripts and the regulation of their fates by RBPs . In addition , we found that the correlation can reflect important biological functions of RBPs . For example , the 5’ poly ( A ) site ( transcription end site ) is used as the reference site for DDX6 in the HepG2 cell line ( S5C Fig ) and the PRAS score is negatively correlated with the LFC of DDX6’s target gene expression ( Fig 6B ) , which indicates that DDX6 may stabilize its targets via binding near to the poly ( A ) site . Interestingly , DDX6 is known to be an important regulator in mRNA decapping and degradation [45 , 46] , which supports our claim that PRAS has the ability to identify the biologically functional targets of the RBP regulators . All these results demonstrate that PRAS has the potential for wide adoption in RBP functional targets identification . In this study , we developed PRAS , a position dependent scoring method for identifying and prioritizing RBP functional targets . Weighting the proximity of RBP binding sites to a given reference position exponentially and combining the strengths of the binding signals , we obtained the PRAS scores and the ranking of all the mRNAs that have reliable binding sites of the RBP . We applied this approach to the iCLIP dataset of a neuronal disease-related RBP , CELF4 and to the CLIP dataset of a DM disease-related RBP , CELF1 –both belonging to the CELF family of RBP . We report a much stronger association between CELF4 and its targets at the distal 3’ UTRs compared to internal 3’ UTR positions . We also demonstrate that PRAS performs much better in predicting the mRNA targets stabilized by CELF4 , compared to the other existing methods such as expressRNA and PPK . We further observe that PRAS performs much better at predicting the mRNA targets degraded by CELF1 . These results not only suggest the importance of incorporating the positional information of the binding sites into target identification , but also suggest the important roles of the distal 3’ UTRs in CELF protein regulated mRNAs . The binding preferences of RBPs have been noticed in previous studies [3 , 29] . However , the link between positional biases of RBP binding sites and their functional consequences has not been well established . PRAS reveals that the distal end of 3’ UTR binding is predictive of CELF4-stabilized targets . The distal end bias of CELF4-stabilized targets suggests possible molecular mechanism ( s ) by which CELF4 regulates its mRNAs . It has been reported that poly ( A ) tails enhance the stability of mRNAs [47] . The proximity between poly ( A ) tails and the distal 3’ UTRs suggests possible connections with poly ( A ) tail functions , such as mRNA stability , polyadenylation itself or promotion of translational reinitiation–possibilities to be explored in future experimental studies . CELF1 is known to recruit cytoplasmic deadenylases [48] and the extent of mRNA degradation is positively correlated to CELF1’s binding magnitude to the 3’ UTRs [34] . Based on the finding in the previous study [34] that CELF1 binding is enriched in the 3’ end of the 3’ UTR , we further found that this binding bias shows strong predictive ability to CELF1-degraded targets ( Fig 3B ) . We also demonstrated the potential of PRAS in the large-scale applications by showing the better performance of PRAS than other methods in identifying the targets of RNA decay related RBPs from ENCODE [43] . These results again strengthen the relationship between the regulatory functions of the RBPs and their binding positions .
PRAS is implemented in Python and R and is freely available at http://ouyanglab . jax . org/pras/ . PRAS can be applied widely to identify the functional targets of any RBPs with CLIP-seq peaks . For RBPs with a known post-transcriptional function , the functional targets may be identified with a corresponding reference position that is related to that function ( e . g . splicing sites for alternative splicing ) . PRAS can also be combined with other types of information , such as sequence motifs , conservation , and perturbation data to predict RBP functional targets using integrative approaches such as [49] . In addition , future versions of PRAS can be extended to study the co-regulations of multiple RBPs by being applied to a set of interested RBPs simultaneously and evaluating the importance of different reference sites on the targets .
|
It is important to identify the functional targets of RBPs , which are essential regulators in post-transcriptional processes . PRAS aims to predict RBP targets based on the intensities and positions of the binding peaks obtained from CLIP-seq studies . We demonstrate that PRAS score outperforms other existing methods not only in the prediction of the PCR-validated targets of the RBP CELF4 , but also in the correlation with the global expression change induced by CELF proteins . The better performance of PRAS on a group of RBPs associated with RNA decay in comparison to the existing methods indicates its potential for large-scale applications in detecting functional targets . Leveraging the position information of the binding peaks , PRAS is a bridge linking peak-calling methods and the interpretation of RBPs’ biological functions , which strengthens the analysis of CLIP-seq data .
|
[
"Abstract",
"Introduction",
"Design",
"and",
"implementation",
"Results",
"and",
"discussions",
"Availability",
"and",
"future",
"directions"
] |
[
"rna-binding",
"proteins",
"chemical",
"bonding",
"3'",
"utr",
"messenger",
"rna",
"dna",
"transcription",
"untranslated",
"regions",
"genome",
"analysis",
"physical",
"chemistry",
"proteins",
"gene",
"expression",
"chemistry",
"cross-linking",
"gene",
"ontologies",
"biochemistry",
"rna",
"polyadenylation",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"computational",
"biology"
] |
2019
|
PRAS: Predicting functional targets of RNA binding proteins based on CLIP-seq peaks
|
Tsetse flies transmit trypanosomes , the causative agent of human and animal African trypanosomiasis . The tsetse vector is extensively distributed across sub-Saharan Africa . Trypanosomiasis maintenance is determined by the interrelationship of three elements: vertebrate host , parasite and the vector responsible for transmission . Mapping the distribution and abundance of tsetse flies assists in predicting trypanosomiasis distributions and developing rational strategies for disease and vector control . Given scarce resources to carry out regular full scale field tsetse surveys to up-date existing tsetse maps , there is a need to devise inexpensive means for regularly obtaining dependable area-wide tsetse data to guide control activities . In this study we used spatial epidemiological modelling techniques ( logistic regression ) involving 5000 field-based tsetse-data ( G . f . fuscipes ) points over an area of 40 , 000 km2 , with satellite-derived environmental surrogates composed of precipitation , temperature , land cover , normalised difference vegetation index ( NDVI ) and elevation at the sub-national level . We used these extensive tsetse data to analyse the relationships between presence of tsetse ( G . f . fuscipes ) and environmental variables . The strength of the results was enhanced through the application of a spatial autologistic regression model ( SARM ) . Using the SARM we showed that the probability of tsetse presence increased with proportion of forest cover and riverine vegetation . The key outputs are a predictive tsetse distribution map for the Lake Victoria basin of Uganda and an improved understanding of the association between tsetse presence and environmental variables . The predicted spatial distribution of tsetse in the Lake Victoria basin of Uganda will provide significant new information to assist with the spatial targeting of tsetse and trypanosomiasis control .
Tsetse flies are responsible for the transmission of human African trypanosomiasis ( HAT ) , also known as sleeping sickness and its animal form ( nagana ) . Trypanosomiasis occurs in 38 sub-Saharan African countries with an average of 15 , 000 human cases reported annually ( period 2000–2012 [1] ) , and 70 million people at risk of contracting the infection [2] . Uganda reports approximately 500 cases of sleeping sickness annually [1] , and it is the only country reporting the presence of both forms of HAT: the gambiense form in the north-west and the rhodesiense form in the south-east and , more recently , in the centre of the country [3 , 4] . Animal trypanosomosis presents major constraints to livestock production among many livestock keeping communities in Africa . The disease is widely reported in Uganda [5] , and the removal of African animal trypanosomiasis ( AAT ) could generate direct economic benefits in the region of 400 million US$ in a 20-year period [6] . Glossina fuscipes fuscipes is known to be present in several parts of Uganda , with its geographical extent stretching from Lake Victoria’s shores through central Uganda up to the West Nile region . In addition , G . f . fuscipes is assumed to be present around Lakes Albert , Edward and George in western Uganda . The islands of Kalangala and Buvuma located within Lake Victoria have also been identified as having G . f . fuscipes [7] . The major drivers of tsetse fly habitation are generally known to be temperature , humidity , rainfall , vegetation and presence of host animals [8 , 9 , 10] . This implies that tsetse flies are found in ecologically suitable habitats , represented through a set of conditioning environmental variables . Such variables determine: feeding behaviour; infection rates; fly movements; fly density; species-diversity; and fly reproduction [10] . Therefore , spatial information on such environmental variables can be helpful in predicting the relative distribution of tsetse flies in an area . Tsetse distribution maps are crucial in the control and management of human and animal trypanosomiasis in affected areas [11 , 12] . Accurate maps should ideally be based on high precision fly data derived from field investigations . In the absence of such data , tsetse distribution maps may be constructed using partial district-level entomological reports , existing publications , sector reports and modelled environmental covariates . Given scarce resources to carry out regular field tsetse surveys , there is a need to devise inexpensive means for periodically obtaining reliable large area and high precision tsetse information across target areas . A potential solution is provided by spatial statistical modelling ( e . g . , spatial regression analysis ) using tsetse presence or abundance data acquired from field survey and fine spatial resolution satellite-generated environmental variables . Regression is a statistical tool used to quantify the association between an outcome measure and predictor variables [13] . Logistic regression , in particular , is commonly used to explain or predict a binary variable response using a set of predictor variables or covariates [14] . This approach has been used in the predictive mapping of various vectors and associated vector-borne diseases including malaria and Rift valley fever , with broad applications in environmental disease risk [15] . The use of GIS and temporal Fourier-processed surrogates for vegetation , temperature and rainfall derived from satellite sensor data in predicting tsetse distributions has been investigated with significant utility [16] . Further use of GIS and remote sensing in attempting to explain tsetse vector distributions is described in Rogers et al . [17 , 18] and Wint et al . [19 , 20 , 21] . Wint and Rogers [19 , 21] , at a spatial resolution of 5 km , predicted tsetse presence at the continental level using logistic regression , targeting 23 tsetse sub-species from the three major species groups ( Fusca , Palpalis and Morsitan ) . The process involved fitting statistical regression models between tsetse data and remotely sensed predictor variables . The tsetse data used were derived from the Ford and Katondo tsetse maps [22 , 23] , through systematic extraction of 12 , 000 points across the entire continent . Predictor variables included; NDVI , surface temperature , middle-infrared reflectance , vapour pressure deficit and surface rainfall [19 , 21] . Wint [20] , in an effort to provide more accurate tsetse maps , derived sub-continental tsetse fly distribution maps at a spatial resolution of 1 km for East Africa ( Uganda ) and selected parts of some countries in West Africa . This approach made use of; ( i ) modified Ford & Katondo presence/absence maps , ( ii ) 5 km-continental tsetse predictions in 2000 , ( iii ) 17 , 000 data points extracted for East Africa and satellite-derived data . According to these maps , Uganda is approximately 80% tsetse infested . Although an improvement from the Wint continental version [19 , 21] , these sub-continental tsetse distribution maps are associated with low precision . The lack of up-to-date field data on tsetse is a key concern , while the absence of land cover data as a predictor , which is known to be important in determining tsetse distributions , is another . In Uganda , there is a need to produce dependable and up-to-date tsetse distribution information , preferably at sub-national level , to support decision-making and improved planning of tsetse control interventions . Relatively few studies have used recently gathered data from traps . The purpose of this study was to quantify the relationships between tsetse presence/absence and external factors in the study area and also to predict the spatial distribution of G . f . fuscipes in the Lake Victoria basin of Uganda .
The study area is predominantly a lake basin stretching for approximately 50 to 100 km from the Lake Victoria shoreline in Uganda . This region is characterized by high annual rainfall ( 1000–1500 mm ) with two distinct rainfall peaks in April and November . Tsetse data were obtained from a systematic entomological survey conducted from May to June , 2010 , to ascertain tsetse presence and abundance . Biconical traps [24 , 25] were used to capture tsetse flies during the survey . Five thousand geo-referenced tsetse trap sites were spread uniformly over a ground area of approximately 40 , 000km2 within the target region [26] . Trapping at each site lasted 72 hours and was conducted by teams led by district entomologists . Single collection was made at the end of this 72 hour period . The parameters recorded in the entomological survey sheet included: trap code , latitude , longitude , altitude , vegetation type around the trap site , start date and time , end date and time , species trapped , number of females , males and flies of un-identified sex , and number of other biting insects . Data were collated and entered into a database . These tsetse data were used as the dependent variable in the regression modelling , while all other variables were used as independent variables . Several covariates ( Table 1 ) were used in the analysis , based on an understanding of the factors important for tsetse reproduction and survival [11 , 27 , 28] . These included; ( i ) land cover , ( ii ) temperature , ( iii ) normalised difference vegetation index ( NDVI ) , ( iv ) elevation , and ( v ) rainfall . The land cover data were extracted from the fine spatial resolution , multi-purpose land cover dataset GlobCover for 2009 [29] . This global land cover series is described by a legend of 22 core land cover categories in total . The region under study contained only 19 of the 22 classes presented . Land cover variables used in the analysis were estimated through the creation of buffers of 1000 m ( catchment ) around each entomological tsetse survey point . Within each buffer , area percentages of the different land cover types were computed and used as the set of land cover predictor variables . NDVI , as a measure of vegetation cover or biomass production , was derived from the National Oceanic and Atmospheric Administration ( NOAA ) Global Inventory Monitoring and Modelling Studies group ( GIMMS ) dataset [29] . The temperature and precipitation data used were obtained as interpolated raster data at a spatial resolution of 30 arc-seconds from the WorldClim—Global Climate Data facility [29] . Elevation data were obtained from the Shuttle Radar Topography Mission ( SRTM ) . Tsetse survey count data were transformed to a binary variable representing tsetse fly presence or absence ( 0 , 1 ) . Presence of tsetse flies was represented by a “1” while absence was represented by “0” . Preliminary visualisation of the geographical distribution of tsetse presence was carried out using the ArcMap10 GIS software ( ESRI , Redlands ) . Exploratory analysis was performed as a means to check for outliers , and aspects of homogeneity , normality and collinearity within the predictor variables . A forward step-wise approach was applied to select the final multivariate logistic regression model . Covariates were added one after the other cumulatively and were retained if they retained statistical significance ( p < 0 . 05 ) . Estimated multivariate regression model coefficients were compared with those obtained at the univariate analysis stage to ascertain the consistency of final covariates in influencing the outcome variable . A residual variogram was constructed to assess the presence of spatial autocorrelation in the model residuals . Autologistic regression was applied to account for the residual spatial autocorrelation [30 , 31 , 32 , 33 , 34 , 35] . This process involves the introduction of a new explanatory variable ( autocovariate ) . Autologistic regression involving several covariates is determined using the formula; lnπ1−πi=α+βs ( yi ) +∑kykxki+εi 1 Where; ɑ is the model intercept β is the coefficient that relates to the autocovariate s ( yi ) is the autocovariate and is a function that summarises the y-values in the neighbourhood of i . It is calculated from the observed data only once and used throughout . γk are the coefficients relating to the k different environmental covariates xki are the k different environmental covariates at location i . Ԑi is the error The spatial autocorrelation was quantified by the Global Morans’s I index as extending up to a distance of 20 km [31 , 32] . Thus , a spatial range of 20 km was used for the calculation of the autocovariate . Receiver operating characteristic ( ROC ) curves were generated to evaluate model performance based on suggested cut-off points . Sensitivity and specificity were used to assess the predictive ability of the model . The area under the ROC curve ( AUC ) was calculated to provide an assessment of how accurately the model can classify the study area into tsetse presence and absence [13 , 36] . Spatial prediction was carried out using the final multivariate model parameters , along with spatially continuous covariate datasets , to enable visualisation of predicted probability of occurrence for both the sampled and unsampled locations . The unsampled locations were represented on a regular grid and the predictions were used to produce continuous surface maps . The probabilities were derived from the regression equation in which the linear predictor was transformed using the logit function into a value between 0 and 1 . Values close to ‘0’ represent a high probability of tsetse absence while ‘1’ represents a high probability of tsetse presence . All analyses were performed using the software R , version: Rstudio2011 , with additional packages; geoR , gstat , MASS and spdep .
A map of tsetse abundance based on the tsetse sampling points is presented in Fig 1 . These data indicate spatially heterogeneous distributions , with high tsetse abundance particularly in the Kalangala islands , along the river Nile , and in the south eastern regions of the study area . In an initial univariate logistic regression stage , 44% of the land cover variables had a statistically significant association with tsetse presence-absence ( p<0 . 05 ) . Covariates; cropland , forest , riverine vegetation , woody vegetation , NDVI , elevation , temperature and rainfall were all positively correlated ( p<0 . 05 , Odds ratio ( OR ) >1 ) , while savannah vegetation , herbaceous vegetation and built-up area were negatively correlated ( p<0 . 05 , OR<1 ) with tsetse presence . Seven covariates were included in the multivariate logistic regression model . The significant covariates were; rainfall , elevation , temperature , cropland , savannah vegetation , forest , and riverine vegetation . Parameter estimates are given in Table 2 . The presence of tsetse flies was negatively correlated with savannah vegetation , and positively correlated with the remainder of the model covariates . However , the covariates cropland , riverine vegetation , elevation and rainfall presented only very small positive associations , with wide confidence intervals . The map of residuals and the residual variogram based on the multivariate logistic regression model revealed the existence of residual spatial autocorrelation . This situation is a problem as it violates the assumption of independence of residuals and can result in biased parameter estimates , leading to inflation of significance . Since non-spatial models fail to account for the autocorrelation effect , there was a need to apply a spatial model: in this case , autologistic regression [37 , 38] . Autologistic regression was applied based on the seven significant variables obtained from the multivariate logistic model together with the computed autocovariate . The resultant statistics are presented in Table 3 and the residual variogram from the autologistic model is shown in Fig 2 . The residual variograms for the two models were compared . The autologistic regression model reduced the spatial autocorrelation in the residuals compared to the multivariate logistic model . In the autologistic model , forest ( p<0 . 05 , OR = 1 . 105 ) and riverine vegetation ( p<0 . 05 , OR = 1 . 008 ) were positively correlated with tsetse presence . Savannah vegetation ( p<0 . 05 , OR = 0 . 993 ) and elevation ( p<0 . 05 , OR = 0 . 997 ) were negatively correlated . These three land cover classes and elevation are , thus , considered to be important determinants of tsetse presence and absence in the study area . Cropland , temperature and rainfall failed to retain their significant association with tsetse presence ( p>0 . 05 ) after accounting for spatial autocorrelation . The Pearson X2 test parameter and Deviance parameter were evaluated as measures of goodness-of-fit . These measures were statistically non-significant ( Pearson X2 = 4654 , p = 0 . 196 ( i . e p>0 . 05 ) and Deviance = 4890 ) , indicating that the model fits the data appropriately and , therefore , could be used to predict probabilities of tsetse presence across the study area . Model evaluation was conducted to assess prediction accuracy . The area under the curve ( AUC ) was computed as 72 . 7% , indicating adequate predictive ability . The plot of sensitivity and false positives ( 1-specificity ) against expected probabilities ( Fig 3 ) indicates a probability cut-off point of 0 . 28 , leading to a sensitivity and specificity of 53% . This is the threshold value for the prediction of tsetse presence where both sensitivity and specificity are maximised , and can be used to classify areas as containing tsetse or not [39] . At a probability cutoff of 0 . 5 , the sensitivity is 10% while specificity is 90% ( Fig 3 ) . This implies that at this cutoff approximately 90% of all the true positive cases ( tsetse presence ) will be missed . As the threshold increases , the sensitivity decreases and the specificity increases . Fig 4 shows the predicted probability of tsetse presence across the study area , based on the multivariate logistic regression model ( non-spatial model ) , while Fig 5 shows the predicted probability of tsetse presence across the study area , based on the autologistic regression model ( spatial model ) . The two models identify areas of scaled potential tsetse fly risk with estimated probabilities of tsetse presence ranging from 0 to 1 . The outcome reflects the presence of a clear tsetse infestation corridor in the Eastern part of the study area . High probability of tsetse occurrence ( predicted probability of occurrence > 75% ) was predicted in the eastern sections of the study area close to the Kenya-Uganda border ( Bugiri , Busia , Tororo Kaliro , Kamuli and Pallisa districts ) as well as on islands located in Lake Victoria . Low probability of tsetse occurrence ( below 20% ) was predicted in the western and north-western parts of the Lake Victoria basin .
Several tsetse sub-species have long been associated with the Lake Victoria basin . The location-specific entomological data gathered for this study provide further evidence of the extensive distribution of tsetse in the area . Using logistic and autologistic regression models coupled with extensive field survey entomological data and a set of environmental covariates , a tsetse distribution map for the lake basin was constructed . These regression models enabled the identification of the important environmental variables determining tsetse presence across the study area . Notably , the final model identified forests and riverine vegetation ( positive ) and savannah vegetation and elevation ( negative ) as the key covariates associated with tsetse presence in the study area . Knowledge of the influential factors and availability of detailed sub-national tsetse distribution maps offers a platform for making meaningful decisions when planning tsetse control interventions . The findings are based on data from Uganda , but the approach is certainly of much broader interest and application .
|
Trypanosomiasis is a vector-borne disease transmitted to both humans and animals by the tsetse fly . The tsetse vector is distributed across sub-Saharan Africa . Trypanosomiasis maintenance is determined by the interrelationship of three elements: vertebrate host , parasite and the vector responsible for transmission . Mapping the distribution and abundance of tsetse flies assists in predicting trypanosomiasis distributions and developing rational strategies for disease and vector control . This study makes available dependable tsetse fly distribution data ( maps ) for use by decision makers . The approach makes use of modelling techniques involving limited field-sampled tsetse data points distributed across an area of approximately 40 , 000km2 within the Lake Victoria basin of Uganda . Precipitation , temperature , landcover , normalised difference vegetation index ( NDVI , a measure of the amount of green vegetation ) and elevation data were used as environmental covariates . We used logistic regression to analyse the relationships between presence of tsetse and the environmental covariates . The results indicated that tsetse are more likely to be present in areas with a greater proportion of riverine vegetation and forest cover . The key outputs are a predicted tsetse distribution map for the Lake Victoria basin of Uganda and an increased understanding of the association between tsetse presence and environmental variables . This will provide a vital resource for the planning and spatial targeting of future tsetse control activities .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Tsetse Fly (G.f. fuscipes) Distribution in the Lake Victoria Basin of Uganda
|
An increased incidence of skin inflammatory diseases is frequently observed in organtransplanted patients being treated with calcineurin inhibitor-based immunosuppressive agents . The mechanism of increased skin inflammation in this context has however not yet been clarified . Here we report an increased inflammation following inhibition of calcineurin signaling seen in both chemically induced mouse skin tumors and in tumors grafted from H-rasV12 expressing primary human keratinocytes ( HKCs ) . Following UVB or TPA treatment , we specifically found that deletion of the calcineurin gene in mouse keratinocytes ( MKCs ) resulted in increased inflammation , and this was accompanied by the enhanced production of pro-inflammatory cytokines , such as TNFα , IL-8 and CXCL1 . Furthermore , expression of the RNA-binding protein , tristetraprolin ( TTP ) was down-regulated in response to calcineurin inhibition , wherein TTP was shown to negatively regulate the production of pro-inflammatory cytokines in keratinocytes . The induction of TTP following TPA or UVB treatment was attenuated by calcineurin inhibition in keratinocytes , and correspondingly , disruption of calcineurin signaling down-regulated the amounts of TTP in both clinical and H-rasV12-transformed keratinocyte tumor models . Our results further demonstrated that calcineurin positively controls the stabilization of TTP in keratinocytes through a proteasome-dependent mechanism . Reducing the expression of TTP functionally promoted tumor growth of H-rasV12 expressing HKCs , while stabilizing TTP expression counteracted the tumor-promoting effects of calcineurin inhibition . Collectively these results suggest that calcineurin signaling , acting through TTP protein level stabilization , suppresses keratinocyte tumors by downregulating skin inflammation .
Calcineurin is the only known serine/threonine phosphatase under calcium/calmodulin control [1] . The active enzyme is a heterotrimeric complex formed by a larger catalytic subunit ( calcineurin A , CnA ) , a Ca2+-binding regulatory subunit ( calcineurin B , CnB ) and calmodulin [2] . CnB is expressed in two isoforms , CnB1 ( ubiquitous ) and CnB2 ( testis-specific ) . Therefore , the inhibition of CnB1 in the skin could completely block calcineurin signaling since only CnB1 is expressed in keratinocytes [3] . The activation of calcineurin dephosphorylates and thus activates Nuclear Factors of Activated T cells ( NFATs ) , which controls the expression of multiple genes related to development and growth , immune responses and inflammatory responses [4 , 5] . The function of calcineurin in the immune system has been elucidated in great detail , where it plays a key role in regulating T and B cell development , and is an important pharmacological target for immune suppression [6] . Cyclosporin A ( CsA ) , which binds and suppresses calcineurin activity in a complex with the cellular protein cyclophilin A [7 , 8] , has been mainly used as an anti-rejection drug for organ transplanted patients [9] . It is known that the higher incidence of skin squamous cell carcinoma ( SCC ) development , as well as increased chronic skin inflammation , occurs in organ transplant patients with long-term use of CsA [10–12] . Several mechanisms have been shown to contribute to the development of SCC in these patients , and increased levels of smoldering inflammation may also be important factors [10 , 11] . However , the mechanism underlying increased skin inflammation with long-term administration of immunosuppressive drugs has not yet been clarified . Inflammation has been shown to play an important role in different stages of tumor development , including initiation , promotion and metastasis [13 , 14] . Inflammation is mainly mediated by the production of cytokines , which recruit immune cells to trigger the inflammatory response . Epidermal keratinocytes , due to their strategic position at the interface between the body and the environment , are the most important type of cell that responds to external stimuli , and usually initiates the inflammatory response by regulating cytokine production during the development of skin inflammation . Cytokine production is controlled at various stages of gene expression , including transcription , mRNA export , and post-transcriptional and translational levels . Recent studies have suggested that the post-transcriptional regulation of RNA stability plays a vital role in controlling the expression of cytokines [15] . The stability of mRNA is mediated mainly by proteins that bind to AU-rich elements ( AREs ) located in 3’-untranslated regions ( 3’UTRs ) of mRNA . A number of ARE-binding proteins have been identified that regulate the stability of mRNA , including tristetraprolin ( TTP ) , human antigen-related protein ( HuR , T-cell-restricted intracellular antigen-1 ( TIA-1 ) , and TIA-related protein ( TIAR ) [15 , 16] . Among those , the zinc finger protein , TTP ( ZFP36 ) encoded by the ZFP36 gene , which belongs to the TIS11/TTP gene family , is the best-characterized [16] . TTP has been reported to regulate the stability of multiple target mRNAs , including pro-inflammatory cytokines and chemokines , such as TNFα , IL-6/IL-8 , GM-CSF and CXCL1 [16–18] . TTP-deficient mice exhibit a severe inflammation syndrome characterized by growth retardation , cachexia , polyarthritis , autoimmunity , myeloid hyperplasia and dermatitis [19] . TTP is believed to promote the ARE-mediated decay of cytokine mRNA [16 , 20] . Recent studies have suggested that TTP plays a tumor suppressing function in cancer development [18 , 21–23] . The deregulation of TTP has been found to play an important role in the progression of various cancers , including inflammation-related cancers , as well as in processes of proliferation , apoptosis , angiogenesis , metastasis , invasion and chemotherapy resistance [18 , 22] . Heretofore , the physiological role of TTP has been mainly assessed in myeloid cells , such as macrophages or dendritic cells ( DCs ) . Recently TTP expression was found in keratinocytes , and a TTP deficiency in keratinocytes resulted in the increased expression of transcripts encoding pro-inflammatory cytokines and chemokines in the skin . The sum of these data suggests that TTP plays a crucial role in the regulation of cytokine production by keratinocytes [24] . However , it still not clear how TTP is activated and what its functions in keratinocytes might be . Emerging data suggest that calcineurin signaling plays a tumor suppressing function in keratinocytes [25] . Our previous study showed that the disruption of calcineurin/NFAT signaling promotes keratinocyte tumorigenesis by blocking p53-dependent cellular senescence , and the accompanying histological findings revealed that there is an increased inflammation response surrounding tumors when calcineurin signaling is inhibited [26] . The goal of the present study was to test whether calcineurin controls keratinocyte tumorigenesis through the regulation of skin inflammation and to explore its underlying mechanism . We now report that calcineurin negatively controls cytokine production by stabilizing TTP , a novel target of calcineurin , and thereby suppresses keratinocyte tumor formation .
We previously showed that disruption of calcineurin signaling , such as by deletion of the CnB1 gene , or treatment with the calcineurin inhibitor CsA , promotes keratinocyte tumorigenesis by blocking p53-mediated cellular senescence [26] . We also noted an increased inflammatory response surrounding these tumors . To confirm that observation , immunofluorescence ( IF ) staining of inflammatory cells using the Gr-1 ( Ly6 ) antibody for granulocytes ( neutrophils ) and the F4/80 antibody for macrophages , showed significantly increased infiltration of inflammatory cells into tumors formed in mice with or without the CnB1 deletion ( KO ) ( S1A and S1B Fig ) . For analysis of the possible infiltration of adaptive immune cells in the tumors , we performed IF staining of different types of T cells with antibodies to CD4 and to CD8 for T cells , to CD16 for NK cells and to CD161 for Treg cells ( S2A–S2D Fig ) . We observed the presence of T cells both in wild-type control ( Ctrl ) and in CnB1 KO ( KO ) tumors , and there was no significant difference between them , indicating that adaptive immune cells didn’t play a major role in the increased inflammation of CnB1 KO skin tumors . Increased inflammation was also identified by analysis of tumor xenografts , derived from our previous study [26] , formed by H-rasV12 expressing primary human keratinocytes ( HKCs ) with concomitant CnB1 knockdown ( siCnB1 ) or in mice subsequently treated with CsA ( S3A and S3B Fig ) . These data suggested that the disruption of calcineurin signaling could result in the up-regulation of innate cells-mediated inflammation in skin tumors . In order to verify that the increased inflammatory response in the surrounding tissue of the generated tumors was directly due to the disruption of calcineurin signaling , TPA or UVB radiation were used to induce acute skin inflammation both in wild-type mice ( Ctrl ) and in mice with keratinocyte-specific deletion of the CnB1 gene ( KO ) . Firstly , the pro-inflammatory compound TPA was applied on the ears of mice , after which the thickness of each ear was measured at different time points ( Fig 1A ) . Ear thickness increased significantly after TPA treatment in KO mice ( KO+TPA ) compared to the control group ( Ctrl+TPA ) ( Fig 1A ) . Histological analysis ( HE stain ) showed that TPA treatment increased ear thickness with an enhanced infiltration of inflammatory cells ( Fig 1B ) . Similar results were obtained when mouse dorsal skin was exposed to UVB and the skin was collected at different time points to analyze its thickness ( Fig 1C ) and its histology ( Fig 1D ) . We also found an increased thickness of subcutaneous edema and a more severe inflammation in CnB1 KO skin compared to the control group ( Fig 1C and 1D ) . To further confirm the increased infiltration of inflammatory cells in the skin of mice with a CnB1 deletion , IF analysis showed significantly more Gr-1 or F4/80 positive cells in CnB1KO skin after UVB exposure for 48 hr compared to the wild-type ( Ctrl ) controls ( Fig 1E–1H ) . No increase of adaptive immune response was observed in Ctrl or KO dorsal skin treated with TPA for 48 hr by IF analysis of different T cell populations ( S4A–S4D Fig ) . These data suggest that calcineurin signaling negatively controls skin inflammation , which is mainly mediated by innate immune cells in keratinocytes . Since inflammatory cell infiltration is mediated by cytokines , we next analyzed whether the increased skin inflammation found after disrupting calcineurin signaling was associated with an increased production of cytokines by keratinocytes . First , we collected ear skin treated with TPA , and analyzed the expression levels of mRNAs encoding pro-inflammatory cytokines , such as TNFα , IL-8 and CXCL1 , using qRT-PCR analysis . Fig 2A shows that treatment with TPA increased the mRNA levels of all 3 pro-inflammatory cytokines , and much higher mRNA levels of cytokines were seen in CnB1 KO skin , which correlated exactly with the increased inflammatory response seen in Fig 1A and 1B . Secondly , we isolated and cultured primary CnB1flox/flox mouse keratinocytes ( MKCs ) and then infected them with a Cre expressing adenovirus ( Ad-Cre ) to obtain CnB1 KO cells . The control cells were infected with a GFP expressing virus ( Ad-GFP ) , then both types of cells were exposed to UVB for 48 hr and analyzed by qRT-PCR for pro-inflammatory cytokines . Fig 2B shows that Cre-mediated deletion of the CnB1 gene enhanced TNFα , IL-8 and CXCL1 expression induced by UVB . Similar results were found using HKCs after the disruption of calcineurin signaling either by siRNA mediated knockdown of CnB1 ( siCnB1 ) ( Fig 2C ) or by CsA to block calcineurin activity ( Fig 2D ) . To further confirm that the inhibition of calcineurin signaling increases the production of cytokines , we performed ELISA assays to measure the protein concentrations of secreted cytokines . We found an increased TNFα protein level in MKCs after the deletion of CnB1 ( Fig 2E ) under both basal conditions and after UVB treatment . Levels of TNFα and IL-8 were significantly increased in HKCs after the knockdown of CnB1 under basal conditions and after TPA treatment ( Fig 2F and 2G ) . Taken together , these data suggested that the inhibition of calcineurin signaling could enhance the production of pro-inflammatory cytokines in keratinocytes in vivo and in vitro . The regulation of cytokines can occur at various levels , and recent studies have suggested that post-transcriptional regulation plays a vital role in controlling their production by modulating mRNA stability through RNA binding proteins . We hypothesized that the negative regulation of cytokine expression by calcineurin signaling in skin keratinocytes is probably through post-transcriptional regulation . To test that hypothesis , primary CnB1flox/flox MKCs infected with Ad-GFP or Ad-Cre viruses and irradiated with UVB were analyzed by Western-blot to determine levels of RNA binding proteins that have been shown to bind ARE and control the stability of 4 cytokine mRNAs: TTP , TIA-1 , TIAR and HuR . We found that , although UVB strongly induces TTP expression in both genotypes , CnB1-deficient keratinocytes express a significantly lower amount of TTP , but the 3 other ARE-binding proteins TIA-1 TIAR and HuR were not affected ( Fig 3A and S5A Fig ) . The down-regulation of TTP in CnB1-deficient primary keratinocytes was also confirmed by immunofluorescence staining with a TTP antibody ( Fig 3B and S5B Fig ) . To further test whether TTP down-regulation in HKCs is caused by the disruption of calcineurin signaling , we performed western-blot analysis of TTP in HKCs with blocking of calcineurin by siRNA of CnB1 ( siCnB1 ) or with CsA or by inhibiting the downstream target of calcineurin by knockdown of NFATc1 ( siC1 ) or NFATc4 ( siC4 ) or by the NFAT inhibitor VIVIT . As shown in Fig 3C , the direct inhibition of calcineurin with a siRNA of CnB1 or with CsA significantly down-regulated the TTP level . In contrast , blocking the calcineurin downstream target NFATs also reduced the level of TTP , but not at a statistically significant level , suggesting that the regulation of TTP level in keratinocytes is likely through calcineurin B activity . Consistent with that observation in vitro , we found that TTP expression was reduced in tumors formed by injection of ras-expressing HKCs in mice treated with CsA ( Fig 3D ) and in tumors formed by ras-expressing HKCs with knockdown of CnB1 by siRNA ( siCnB1 , Fig 3E ) . We also found a lower expression of TTP in most skin and oral SCC cell lines compared with normal primary keratinocytes in vitro ( Fig 3F , S6 Fig ) . Importantly , lower TTP levels were found in tissue microarrays ( TMAs ) of clinically excised SCC tumors from immunocompetent patients undergoing CsA treatment compared to those without CsA treatment , and also compared to normal skin ( NS ) derived from immunocompetent patients without SCCs ( Fig 3G and 3H ) . In summary , these data suggest that the disruption of calcineurin in keratinocytes or in SCC cells , in vitro and in vivo , down-regulates the expression of TTP . Next , we tested whether TTP controls cytokine production by keratinocytes . First , when TTP expression was knocked down by siRNA in HKCs , mRNA levels of TNFα , IL-8 and CXCL1 were up-regulated ( Fig 4A ) . The efficiency of TTP knockdown was confirmed by western-blot analysis ( Fig 4B and S7 Fig ) . The results recapitulated what was found in keratinocytes after calcineurin was inhibited . Second , when HKCs were infected with a lentivirus expressing a full length TTP cDNA devoid of the 3’ UTR region to over-express TTP ( TTP-OE ) ( Fig 4C and 4D ) , the high expression of TTP caused a decreased production of pro-inflammatory cytokines ( Fig 4E ) . We confirmed that protein levels of cytokines ( IL-8 and TNFα ) were induced in keratinocytes following the knockdown of TTP and were reduced in keratinocytes that over-expressed TTP using ELISA analysis ( Fig 4F and 4G ) . In order to investigate the underlying mechanism by which calcineurin inhibition down-regulates TTP , we first analyzed the levels of TTP mRNA using qRT-PCR analysis . We found that TTP mRNA levels actually increased in MKCs with the deletion of CnB1 ( Fig 5A and 5B ) , and this result was confirmed in HKCs by knockdown of CnB1 ( siCnB1 ) and by blocking CnB1 activity with CsA ( Fig 5C and 5D ) . However , the inhibition of NFATs either by Vivit or by knockdown of NFATc1 or NFATc4 didn’t significantly affect the mRNA level of TTP . These data suggested that calcineurin likely regulates the stabilization of TTP protein . In order to test that possibility , we sequentially treated MKCs with UVB to induce TTP expression and with puromycin to prevent novel protein synthesis . We found that in CnB1-deleted cells , TTP protein was significantly destabilized ( Fig 5E ) , and its half-life was clearly reduced ( 94 min; controls: 147 min ) ( S8 Fig ) . To assess the involvement of proteasome-mediated degradation in that process , we analyzed TTP expression in the presence of the potent proteasome inhibitor Epoxomicin ( Fig 5F ) . The differential expression of TTP observed in the control versus KO cells was completely abolished by treatment with Epoxomicin , which suggests that calcineurin controls TTP stability through a mechanism involving proteasomes ( Fig 5F ) . In further support of this mode of regulation , we infected HKCs with a TTP over-expressing lentivirus ( TTP-OE , see Fig 4D ) , after which the stably infected keratinocytes were treated with CsA or with DMSO vehicle alone , followed by immunoblotting to determine TTP protein levels . As shown in Fig 5G , even in TTP over-expressing keratinocytes , CsA treatment caused a strong down modulation of TTP , which was counteracted by the proteasome inhibitor MG132 . Finally , qRT-PCR analysis showed no significant difference in the TTP mRNA level between CsA treated or CsA plus MG132-treated HKCs ( Fig 5H ) . Taken together , these data suggested that calcineurin regulates the stabilization of TTP protein likely through the proteasome degradation pathway . To assess the biological significance of the down-modulation of TTP expression on CsA-tumor promoting effects , a dual approach was undertaken . The siRNA-mediated knockdown of TTP expression was found to enhance the tumorigenicity of H-rasV12 expressing HKCs , tested by intradermal injection into nu/nu mice . Grafts with injected H-rasV12 expressing HKCs with siTTP showed higher cellularity , an increased Ki67 labeling index and increased surrounding cellular inflammation compared to the controls ( Fig 6A–6C ) . Conversely , H-rasV12 expressing HKCs infected with a TTP over-expressing lentivirus and injected intradermally into nu/nu mice followed by CsA treatment exhibited a significantly lower tumorigenic phenotype ( decreased cellularity , decreased Ki67 labeling index and decreased surrounding cellular inflammation ) than H-rasV12 expressing HKCs infected with an empty vector control virus ( Fig 6D–6F ) . HKCs that over-expressed TTP did not give rise to lesions ( Fig 6G ) . These results suggest that the over-expression of TTP at least partially counteracts the tumor formation induced by rasV12 expressing HKCs treated with CsA . Taken together , these data indicate that TTP is a downstream target of calcineurin signaling and plays a tumor-suppressing role in keratinocytes . Next , we sought to further understand the mechanism of the tumor-suppressing function of TTP in the skin since increased inflammation was observed in grafts with knockdown of TTP in keratinocytes . We analyzed changes of cytokine expression and of metabolic genes , which have been reported to play crucial roles in TTP tumor-suppression function[18 , 23] , in SCC13 skin tumor cells with modulations of TTP expression levels . We found that the knockdown of TTP in SCC13 cells induced the expression of the cytokines IL-8 , TNFα , VEGF and COX2 ( S9A Fig ) as well as genes involved in metabolic pathways including the pyruvate dehydrogenase complex ( PDK1 ) , the citric acid cycle ( IDH3A ) , the electron transport chain ( GPD2 ) , branched-chain amino acid metabolism ( BCKDHB ) , purine biosynthesis ( ADSS ) and the pyrimidine salvage pathway ( CMPK1 ) ( S9B Fig ) . In contrast , the over-expression of TTP reduced the expression of cytokines and metabolic genes in SCC13 cells ( S9C and S9D Fig ) .
First we observed increased inflammation surrounding the experimental mouse and human skin tumors derived in our previous study [26] that were induced by the disruption of calcineurin signaling . We further found that the pro-inflammation stimuli of TPA treatment and of UVB exposure induced more acute skin inflammation in mice with the CnB1 gene deleted in keratinocytes . Calcineurin , through its downstream transcriptional factor NFATs , plays a crucial role in regulating T cell and B cell development [1 , 6] . However , our results show that the increased inflammation is mainly mediated by innate immune cells , not by adaptive immune cells in skin with deletion of the CnB1 gene in keratinocytes . The skin inflammation is initiated and mediated by keratinocytes due to their strategic position at the outermost layer of the skin , and keratinocytes communicate with other cells by means of cytokines and chemokines to play an important role in inflammatory skin diseases [27 , 28] . Our results demonstrated a higher expression level of pro-inflammatory cytokines such as TNFα , IL-8 and CXCL1 in MKCs and in HKCs after the disruption of calcineurin signaling ( Fig 2A–2D ) , which was confirmed by ELISA analysis for the secretion of cytokines in the medium ( Fig 2E–2G ) . Interestingly , keratinocytes showed an increased production of pro-inflammatory cytokines caused by the disruption of calcineurin signaling even without TPA or UVB treatment ( Fig 2 ) . These results suggest that calcineurin signaling plays an intrinsic role in negatively regulating cytokine production in keratinocytes . The production of cytokines is regulated at various cellular levels including transcriptional , post-transcriptional and translational levels [29] . Since calcineurin signaling activates NFAT to positively control the transcription of immune response genes , including cytokines , in T cells [6] , the negative regulation of cytokine production by calcineurin in keratinocytes likely acts at the post-transcriptional level . Indeed , our results demonstrate that the well-characterized RNA binding protein TTP is significantly down-regulated in MKCs and in HKCs by the inhibition of calcineurin ( Fig 3A and 3B ) . The down-modulation of TTP by calcineurin was observed both in experimental tumors and in clinical SCCs . Importantly , regulation of the TTP expression level by calcineurin in keratinocytes is specific , because other RNA-binding proteins were not affected by the deletion of CnB1 . Moreover , we found that the down-modulation of TTP was sufficient to recapitulate the effects of calcineurin signaling inhibition , and the over-expression of TTP decreased the production of those cytokines . This suggests that calcineurin controls cytokine production through TTP in keratinocytes . The regulation of TTP occurs at multiple levels of cellular signaling events from transcription , mRNA turnover , and phosphorylation status to proteasomal degradation [16] . At the level of transcription , the gene encoding TTP has been reported to be under the control of TGF-β/Smad signaling [30] . We found that the inhibition of calcineurin by knockdown of CnB1 or CsA , but not by blocking calcineurin downstream targets NFATs , caused a decrease of TTP protein levels while it increased rather than decreased TTP mRNA levels , which might be due to the fact that TTP negatively regulates the stability of its own mRNA [31] . This suggests that the inhibition of calcineurin affects the stabilization of TTP rather than transcriptionally regulating its expression in keratinocytes . However , we can’t completely exclude the possible role of NFAT controlling TTP expression since there are four NFAT isoforms that have functional redundancy . We further found that in cells with the CnB1 deletion , the TTP protein is destabilized and shows a reduced half-life . This result was confirmed by the fact that the destabilization of TTP by the inhibition of calcineurin could be blocked by the proteasome inhibitors epoxomicin or MG132 ( Fig 5D and 5E ) , and proteasome inhibitors didn’t alter the mRNA level of TTP in keratinocytes treated with CsA . These data suggest that calcineurin signaling regulates TTP stabilization through a proteasome-mediated degradation mechanism . Current evidence shows that the phosphorylation status of TTP is important for its stability and activity [16] . The TTP protein can be extensively phosphorylated by multiple signaling pathways: ERK/MAPK , p38 MAPK , JNK and AKT . p38 MAP-activated protein kinase 2 ( MK-2 ) in vitro and in vivo , and protein phosphatase 2 ( PP2A ) can dephosphorylate TTP [32 , 33] . Taken together , we conclude that calcineurin can control the stabilization of TTP through proteasome-mediated degradation , which likely results from a change of its phosphorylation profile . However , the detailed mechanism involved still needs further studies . The aberrant over-expression of ARE-containing genes plays a crucial role in the initiation and progression of tumorigenesis [22 , 34 , 35] . A loss or decreased expression of TTP was observed in various epithelial tumors including breast , cervix , colon , liver , prostate and other cancers [18 , 22 , 23 , 35 , 36] . More and more evidence shows that the loss of TTP promotes tumorigenesis through multiple cancer-associated progressions including enhancing cancer cell proliferation by regulating IL-8 , VEGF and COX2 production , and accelerating the cell cycle [18] . Recently , it was reported that TTP regulates the metabolic process of prostate tumor cells . In agreement with previous reports , our study also demonstrates that TTP is down-regulated in clinical tumor-derived skin and in oral SCC keratinocyte cell lines , as well as in clinical SCCs compared to normal skin . The knockdown of cellular TTP promotes keratinocyte tumorigenesis , possibly through multiple pathways including increased IL-8 , VEGF and COX2 production , and the alteration of metabolic processes of skin tumor cells . These findings support the notion that TTP plays a tumor-suppressing role in the skin . Importantly , down-regulated levels of TTP in skin tumors correspond to the inhibition of calcineurin signaling both in experimental tumors and in clinical SCCs . Notably , the overexpression of TTP could partially block tumorigenesis by inhibiting calcineurin signaling . This suggests that the negative regulation of skin inflammation through TTP stabilization could also contribute to the tumor suppressing function of calcineurin . Future studies will focus on exploring the detailed molecular mechanisms of TTP as a tumor suppressor in keratinocytes . In summary , we discovered that calcineurin negatively controls pro-inflammatory cytokine production by controlling the stabilization of TTP both in MKCs and in HKCs . The disruption of calcineurin signaling facilitates the degradation of TTP , which results in an enhanced pro-inflammatory cytokine/chemokine production , increased skin inflammation and keratinocyte tumorigenesis . The down-regulation of TTP by calcineurin inhibition in keratinocytes could contribute to increased skin inflammation and SCC development in organ-transplanted patients . As a downstream target , TTP plays an important role in the tumor suppressing function of calcineurin .
The procedure for obtaining human foreskin tissues from discarded hospital specimens was reviewed and approved by the Medical Ethical Committee of the School of Stomatology , Shandong University ( Protocol NO: 2015120401 , Date: 12-05-2015 ) . Specimens were analysed anonymously and no patient consent was required . All animal studies were approved by the ethics committee of Stomatological Hospital Shandong University , ( Protocol NO . 2015120402 , Date: 12-05-2015 ) . All the animal procedures in this study were carried out in accordance with National Institutes of Health Guidelines for the Care and Use of Laboratory Animals . MKCs were prepared from newborn mice and were cultured following standard protocols , as previously described [37 , 38] . In order to obtain CnB1 KO MKCs , MKCs from CnB1flox/flox transgenic mice ( see below ) were infected with an adenovirus expressing Cre ( Ad-Cre ) or GFP ( Ad-GFP , as a control ) . Primary HKCs were isolated from foreskin tissues and were cultured in serum-free keratinocyte medium ( SFM , Invitrogen ) as previously described [39–41] . All SCC lines ( SCC12 , 13 , 15 , 25 were derived from the skin , and SCCO11 , O12 , O22 , O23 , O28 were derived from the oral epithelium ) were cultured in SFM . Human foreskin tissues were obtained from discarded hospital specimens following an institutional protocol ( NO . 2015120401 , Date: 12-05-2015 ) . Clinical SCC samples were obtained from the Department of Dermatology , Zurich University Hospital , Zurich , Switzerland . Participants were provided verbal and written informed consent; the protocol was approved by the Swiss Ethics Committee as described previously [26 , 42] . Methods for infecting adenoviruses , lentivirus or retroviruses followed described protocols [26 , 43] . For transient transfection of siRNA , Lipofectamine 2000 ( Invitrogen ) was employed and the final concentration of siRNA in the transfection medium was 200 nM . All siRNA oligo sequences are listed in S2 Table . CsA ( 30024 , Sigma-Aldrich ) was dissolved in DMSO and stored as a stock solution ( 20 mM ) ; TPA ( P8139 , Sigma-Aldrich ) was dissolved in acetone and stored as a stock solution ( 10 mM ) . Both CsA and TPA were kept at -80°C . The Vivit peptide , an inhibitor of calcineurin mediated NFAT activation [44] , and its negative control , the Veet peptide , were chemically synthesized by the Peptide Core facility of the University of Lausanne and were then dissolved in H2O and stored as stock solutions ( 20 mM ) at -80°C . The final concentration of CsA and Vivit used to treat in vitro keratinocyte cultures was 5 μm . For UVB treatment In vitro: After removal of culture medium and two washes with PBS , confluent keratinocytes were covered with PBS and exposed to 35 mJ/cm2 UVB . The UVB dose was measured each time using an IL 1400A photometer ( International Light Inc . , Newburyport , MA ) equipped with a SEL240 probe . After UVB exposure , PBS was removed from the cells and replaced with culture medium . Cells were harvested at different times ( as indicated in each experiment ) after UVB treatment for RNA isolation and total protein extract preparation . Procedures for real time RT-PCR analysis , immunoblotting and immunofluorescence ( IF ) analysis were executed as previously described [47–49] . The list of gene-specific primers is provided in S1 Table . We used the following primary antibodies: polyclonal rabbit anti-TTP ( ab33058 , for IF staining ) , monoclonal rat anti-F4/80 ( ab16911 ) , monoclonal rat anti-F4/80 ( Ly6c ) ( ab15627 ) , polyclonal rabbit anti-CD4 ( ab183685 ) and rabbit monoclonal anti-Ki67 ( ab16667 ) were purchased from Abcam; polyclonal rabbit anti-CD8 ( 98941 ) was purchased from Cell Signalling: polyclonal goat anti-TIA-1 ( sc-1751 ) , mouse monoclonal anti-TIAR ( sc-398372 ) , mouse monoclonal anti-HuR ( sc-5261 ) , mouse monoclonal anti-TNFα ( sc52746 ) and mouse monoclonal anti-γ-tubulin ( sc-17787 ) were purchased from Santa-Cruz; mouse monoclonal anti-CD16 ( NBP2-42228 ) and mouse monoclonal anti-CD161 ( NB100-77528 ) were purchased from Novus Biological . Mouse monoclonal anti-CnB1 ( C0581 ) and anti-ZFP36/TTP ( SAB4200565 , for Western-blot ) ) were purchased from Sigma . Tissue Microarray ( TMA ) analysis was performed as previously described [26] . Eighteen in situ SCC samples from immunocompetent patients , 16 in situ SCCs from CsA treated patients and 12 normal skin samples form immunocompetent patients were analyzed for TTP expression by immunohistochemistry . All staining was performed twice and was evaluated twice by two Independent persons . Evaluations were based on arbitrary units as follows: 1—no or weak staining , 2—intermediate staining and 3—strong staining . For the relative quantification of TTP expression , the mean value of the independent measurements was taken as the final score . Elisa kits to measure secreted cytokines human IL-8 ( VAL103 ) , human TNFα ( VAL105 ) and mouse TNFα ( MTA00B ) were purchased from R&D Systems , Inc ( Minneapolis , MN , USA ) , and the detection procedure followed the manufacturer’s protocol . For protein stability assays , puromycin ( 30 μg/ml in DMSO ) was added to the cell medium to inhibit translation starting 3 hr after UVB treatment . Cells were collected immediately before and 1 , 2 and 4 hr after the addition of puromycin . Cells were then lysed in loading buffer for Western blot analysis , and protein half-lives were extrapolated from densitometric analysis of the electrophoretic bands through linear regression . To duplicate 60 mm dishes , either 1 μM Epoxomicin ( #24801 , EMD ) , 10 μM MG132 or vehicle ( DMSO ) was added directly to the cell medium , starting 3 hr after UVB treatment and subsequently after an additional 2 and 5 hr . Cells were collected at the indicated times and then lysed in loading buffer for Western blot analysis .
|
It is known that a higher incidence of skin squamous cell carcinoma ( SCC ) development , as well as increased chronic skin inflammation , occurs in organ transplant patients with long-term use of calcineurin inhibitors as immunosuppressive drugs . Whether the increased skin inflammation contributes to the development of SCC in these patients is not clear , and importantly , the mechanism of increased skin inflammation with long-term administration of immunosuppressive drugs has not yet been clarified . Here we report that the disruption of calcineurin signaling results in an increased inflammatory response in the skin . The increased skin inflammation was associated with the down-regulation of a RNA-binding protein , Tristetraprolin ( TTP ) , also known as zinc finger protein 36 homolog ( ZFP36 ) . We further demonstrated that calcineurin controls the stabilization of TTP in the skin and the expression of TTP promotes keratinocyte tumorigenesis , while high expression levels of TTP counteract the pro-tumorigenic effects of calcineurin inhibition . Our study is the first to discover that TTP is a novel target of calcineurin and plays a tumor-suppressing role in the skin . Further , our data suggest that inflammation may be an important factor in the development of SCC in organ transplant patients .
|
[
"Abstract",
"Introduction",
"Results",
"Discussions",
"Methods"
] |
[
"dermatology",
"innate",
"immune",
"system",
"keratinocytes",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"cytokines",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"skin",
"tumors",
"immunology",
"cancer",
"treatment",
"cancers",
"and",
"neoplasms",
"epithelial",
"cells",
"oncology",
"skin",
"neoplasms",
"developmental",
"biology",
"signal",
"inhibition",
"signs",
"and",
"symptoms",
"molecular",
"development",
"small",
"interfering",
"rnas",
"inflammation",
"animal",
"cells",
"gene",
"expression",
"biological",
"tissue",
"immune",
"response",
"immune",
"system",
"biochemistry",
"signal",
"transduction",
"rna",
"calcineurin",
"signaling",
"cascade",
"diagnostic",
"medicine",
"cell",
"biology",
"anatomy",
"nucleic",
"acids",
"physiology",
"genetics",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"non-coding",
"rna",
"cell",
"signaling",
"signaling",
"cascades"
] |
2018
|
The ARE-binding protein Tristetraprolin (TTP) is a novel target and mediator of calcineurin tumor suppressing function in the skin
|
Tens of millions of dengue cases and approximately 500 , 000 life-threatening complications occur annually . New tools are needed to distinguish dengue from other febrile illnesses . In addition , the natural history of pediatric dengue early in illness in a community-based setting has not been well-defined . Data from the multi-year , ongoing Pediatric Dengue Cohort Study of approximately 3 , 800 children aged 2–14 years in Managua , Nicaragua , were used to examine the frequency of clinical signs and symptoms by day of illness and to generate models for the association of signs and symptoms during the early phase of illness and over the entire course of illness with testing dengue-positive . Odds ratios ( ORs ) and 95% confidence intervals were calculated using generalized estimating equations ( GEE ) for repeated measures , adjusting for age and gender . One-fourth of children who tested dengue-positive did not meet the WHO case definition for suspected dengue . The frequency of signs and symptoms varied by day of illness , dengue status , and disease severity . Multivariable GEE models showed increased odds of testing dengue-positive associated with fever , headache , retro-orbital pain , myalgia , arthralgia , rash , petechiae , positive tourniquet test , vomiting , leukopenia , platelets ≤150 , 000 cells/mL , poor capillary refill , cold extremities and hypotension . Estimated ORs tended to be higher for signs and symptoms over the course of illness compared to the early phase of illness . Day-by-day analysis of clinical signs and symptoms together with longitudinal statistical analysis showed significant associations with testing dengue-positive and important differences during the early phase of illness compared to the entire course of illness . These findings stress the importance of considering day of illness when developing prediction algorithms for real-time clinical management .
Dengue virus ( DENV ) causes the most prevalent mosquito-borne viral disease affecting humans , with 2 . 5–3 billion people at risk for infection and approximately 50 million cases of dengue each year [1] , [2] . The four DENV serotypes are transmitted to humans by Aedes aegypti and Ae . albopictus mosquitoes , primarily in urban and peri-urban areas in tropical and subtropical countries worldwide . Most cases present as classic dengue fever ( DF ) , a debilitating but self-limited illness that manifests with high fever , retro-orbital pain , severe myalgia/arthralgia , and rash . However , in some cases , mainly children , illness progresses to life-threatening dengue hemorrhagic fever/dengue shock syndrome ( DHF/DSS ) , characterized by vascular leakage leading to hypovolemic shock and a case fatality rate up to 5% [1] , [3] , [4] . Currently , no licensed vaccine or antiviral therapy exists for dengue . Early identification of patients at risk of developing severe dengue is critical to provide timely supportive care , which can reduce the risk of mortality to <1% [1] , [2] . However , distinguishing dengue from other febrile illnesses ( OFIs ) early in illness is challenging , since symptoms are non-specific and common to other febrile illnesses such as malaria , leptospirosis , rickettsiosis , and typhoid fever [5]–[7] in dengue-endemic countries . In addition , many distinguishing clinical features of DHF/DSS generally emerge only after 4–5 days , at defervescence , when the patient is already critically ill . Although the World Health Organization ( WHO ) has recently established new clinical guidelines to classify dengue severity [1] , serological , virological , and molecular biological tests are required to definitively diagnose DENV infection . In many endemic countries , laboratory diagnosis of dengue is often problematic due to lack of reagents , expense , or delay in obtaining results . Patients with suspected dengue are often hospitalized for close monitoring to ensure proper treatment if they begin to develop severe dengue; however , up to 38–52% are later diagnosed with OFIs [8] , [9] and thus were hospitalized unnecessarily at great financial cost to their family and society [10] . New tools are therefore needed to distinguish dengue from OFIs to prevent deaths from severe dengue and to mitigate the economic burden of excess hospitalization . Recent approaches using multivariable logistic or linear regression models have shown that petechiae , thrombocytopenia ( platelet count ≤100 , 000 cells/mm3 ) , positive tourniquet test , rash , and other signs and symptoms can distinguish dengue from OFIs [11]–[17]; however , results were not consistent across studies . Only two studies considered clinical and laboratory features according to day of illness [18]–[20] , but as these were hospital-based studies , the results likely reflect patients with more severe symptoms and not the clinical spectrum of all symptomatic cases in dengue-endemic populations . Furthermore , none of these studies analyzed data using longitudinal statistical methods , which account for correlations between repeated measures on individuals over time . The use of longitudinal statistical methods to analyze cohort data is essential to utilize all of the data available for analysis and appropriately estimate the within-person and between-person variance in measures over time . In this study , we used five years of data from an ongoing prospective cohort study of approximately 3 , 800 children aged 2–14 years in Managua , Nicaragua , to examine the frequency of clinical signs and symptoms by day of illness and to generate models for the association of signs and symptoms during the early phase of illness and over the entire course of illness with testing dengue-positive . In order to account for the longitudinal structure of the data , odds ratios ( ORs ) and 95% confidence intervals were calculated using generalized estimating equations ( GEE ) , adjusting for age and gender .
In August and September 2004 , a community-based pediatric cohort was established in District II of Managua , a low-to-middle income area with a population of approximately 62 , 500 [21] . Study activity was based in the Health Center Sócrates Flores Vivas ( HCSFV ) , a public facility that is the primary source of health care for District II residents . Briefly , participants aged 2–9 years were recruited through house-to-house visits , and additional two year-olds were enrolled each year to maintain the age structure of the cohort [21] . Children were eligible to remain in the study until age 12 or until they moved from the study area . The parent/legal guardian of each participant signed an informed consent form , and children ≥6 years old provided verbal assent . In 2007 , participants ≤11 years old were given the opportunity to continue for an additional 3 years , and a second informed consent was performed . The study was approved by the Institutional Review Boards of the University of California , Berkeley , the Nicaraguan Ministry of Health , and the International Vaccine Institute in Seoul , Korea . Parents or legal guardians of all subjects in both studies provided written informed consent , and subjects 6 years of age and older provided assent . Upon enrollment , parents/legal guardians of all participants were encouraged to bring their child ( ren ) to the HCSFV at first sign of illness or fever . Study physicians and nurses , trained in identification of possible dengue cases , provided medical care for study participants . Febrile illnesses that met the WHO criteria for suspected dengue ( Table 1 ) and those without other apparent origin ( undifferentiated febrile illnesses ) were treated as possible dengue cases and followed daily while fever or symptoms persisted through visits with study medical personnel ( Figure 1 ) . Complete blood counts ( CBCs ) were completed every 48 hours or more frequently as necessary , as indicated by the physician . Cases were monitored closely for severe manifestations and were transferred by study personnel to the Infectious Disease Ward of the Manuel de Jesús Rivera Children's Hospital , the national pediatric reference hospital , when they presented with any sign of alarm ( Table 1 ) . In addition , an annual healthy blood sample was collected to identify all DENV infections during the previous year and for baseline CBC values . Study physicians in both the hospital and HCSFV completed systematic data collection forms that contained approximately 80 variables ( Table 1 ) . In the hospital , additional clinical data , including fluid balance and treatment , were collected daily during hospitalization or through ambulatory follow-up visits by a team of study physicians and nurses . Data were also recorded on medical tests ordered and treatments prescribed . A case was considered laboratory-confirmed dengue when acute DENV infection was demonstrated by: detection of DENV RNA by RT-PCR; isolation of DENV; seroconversion of DENV-specific IgM antibodies observed by MAC-ELISA in paired acute- and convalescent-phase samples; and/or a ≥4-fold increase in anti-DENV antibody titer measured using Inhibition ELISA [22]–[25] in paired acute and convalescent samples . DENV serotypes were identified by RT-PCR and/or virus isolation . Laboratory-confirmed dengue cases were further classified by severity . DHF and DSS were defined according to the traditional WHO criteria ( Table 1 ) [26] . Additional categories of severity were included for those cases presenting with shock without thrombocytopenia and/or hemoconcentration ( dengue with signs associated with shock ( DSAS ) ) [23] or dengue fever with compensated shock ( DFCS ) [27] ( Table 1 ) . Laboratory-confirmed cases were defined as primary DENV infections if acute-phase antibody titer , as measured by Inhibition ELISA , was <1∶10 or if convalescent phase antibody titer was <1∶2560 , and as secondary infections if the acute titer was ≥1∶10 or convalescent titer was ≥1∶2560 [22]–[25] . Data from the first five years of the study ( August 30 , 2004–June 30 , 2009 ) were used for analysis . The first three days after onset of fever were considered the early febrile phase of illness . Day of illness at presentation was determined by the date of fever onset , which was defined as the first day of illness as reported by the parent/guardian . Variable definitions are described in Table 1 . Positive tourniquet test was examined using cut-offs of ≥10 petechiae/in2 and ≥20 petechiae/in2 . Platelet count was dichotomized using a cut-off of ≤150 , 000 cells/mm3 to enable comparisons during days 1–3 . Only data from days 1–8 of illness were included for analysis . The frequency of dengue testing results ( laboratory-confirmed dengue-positive versus dengue-negative ) and disease severity ( DF versus severe dengue ) was examined by year , demographics , serotype and immune response . The frequency of the WHO case definition for suspected dengue was examined by dengue testing results and age , and a chi-square test for trend was performed . The frequency of clinical signs and symptoms by day of illness and dengue severity was also examined using chi-square tests . To examine the association between clinical signs and symptoms and the odds of testing dengue-positive versus dengue-negative , odds ratios ( ORs ) and 95% confidence intervals ( CIs ) were calculated using GEE models assuming an exchangeable correlation structure with robust standard errors to account for the correlations between repeated measures on the same patients over time . First , ORs were calculated using bivariable models that included only dengue testing results and each of the signs or symptoms . All signs and symptoms were then examined in multivariable models that adjusted for age and gender . Data from the first three days of illness and from all days of illness only were analyzed separately . Finally , for comparison , we used traditional logistic regression models to analyze the association between signs and symptoms and testing dengue-positive with data collapsed by illness episode to disregard repeated measures on the same patients , using the same model generation process as for the GEE models . All analyses were conducted using STATA 10 ( StataCorp LP , College Station , TX ) .
From August 2004 to June 2009 , 22 , 778 episodes of febrile illness were evaluated , of which 1 , 974 episodes were suspected dengue or undifferentiated fever ( Figure 1 ) . Of the 1 , 974 possible dengue cases , 1 , 793 ( 91% ) tested negative and 181 ( 9% ) were laboratory-confirmed as dengue-positive , of which 161 ( 89% ) were classified as DF , 9 ( 5% ) as DHF , 4 ( 2% ) as DSS , 3 ( 2% ) as DSAS and 4 ( 2% ) as DFCS ( Table 1 ) . Nearly all ( 95% ) of the severe dengue cases but only 116 ( 72% ) of the DF cases met the WHO case definition for dengue . The proportion of laboratory-confirmed DENV infections that met the WHO case definition significantly increased by age ( chi-square test for trend 5 . 977 , p = 0 . 01 ) , while younger children experienced significantly more undifferentiated febrile illness due to DENV infection ( Figure 2 ) . The median age for cases meeting the dengue case definition was 8 years ( range 2–13 ) and that of undifferentiated febrile illness due to DENV infection was 6 years ( range 2–10 ) . The number of confirmed dengue-positive cases varied by year , as expected ( Table 2 ) [28] . Both genders were equally represented , with a slightly higher percentage of females experiencing severe dengue , though this difference was not statistically significant . The majority of DF cases were DENV-2 ( 58% ) , followed by DENV-1 ( 21% ) and DENV-3 ( 9% ) , while 60% of severe dengue cases were DENV-2 , followed by DENV-3 ( 25% ) and DENV-1 ( 10% ) . In addition , there were nearly equal proportions of primary and secondary immune responses among DF cases , whereas the majority ( 70% ) of severe dengue cases were secondary DENV infection ( Table 2 ) . The median day of illness at presentation was day 2 for all patients , and almost all presented on days 1–3 of illness ( 90% ) . The total follow-up time of all children in the cohort was 17 , 931 person-years with a median follow-up of 3 . 9 years per child . As shown in Figure 3 , several signs and symptoms appeared to differentiate OFIs from DF cases , and DF cases from severe dengue cases , according to day of illness . In particular , higher proportions of DF and severe dengue cases experienced petechiae , platelets ≤150 , 000 cells/mm3 , leukopenia , and positive tourniquet test compared to patients with OFIs . Higher proportions of severe cases experienced petechiae , platelets ≤150 , 000 cells/mm3 , myalgia/arthralgia and abdominal pain compared to DF cases and patients with OFIs . Abdominal pain differentiated severe dengue cases from DF and OFI only beginning on day 3 of illness ( for severe dengue compared to DF: chi-square 0 . 144 , p = 0 . 70 for days 1–2 versus chi-square 16 . 910 , p<0 . 0001 for day ≥3 ) . Bivariable and multivariable analyses were performed using GEE models to examine signs and symptoms early in illness and over the course of illness ( Table 3 ) . On days 1–3 of illness , dengue-positive cases had up to 2-fold increased odds of fever , headache , retro-orbital pain , myalgia , arthralgia , and vomiting compared to patients with OFIs . They also had from 3-fold to 9-fold increased odds of rash , petechiae , positive tourniquet test with cut-offs of ≥10 and ≥20 petechiae/in2 , leukopenia , platelets ≤150 , 000 cells/mm3 , poor capillary refill , cold extremities and hypotension compared to patients with OFIs . In contrast , they had decreased odds of abdominal pain , likely because this feature appears later in the entire course of dengue illness . On all days of illness , dengue-positive cases had increased odds of the same signs and symptoms as on days 1–3 of illness; however , the magnitude of the point estimates tended to be higher . This difference was most pronounced for rash and platelets ≤150 , 000 cells/mm3 , which had ORs approximately double in magnitude . In addition , dengue-positive cases had increased odds of three additional signs and symptoms: poor appetite , absence of cough , and increased hematocrit . When GEE analyses on data with the longitudinal structure preserved were compared to traditional logistic regression analyses on data collapsed on febrile episode , the point estimates for the ORs were similar , although the 95% confidence intervals for the logistic regression models tended to be slightly narrower ( data not shown ) .
In this study , we describe the clinical spectrum of pediatric dengue starting early in illness in a community setting . Longitudinal statistical analysis of day-by-day clinical signs and symptoms revealed significant associations with testing dengue-positive and important differences during the early phase of illness compared to the entire course of illness . These results stress the importance of considering day of illness when developing prediction algorithms for real-time clinical management . The early identification of dengue cases and particularly those at risk for severe dengue is critical for preventing severe illness and death . We found that 25% of laboratory-confirmed dengue cases did not meet the WHO case definition , suggesting that the WHO criteria are not sufficient to identify dengue at younger ages . Younger children may experience different signs and symptoms from adults or may be unable to communicate their symptoms to their parents , health care providers , or both . Previous studies demonstrated that children may experience significantly more cough , vomiting , abdominal pain , rash , epistaxis , oliguria , thrombocytopenia , hepatomegaly , and shock compared to adults , although the direction of these differences was not consistent across studies [13] , [15] , [29]–[34] . A recent study of dengue in adults showed significant differences in clinical features and outcomes across ten-year age groups , indicating that signs and symptoms associated with DENV infection may continue to evolve past childhood [12] . If these differences are confirmed , the WHO case definition may need to be adjusted to be age-specific to function effectively for younger children and older age groups . Retro-orbital pain and low platelets were among the clinical features independently associated with DENV infection in this study . These results are supported by a study of dengue patients in Puerto Rico in which data were recorded at the time of initial consult rather than at hospitalization [15] , and by a study of Thai children [11] . Moreover , our results showing increased frequency of abdominal pain in patients beginning at day 3 of illness are consistent with a prospective study of adults admitted to an emergency department in Martinique [35] . A positive tourniquet test using cut-offs of ≥10 and ≥20 petechiae/in2 was also independently associated with DENV infection . Both cut-offs were used because studies have indicated that a cut-off of ≥10 may improve discrimination of DENV infection [20] , [36]; however , the 1997 WHO classification scheme specified a cut-off of ≥20 [26] . Our results support using a cut-off of ≥10 petechiae/in2 , and this cut-off has been specified in the 2011 WHO clinical guidelines [37] . A major strength of this study is the use of statistical models designed for analysis of longitudinal data . Few other prospective community-based cohort studies have analyzed early clinical features in pediatric dengue compared to OFI [20] , [38]–[40] , and none that we are aware of were analyzed using longitudinal statistical methods that account for correlations between repeated measures on patients . Here , we preserved the longitudinal structure of the dataset by using statistical models that support repeated measurements on subjects over time and account for correlations between signs and symptoms experienced within the same individual on different days of illness and in multiple episodes . Longitudinal data have long been collected in dengue research but have rarely been analyzed using appropriate statistical methods . This may introduce bias into findings , as studies may overestimate the magnitude of association or reduce the statistical power of the study as data are lost when they are collapsed for non-longitudinal analysis . An additional strength of this study is that it is community-based [21] , enabling day-by-day capture of information on the early course of illness and on the full clinical spectrum of symptomatic dengue . In contrast , nearly all previous studies enrolled patients upon presentation to a hospital [18] , where patients present later; thus , these studies were unable to capture information on the early days of illness or on mild disease . By examining the clinical spectrum of dengue by day of illness , we were able to detect differences in the prevalence of signs and symptoms that could not be revealed by simply analyzing whether they ever occurred over the course of illness . In addition , through multivariable longitudinal models , we were able to identify distinguishing features of dengue during the early phase of illness compared to the entire course of illness . These findings are important for clinical practice since outside of the hospital setting , clinicians may see dengue patients toward the beginning of their illness and utilize that information to decide whether their patient has dengue or another febrile illness . The results of these models should be extended for the development of prediction algorithms to aid clinicians in diagnosing suspected dengue . This study was not without its limitations . Some participants migrated out of the study area or withdrew from the study; however , our retention rate was approximately 95% per year [21] , suggesting that any bias from loss to follow-up would be minimal . It is also possible that we did not capture all symptomatic dengue cases . However , in yearly participant surveys , only an average of 2–3% of participants reported having attended a health-care provider outside of the study or having an illness and not attending any medical provider [21] , and approximately 20-fold more laboratory-confirmed dengue cases were captured in the cohort study than by the National Surveillance System [41] . Unfortunately , due to the low number of severe dengue cases , this study did not have sufficient statistical power to compare severe dengue cases to DF cases using GEE models , and these low numbers may have influenced the lack of significant association of signs of severe dengue with testing dengue-positive . For this study , we used the 1997 WHO classification scheme for disease severity . In 2009 , the WHO updated its guidelines for classification of dengue disease severity [1] , [37]; it would be interesting to re-analyze the data in a future study using the new classification scheme . Studies of the usefulness and applicability of the revised guidelines have been recently performed [42] , [43] . In summary , this study is one of the few cohort studies to provide early data on the full clinical spectrum of pediatric dengue . Though we found significantly increased odds for association of several clinical signs and symptoms with testing dengue-positive , these increases were more modest for the early phase of illness compared to the course of illness , suggesting that caution should be taken when using the results from the entire course of illness to develop prediction algorithms . Non-parametric methods such as decision tree analysis overcome some of the limitations of traditional logistic regression models and have recently been applied to develop algorithms for prediction of dengue diagnosis and disease severity [9] , [44] , [45] . These and other data-adaptive approaches such as Super Learner [46] that are less subject to bias should be further explored to develop prediction algorithms for early identification of dengue cases and improved clinical management .
|
Dengue virus causes an estimated 50 million dengue cases and approximately 500 , 000 life-threatening complications annually . New tools are needed to distinguish dengue from other febrile illnesses . In addition , the natural history of pediatric dengue early in illness in a community-based setting has not been well-defined . Here , we describe the clinical spectrum of pediatric dengue over the course of illness in a community setting by using five years of data from an ongoing prospective cohort study of children in Managua , Nicaragua . Day-by-day analysis of clinical signs and symptoms together with longitudinal statistical analysis showed significant associations with testing dengue-positive and important differences during the early phase of illness compared to the entire course of illness . These findings are important for clinical practice since outside of the hospital setting , clinicians may see dengue patients toward the beginning of their illness and utilize that information to decide whether their patient has dengue or another febrile illness . The results of these models should be extended for the development of prediction algorithms to aid clinicians in diagnosing suspected dengue .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"clinical",
"research",
"design",
"epidemiology"
] |
2012
|
Early Clinical Features of Dengue Virus Infection in Nicaraguan Children: A Longitudinal Analysis
|
The d-arabinan-containing polymers arabinogalactan ( AG ) and lipoarabinomannan ( LAM ) are essential components of the unique cell envelope of the pathogen Mycobacterium tuberculosis . Biosynthesis of AG and LAM involves a series of membrane-embedded arabinofuranosyl ( Araf ) transferases whose structures are largely uncharacterised , despite the fact that several of them are pharmacological targets of ethambutol , a frontline drug in tuberculosis therapy . Herein , we present the crystal structure of the C-terminal hydrophilic domain of the ethambutol-sensitive Araf transferase M . tuberculosis EmbC , which is essential for LAM synthesis . The structure of the C-terminal domain of EmbC ( EmbCCT ) encompasses two sub-domains of different folds , of which subdomain II shows distinct similarity to lectin-like carbohydrate-binding modules ( CBM ) . Co-crystallisation with a cell wall-derived di-arabinoside acceptor analogue and structural comparison with ligand-bound CBMs suggest that EmbCCT contains two separate carbohydrate binding sites , associated with subdomains I and II , respectively . Single-residue substitution of conserved tryptophan residues ( Trp868 , Trp985 ) at these respective sites inhibited EmbC-catalysed extension of LAM . The same substitutions differentially abrogated binding of di- and penta-arabinofuranoside acceptor analogues to EmbCCT , linking the loss of activity to compromised acceptor substrate binding , indicating the presence of two separate carbohydrate binding sites , and demonstrating that subdomain II indeed functions as a carbohydrate-binding module . This work provides the first step towards unravelling the structure and function of a GT-C-type glycosyltransferase that is essential in M . tuberculosis .
Tuberculosis ( TB ) affects large parts of the world's population , particularly in developing countries [1] . The antibiotics isoniazid ( INH ) and ethambutol ( EMB ) [2] have been used for decades as frontline drugs to treat Mycobacterium tuberculosis infections , the causative agent of TB , but the rise of multi-drug resistant ( MDR ) and extensively drug resistant ( XDR ) strains poses a serious threat to present treatment options [3] . Both , INH and EMB inhibit the synthesis of essential components of the mycobacterial cell wall . This unique and highly impermeable barrier surrounds a single phospholipid bilayer membrane and is composed of an outer segment of solvent-extractable lipids , glycans and proteins , and a covalently linked inner segment , known as the mycolyl-arabinogalactan-peptidoglycan ( mAGP ) core [4] . Perturbations to the mAGP core tend to undermine viability of M . tuberculosis , a major reason why mAGP biosynthesis constitutes an attractive target for drug design efforts . The mycobacterial cell wall also encompasses various membrane-anchored lipoglycans , a group that includes lipoarabinomannan ( LAM ) , which plays a key role in modulating the host immune response [5] . The arabinogalactan ( AG ) segment of the mAGP core and LAM both contain d-arabinan polymer , composed of α ( 1→5 ) , α ( 1→3 ) and β ( 1→2 ) -linked arabinofuranosyl ( Araf ) residues that are assembled in distinct structural motifs ( Fig . 1A ) [4] , [5] . In recent years , substantial progress has been made in defining the enzymatic processes resulting in the complete synthesis of AG and LAM [6]–[14] . Probing susceptibility to EMB , initial studies established that this inhibitor acted on a set of closely related arabinofuranosyl ( Araf ) transferases , EmbC ( Rv3793 ) , EmbA ( Rv3794 ) and EmbB ( Rv3795 ) [6] , [7] , collectively referred to as the Emb enzymes . These three proteins belong to the glycosyltransferase superfamily C ( GT-C ) , which encompasses a diverse set of membrane-embedded glycosyltransferases that utilise lipid-linked as opposed to nucleotide-linked sugars as donor substrates ( Fig . 1A ) [15] . The Emb enzymes of M . tuberculosis display a common architecture of 13 transmembrane helices in conjunction with a hydrophilic C-terminal domain [10] , [14] ( Fig . 1B ) , and share the same polyprenyl donor-substrate , β-D-arabinofuranosyl-1-monophosphoryldecaprenol ( DPA ) [16] , [17] . Owing to their hydrophobic nature , generating recombinant Emb proteins in soluble form has proved difficult , hampering in vitro characterisation . As a result , the function of the Emb enzymes has been delineated by genetics , phenotypic analysis of the cell envelope and cell-free assays . Single gene deletions of embC , embB in M . tuberculosis are lethal [18] , [19] , but corresponding knock-outs in Mycobacterium smegmatis or Corynebacterium glutamicum yield viable , albeit slow growing mutants , whose cell wall defects can be analysed [8] , [9] . Following attachment of the initial Araf residue to the linear galactan polymer [Galf-β ( 1→5 ) Galf-β ( 1→6 ) ]n , catalysed by the Araf-transferase AftA [12] , EmbA and EmbB extend the arabinan chain in AG synthesis , transferring Araf residues from DPA to polysaccharide acceptors [8] , [9] . Highly similar in amino acid sequence ( ∼40% identity , see also Supporting Fig . S1 ) , EmbA , B and EmbC have differential roles: the ΔembA , B deletions inhibit AG synthesis , but leave LAM synthesis intact , whereas the ΔembC deletion only affects LAM synthesis . Chimaeric forms of the Emb enzymes , where the hydrophilic C-terminal domain of EmbC was swapped for that of EmbB led to a hybrid-LAM , bearing an AG-specific , branched Araf6 group instead of the characteristic LAM-specific linear Araf4 [9] . These data indicated that the hydrophilic C-terminal domain makes a critical contribution to determining the structure of the resulting AG or LAM segments . To date , the Emb enzymes have remained poorly characterised in structural terms , despite their central significance as targets of the TB antibiotic EMB and their link to drug resistance [20] . Herein , we present the crystal structure of the C-terminal hydrophilic domain of M . tuberculosis EmbC ( residues 719–1094 , henceforth EmbCCT ) , as a first step towards the elucidation of the 3D structure of the full-length enzyme .
EmbCCT crystallised in space group P6522 over a diverse range of reservoir conditions , with one molecule in the crystallographic asymmetric unit . Crystals were generated with or without an Araf acceptor analogue ( see below ) present in the crystallisation droplet . The experimental density , phased by multi-wavelength anomalous dispersion ( 2 . 7 Å , Table 1 ) , was of very good quality ( Fig . S2A ) , defining the structure for residues 735–1067 , except for two disordered loops ( 795–824 and 1016–1037 , Fig . 2A ) . EmbCCT is composed of two distinct subdomains , separated by a deep crevice marked by the disordered loops ( residues 795–824 and 1016–1037 ) . Subdomain I , which encompasses residues 746–760 and 967–1067 , displays a mixed α/β structure , with a 5-stranded β-sheet forming a semi-barrel ( Fig . 2A ) . The long H6-S13 loop , which forms a minor crystal packing interface , protrudes from the core of subdomain I with a helical half-turn at its tip ( Fig . 2A ) . Subdomain II ( residues 761–966 ) forms an anti-parallel β-sandwich structure , of which the ‘outer’ sheet ( S2 , S4 , S10 , S6 , S7 ) faces solvent while the ‘inner’ sheet ( S3 , S11 , S5 , S9 , S8 ) packs against the core of the domain ( Fig . 2A ) . The β-sandwich of subdomain II assumes a jellyroll fold ( Fig . 2B ) , a fold typical for polysaccharide binding units in plant lectins and carbohydrate active enzymes [21] . Although not part of the formal jellyroll description , strands S2 and S8 extend the ‘outer’ and ‘inner’ sheet , respectively , while helix H4 forms a boundary to the ‘outer’ sheet . A high-density peak ( 14σ , anomalous density difference map , Fig . 3A ) is embedded between loops S3–S4 and S10–S11 . Quasi-octahedral coordination geometry and the distribution of peak-ligand distances from 2 . 40 to 2 . 63 Å ( Fig . 3A ) suggest a bound Ca2+ ion [22] . The metal ion appears shielded from solvent , although including 10 mM EDTA in the cryoprotectant buffer significantly diminished the height of the density peak ( Fig . S2B ) . Substitution of Asp949 by serine in EmbCCT , the only side chain in direct contact with the Ca2+ ion ( 2 . 6 Å , bidentate , Fig . 3A ) , resulted in very poor recombinant expression compared to wild-type and other point mutants probed in this study ( see below ) . Together these observations suggest that the Ca2+ ion is important for the structural integrity of EmbCCT . The fold of subdomain II is consistent with the proposed role of EmbCCT as an acceptor saccharide recognition module . The comparison with structural homologues , identified via distance matrix alignment using the DALI program ( http://ekhidna . biocenter . helsinki . fi/dali_server/ , [23] ) reinforces this notion . The vast majority of PDB entries retrieved by DALI ( over 300 entries above the default significance threshold of Z = 2 ) match the β-sandwich fold of subdomain II and represent ‘carbohydrate binding modules’ ( CBM ) , structural domains that confer carbohydrate-binding specificity , but that lack intrinsic catalytic activity [21] . CBMs occur frequently as a part of glycoside hydrolase enzymes and fall into ( to date ) 61 distinct CBM families ( http://www . cazy . org/ ) . While none of the structural homologues is particularly close to subdomain II ( Z-scores≤6 . 9 , root mean square deviation ( RMSD ) ≥3 . 0 Å ) , the top 10 hits include the calcium-containing CBM families 6 and 36 ( Fig . S3A–C ) . Interestingly , in the DALI-generated superposition of EmbCCT with Paenibacillus polymyxa endo-1 , 4-β-xylanase ( PDB entry 1UX7 , CBM 36 ) , the Ca2+ sites match to within 0 . 9 Å , and in the latter , the Ca2+ ion makes direct contact with the bound xylobiose ligand ( Fig . S3A ) . In contrast , only three hits were obtained for subdomain I of which only the best ( PDB entry 2ZAG , Z = 3 . 0 , RMSD 3 . 4 Å for 66 Cα pairs ) showed weak similarity in terms of secondary structure topology in a limited region of overlap ( Fig . S4 ) . This PDB entry describes the hydrophilic C-terminal domain of oligosaccharyltransferase STT3 from Pyrococcus furiosus [24] , a membrane-embedded glycosyltransferase of the GT-C superfamily that catalyses transfer of glycosyl groups from a lipid donor to Asn-glycosylation sites of the acceptor protein . Crystal packing contacts , analysed using the PISA server ( http://www . ebi . ac . uk/msd-srv/prot_int/pistart . html ) , highlighted three prominent interaction surfaces burying 390 Å2 , 670 Å2 and 1100 Å2 of solvent accessible surface ( SAS ) per monomer , respectively ( Fig . S5 ) . We probed self-assembly of EmbCCT by sedimentation velocity at three different protein concentrations ( Fig . 4A ) . The distribution C ( S ) of the sedimentation coefficient S indicates a dynamic equilibrium between three different molecular species at 3 . 1S , 4 . 6S and 7S , which correspond to apparent molecular weights of 46 . 5 kDa , 75 . 8 kDa and 138 . 0 kDa , respectively , compared to the calculated monomer mass of 39 . 9 kDa . Bearing in mind that under- or overestimates of apparent masses can occur as a result of fitting a single frictional coefficient for an ensemble of species with different frictional ratios , the dominant peak at 4 . 6S most likely represents a dimer . The higher molecular weight peak at 7 . 6S , could be a trimer or tetramer , but strongly suggests that more than one of the crystal packing interfaces is able to mediate oligomerisation of EmbCCT in vitro . Previous studies had attributed to the C-terminal domain of the Emb proteins a critical role in arabinan chain extension [9] , [11] . Therefore , we asked whether the isolated domain is able to bind synthetic acceptor analogues . As the physiological substrate is chemically complex and diverse , using synthetic acceptor analogues offered the best chance to obtain an experimental acceptor-bound complex structure . In previous work , our laboratory had chemically synthesised neo-glycolipid acceptors that were modelled on motifs found in mycobacterial AG and LAM . When incubated with [14C]-labelled Araf-donor substrate DPA and isolated mycobacterial membranes in a cell-free Araf transferase , these molecules acted as potent acceptor mimics [25] . One of these acceptors was the di-arabinoside α-D-Araf- ( 1→5 ) -α-D-Araf-O- ( CH2 ) 7CH3 ( for short: Ara ( 1→5 ) Ara-O-C8 , Fig . 4B ) . The O-linked octyl tail allowed extraction of the reaction products for qualitative characterisation in vitro . Importantly , the closely related di-arabinoside α-D-Araf- ( 1→5 ) -α-D-Araf-O-CH3 ( Ara ( 1→5 ) Ara-O-C1 ) exhibited similar levels of acceptor activity , demonstrating the O-linked octyl was dispensable for activity [25] . By way of intrinsic tryptophan fluorescence , we probed binding of Ara ( 1→5 ) Ara-O-C8 to EmbCCT , as well as that of analogous tri- and penta-arabinofuranosides , [α-D-Araf- ( 1→5 ) ]2-α-D-Araf-O- ( CH2 ) 7CH3 ( Ara-α ( 1→5 ) 2-Ara-O-C8 ) and [α-D-Araf- ( 1→5 ) ]4-α-D-Araf-O- ( CH2 ) 7CH3 ( Ara-α ( 1→5 ) 4-Ara-O-C8 , Fig . 4B ) . Fitting the binding curves to a single-site saturation model , yielded an equilibrium dissociation constant Kd of 3 . 6 µM for the di-arabinofuranoside Ara ( 1→5 ) Ara-O-C8 ( Table 2 ) , while the disaccharide lacking the octyl chain , Ara ( 1→5 ) Ara-O-C1 , resulted in a Kd of 11 . 0 µM . These data confirmed that in the solution state the octyl chain is not essential for binding , although it may enhance affinity . Soaking EmbCCT crystals in cryoprotectant solution containing 27 mM Ara ( 1→5 ) Ara-O-C8 ( ∼3-fold excess of ligand relative to protein concentration in the crystal ) reproducibly resulted in defined ligand density ( Fig . 3B ) , allowing us to unequivocally build one Araf unit and the octyl chain of Ara ( 1→5 ) Ara-O-C8 , while the second Araf ring remained invisible , even when contouring the map at near-noise level . Soaking experiments using the other acceptor analogues , for which solution binding was examined , failed to reveal electron density for the ligand . The soaked di-arabinofuranoside ligand is positioned between two symmetry-related copies of EmbCCT , forming non-covalent contacts only with residues in subdomain I , but not with the CBM-like subdomain II , in contrast to our expectation . The Araf moiety packs against helix H6 and the H6-S13 loop ( Fig . 2 ) , forming three direct H-bond contacts with protein: O2 binds to carbonyl O of Trp985 ( 2 . 53 Å ) , O1 to Nε1 of Trp985 ( 2 . 99 Å ) , and O3 to Nδ2 of Asn740′ ( primed residues indicating the symmetry mate ) . In contrast , the octyl chain binds between helix H0 and the S13–S14 loop of the symmetry mate ( Fig . 3B ) . Ligand binding promotes ordering of the N-terminus of helix H0 , where 3 additional residues become visible compared to apo , and induces a conformational shift of aspartate residues 1051 and 1052 in the S13–S14 loop ( Fig . S6 ) . While this crystallographic complex structure did not reveal binding to the CBM-like subdomain II , it is possible that crystal lattice formation of EmbCCT interferes with binding at a site on subdomain II . We , therefore , asked whether the structural superimposition with saccharide-bound CBM domains could be exploited to predict potential additional binding sites . We note that ligand binding modes and substrate specificity of CBM domains can differ even within the same CBM family [21] , [26] . Thus , structural alignments of the protein scaffolds are unlikely to accurately predict the precise modes of binding and potential specificity-determining interactions . Nevertheless , superimposing carbohydrate-bound structures of CBM domains with the 10-highest DALI Z-scores ( with respect to the non-redundant PDB90 subset ) shows two clusters of putative ligand binding sites in subdomain II ( Fig . 3C ) : ( 1 ) near the Ca2+ site and the S3–S4 loop , and ( 2 ) on the open surface of the ‘outer’ β-sheet ( strands S2 , S4 , S10 , S6 , S7 ) . Virtually all ligands in the first cluster sterically clash with the loops that coordinate the Ca2+ site . Without invoking a conformational change that exposes the Ca2+ to solvent , this site appears unable to accommodate a ligand . In contrast , in the second cluster , only minor steric hindrance occurs between EmbCCT and the superimposed ligands , and thus this site appeared more plausible as a carbohydrate-binding site . The crystallographic complex of EmbCCT bound to Ara ( 1→5 ) Ara-O-C8 and the structural superposition with carbohydrate-bound homologues had indicated two distinct regions in EmbCCT as potential sites for carbohydrate binding ( Fig . S7A ) . In order to probe the relevance of these two sites , we asked whether replacement of endogenous EmbC with recombinant EmbC carrying appropriate point mutations would alter the cell wall composition of M . smegmatis . Aromatic residues frequently mediate binding of carbohydrate ligands to CBMs [21] . Given the H-bond contacts between Trp985 and Ara ( 1→5 ) Ara-O-C8 in subdomain I , and the central position of Trp868 of the ‘outer’ ( solvent-exposed ) β-sheet of subdomain II ( Fig . 3C and Fig . S7A ) , we probed these two residues in the first instance . Using a phage-mediated transduction method for allelic exchange [27] , we generated an EmbC-deficient strain of M . smegmatis ( M . smegmatis ΔembC ) , which was complemented with plasmids encoding either wild-type ( full length ) M . tuberculosis EmbC or mutant forms thereof . In accordance with previously reported data [9] , our M . smegmatis ΔembC strain retains lipomannan ( LM ) synthesis , but is deficient in LAM ( Fig . 4C – lane 2 ) . The abrogation of LAM biosynthesis can be directly attributed to the loss of EmbC , which is involved in the early synthesis of α ( 1→5 ) -Araf arabinan elongation of LM , the immediate LAM precursor ( Fig . 1A ) [9] . We utilised this phenotype by analysing LM/LAM resulting from complementation of M . smegmatis ΔembC with plasmid pVV16-Mt-embC , encoding full-length M . tuberculosis EmbC , and plasmids pVV16-Mt-embCW868A or pVV16-Mt-embCW985A , which encode point mutants W868A and W985A of full-length M . tuberculosis EmbC , respectively . Complementation with wild type EmbC largely restored the normal phenotype ( Fig . 4C – lane 3 ) , whereas complementation with the point mutants failed to re-establish LAM synthesis ( Fig . 4C – lanes 4 , 5 ) . We verified by Western blot that loss of LAM synthesis was not due to failure of the plasmid-encoded protein to incorporate into the membrane of M . smegmatis ΔembC ( Supporting Fig . S7B ) . These results suggest that the structural perturbations caused by the individual single-site mutations are sufficient to disrupt the function of EmbC . In order to establish whether loss of activity was linked to compromised acceptor binding , we introduced the single-residue mutations W868A or W985A into expression plasmids encoding EmbCCT . In addition , we prepared analogous expression plasmid constructs bearing mutations on Asn740 ( to Ala , binding site subdomain I ) , Gln899 ( to Ser ) and His911 ( to Ala , binding site subdomain II ) and Asp949 ( to Ser , Ca2+ binding site , see Supporting Fig . S7A ) . Two constructs ( Q899S , D949S ) did not express well enough to yield protein suitable for in vitro assays . For those proteins that were produced successfully , proper folding was verified by far-UV circular dichroism spectroscopy ( Supporting Fig . S7C ) . When comparing binding of the di- and penta-arabinoside acceptor analogues ( Fig . 4B and Fig . 5 ) that both carry the O-linked octyl tail , it was striking that the substitutions W868A and W985A affected binding of these ligands in a differential fashion . While the W985A mutation virtually abrogated binding of the disaccharide Ara ( 1→5 ) Ara-O-C8 , the W868A substitution preserved binding of this particular ligand , with only a modestly higher Kd ( Table 2 , Fig . 5A ) . In contrast , binding of the penta-arabinoside Ara ( 1→5 ) 4Ara-O-C8 was insensitive to the W985A mutation , but completely inhibited in response to the W868A mutation . Likewise , mutating Asn740 to Ala weakened binding of the disaccharide ( Table 2 ) , consistent with its position within H-bond distance of the ordered Araf in subdomain I , whereas the distant H911A mutation in subdomain II had no effect on this ligand . Thus , the differential effect of mutations in the putative binding sites in subdomain I and II on binding of acceptor analogues that differ only in length , strongly suggests that these bind preferentially to distinct sites on EmbCCT .
Polyprenyl-dependent glycosyltransferases of superfamily GT-C are still awaiting the determination of a structure of an intact , full-length enzyme , but structures of individual hydrophilic domains have begun to emerge [24] ( see also PDB entry 3BYW ) . As a first step towards the complete structural characterisation of the Emb Araf-transferases in M . tuberculosis , we have determined the crystal structure of the hydrophilic C-terminal domain of EmbC , the enzyme responsible for arabinan chain elongation in LAM synthesis and a target for the front line antibiotic EMB [5] . We found that the architecture of this domain comprises two subdomains , one of which folds as a lectin- or CBM-like domain , the other one shows weak similarity to the C-terminal hydrophilic domain of an unrelated GT-C glycosyltransferase , oligosaccharyl transferase STT3 [24] . The match between subdomain I and the so-called CC region of STT3 is poor ( Fig . S4 ) , and is limited to core secondary structure elements . Nevertheless , the DALI-derived superposition aligns the second Trp in STT3's highly conserved WWDYG motif with EmbC's Trp985 , a side chain we showed is critical for enzymatic activity . Thus the alignment lends additional support to the notion of Trp985 sitting at a critical junction of the C-terminal domain of EmbC . Sequence comparison of the Emb C-terminal domains ( Fig . S1 ) strongly suggests that the disulfide bond Cys749-Cys993 is a conserved structural feature . Forming a topologically intuitive demarcation of this domain , this covalent link presumably enhances the stability of the C-terminal domain at physiological conditions in the host . The disordered loops ( residues 794–825 , 1016–1037 ) encompass regions of high sequence diversity as opposed to otherwise remarkably conserved regions of the structure . Given the latter , one could speculate that these disordered regions are linked to acceptor discrimination , and/or that ordering might be induced by contacts with adjacent structural elements in the context of the full-length enzyme . It has previously been proposed that the Emb enzymes may function as dimers , possibly in the combination EmbA/EmbB and EmbC/EmbC [11] , [28] . Our sedimentation velocity data now provide supporting evidence for self-assembly of EmbC , although we cannot rule out that the observed oligomerisation occurs solely as a result of separating EmbCCT from the rest of the protein . However , the presence of dimers and trimers ( or tetramers ) ( Fig . 4A ) in solution demonstrated that at least two of the observed crystal packing interfaces were able to mediate self-assembly of EmbCCT . While thile the most-extended packing interface ( SAS buried 1100 Å2 ) is mediated by structural elements ( helices H0 and H6 ) that are close the truncation site , the second-largest interface ( SAS buried 670 Å2 ) is mediated by strand S2 , and distant to the truncation site . Indeed , the latter self-assembly interface generates a continuous β-sheet that extends across the monomer-monomer boundary ( Fig . S5C ) , hinting that it could be preserved in the full-length enzyme . The presence of a CBM-like subdomain in EmbCCT is consistent the proposed role of the C-terminal domain in acceptor substrate recognition [10] , [11] . Among these structurally diverse carbohydrate binding modules , the β-sandwich fold seen in EmbCCT is most common [21] . The differential response of the ligands of different length to the Trp mutations in subdomains I and II provides compelling evidence for the presence of two separate ligand binding sites in EmbCCT . This response also links the loss of Araf transferase activity in the Trp mutants to compromised acceptor binding . Although we were not successful in crystallising a complex structure that directly demonstrates binding of an acceptor analogue to the CBM-like subdomain II , the dramatic loss of binding affinity of the penta-arabinoside acceptor for the mutant EmbCCT ( W868A ) ( Fig . 5B , Table 2 ) and the corresponding loss of LAM synthesis , are strong indications that subdomain II indeed functions as a carbohydrate binding module . We note that the W868A mutation has also a modest effect on binding of Ara ( 1→5 ) Ara-O-C8 ( ∼2 . 5-fold increase in Kd , Table 2 ) , despite the obvious preference of this ligand for binding to subdomain I , as shown by the structure and the response to the W985A mutation . This observation could indicate that Ara ( 1→5 ) Ara-O-C8 also associate with the CBM-like subdomain II , albeit with considerably lower affinity . The converse may be true for the penta-saccharide as well , although the affinities we measured show no corresponding signature . Comparison of the affinities for binding of the tri- and pentasaccharide to wild type EmbCCT clearly indicates that binding to subdomain II is tighter for longer polysaccharides , as these can be expected to make additional contacts . However , the apparent switch in binding preference from the site in subdomain I to that in subdomain II on going from two to five Araf units is less straightforward to explain . If , as the structure suggests , only the octyl tail and the first Araf unit were the major determinants of binding to subdomain I , one would expect to see evidence for binding of Ara ( 1→5 ) 4Ara-O-C8 to subdomain I , that is , a significant change in affinity when mutating Trp985 . Thus , while the octyl tail clearly influences binding of the di-saccharide , this appears to be less the case for the tri- and penta-saccharides . This observation is in line with the dispensable nature of the octyl chain when the above ligands are used as acceptor mimics in cell-free Araf transferase assays [25] . Overall , a string of genetic and biochemical evidence consistently indicated that enzymatic activity of the Emb Araf-transferases is associated with loops displayed on the extra-cellular face of the membrane . For instance , the most frequent point mutation present in EMB-resistant clinical isolates of M . tuberculosis concerns residue Met306 in EmbB ( = Met300 in EmbC , see Fig . 1 ) [20] , only a few residues downstream of the GT-C-specific , strictly conserved DDX motif in the E2 loop [15] . Berg et al . showed that loop E6 carries a functionally relevant , conserved proline-containing sequence motif [10] , consistent with findings in the Emb protein of C . glutamicum [14] . Moreover , a crystal structure of the first extracellular loop of the Emb Araf-transferase of the related organism Corynebacterium diphtheriae has become available very recently ( PDB entry 3BYW; Tan K . , Hatzos C . , Abdullah J . , Joachimiak A . , unpublished ) . The domain of the E1 loop displays a β-sandwich fold with similarity to the fold of galectin [29] , but is not superimposable on that of subdomain II of EmbCCT . The galectin-like fold again hints to a potential function in carbohydrate binding – perhaps the sugar moiety of the Araf-donor DPA . In conclusion , the present structure of the C-terminal domain of M . tuberculosis EmbC provides a first corner stone towards assembling the structure of the full-length enzyme , and allows us to begin probing this essential enzyme in a rational and targeted fashion .
Plasmids were propagated during cloning in E . coli Top10 cells ( Invitrogen ) . All restriction enzymes , T4 DNA ligase and Phusion DNA polymerase enzymes were sourced from New England Biolabs . Oligonucleotides were from MWG Biotech Ltd and PCR fragments were purified using the QIAquick gel extraction kit ( Qiagen ) . Plasmid DNA was purified using the QIAprep purification kit ( Qiagen ) . A 1125-bp region coding for the C-terminal domain ( residues 719–1094 ) of EmbC was cloned from genomic DNA of M . tuberculosis H37Rv using PCR primers ( restriction sites underlined ) GATCGATCCATATGGAGGTGGTATCGCTGACCCAG ( forward ) and GATCGATCCTCGAGCTAGCCTCTGCGCAACGGC ( reverse ) . The PCR product was ligated into plasmid pET23b ( NdeI , XhoI restriction sites ) , yielding the His6-tagged pET23b-EmbCCT construct , whose sequence was verified ( School of Biosciences Genomics Facility , University of Birmingham ) . For expression , E . coli C41 ( DE3 ) cells were transformed with pET23b-EmbCCT using the rubidium chloride method . Overnight cultures ( 5 ml LB medium , 100 µg/ml ampicillin ) were used to inoculate bulk cultures ( 4×1 litre LB , 100 µg/ml ampicillin , 37°C , 200 rpm ) . Seleno-methionine derivatised EmbCCT was produced using the same expression plasmid and host , but following the feedback inhibition protocol described in [30] . Cultures were induced at OD600 = 0 . 5 using 1 mM IPTG ( 12 h , 16°C ) . Cells were harvested ( 6000×g , 15 min ) , washed with 20 ml phosphate buffered saline , and frozen . Pellets were re-suspended in 50 mM KH2PO4 ( pH 7 . 9 ) , 300 mM NaCl , 1 mM PMSF , 15 µg/ml benzamidine , DNAse and RNAse ( 50 µg/ml ) , and sonicated ( 30 sec ON/OFF cycles , total of 8 cycles ) . The lysate was cleared ( 30 min , 28000×g , 4°C ) and passed over a HiTRAP Ni2+-NTA column ( GE Healthcare ) , equilibrated in 50 mM KH2PO4 ( pH 7 . 9 ) , 300 mM NaCl , and eluted using a step-gradient of 50–500 mM imidazole . The purification was monitored by 12% SDS-PAGE . Fractions containing EmbCCT ( 250 , 500 mM imidazole ) were pooled and dialysed against 50 mM KH2PO4 ( pH 7 . 9 ) , 300 mM NaCl , and concentrated by ultrafiltration to ∼15 mg/ml . Hanging drop vapour diffusion was used to grow crystals of EmbCCT over a reservoir of 0 . 1 M sodium acetate pH 4 . 4 , 80 mM ammonium phosphate , mixing 1 µl of protein with 1 µl of reservoir solution . Crystals were cryoprotected in reservoir solution , adding up to 12% ethylene glycol and 12% glycerol , and flash frozen in liquid nitrogen . Native and 3-wavelength SeMet MAD data were recorded on beamline ID23-1 ( ESRF , Grenoble , France ) . Diffraction images were processed using XDS and XSCALE [31] ( Table 1 ) . Selenium sites and phases were obtained using standard procedures ( SHELXD [32] , SHARP v2 . 2 [33] SOLOMON [34] ) leading to a readily interpretable electron density map ( Fig . S2A ) . The ARP/wARP-built [35] initial model was rebuilt in COOT [36] , with intermittent refinement against native data ( REFMAC5 [37] , PHENIX . REFINE [38] ) . Temperature factor modelling included TLS refinement [39] . The final model has good stereochemistry and comprises EmbC residues 735–794 , 825–1015 and 1038–1067 , 113 water molecules , one molecule of Ara ( 1→5 ) Ara-O-C8 , one Ca2+ and one phosphate ion ( Table 1 ) . Intrinsic tryptophan fluorescence ( ITF ) experiments were carried out using a PTI QuantaMaster 40 spectrofluorimeter , recording data with the FeliX32 software package ( PTI , Birmingham , New Jersey , USA ) . The excitation wavelength was set to 294 nm and the fluorescence emission ( Femission ) was recorded between 300–400 nm for each ligand aliquot added to a 200 µl solution containing 20 µM EmbCCT in 50 mM KH2PO4 ( pH 7 . 9 ) , 300 mM NaCl . For EmbCCT , the emission maximum ( Femissionmax ) was at λ = 338 nm , providing a basal Femission coordinate for the collection of subsequent ITF data . The change in fluorescence emission ( ΔFemission ) was calculated by subtracting Femission ( recorded 2 min after each ligand addition ) from Femissionmax , and the data was then plotted against ligand concentration , [L] ( 3 independent experiments ) . A plot of ΔFemission at λ = 338 nm vs . [L] was fitted to the saturation binding equation using GraphPad Prism software: Far-UV circular dichroism ( CD ) spectra were recorded at 25°C using a Jasco J-715 spectropolarimeter and a cell of 0 . 01 cm path length . Proteins EmbCCT , EmbCCT ( N740A ) , EmbCCT ( W868A ) , EmbCCT ( H911A ) and EmbCCT ( W985A ) were dialysed into 50 mM KH2PO4 ( pH 7 . 9 ) , 50 mM NaF to a final concentration of 0 . 5 mg/ml each . Spectra were recorded of 250 µl aliquots of each protein by measuring ellipticity from 195–260 nm , using a bandwidth of 2 nm and a scan speed of 100 nm/min . Spectra were normalised by subtracting the spectrum of buffer alone ( baseline ) . Sedimentation velocity experiments were performed using a Beckman Proteome XL-I analytical ultracentrifuge equipped with absorbance optics . EmbCCT was dialysed into 50 mM KH2PO4 ( pH 7 . 9 ) , 300 mM NaCl , and loaded into cells with two channel Epon centre pieces and quartz windows . A total of 100 absorbance scans ( 280 nm ) were recorded ( 40 , 000 rpm , 4°C ) for each sample , representing the full extent of sedimentation of the sample . Data analysis was performed using the SEDFIT software , fitting a single friction coefficient [40] . Approximately 1 kb of upstream and downstream flanking sequences of the embC gene ( MSMEG2785 ) were PCR amplified from M . smegmatis mc2155 genomic DNA using the primer pairs MSEMBCLL , MSEMBCLR , MSEMBCRL and MSEMBCRR , respectively ( sequences listed in Supporting Information Table S1 ) . Following restriction digestion of the primer incorporated AlwNI sites , the PCR fragments were cloned into AlwNI-digested p0004S to yield the knockout plasmid pΔMSMEGEMBC which was then packaged into the temperature sensitive mycobacteriophage phAE159 as described previously [27] to yield phasmid DNA of the knockout phage phΔMSMEGEMBC . Generation of high titre phage particles and specialized transduction were performed as described earlier [27] , [41] . Deletion of MSMEGEMBC in one hygromycin-resistant transductant was confirmed by Southern blot . For complementation , M . tuberculosis embC was cloned using primer pairs Mt-embC-forward and Mt-embC-reverse ( sequences listed in Supporting Information Table S1 ) and blunt-end ligated into SmaI digested pUC18 . For QuikChange mutagenesis ( Stratagene ) of pUC18-Mt-embC W868A and W985A codons , primer pairs W868A-sense/-antisense and W985A-sense/-antisense ( sequences in Supporting Information Table S1 , each with 5′-phosphate modifications ) were used . The 3301 bp product was extracted from plasmids ( pUC18-Mt-embC , pUC18-Mt-embCW868A and pUC18-Mt-embCW985A ) digested with NdeI and HindIII , and sub-cloned into the similarly digested mycobacterial shuttle vector pVV16 to yield pVV16-Mt-embC , pVV16-Mt-embCW868A and pVV16-Mt-embCW985A . These plasmids were then used to transform M . smegmatisΔembC to yield clones resistant to both hygromycin and kanamycin . QuikChange mutagenesis ( Stratagene ) was carried out using pET23b-Mt-embCCT ( generated as described above ) . Primer pairs used for the codon alterations N740A , W868A , Q899S , H911A and W985A are listed in the Supporting Information Table S1 . Mutant plasmids were subsequently transformed individually into E . coli C41 ( DE3 ) . Mutant proteins were expressed and purified as described above . Lipoglycans form M . smegmatis strains were extracted as described previously [42] . Dried cells were resuspended in de-ionized water and disrupted by sonication ( MSE Soniprep 150 , 12 µm amplitude , 60 s on , 90 s off for 10 cycles , at 4°C ) . An equal volume of ethanol was added to the cell suspension and the mixture was refluxed at 68°C , for 12 h intervals , followed by centrifugation and recovery of the supernatant . The C2H5OH/H2O extraction process was repeated five times and the combined supernatants dried . The dried supernatant was then subjected to hot-phenol treatment by addition of phenol/H2O ( 80% , w/w ) at 70°C for 1 h , followed by centrifugation and the aqueous phase was dialyzed using a 1500 MWCO membrane ( Spectrapore ) against de-ionized water . The retentate was dried , resuspended in water and sequentially digested with α-amylase , DNase , RNase , chymotrypsin and trypsin . The retentate was further dialyzed using a 1500 MWCO membrane ( Spectrapore ) against deionized water . The eluates were collected , extensively dialysed against deionized water , concentrated and analyzed by 15% SDS-PAGE using a Pro-Q emerald glycoprotein stain ( Invitrogen ) . The accession number for the coordinates and structure factors of the C-terminal domain of EmbC in the Protein Data Bank ( http://www . rcsb . org ) is 3PTY .
|
Tuberculosis ( TB ) , an infectious disease caused by the bacillus Mycobacterium tuberculosis , burdens large swaths of the world population . Treatment of active TB typically requires administration of an antibiotic cocktail over several months that includes the drug ethambutol . This front line compound inhibits a set of arabinosyltransferase enzymes , called EmbA , EmbB and EmbC , which are critical for the synthesis of arabinan , a vital polysaccharide in the pathogen's unique cell envelope . How precisely ethambutol inhibits arabinosyltransferase activity is not clear , in part because structural information of its pharmacological targets has been elusive . Here , we report the high-resolution structure of the C-terminal domain of the ethambutol-target EmbC , a 390-amino acid fragment responsible for acceptor substrate recognition . Combining the X-ray crystallographic analysis with structural comparisons , site-directed mutagenesis , activity and ligand binding assays , we identified two regions in the C-terminal domain of EmbC that are capable of binding acceptor substrate mimics and are critical for activity of the full-length enzyme . Our results begin to define structure-function relationships in a family of structurally uncharacterised membrane-embedded glycosyltransferases , which are an important target for tuberculosis therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biochemistry",
"infectious",
"diseases/antimicrobials",
"and",
"drug",
"resistance",
"microbiology"
] |
2011
|
The C-Terminal Domain of the Arabinosyltransferase Mycobacterium tuberculosis EmbC Is a Lectin-Like Carbohydrate Binding Module
|
The stable infection of host macrophages by Mycobacterium tuberculosis ( Mtb ) involves , and depends on , the attenuation of the diverse microbicidal responses mounted by the host cell . This is primarily achieved through targeted perturbations of the host cellular signaling machinery . Therefore , in view of the dependency of the pathogen on host molecules for its intracellular survival , we wanted to test whether targeting such factors could provide an alternate route for the therapeutic management of tuberculosis . To first identify components of the host signaling machinery that regulate intracellular survival of Mtb , we performed an siRNA screen against all known kinases and phosphatases in murine macrophages infected with the virulent strain , H37Rv . Several validated targets could be identified by this method where silencing led either to a significant decrease , or enhancement in the intracellular mycobacterial load . To further resolve the functional relevance of these targets , we also screened against these identified targets in cells infected with different strains of multiple drug-resistant mycobacteria which differed in terms of their intracellular growth properties . The results obtained subsequently allowed us to filter the core set of host regulatory molecules that functioned independently of the phenotypic variations exhibited by the pathogen . Then , using a combination of both in vitro and in vivo experimentation , we could demonstrate that at least some of these host factors provide attractive targets for anti-TB drug development . These results provide a “proof-of-concept” demonstration that targeting host factors subverted by intracellular Mtb provides an attractive and feasible strategy for the development of anti-tuberculosis drugs . Importantly , our findings also emphasize the advantage of such an approach by establishing its equal applicability to infections with Mtb strains exhibiting a range of phenotypic diversifications , including multiple drug-resistance . Thus the host factors identified here may potentially be exploited for the development of anti-tuberculosis drugs .
Successful parasitization of macrophages by Mycobacterium tuberculosis ( Mtb ) reflects the equilibrium between host and pathogen , which is established and maintained through the modulation of macrophage-signaling mechanisms . This leads to the attenuation of several cellular processes that include fusion of phagosomes with lysosomes , antigen presentation , apoptosis , and the bactericidal responses initiated by the macrophage [1] , [2] , [3] . This attenuation represents the outcome of a dynamic process wherein bacterial molecules interfere with the signaling machinery of the host cell . Although a detailed picture is yet unavailable , many bacterial mediators of virulence have been identified and the strategies employed by them are currently being elucidated [4] , [5] , [6] , [7] , [8] , [9] . The emerging theme , however , suggests that pathogen-directed manipulation of host processes is achieved through targeted perturbations in the host cell-signaling network [10] , [11] , [12] , [13] , [14] . It is these perturbations that then reorient the cellular response to support intracellular survival and growth of Mtb . A better understanding of the mechanisms involved in these perturbations should , therefore , greatly aid current efforts at developing new drugs for tuberculosis ( TB ) . We undertook this study to identify components of the host cell signaling machinery that are important for regulating an Mtb infection . For this , we performed an siRNA screen against all kinases and phosphatases , in mouse macrophages infected with a virulent strain of Mtb ( H37Rv ) . Such experiments identified several host molecules that were involved either in facilitating , or suppressing , the intracellular infection . By then applying a filter of phenotypic variations , at the level of both drug resistance and intracellular growth properties , we could further distinguish those host molecules that regulated intracellular pathogen load in an Mtb strain-independent manner . A combination of in vitro and in vivo experimental approaches subsequently enabled us to establish that targeting such host factors indeed provides an attractive alternate strategy for the development of anti-TB drugs . Importantly , our results suggest that such an approach could also potentially address the problem of multiple drug resistance in TB infections .
The siRNA library employed in the primary screen consisted of a pool of two siRNAs per target gene , and the targets included 744 kinases and 288 phosphatases . Cells of the murine macrophage line , J774 . 1 , were first infected with a virulent strain of Mtb ( H37Rv ) , and then transfected with individual target-specific siRNA . The siRNA transfection was performed after infection of cells to ensure an un-hindered uptake of pathogen and , therefore , to select for only those host proteins that were involved in the maintenance of an established infection . The effect of specific silencing on intracellular mycobacterial load was then determined from cell lysates in terms of the colony forming units ( CFU ) subsequently obtained ( Methods ) . At one level , silencing of several macrophage proteins resulted in a significant decrease in titers of the intracellular bacteria obtained ( Table S1 ) . However , there were also many instances where a targeted protein-knockdown led to a marked increase in mycobacterial levels ( Table S1 ) . Selection of only those effects as significant , where intracellular bacterial load varied by greater than two standard deviations ( 2SD ) of the mean CFU values obtained for the control wells ( Methods ) , identified 203 target-specific siRNAs ( Table S1 ) . To next filter out any non-specificity arising from ‘off-target’ effects of the siRNA , we performed a second screen where the primary ‘hits’ were now silenced with an alternate pool of siRNAs ( Methods ) . Although such an approach likely leads to an increased number of false negatives due to differences in silencing efficiency between the two target-specific siRNA pools , a high degree of confidence is – nonetheless - associated with any reproducible effects that are consequently obtained . This procedure led to a further reduction in the number of significant targets to forty one ( Table S1 ) . Then , as the final validation step , we tested whether these target-specific siRNA pools had any effect on the viability of either un-infected , or infected , J774 . 1 cells over the time course of our experiment ( Text S1 ) . In neither of these cases , however , did any of the siRNA pools negatively influence cell viability to an extent that was greater than 1 . 5 SD of the mean values obtained for the control wells ( Table S2 ) . Therefore , these forty one targets identified by the corresponding set of siRNA pools were taken as the validated target group . The effect of silencing of each of these on the intracellular bacterial load is shown in Figure 1A , whereas Figure 1B confirms that the siRNA pools used in the primary and in the validation exercise both yielded effects that were greater than 2SD of the mean CFU value obtained for the corresponding control wells . As is evident , the effects of target-suppression ranged from a near complete elimination of the infection , to a significant enhancement in intracellular bacterial titers ( Fig . 1A ) . Thus , adaptation of mycobacteria in the intracellular milieu likely involves both facilitating and inhibitory contributions from some of the components of the host cell signaling machinery . The validated targets are listed in Table 1 , which also includes a gene ontology-based classification of their functional roles . A more detailed description of these genes and their known association with disease and other cellular functions is provided in Table S3 . We next took four representative examples from each of the two distinct groups in Figure 1 , and examined – by confocal microscopy - whether the observed modulation in CFU values also correlated with alterations in the level of mycobacteria localization , within the lysosome of the host cell . Interestingly , in infected cells treated with non-silencing siRNA , the co-localization of H37Rv with acidified lysosomes could be detected – with a mean overlap coefficient of 0 . 20 - in only a little over 50% of the cells , ( Fig . 2 ) . In contrast , those instances where silencing yielded a reduced CFU count in Figure 1 ( Csnk1d , Adrbk1 , Prkacb , and TgfβrI ) , showed a marked increase in mycobacterial co-localization with acidified lysozomes , both at the level of individual cells , as well as at that of the proportion of cells showing such co-localization ( Fig . 2 ) . On the other hand , values for both of these parameters were significantly reduced for the cases ( Wee1 , Abl1 , Dgkz , and Chek1 ) where target gene silencing resulted in an increase in the CFU counts ( Fig . 2 , compare with Fig . 1 ) . Thus , at least for the representative cases shown in Figure 2 , the observed alterations in the CFU counts constitute a true reflection of shifts in the fate of the intracellular bacteria . Finally , an examination of the gene expression profiles both in un-infected J774 . 1 cells , and in cells infected with H37Rv either for 16h , 48h , or 96h , confirmed that the genes coding for all of the proteins listed in Figure 1 were indeed expressed in uninfected cells , albeit to different levels . Further , expression levels of thirty-four of these remained largely unaffected following infection of the cells ( Table S4 ) . We next asked whether the proteins identified against H37Rv in Figure 1 would also be relevant for infections with field isolates of Mtb , particularly for those exhibiting multiple drug resistance ( MDR ) . To address this we took two independent clinical isolates of MDR-Mtb , each displaying a distinct drug-sensitivity profile ( Fig . 3A ) . The relative growth rates of these two isolates with , that of H37Rv , was first compared both in extracellular cultures , as well as in infected cells . Interestingly , while all the three isolates displayed similar growth properties under extracellular conditions ( Fig . 3B ) these , however , differed markedly when measured within infected J774 . 1 cells ( Fig . 3C ) . In this latter experiment , a progressive increase in CFU counts was obtained for H37Rv over the time course studied whereas the MDR strain 1934 displayed an accelerated growth rate between the 48h and the 90h time points . In stark contrast , however , virtually no increase in bacterial titers could be observed for the other MDR strain , JAL2261 , at any of the time points studied . A reduced intracellular growth rate for JAL2261 , in comparison with the other two strains , was also similarly observed in primary murine macrophages ( Fig . S1 ) . Thus the three Mtb strains studied exhibit diverse growth properties within J774 . 1 cells . We next infected J774 . 1 cells with each of the MDR-Mtb strains and tested for the effects of treatment with the siRNAs described in Figure 1 . The results obtained are shown in Figure 4 . While the overall profiles obtained for both 1934 and JAL2261 show distinctions from that for H37Rv , some overlaps were – nonetheless – clearly evident . This was particularly true in the case of 1934 where , of the 41 targets tested , comparable effects of target silencing ( i . e . within a 30% deviation ) on both 1934 and H37Rv were observed in 31 cases ( Fig . 4 ) . The majority of these instances ( 25 of 31 ) represented the group wherein siRNA treatment yielded a reduction in the intracellular bacterial load , whereas a relatively poorer correspondence was obtained for the group where increased bacterial titers were observed ( 6 out of 11; Fig . 4 ) . Discrepant results were obtained for Trpm7 and Prkag3 although the cause for this is presently unclear . In contrast to 1934 , a comparison between the results obtained for JAL2261 and H37Rv revealed a significantly reduced overlap , with the similarity in effects being restricted to only 11 instances of siRNA treatment ( Fig . 4 ) . Further , all of these cases derived exclusively from the subset involving a consequent reduction in the level of the intracellular mycobacteria ( Fig . 4 ) . This reduced sensitivity of JAL2261 , to perturbations in host protein levels , is probably consistent with the fact that it exists in a non-dividing , or slow-dividing , state within the host cells . Thus the results in Figure 4 identify 11 host targets , whose depletion uniformly led to a reduction in intracellular load of all the three Mtb strains tested . The identification of host factors commonly implicated in regulating intracellular levels of all the three Mtb strains tested raised the possibility that at least some of these may serve as targets for the development of drugs against both drug-sensitive and drug-resistant forms of Mtb infection . At least at the level of ‘proof of principle’ , employing existing chemical inhibitors of any of these targets can readily test such a possibility . Therefore , for the present purposes , we chose to examine the effects of inhibition of TGF-β type-1 receptor ( TGFβRI ) . Our choice of this target was based on the results of our microarray experiments which indicated that transcript levels of both this receptor , as well as that of its isoform TGF-β type-2 receptor ( TGFβRII ) were significantly increased in cells infected with Mtb for 16h , although these levels then returned to their basal values at the later times ( Fig . 5A ) . Further , expression of the genes for the cognate ligands , TGFβI and TGFβII were also induced ( Fig . 5A ) . Here , although the extent of induction of TGFβI was not significant ( i . e . below the cut-off threshold ) this , nonetheless , translated into a significant increase in the levels of this cytokine in the culture supernatants ( Fig . 5B ) . It was , therefore , logical to infer that secretion of TGFβ by infected cells would lead to the corresponding activation of its receptor , either through autocrine or paracrine mechanisms . Further , at least based on the results of our siRNA screen , it seemed likely that this activation was somehow relevant for maintenance of the intracellular pathogen . To experimentally verify the inferred role for TGFβ-dependent activation of its receptor , we added neutralizing anti-TGFβ antibodies to cultures of H37Rv-infected J774 cells and then determined its consequent effect in terms of the mycobacterial CFUs obtained . Figure 5C shows that addition of anti-TGFβ antibodies resulted in a substantial reduction in intracellular Mtb load , thus confirming the relevance of this cytokine in regulating intracellular Mtb . To further establish this , we also performed experiments in mice that were transgenic for the dominant negative form of the TGFβ receptor type II ( TGFβRIIDN ) . Although our screen identified TGFβ receptor type-I as the hit , signaling through TGFβ1 involves the formation of an active heteromeric complex between the type-I and type-II receptors . Thus , binding of TGFβ to the type-II receptor leads to the recruitment and phosphorylation of the type-I receptor , with the subsequent activation of downstream pathways [15] . Consequently , TGFβ-dependent signaling would be similarly affected by either silencing the type-1 receptor , or over–expressing the dominant negative form of the type-II receptor . As shown in Figure 5D , TGFβRIIDN mice were significantly more resistant to an aerosol route of infection with H37Rv , than their wild type counterparts . This was consistent with the fact that macrophages isolated from the transgenic mice indeed showed over-expression of TGFβRIIDN ( Fig . 5D ) . Collectively then , the results in Figure 5 provide strong experimental support for the likelihood that TGFβ-dependent activation of its receptor on macrophages is critical for the survival of intracellular Mtb . To inhibit TGFβRI activation , we employed the compound 4-[4- ( 2 , 3-Dihydro-1 , 4-benzodioxin-6-yl ) -5- ( 2-pyridinyl ) -1H-imidiazol-2-yl]benzene ( D4476 ) . Our choice of this inhibitor was guided by the fact that , in addition to TGFβRI , this compound is also known to inhibit casein kinase 1 ( CSNK1 ) [16] , [17] , another member of our validated target list in Figure 4 . Interestingly , CSNK1 also represents a downstream intermediate in the TGFβR signaling pathway , and has been shown to regulate both ligand-independent ( i . e . basal ) , as well as ligand-induced signaling processes [18] . Thus , it was anticipated that the simultaneous inhibition of TGFβRI and one of its downstream signaling intermediates would yield a more effective inhibition of TGFβ-dependent signaling and , therefore , a more potent effect on intracellular Mtb levels . Cells infected either with H37Rv , 1934 , or JAL2261 were treated with increasing doses of D4476 , and the consequent effect on intracellular pathogen load was then determined as the resulting CFU values obtained ( Fig . 6A ) . A dose-dependent decrease in mycobacterial CFUs , with increasing inhibitor concentrations , was clearly obtained for all the three isolates tested ( Fig . 6A ) . Importantly , this effect was specific for the mycobacteria and this drug displayed no toxicity towards the host cell ( Fig . 6B ) . In parallel studies employing confocal microscopy we observed that treatment with the inhibitor also led to a corresponding increase in localization of each of the mycobacterial isolates within acidified lysosomes of the cell ( Fig . 6C ) . This further supports that treatment with the inhibitor leads to an enhanced clearance of the intracellular pathogen . Importantly , the effect of D4476 addition was specific for intracellular mycobacteria , with no significant effect on mycobacterial growth in extracellular cultures ( Fig . 6D ) . This confirms that the inhibitor did not directly target Mtb . Rather it more likely interfered with the intracellular survival mechanisms of the pathogen . Importantly , from the standpoint of pharmacological intervention , that simultaneous inhibition of TGFβRI and its downstream signaling intermediate leads to a more potent effect on intracellular Mtb could be demonstrated by comparing the relative efficacy of D4476 with that of inhibitors specific for only either TGFβRI ( LY364947 [19] ) or CSNK1 ( IC261 , [17] ) ( Fig . 6E ) . The inhibitory effect of D4476 on intracellular Mtb survival could also be demonstrated in primary murine macrophages infected with each of the three Mtb strains . Thus , consistent with the findings in J774 cells , addition of D4476 to infected primary macrophages resulted in enhanced co-localization of all three mycobacterial strains with the lysosome , and also a significant reduction in CFU count in each of these cases ( Figure S1 ) . Thus the cumulative results in Figure 6 confirm a key role for TGFβR signaling in regulating intracellular Mtb survival , but in a manner that is independent of at least the spectrum of phenotypic variations encompassed by this group of isolates . To further validate the relevance of targeting host factors , we also examined the efficacy of D4476 in the murine model of TB infection . For this , BALB/c mice were intravenously infected with H37Rv and these mice were then treated with two different concentrations of D4476 . Treatment was initiated at ten days after the infection , and included six administrations of the relevant dose of the inhibitor ( Text S1 ) . As shown in Figure 7A , a clear reduction in CFU counts was obtained from the lungs of infected mice treated with D4476 . Further , the magnitude of this inhibition was sensitive to the dose of the compound with a dose of 4 nmol/g of body weight yielding a nearly 80% reduction in CFU counts . Similar results were also obtained in mice infected through the aerosol route ( Fig . 7B ) , confirming that the efficacy of D4476 was independent of the route of infection employed . D4476-dependent clearance of infection was also evidenced through recovery from the infection-induced splenomegaly ( Fig . 7C ) . Further , histochemical staining of lung sections also revealed a significant reduction in the number of epitheloid cell granulomas , and in that of acid-fast staining bacilli , in treated versus the untreated mice ( Figure 7D ) . The collective results in Figure 7 , therefore , demonstrate the in vivo efficacy of D4476 in terms of eliminating an Mtb infection . While its potency may be considered to be low we emphasize , however , that our objective here was not to propose this compound as a drug against TB . Rather , these and the preceding experiments were merely intended to reveal and highlight the feasibility of targeting relevant host factors , as an alternate strategy for the chemotherapy of TB .
Recent years have witnessed a resurgence of efforts directed at tuberculosis ( TB ) drug research and development . This has been spurred by the increasing incidence of drug resistance in the infected population , engendering an urgent need for the development of improved regimens for TB treatment . Although drug resistant - particularly multi- and extremely-drug resistant - TB poses a daunting challenge , the situation is further complicated by the predominance of latent and/or persistent forms of TB infection in the population [20] . These latter infections are characterized by phenotypic resistance - or tolerance - to drugs , despite being genotypically sensitive to them [20] . Indeed , latent/persistent infection constitutes over ninety nine percent of the TB infected population worldwide [20] . Conventional approaches to drug research have targeted unique processes or enzymes of the pathogen . However , despite its many successes , this approach suffers from the risk of generating newer variants exhibiting drug resistance [21] . Further , this strategy is also limited by its inability to address infection states where the pathogen is metabolically less active such as in persistent or latent infections . This issue is especially relevant in the case of Mtb infections . An alternate paradigm for drug discovery has recently emerged , at least for intracellular pathogens [22] . This is based on the fact that the survival of an intracellular pathogen in the hostile cellular environment requires it to exploit and subvert various host factors . Therefore , identifying - and then targeting – such host factors should provide an additional route for the therapeutic management of these infectious diseases . Here , the anticipations also are that such a strategy will be less likely to induce microbial resistance [22] . To explore the possibilities offered by this avenue , siRNA-based screens are now being widely employed to identify such host factors for a range of viral , bacterial , and parasitic infections [23] , [24] , [25] , [26] , [27] , [28] , [29] . In this connection , a recent screen of the human kinome identified a kinase cluster around protein kinase B ( PKB ) as being obligatory for the intracellular survival of Salmonella typhimurium [30] . Interestingly , while inhibitors of PKB were effective against this pathogen , they were also capable of partially inhibiting infection of human macrophages with a strain of MDR-Mtb . We , therefore , undertook the present study to further explore the potential of such an approach in the specific context of Mtb . Our siRNA screen successfully identified several host factors involved in the regulation of H37Rv survival within the milieu of the host macrophage . Interestingly , this group of target proteins included those that facilitated , as well as those that impeded Mtb survival . While a mechanistic resolution of the biochemical pathways involved is clearly needed , this identification of molecular components with opposing functional roles strongly supports the existence of a host-specified axis that regulates the fate of intracellular Mtb . Such an axis probably reflects the level of equilibrium achieved between the intracellular processes initiated to eliminate the infection , and the pathogen-mediated manipulation of the host machinery in its own favor [31] . In this connection , it is pertinent to note that previous studies have identified several host proteins that play an important role during the course of an Mtb infection . However , with the exception of a few examples [32] , the majority of these proteins are involved in processes that occur either during , or soon after , the endocytic uptake of Mtb . Thus , for example , the role of the calcium signaling pathway- involving PI3K , PKB , CaMKII and Coronin 1 – in regulating endocytosis and subsequent fusion of phagosomes with lysosomes has been well studied [1] , [12] . Similarly , there is also information available on the biochemical pathways that mediate microbicidal responses of the infected macrophage [33] . In contrast to these early responses , however , our screen was designed to specifically capture those host molecular components that are involved in the later time window of the infection process . That is , in the window where the pathogen is either in the process of establishing , or has established , a dynamic equilibrium with the intracellular machinery of the host cell . Consequently , our present study examined a relatively less explored facet of macrophage infection by Mtb and , therefore , identified several novel host molecules whose role in regulating intracellular Mtb has not been hitherto suspected . From the standpoint of targeting host factors as a drug development strategy for TB , we believe that it is such factors that regulate the maintenance of an established infection that would be more relevant . An important aspect of our studies was the additional filtration , of the identified H37Rv-specific host factors , against MDR-Mtb variants that not only exhibited altered drug-sensitivity profiles , but also altered properties of intracellular growth . Interestingly , a significant proportion of these factors could be validated against the other rapidly growing strain , 1934 . In contrast , only a small subset of these targets retained efficacy against JAL2261 , a strain whose levels did not significantly increase within infected cells . These differences are consistent with at least a priori expectations that the extent of cross-regulatory interactions between host and pathogen would correlate directly with state of replication activity of the pathogen . The net outcome of our experiments was the identification of a core list of host targets that were involved in regulating survival of Mtb , independent of variations in either drug sensitivity profiles or growth properties in the host cell . Further , we could also demonstrate , at the level of ‘proof of principle’ , that at least some of these host factors may provide novel targets for the development of anti-TB drugs . This was exemplified by our findings that the simultaneous inhibition of TGFβRI and CSNK1 substantially inhibited intracellular survival of both drug-sensitive and multiple drug-resistant strains of MTB . In addition to J774 . 1 cells , this effect was also observed in primary murine macrophages . More importantly though , our experiments involving D4476 administration to H37Rv-infected mice also provided in vivo validation for the possibility of targeting host factors as a possible approach for TB therapy . Here , it will be important to test the potential applicability of the remaining host factors identified by our screen in this regard . Further , the mechanism by which inhibition of TGFβRI and CSNK1 induces elimination of the infection is also of interest . Thus , in summary , our present results highlight the existence of host factors that regulate the intracellular survival of Mtb in a manner that is insensitive to variations in either the drug-sensitivity profile , or the intracellular growth properties . In addition , we also provide proof-of-concept demonstration that targeting at least some of these molecules can provide an alternate approach for the chemotherapy of TB . Again as demonstrated here , a highlight of such a strategy would be that it holds the promise of eventually being able to develop suitable drugs that function in a manner that is independent of the phenotypic and genotypic diversification exhibited by Mtb in the field [20] . However , further validation in the context of human infections will be necessary before such a promise can be realized . Further , the issue of possible toxicity to the host cell - as a result of inhibition of key host molecules - may also require to be addressed .
All animal experiments were carried out in accordance with guidelines approved and created by the ICGEB animal ethics committee . Female BALB/c mice 4–6 wk of age kept in pathogen free environment . TGFβRDN ( TGFβ receptor dominant negative , BALB/C background ) mice , 6 to 8 weeks of age , were purchased from Jackson Laboratory , Bar Harbor , Maine . These mice were maintained and breed in a specific-pathogen-free biosafety level-3 facility . Murine macrophage cell line J774 . 1 ( American Type Culture Collection ) was used in this study . J774 . 1 cells were cultured in RPMI 1640 ( Gibco Laboratories ) supplemented with 10% FCS ( Hyclone ) and were maintained between 2 and 10×105 cells per mL at 37°C in a humidified , 5% CO2 atmosphere . Before infection cells were plated in 96 well plates at 2 . 5×104 cells per well overnight . Mouse Phosphatase siRNA set V 1 . 0 and Mouse kinase siRNA set V 1 . 0 library from Qiagen ( two siRNAs/target ) were used for the study . For validation experiments siRNAs against the kinases were obtained from Sigma siRNA ( MISSION siRNA Mouse Kinase Panel , three siRNAs/target ) , whereas the phosphatase-specific siRNAs were procured from Dharmacon ( SMARTpool ) ( four siRNAs/target ) . J774 . 1 cells ( 25 , 000 cells/well in 96-well plates ) were infected with mycobacteria at an MOI ∼10 ( 10 bacteria /cell , Figure S2 ) . After 4h , infected cells were washed twice with warm RPMI and treated with gentamicin ( 100 µg/ml ) for 2h to remove any remaining extra cellular bacteria , and then in complete RPMI containing 10 µg/mL gentamicin for the rest of the experiment . These cells were then transfected with siRNA at a final concentration of 100nM using hiperfect transfection reagent ( Qiagen ) according to manufacturer's protocol . At 48h a second siRNA treatment was performed ( see Text S1 ) , and the cells cultured for an additional 36h ( i . e . a total culture period of 90h after initiation of infection ) . At this point cells were solubilized in 50 µL of 0 . 06% SDS ( in 7H9 medium ) , and CFUs determined using either the undiluted lysate , or from lysate dilutions of 1∶10 , or 1∶100 . Each siRNA pool was evaluated in duplicate and the mean of the CFU values , obtained at the two cell lysate dilutions was determined . Each 96-well plate also included six negative control wells . In two of these cells were treated with scrambled siRNA following infection , whereas another two included infected cells that were transfected with GFP-specific siRNA . In the remaining two wells , infected cells were treated only with the transfection reagent ( Hiperfect ) . The mean CFU count obtained from all of these six wells , in each 96-well plate , was taken as the control value for that plate for determining the effects of target-specific siRNAs . The standard deviation ( SD ) for the control values was also calculated . A cut off of two SD from the mean value of control wells was employed to designate the effects of a given siRNA treatment as significant . The ‘hits’ obtained from this primary screen were then further validated with siRNA pools obtained from alternate sources as described above . Here , ‘hits’ were confirmed on the basis of the percent change in CFUs from the mean control value , in addition to an SD cut off value of two . That is , an increase or decrease of 50% from the control value for an siRNA with the additional caveat that this deviation from the mean control value was greater than 2SD was selected as validated ( see Fig . 1 ) . The validation exercise involved two separate experiments , whereas the subsequent comparison of the effects of the validated siRNA pools in cells infected with H37Rv , with that in cells infected with 1934 and JAL2261 was also ascertained in two additional independent experiments . In two separate experiments , cells were plated in six-well plate ( 2×106 cells per well ) and infected with H37Rv as described above . At 16h , 48h , and 96h later these , and uninfected , cells were lysed and RNA was isolated with trizol . One-color microarray-based gene expression analysis was performed by hybridizing against a mouse whole genome array consisting of probes for 44 , 000 genes ( Agilent ) . Bacteria were stained using the membrane stain PKH67 ( Sigma ) according to the manufacturer's protocol . J774 . 1 cells were seeded onto #1 thickness , 12 mm diameter glass cover-slips pre-coated with fibronectin in 24-well tissue culture plates at a density of 0 . 07×106 cells per cover-slip , respectively and infected with stained bacteria using the protocol as described above . Cells were incubated with 100nM Lysotracker during the last hour of the 72 hr chase at 37°C and then fixed with 4% para-formaldehyde ( Sigma ) . The cover slips were washed thoroughly with PBS and were mounted on slides with Antifade ( Biorad ) . Stained cells were observed with a Nikon TE 2000E laser scanning confocal microscope equipped with 60×/1 . 4 NA PlanApochromat DIC objective lens , and the extent of bacterial co-localization with acidified lysosomes was determined as the Overlap Coefficient ( see Text S1 ) . Groups of naive mice ( Female BALB/c mice 4–6 wk of age at 4/group ) were infected with 1×106 M . tuberculosis H37Rv via the tail vein . One group of mice was sacrificed 24 h later and lung homogenates were plated onto 7H11 agar plates for confirming infection . At ten days post infection , compound D4476 a final concentration of either 25 µM or 50 µM was injected into the tail vein of mice for treatment . The i . v . injection was repeated on seventh day of treatment . In addition the inhibitor was also given intraperitoneally on the third , fifth , tenth and twelfth day after initiation of treatment . At fourteen days following treatment , mice were sacrificed by carbon dioxide narcosis . For experiments involving the aerosol route of infection , mice ( 8 per group ) were infected with H37Rv by delivering between 100–150 bacteria per lung – as determined by the culture of lung homogenates at 24 h later - during 30 min of exposure . At the relevant times after infection , mice were sacrificed by carbon dioxide narcosis . The lungs were perfused and removed aseptically and weighed . The lungs were then homogenized and dilutions of these homogenates were plated on 7H11 agar plates for subsequent enumeration of the CFU .
|
The adaptation of Mycobacterium tuberculosis ( Mtb ) involves dynamic interactions with the molecular components of the host cellular machinery . Therefore , targeting relevant host factors may provide an alternate approach for the chemotherapy of tuberculosis ( TB ) . To test this , we first performed an siRNA screen targeting all known kinases and phosphatases in murine macrophages infected with a virulent strain of Mtb . A subsequent validation of this screen then identified several host molecules whose depletion severely affected the intracellular survival of mycobacteria . We also then screened against the identified host targets in cells infected with independent isolates of MDR-Mtb . This exercise identified those host molecules that were indispensable for supporting infection , independent of the phenotypic variations exhibited by the pathogen . Then , by using a pharmacological inhibitor that simultaneously targeted two of these molecules , we were able to demonstrate clearance of both drug-sensitive and drug-resistant strains of Mtb from infected cells . Importantly , this inhibitor was also effective in mice infected with the virulent strain of Mtb . Thus , in addition to demonstrating the feasibility of targeting host molecules involved in supporting intracellular persistence of pathogen for TB therapy , our studies also identify several such molecules that may be exploited for the purposes of drug development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"infectious",
"diseases/bacterial",
"infections",
"cell",
"biology/cell",
"signaling"
] |
2010
|
Identification of Host-Dependent Survival Factors for Intracellular Mycobacterium tuberculosis through an siRNA Screen
|
The importance of Zika virus ( ZIKV ) has increased noticeably since the outbreak in the Americas in 2015 , when the illness was associated with congenital disorders . Although there is evidence of sexual transmission of the virus , the mosquito Aedes aegypti is believed to be the main vector for transmission to humans . This species of mosquito has not only been found naturally infected with ZIKV , but also has been the subject of study in many vector competence assays that employ different strains of ZIKV around the world . In Argentina , the first case was reported in February 2016 and a total of 278 autochthonous cases have since been confirmed , however , ZIKV virus has not been isolated from any mosquito species yet in Argentina . In order to elucidate if Argentinian Ae . aegypti populations could be a possible vector of ZIKV , we conducted vector competence studies that involved a local strain of ZIKV from Chaco province , and a Venezuelan strain obtained from an imported case . For this purpose , Ae . aegypti adults from the temperate area of Argentina ( Buenos Aires province ) were fed with infected blood . Body , legs and saliva were harvested and tested by plaque titration on plates of Vero cells for ZIKV at 7 , 11 and 14 days post infection ( DPI ) in order to calculate infection , transmission , and dissemination rates , respectively . Both strains were able to infect mosquitoes at all DPIs , whereas dissemination and transmission were observed at all DPIs for the Argentinian strain but only at 14 DPI for the Venezuelan strain . This study proves the ability of Ae . aegypti mosquitoes from Argentina to become infected with two different strains of ZIKV , both belonging to the Asian lineage , and that the virus can disseminate to the legs and salivary glands .
Zika virus ( ZIKV ) is a single-stranded positive sense RNA virus that was first isolated in 1947 in the Ziika Forest in Uganda from a sentinel Rhesus monkey [1 , 2] . One year later , this arthropod-borne virus ( arbovirus ) member of the genus Flavivirus was isolated from Aedes africanus mosquitoes in the same forest , suggesting the mosquito as vector of the virus [2] . Outside Africa , ZIKV was isolated for the first time from Aedes aegypti in Malaysia in 1966 , providing evidence of transmission by an urban vector [3] . Since then , human cases were reported occasionally in Africa and Asia , until 2007 , when a massive outbreak was reported in Yap Island where the virus seems to have emerged from its sylvatic cycle to a rural habitat , causing fever , rash , conjunctivitis , and arthralgia [4 , 5] . Successively , a ZIKV outbreak that affected approximately 11% of the population occurred in French Polynesia in 2013 . During this outbreak , the Guillain-Barré syndrome ( GBS ) was associated with ZIKV for the first time [5] . In the Americas , the virus was introduced in Brazil [6] , probably after the World Cup soccer games held between June and July 2014 , or during the 2014 World Sprint Championship held in Rio de Janeiro in August , in which four Pacific countries participated [7] . The first cases in patients were reported in 2015 [8] . Since this outbreak , the virus had spread all over the country by the end of 2015 , reaching other 28 countries in South and Central America by February 2016 [5] . The interest in ZIKV has increased when the infection was correlated with severe congenital disorders such as microcephaly ( MC ) and other neurological malformations in fetuses and newborns [9] , especially when mothers are infected during the first trimester of pregnancy [10] . Due to the consequences caused by MC that severely affects cognitive and motor skills , many families are force to leave their jobs to care for their children having an impact in their socioeconomic status [5] . Although there is evidence that sexual intercourse is a route of transmission between humans [11 , 12] , the mosquito bite is still believed to be responsible for the dispersion of the virus , with humans as amplification hosts in endemic and epidemic zones during the urban cycle [13] . Sixteen different Aedes spp . mosquitoes were found naturally infected in the field with ZIKV . Among all , Ae . aegypti is considered to be the predominant species in the transmission of the virus , probably because it is close associated with humans in urban areas [14] . Additionally , ZIKV has been isolated from Ae . aegypti in field-caught specimens , although very occasionally , with evidence of vertical transmission detected in a few cases [3 , 15–20] . Furthermore , Ae . aegypti was confirmed to be competent in the transmission of ZIKV in a large number of experimental assays [4 , 14] , with differences in transmission efficiency been attributed to the genetic background of the vector population and the virus strain utilized [4 , 14 , 21–24] . In Argentina , ZIKV was first detected in the Córdoba province ( temperate central area ) in February 2016 , with the case being attributed to the sexual transmission of the virus ( Fig 1 ) . A few weeks later , the first outbreak occurred in Tucumán province resulting in 25 confirmed autochthonous cases . During October 2016 , the first case of congenital syndrome caused by ZIKV infection was confirmed in a newborn in Tucumán [25] . During the first semester of 2017 , 251 autochthonous cases of Zika were registered in Formosa , Salta and Chaco , all provinces located in North Argentina close to the Bolivian and Paraguayan borders ( Fig 1 ) [25] . Additionally , a MC case was detected in a newborn in Santa Fe province . Although the mother had no recent travel records , additional studies confirmed her ZIKV infection status [25] . Interestingly , during 2018 ZIKV circulation seems to be confined to Salta province where 54 locally acquired cases where confirmed . Additionally , in Buenos Aires province the only ZIKV case without travel antecedent was registered until now [26] . Due to the lack of entomological surveillance studies in Argentina during the ZIKV outbreak , as well as no vector competence studies in Argentinian mosquitoes , we aim to evaluate the potential role of Ae . aegypti in ZIKV transmission . To do so , we challenged a population of Ae . aegypti from La Plata in the Buenos Aires province against two different strains of ZIKV , in order to determine not only if the mosquito population from this region would be capable of transmitting the virus , but also to test for different levels of vector competence using distinct viral strains . One strain derives from an imported case from Venezuela , whereas the other strain was isolated from an autochthon case from Chaco province in Argentina .
Aedes aegypti mosquitoes employed in this study are derived from a laboratory colony established from La Plata in Buenos Aires province ( Argentina ) . The colony was originated in 2014 from larvae mosquitoes originally collected from La Plata cemetery , and since then it has been maintained at Centro de Estudios Parasitológicos y de Vectores ( CEPAVE ) in La Plata . Periodical introgressions of field mosquitoes from the same area of the city are incorporated seasonally to the colony in order to keep the genetic background as close as possible to the field . Multilocus genotype analysis was carried out with some individuals from the same area of collection [27] . For this study we selected two Zika virus strains: one strain ( ZIKV-VEN ) was isolated from a patient who travelled from Venezuela to Argentina ( strain ARCB116141 , GenBank accession no . MK637519 ) , whereas the other strain ( ZIKV-ARG ) was isolated from a patient in Chaco where the virus circulated among the population ( strain ARCH125797 , GenBank accession no . MK637518 ) . These strains show a nucleotide identity of 99 . 5% and an amino acid identity of 100% for a fragment that enclosed the last part of the capsid ( C ) , the precursor membrane segment ( prM ) , and the first part of the envelope ( E ) . The viral stock was prepared after four passages on Vero cells and frozen at −86°C before being employed in oral infections . The titers of each strain were analyzed by plaque assays on 12-well plates of Vero cells . The titer for the ZIKV-VEN strain was 7 . 17 log10 PFU/ml whereas the titer for the ZIKV-ARG was 5 . 3 log10 PFU/ml . These values are close to the range of viremia that has been previously reported for ZIKV infection [28] . Adult mosquitoes were maintained via incubation at 27 ± 1°C , 70 ± 10% RH , 16:8 hours light:dark cycle , and supply with sugar and water , except for a period of 24 hours of starvation prior to the oral infection . Four-five days after emergence , mosquitoes were offered a blood meal supplied by a glass artificial feeder which allows maintaining the blood temperature at 37°C . Blood meals were comprised of 4 ml of ovine blood ( Laboratorio Alfredo Gutierrez , C . A . B . A . , Argentina ) , 0 . 5 ml of sucrose 50% , and 0 . 5 ml of previously frozen cell culture supernatant containing ZIKV . Mosquitoes were fed for one hour . Engorged females were counted and separated in three cardboard cages to be analysed at 7 , 11 , and 14 days post infection ( DPI ) . The cages were placed back into the incubator at the same conditions of temperature and humidity until sample collection at each time point . Each group was offered sugar and water ad libitum . All the infections assays were performed in biosafety level 3 facilities at Instituto Nacional de Enfermedades Virales Humanas ( INEVH ) in Pergamino . At 7 , 11 and 14 DPI mosquitoes were anesthetized with triethylamine as previously described [29] , and bodies , legs , and saliva were harvested from each mosquito . Different time points were selected in order to determine the extrinsic incubation period ( EIP ) , which corresponds to the time between oral infection and presence of virus in saliva . The proboscis of each immobilized mosquito was inserted into a capillary tube containing 5 μl of Minimum Essential Medium ( MEM ) supplemented with 20% of fetal bovine serum ( FBS ) . After 30 min of salivation , the proboscis was removed from the capillary tubes , and legs and bodies separated into individual tubes . Each capillary tube containing salivary expectorate was collected from the capillary into a tube containing 300 μl of MEM supplemented with 20% FBS . All samples were stored at -86 °C until processing . Bodies and legs were each homogenized separately in microcentrifuge tubes containing 1 . 4 mm ceramic beads and 1 ml MEM with 20% FBS , for one min at 20 cycles per second using a Bead Ruptor 24 Elite ( OMNI international , Kennesaw , Georgia , USA ) . Homogenates were clarified by centrifugation at 5000xG for 10 min at 4 °C . In order to detect ZIKV infectious virions , all samples were analysed by plaque titration on 12-well plates of Vero C76 cells . Titration was performed as previously described [30] . Briefly , tenfold serial dilutions of each sample in MEM supplemented with 2% FBS and antibiotics were added in a confluent Vero C76 monolayers attached to 12-well plates and incubated for 1 hour with periodic gentle rocking to facilitate virus adsorption at 37°C . The volume of the inoculums was 100 ul in each well . Plaques were incubated undisturbed for 5 days at 37°C . Vital dye neutral red was used at 2% for plaque visualization . The mosquito body was examined to estimate the infection rate ( IR ) , the legs to estimate the dissemination rate ( DR ) , and saliva for the transmission rate ( TR ) of the virus . IR is defined as the percentage of mosquitoes with infected body among total engorged mosquitoes . DR corresponds to the percentage of mosquitoes that contained infectious virus in their legs among the previously infected mosquitoes detected . TR is reported as the percentage of mosquitoes that contained infectious virus in the saliva , among mosquitoes with disseminated infection . Transmission efficiency ( TE ) refers to the proportion of mosquitoes with infectious saliva among the total number of engorged mosquitoes . Differences in the IR , DR and TR between the two strains ( ZIKV-VEN and ZIKV-ARG ) were compared by a Fisher exact test , considering statistically significant p-value < 0 . 05 . Comparisons of viral titers in body and legs between both strains were performed at 14 DPI by using Student’s t-test or permutation test according to data normality . All analyses were performed using R software ( v . 3 . 5 . 0 ) [31–34] .
The total number of mosquitoes employed in the assay for both strains were similar ( 60 and 61 specimens ) ( Table 1 ) . Infection was successful for both ZIKV strains in all three DPIs , and IR varied from 15 . 8% to 50% for ZIKV-ARG and from 11 . 1% to 61 . 8% for the ZIKV-VEN strain ( Table 1 ) . Virus dissemination was found at 7 , 11 and 14 DPI for the ZIKV-ARG strain , with a total of 14 mosquitoes confirmed to have disseminated virus during the experiment ( 60 . 9% ) . On the other hand , for the ZIKV-VEN strain , dissemination was observed only at 14 DPI , where virus dissemination was detected in 52 . 4% of the mosquitoes . Although the proportion of ZIKV-infected saliva from Ae . aegypti was low for both strains , transmission was still observed . One of 11 samples was found positive for ZIKV-VEN strain in saliva at 14 DPI , whereas one sample of 6 was found positive for ZIKV-ARG in saliva by the same DPI . However , unlike the imported strain , TR was detected at all DPIs for the ZIKV-ARG strain . The ZIKV-ARG strain exhibited a minimum extrinsic incubation period ( EIP ) of 7 days , while the EIP for ZIKV-VEN was 14 days . Finally , the total TE was 6 . 6% for ZIKV-ARG and 1 . 7% for ZIKV-VEN . No significant differences were detected for the total IR ( p-value = 0 . 71 ) , DR ( p-value = 0 . 61 ) , TR ( p-value = 0 . 34 ) and TE ( p-value = 0 . 71 ) between both strains . Virus titer in saliva was detected by plaque assay in all samples despite the results of IR and DR . We did not find any saliva sample positive where infection and dissemination were negative . For both ZIKV strains the mean titers in body , legs and saliva were calculated for each DPI ( Table 2 ) . Furthermore , comparisons between two strains were performed at 14 DPI for body and legs , when data was available . There were not significant differences in the body between both strains ( Student’s t-test , t = -0 . 47 , df = 26 , p-value = 0 . 64 ) . When comparing viral titers between legs , a permutation test was preferred due to the low number of samples . In this case , significant differences were found between both strains , being the viral titers in legs for ZIKV-VEN higher than those for ZIKV-ARG ( Z = -2 . 04 , p-value = 0 . 04 ) . Additionally , comparisons between legs and saliva showed that the average viral titers in saliva dropped respect to the legs 8 . 4% for ZIKV-ARG , and 40 . 7% for ZIKV-VEN .
The ZIKV outbreak in Brazil in 2015 triggered an international alarm , especially when neurological disorders and microcephaly in newborns were associated with the infection [5] . Due to its proximity to Brazil , and the presence of the implicated vector , Argentina also focused the attention on this neglected disease . Interestingly , 137 , 288 Zika autochthonous cases were reported by the Brazilian Ministry of Health by January 2018 , whereas in Argentina 278 autochthonous cases were confirmed by the Argentinian Ministry of Health in the same period , confined to five provinces in the Northern region of the country [35] . In America , Ae . albopictus was found naturally infected with ZIKV in Brazil , while Ae . aegypti-infected mosquitoes were detected in Brazil , Ecuador and Mexico [17 , 20 , 36 , 37] . Vector competence studies that involved Ae . aegypti population from different countries corroborated the efficiency of this species in transmitting different strains of ZIKV , however competence varies greatly , and depending mainly on mosquito origin , Zika strain and type of blood meal used [21–24 , 38–41] . Vector competence of Ae . aegypti for ZIKV has been evaluated through all five continents . In Africa , mosquitoes populations from Senegal and Nigeria were tested for 14 different ZIKV strains , all of them infected this species but only two strains reached the saliva . In Asia , mosquitos from Singapore were able to transmit three strains of ZIKV showing an EIP of 3 and 4 days . Three ZIKV strains were also transmitted by Ae . aegypti from Australia and French Polynesia . In Europe a population of Ae . aegypti from Madeira Island was tested against two different strains of ZIKV; both strains infected the mosquitoes but only one of them was transmitted ( EIP = 9 ) . Finally , vector competence studies were also carried out in Mexican and Brazilian mosquitoes , which were able to transmit seven and three different strains of ZIKV , respectively [13] . These studies , together with the isolation of ZIKV from field-collected mosquitoes , confirm Ae . aegypti as the main vector of ZIKV [13 , 14] . In Argentina , ZIKV has not yet been isolated in the field from any mosquito species , and no vector competence studies were performed so far with local ZIKV strains . In this study we evaluated the vector competence of a local population of Ae . aegypti from the temperate area of Argentina ( La Plata ) , for two different Zika virus strains both belonging to the Asian lineage . One strain was isolated from a patient who has travelled to Venezuela ( ZIKV-VEN ) , whereas the other strain was isolated in Chaco , Argentina , during the outbreak in 2017 ( ZIKV-ARG ) . We demonstrated that the Argentinian Ae . aegypti population is able to be infected by both ZIKV-VEN and ZIKV-ARG strains . Despite the fact that the proportion of mosquitoes infected with these strains was relatively high , the TR remains very low for both strains . Moreover , the titers in the transmitting mosquitoes were also very low . However , it should be considered that due to the low number of infected mosquitoes at 7 DPI for both strains , and 11 DPI for ZIKV-VEN , DR and TR at these time points may not reflect the actual susceptibility of the population used in this study . For the Argentinian strain we detected an overall TE ( 6 . 6% ) that was slightly higher than the overall TE for the Venezuelan strain ( 1 . 7% ) . These data are closer to the TE detected in Ae . aegypti from Rio de Janeiro ( 10% ) , than the TE found for ZIKV strains transmitted by Ae . aegypti populations from Los Angeles ( 53–75% ) [22 , 42] . One possible explanation for the low levels of transmission could be due to the employment of frozen virus stocks . The difficulty in conducting the experiment with fresh virus leaves us questions about whether the susceptibility of the strains may vary according to a different response to the virus freeze/thaw , leading to a decrease of transmission efficiency [23 , 43 , 44] . Another factor that could influence on the transmission is the genetics of the mosquito . In fact , Ae . aegypti s . l . is divided into two genetic units which correspond to the standard defined subspecies: Ae . aegypti formosus , in Africa , and Ae . aegypti aegypti outside Africa . In Argentina , a mixture between both subspecies was detected in four different populations , including that from La Plata [45] . The fact that Argentinian populations have an African background could be an explanation for the low transmission of ZIKV , since some Ae . aegypti populations from Senegal were not competent for the transmission of this virus [46] . The circulation of the virus in Argentina was very low compared to Brazil . Interestingly , during 2016 , Argentina had the greatest Dengue outbreak in magnitude and geographical dissemination to date . The infection spread through 15 provinces of the center and north of the country , with 41 , 207 confirmed cases . This situation could have affected the detection of ZIKV infection from both the clinical aspect and the serological studies ( IgM cross reactivity ) [47] . In Argentina , laboratory assays indicates high vector competence for DENV-2 of Ae . aegypti populations from both subtropical and temperate areas . Mosquitoes from the subtropics were even more efficient that those of temperate Argentina [48] . Additionally , vector competence studies for DENV and chikungunya virus ( CHIKV ) in genetically distinct populations of Ae . aegypti from Argentina ( one of these belong to the same population that we used in this study ) , showed a large variability in vector competence for these viruses . In particular , La Plata Ae . aegypti were highly refractory to CHIKV infection and even at mean temperatures ( higher than the specific-site temperature for La Plata ) the population was more refractory than other populations for both pathogens [27] . It is remarkable that although Buenos Aires is the most populated province in the country , and many imported cases were diagnosed from travellers who arrived to the province from all over the world , and considering that the abundance of Ae . aegypti has increased during the last 20 years , only one autochthonous ZIKV case was reported by the Ministry of Health [26 , 49 , 50] . Additionally , Ae . aegypti is the only invasive species in the province since Ae . albopictus is present only in the Misiones province bordering Brazil [51 , 52] . The low TE of the La Plata Ae . aegypti population observed here might be an explanation for the absence of an outbreak in the region . The various barriers encountered during the extrinsic incubation period could probably influence the infection and replication in different tissues [53] . Our results show that the decrease in the titers in saliva related to those in legs is four times higher in the Venezuelan strain than in the Argentinian , although mean viral titers were significantly higher in legs for ZIKV-VEN than ZIKV-ARG , suggesting differences in the viral strain fitness . Another difference in the fitness between both viral strains is observed by the shorter EIP of the strain that circulated in the country ( ZIKV-ARG ) compared to the imported strain . As was mentioned before , ZIKV outbreak in Argentina was simultaneous with DENV outbreak . Further studies will be necessary in order to evaluate the impact of arboviruses coinfections on the epidemiology of these diseases . Moreover , we will explore in future studies the transmission of the virus among other populations of Ae . aegypti belonging to areas of Argentina where Zika outbreaks were reported . Because of the genetic diversity observed among Ae . aegypti in Argentina , these populations could be more susceptible to ZIKV transmission than those from La Plata [27] . Furthermore , albeit the attention is focused on Ae . aegypti , we do not preclude that other species of Aedes could be involved in the transmission of the virus , especially in other provinces affected by the outbreak with a higher diversity of Aedes spp . mosquitoes than Buenos Aires [54] . Another factor that should be considered is that RNA viruses replicate with low genetic fidelity that results in high mutation rates . If these genetic changes are able to generate new variants , there could be variability in epidemiological fitness and therefore an increment of vector competence [55 , 56] . Evidence of the importance of mutation in arbovirus is provided by Chikungunya virus , which adapted to the vector Ae . albopictus after a single adaptive mutation [57] , and for West Nile virus ( WNV ) which with a single-amino acid substitution became resistant to lycorine [58] . Also , the spread and continuous evolution of WNV led to the change from attenuated to virulent phenotype for lineage 2 [59] . For these reasons we could expect ZIKV to adapt further to local Ae . aegypti populations . Therefore active surveillance on circulating strains and other vector competence studies could contribute to elucidate the dynamics of the virus in this region .
|
Zika virus is a flavivirus transmitted by mosquitoes , isolated for the first time in the Ziika Forest in Uganda in 1947 from a rhesus macaque monkey . The disease is usually asymptomatic , but sometimes it causes a mild illness that comes with fever , rash , joint pain , and conjunctivitis . The World Health Organization focused the attention on this virus after the outbreak in the Americas , when the virus was linked to microcephaly and serious neurological diseases , including Guillain-Barré syndrome . Aedes aegypti was incriminated as the main vector of the virus as it was found both naturally and experimentally infected . This mosquito species was declared eradicated in Argentina by 1970 but re-emerged in 1989 . Recent studies found a peculiarity in the genetics of Argentinian Ae . aegypti populations that consists in a combination between both subspecies: Ae . aegypti formosus and Ae . aegypti aegypti . Our study tries to elucidate if Ae . aegypti from Argentina are able to transmit the virus in order to add these mosquitoes to the list of possible vectors of ZIKV and , in future prospect , orient to fight the virus by controlling the vector .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"pathogens",
"geographical",
"locations",
"microbiology",
"argentina",
"saliva",
"animals",
"viral",
"vectors",
"viruses",
"rna",
"viruses",
"insect",
"vectors",
"infectious",
"diseases",
"south",
"america",
"aedes",
"aegypti",
"medical",
"microbiology",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"brazil",
"arthropoda",
"people",
"and",
"places",
"mosquitoes",
"eukaryota",
"blood",
"anatomy",
"flaviviruses",
"virology",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"organisms",
"zika",
"virus"
] |
2019
|
Vector competence of Aedes aegypti for different strains of Zika virus in Argentina
|
Finely tuned changes in cytosolic free calcium ( [Ca2+]c ) mediate numerous intracellular functions resulting in the activation or inactivation of a series of target proteins . Palmitoylation is a reversible post-translational modification involved in membrane protein trafficking between membranes and in their functional modulation . However , studies on the relationship between palmitoylation and calcium signaling have been limited . Here , we demonstrate that the yeast palmitoyl transferase ScAkr1p homolog , AkrA in Aspergillus nidulans , regulates [Ca2+]c homeostasis . Deletion of akrA showed marked defects in hyphal growth and conidiation under low calcium conditions which were similar to the effects of deleting components of the high-affinity calcium uptake system ( HACS ) . The [Ca2+]c dynamics in living cells expressing the calcium reporter aequorin in different akrA mutant backgrounds were defective in their [Ca2+]c responses to high extracellular Ca2+ stress or drugs that cause ER or plasma membrane stress . All of these effects on the [Ca2+]c responses mediated by AkrA were closely associated with the cysteine residue of the AkrA DHHC motif , which is required for palmitoylation by AkrA . Using the acyl-biotin exchange chemistry assay combined with proteomic mass spectrometry , we identified protein substrates palmitoylated by AkrA including two new putative P-type ATPases ( Pmc1 and Spf1 homologs ) , a putative proton V-type proton ATPase ( Vma5 homolog ) and three putative proteins in A . nidulans , the transcripts of which have previously been shown to be induced by extracellular calcium stress in a CrzA-dependent manner . Thus , our findings provide strong evidence that the AkrA protein regulates [Ca2+]c homeostasis by palmitoylating these protein candidates and give new insights the role of palmitoylation in the regulation of calcium-mediated responses to extracellular , ER or plasma membrane stress .
In all eukaryotic cells , the cytosolic free calcium ( [Ca2+]c ) concentration is strictly and precisely controlled by complex interactions between various calcium-channels , calcium-pumps and calcium-antiporters and by calcium buffering in the cytoplasm . Finely tuned changes in [Ca2+]c mediate a variety of intracellular functions , and disruption of [Ca2+]c homeostasis can lead to various pathological conditions [1] . In fungi , numerous studies have shown that calcium signaling is involved in regulating a wide range of processes including cell morphogenesis , cell cycle progression , stress responses and virulence [2] . Two different calcium uptake systems in the plasma membrane have been identified in most fungal species: the high-affinity Ca2+ influx system ( HACS ) and the low-affinity calcium influx system ( LACS ) [3–5] . The main components of the HACS are primarily composed of an α-subunit of the mammalian voltage-gated Ca2+-channel homolog Cch1 and a stretch-activated β-subunit called Mid1 . Loss of the HACS results in an inability to grow under low-calcium conditions . In addition , fungi possess a range of other calcium P-type ATPases and calcium transporters that play important roles in calcium signaling and homeostasis [6] . Upon stimulation , calcium is rapidly taken up from the extracellular environment or released from these intracellular calcium stores and either interacts with the primary intracellular calcium sensor/receptor calmodulin or directly regulates that activity of other proteins . When the calcium signal binds to calmodulin this results in a conformational change in the protein allowing it to interact with and regulate the activity of various target proteins involved in converting the original stimuli into cellular responses . The [Ca2+]c increase is transient because various calcium-pumps and calcium-antiporters , as well as the cytoplasmic calcium buffering , subsequently return the [Ca2+]c to its normally low resting level within the cytosol [7 , 8] . The phosphatase calcineurin is an important [Ca2+]c transient effector and is conserved from yeast to humans . Its most well known target in fungi is the transcription factor Crz1 ( calcineurin responsive zinc finger 1 ) [9 , 10] . In vegetatively growing S . cerevisiae cells , [Ca2+]c concentrations are normally maintained at low non-signaling levels . During this stage , Crz1 is fully phosphorylated , localized to the cytoplasm , and transcriptionally inactive [11 , 12] . When fungal cells are exposed to chemicals that induce plasma membrane stress ( e . g . by azole antifungals ) or endoplasmic reticulum ( ER ) stress ( e . g . by tunicamycin ) , or are under low calcium conditions , the HACS is activated . These stimuli result in calcium uptake and a transient increase in [Ca2+]c which leads to calcineurin activation and subsequent Crz1 de-phosphorylation . Crz1 is then recruited to nuclei where it transcriptionally regulates downstream signaling pathways to alleviate cellular stress and promote cell survival [13 , 14] . Interestingly , there are no known mammalian Crz1 orthologs , but mammals express another calcineurin sensitive transcription factor target , known as NFAT ( nuclear factor of activated T-cells ) . Crz1 does not belong to the NFAT family , but the Zn-finger domains in Crz1 and NFAT bind specific DNA sequences within the promoter regions of calcineurin-dependent response elements ( CDREs ) to activate transcription [15 , 16] . In the filamentous fungus Aspergillus nidulans , there is a calcineurin-dependent Crz1 homolog , known as CrzA . Interestingly , calcineurin deletion causes more severe growth defects than CrzA deletion in this species , suggesting that calcineurin has additional target proteins other than CrzA [17 , 18] . Palmitoylation is a reversible posttranslational modification that catalyzes the attachment of palmitate to cytoplasmic cysteine residues of soluble and transmembrane proteins . Palmitoyl transferases ( PATs ) are known to be responsible for palmitoylation . The defining feature of PATs is the presence of a cysteine-rich domain ( CRD ) with an Asp-His-His-Cys ( DHHC ) motif , which is required for PAT activity . Many proteins that require palmitoylation are involved in cellular signaling , membrane trafficking and synaptic transmission [19–21] . There are more than 20 encoded DHHC proteins in mammalian genomes , and there is now a major effort to verify DHHC-substrate partners and determine how their interaction specificity is encoded [22] . Several lines of recent evidence have shown that protein palmitoylation influences various cell functions , physiology and pathophysiology [23–25] . In this study , we have demonstrated that AnAkrA in A . nidulans and AfAkrA in A . fumigatus , which are homologs of the yeast palmitoyl transferase ScAkr1p , have similar function to the HACS in the presence of low extracellular calcium . The akrA deletion resulted in marked defects in hyphal extension and conidiation , especially under low calcium conditions . Moreover , using codon-optimized aequorin as a calcium reporter in living cells , we found that AkrA dysfunction significantly decreased the amplitude of the [Ca2+]c transient induced by an extracellular calcium stimulus , ER stress caused by tunicamycin or plasma membrane stress resulting from itraconazole , respectively . Our data suggest that these [Ca2+]c responses are mediated by the palmitoylation of the cysteine residue of the DHHC motif in AkrA . Moreover , we have identified that two new putative P-type ATPases ( Pmc1 and Spf1 homologs ) , a putative proton V-type proton ATPase ( Vma5 homolog ) and three putative CrzA-dependent proteins , are palmitoylated substrates of the AkrA protein . To our knowledge , this is the first report that a palmitoylation protein is involved in regulating eukaryotic calcium signaling .
Based on a NCBI BLASTp search ( http://www . ncbi . nlm . nih . gov/BLAST/ ) , we identified a putative ortholog of NFAT in A . nidulans , AkrA ( AN5824 . 4 , Accession: XP_663428 . 1 ) , which encodes a putative palmitoyltransferase . However , it showed low identity ( less than 20% ) or similarity ( less than 30% ) to mammalian NFAT based on full-length sequences . Interestingly , a bioinformatic analysis revealed that the promoter region contains a putative calcineurin-dependent-response-element ( CDRE-like ) motif . As shown in Fig 1A , we identified a CDRE-like sequence at 398 bp ( akrA , AN5824 . 4 ) , upstream of this gene’s start codon [26 , 27] . These data suggest that AkrA may be a component of the calcium signaling machinery . To further explore the function of the akrA gene and its relationship to calcineurin , the full-length deletion strain was constructed by homologous gene replacement employing a self-excising recyclable cassette that contains an AfpyrG gene as a selectable marker . Diagnostic PCR analysis of the resulting strain ΔakrA confirmed the homologous replacement ( S1A Fig ) . We also generated ΔakrAΔcnaA double mutants through genetic crosses ( the cnaA gene encodes the catalytic subunit of calcineurin ) . The ΔakrA mutant produced smaller colonies compared to that of the parental wild-type strain , when grown on minimal medium . In comparison , the ΔcnaA mutant exhibited severe growth defects on minimal medium . Moreover , the double mutant had a smaller colony size and underwent less conidiation than the single mutants ( Fig 1B ) . These results suggest that akrA and cnaA may have different functions in A . nidulans . Therefore , the double deletion mutant exacerbates the growth defects on minimal medium . We next tested whether low external calcium conditions could affect the colony phenotype in the akrA deletion mutant . When conidia were spot inoculated onto the solid minimal medium containing the calcium chelator EGTA and were allowed to grow at 37°C for 2 . 5 days , the ΔakrA mutant exhibited increased EGTA sensitivity compared to the parental wild-type strain . As shown in Fig 1C , the akrA deletion exhibited markedly reduced conidial formation and colony growth under low-calcium conditions . Since , mutants of the HACS components have been previously shown to exhibit similar defects under low calcium conditions [28–30] , we next examined whether AkrA was a potential novel HACS component . To determine whether the defects in the ΔakrA mutant could be rescued by high extracellular calcium , we inoculated ΔakrA mutant conidia on minimal medium supplemented with 20 mM Ca2+ . We found that the colony diameter of the ΔakrA mutant was restored almost to the same diameter of the parental wild-type strain by the addition of extracellular calcium ( Fig 1C ) , indicating that exogenous calcium could completely rescue the colony growth defect caused by AkrA loss . We further examined conidiation in the ΔakrA mutant in a calcium-limited environment ( i . e . in the presence of EGTA ) with a stereomicroscope ( Fig 1D left panels ) . The results showed that the vegetative mycelia from the parental wild-type strain were capable of producing numerous conidia under low-calcium conditions . In contrast , conidiation was almost completely abolished in the ΔakrA mutant on minimal media supplemented with EGTA ( 1 mM ) ( Fig 1D left panels ) . In submerged liquid culture , the wild-type strain displayed robust polarized hyphal growth around the margins of mycelial balls , whereas the ΔakrA mutant showed smooth margins around small mycelial balls ( Fig 1D right panels ) . Consistently , the ΔakrA mutant had a significantly reduced biomass , germination rate , and colony size compared to the parental strain on minimal media ( S3 Fig ) . Moreover , ectopically expressed akrA was able to completely rescue these defects in the akrA deletion strain ( Fig 1D ) , establishing that these phenotypes were specific to the loss of akrA . In addition , we deleted the akrA homolog gene in A . fumigatus . Similar to the ΔakrA phenotypes in A . nidulans , the ΔAfakrA mutant displayed hypersensitivity to the low calcium conditions , and its phenotypic defects could be rescued by high extracellular calcium ( S2 Fig ) . Thus , these data are consistent with AkrA being involved in calcium uptake especially in a calcium-limited environment . To further confirm and assess the localization and the molecular mass of AkrA , we generated a conditional expression allele , alcA ( p ) ::GFP-akrA , referred to here as ZYA09 ( S1B Fig ) . In this conditional allele , akrA expression was assumed to be regulated by the carbon source , as it was not induced by glucose , induced by glycerol , and overexpressed to high levels by L-threonine [31] . To determine whether this conditional allele behaved as predicted , we inoculated the ZYA09 strain in liquid media for 18 h , which promoted induction , non-induction or overexpression . As expected , the akrA mRNA level was approximately 20-fold higher when grown in overexpressing medium compared to that grown in non-inducing medium , which was 12-fold higher than that in inducing medium ( S4B Fig ) . Moreover , the conditional strain ZYA09 displayed an identical phenotype to the parental wild-type strain when grown on the inducing or the overexpressing media , indicating that the fusion GFP-AkrA protein was functional and that the assumed akrA over-expression had no detectable effects in A . nidulans . In comparison , when grown on the non-inducing medium , the conditional allele alcA ( p ) ::GFP-akrA exhibited an identical phenotype to the ΔakrA mutant , confirming a consistent phenotype for the loss of AkrA and for the knock-down of AkrA ( Figs 2A and 1C ) . Western blotting showed a band at approximately 110 kDa in the GFP-AkrA strain grown under inducing or overexpressing conditions using an anti-GFP antibody but no such a band appeared in the parental wild-type strain or the conditional allele ( ZYA09 ) under the non-inducing condition ( Fig 2B ) . These results indicate that the molecular mass of AkrA is approximately 80 kDa because GFP is a 27 kDa protein . Microscopic examination showed that the AkrA-GFP localization pattern resembled that of the Golgi previously reported in A . nidulans [32] . To confirm this we generated the strain ZYA13 by genetically crossing the alcA ( p ) ::GFP-akrA strain ZYA09 with the MAD2013 strain in which the late Golgi marker ( gpdAmini::mRFP-PHOSBP ) , consisting of the pleckstrin homology domain of the human oxysterol binding protein ( PHOSBP ) fused to mRFP was included [33 , 34] . Spores of the ZYA13 strain were incubated in non-inducing medium at 37°C for 10 h and were then shifted to the overexpression medium for 6 h . Microscopic examination of the young germlings produced under these conditions showed the majority of GFP-AkrA proteins colocalized with mRFP-PHOSBP late Golgi marker ( Fig 2C ) . Because the bioinformatic analysis showed that AkrA contains a conserved DHHC motif required for its palmitoylation activity [19–21] , we next investigated whether the DHHC motif was required for the normal function of AkrA under low calcium conditions . We first constructed a C-terminal AkrA truncation lacking the region from the DHHC motif through to the stop codon by homologous gene replacement ( Fig 3A ) . The colony phenotype of the truncation mutant was similar to that resulting from the complete deletion of the akrA gene when grown in minimal medium plus EGTA , indicating that the DHHC motif is required for AkrA function ( Fig 3B ) . To rule out the possibility that a loss of function in the truncated mutant might result from a conformational change that prevented a true reflection of the function of the DHHC motif , we performed site-directed mutagenesis . Since Cys487 in the DHHC motif has previously been shown to be crucial for palmitoyl transferase activity , we therefore mutated Cys487 to Ser487 in the DHHC motif ( Fig 3A ) [35 , 36] . Consequently , we found that the C487S site-mutated DHHS fragment could not rescue the defect of the akrA deletion mutant under either the control of a native promoter ( native ( p ) ::akrAC487S ) or a GPD promoter ( GPD ( p ) ::akrAC487S ) ( Fig 3B ) . In comparison , the wild-type akrA gene completely rescued the growth defects in the akrA deletion recipient strain . To confirm that these fusion cassettes were transcribed in the transformant , we performed quantitative real-time PCR to verify the akrA mRNA levels . The results showed that both the GPD and native promoters induced normal akrA mRNA expression , even though the mRNA expression level under the control of the GPD promoter was higher than that with the native promoter ( S4D and S4E Fig ) , indicating that the AkrA-DHHS cassettes were fully transcribed . Next , we generated Flag-tagged AkrA and the site mutated AkrAC487S strains to further confirm the expression of the AkrA protein . As shown in Fig 3C , the predicted bands on a Western blot were observed clearly , suggesting that both Flag-AkrA and Flag-AkrAC487S proteins were fully expressed in vivo . In addition , the Flag-tagged AkrAC487S strain displayed an identical phenotype to that of the Flag-untagged ( native ( p ) ::akrAC487S ) mutant , suggesting that the Flag tag could not phenotypically change the function of the targeted protein AkrA ( Fig 3B and 3D ) . These data suggest that the growth defect caused by akrA deletion was closely associated with the Cys487 site in the DHHC motif . Because the loss of akrA caused a similar defect phenotype to that of deletion mutants of the HACS components cchA and midA under the low calcium conditions , we hypothesized that AkrA forms a complex with CchA or MidA to perform its function . To assess whether AkrA physically interacts with CchA or MidA , we performed yeast two-hybrid assays . We cloned the cDNA fragments corresponding to the cytosolic C-terminus of cchA and the full-length cDNA of midA , respectively . They were then amplified and cloned into the pGADT7 vector , which contains the GAL4 DNA-AD and the LEU2 marker . In addition , a full-length cDNA of akrA was cloned into the pGBKT7 vector , which contains the GAL4 DNA-BD and TRP1 marker . As a result , some small colonies of pGBKT7-akrA with pGADT7-cchA were obtained , and there was no detectable growth of colonies of pGBKT7-akrA with pGADT7-midA under the high stringency screening conditions compared to the positive colonies of pGADT7-T and pGBKT7-53 , which showed robust growth ( S4A Fig ) . These data suggest that AkrA and MidA do not directly interact , and that AkrA and CchA might weakly and transiently interacted . We next investigated the functional interaction ( s ) between AkrA and CchA and between AkrA and MidA by a genetic phenotypic analysis . The ΔakrAΔmidA , ΔakrAΔcchA double mutants were generated by genetic crossing . As shown in Figs 4A and S6 , phenotypic defects in colony size and conidiation were exacerbated in the double mutants compared to the parental single mutants , especially in the presence of EGTA . Notably , the growth retardation of the ΔakrAΔmidA and ΔakrAΔcchA double mutants under low calcium conditions was reversed by the addition of 20 mM calcium to the minimal medium . These results suggest that AkrA , CchA , and MidA are all required under the calcium-limited condition , but may have some non-overlapping roles in growth . To determine whether overexpression of cchA could rescue the ΔakrA defects under the low calcium condition , we crossed ΔakrA ( ZYA02 ) and alcA ( p ) ::GFP-cchA ( ZYA11 ) to generate the ZYA12 strain . Real-time PCR verified that the mRNA level of cchA in ZYA12 was approximately 15-fold higher in the overexpressing medium than in the inducing medium when cultured for 18 h ( S4C Fig ) . However , overexpression of cchA did not rescue the ΔakrA defects under low calcium conditions ( Fig 4B ) . Previous studies have demonstrated that pmr1 , which encodes a Ca2+/Mn2+ P-type ATPase and is involved in Ca2+ homeostasis , localizes to the Golgi in yeast [37] . In A . nidulans , ΔpmrA had no discernible effect on fungal physiology , but the cells were hypersensitive to low extracellular calcium [38] . To investigate the link between AkrA and PmrA , we crossed the ΔakrA and ΔpmrA mutants . Surprisingly , the double mutant had no detectable defect when grown in minimal medium compared to the ΔakrA strain , which had a reduced-colony size ( Fig 4A ) . These data suggest that the pmrA deletion suppressed the ΔakrA growth defect . However , when cultured on minimal medium with 1 mM EGTA , the double mutant showed an exacerbated growth retardation phenotype compared to the parental single mutants . In addition , the phenotypic defects of ΔakrAΔpmrA were completely suppressed by the addition of 20 mM calcium . These results suggest that AkrA and PmrA may operate together in regulating cellular calcium homeostasis in a reverse way . Previous studies with yeast reported that Cch1 and Mid1 mutations reduced calcium uptake and affected [Ca2+]c accumulation under both stimulating and non-stimulating conditions [5 , 39–41] . We monitored the extracellular calcium-induced [Ca2+]c changes in living cells of A . nidulans wild type and mutant strains in which we expressed codon-optimized aequorin [42–44] . When treated with 0 . 1 M CaCl2 , the [Ca2+]c concentration in wild type cells transiently increased from a resting level of approximately 0 . 1 μM to a peak concentration of 1 . 2 μM ( Fig 5 ) . In comparison , cchA or midA mutants showed a reduction of 17 ± 11% or 25 ± 12% in the [Ca2+]c amplitudes , respectively , under the same stimulating conditions . Surprisingly , the decrease in the [Ca2+]c amplitude in akrA mutants was much larger than that observed in the HACS mutants . The [Ca2+]c amplitudes were decreased as follows: 53 ± 13% in the akrA deletion strain ZYA02 , 54 ± 9% in the DHHC truncated mutant ZYA15 , and 55 ± 8% in the site-mutated native ( p ) ::akrAC487S mutant ZYA16 . These data suggest the significant reduction in calcium influx due to the loss of AkrA is mediated by the DHHC motif and , in particular , the cysteine residue within the DHHC motif . The [Ca2+]c amplitude in the ΔpmrA mutant exposed to the 0 . 1 M CaCl2 stimulus was similar to that of the parental wild-type strain , which is different from that previously reported for yeast [45–47] , suggesting that other Ca2+-ATPases may compensate for the loss of PmrA function in response to the extracellular calcium stimulus . However , loss of pmrA in the akrA deletion background was able to recover the decreased [Ca2+]c amplitude in the akrA mutant to a similar level as that in the parental wild-type strain in response this extracellular calcium stimulus , indicating that the perturbation of calcium homeostasis induced by AkrA could be rescued by loss of pmrA . The protein palmitoylation inhibitor 2-bromopalmitate ( 2-BP ) is a palmitate analog that blocks palmitate incorporation into proteins [48 , 49] . To determine whether inhibition of palmitoyl transferase activity influences calcium influx into the cytoplasm , we measured the [Ca2+]c amplitude of the wild type pre-incubated in 2-BP ( 20 μM ) for 2 h . Following this drug treatment , the amplitude of the [Ca2+]c increase following stimulation with 0 . 1 M CaCl2 was significantly reduced by approximately 40% of the untreated cells in response to stimulation with 0 . 1 M CaCl2 ( Fig 5 ) . These data suggest that the inhibition of palmitoyl transferase activity can significantly block calcium influx . Activation of Ca2+ channels , calmodulin , calcineurin and other factors is necessary for the long-term survival of cells undergoing ER stress , and during this process the HACS components , CchA and MidA , are required for Ca2+ influx from the extracellular environment [41 , 50 , 51] . To verify whether AkrA is involved in the calcium influx response during ER stress , we measured the influence of the ER-stress agents , tunicamycin ( TM ) and dithiothreitol ( DTT ) on [Ca2+]c . When the parental wild-type strain was treated with 5 μg/mL tunicamycin , we observed an immediate transient increase in [Ca2+]c with an amplitude of 0 . 60 ± 0 . 03 μM ( Fig 6B ) . In comparison , the [Ca2+]c amplitude in the ΔcchA mutant ( but not the ΔmidA mutant ) in response to tunicamycin was decreased by approximately 32 ± 6% , suggesting that the loss of CchA but not MidA mediates the ER stress-induced calcium influx in A . nidulans . Furthermore , in response to tunicamycin treatment the [Ca2+]c amplitude decreased by 40 ± 5% , 34 ± 8% and 34 ± 6% in the ΔakrA , akrAΔC , native ( p ) ::akrAC487S mutants , respectively . We next examined the [Ca2+]c response after addition of DTT , another agent causing ER-stress . 10 mM DTT induced a rapid increase in [Ca2+]c which peaked at approximately 0 . 40 μM in the wild-type and ΔmidA strains , but the [Ca2+]c amplitudes decreased by approximately 40% in the ΔakrA ( 36 ± 10% ) , akrAΔC ( 37 ± 7% ) , and native ( p ) ::akrAC487S ( 36 ± 8% ) mutants , and by 15 ± 9% in the ΔcchA mutant ( S7 Fig ) . These data suggest that CchA , but not MidA , influences the ER stress-induced calcium influx in A . nidulans , which is different from that previously reported in yeast [41 , 51] . Furthermore , loss of AkrA , or mutations in its DHHC significantly decreased the ER stress-induced calcium influx . We further tested whether the amplitude of the [Ca2+]c increase in response to tunicamycin was dependent on the extracellular calcium concentration . We found that there was no significant change when mycelia were cultured in media with or without 5 mM calcium ( S8A Fig ) . In contrast , exposure of cells to 1 mM EGTA prior to tunicamycin treatment completely abolished the increase in [Ca2+]c in the ΔakrA , akrAΔC and native ( p ) ::akrAC487S mutants , but not in the parental wild-type , ΔcchA or ΔmidA strains ( Fig 6A ) . Similar data was obtained when we used the more selective , calcium chelator BAPTA ( S9 Fig ) . These data suggest that intracellular calcium stores contribute to the transient increase in [Ca2+]c induced by agents causing ER stress . Because azole antifungal drugs induce plasma membrane stress [13 , 14 , 52] , we next compared the differences in the [Ca2+]c transient between wild-type and relevant mutant strains after treatment with the azole antifungal agent itraconazole ( ITZ ) , which is currently used as a primary antifungal drug in the clinic . In all the tested mutants and the wild-type strain , the [Ca2+]c resting levels were similar at approximately 0 . 05 μM . After addition of 1 μg/mL ITZ to the medium , all strains responded with a transient increase in [Ca2+]c ( Fig 7B ) . However , all the akrA defective mutants exhibited significantly lower increases in [Ca2+]c compared to their parental wild-type strain: the amplitudes of the [Ca2+]c transients were reduced by 36 ± 11% in the ΔakrA , 29 ± 10% in the akrAΔC , 24 ± 8% in the native ( p ) ::akrAC487S and 27 ± 8% in the ΔcchA mutants , respectively , compared to that of the parental wild-type strain . In marked contrast to these mutants , the ΔmidA mutant exhibited a similar [Ca2+]c amplitude in response to ITZ as observed in the wild-type strain . In addition , the amplitude of the ITZ-induced [Ca2+]c elevation increased when mycelia were cultured in media containing 5 mM CaCl2 ( S8B Fig ) . We next examined whether the [Ca2+]c transient induced in response to ITZ was dependent on external calcium or internal calcium stores . We exposed hyphal cells to media supplemented with EGTA ( 1 mM ) prior to ITZ treatment , and found that [Ca2+]c transients were dramatically abolished in all the ΔakrA mutants , whereas the [Ca2+]c transients in the wild type , and the ΔcchA and ΔmidA mutants , were still observed ( Fig 7A ) . Similar data were obtained when we used the calcium chelator BAPTA ( S9 Fig ) . These data indicate that the loss of AkrA or disruption of its DHHC motif in the absence of extracellular calcium completely block calcium influx after treatment with chemicals that induce ER or plasma membrane stress from both extracellular and intracellular sources . Furthermore , both extracellular calcium and intracellular calcium stores play roles in generating these [Ca2+]c transients induced by these stress treatments . Our evidence above indicates that the cysteine residue in the DHHC motif of AkrA is involved in regulating the calcium response to high extracellular calcium- , ER- and plasma membrane-stress . To test whether the cysteine residue of DHHC is required for AkrA palmitoylation , we set up an acyl-biotin exchange ( ABE ) chemistry assay to detect palmitoylation in potential target proteins based on selective thioester hydrolysis by hydroxylamine ( HA ) ( Fig 8A ) . Compared to the control , the treatment of hydroxylamine combined with N-ethylmaleimide ( NEM ) ( which blocks free sulhydryls ) , efficiently enriches palmitoylated proteins . Subsequent treatment with HA cleaves the thioester bond between palmitate and cysteine residues , exposing bound thiols , which are then covalently linked to HPDP-biotin . The controls were protein samples not treated with HA . Lastly , the biotinylated proteins were bound to streptavidin agarose , washed with buffer , and eluted by cleavage of the cysteine-biotin disulfide linkage following by SDS-PAGE . Several previous reports have suggested that the process of palmitoylation involves in a two-step mechanism in which palmitoyl transferase is auto-acylated by itself to create an intermediate followed by the transfer of the palmitoyl moiety to its substrate [53 , 54] . Therefore , to investigate whether the cysteine residue in the DHHC motif is responsible for AkrA auto-acylation , we used the ABE assay to detect whether AkrA palmitoylates itself [20] . As shown in Fig 8B , when HA was present , Flag-AkrA can be clearly detected with an anti-Flag antibody . However , a site-directed mutation of the cysteine residue in the DHHC motif and the parental wild-type strain pre-cultured with 2-bromopalmitate ( 2-BP ) completely abolished palmitoylation of AkrA , which resulted in no signal being detected in the enriched pamitoylated proteins . These results indicate that AkrA is able to be auto-acylated and the cysteine residue in the DHHC motif is required for this process . In addition , we found that treatment with 2-BP ( 24 h , 50 and 100 μM ) virtually abolished the Golgi localization of GFP-labelled AkrA ( Fig 8D ) and resulted in a similar defective growth defect phenotype to the ΔakrA mutant on minimal medium ( S10 Fig ) . We constructed another alcA ( p ) ::GFP-akrAC487S mutant and confirmed by Western blotting ( Fig 8C ) to further check whether site directed mutagenesis of the Cys487 in the DHHC motif disrupted the normal localization of AkrA in the Golgi . The GFP-AkrAC487S was less distinctly localized in the punctate Golgi structures characteristic of wild-type GFP-AkrA and some appeared to be localized in the cytoplasm ( Fig 8D ) . These data collectively suggest that the cysteine residue in the DHHC motif of AkrA and the palmitoylation activity are closely associated with AkrA auto-acylation , which is required for normal AkrA localization and palmitoylation . To further explore palmitoylated protein substrates specifically mediated by AkrA , total proteins of the wild-type and ΔakrA strains were treated and analyzed using the ABE chemistry assay combined with liquid chromatograpy-mass spectrometry ( LC-MS ) for comparative proteomics ( Fig 8E ) . Using this approach , 334 proteins were identified as potential AkrA substrates in the parental wild-type strain because they were completely absent in the ΔakrA strain . As shown in Table 1 , AkrA belonged to one of the AkrA-mediated pamitoylated substrates suggesting it is able to auto-acylate itself . Among the palmitoylated protein candidates identified , Yck2 , Lcb1 , Ras2 , Cdc48 and Pab1 have been previously identified as palmitoylated proteins in S . cerevisiae but only Yck2 has been characterized as an Akr1 substrate [20 , 55–57] . These data indicated that the ABE chemistry assay combined with LC-MS was a valid approach to identify proteins palmitoylated by AkrA and it also indicated that A . nidulans may palmitoylate some of the substrates previously reported in S . cerevisiae . In our study we notably identified the following protein substrates palmitoylated by AkrA: a vacuolar Ca2+-ATPase Pmc1 homolog ( AN5088 . 4 ) ; a P-type ATPase Spf1 homolog ( AN3146 . 4 ) involved in calcium homeostasis [58]; a putative V-type H+-ATPase Vma5 homolog ( AN1195 . 4 ) that has been linked to Ca2+-ATPase function [59] , and three uncharacterized proteins ( AN8774 . 4 , AN3420 . 4 and AN2427 . 4 ) , the transcripts of which have previously been shown to be induced by extracellular calcium stress in a CrzA-dependent manner [53] . These results provide strong evidence that the AkrA protein regulates [Ca2+]c homeostasis in A . nidulans by palmitoylating these protein candidates . Other candidate substrates of AkrA that we identified included the P450 enzymes , Cyp51A ( Erg11A ) , Cyp51B ( Erg11B ) and Erg5 homologs , which are all involved in ergosterol biosynthesis and azole resistance . Thus AkrA may influence the azole resistance by these biosynthetic enzymes .
Deletion of the akrA gene exhibited marked growth and conidiation defects under low calcium conditions , which is similar to the defects caused by mutations in the CchA/MidA HACS [28–30] . In addition , the akrA deletion conferred increased sensitivity to Li+ , Na+ , K+ , Mg2+ , but slightly increased resistance to the cell wall disrupting agents compared to the parental wild-type strain ( S5 Fig ) . Moreover , the ΔakrAΔcchA and ΔakrAΔmidA double mutants exacerbated the ΔakrA defects under calcium-limited conditions , suggesting that AkrA may have independent functions to those of the CchA-MidA complex . AkrA localized to trans Golgi structures ( Fig 2C ) , while the CchA-MidA complex probably localizes to the plasma membrane as reported for yeast [40 , 66 , 67] . In addition , results from the Y2H assays ( S4A Fig ) suggested that there were no direct , or only very weak , interactions between AkrA and CchA and between AkrA and MidA . Nevertheless , the [Ca2+]c transient in the ΔakrA mutant had a much lower amplitude ( approximately 53 ± 13% lower ) than the wild-type control following treatment with a high extracellular calcium stress stimulus , suggesting that the loss of AkrA reduced calcium influx into the cytoplasm . In contrast , loss of CchA and MidA caused a 25% decrease in the [Ca2+]c amplitude in response to this treatment with high external calcium , consistent with the results from previous studies on yeast cells lacking either Cch1 or Mid1 , which exhibited a low calcium uptake [5 , 39–41] . The akrA deletion also had a bigger impact on inhibiting calcium influx in response to ER stress than observed in the ΔcchA and ΔmidA mutants . Overall our data suggests that AkrA regulates calcium uptake from the external medium as well and its release from intracellular Ca2+ stores through a pathway that is independent of the previously identified CchA/MidA HACS as shown in Fig 9 . PmrA is an A . nidulans homolog of yeast Pmr1 , which is a P-type Golgi Ca2+/Mn2+ ATPase responsible for Ca2+ transport into the Golgi and widely accepted as responsible for Ca2+ efflux from the cytoplasm into the Golgi to regulate calcium signaling and homeostasis and prevent calcium toxicity . Loss of Pmr1 function in budding yeast is believed to inhibit the return of [Ca2+]c to its resting level following stimulus-induced [Ca2+]c increases [37 , 45–47] . In contrast , our data showed that the pmrA deletion in A . nidulans exhibited no significant change in the calcium signature following a high extracellular calcium stress stimulus compared with the wild-type strain , suggesting that other paralogs of pmrA ( e . g . other Ca2+-ATPases ) may compensate or play more important roles in returning the elevated [Ca2+]c back to its resting level . Surprisingly , loss of pmrA alleviated the decreased response of the ΔakrA mutants to the external calcium stimulus , resulting in the amplitude of the [Ca2+]c increase of the double mutant ΔpmrAΔakrA being almost back to the normal level of the wild type . Thus deletion of PmrA reverses the effects of the AkrA deletion in regulating calcium influx following extracellular calcium stress . The lower amplitude of the [Ca2+]c increase of the ΔakrA mutant in response to the high extracellular calcium stimulus indicate that AkrA and its pamitoylated targets play a role in mediating the calcium influx into the cytoplasm and then PmrA may store cytoplasmic calcium into Golgi . When both PmrA and AkrA were absent , the increase in [Ca2+]c following extracellular calcium stimulation was back to almost the normal level in the wild-type ( Fig 5 ) . This suggests that the [Ca2+]c increase in the ΔpmrAΔakrA double mutant following treatment with high extracellular calcium is compensated by some other unknown component ( s ) of the calcium signaling/homeostatic machinery . Furthermore , our data ( Fig 4A ) showed that loss of pmrA suppressed the colony growth defect of ΔakrA mutants , providing further evidence to support interactive regulatory roles of PmrA and AkrA in A . nidulans . Previous studies have verified that exposure of fungi to ER or plasma membrane stress stimulates store-operated calcium influx through the HACS to promote fungal cell survival [13 , 14 , 41 , 50–52] . Consistent with previous studies , in A . nidulans we observed a transient increase in [Ca2+]c after treatment with the ER-stress agents tunicamycin ( TM ) or dithiothreitol ( DTT ) . The ΔcchA mutant exhibited reduced [Ca2+]c amplitudes by 32 ± 6% and 15 ± 9% upon treatment with TM or DTT , respectively ( Figs 6 and S7 ) . In contrast , we did not detect a change in the [Ca2+]c response to the ER stress agents in the ΔmidA mutant compared to its parental wild-type strain . This suggests that as a complex of CchA and MidA , CchA may have a more predominant role than MidA during the ER stress response . Moreover , the ΔakrA mutant displayed a decreased response to ER and plasma membrane stress inducing drugs , as the [Ca2+]c amplitude of ΔakrA mutants decreased by approximately 36–40% of the wild-type strain following treatment with these drugs ( Figs 6 and S7 ) . These data suggest that , in addition to HACS components , AkrA is also involved in ER and plasma membrane stress-induced calcium influx . Moreover , these responses were completely abolished in the ΔakrA mutant but not in the wild-type strain in the presence of EGTA or BAPTA that chelate external calcium . These results indicate that both extracellular calcium and calcium stores contribute to the transient [Ca2+]c changes following ER or plasma membrane stress . Because calcium release from intracellular stores in response to these types of stress was abolished in the akrA mutants ( Figs 6 , 7 and S9 ) , our results are consistent with AkrA regulating calcium influx across the plasma membrane , which in turn activates the release of calcium from intracellular pools . Altogether , our results provide the first report that AkrA is a putative palmitoyl transferase in A . nidulans , and it mediates calcium influx in a DHHC-dependent mechanism to perform an essential function in calcium homeostasis/signaling for survival under high extracellular calcium- , ER- or azole antifungal-stress conditions . Calcium signaling regulators have been previously identified as antifungal target candidates , including FK506 , which targets calcineurin [8] . However , most of the fungal homologs of known calcium signaling components in mammalian cells are of proteins also required for mammalian cell growth and metabolism [68] . Thus , potential antifungals against these components may cause side effects in mammalian hosts . The use of drugs that target regulators of posttranslational modification of calcium signaling that show significant differences to their mammalian homologs ( e . g . AkrA only exhibits 24 . 8% identity to the human AkrA homolog HIP14 ) , may circumvent this problem . The potential for developing novel antifungal drugs of this type has been greatly facilitated by our study that has shown a critical link between palmitoylation and calcium signaling . Previous studies have shown that all AkrA homologs across different species require the DHHC motif to be active and function normally as palmitoyl transferases [69–71] . Three approaches were initially employed to determine AkrA function: deletion of the DHHC motif; site-directed mutagenesis of the cysteine residue in the DHHC motif; and use of a specific palmitoyl transferase analogue inhibitor ( 2-bromopalmitate ) , to determine AkrA function [48 , 49] . Our data from these experiments suggested that the DHHC motif and its cysteine residue are required for the function of AkrA , especially when extracellular calcium is limited . To further test whether the cysteine residue in the DHHC motif , is correspondingly required for AkrA palmitoylation , we used the acyl-biotin exchange ( ABE ) chemistry assay to detect palmitoylation based on selective thioester hydrolysis by hydroxylamine . Compared to the treatment without hydroxylamine , the newly exposed cysteine residues are disulfide-bonded to a biotin analogue , affinity purified and digested into peptides , leaving the labeled peptides on the affinity beads so that palmitoylated proteins have been enriched . As the ABE chemistry detects palmitoylation through identification of all the thioester linkages . A subsequent Western experiment was used to further confirm palmitoylated proteins by specific antibodies . Consequently , among these enriched palmitoylated proteins , Flag-AkrA was clearly detected with an anti-Flag antibody . Site-directed mutation of the cysteine residue in the DHHC or treatment of the parental wild-type strain with the palmitoyl transferase analogue inhibitor 2-BP completely abolished palmitoylation of AkrA ( Fig 8B ) . Previous studies have demonstrated that although the exact mechanism of S-acylation is not known , palmitoylation of the purified DHHC-CRD palmitoylated proteins zDHHC2 , zDHHC3 and yeast Erf2 , involves a two-step mechanism , in which the zDHHCs form an acyl-enzyme intermediate ( auto-acylation ) , with the acyl group later transferred to the target protein [53 , 54] . Our results indicated that AkrA auto-acylated itself before palmitoylating its target proteins . In mammalian cells , any protein that contains a surface-exposed and freely accessible cysteine that has transient access to Golgi membranes is susceptible to palmitoylation . Our data suggests AkrA both auto-acylated itself and palmitoylates target proteins in association with Golgi membranes . Moreover , we found that site directed mutagenesis of the Cys487 in the DHHC motif significantly affect normal localization of AkrA in the Golgi . When we treated cells with a specific palmitoyl transferase analogue inhibitor 2-BP , AkrA localization within the Golgi localization was completely lost ( Fig 8D ) , suggesting that the 2-BP treatment not only prevented AkrA auto-acyltation but also prevented the normal subcellular localization of AkrA . The reason for the different localization pattern , if any , caused by the site directed mutagenesis and the treatment of 2-BP as shown in Fig 8D is likely to be due to a side effect of the 2-BP reagent . In conclusion , our results provide the first report that AkrA is a palmitoyl transferase in A . nidulans , and that it mediates calcium influx in a DHHC-dependent mechanism to perform an essential role in calcium homeostasis to survive high extracellular calcium- , ER- and plasma membrane-stress conditions . A working model of AkrA function in regulating [Ca2+]c homeostasis in A . nidulans is presented in Fig 9 . Our findings provide new insights into the link between palmitoylation and calcium signaling that may be of relevance for understanding the mechanistic basis of human PAT-related diseases . Regulators of posttranslational modification in fungi may provide promising targets for new therapies against life threatening fungal diseases .
All fungal strains used in this study are listed in S1 Table . Minimal media ( MM ) , and MMPDR ( minimal media + glucose + pyrodoxine + riboflavin ) , MMPDR+UU ( minimal media + glucose + pyrodoxine + riboflavin+ uridine + uracil ) , MMPGR ( minimal media + glycerol + pyrodoxine + riboflavin ) have been described previously [29 , 72] . MMPGRT was MMPGR with 100 mM threonine . Fungal strains were grown on minimal media at 37°C , harvested using sterile H2O and stored for the long-term in 50% glycerol at −80°C . Expression of tagged genes under the control of the alcA promoter was regulated by different carbon sources: non-induced by glucose , induced by glycerol and overexpressed by glycerol with threonine . Growth conditions , crosses and induction conditions for alcA ( p ) -driven expression were as previously described [73] . In order to generate constructs for akrA null mutant ( ΔakrA ) , the fusion PCR method was used as previously described [74] . Primers used to design constructs are listed in S2 Table . The A . fumigatus pyrG gene in plasmid pXDRFP4 was used as a selectable nutritional marker for fungal transformation . The transformation was performed as previously described [75] . For creating an ΔakrA construct , a 5′ flank and a 3′ flank DNA fragments were amplified using the primers akrA-P1 and akrA-P3 , akrA-P4 and akrA-P6 , respectively , using genomic DNA ( gDNA ) of the A . nidulans wild-type strain TN02A7 as the template for PCR . As a selectable marker , a 2 . 8 kb DNA fragment of A . fumigatus pyrG was amplified from the plasmid pXDRFP4 using the primers pyrG-5’ and pyrG-3’ . The three PCR products were combined and used as a template to generate a 4 . 8 kb DNA fragment using the primers akrA-P2 and akrA-P5 . The final PCR product was transformed into a wild-type strain . A similar strategy was used to construct akrA-truncated mutants . To design the revertant strain construct , a 3 . 7 kb DNA fragment , which included a 1 . 2 kb promoter region , a 2 . 4 kb coding sequence , and a 3′ flank was amplified using the primers primer A and primer D from A . nidulans gDNA . As a selectable marker , a 1 . 7 kb pyroA fragment was amplified from the plasmid pQa-pyroA using the primers pyro-5’ and pyro-3’ . The two PCR products were co-transformed into the ΔakrA strain to produce the revertant strain . To generate the alcA ( p ) ::GFP-akrA vector , a 1 kb akrA fragment was amplified from the gDNA in the wild-type strain TN02A7 with primers akrA-5’ and akrA-3’ ( S2 Table ) and then ligated into the plasmid vector pLB01 yielding plasmid pLB-alcA ( p ) ::GFP-akrA which contains GFP-N under the control of the alcA promoter with the N . crassa pyr4 as a marker . For site-directed mutation , a 3 . 7 kb akrA DNA fragment with a site directed mutation in which cysteine487 was replaced by serine and a selective marker pyroA were co-transformed into the ΔakrA strain to obtain native ( p ) ::akrAC487S strain . The fragment containing the site mutation was amplified with two steps . First , fragment AB and fragment CD were amplified from A . nidulans gDNA with primers A and B , primers C and D , respectively , and complementary regions contained the desired mutation ( cysteine487 to serine487 ) . Second , using fragment AB and fragment CD as a template , the final 3 . 7 kb fragment was generated through fusion PCR using primer A and primer D . The GPD ( p ) ::akrAC487S and alcA ( p ) ::GFP-akrAC487S strains were constructed using a similar strategy . In brief , the GPD promoter was amplified with the GPD-5’ and GPD-3’ , and 2 . 4 kb akrA DNA fragment including a 2 . 4 kb coding sequence , and a 0 . 5 kb 3’ flanking was amplified with akrA-GPD-5’ and primer D . These two fragments were combined using GPD-5’ and primer D , Lastly , the aboved fusion PCR products and the selective marker pyroA were co-transformed into the ΔakrA strain to obtain the GPD ( p ) ::akrAC487S strian . For the alcA ( p ) ::GFP-akrAC487S construction , a 5′ flank and a 3′ flank DNA fragments were amplified from genomic DNA of alc-akrA mutant using the primers alc-up and primer B , primer C and new primer D , respectively . Then the two PCR products were combined and used as a template to generate a 3 . 9 kb DNA fragment using the primers alc-up and new primer D , and then this fragment was ligated into a plasmid vector yielding the pEA-C487S . The pyroA fragment was amplified from the pQa-pyroA using the primers pyro-cre-5’ and pyro-cre-3’ , then recombined into the plasmid pEA-C487S . Finally the plasmid was transformed into the ΔakrA strain to obtain the alcA ( p ) ::akrAC487S strian . All N-terminal Flag constructs were designed and fabricated using restriction-free cloning protocols outlined at http://www . rf-cloning . com using PrimerSTAR MAX DNA polymerase ( TAKARA , R045A ) [76] . Then , N-Flag tagged cassettes and selective marker pyroA were co-transformed into the ΔakrA strain . For the mutants expressing the codon-optimized aequorin , the plasmid pAEQS1-15 containing codon-optimized aequorin and selective markers pyroA or riboB genes were co-transformed into the indicated mutants . Transformants were screened for aequorin expression using methods described previously [77] and high aequorin expressing strains were selected after homokaryon purification involving repeated plating of single conidia . For each experiment , at least three replicate plates were used to test phenotypes for each strain . To assess the influence by the extracellular calcium to the colony phenotype , minimal medium was supplemented with 20 mM CaCl2 or 1 mM EGTA , respectively . The influence of osmotic stress or ionic stress was tested by adding 600 mM NaCl , 600 mM KCl , 10 mM MnCl2 , 400 mM MgCl2 , 400 mM CaCl2 or 300 mM LiCl into minimal medium , respectively . For the cell wall integrity test , the reagent of 60 μg/mL Calcofluor White or 100 μg/mL Congo Red was added to the minimal medium , respectively . 2 μL of conidia from the stock ( 1×106 conidia/mL ) for indicated strains were spotted onto relevant media and cultured for 2 . 5 days , at 37°C , and then the colonies were observed and imaged . For microscopic observations , conidia were inoculated onto pre-cleaned glass coverslips overlaid with liquid media . To observe co-localization of GFP-AkrA and mRFP-PHOSBP , strain ZYA13 ( S1 Table ) was cultured at 37°C for 10 h in non-inducing medium ( non-inducing conditions for the alcA ( p ) driving expression of AkrA ) and shifted for 6 h to the inducing medium ( in which the alcA promoter was induced ) before microscopic observation [34] . Differential interference contrast ( DIC ) and fluorescence images of the cells were captured with a Zeiss Axio imager A1 microscope ( Zeiss , Jena , Germany ) equipped with a Sensicam QE cooled digital camera system ( Cooke Corporation , Germany ) . The images were processed with MetaMorph/MetaFluor software ( Universal Imaging , West Chester , PA ) and assembled in Adobe Photoshop ( Adobe , San Jose , CA ) . Germination was assessed in liquid non-inducing medium at 37°C with a total number of 106 conidia/mL for each strain in their stationary phase [78] . The percentage rate of germination was measured at 4 , 5 , 6 , 7 and 8 h by microscopic examination . Spores were considered as germinated ones when length of the germ tube was almost equal to the conidium in diameter . For each strain , three replicates of 100 spores were quantified at each time point to determine the germination rate . Saccharomyces cerevisiae strain AH109 ( Clontech , Palo Alto , CA ) was used as the host for the two-hybrid interaction experiments . The analysis was performed using the Matchmaker Library Construction & Screening system ( BD Clontech ) . For strain generation , a cDNA fragment corresponding to the cytosol C-terminus of cchA and the full-length cDNA of midA were amplified and cloned into the pGADT7 vector , which contains the GAL4 DNA-AD and the LEU2 marker ( BD Clontech ) . Full-length cDNA of akrA were used for the pGBKT7 vector ( Clotech , Palo Alto , CA ) . The strains expressing the codon-optimized aequorin gene were grown on minimal media for 2 . 5 days to achieve maximal conidiation . 106 spores with liquid media were distributed to each well of a 96-well microtiter plate ( Thermo Fischer , United Kingdom ) . Six wells were used in parallel for each treatment . The plates were incubated at 37°C for 18 h . The medium was then removed and the cells in each well were washed twice with PGM ( 20 mM PIPES pH 6 . 7 , 50 mM glucose , 1 mM MgCl2 ) . Aequorin was reconstituted by incubating mycelia in 100 μL PGM containing 2 . 5 μM coelenterazine f ( Sigma-Aldrich ) for 4 h , at 4°C in the dark . After aequorin consititution , mycelia were washed twice with 1 mL PGM and allowed to recover to room temperature for 1 h [79 , 80] . To chelate extracellular Ca2+ , 1 mM EGTA or 8 mM BAPTA was added to each well 10 min prior to stimulus injection . At the end of each experiment , the active aequorin was completely discharged by permeabilizing the cells with 20% ( vol/vol ) ethanol in the presence of an excess of calcium ( 3 M CaCl2 ) to determine the total aequorin luminescence of each culture . Luminescence was measured with an LB 96P Microlumat Luminometer ( Berthold Technologies , Germany ) , which was controlled by a dedicated computer running the Microsoft Windows-based Berthold WinGlow software . Conversion of luminescence ( relative light units [RLU] ) into [Ca2+]c was done using Excel 2007 software ( Microsoft ) . The relative light units ( RLU ) values were converted into [Ca2+]c concentrations by using the following empirically derived calibration formula: pCa = 0 . 332588 ( -log k ) + 5 . 5593 , where k is luminescence ( in RLU ) s-1/total luminescence ( in RLU ) [77] . Error bars represent the standard error of the mean of six independent experiments , and percentages in the figures represent peak of [Ca2+]c compared to that of the wild-type ( 100% ) . ABE was performed as described previously with some modifications [81] . Briefly , the strain mycelium was ground to a fine powder in liquid nitrogen and resuspended in 5 mL lysis buffer . Samples were incubated for 1 h at 4°C followed by centrifugation at 4°C , 13 , 000 g to remove insoluble material . 5 mg of protein was incubated overnight with 50 mM N-ethylmaleimide ( NEM ) at 4°C to reduce proteolysis while allowing free sulhydryls to be blocked . Proteins were precipitated at room temperature using methanol/chloroform . The pellet was resuspended in 200 μL resuspension buffer and the solution divided into two equal aliquots . One aliquot was combined with 800 μL of 1 M fresh hydroxylamine ( HA ) , 1 mM EDTA , protease inhibitors and 100 μL 4 mM biotin-HPDP ( Thermo Scientific ) . As a control the remaining aliquot was treated identically but hydroxylamine ( HA ) was replaced with 50 mM Tris pH 7 . 4 . Proteins were precipitated and resuspended in 100 μL of resuspension buffer . 900 μL PBS containing 0 . 2% Triton X-100 was added to each sample , aliquots were removed as a loading control , and the remaining reactions were incubated with 30 μL of streptavidin-agarose beads ( Thermo scientific ) . The streptavidin beads were washed four times with 1 mL PBS containing 0 . 5 M NaCl and 0 . 1% SDS . Proteins were eluted by heating at 95°C in 40 μL 2× SDS sample buffer containing 1% 2-mercaptoethanol v/v . Samples were analyzed by silver staining or Western blotting as described below . In some cases , cells were treated with 50 or 100 μM of the palmitoylation inhibitor 2-bromopalmitate ( 2-BP ) before the ABE assay . For mass spectrometry ( MS ) , total protein ( 100 μg ) extracted from each sample was chemically reduced for 1 h at 60°C by adding DTT to 10 mM and carboxyamidomethylated in 55 mM iodoacetamide for 45 min at room temperature in the dark . Then trypsin gold ( Promega , Madison , WI , USA ) was added to give a final substrate/enzyme ratio of 30:1 ( w/w ) . The trypsin digest was incubated at 37°C for 16 h . After digestion , the peptide mixture was acidified by 10 μL of formic acid for further MS analysis . After protein digestion , each peptide sample was desalted using a Strata X column ( Phenomenex ) , vacuum-dried and then resuspended in a 200 μL volume of buffer A ( 2% ACN , 0 . 1% FA ) . After centrifugation at 20000 g for 10 min , the supernatant was recovered to obtain a peptide solution with a final concentration of approximately 0 . 5 μg/μL . 10 μL supernatant was loaded on a LC-20AD nano-HPLC ( Shimadzu , Kyoto , Japan ) by the autosampler onto a 2 cm C18 trap column . The peptides were then eluted onto a 10 cm analytical C18 column ( inner diameter 75 μm ) packed in-house . The samples were loaded at 8 μL/min for 4 min , then the 35 min gradient was run at 300 nL/min starting from 2 to 35% buffer B ( 95% ACN , 0 . 1% FA ) , followed by a 5 min linear gradient to 60% , then followed by a 2 min linear gradient to 80% , and maintenance at 80% buffer B for 4 min , and finally returned to 5% in 1 min . Data acquisition was performed with a TripleTOF 5600 System ( AB SCIEX , Concord , ON ) fitted with a Nanospray III source ( AB SCIEX , Concord , ON ) and a pulled quartz tip as the emitter ( New Objectives , Woburn , MA ) . Data was acquired using an ion spray voltage of 2 . 5 kV , curtain gas of 30 psi , nebulizer gas of 15 psi , and an interface heater temperature of 150 . The MS was operated with a RP of greater than or equal to 30 , 000 FWHM for TOF MS scans . Raw data files acquired from the Orbitrap were converted into MGF files using Proteome Discoverer 1 . 2 ( PD 1 . 2 , Thermo ) , [5 , 600 msconverter] and the MGF file were searched . Protein identification was performed by using Mascot search engine ( Matrix Science , London , UK; version 2 . 3 . 02 ) against a database containing 13 , 597 sequences . To extract proteins from A . nidulans mycelia , conidia from alcA ( p ) ::GFP-akrA and the wild-type strains were inoculated in the liquid inducing medium , then shaken at 220 rpm on a rotary shaker at 37°C for 24 h . The mycelium was ground in liquid nitrogen with a mortar and pestle and suspended in ice-cold extraction buffer ( 50 mM HEPES pH 7 . 4 , 137 mM KCl , 10% glycerol containing , 1 mM EDTA , 1 μg/mL pepstatin A , 1 μg/mL leupeptin , 1 mM PMSF ) . Equal amounts of protein ( 40 μg ) per lane were subjected to 10% SDS–PAGE , transferred to PVDF membrane ( Immobilon-P , Millipore ) in 384 mM glycine , 50 mM Tris ( pH 8 . 4 ) , 20% methanol at 250 mA for 1 . 5 h , and the membrane was then blocked with PBS , 5% milk , 0 . 1% Tween 20 . Next , the membrane was then probed sequentially with 1:3000 dilutions of the primary antibodies anti-GFP or anti-FLAG or anti-actin and goat anti-rabbit IgG-horseradish peroxidase diluted in PBS , 5% milk , 0 . 1% Tween 20 . Blots were developed using the Clarity ECL Western blotting detection reagents ( Bio-Rad ) , and images were acquired with the Tanon 4200 Chemiluminescent Imaging System ( Tanon ) . The mycelia were cultured for 18 h in liquid media and were then ground to a fine powder in liquid nitrogen . Total RNA was isolated using Trizol ( Invitrogen , 15596–025 ) following the manufacturer’s instructions . 100 mg of mycelia per sample was used as the starting material for the determination of total RNA . The reverse transcription polymerase chain reaction ( RT-PCR ) was carried out using HiScript Q RT SuperMix ( Vazyme , R123-01 ) , and then cDNA was used for the real-time analysis . For real-time reverse transcription quantitative PCR ( RT-qPCR ) , independent assays were performed using SYBR Premix Ex Taq ( TaKaRa , DRR041A ) with three biological replicates , and expression levels normalized to the mRNA level of actin . The 2-ΔCT method was used to determine the change in expression .
|
Palmitoylation is a reversible post-translational modification catalyzed by palmitoyl acyltransferases ( PATs ) and proteins that undergo this modification are involved in numerous intracellular functions . Yeast Akr1p was the first characterized PAT whilst HIP14 , an Akr1p homolog in human , is one of the most highly conserved of 23 human PATs that catalyze the addition of palmitate to the Huntington protein which is of major importance in Huntington’s disease . Calcium serves numerous signaling and structural functions in all eukaryotes . However , studies on the relationship between calcium signaling and palmitoylation are lacking . In this study , we demonstrate that the palmitoyl transferase Akr1 homolog in the filamentous fungus Aspergillus nidulans , similar to the high-affinity calcium uptake system ( HACS ) , is required for normal growth and sporulation in the presence of low extracellular calcium . We find that AkrA dysfunction decreases the transient increase in cytosolic free calcium induced by a high extracellular calcium stress , tunicamycin ( which induces endoplasmic reticulum stress ) or the antifungal agent itraconazole ( which induces plasma membrane stress ) . The influence of AkrA on all of these processes involves its DHHC motif , which is required for palmitoylation of various proteins associated with many processes including calcium signaling and membrane trafficking . Our findings provide evidence for a crucial link between calcium signaling and palmitoylation , suggesting a possible role in the mechanistic basis of human PAT-related diseases . These results also indicate that regulators of posttranslational modification may provide promising antifungal targets for new therapies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"aspergillus",
"organic",
"compounds",
"membrane",
"proteins",
"aspergillus",
"nidulans",
"physiological",
"processes",
"fungi",
"model",
"organisms",
"homeostasis",
"amino",
"acids",
"calcium",
"signaling",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"cysteine",
"proteins",
"chemistry",
"molds",
"(fungi)",
"metabolism",
"cell",
"membranes",
"sulfur",
"containing",
"amino",
"acids",
"biochemistry",
"signal",
"transduction",
"palmitoylation",
"cell",
"biology",
"post-translational",
"modification",
"organic",
"chemistry",
"physiology",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"physical",
"sciences",
"cell",
"signaling",
"organisms",
"bone",
"and",
"mineral",
"metabolism"
] |
2016
|
Palmitoylation of the Cysteine Residue in the DHHC Motif of a Palmitoyl Transferase Mediates Ca2+ Homeostasis in Aspergillus
|
Understanding the control of epigenetic regulation is key to explain and modify the aging process . Because histone-modifying enzymes are sensitive to shifts in availability of cofactors ( e . g . metabolites ) , cellular epigenetic states may be tied to changing conditions associated with cofactor variability . The aim of this study is to analyse the relationships between cofactor fluctuations , epigenetic landscapes , and cell state transitions . Using Approximate Bayesian Computation , we generate an ensemble of epigenetic regulation ( ER ) systems whose heterogeneity reflects variability in cofactor pools used by histone modifiers . The heterogeneity of epigenetic metabolites , which operates as regulator of the kinetic parameters promoting/preventing histone modifications , stochastically drives phenotypic variability . The ensemble of ER configurations reveals the occurrence of distinct epi-states within the ensemble . Whereas resilient states maintain large epigenetic barriers refractory to reprogramming cellular identity , plastic states lower these barriers , and increase the sensitivity to reprogramming . Moreover , fine-tuning of cofactor levels redirects plastic epigenetic states to re-enter epigenetic resilience , and vice versa . Our ensemble model agrees with a model of metabolism-responsive loss of epigenetic resilience as a cellular aging mechanism . Our findings support the notion that cellular aging , and its reversal , might result from stochastic translation of metabolic inputs into resilient/plastic cell states via ER systems .
Aging is associated with profound changes in the epigenome involving large disturbances of the epigenetic landscape and genome architecture [1 , 2] . Studies in model organisms have not only revealed the complex changes occurring in chromatin structure and functioning during aging , but also the remarkable plasticity of age-associated epigenetic marks [3–5] . Thus , whereas epigenetic alterations in DNA methylation , post-translational modification ( PTM ) of histones and chromatin remodelling are considered highly conserved hallmarks of aging [4 , 6] , the ability of cellular reprogramming-driven epigenetic remodelling to ameliorate age-associated phenotypes has been described recently . This finding unequivocally supports the causative role of epigenetic dysregulation as a driver of aging [7] . The reversible nature of epigenetic regulation of aging is receiving increasing attention as it might offer a revolutionary strategy to simultaneously delay or reverse a spectrum of diseases , including cancer , clustered in older individuals [8 , 9] . A mechanistic understanding of the dependence and inter-relationship between aging and the functional status of specific epigenetic modifiers , for example histone demethylases ( HDMs ) and histone deacetylases ( HDACs ) , is largely lacking . There is an increasing awareness of the relationship between epigenetic modifiers and metabolism . Common metabolites of intermediary metabolism , such as acetyl-CoA , NAD+ , α-ketoglutarate , succinate , FAD , ATP or S-adenosylmethionine , drive epigenetic processes by directly regulating epigenetic modifiers . The usage of these intermediates as substrates and regulators of chromatin-modifying enzymes provides a direct link between the metabolic state of the cell and epigenetics [10–17] . However , it remains intriguing how aging-related changes in cellular metabolism ( e . g . , loss of NAD homeostasis [18–20] ) might control the layers of epigenetic instructions that influence cell fate without involving changes in the DNA sequence . The capacity of the chromatin structure to affect cellular identity and cellular state transitions can differ as a function of metabolic conditions that change during aging . However , the possibility that cellular aging might result from the stochastic translation of metabolic signals into cellular epigenetic states has not been formally evaluated . In this paper , we explore the causative relationship between cofactor ( e . g . metabolite ) variability and chromatin modification state underpinning the aging-associated loss of epigenetic resilience , which leads to a gain of more plastic cell and tissue features . This fact might predispose aging tissues to cancer [21 , 22] . To this end , we generated an ensemble of epigenetic regulation ( ER ) systems by means of Approximate Bayesian Computation ( ABC ) whose heterogeneity reflects the inhomogeneous abundance of cofactors used by epigenetic modifiers . By analysing the robustness of ER systems in response to the regulation of HDM and HDAC activity , we present a model of ER capable of formulating strategies aimed at modifying the aging process and the aging-dependency of cancer , based on the control of epigenetic resilience and plasticity . Recent advances in experimental determination of the mechanisms of ER have triggered an interest in developing mathematical models capable of reducing their intrinsic complexity to essential components such as ER of gene expression [17 , 23–27] and epigenetic memory [24 , 25 , 27–32] . For comprehensive reviews , we refer the readers to [25 , 27] . In order to put our model into context , we briefly summarise the current state of the art in ER modelling . Models of ER were originally formulated in order to shed light onto the mechanisms of epigenetic memory; since DNA during cell cycle is duplicated and , therefore , the epigenetic marks diluted , early ER models were aimed at explaining how epigenetic-regulatory states remain stable upon cell division and transmitted to daughter cells . Such models must satisfy two essential properties , namely , they must be bistable , i . e . , each steady state corresponding to an alternative epigenetic state , and the basin of attraction of such states must allow that large perturbations of the ER systems undergoing DNA replication should not change the epigenetic state thus allowing mitotic heritability [29] . Dodd et al . [28] developed the first of such ER models . The authors considered a region of DNA consisting of N nucleosomes , each assumed to be in either of three states , namely unmodified ( U ) , methylated ( M ) , and acetylated ( A ) . Because modifying and de-modifying enzymes carry out nucleosome modifications and removal of marks , a crucial ingredient of the model by Dodd et al . [28] is that histone-modifying enzymes are recruited by modified nucleosomes , thereby providing the necessary positive feed-back for the system to be bistable . However , recruitment based on next-neighbours interactions is not enough to produce robust bistability . Long-range correlations are necessary . The model by Dodd et al . [28] has been modified and extended in several ways [31] . Sneppen and Dodd have successfully applied the same ideas [32] to modelling the patterns of epigenetic regulation in CpG islands [33] . Another interesting feature of the model developed by Sneppen and Dodd [31] is that medium-length correlations are provided by the size of nucleosomes , which allows relaxing the requirement for recruited demethylation . Angel et al . [30] have proposed an ER model to explain quantitative epigenetic control associated with the phenomenon of vernalisation , i . e . the perception and epigenetic memory of a period of cold temperatures to initiate flowering later . This model is capable of reproducing both the patterns of flowering locus C ( FLC ) and the quantitative dependence with respect to the duration of the exposition to low temperatures . Besides the issue of maintaining stable epigenetic memory , recent efforts have been dedicated to the study of the regulation of epigenetic modifications by transcription factors [23 , 26] . Based on the experimental observation that transcription factors ( TFs ) can recruit histone-modifying enzymes , Sneppen et al . [23] proposed a model where transcription factors are coupled to ER . A similar approach , although with rather significant differences , has been recently proposed by Berry et al . [26] . An essential feature of this model is the proposed feedback between transcription and epigenetic chromatin modification: activation of transcription depends on the balance between positive and negative modifications , and , in turn , each passage of RNA polymerase II , which is modelled as a discrete event , causes demethylation ( see [26] for details ) . An important feature that distinguishes this model from its predecessors is the assumption of next-neighbour recruitment as exclusively opposed to long-distance recruitment . Bintu et al . [24] have recently proposed a more phenomenological ER model capable of explaining experimental data obtained by using a reporter gene that expresses a fluorescent protein with induced recruitment of a number of epigenetic-modifying enzymes . The model by Bintu et al . [24] considers active , reversible silent , and irreversible silent states and is able to predict the rates of transition between states .
The stochastic model of epigenetic regulation is formulated in terms of the associated Chemical Master Equation ( CME ) , which , in general , is given by: ∂ P ( X , t ) ∂ t = ∑ i ( W i ( X - r i ) P ( X - r i , t ) - W i ( X ) P ( X , t ) ) ( 1 ) where X = ( X1 , … , Xn ) is the vector containing the number of molecules of each molecular species at time t , Wi ( X ) is the transition rate corresponding to reaction channel i and ri is a vector whose entries denote the change in the number of molecules of each molecular species when reaction channel i fires up , i . e . P ( X ( t + Δt ) = X ( t ) + ri|X ( t ) ) = Wi ( X ) Δt . Our model ( see Table 1 ) is based on the stochastic models by Dodd et al . [28] and Menéndez et al . [34] . Dodd et al . [28] consider that direct transitions between M and A are very unlikely . Instead , they assume that transitions occur in a linear sequence given by M ⇌ U ⇌ A . They further put forward the hypothesis that such nucleosome modifications are of two types , namely , recruited and unrecruited . Mathematically , recruited modifications are represented by non-linear dependence on the number of M-nucleosomes and A-nucleosomes of the corresponding transition rates ( see Table 1 ) . Specifically , the reactions involved in our model are: All these reactions can be both recruited or unrecruited . The associated reactions rates are reported in Table 1 . We consider the scenario where both hyper- ( hypo- ) abundance of A ( M ) marks allows for genes to be expressed , insofar the associated transcription factors are present [10] . On the contrary , we associate hypo- ( hyper- ) abundance of A ( M ) marks with silent states where genes are not expressed even in the presence of the appropriate transcription factors . We here focus on the conditions for bistability to arise and the robustness of the associated open and closed states particularly in connection with the abundance or activity of HDMs and HDACs . Our aim is to analyse the effects of varying the concentration of these enzymes as well as possible synergies between them . In more detail , we focus our analysis on plastic behaviour of the epigenetic regulatory states when the activity of histone-modifying enzymes ( HMEs ) is down-regulated against the background of heterogeneity due to variability in the pool of cofactors for chromatin-modifying enzymes . We proceed by first defining a base-line scenario ( which we categorise as normal cell ) in which the associated epigenetic regulatory system is such that , for average values of HDM and HDAC activities , the differentiation-promoting gene ER is open and the pluripotency-promoting gene ER is closed . We then proceed to generate an ensemble of ER systems that satisfy the requirements imposed by this base-line scenario; the necessary variability to generate this ensemble is provided by heterogeneity in abundance of epigenetic cofactors . Analysis of this ensemble reveals that the requirements of the base line scenario restrict the values of a few parameters only , leaving ample flexibility to fix the rest of them . This behaviour is typical of the so-called sloppy models [35] , where available data constrains a limited number of parameters ( or parameter combinations ) , the system being robust to the choice of a large number of model parameters . In our case , this feature is absolutely essential since , nested within this heterogeneous ensemble of ER systems , there exists a sub-ensemble of plastic ER systems . In order to gain some insight into the behaviour of the stochastic ER model , we analyse its mean-field limit regarding time scale separation and the quasi-steady state approximation . For a full account of the technicalities we refer the reader to our previous work [36 , 37] . The mean-field equations , which describe the time evolution of the ensemble average of the variables Xi , associated to the stochastic system with rates given in Table 1 are: d Q i d t = ∑ j = 1 16 r j , i W j ( Q ) ( 2 ) where Q is a vector whose entries , Qi , are Qi ≡ 〈Xi〉 . In order to proceed further , we assume that the variables describing the system are divided into two groups according to their characteristic scales . More specifically , we consider the situation where the subset of chemical species Xi , with i = 1 , 2 , 3 , scale as Xi = Sxi , where xi = O ( 1 ) , whilst the remaining species are such that Xi , with i = 4 , 5 , 6 , 7 , scale as Xi = Exi , where xi = O ( 1 ) . Key to our approach is the further assumption that S and E must be such that ϵ = E S ⪡ 1 . The averaged variables , Qi , are similarly divided into two groups: slow variables , i . e . Qi = Sqi ( i = 1 , 2 , 3 ) , and fast variables , i . e . Qi = Eqi ( i = 4 , 5 , 6 , 7 ) . Under this rescaling , we define the following scale transformation for the transition rates in Table 1: Wj ( Q ) = k4S2Eωj ( q ) . We further rescale the time variable so that a dimensionless variable , τ , is defined as τ = k4SEt . It is now straightforward to verify that , upon rescaling , the mean-field equations become: d q i d τ = ∑ j = 1 16 r j , i ω j ( q ) , i = 1 , 2 , 3 , ( 3 ) ϵ d q i d τ = ∑ j = 1 16 r j , i ω j ( q ) , i = 4 , 5 , 6 , 7 . ( 4 ) with ϵ = E/S . If ϵ = E/S ≪ 1 holds , Eqs ( 3 ) and ( 4 ) naturally display multiple scales structure , which we will exploit to simplify our analysis by means of a quasi-steady state approximation ( QSSA ) [38] , which is given by: d q 1 d τ = e H D M ( κ 1 + q 3 ) ( κ 3 + κ 6 q 3 ) q 2 ( κ 2 + κ 3 ) + ( κ 1 + q 3 ) q 2 + ( κ 5 + κ 6 ) q 3 + e H D A C ( κ 9 + κ 12 q 2 ) ( κ 11 + κ 14 q 2 ) q 3 ( κ 10 + κ 11 ) + ( κ 9 + κ 12 q 2 ) q 3 + ( κ 13 + κ 14 ) q 2 - ( κ 8 q 2 + κ 7 + κ 16 q 3 + κ 15 ) q 1 ( 5 ) d q 2 d τ = - e H D M ( κ 1 + q 3 ) ( κ 3 + κ 6 q 3 ) q 2 ( κ 2 + κ 3 ) + ( κ 1 + q 3 ) q 2 + ( κ 5 + κ 6 ) q 3 + ( κ 8 q 2 + κ 7 ) q 1 ( 6 ) d q 3 d τ = - e H D A C ( κ 9 + κ 12 q 2 ) ( κ 11 + κ 14 q 2 ) q 3 ( κ 10 + κ 11 ) + ( κ 9 + κ 12 q 2 ) q 3 + ( κ 13 + κ 14 ) q 2 + ( κ 16 q 3 + κ 15 ) q 1 ( 7 ) q 4 = e H D M κ 2 + κ 3 + ( κ 5 + κ 6 ) q 3 ( κ 2 + κ 3 ) + ( κ 1 + q 3 ) q 2 + ( κ 5 + κ 6 ) q 3 ( 8 ) q 5 = e H D M ( κ 1 + q 3 ) q 2 ( κ 2 + κ 3 ) + ( κ 1 + q 3 ) q 2 + ( κ 5 + κ 6 ) q 3 ( 9 ) q 6 = e H D A C κ 10 + κ 11 + ( κ 13 + κ 14 ) q 2 ( κ 10 + κ 11 ) + ( κ 9 + κ 12 q 2 ) q 3 + ( κ 13 + κ 14 ) q 2 ( 10 ) q 7 = e H D A C ( κ 9 + κ 12 q 2 ) q 3 ( κ 10 + κ 11 ) + ( κ 9 + κ 12 q 2 ) q 3 + ( κ 13 + κ 14 ) q 2 ( 11 ) where the re-scaled parameters κj are defined in Table 2 , and the conservation laws q4 ( τ ) + q5 ( τ ) = eHDM and q6 ( τ ) + q7 ( τ ) = eHDAC hold . These conservation laws account for the fact that the total number of enzyme molecules , i . e . the enzyme molecules in their free form and those forming a complex must be constant . Hence , the quantities eHDM and eHDAC are defined as e H D M = z 0 E and e H D A C = v 0 E , respectively , where z0 and v0 are the numbers of HDM and HDAC enzyme molecules , respectively . E is the characteristic scale ( i . e . average ) of abundance of the histone-modifying enzymes which , for simplicity , has been taken to have the same value for both HDMs and HDACs . This result opens interesting avenues to investigate , since both oncometabolic transformation and aging appear to reduce the number of both types of enzymes . Our theory thus allows us in a natural manner to explore the effects of these anomalies on the stability of epigenetic regulatory states .
We first focus on a bifurcation analysis of the mean-field QSSA Eqs ( 5 ) – ( 11 ) , to investigate the qualitative behaviour of the ER system as the relative abundances of HDMs and HDACs are varied . Results are shown in Fig 3 ( a ) and 3 ( b ) . In particular , the phase space of both ER systems obtained by varying the parameters eHDM and eHDAC . Both these diagrams display three differentiated regions: one in which the only stable steady-state is the one associated with a silenced gene , another one in which the only stable steady-state is the corresponding to an open gene , and a third one where the system is bistable . Fig 3 ( a ) is associated with the differentiation-promoting gene , and Fig 3 ( b ) corresponds to the pluripotency-promoting gene ( parameters as per Table A , Table B in S1 File , respectively ) . In order to clarify the three regions ( open , closed and bistable ) displayed in Fig 3 ( a ) , a 3D plot is shown in Fig 4 ( a ) , where the vertical axis shows the level of positive marks ( q3 ) . This plot shows that the system dysplays bistable behaviour: depending on the parameter values eHDM and eHDAC , the system may be both in the open state ( high levels of q3 , top of the plot ) , or in the closed state . Fig 4 ( b ) displays the projection on the xy-plane of the plot shown in Fig 4 ( a ) , where we can clearly identify the three regions described in Fig 3 ( a ) . A more detailed picture of the situation illustrated in Figs 3 ( a ) and 4 is given in Fig 3 ( c ) , which shows the bifurcation diagram where eHDM , i . e . HDM concentration , is taken as the control parameter , whilst keeping eHDAC constant . In particular we show the steady state value of q3 , i . e . the variable with positive marks , as a function of HDM concentration . This allows to distinguish the three regions displayed in Fig 3 ( a ) . We observe , that a decrease in HDM makes the corresponding gene inaccessible to the transcription machinery ( corresponding to the closed region , Fig 3 ( a ) ) . As HDM concentration recovers , the system enters a bistable regime where both the active and silent states coexist ( region marked as bistable in Fig 3 ( a ) ) . Further increase of the demethylase concentration drives the system through a saddle-node bifurcation , beyond which the only stable steady-state is the active state ( region labelled as open in Fig 3 ( a ) ) . It is noteworthy that these results are in agreement with the oncometabolic transformation scenario associated with IDH mutations proposed by Thompson and co-workers [10 , 42] in which downregulation of HDM activity locks differentiation genes into a silenced state which favours reprogramming of the differentiated state of somatic cells into a pluripotent phenotype [17] . The association between IDH mutations and cancer progression has been well established in the case of glioblastomas and acute myelogenous leukaemia [43–46] . In Fig 3 ( e ) , we show the bifurcation diagram associated with fixing eHDM and varying eHDAC . Within the scenario we are considering , i . e . the epigenetic regulation of a differentiation-regulating gene , reduced HDAC concentration recovers the base-line state where the epigenetic regulatory machinery is set to the open state . As HDAC concentration recovers , the system enters a bistable regime in which both the active and silent states coexist . Further increase in HDAC activity locks the system into the close chromatin state so that the gene is silenced . This implies that reduced HDAC activity may help to rescue differentiation-regulating genes from the effects of IDH mutation . Numerical results which verify the predictions of the bifurcation analysis are presented and discussed in Section I in S1 File . We now proceed to analyse in more detail the implications of the bifurcation analysis , regarding robustness of the epigenetic regulatory state . In Fig 3 ( d ) , which shows the phase diagram of both modes of epigenetic regulation ( differentiation- and pluripotency-promoting ) in the same phase space , the region between the solid red line and the dashed blue line represents the part of the phase space where the differentiation genes are open and the pluripotency genes are closed ( region marked as Normal Cell in Fig 3 ( d ) ) . This sub-space is therefore associated with normal , differentiated somatic cells . As we have previously shown [17] , efficient reprogramming requires both closed differentiation genes and open pluripotency genes . Such situation is not viable under the scenario shown in Fig 3 ( d ) because these two conditions cannot hold simultaneously , which we therefore dubb as the refractory scenario . By contrast , Fig 3 ( f ) corresponds to a plastic scenario , where , under appropriate conditions , cells become poised for reprogramming . The main difference with the refractory scenario is the intersection between the bistability regions of both the differentiation regulator and the pluripotency gene . In Fig 3 ( f ) , the regime where both bistability regions overlap is the one between the red solid line and the blue dashed line ( region marked as Rep . in Fig 3 ( f ) ) . Within this region , since both genes are in the bistable epigenetic regulatory regime , it is possible to find the differentiation gene in its closed state and the pluripotency gene in the open state . Such situation makes reprogramming much more likely to occur [17] and therefore we identify this feature of the phase space with plastic behaviour . By driving the ER system into this region by means of down-regulation of both HDM and HDAC activity , cells become epigenetically poised to undergo reprogramming . This is consistent with evidence according to which both oncometabolic transformation ( e . g . IDH mutation leading to down-regulation of JHDM activity [10 , 42] ) and aging ( e . g . down-regulation of SIRT6 [5 , 19 , 47] ) induce loss of HDM and HDAC activity thus facilitating reprogramming . In order to study the robustness of the refractory and plastic scenarios with respect to variations of the model parameters , kj ( see Table 1 ) , we first generate an ensemble of parameter sets θ = ( kj , j = 1 , … , 16 ) compatible with simulated data for the epigenetic regulation systems . Such ensemble is generated using Approximate Bayesian Computation [48] ( for further details see Section III in S1 File ) . Our approach is as follows . For each mode of epigenetic regulation , we have generated simulated data ( denoted as “raw data” in Fig 2 ) using the stochastic simulation algorithm on the model defined by the transition rates Table 1 . This simulated data will play the role of the experimental data , x0 , to which we wish to fit our model . We consider two different data sets x 0 d and x 0 p , corresponding to the differentiation gene ( reaction rates from Table A in S1 File ) and the pluripotency gene ( reaction rates from Table B in S1 File ) , respectively . Each data set consists of 10 realisations and 25 time points per realisation . For each time point , ti , we consider two summary statistics: the mean over realisations , x ¯ ( t i ) , and the associated standard deviation , σ ( ti ) . We then run the ABC rejection sampler method until we reach an ensemble of 10000 parameter sets which fit the simulated data , x0 , within the prescribed tolerances for the mean and standard deviation . Fig 2 ( a ) & 2 ( b ) shows results comparing the reference ( raw simulated ) data to a sub-ensemble average ( full posterior distributions are shown in Fig . C in S1 File , differentiation-promoting gene , and Fig . D in S1 File , pluripotency-promoting gene ) . The above procedure provides us with an ensemble of parameter sets that are compatible with our raw data , i . e . such that they fit the data within the prescribed tolerances . The heterogeneity associated with the variability within this ensemble has a clear biological origin . The rates kj are associated with the activity of the different enzymes that carry out the epigenetic-regulatory modifications ( HDMs , HDACs , as well as , histone methylases ( HMs ) and histone acetylases ( HACs ) ) , so that variation in these parameters can be traced back to heterogeneity in the availability of cofactors , many of them of metabolic origin such as NAD+ , which are necessary for these enzymes to perform their function ( as illustrated in Fig 1 ) . We first consider the differentiation ER system . In particular , we focus on the sub-ensemble of the 400 parameter sets that best fit the raw data . Within such sub-ensemble , we proceed to evaluate the robustness of the different scenarios we study . We consider that a particular scenario is sensitive to a specific parameter , kj , if its distribution is significantly different from the uniform distribution [49] . We first analyse the base-line scenario for the epigenetic regulation of a differentiation-regulated gene , namely , ( i ) when eHDM = eHDAC = 1 , the regulatory system is mono-stable ( only the open chromatin state is stable ) , and ( ii ) for eHDM < 1 , eHDAC < 1 there exists a region of bistability . Out of all the parameter sets of the considered sub-ensemble , only 94 fulfill these requirements . We refer to these as the viable set . The remaining 307 are bistable at eHDM = eHDAC = 1 , and they will be referred to as the non-viable set . In Fig 5 , we present the cumulative frequency distributions ( CFDs ) of each kj within both sets . The rationale for looking into this is that the requirements upon system behaviour associated with both sets should reflect themselves on the corresponding CFDs . Regarding the viable set , we seek to assess which kinetic constants have distributions which deviate in a statistically significant manner from the uniform distribution [49] . Such parameters are deemed to be the essential ones for the ER system to exhibit the behaviour associated with the viable set . We perform this analysis by means of the Kolmogorov-Smirnov ( KS ) test [50 , 51] , which we use to compare our samples with the uniform distribution . According to such analysis , the kinetic constants k1 , k3 , k6 , k7 , k12 , k14 , and k16 are not uniformly distributed ( p-values are reported in Table E in S1 File ) . Nested within the viable set , there are parameter sets which exhibit plastic behaviour , as characterised by a phase diagram as per Fig 3 ( f ) . We thus continue by studying the plastic subset regarding both its frequency within the viable subset and further restrictions imposed on parameter variability . We first check the number of the plastic parameter sets within the viable set relative to the pluripotency-gene ER system defined by Table D in S1 File . Somehow unexpectedly , the plastic scenario is rare , but not exceptional: amongst the 94 parameter sets that we have identified as viable , 10 exhibit plasticity ( see Fig 5 for their CFDs ) . Further restrictions on parametric heterogeneity imposed by the plastic scenario are analysed regarding the variation of the CFDs of kinetic constants when compared to those associated with the whole viable subset . The results of KS analysis performed on the data shown in Fig 5 show that only the distributions of k1 ( associated with recruited demethylation ) , k9 ( unrecruited deacetylation ) , and k14 ( recruited deacetylation ) are significantly modified by the plasticity requirement ( p-values reported in Table G in S1 File ) . From a more mechanistic perspective , we observe that , within the plastic set , the mass of the CFDs of k1 , k9 and k14 is displaced towards the large-value end of their intervals with respect to their behaviour within the full viable set . In other words , k1 , k9 and k14 tend to be larger for plastic ER systems than for non-plastic , viable ER systems . In essence , we observe that ER systems exhibiting plastic behaviour tend to have increased activity in the enzymes performing histone deacetylation . This is consistent with recent evidence that aging decreases histone acetylation and promotes reprograming [5 , 19 , 47] . The same analysis has been conducted regarding the ensemble of parameter values generated using ABC for the pluripotency gene ER system ( full posterior distribution in Fig . D in S1 File ) . The results of this analysis are shown in Fig 6 . Detailed analysis using the KS test of the ensemble viable pluripotency ER systems shows that k3 , k8 , k12 , k14 , k15 , and k16 are significantly constrained by the requirements of such scenario ( i . e . their CDF departs significantly from the uniform distribution , as shown by the p-values from Table F in S1 File ) . We then move on to investigate further restrictions within the plastic set . We observe that only the CDFs associated with k2 and k6 are significantly different ( p-values reported in Table H in S1 File ) . In both cases , values of k2 and k6 associated with plasticity are larger than in the general viable population . Both parameters are associated with demethylation activity . Our ensemble analysis thus provides a rationale for the coupling between variations in the size of the pool of epigenetic cofactors and increased reprogramming in a heterogeneous cell population . A notable case in point is provided by metabolic changes during aging: those cells where key metabolites such as acetyl-CoA and NAD+ are less abundant lose acetylation capability ( in our model , this is reflected through the dependence of histone-modifying enzyme activity on the concentration of these cofactors ) , leading to cells poised for reprogramming . This analysis provides a rationale for a strategy to interfere with the epigenetic regulatory system , regarding the ability to either drive the system away from plastic behaviour or to drive it to the plasticity scenario , while keeping it functional ( i . e . within the restrictions of the base-line scenario ) . An example illustrating the effectiveness of this strategy is shown in Fig 7 . Consider the viable set of the ER differentiation-promoting gene , Fig 5 , which is neutral with respect to the value of k9: k9 remains uniformly distributed within the viable subset . By contrast , when plasticity is required , the admissible values of k9 accumulate mostly towards the large-value end . This suggests that decreasing the value of k9 might be a viable strategy to restore resilience . To check this , we consider the parameter set , θ = kj/k4 , j = 1 , … , 16 , that gives rise to the plastic behaviour depicted in Fig 3 ( f ) ( Table C in S1 File , for the differentiation-promoting gene ) . We then analyse the effect of modifying the value of k9 for the differentiation-promoting gene on system behaviour . The new parameter set , θ ′ = k j ′ / k 4 , j = 1 , … , 16 , is such that k 9 ′ = k 9 / 4 and k j ′ = k j for all j ≠ 9 ( kj values as per Table C in S1 File ) . Parameter values for the pluripotency gene remain unchanged ( as per Table D in S1 File ) . The corresponding phase space is shown in Fig 7 ( a ) . We observe that by reducing deacetylase activity in this fashion , the ER system reverts to resilient behaviour . This suggests that , by regulating the abundance of cofactors associated with ( de ) acetylation , we can drive the system off the plastic regime into the base-line behaviour . Similarly , we can seek for complex , combined strategies to increase the robustness of plastic behaviour . An example of such strategy is shown in Fig 7 ( b ) . Based on the results of the KS test for the differentiation-promoting gene , we observe that deacetylation-related rates k9 and k14 are significantly increased in plastic scenarios . Taking parameter sets from a resilient scenario ( Tables A & D in S1 File , which lead to a combined phase diagram qualititatively similar to that shown in Fig 3 ( d ) ) and modifying k9 and k14 for the differentiation-promoting gene so that k 9 ′ = 3 k 9 and k 14 ′ = 3 k 14 while keeping all the others at the same value , the resulting ER system corresponds to a plastic system . Futhermore , this combined strategy results in more robust plasticity ( as compared to e . g . the case shown in Fig 3 ( f ) ) , as measured by the area of the phase space region where reprogramming is feasible . This indicates that by combining the strategies suggested by the statistical analysis of the plastic sub-ensemble , we can find conditions for optimal conditions to achieve robust reprogramming . This , in turn , highlights the importance of cofactor levels , since as it has been shown in Fig 7 , depending on its availability , the same ER system can be driven to the plastic or resilient state . These strategies require close attention to be payed to the correlations between parameters . Parameters in complex systems biology models exhibit strong correlations which confer the system with essential properties such as sloppiness , which refers to the property exhibited by many multi-parameter systems biology models , whereby the system’s behaviour is insensitive to changes in parameter values except along a small number of parameter combinations [35] . In order to quantify such correlations , we have used hierarchical clustering . The results are shown in Fig . E ( a ) & E ( b ) in S1 File for the base-line and the plastic scenarios of the differentiation-regulating ER system , respectively . Not unexpectedly , we observe that , with respect to the base-line scenario , correlations substantially change when the plastic scenario is considered . Although the strategies illustrated in the results shown in Fig 7 changed one or two parameters alone independently of all the others , more general situations will require to closely monitor these correlations to understand which combinations of parameters are relevant to control the system’s behaviour [35] .
We here provide computational evidence for the role of stochastic translation of epigenetic cofactors into resilient/plastic cell states via ER systems as a mechanistic facilitator of cellular aging , and its reversal . When changes in levels of such cofactors operate as regulators of the kinetic parameters associated with chromatin-modifying enzymes such as HDMs and HDACs , the ensemble of ER configurations reveals the occurrence of cell-to-cell phenotypic variability in terms of different epi-states ( see Fig 8 ) . This model provides a rationale for the responsiveness of cellular phenotypes to metabolic signals , as metabolic pools serve as epigenetic cofactors . The metabolic control of epigenetic landscapes and cell state transitions might therefore operate as a common hub capable of facilitating the pathogenesis of aging-related diseases including cancer . Several layers of molecular communication exist between cell metabolism and chromatin remodelling [16 , 52–56] . A first layer of metabolo-epigenetic regulation includes metabolites/nutrient-responsive TF-dependent transcriptional regulation of chromatin regulators ( HMT , HAT , DNMTS , etc ) , which can lead to global changes on chromatin structure . Second , metabolites can modulate chromatin modifications at specific genomic loci by affecting the activity/localisation of proteins that recruit or regulate chromatin-modifying enzymes during , for example , transcriptional activation phenomena . Third , chromatin-modifying enzymes employ many metabolites as donor substrates and cofactors , and changes in levels of these bona fide epigenetic metabolites can in turn lead to changes not only in the global status of chromatin modifications but also to gene specific regulation under different metabolic conditions . Our mathematical model only incorporates the third such layer through cofactor-induced heterogeneity . Because any metabolic input has the potential to affect various chromatin marks via its effects on transcription , our model ignored metabolic regulation of TF activity . In contrast to other metabolically-regulated enzymatic activities such as phosphorylation in which the substrate ( ATP ) is present in cellular concentrations far greater than the enzyme Km values , i . e . , the concentration of metabolite at half maximum velocity of enzyme-mediated reaction , the physiological cellular concentrations of donors and cofactors that are employed by histone-modifying enzymes ( e . g . , organic ketoacids such as the demethylase cofactor α-ketoglutarate for HDMs or the NAD+ deacetylase cofactor for HDACs ) are close to HDM and HDAC Km values [16 , 57]; consequently , based solely on the intrinsic biochemical characteristics of chromatin-modifying enzymes such as HDMs and HDACs , small fluctuations in the concentrations of such metabolites could significantly alter HDM and HDAC activities , either increasing or decreasing their respective histone-modifying activities . This layer of metabolo-epigenetic regulation is commonly viewed as a direct link from cell metabolism to chromatin-modification status , which could be mathematically modelled and tested as has been confirmed in our current computational model ( see Fig 8 ) . Evidence accumulates demonstrating that differing metabolomes can be found in distinct cell states , thereby suggesting how changes in metabolism can impact and probably specify cell fate via alteration of the chromatin landscape [58–63] . Yet , there is a scarcity of examples showing that metabolic changes can restructure the epigenetic landscape and lead to different cell states regardless of other global changes in cell physiology occurring in response to this variation in metabolite levels . Our findings support the notion that changes in the abundances of certain metabolites would alter specific chromatin marks , thereby determining both the stability of cell types and the probability of transitioning from one epi-state to another [64] . Our model infers that such a change in metabolite level would be sufficient to either impede or allow cell epi-state transitions by regulating the height of the phenotypic barriers in the context of Waddington’s landscape ( Fig 8 ) . However , we should acknowledge that the necessary involvement of cellular metabolism on the structure of the epigenetic landscape will require the experimental coupling of defined metabolic conditions with epigenome editing systems ( e . g . , CRISPR-Cas9 ) capable of targeting specific histone PTMs playing important roles in chromatin structure [65] . Our ensemble approach provides mechanistic support to the notion that emergence of the cellular and molecular hallmarks of aging including cancer might result from a metabolically driven loss of epigenetic resilience . Flavahan et al . [57] have recently proposed that non-genetic stimuli including aging and metabolic insults can induce either overly restrictive chromatin states , which can block tumor-suppression and/or differentiation programs , or overly permissive/plastic chromatin states , which might allow normal and cancer cells to stochastically activate oncogenic programs and/or nonphysiologic cell fate transitions . Our ensemble approach provides a framework that supports heterogeneity of epigenetic states as an engine that facilitates cancer hallmarks and other aging diseases . On the one hand , the ability of resilient states to maintain large epigenetic barriers refractory to non-physiologic cell fate transitions might explain why the NAD+-dependent HDAC/sirtuin pathway is one of the few mechanisms described to mediate the correction or resetting of the abnormal chromatin state of aging cells induced by calorie restriction , the most robust life span-extending and cancer preventing regimen [2 , 66–68] . On the other hand , the ability of plastic states to lower epigenetic barriers , and increase the sensitivity of primed cells to undergo reprogramming-like events leading to loss of cell identity is consistent with the ability of certain metabolites to promote oncogenesis by epigenetically blocking the HDM-regulated acquisition of differentiation markers [17 , 69–71] . The traditional view of cancer formation ( i . e . , the Knudson model [72] ) exclusively involves the binary acquisition and accumulation of genetic alterations as the principal driver mechanism for the age-dependency of multistage cancer development . Our ensemble approach suggests an alternative , namely , that oncogenic chromatin aberrations might also occur via purely epigenetic stimuli . Our model shows that , nested within the ensemble of ER systems , those that prime cells for reprogramming exhibit properties associated with age-induced epigenetic dis-regulation [73 , 74] . Aging-responsive ER reprogramming might thus operate in a more progressive and graded manner to increase cancer susceptibility without the need to induce genetic mutations . Our ensemble model is mechanistically consistent with the fact that those cancers in which the sole presence of epigenetic metabolites ( e . g . , oncometabolites ) suffices to stabilise undifferentiated cellular states by preventing demethylation of genes implicated in differentiation have accelerated models of oncogenesis [44 , 75–82] . Whereas the epigenetic signature of adult somatic cells must be partially and acutely erased to adopt a more plastic epigenome , such cellular plasticity , which might occur via metabolically driven epigenetic activation of promoter regions of pluripotency genes , could impose a chronic , locked gain of stem cell-like states disabled for reparative differentiation . The existence of metabolism-permissive resilient and plastic epigenetic landscapes might have predictive power on the susceptibility of a cell to lose its normal cellular identity through reprogramming-like resetting phenomena . The beneficial or deleterious decision paths during the maintenance of cell and tissue homeostasis might be closely related to the ability of epigenetic landscapes to modulate the intrinsic responsiveness to reprogramming cellular identity . The incapability of finishing cellular reprogramming , or at least to increase cellular epigenetic plasticity , might impede tissue self-repair in response to injury , stress , and disease , thus driving the observed aging phenotypes . Accordingly , the infliction of chronic injury and the aging phenotype have been shown to render tissues highly permissive to in vivo reprogramming [47] while the cyclic , transient expression of reprogramming factors has recently been shown to increase lifespan in a murine model of premature aging via remodeling of the chromatin landscape [7] . Because our model suggests that the fine-tuning of metabolic epigenetic cofactors might direct plastic epigenetic states to re-enter into epigenetic resilience , and vice versa , it would be relevant to experimentally evaluate whether specific metabolic interventions might either mimic transient reprogramming and revert some age-associated features without promoting complete undifferentiation , or prevent the occurrence of unrestricted/uncontrolled plasticity in chronically injured tissues such as those occurring in aging and cancer . In summary , by integrating the ability of chromatin epigenetic modifiers to function as sensors of cellular metabolism , our ensemble model provides computational support to the notion that a metabolism-responsive loss of epigenetic resilience might mechanistically facilitate cellular aging . The stochastic translation of metabolic signals into resilient/plastic cell states via ER systems might be viewed as a metabolo-epigenetic dimension that not only facilitates cellular aging , but that also offers new therapeutic and behavioural avenues for its reversal . Our findings strongly suggest that the development of predictive mathematical models and computational simulation platforms capable of operatively integrate the metabolic control of epigenetic resilience and plasticity and its combination with confirmatory lab-based testing might accelerate the discovery of new strategies for metabolically correcting the aberrant chromatin structure that affects cellular identity and epi-state transitions in aging and aging-related diseases .
|
Cell reprogramming , a process that allows differentiated cells to re-acquire stem-like properties , is increasingly considered a critical phenomenon in tissue regeneration , aging and cancer . In light of the importance of metabolism in controlling cell fate , we designed a computational model capable of predicting the likelihood of cell reprogramming in response to changes in aging-related metabolites . Our predictive mathematical model improves our understanding of how pathological processes that involve changes in cell plasticity , such as cancer , might be accelerated or attenuated by means of metabolic reprogramming .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"enzymes",
"cancer",
"risk",
"factors",
"enzymology",
"cell",
"metabolism",
"oncology",
"stem",
"cells",
"enzyme",
"metabolism",
"epigenetics",
"enzyme",
"chemistry",
"chromatin",
"cell",
"potency",
"chromosome",
"biology",
"proteins",
"animal",
"cells",
"gene",
"expression",
"pluripotency",
"biochemistry",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"cofactors",
"(biochemistry)",
"aging",
"and",
"cancer"
] |
2018
|
Epigenetic regulation of cell fate reprogramming in aging and disease: A predictive computational model
|
One ultimate goal of metabolic network modeling is the rational redesign of biochemical networks to optimize the production of certain compounds by cellular systems . Although several constraint-based optimization techniques have been developed for this purpose , methods for systematic enumeration of intervention strategies in genome-scale metabolic networks are still lacking . In principle , Minimal Cut Sets ( MCSs; inclusion-minimal combinations of reaction or gene deletions that lead to the fulfilment of a given intervention goal ) provide an exhaustive enumeration approach . However , their disadvantage is the combinatorial explosion in larger networks and the requirement to compute first the elementary modes ( EMs ) which itself is impractical in genome-scale networks . We present MCSEnumerator , a new method for effective enumeration of the smallest MCSs ( with fewest interventions ) in genome-scale metabolic network models . For this we combine two approaches , namely ( i ) the mapping of MCSs to EMs in a dual network , and ( ii ) a modified algorithm by which shortest EMs can be effectively determined in large networks . In this way , we can identify the smallest MCSs by calculating the shortest EMs in the dual network . Realistic application examples demonstrate that our algorithm is able to list thousands of the most efficient intervention strategies in genome-scale networks for various intervention problems . For instance , for the first time we could enumerate all synthetic lethals in E . coli with combinations of up to 5 reactions . We also applied the new algorithm exemplarily to compute strain designs for growth-coupled synthesis of different products ( ethanol , fumarate , serine ) by E . coli . We found numerous new engineering strategies partially requiring less knockouts and guaranteeing higher product yields ( even without the assumption of optimal growth ) than reported previously . The strength of the presented approach is that smallest intervention strategies can be quickly calculated and screened with neither network size nor the number of required interventions posing major challenges .
Stoichiometric and constraint-based modeling techniques such as flux balance analysis or elementary modes analysis have become standard tools for the mathematical and computational investigation of metabolic networks [1]–[4] . Although these methods rely solely on the structure ( stoichiometry ) of metabolic networks and do not require extensive knowledge on mechanistic details , they enable the extraction of important functional properties of biochemical reaction networks and deliver various testable predictions . The steadily increasing number of reconstructed and examined genome-scale metabolic network models of diverse organisms proves that methods for constraint-based modeling can deal with networks comprising up to several thousands of metabolites and reactions [1] . Metabolic networks consisting of m internal metabolites and n reactions can be formalized by an m×n stoichiometric matrix N . A common assumption of constraint-based methods is that the network is in steady state ( i . e . , the metabolite concentrations do not change ) resulting in a system of homogeneous linear equations ( 1 ) where r is the vector of ( net ) reaction fluxes or reaction rates . In addition , the non-negativity constraints on fluxes through irreversible reactions must be fulfilled: ( 2 ) ( Irrev comprises the indices of the irreversible reactions ) . The two constraints ( 1 ) and ( 2 ) form a convex polyhedral cone ( the flux cone ) in the n-dimensional space of the rate vectors r . Flux Balance Analysis ( FBA; [3] ) searches for optimal flux distributions within this cone that maximize a given linear objective function ( 3 ) Typical objective functions are maximization of growth ( or biomass yield ) or of the yield of a certain product . For FBA , the irreversibility constraint ( 2 ) can be refined to general upper and lower boundaries for each reaction rate ri: ( 4 ) Elementary-modes analysis [2] , [5] is another stoichiometric technique facilitating the exploration of the space of feasible steady state flux distributions by means of particular flux vectors e fulfilling the basic constraints ( 1 ) and ( 2 ) and in addition a non-decomposability property . The latter demands that an elementary mode e is irreducible ( or support-minimal ) , hence , there is no vector r≠0 obeying ( 1 ) and ( 2 ) and ( 5 ) Here , P ( r ) and P ( e ) represent the support of r and e , respectively , i . e . , they contain the indices of the vector elements being non-zero: P ( t ) = {i | ti≠0} . Elementary modes ( EMs ) represent stoichiometrically balanced metabolic pathways or cycles and several important properties of a metabolic network can be analyzed by its unique set of EMs [2] , [5] . EMs correspond to extreme rays of convex cones and can be computed as such [6] , [7] . One ultimate goal of metabolic network modeling is the targeted manipulation of the network behavior . A typical application is metabolic engineering where one is interested in the optimization of the production of a certain compound by a given host organism . A number of constraint-based optimization techniques have been proposed for this purpose [2] , [8] , [9] , [10] , [11] , [12] . FBA can directly be used to determine the optimal ( maximal ) value for a given optimization problem ( e . g . , maximal yield of biomass or of a certain chemical when growing on a certain substrate ) . This approach , however , cannot yet explain which manipulations will eventually drive the cell towards this optimum . A simple approach would be to use flux-variability analysis ( FVA , [13] ) to analyze how the feasible ranges of stationary fluxes in a metabolic network would change when switching from the wild-type to a desired phenotype . More sophisticated and directed FBA-based optimization routines operate on the principle put forward by the OptKnock approach [8] . Here , the key idea is to search for interventions that lead to obligatory coupling between the production of biomass and of a desired compound . Mathematically , OptKnock is a bilevel optimization problem where the inner problem defines biomass optimization as the cellular objective and where the outer optimization problem is to search for reaction removals ( represented by integer variables ) that lead , under consideration of the inner problem , to maximal product formation . The bi-level optimization coupling can be reformulated as a single level mixed integer linear program ( MILP ) . Successful applications ( e . g . [14] ) and several refined variants of OptKnock ( including , for example , RobustKnock [9] and OptORF [11] ) have been published ( for a review see [12] ) . The advantage of FBA-based approaches is that they can readily be applied to genome-scale networks . However , a potential disadvantage is that they deliver particular solutions only where often multiple alternate solutions exist which might be equally or even more relevant than the found solutions . Some methods have therefore been proposed to enumerate intervention strategies . A brute-force approach would be to test all single , double , triple … reaction knockouts with respect to their impact on the objective function [15] , [16] . Suthers et al . [15] used this method to enumerate synthetically lethal reaction sets and found that this search becomes prohibitive in genome-scale networks for interventions with more than two or three reaction knockouts ( the upper limit set in [16] was also three ) . They designed therefore a more directed search algorithm based on a bi-level optimization method formulated as a mixed integer linear program ( MILP ) [15] . However , to the best of our knowledge , enumerated knockout sets in genome-scale networks did not exceed a cardinality of three . This is a serious limitation because complex interventions problems may require 5 , 6 , 7 or more knockouts , even in medium-scale networks ( see [17] and the examples in the Results section ) . The method of Minimal Cut Sets ( MCSs ) directly addresses the enumeration of metabolic intervention strategies [10] , [18] , [19] . MCSs specify minimal sets of reactions whose removal ( knockout ) will block certain undesired ( target ) flux distributions . For example , one can compute ( i ) MCSs that block growth; ( ii ) MCSs that disable the production of a certain compound; ( iii ) MCSs that block all flux vectors where a certain compound is produced with a low ( including zero ) yield . In the context of MCSs , the term “minimal” refers to the property that reaction cuts specified by any proper subset of an MCS are insufficient to ensure the full repression of the undesired behaviour . In this regard , the minimality of MCSs is similar to the minimality or non-decomposability property of elementary modes specified by equation ( 5 ) . In fact , there is a dual relationship between MCSs and EMs: the MCSs blocking a certain set of target flux vectors are the minimal hitting sets of the set of ( target ) EMs that generate these behaviors [19] , [20] . By this property , each MCSs must hit ( knockout ) at least one utilized reaction from each EM . As a consequence , MCSs can be computed as minimal hitting sets ( or so-called hypergraph transversals ) of the target modes , for instance , by the Berge algorithm ( see [20] ) or by Binary Linear Programming [21] . Another approach to compute MCSs , which exploits the inherent dual relationship between EMs and MCSs , was recently presented by Ballerstein et al . [22] . Briefly , the MCSs of a given metabolic network can be computed as certain EMs of a dual network; the latter being derived by a simple transformation of the ( primal ) network . This finding makes it possible to calculate MCSs by using optimized algorithms for EM computation [7] . However , there are two potential problems related to MCSs . First , when the reactions contained in an MCS are removed , we are sure that the targeted network functions are disabled but other ( desired ) functions might be blocked as well . For instance , it can occur that an MCS which disables low-yield pathways synthesizing a desired product also blocks growth of the organism making this MCS impractical . To prevent such side effects , the concept of constrained minimal cut sets ( cMCSs ) was introduced by Hädicke and Klamt [10] where not only undesired but also desired functionalities ( to be preserved ) can be specified . When the EMs are available , an adapted Berge algorithm can be used to conveniently compute cMCSs by specifying in addition to the target modes ( expressing the unwanted behaviour ) a set of desired modes expressing the functionality that must be preserved . A cMCS will hit all target EMs and keep a ( user-specified ) minimal number of desired EMs . As shown in [10] , cMCS provide a very flexible and powerful approach to enumerate intervention strategies; many other techniques such as Minimal Metabolic Functionality [2] , [17] , and the aforementioned OptKnock and RobustKnock may be reformulated as special cMCSs problems . cMCSs are also well-suited to identify knockout combinations leading to coupled growth and product formation . The second and more serious problem of ( c ) MCSs is that their full enumeration in large/genome-scale networks becomes prohibitive . The algorithms requiring as inputs the target ( and possibly desired ) EMs are usually not applicable: despite large progress in algorithmic design [7] the full set of EMs is often not computable at genome-scale . For the same reason , the dual approach of Ballerstein et al . [22] cannot be applied either . On the other hand , for the purpose of applying MCSs in real networks , those with the smallest number of elements are usually the most relevant . Thus , it is worthwhile to consider computing only the ( c ) MCSs with low cardinality . The effective enumeration of the smallest cut sets is therefore the key goal of the present work . Usually , the unwanted/desired functionalities to be disabled/kept in a metabolic network can be described by sets of linear equalities and inequalities over the fluxes . For the purpose of computing MCSs , we could therefore use an exhaustive FBA-based scheme by testing all single , double , triple and higher knockout sets whether they are suitable cut sets or not . The formulation of FBA problems would circumvent the problem to enumerate the EMs first . However , as discussed above , this approach becomes problematic if larger knockout sets are required to solve an intervention problem , as it must test a large number of candidate sets with increasing MCS size ( the number of candidates grows with where n is the number of possible cuts and k the size of cut set candidates ) . Therefore , it is not normally possible to find genome-scale MCSs in reasonable time with more than 4 knockouts using this scheme . Whereas the direct calculation of smallest MCSs in large-scale networks cannot be properly addressed yet by current methods , a method for computing the smallest ( or shortest ) EMs in genome-scale networks was recently presented by de Figueiredo et al . [23] . This algorithm formulates the search for the EMs with fewest elements as a Mixed Integer Linear Programming ( MILP ) problem and delivers in the k-th iteration the k-th shortest EM ( hence , it starts with shortest EM , delivers then the second shortest and so forth ) . As shown by the authors , this approach can readily be applied to genome-scale networks to find the first hundred or even thousand shortest EMs involving the fewest number of reactions . The goal of the present work is to realize a similar approach for computing the k-smallest MCSs from a given network structure . We show that this can be achieved in two steps . First , the original network and the actual intervention goal are converted to its dual representation using the approach of Ballerstein et al . [22] . We then compute the shortest EMs ( up to a certain size or number ) in the dual network by employing a modified algorithm of de Figueiredo et al . [23] . As the EMs in the dual network correspond to the MCSs of the primal , the shortest EMs in the dual system will represent the smallest MCSs of the original network . The paper is organized as follows: we will first briefly review the approach of de Figueiredo et al . for computing k-shortest EMs and introduce several modifications improving the performance of this algorithm . In particular , we will make use of certain features of MILP solvers provided for effective enumeration of solutions of a MILP problem . Thereafter we will describe how the network constraints ( including inhomogeneous constraints ) and the intervention goal have to be translated into their dual description in which we can then enumerate the shortest EMs to obtain the smallest MCSs in the primal network . We shall also explain how constrained MCSs can be computed within this framework . Finally , to demonstrate the power of our new approach we will exemplify its use by computing relevant intervention strategies ( of different complexities ) in iAF1260 , a genome-scale metabolic model for E . coli [24] . These benchmarks demonstrate , for example , that our approach enables us to enumerate synthetic lethals of E . coli up to size 5 which was not possible before . Moreover , we show that the algorithm facilitates the calculation of thousands of the minimal intervention strategies that lead to growth-coupled synthesis of certain compounds by E . coli . For the sake of simplicity , throughout the manuscript we will deal with reaction cut ( or knockout ) sets , which must in practice be translated to gene knockout sets to construct the corresponding mutants . This transformation can be easily achieved if the corresponding gene-enzyme-reaction associations are available . The latter could also directly be included in the problem formulations given below to compute gene ( instead of reaction ) cut sets .
We present now the key methodological development of this work showing that the basic algorithm for enumerating shortest EMs introduced in the previous section can also be used to compute smallest MCSs . The procedure is based on the duality properties of EMs and MCSs presented by Ballerstein et al . [22] which we outline in the following . A necessary first step to establish the scheme is to describe the undesired network functionality ( the “target flux vectors” r to be disabled by the MCSs ) by a suitable inequality constraint ( 13 ) where t is a ( n×1 ) vector . Usually , t corresponds to a single row with zeros except a single 1 for a target reaction ( rate ) whose operation is to be blocked ( e . g . biomass formation if we searched for synthetic lethals ) . Setting in addition b to 1 we would target all flux vectors in which the rate of the target reaction is non-zero ( in our context we can again set b to an arbitrary value greater than zero without loss of generality ) . Constraint ( 13 ) specifying the target flux vectors can be generalized to: ( 14 ) Here , matrix T ( of size t×r ) together with poses t inhomogeneous inequality constraints defining the target flux polyhedron ( which may be bounded becoming then a polytope ) . It must be made sure that the zero flux vector is not contained in the target flux polyhedron as it can not be blocked by reaction knockouts . A nice feature of ( 14 ) is that we may directly include inhomogeneous constraints to characterize target flux vectors ( with maintenance ATP demand as a typical example ) . In addition to ( 14 ) and to the standard network constraints ( 1 ) and ( 2 ) , Ballerstein et al . augmented the system by equality constraints setting all reaction rates to zero ( 15 ) ( I is the ( n×n ) identity matrix ) . These constraints ensure that the system becomes infeasible as the zero flux vector implied by ( 15 ) contradicts ( 14 ) . Note that ( 15 ) can be seen as the maximal ( trivial ) cut set knocking out every reaction in the network . In fact , the MCSs correspond to minimal subsets of the homogeneous equations in ( 15 ) which ensure ( induce ) inconsistency of the inequality system posed by constraints ( 1 ) , ( 2 ) , ( 14 ) and ( 15 ) . Minimal subsets of constraints that induce inconsistency of an inequality system are also known as irreducible inconsistent subsets ( IISs; [25] ) . Generally , IISs can be calculated as follows: using the Farkas Lemma , the infeasible primal system is converted to its dual system which is ensured to be consistent . It can be shown that the IISs of the primal system correspond to extreme rays ( and thus EMs ) in the dual system which makes it possible to calculate them using methods from EM computation . Since IISs in our particular case may , in general , also contain constraints from ( 1 ) or ( 2 ) , a modified algorithm was introduced in [22] to ensure that only those IISs ( = EMs in the dual system ) are computed which are minimal with respect to the constraints in ( 15 ) and correspond thus to the MCSs . We thus need to transform the primal system defined by ( 1 ) , ( 2 ) , ( 14 ) , ( 15 ) into its dual which can be written as follows ( cf . equation ( 8 ) in [22]; Ndual is the “dual stoichiometric matrix” and rdual the dual “rate” vector ) : ( 16 ) The ( sub- ) matrix contains the identity matrix for irreversible reactions of the primal system and is filled with n-|Irrev| zero rows at the position of reversible reactions ( note that reversible reactions of the primal system need not to be split before dualizing the system; however , reversible reactions affected by ( 14 ) must sometimes be split to properly describe the target flux polyhedron ) . As described above , the MCSs in the primal correspond to particular EMs of the dual system ( 16 ) which have minimal support in the v variables . The dual variables vi , i∈{1 … n} are thus of particular importance as their values indicate which reactions participate in an MCS . Concretely , if vi≠0 then reaction i is part of the MCS ( irrespective of the sign of vi ) , if vi = 0 then it is not . Therefore , similar as we did for reversible reactions when computing shortest EMs , both positive and negative values of vi must be checked with indicators and in order to facilitate this each vi is split into two variables , vpi and vni , both with the lower bound 0 . Furthermore , since h≥0 and because the MILP can directly operate on inequalities , we can rewrite ( 16 ) to: ( 17 ) ( the sub-matrices with subscript i refer to the part of the irreversible reactions and subscript r to the part of the reversible reactions of the primal system ) . As mentioned above , for the vpi and vni we introduce the associated indicators zpi and zni , and ( in equivalence to ( 8 ) ) the constraints ( 18 ) stating that vpi and vni cannot be active simultaneously . The constant c in ( 17 ) can again be set to any positive value ( e . g . , to 1 ) ; this will not change the set of minimal non-zero combinations of vpi and vni fulfilling ( 17 ) which are relevant for the optimization problem formulated below ( eq . ( 19 ) ) . After dualization , we can now compute the smallest MCSs of the primal system by applying algorithm ALGO2 in the dual system . As constraints we need to consider ( 17 ) ( replacing ( 1 ) and ( 2 ) from the primal system ) as well as ( 18 ) and as objective function we exchange ( 10 ) with ( 19 ) Furthermore , because the presence of one reaction in a concrete solution is now indicated by two separate variables , the exclusion constraints ( 11 ) must be adapted accordingly to ( and are the values of a given concrete solution and is a shortcut for + ) : ( 20 ) In this way both positive and negative values of the original vi are counted in the same way towards reaction participation in the MCS . Finally , for the same reason , the size control constraint ( 12 ) sums here over zpi+zni as in the objective function ( 19 ) . The MCSs of the primal network are eventually obtained by taking the z-vector of the solutions found in the dual; z is obtained by collapsing zpi and zni: zi = zpi+zni . In the previous subsection we dealt with enumeration of smallest MCSs , however , we have not yet clarified how constrained MCSs can be computed by this approach . As it turns out , this is straightforward: one first enumerates the smallest MCSs blocking the undesired flux vectors as described above . We can assume that the desired flux vectors ( of which at least one has to be kept ) is formulated by appropriate inequalities - similar as for the targeted undesired flux vectors in ( 14 ) : ( 21 ) We can then filter the true cMCS from the set of ( unconstrained ) MCSs by testing for each MCS with a separate linear program whether the removal of the reactions in the MCS still allows the network to perform the desired function , i . e . , whether the system given by ( 1 ) , ( 2 ) , and ( 21 ) is feasible when setting the rates of the reactions contained in the MCS to zero . From our experience , the computational costs for these tests are negligible compared to the calculation of the smallest MCSs , even if hundred thousand MCSs have to be tested ( see Results section ) . The MCSEnumerator method has been integrated as a new functionality in the CellNetAnalyzer package , a MATLAB toolbox for biological network analysis [26] , [27] . The implementation uses the IBM ILOG CPLEX Optimization Studio V12 . 4 for solving the respective MILP and LP problems . Arbitrary intervention problems can be defined by providing the respective matrices and vectors describing the network and the desired and undesired flux vectors . The resulting MILPs are set up via the JAVA-CPLEX API and MATLAB's integrated JVM while for running the LPs the MATLAB-CPLEX interface is used . A separate package containing the data and script files needed for running the iAF1260 examples discussed herein can be downloaded from http://www . mpi-magdeburg . mpg . de/projects/cna/etcdownloads . html .
In order to compare our MILP-based MCS enumeration scheme to other approaches the same benchmark problems as in Table 1 in [22] were used . The target of the ( unconstrained ) MCSs in these problems is the deactivation of biomass synthesis in a smaller model of the central metabolism of E . coli for growth under different substrates ( acetate , succinate , glycerol , glucose ) . The MCSs determined in this way will thus correspond to the synthetic ( reaction ) lethals for E . coli ( whose compositions depend strongly on the provided substrate ) . Before using the different MCS calculation routines the metabolic network is compressed by combining correlated reactions ( operating with a fixed ratio under steady state conditions ) to single cumulated reactions [6] . The compression in the primal system can also conducted if the computation is done in the dual system . MCSs found in the compressed network must be decompressed after calculation [18] . The number of calculated MCSs and computation times are shown in Table 1 . As a first observation , it is apparent that calculation of EMs followed by the Berge algorithm ( computing MCSs as the minimal hitting sets of the selected target EMs; Haus et al . 2008 ) is the most efficient of the shown MCSs calculation methods . The approach of Ballerstein et al . to compute primal MCSs as EMs in the dual system performs similar to EM calculation+Berge algorithm in the ( primal ) network but in its current implementation it requires a lot of memory . For this reason , the MCSs for glucose could not exhaustively be enumerated by this approach on the computer used ( with an effective memory limit of 2GB per process ) . Although the MILP algorithm developed herein was actually developed to compute the smallest MCSs , we can use it here even for enumerating all of them . The EMs in the dual network ( the MCSs in the primal ) where computed with both MILP formulations for shortest EM calculation: ALGO1 ( the original approach by de Figueiredo et al . [23] implemented with indicator variables ) and the ALGO2 approach calling the populate sub-routine for fixed EM sizes . Generally , applying the MILP formulations to the dual system is at first sight comparatively slow even when using multiple threads . Nonetheless , it is apparent that solving the dual system with our new ALGO2 is more efficient ( ∼17 times faster ) than ALGO1 based on the scheme used by de Figueiredo et al . [23] . As can be seen for the MCSs with glucose as substrate , increasing the number of threads from 1 to 4 on the same CPU decreases the time needed for computation to some extent when using ALGO1 or ALGO2 . Using 12 threads on a compute cluster node yields a more noticeable speed improvement but , as in the case of 4 threads , the combined computation times of all threads is still larger than in the case where a single thread is used . The main advantage of our new approach can be seen in the case where only the MCSs up to size 4 have to be calculated ( fifth row in Table 1 ) : here the dual approach in combination with ALGO2 is clearly the fastest way to determine small MCSs among the approaches compared . As described in the Introduction section , the direct calculation of EMs and MCSs in genome-scale networks is normally infeasible . For this reason , the Berge algorithm and the dual system approach by Ballerstein et al . used in the previous example become impractical . In contrast , with the MILP approach enumerating shortest EMs in the dual system as proposed here , calculation of small MCSs becomes possible . To demonstrate this , we use the E . coli genome-scale network iAF1260 [24] that accounts for 1260 ORFs and defines the reversibilities of the included reactions . In total , this network comprises 1668 internal metabolites and 2382 reactions including 304 exchange reactions with the environment and 29 spontaneous reactions . The intervention goal for the MCSs to be computed is again to disable growth ( biomass formation ) when glucose is available as sole carbon source . The glucose uptake rate was fixed to = 10 mmol/ ( gDW⋅h ) and the ATP maintenance requirement was set to the standard value of = 8 . 39 mmol/ ( gDW⋅h ) . Analogous to Suthers et al . [15] we considered a cell viable if it has a growth rate larger than μmin≥0 . 01⋅μmax = 0 . 0093 h−1 . With these inhomogeneous conditions , the MCSs will thus correspond to synthetic reaction lethals as also computed by Suthers et al . [15] , where full enumeration for MCSs up to size 3 was achieved ( some MCSs of size 4 could also be determined ) . With glucose and oxygen available 152 reactions are disabled as suggested by the gene-regulatory model included in the iAF1260 reconstruction . A subsequent flux variability analysis revealed 991 blocked reactions in total and these were removed from the network . In addition , spontaneous and exchange reactions , of which 23 resp . 97 remain after removing blocked reactions , were not allowed to take part in any MCS . After removing the blocked reactions network compression by combining correlated reaction sets was again applied by which the ( primal ) network could be reduced to 562 metabolites and 936 ( lumped ) reaction subsets of which 816 can be knocked out . By using ALGO2 in the dualized system , for the first time it was possible to fully enumerate all synthetic ( reaction ) lethals of sizes 1 to 5 as shown in Table 2 yielding a total set of 2486 MCSs . Although the last iteration ( MCSs with 5 knockouts ) took several days all of them could be determined . Comparison of the runtimes of our MCSEnumerator implementation and of SL Finder ( used in [15] ) for the calculation of MCSs of size two and three indicates that our algorithm is more than 100 times faster therefore allowing full enumeration of synthetic reaction lethals also of size 4 and 5 . We also tested the homogeneous version of the above intervention problem , that is , we calculated the MCSs blocking growth without the additional constraints for ATP maintenance ( is a free flux ) , without restriction on glucose uptake and without the minimum threshold for the growth rate ( all flux vectors with biomass production >0 have thus to be blocked ) . As expected , for we found less MCSs of size 1–5 ( 1933 in total ) because the target polyhedron containing the target flux vectors was expanded leading to larger MCSs with more than 5 reaction deletions . We also observed that the computation of the MCSs in the homogeneous problem was much faster ( ∼17 hours ) than for the inhomogeneous scenario ( ∼430 hours ) indicating that inhomogeneous constraints may complicate the whole calculation procedure . The following third example relates to a typical problem of finding rational intervention strategies for metabolic engineering purposes . We here focus on a biotechnologically relevant application , namely to let E . coli produce a biofuel ( ethanol ) from glucose . The intervention goal is thus to disable flux vectors with a low ethanol yield in E . coli ( undesired behavior ) while retaining the capability for both maintenance and growth of the bacterium under anaerobic conditions ( desired functionality ) . This forms a constrained MCS problem . All cMCSs that fulfill the stated requirements will lead to obligatory coupling between growth and ethanol formation . We used again the iAF1260 genome-scale network model of E . coli metabolism but this time with the oxygen uptake removed to establish anaerobic conditions on the network . As before , glucose is the only available carbon source . To study the effect of different capacities for substrate uptake , we considered two possible limits for the glucose uptake rate: = 10 mmol/ ( gDW⋅h ) and = 18 . 5 mmol/ ( gDW⋅h ) . The latter value has been measured under anaerobic conditions where E . coli tends to exhibit higher substrate uptake rates [28] . The ATP maintenance requirement was set to ≥8 . 39 mmol/ ( gDW⋅h ) . With these values in mind , we formulated the following intervention goal: the task is to identify cMCSs that guarantee a minimal ethanol yield of or , in a second scenario , of . In addition , a minimum growth rate of at least μmin≥0 . 001 h−1 was demanded . With these inhomogeneous constraints we can now specify the target flux polyhedron containing all undesired network behaviors to be eliminated by the cMCSs: ( 22 ) ( YEth/Glc ( r ) denotes the ethanol yield of the reaction rate vector r ) . The set of desired behaviors from which we want to keep at least some flux vectors is given by: ( 23 ) ( The constraints due to anaerobic growth ( e . g . , oxygen uptake is zero ) were not restated in ( 22 ) and ( 23 ) . ) With these values , several linear programs were run in a preprocessing step to explore network capabilities . For and , the maximal ethanol yield is 2 ( molecules ethanol per molecule glucose ) . The maximum growth rate is 0 . 1955 h−1 ( for ) and 0 . 4954 h−1 ( for ) if we want to achieve an ethanol yield of at least 1 . 4 ; these values drop to 0 . 1356 h−1and 0 . 4827 h−1 , respectively , for a minimal ethanol yield of 1 . 8 . Hence , we can be sure that the set of desired behaviors is not empty . We then computed the cMCSs . As described in the Methods section , the calculation of cMCSs ( accounting for undesired and desired behavior ) based on our approach requires to first compute the MCSs blocking the undesired behavior and to keep afterwards only those MCSs that admit the desired behavior . This test is done for each found MCS by solving a separate linear program ( LP ) which verifies whether the remaining network supports the desired behavior . To reduce the search space , blocked reactions for the network under desired ethanol production conditions were determined and removed in a preprocessing step using flux variability analysis [13] . In addition , 104 reactions were disabled for growth on glucose as suggested by the gene-regulatory model included in the iAF1260 reconstruction . The FVA then identifies 996 blocked reactions in total , which are removed from the network . Furthermore , the remaining 19 spontaneous and 94 exchange reactions were again not allowed to take part in the MCSs . The latter can be easily achieved by setting the upper bounds of the corresponding zpi and zni indicator variables to zero . After network compression , the ( primal ) network could be reduced to 562 metabolites and 958 ( lumped ) reaction subsets of which 845 can be knocked out . Note also that the disruption of glucose uptake or ATP maintenance are valid MCSs deleting all undesired behaviors but they violate for trivial reasons the desired functionality ( growth not possible ) and can thus not be contained in any valid cMCSs . Such reactions being essential for the desired flux space could also be identified at an early stage and then be excluded from the search space . Table 3 shows the results for the computation of the ( c ) MCSs for this problem . As we considered two different maximal glucose uptake rates and two different minimal ethanol yields we obtained four scenarios . We were able to enumerate all cMCSs up to size 7 in all four scenarios within 21 hours . For each scenario , after calculating first the ( unconstrained ) MCSs up to size 7 , each MCS was tested with a LP whether the solution space of ( 23 ) is non-empty ( i . e . , whether the MCS is a valid cMCS ) . These tests took less than 7 minutes running time ( single-threaded; on the same computer that was used for MCS calculation ) for each of the four scenarios . Hence , the LPs account only for a negligible part of the overall computational costs . As can be seen in Table 3 , only a fraction ( between 1 . 3% and 6% ) of the computed MCSs up to size 7 turned out to be valid cMCSs . However , a large number of several thousand cMCSs could eventually be computed for each scenario . We then analyzed the cMCSs in more detail . A first observation in Table 3 is that in three of the four scenarios considered cMCSs were found comprising only three reaction deletions; whereas for the case with smaller glucose uptake and higher demanded ethanol yield ( scenario 2 in Table 3 ) at least 5 reaction removals are required . Generally , it is intuitive that expanding the space of undesired flux vectors in ( 22 ) and reducing the space of desired solutions in ( 23 ) by increasing can lead to larger cMCSs since ( i ) a larger set of undesired flux vectors must be suppressed , and ( ii ) due to the reduced set of desired behaviors a smaller number of MCSs become admissible cMCSs . Hence , there is no cMCS in scenario 2 that is a subset of any cMCSs in scenario 1 in Table 3 , but the other way around can occur . The same relationship exists between scenarios 3 and 4 . Thus , the higher the yield that we want to guarantee by an intervention strategy , the larger is the required effort in terms of number of reaction knockouts . The situation is different in the case of increasing . While the target flux polyhedron in ( 22 ) increases potentially demanding more cuts , the space of desired behaviors in ( 23 ) expands as well meaning that an MCS that was not a suitable constrained MCS in the case with smaller could now become a suitable cMCS . Hence , when increasing , some cMCSs of a given size might disappear whereas others may arise as new solutions . This is also reflected by the cMCSs of size three which are depicted in Figure 1 . All these cMCSs block central pathways for glucose degradation . An essential cut ( red cross in Figure 1 ) for all cMCSs is that of the glucose-phosphate isomerase blocking upper glycolysis . In addition , all the considered cMCSs block the Entner-Doudoroff pathway by either cutting the phosphogluconate dehydratase or the 2-keto-3-deoxyphosphogluconate aldolase ( blue crosses in Figure 1 ) . In addition , for scenario 1 ( with the smaller values for and ) , we have to cut one additional reaction out of 4 reactions of the pentose phosphate pathway ( dark green crosses in Figure 1 ) whereas for scenarios 3 and 4 ( whose two cMCSs of size three are identical ) the third cut is given by the pyruvate-formate lyase reaction ( light green cross in Figure 1 ) . This result confirms that increasing ( from scenario 1 to scenario 3 ) may remove existing cMCSs but also produce new ones . The cMCSs for scenario 1 ( the red cut , one of the two blue cuts and one of the four dark green cuts in Figure 1 ) also illustrate the difference between reaction and enzyme/gene cut sets . Since two of the four reactions with a green cross are catalyzed by the same enzyme ( transketolase ) knocking out the corresponding two genes ( there are two different transketolases in E . coli ) would actually cut two reactions at the same time for which the model predicts that E . coli can not grow anymore . Thus , only four of the eight cMCSs remain valid on gene basis . However , as already explained earlier , those effects can be taken into account based on gene-enzyme-reaction associations . The fact that three reaction or gene knockouts may suffice to induce a high ethanol yield of more than 1 . 8 ( scenario 4 ) is a surprising fact on its own . Previous work on computing intervention strategies for ethanol overproduction in a smaller ( core ) network of E . coli showed that more than three reaction knockouts would be required to ensure a large ethanol yield ( see , e . g . , [10] ) . Given the results with three knockouts made herein , this might be a bit confusing since much more inefficient pathways will exist in a genome-scale network which must all be blocked . However , similar as discussed above for a scenario with increased substrate uptake rates , a larger network may also have additional high-yield metabolic routes ( allowing coupled biomass and ethanol synthesis ) not contained in the smaller network which could ‘survive’ a cut set for blocking the low-yield pathways . We can thus conclude that genome-scale network models may reveal metabolic engineering strategies that are smaller than those found in small- or medium-scale subnetworks . Importantly , one always has to keep in mind that an MCS predicts an intervention purely from stoichiometric relationships . Whereas blockage of the undesired flux vectors can be guaranteed if the network structure is correct , it can not ensure that the remaining pathways will have the capacity to carry a flux that is large enough to fulfil the requirements of the desired flux vectors . In addition , unknown regulatory constraints may further reduce the space of desired behaviors by which some cMCSs may become invalid . We mention here that two other intervention strategies with three knockouts for production of ethanol by E . coli were presented in [9] . However , these solutions ensure high ethanol yield only if the cell grows at maximal growth rate whereas our interventions are more stringent since they guarantee a high ethanol yield for any growth rate of the mutant . Having exhaustively enumerated the cMCSs up to a given size enables one to analyze essential features and performance measures of all found intervention strategies by which eventually the optimal knockout strategy can be selected . Figure 2 shows exemplarily two such performance studies . Figure 2A displays for each cMCS of scenario 3 the relationship between ( i ) maximal growth rate , ( ii ) minimal ( guaranteed ) product yield ( shown for maximal substrate uptake rate; the lower boundary for arbitrary substrate uptake rates still holds to be 1 . 4 ) , and ( iii ) number of required reaction deletions ( cut set size ) . It can be seen that most cMCSs ( including those with the smallest size 3 ) achieve relatively low growth rates ( lower than 0 . 1 h−1 ) and that in order to have a growth rate larger than 0 . 1 h−1 it is necessary to use cut sets with a least 6 knockouts . If higher growth rates and/or smaller cut sets are required the minimal product yield would have to be lowered . Other performance measures of designed mutant strains can be studied as well . One such proposed measure is substrate-specific productivity ( SSP ) which is the product of the growth-rate and the product yield [29] . Figure 2B shows the SSP of all cMCSs computed for scenario 3 . It can be seen that highest SSP values can only be achieved with cut sets of size 6 or 7 . This illustrates again that a trade-off between number of knockouts and certain performance measures has sometimes to be made when eventually selecting an intervention strategy for implementation . Such a screen is greatly facilitated if all cut sets have been enumerated up to a certain size . More advanced screening methods for evaluating strain design strategies have been suggested in [30] and could readily be applied to calculated cMCSs . As a technical note , it is not absolutely mandatory to have all MCSs ( up to a maximal size ) enumerated before running the LP checks for testing the “survival” of some desired flux vectors: these checks could be ( independently ) performed as soon as an MCSs has been found by the MILP solver . In fact , it is in principle possible to integrate the LP into the MILP so that the cMCSs are computed directly which offers the advantage that far fewer exclusion constraints need to be integrated while the enumeration proceeds . In practice , however , this approach showed a markedly inferior performance for the system studied here . One reason is that the LP adds further degrees of freedom to the solution space and leads to redundant solutions for the cMCSs which requires a more intricate control of the populate procedure to suppress these redundant solutions . Whether the integrated approach can be reformulated in a manner that facilitates a more efficient calculation of its cMCSs solutions is a potential topic for further investigation . To summarize the results of this sub-problem , our algorithm enabled the enumeration of all reaction knock-out sets up to size 7 that lead to coupled ethanol and biomass synthesis in E . coli . To the best of our knowledge , this exceeds by far other attempts to enumerate such metabolic engineering strategies in large-scale networks . If more computational capacity is available , one might try to find even larger cMCSs . However , the best knockout strategy to be implemented is likely to be contained among the up to 8819 smallest cMCSs found as the number of required interventions will be one ( though not the only ) key criterion when deciding for a concrete strain design . One large-scale study to evaluate the growth-coupled production potential in E . coli has been presented by Feist et al . [29] . The aim was to identify strain designs based on reaction knockouts with a maximum production rate at optimal growth for a number of substrate/product pairs . This was achieved by first applying OptKnock [8] with a knockout limit of either three or five and then using the results in the initial population for OptGene which employs genetic programming as optimization method [31] . OptGene was then run with a time limit of one week to find additional strain designs with up to 10 knockouts . As underlying E . coli model the iAF1260 reconstruction [24] was taken and in order to reduce the search space the knockouts were restricted to a subset of about 150 reactions in the network . As minimum growth rate for the strains a limit of 0 . 1 h−1 was chosen and an ATP maintenance of 8 . 39 mmol/ ( gDW⋅h ) required . Both glucose and oxygen uptake were limited to 20 mmol/ ( gDW⋅h ) . Given this setup it was possible to calculate strain designs for many substrate/product pairs but for some of them strains with only low productivity or even no strains with growth-coupled product synthesis were found . Here we wanted to test the potential of our method for some of the intervention problems . We focused on the aerobic production of either fumarate or serine from glucose which both have a potential for high yield as calculated by FBA . However , growth-coupled strains for the production of fumarate only achieved 20% ( 5 knockouts , OptKnock ) respectively 23% ( 7 knockouts , OptGene ) of the theoretical maximum while for serine no growth-coupled strains could be identified in [29] . We therefore applied our approach to look for ( additional ) strain designs for these two configurations . To demonstrate the power of our method in dealing with large-scale systems , we increased the search space drastically compared to [29] by allowing all reactions to be knocked out except for those that are either spontaneous or essential for the production condition . Since glucose is taken up under aerobic conditions , the same 152 reactions as for the calculation of the synthetic lethals above have also been removed . This results in 718 ( fumarate ) resp . 719 ( serine ) knockout candidates ( compared to 150 candidates used in [29] ) . As the results in [29] suggested that growth coupling will be difficult for fumarate and serine production we chose a comparatively low minimal product yield of 0 . 5 . This constraint together with the ATP maintenance requirement und the uptake limits was used to calculate MCSs that disable flux vectors with product yields below 0 . 5 . Afterwards , only those ( constrained ) MCSs were kept that fulfil the minimal growth rate requirement . For fumarate production , the MCSs up to size 7 were calculated ( taking 13 . 6 h ) from which 30 cMCSs ( all of size 7 ) could be extracted . Applying those cMCSs would result in production strains exhibiting – at maximal substrate uptake rates – a guaranteed ( minimal ) fumarate yield between 0 . 71 and 0 . 89 corresponding to minimal production rates between 40 . 9% and 51 . 3% of the theoretical maximum of 34 . 68 mmol/ ( gDW⋅h ) ( note that the minimal yield for any substrate uptake rate is ensured to be 0 . 5 as demanded by the constraints for the desired flux vectors ) . As for the ethanol study , all these values are independent of the assumption of optimal growth . Likewise , in the case of serine production , the MCSs up to size 6 were calculated ( taking 3 . 1 h ) from which 140 cMCSs ( all of size 6 ) could be extracted . These would result in strains with with a guaranteed serine yield between 0 . 71 and 0 . 91 ( at maximal substrate uptake rate ) corresponding to minimal production rates between 36 . 6% and 47 . 0% of the theoretical maximum ( 38 . 71 mmol/ ( gDW⋅h ) ) . Hence , our results show that significantly larger fumarate production rates can be achieved with 7 knockouts than computed by OptGene . In case of serine where no suitable knockout strategy could be identified in [29] , our method proves the existence of strain designs for coupled biomass and product synthesis and that 6 reaction knockouts would be theoretically sufficient to guarantee a serine yield of 47% of the theoretically maximal value . Moreover , tens of the smallest strain designs with 6 knockouts could be identified by our algorithm in a comparably fast way and larger ones could also be determined if desired .
In this work we presented MCSEnumerator , a new algorithmic approach to enumerate the smallest ( c ) MCSs up to a given size in genome-scale networks . This approach is based on a MILP problem calculating the shortest EMs in the dual representation of the metabolic network eventually yielding the smallest cMCSs . The whole procedure can be summarized by five steps: With these five steps , MCSEnumerator provides a generic approach for enumerating smallest intervention strategies; one just has to plugin the corresponding matrices in equation ( 17 ) and can then start the calculation using ALGO2 . Apart from the combination of dualization and shortest EM calculation in step 3 , another key development made herein is the improvement of the required sub-routine for computing shortest EMs ( ALGO2 ) which is now based on a more efficient enumeration of feasible EMs with fixed size and which consequently makes use of available enumeration features of modern MILP solvers . Appropriate integration of such functionalities could also be useful to effectively solve other enumeration problems in the field . Despite the fact that calculation of all ( c ) MCSs with our approach is slower compared to other approaches requiring EMs to be calculated in a first step , it has the advantage that the smallest ( c ) MCSs , which are often the most interesting ones , can be found first and that no EMs need to be calculated beforehand . This property renders ( c ) MCSs calculation feasible in genome-scale networks . Also , the number of elements in an MCS has no major impact on the performance as it would have in brute-force enumerations ( that exhaustively test all reaction subsets ) and as it has been observed also for several directed search algorithms . The main drawback of using a MILP stems from the fact that constraints have to be continuously added to remove already found MCSs and their supersets from the solution space . Hence this method is bound to slow down with increasing number of constraints which explains the inferior performance when computing all MCSs . However , the shown application examples demonstrated that our approach is capable to compute hundreds of thousands of smallest MCSs and several thousand smallest constrained MCSs in genome-scale networks ( Table 3 ) which has not been achieved before . The large set of smallest cMCSs should suffice to characterize the space of the most efficient intervention strategies from which , in metabolic engineering applications , the most promising ones can be selected , possibly by screening the cMCSs via certain performance parameters . The algorithmic advantage of the presented approach lies thus in the possibility to quickly ( compared to other approaches ) calculate the smallest ( c ) MCSs with neither network size nor the number of elements in the ( c ) MCSs posing major challenges . With these results and due to the fact that the approach of ( c ) MCSs allows the setup of complex intervention problems in a flexible and convenient way , we expect that a large number of metabolic network studies can benefit from our conceived framework . An interesting aspect for future work will be to investigate how far ALGO2 ( the sub-routine used for shortest EM calculation ) can be generalized to enumerate also other elementary sets arising in different contexts of computational biology ( e . g . , for calculating minimal intervention sets in signaling or regulatory networks [32] ) .
|
Mathematical modeling has become an essential tool for investigating metabolic networks . One ultimate goal of metabolic network modeling is the rational redesign of biochemical networks to optimize the production of certain compounds by cellular systems . Accordingly , several optimization techniques have been proposed for this purpose . However , for large-scale networks , an effective method for systematic enumeration of the most efficient intervention strategies is still lacking . Herein we present MCSEnumerator , a new mathematical approach by which thousands of the smallest intervention strategies ( with fewest targets ) can be readily computed in large-scale metabolic models . Our approach is built upon an extended concept of Minimal Cut Sets , the latter being minimal ( irreducible ) combinations of reaction ( or gene ) deletions that will lead to the fulfilment of a given intervention goal . The strength of the presented approach is that smallest intervention strategies can be quickly calculated with neither network size nor the number of required interventions posing major challenges . Realistic application examples with E . coli demonstrate that our algorithm is able to list thousands of the most efficient intervention strategies in genome-scale networks for various intervention problems .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2014
|
Enumeration of Smallest Intervention Strategies in Genome-Scale Metabolic Networks
|
Hepatitis C virus ( HCV ) is an oncogenic virus associated with the onset of hepatocellular carcinoma ( HCC ) . The present study investigated the possible link between HCV infection and Netrin-1 , a ligand for dependence receptors that sustains tumorigenesis , in particular in inflammation-associated tumors . We show that Netrin-1 expression is significantly elevated in HCV+ liver biopsies compared to hepatitis B virus ( HBV+ ) and uninfected samples . Furthermore , Netrin-1 was upregulated in all histological stages of HCV+ hepatic lesions , from minimal liver fibrosis to cirrhosis and HCC , compared to histologically matched HCV- tissues . Both cirrhosis and HCV contributed to the induction of Netrin-1 expression , whereas anti-HCV treatment resulted in a reduction of Netrin-1 expression . In vitro , HCV increased the level and translation of Netrin-1 in a NS5A-La-related protein 1 ( LARP1 ) -dependent fashion . Knockdown and forced expression experiments identified the receptor uncoordinated receptor-5 ( UNC5A ) as an antagonist of the Netrin-1 signal , though it did not affect the death of HCV-infected cells . Netrin-1 enhanced infectivity of HCV particles and promoted viral entry by increasing the activation and decreasing the recycling of the epidermal growth factor receptor ( EGFR ) , a protein that is dysregulated in HCC . Netrin-1 and HCV are , therefore , reciprocal inducers in vitro and in patients , as seen from the increase in viral morphogenesis and viral entry , both phenomena converging toward an increase in the level of infectivity of HCV virions . This functional association involving a cancer-related virus and Netrin-1 argues for evaluating the implication of UNC5 receptor ligands in other oncogenic microbial species .
Cancers triggered by microbial oncogenes account for approximately 16% of cancer occurrences [1] . Hepatitis C virus ( HCV ) is a major etiologic agent of hepatocellular carcinoma ( HCC ) , the fifth most common cancer worldwide [2] . The epidermal growth factor receptor ( EGFR ) is a host factor for entry of HCV [3 , 4] as well as for the influenza virus [5] and adeno-associated virus 6 [6] . EGFR signaling is involved in HCC development [7] and possibly in the resistance to the HCC drug sorafenib [8] . An interesting advance in developmental biology and oncology in the last decade was the discovery of dependence receptors ( DRs ) [9–13] , a class of receptors that auto-activate and trigger apoptosis in the absence of their ligands . One such ligand is Netrin-1 . Netrin-1 is a secreted protein that was initially identified as the canonical soluble partner of the uncoordinated receptor-5 ( UNC5 ) DR family in the field of neuroembryogenesis . It inactivates UNC5-mediated intrinsic signals , including cell death , unlike most ligands that exert positive pharmacology on their cognate receptors . Recent data support the implication of Netrin-1 and its main receptors in epithelial tissues and suggested its role in the morphogenesis of “branched” organs [11] . In cancer , the model of dependence receptors predicts that instead of losing Netrin-1 receptors , a second potential selective advantage for tumor cell survival could be an autocrine expression of the ligand that inhibits this receptor . Accordingly , Netrin-1 , a reprogramming modulator [14] , is upregulated in several cancer types [13 , 15–18] as well as in cancer-associated inflammatory diseases such as colitis and Crohn’s disease [13 , 19 , 20]—for review , see Paradisi and Mehlen [21] . The inflammatory response associated with several epithelial disorders thus appears to play a key role in Netrin-1 induction . As is the case for most viral infections , chronic hepatitis C also bears an important inflammatory component , thought to strongly participate in the worsening of the liver structure and function , which could ultimately result in cancer promotion within hepatocytic compartments . Taken together , these data argue in favor of the establishment of a more complete landscape regarding the interplay between inflammation and cancer . Long-term infections involving inflammation-inducing oncogenic viruses may represent potential factors for the regulation of Netrin receptors . To our knowledge , data linking such factors and oncogenic viruses are nonexistent . In addition , neither the regulation of the Netrin-1 transcript nor that of the protein have , so far , been identified in association with HCV via high-throughput studies . We therefore decided to investigate the possible interplay between the DR system and HCV , focusing on Netrin-1 . In this study , we show that Netrin-1 is upregulated by HCV and that it participates in a mutual amplification loop with HCV , in turn leading to an enhancement of viral infectivity . Our results indicate that induction of Netrin-1 by HCV represents an important mechanism by which the virus establishes persistent infection and may contribute to neoplastic transformation .
HCV , along with several other liver conditions , is known to trigger hepatic inflammation . To establish a connection between the expression of Netrin-1 ( Uniprot Acc . # O95631 ) and viral infection of the liver , we first measured the level of Netrin-1 mRNA ( GenBank Acc . # NM_004822 ) in 418 liver biopsies , taken either from virus-free patients ( 165 samples ) , from HCV-infected patients ( 223 samples ) , or from HBV-infected patients ( 30 samples ) ( S1 Table ) . The latter were included as a positive control for chronic viral infection of the liver , and tissue biopsies revealed an 11-fold increase in the level of Netrin-1 mRNA compared to uninfected controls ( Fig 1A ) . Interestingly , the HCV-infected samples displayed a further 2-fold increase in Netrin-1 transcripts versus HBV+ samples , totaling a 23-fold increase in Netrin-1 mRNA levels compared to the uninfected controls . Moreover , a positive correlation was found between the levels of Netrin-1 mRNA and HCV RNA in those liver biopsies ( Fig 1B ) . Similarly , HCV RNA and Netrin-1 mRNA levels were measured in patients before and after first-time treatment with interferon and ribavirin , two antiviral compounds , in biopsies obtained from 18 HCV+/HBV- patients . Of these , 16 showed a partial treatment response ( i . e . , presented a decrease in viral load; Fig 1C , left panel ) , accompanied in all but one with a clear decrease in Netrin-1 mRNA levels ( Fig 1D , left panel ) . The two patients who failed to respond to treatment ( Fig 1C , right panel ) showed stable or increased levels in Netrin-1 mRNA , which paralleled their stable or increased HCV RNA load ( Fig 1D , right panel ) . These data support the HCV-dependent status of Netrin-1 upregulation in HCV-positive patients . Chronic HCV infection features gradual worsening of the liver through the replacement of functional hepatocytes by nonfunctional connective tissue , a process called fibrosis . To determine whether an increase in the level of Netrin-1 mRNA resulted in a concurrent increase at the protein level , immunostaining for Netrin-1 and the HCV E2 envelope glycoprotein antigens was performed on liver samples matched for fibrosis score . Netrin-1 was clearly observed in HCV+ samples compared to their uninfected counterparts , and , furthermore , the use of the well-characterized anti-E2 antibody [23] confirmed that only hepatocytes exhibit positive Netrin-1 staining in infected samples ( Fig 1E and S1 Supporting Information ) . The additional comparison of protein levels in HCV- and HCV+ liver tissues , by immunoblotting , corroborated our findings ( Fig 1F ) . Together , these data indicate that hepatocytes of HCV+ patients express increased levels both of Netrin-1 mRNA and protein and further strengthen the likelihood that HCV is a hepatocytic Netrin-1 inducer . It is well known that Netrin-1 induction can result from inflammation , in particular in the gastrointestinal system [21] . To address the specific role of HCV in our model and distinguish fibrosis-associated Netrin-1 induction from HCV-associated Netrin-1 induction , we plotted Netrin-1 mRNA levels in HCV+ versus HCV- samples against all ( F0 to F4/cirrhosis ) histological stages . Netrin-1 mRNA had increased significantly by 25-fold ( F0 ) , 15-fold ( F1 ) , 17-fold ( F2 ) , 12-fold ( F3 ) , and 4-fold ( F4 ) in all HCV-infected samples compared to their HCV-uninfected counterparts ( HCC: 1 . 4-fold ) ( Fig 1G ) . In addition , levels of Netrin-1 mRNA were further elevated ( >30-fold ) in HCV-infected cirrhotic patients compared to control patients ( samples F0–F3 ) . Of note , no association could be observed between non-HCV clinical parameters and levels of Netrin-1 mRNA ( S1 Fig; see also S2 Supporting Information for more insight on Netrin-1 transcript levels in patients ) . Importantly , a comparison of HCV ( - ) biopsies revealed that HCV-negative cirrhosis ( i . e . , F4 ) samples already displayed the highest 4-fold to 12-fold increase in Netrin-1 mRNA compared to all other HCV-negative samples ( S2C Supporting Information ) . Taken together , these data suggest that Netrin-1 expression is induced in patients chronically infected with HCV across all stages of the disease , and that HCV and cirrhosis cooperate for higher Netrin-1 induction . Next , we investigated whether induction of Netrin-1 by HCV could also be seen in vitro , in a tissue inflammation-free environment . Primary human hepatocytes ( PHH ) were infected with an HCV japanese fulminant hepatitis 1 ( JFH1; genotype 2 ) selected for its higher rate of propagation in cell cultures [24] . Results revealed a peak in Netrin-1 upregulation ( by 5-fold to 20-fold ) in the HCV-infected cultures at day two or day three ( Fig 2A–2D ) . This was also confirmed in endogenously infected PHH with wild-type genotype 3 virus ( Fig 2E ) . Our in vitro experiments were then conducted using a known hepatocytic cell line that is also amenable for mechanistic studies , in order to further examine Netrin-1 induction . We infected proliferating and dimethyl sulfoxide ( DMSO ) -differentiated [25] Huh7 . 5 cells [26] with an HCV JFH1 isolate bearing three adaptive mutations [27] , or with a non-adapted genotype 1 strain [28] . Over the five- to ten-day kinetic follow-up study , HCV induced an 8-fold increase in the levels of Netrin-1 mRNA at day eight post-infection in proliferating cultures infected with an HCV JFH1 isolate ( Fig 2F ) or with a genotype 1 strain ( Fig 2G ) and in differentiated cells ( Fig 2H ) . These results indicate that HCV induces Netrin-1 expression in hepatocytes both in vivo and in vitro . Next , we reasoned that substances other than virions released by HCV-infected cells might contribute to upregulating Netrin-1 . To test this possibility , we incubated naïve Huh7 . 5 cells with conditioned medium obtained from HCV-infected Huh7 . 5 cultures , which had previously undergone ultracentrifugation to remove virions . The HCV-depleted conditioned medium had little effect on Netrin-1 expression in the recipient cultures ( Fig 3A ) , indicating that the increase in Netrin-1 expression described above was most likely mediated by the virus . Overall , these observations indicate that HCV and Netrin-1 levels are linked in individual patients , as well as across the cohort , and that HCV infection is able to induce Netrin-1 expression in vitro . Having established that Netrin-1 expression was strongly induced by HCV , we were interested in studying the mechanisms underlying this expression . As is frequently observed in studies on Netrin-1 using cultured non-neural cell lines , Netrin-1 was difficult to detect at the protein level in Huh7 . 5 cells . In a novel and indirect approach to study Netrin-1 protein production , we investigated the association of Netrin-1 mRNA with endoplasmic reticulum ( ER ) membrane-bound polysomes , in which the translation of this secreted protein takes place , or with free polysomes , upon infection . The partitioning of the glucuronidase ( GUS ) and phosphomannomutase 1 ( PMM1 ) mRNAs , which are translated by membrane-bound and free polysomes , respectively , was also assessed as enrichment controls . Our results showed that HCV infection caused a striking enrichment in membrane-bound Netrin-1 mRNA but did not alter the profiles of the enrichment controls ( Fig 3B ) . Taking advantage of this subcellular fractionation approach , we submitted these previously obtained microsomes to Netrin-1 immunoblotting and observed an increase in the levels of Netrin-1 in HCV+ cells ( Fig 3C ) . These observations indicate that HCV increases Netrin-1 translation and support the pattern of increased levels of Netrin-1 protein observed in the HCV+ clinical biopsies . In order to investigate the HCV-related induction of the Netrin-1 protein , we examined the genetic structure of the Netrin-1 transcript and found that its mRNA has a terminal oligopyrimidine tract ( TOP ) in its 5’UTR using the RegRNA2 database [29 , 30] . As reported previously , TOP mRNAs interact with the La-related protein 1 ( LARP1; GenBank Acc . # NM_015315; Uniprot Acc . # Q6PKG0 ) protein during translation [31 , 32] . We therefore reasoned that Netrin-1 may benefit from such an interaction . Indeed , it is known that the Netrin-1 transcript is a LARP1-binding transcript [30] , and our experimental data using the LARP1-binding transcript ribosomal protein S18 ( RPS18; GenBank Acc . # NM_022551 ) as a positive control confirmed this finding in hepatocytic cells using RNA immunoprecipitation ( S2 Fig ) followed by qPCR ( Fig 4A and 4B ) . We then searched for a specific HCV factor that could implicate LARP1 in the final Netrin-1 phenotype . We screened for potential interactions between individual HCV proteins and LARP1 using a mammalian cell-based protein-fragment complementation assay ( PCA ) . This technique provides a highly reproducible and specific means of measuring protein interactions , including those involving membrane proteins in a cell model and in subcellular compartments [33 , 34] . Open reading frames ( ORFs ) of the ten individual HCV proteins [35] were cloned and recombined into an expression vector containing a fragment of the luciferase reporter ( GLuc1-A ) . These plasmids were expressed in Huh7 . 5 cells along with a plasmid encoding for the complementary luciferase fragment ( GLuc2-A ) fused to LARP1 . This screen revealed a novel interaction between LARP1 and HCV NS5A ( Fig 4C ) . The apparent affinity of the NS5A-LARP1 binding was comparable to or greater than that of NS5A binding to its known partners VAPA , GRB2 , RAF1 , PITX1 , and TP53 ( reviewed in He et al . [36] ) . In contrast , core and NS5B exhibited weak binding close to the detection threshold , and E1 , E2 , P7 , NS2 , NS3 ( expressed either individually or with Flag-tagged NS4A ) , NS4A , and NS4B exhibited an even lower level of binding . Interestingly , NS5A is an ER-bound protein ( reviewed , for instance , in reference [36] ) and is , therefore , theoretically capable of bringing its interacting partners closer to this subcellular compartment . In this context , we verified whether HCV infection induced alterations in the expression pattern of LARP1 in infected cells . Since HCV NS5A appeared to bind to LARP1 with the highest affinity , we conducted immunofluorescence assays to confirm this finding . LARP1 signals were strongly reconfigured following HCV infection , adopted a granular pattern at the expense of their initial homogenous staining profile in naïve cells , and concentrated at the vicinity of lipid droplets visible as spheric structures surrounded by LARP1 and NS5A staining ( Fig 5A ) . We confirmed that the HCV NS5A protein colocalized with LARP1 in infected cells using a plot profile assay ( Fig 5B ) . This was further confirmed using calnexin , an independent ER marker . Indeed , LARP1 underwent general relocalization to ER-positive sites ( i . e . , relevant to translation of secreted proteins ) in HCV+ cells , especially at the vicinity of classically core-decorated lipid droplets ( Fig 6 , zoomed insert ) . Li colocalization parameters presented in diagram Fig 6A and the coefficient ( Fig 6B ) between LARP1 and calnexin were significantly upregulated ( >2 . 5-fold ) following HCV infection ( for more details on Li values , see S1 Text ) . Representative images and plot profiles ( r = 0 . 16; p = 0 . 1 in naïve cells versus r = 0 . 38; p = 0 . 002 in HCV+ cells ) of these colocalization levels are shown in Fig 6C and 6D , respectively . To determine whether LARP1 was localized in translationally active sites within infected cells , we compared LARP1 aggregation sites with puromycin ( + ) areas using the ribopuromycylation method [37] . Accordingly , LARP1 had significantly accumulated in the cytosol of HCV+ cells ( Fig 7 ) . Li diagrams ( Fig 7A ) and coefficient ( Fig 7B ) were significantly upregulated ( 1 . 3-fold ) in HCV+ cells . Corresponding representative images ( Fig 7C ) and plot profiles ( r = 0 . 31 in naïve cells versus r = 0 . 82 in HCV+ cells , Fig 7D ) are shown . These data ( i ) show that LARP1 strongly relocates to ER-associated translationally active sites upon HCV infection , which comprise areas adjacent to lipid droplets , and ( ii ) led us to investigate whether the HCV-induced increase in Netrin-1 translation was , in turn , LARP1-mediated . LARP1 expression is conditioned by NS5A in infected cells . We reasoned that the siRNA-based modulation of LARP1 expression in HCV+ cells should alter the microsomal levels of Netrin-1 in these cells . To test this hypothesis , we first assessed LARP1 knockdown by immunoblotting ( S3 Fig ) and subsequently separated the microsomes from the cytosol by performing subcellular fractionation with the HSP60 marker . We then evaluated Netrin-1 levels in both types of samples after modulating the expression of LARP1 . In agreement with previous immunofluorescence data , levels of LARP1 and Netrin-1 increased in the microsomal compartment upon HCV infection , while Netrin-1 decreased in this subcellular fraction upon depletion of LARP1 ( Fig 8 ) . Therefore , the virus NS5A-mediated relocalization of LARP1 toward ER-associated translationally active sites accounts for the HCV-related increase in Netrin-1 . Consistently with the secreted protein status of Netrin-1 , this HCV-related increase occurs primarily in the secretory , microsomal machinery of the cell . These initial findings prompted us to investigate whether Netrin-1 produced by cultured cells was , in turn , able to promote HCV replication and/or propagation . HCV-infected proliferative Huh7 . 5 cells were transfected with a Netrin-1 expression plasmid . Intracellular HCV RNA , viral RNA release , supernatant infectivity , and virion-specific infectivity were then monitored . The expression vector produced a modified Netrin-1 containing the hemagglutinin ( HA ) epitope at its carboxyl-terminus , while a plasmid expressing HA-tagged vanilloid receptor ( VR1-HA ) served as a control . Immunoblotting analyses confirmed expression of both proteins in the transfected cells ( Fig 9A ) . Plasmid-delivered Netrin-1-HA produced a 3-fold increase in the intracellular HCV RNA compared to cells transfected with VR1-HA ( Fig 9B ) and was accompanied by a 2 . 5-fold increase in supernatant infectivity , measured using the TCID50 protocol ( Fig 9C ) . Furthermore , overexpression of Netrin-1 resulted in a 2-fold increase in the level of intracellular infectivity ( Fig 9D ) and likewise increased the infectivity peak of the released virions ( Fig 9E ) . Since LARP1 intriguingly concentrates at the ER and in the proximity of lipid droplets , an important viral budding site [38] , upon infection , we tested whether accumulation of Netrin-1 in microsomes could foster increased morphogenesis , through the evaluation of the specific infectivity of virions . This viral parameter is defined by the ratio of biological infectivity values , expressed as TCID50 units , and viral RNA copy numbers of the sample . To achieve this , we plotted TCID50/extracellular HCV RNA ratios against their buoyant densities for Netrin-1- and control VR1-transfected samples . Netrin-1 overexpression caused a 4-fold increase in the specific infectivity of released virions ( Fig 9F ) , suggesting that the protein induces alterations in the virus particle . Netrin-1 also resulted in a viral increase of the poorly infectious , high-density fractions ( S4 Fig ) . The DR hypothesis states that , ligand withdrawal induces receptor activation for subsequent death signaling . As Netrin-1 is known to promote cell survival [39] , we performed a set of experiments resulting in the overexpression or depletion of Netrin-1 , to test whether its effect on HCV occurs via a cell death protection-dependent mechanism . We initially performed cell death-related assays on Huh7 . 5 cells during the entire course of the transfection experiments , and observed that neither caspase-3 activity nor cell proliferation ( S5 Fig ) , known to be beneficial for HCV replication in vitro , were altered by the forced expression of Netrin-1 . This suggests that Netrin-1 does not exert its proviral effect through the death-related DR function of its cognate receptors We then incubated recombinant soluble Netrin-1-Fc on HCV in Huh7 . 5 cells . In both cell systems , Netrin-1-Fc induced a significant ( up to 2-fold ) increase in the level of intracellular HCV RNA . TCID50 assays showed that while extracellular release of HCV RNA was not changed , it produced a 2-fold increase in the level of infectivity of the supernatant ( S6 Fig ) . Netrin-1-Fc did not affect caspase-3 activity in Huh7 . 5 cells . Results of neutral red assays indicated that Netrin-1-Fc did not influence the viability of the Huh7 . 5 cells over time regardless of their HCV infection status ( S7 Fig ) . Similar results were obtained when using a distinct recombinant soluble Netrin-1 , Netrin-1-FLAG ( S8 Fig ) , strengthening our previous findings that Netrin-1 does not promote HCV infection by protecting against cell death . In turn , we also studied the effect of Netrin-1 depletion on all previously depicted viral parameters . The efficiency of the Netrin-1 siRNA was assessed by qPCR and immunoblotting ( Fig 9G ) . SiRNA-mediated Netrin-1 knockdown was associated with a 3-to-4-fold decrease in the level of intracellular HCV RNA ( Fig 9H; see also SI3 for dose-dependence data ) and caused a 2-fold decrease in the level of global infectivity of the HCV supernatant ( Fig 9I ) , as well as a decrease in intracellular infectivity ( Fig 9J ) . Infectivity of the released virions was also clearly impaired by Netrin-1 depletion ( Fig 9K ) . Consistently with results generated upon overexpression , Netrin-1-silencing caused a 3-fold decrease in the specific infectivity levels of released virions ( Fig 9L ) , another indication that Netrin-1 induces alterations in the virus particle that are virion-density unrelated ( S9 Fig ) . Finally , Netrin-1-depleted cells were also analyzed for caspase-3 activity and cell viability to rule out the possibility that the positive effects of Netrin-1 on HCV infection were due to protection against cell death . No effect of Netrin-1 depletion on caspase-3 activity ( S10A Fig ) or cell viability ( S10B Fig ) was observed . Importantly , in an approach to deplete Netrin-1 in an RNAi-independent fashion , infected Huh7 . 5 cells were exposed to a recombinant anti-Netrin-1 monoclonal antibody . This treatment resulted in a 3-fold decrease in intracellular HCV RNA ( S11A Fig ) and a 5-fold decrease in supernatant infectivity ( S11B Fig ) . Having shown that the modulation of Netrin-1 was capable of influencing specific infectivity levels of the HCV virions , in a context devoid of alterations in cell integrity , we investigated whether Netrin-1 could potentially be a component of the particles , since it also concentrates in the ER . We performed neutralization assays followed by TCID50 quantification . Anti-E2-based neutralization ( versus its RO4 isotype ) [23] served as a positive control ( Fig 10A ) , while a recombinant form of the DCC ( deleted in colorectal cancer ) receptor of Netrin-1 ( compared with the same heat-inactivated receptor ) and two distinct anti-Netrin-1 antibodies ( compared with their respective isotypic IgG controls ) were used to evaluate the effect of virus production on Huh7 . 5 cells ( Fig 10B ) . Neutralization decreased the initial level of infectivity by up to 80% , and this inhibition was enhanced by Netrin-1 overexpression in the initial virus-producing cells , suggesting that Netrin-1 participates in HCV infectivity as a candidate part of the viral particle . The effects of Netrin-1 were also tested in Huh7-derived cell lines containing subgenomic HCV replicons [40] , a system that does not generate viral particles . Results showed that Netrin-1 did not alter levels of HCV RNA ( S12 Fig ) . These observations confirm that the increase in HCV levels mediated by Netrin-1 occurs at the level of the assembly/morphogenesis , with a specific impact on the level of infectivity of the virions , but show no effect on viral RNA replication . Netrin-1 exerts most of its known activities by interacting with the DRs DCC and UNC5 [41 , 42] . In order to identify the receptor transducing the pro-HCV activity of Netrin-1 , we quantified the levels of expression of UNC5s and DCC in Huh7 . 5 and PHH cells , as well as in tissue biopsies . UNC5A ( GenBank Acc . # NM_133369 ) , B ( GenBank Acc . # NM_170744 ) , and D ( GenBank Acc . # NM_080872 ) transcripts were readily detectable in Huh7 . 5 cells , while UNC5C ( GenBank Acc . # NM_003728 ) levels remained marginal and DCC ( GenBank Acc . # NM_005215 ) mRNA was neither expressed in Huh7 . 5 cells nor in PHH . UNC5 profiles were similar in Huh7 . 5 cells , PHH , and liver tissues ( S13 Fig ) , suggesting that the in vitro setting presented in this study was a representative model for the UNC5 DR profile in patients . Based on these results , we set out to identify which of the UNC5 receptors was responsible for mediating the effects of Netrin-1 by monitoring HCV in Netrin-1-Fc-treated Huh7 . 5 cells , which had previously been subjected to siRNA-mediated depletion of each individual UNC5 transcript . Depletion of UNC5A alone induced an up to 4-fold increase in the levels of HCV ( Fig 11 , left column ) . RT-qPCR conducted to detect the UNC5 transcripts confirmed the efficacy of the siRNAs ( Fig 11 , right column ) . These results were subsequently validated using RNAi-based depletion and plasmid-mediated forced expression approaches of UNC5A ( Uniprot Acc . # Q6ZN44 ) on viral parameters . Indeed , intra/extracellular infectivity parameters increased and decreased by 3-fold to 6-fold upon UNC5A depletion or overexpression , respectively ( S14 Fig ) . These results demonstrate that Netrin-1 exerts its pro-HCV effect via inhibition of the UNC5A receptor that itself decreases Netrin-1’s proviral effect . They also indicate that UNC5A-related functions ultimately condition infectivity of the virus particle , through increased viral propagation inducing enhanced Netrin-1 expression . The EGFR ( GenBank Acc . # K03193; Uniprot Acc . # P00533 ) is a host receptor necessary for HCV entry [3 , 4 , 43] that acts by promoting the formation of the CD81-CLDN1 viral capture complex at the level of the membrane [3] . Since the altered signaling of both Netrin-1 and EGFR plays a widespread role in cancer , we examined whether the two proteins might be functionally connected . In order to study the response of EGFR concomitantly to Netrin-1 modulation , our experiments were conducted in EGF stimulation synchronized settings . In this context , siRNA-mediated knockdown of Netrin-1 induced a decrease in the level of EGFR at the plasma membrane level , while forced expression of Netrin-1 resulted in an increase in EGFR levels ( S15A Fig ) . Netrin-1 had no effect on the levels of plasma membrane-associated CD81 ( Uniprot Acc . # P60033 ) , which is one of the co-receptors of HCV ( S15B Fig ) . Working in similar conditions , we then investigated whether the level of EGFR activation was sensitive to Netrin-1 fluctuations . While knockdown of Netrin-1 caused a decrease in the activation of EGFR , as measured by immunoblotting using an anti-phospho1068 antibody , forced expression of Netrin-1 increased EGFR phosphorylation ( S15C Fig ) , an event necessary for HCV entry [43] . These data were also confirmed in serum-containing conditions ( S16 Fig ) and indicate that Netrin-1 participates in EGFR activation . In order to provide a dataset at the functional level regarding the effect of Netrin-1 on the entry of the HCV and the possible role of the EGFR in this process [4] , we used the HCV pseudoparticles ( HCVpp ) system , a pseudotyped HCV glycoprotein-expressing lentiviral tool , widely used for entry quantification assays in hepatitis C research [44] . Experiments were carried out in cells transfected with the Netrin-1 expression plasmid and/or the EGFR siRNA , while cells transfected with the VR1 expression plasmid served as negative controls . SiRNA-mediated EGFR knockdown was tested at the mRNA and protein levels ( Fig 12A and 12B ) . Forced expression of Netrin-1 caused a 2-fold increase in the entry of HCVpp , whereas values were reversed following depletion of the EGFR ( Fig 12C ) . Netrin-1-depletion led to a 3-fold to 4-fold decrease in the entry of HCVpp , a level of inhibition comparable to that observed when cells were transfected with EGFR siRNA ( Fig 12D ) . The entry of positive control lentivirus pseudotyped with VSV-G was insensitive to either treatment . Entry of HCVpp devoid of envelope glycoproteins served as a negative control ( Fig 12C–12F ) . Treatment of cells with the EGFR inhibitor erlotinib interfered with their entry , similarly to interferences observed upon overexpression of Netrin-1 , but did not affect the entry of particles carrying the VSV-G phenotype ( Fig 12E ) . Erlotinib cooperated with siRNAs directed against EGFR for further entry inhibition , as previously shown ( Fig 12F ) [4] . These data support the hypothesis that Netrin-1 heightens the infectivity of an inoculum by increasing the susceptibility of target cells to the entry of HCV . Immunoblotting experiments conducted to detect the EGFR protein in total cell lysates showed that silencing or overexpression of Netrin-1 did not alter protein levels ( S17 Fig ) . Since , in contrast , Netrin-1 was found to modulate the EGFR at the level of the plasma membrane ( S15 and S16 Figs ) , we verified whether Netrin-1 was able to modulate its recycling upon experimental binding with its cognate ligand EGF . RT-qPCR and immunoblotting results revealed that stimulation of Netrin-1-silenced or Netrin-1-overexpressing Huh7 . 5 cells with EGF , for 5 or 15 min , affected neither EGFR mRNA nor protein levels ( S17A and S17B Fig , respectively ) . In contrast , results of flow cytometry confirmed the positive effect of Netrin-1 on the levels of cell surface-exposed EGFR ( S17C Fig , top panel ) , while EGF was used as an internalization control ( S17C Fig , center and bottom panels ) . These results indicate that the EGFR is functional in the experimental setting used in the present study , and also that its global expression level is not regulated by Netrin-1 . We then investigated the interplay between HCV infection and Netrin-1 on the internalization of EGFR . HCV-infected cells were transfected with anti-Netrin-1 ( siRNA ) or a Netrin-1 expressing plasmid . Cells were subsequently serum-starved and incubated with EGF prior to their fixation and indirect immunofluorescence analysis to detect EGFR along with the endosomal marker EEA1 ( Uniprot Acc . # Q15075 ) . The EGFR and EEA1 signals were assessed by intensity correlation coefficient-based ( ICCB ) analyses ( Fig 13A ) , based on the Li coefficient ( also see detailed explanation in S3 Supporting Information ) [45] . Our results reveal a strong colocalization of EGFR and EEA1 in Netrin-1 depleted cells ( i . e . , almost all of the pixels have positive staining amplitude values ) and inversely display low levels of colocalization in cells overexpressing Netrin-1 . The effect of EGF treatment on the internalization of the EGFR in Netrin-1 expression-modulated cells was then assessed by comparing changes in the Li coefficient , which is a measure of the transfer of EGFR to the early endosome . Treatment with EGF , for 5 or 15 min , resulted in an increase in the Li coefficient in the transfected cells . At the 5-min time point , the Li coefficient increased by 35% in cells depleted of Netrin-1 , whereas it showed a 39% decrease in cells overexpressing Netrin-1 ( Fig 13B ) . We showed a partial ( empty arrows ) and total ( solid arrows ) colocalization of EGFR and EEA1 ( Fig 13C ) by immunofluorescence in Netrin-1 overexpressing and depleted cells , respectively . The plot profiles of the fluorescence intensities of EEA1 and EGFR signals were measured across a five-micron line located at the tip of each arrowhead . Spearman correlation coefficients between EEA1 and EGFR fluorescence intensities were calculated from these graphs and yielded values of 0 . 34 and 0 . 78 for cells transfected with the control siRNA and Netrin-1 siRNA , respectively , and 0 . 43 and 0 . 85 for cells transfected with the Netrin-1 plasmid and control plasmid , respectively ( Fig 13D ) . These results suggest that Netrin-1 increases the amount and activation of EGFR at the level of the plasma membrane by impeding the internalization of this receptor . Netrin-1 , therefore , fosters persistence of activated EGFR at the cell surface . This phenotypic alteration increases susceptibility of target cells to viral entry and , thus , leads to an even higher level of HCV infectivity than that induced by Netrin-1 at the level of the virions produced by infected cells .
In this study , we establish a causative relationship between a pro-oncogenic viral infection , namely HCV , and Netrin-1 , an extensively studied dependence receptor ligand , in the context of cancer development . These observations were made in vitro as well as in chronically infected patients suffering from fibrotic liver disease , cirrhosis , or even HCC . We report that the major molecular mechanism underlying the increase in Netrin-1 translation upon HCV infection is regulated by the NS5A and LARP1 proteins and that this increase is particularly marked in the microsomal compartments close to viral budding sites . In turn , we show that the rise in Netrin-1 leads to an increase in the level of infectivity of HCV particles , through both its presence on virions and its implication in the formation/morphogenesis of viral particles . Of note , LARP1 was recently identified as having a pivotal role in general [30 , 31 , 46 , 47] and in hepatic [48] carcinogenesis . Its implication in HCC now needs to be addressed in light of our findings regarding its sensitivity to an oncogenic virus , HCV , and its role in regulating Netrin-1 . The pro-viral effect mentioned above was further enhanced by the indirect effect of Netrin-1 on the persistence of EGFR at the surface of target cells , thus increasing their susceptibility to HCV viral entry . Indeed , Claudin-1/CD81 interactions , which enable HCV entry into hepatocytes [49] , are mediated by the activation of EGFR [3 , 43] . As an HCV virion candidate component and an EGFR activator , Netrin-1 is , thereby , prone to favor the contribution of EGFRs to virion transfer from CD81 to CLDN1 at the level of the membrane [3] . EGFR signaling is implicated in HCC [7] and possibly in the resistance to the anti-HCC drug sorafenib [8] . The fact that Netrin-1 fosters HCV entry through fostering EGFR activation is in agreement with previous reports on the involvement of Netrin-1 in cancers [11 , 41 , 50] , in which dysregulated EGFR expression and signaling play a major role [51] . As for high-throughput approaches and Netrin-1 biology , transcriptomic studies focusing on the regulation of Netrin-1 have so far not been reported . Although previous research has resulted in the identification of cell death-related signatures in HCV+ cells , in vitro , in either human or chimpanzee samples [52–58] , most transcriptomic screens performed in the last 15 y on HCV-related samples were conducted using non-tiling arrays or RNAseq procedures . It is , therefore , not entirely surprising that Netrin-1 , a difficult-to-target GC-rich transcript , has so far not been identified . Our data revealed the occurrence of a positive feedback loop , represented here by the reciprocal induction of Netrin-1 and HCV . Such a positive feedback is relatively scarce in biological systems because of the irreversible imbalance this type of dynamics can rapidly produce . For this reason , our findings raise questions regarding the consequences of an increase in Netrin-1 on HCV infection , beyond HCV persistence per se . Although Netrin-1 expression did not confer a pro-survival advantage to cells used in this study , elevated levels of intrahepatic Netrin-1 may enhance the survival of hepatocytes previously altered and/or rendered more resistant to apoptosis by genetic or tissular injuries , such as HCV trans-acting factors or HCV-related chronic liver regeneration . Netrin-1-related enhancement of cell survival in a cytotoxic context has been observed in non-hepatic cancers following chemotherapy [59] . It is likely to be most relevant at the cirrhosis level , a harsh environment for hepatocytic survival , which displays the highest level of Netrin-1 signals , and even more so upon HCV infection . Importantly , previous studies demonstrated that the expression and activation of EGFR and its main dimerization partner HER-3 ( ErbB-3 ) are frequently dysregulated in HCC [60–62] . Other investigations also indicated that EGFR and the dimerization inducer HER-3 may provide compensatory signals for cancer cells to escape targeted therapies in the liver [63] . Hence , our observation that EGFR undergoes functional upregulation by Netrin-1 indicates that it may mediate the potentially deleterious effects of Netrin-1 in liver pathologies . A relevant study on the role of Netrin-1 in fostering liver regeneration upon ischemia/reperfusion in mice livers [64] has been most recently published . We believe this study emphasizes the relevance of our work , since it concatenates the consequences of elevated Netrin-1 levels at the hepatic level with the Netrin-1 inducer status of HCV . Indeed , HCV may induce Netrin-1 for its own replicative benefit in a direct manner , but also in a liver maintenance prospective . EGFR-mediated interferon resistance has been evidenced in the context of hepatitis C [65] . From the therapy perspective , it is widely accepted that treatments targeting HCV have undergone major improvements in recent years [66] , although cirrhotic patients may in some cases remain more exposed to treatment failure than a majority of patients [67 , 68] . In this context , elevated levels of Netrin-1 in HCV+ cirrhotic patients should orient future research efforts not only in the direction of the onset of HCC but also of Netrin’s potential ability to condition the efficacy of direct-acting antivirals ( DAAs ) through a dysregulated Netrin-EGFR-endogenous interferon sensitive axis . In summary , the HCV-Netrin-1 amplification loop studied herein is composed of two molecularly distinct Netrin-1-mediated arms , which converge toward a single phenotype of increased infectivity conferred to HCV particles . The involvement of Netrin-1 in hepatic pathobiology and the role of DR ligands in the persistence of infectious agents associated with cancer deserve further investigation .
Samples and PHHs were used according to the French IRB “Comité de Protection des Personnes ( CPP ) Sud-Est IV” agreement #11/040 , obtained in 2011 . Written informed consent was obtained from patients . PHHs were prepared , grown in William’s E medium , and infected with cell culture-adapted HCV as described previously [24] . The human hepatocyte cell line Huh7 . 5 was grown in DMEM ( Life technologies ) and supplemented with 10% fetal bovine serum ( FBS; Thermo Scientific ) and 1% penicillin-streptomycin ( Life technologies ) . Twenty thousand cells per square centimeter were infected with HCV JFH1 [27] at an MOI of 0 . 1 ( proliferative cells ) or 0 . 05 ( differentiated cells ) . In order to induce differentiation , 2% DMSO ( Sigma ) was added to the medium . Netrin-1-Fc ( 125 ng/mL ) was obtained from Apotech Corp . /Axxora . For Netrin-1 mRNA stability assays , cells were treated with either DRB ( 25 μg/mL , Sigma-Aldrich ) or actinomycin D ( 5 μg/mL , Sigma-Aldrich ) in order to inhibit transcription for the various durations of time prior to total RNA isolation using the Extract-all reagent ( Eurobio ) . Total RNA was extracted using the Nucleospin RNA/protein kit ( Macherey-Nagel ) for biopsies and the Extract-all reagent ( Eurobio ) for cultured cells . One μg of RNA was DNAse I-digested ( Promega ) and reverse transcribed with 5% DMSO , using the MMLV enzyme , according to the manufacturer’s instructions ( Invitrogen ) . Real-time quantitative RT-PCR was performed on a LightCycler 480 device ( Roche ) using the iQ™ SYBR®Green Supermix ( BIO-RAD ) . DMSO ( 10% , Sigma-Aldrich ) was added to the PCR reaction for Netrin-1 quantification . All PCR primer sequences and qPCR conditions are reported in the S2 Table . The partitioning of Netrin-1 , GUS , and PMM1 mRNA between the cytosol ( free polysomes ) and ER membrane-bound polysome compartments was performed as described by Stephens et al . [69] . GUS and PMM1 mRNAs were used to assess the quality of each fraction . GUS mRNA is membrane-enriched [70] , whereas PMM1 mRNA is not . Nucleocytoplasmic fractionation was performed as described previously [71] . Two hundred thousand cells in ten-square-centimeter wells were transfected either with 2 . 5 μg Netrin-1 or with the neuronal vanilloid receptor VR1 expressing plasmid , or with the F-Luc-bearing Netrin-1 promoter reporter construct described previously [20] using the TransIT-LT1 reagent ( Mirus Bio ) , following the manufacturer’s instructions . ORF encoding LARP1 was picked from the Human ORFeome library ( Open biosystems ) and recombined into pGLuc vector ( for luciferase tagging ) [33] by gateway technology ( Invitrogen ) . ORFs encoding HCV proteins were amplified from described vectors [72] and recombined into pGLuc vector and/or pFlag vector . Sequences of relevant primers are available upon request . As described [33 , 34] , combinations of plasmids encoding prey ( A ) and bait ( B ) proteins , each fused to a fragment of the Gaussia princeps luciferase protein ( GLuc1 and GLuc2 ) or control vectors , were cotransfected into Huh-7 . 5 cells plated in 96-well plates in triplicates . At 24 hr post-transfection , cells were lysed and subjected to luciferase assays ( Promega ) . Results were expressed as normalized luminescence ratios ( NLR ) : the average luminescence signal in cells transfected with GLuc1-A and GLuc2-B divided by the average signal in wells transfected with GLuc1-A and an empty GLuc2 vector and those transfected with GLuc2-B and an empty GLuc1 vector . We benchmarked the sensitivity and accuracy of this screen by including a random reference set ( RRS ) composed of 53 noninteracting human protein pairs and a set of host factors known to interact with various HCV proteins [33] . Cell culture supernatants collected at the indicated time points were loaded on top of 10%–60% sucrose gradients and centrifuged at 38 , 000 rpm at 4°C for 16 h in a SW41 rotor ( Beckman ) . A total of 12 fractions ( 1 ml each ) were collected and their densities were determined with a refractometer ( Euromex ) . HCV RNA in each fraction ( 150 μl ) was extracted with the Nucleospin RNA Virus kit ( Macherey-Nagel ) and quantified by RT-qPCR . The infectivity of HCV virions in each fraction was determined on Huh7 . 5 cells using the TCID50 protocol [73] as described below . Twenty thousand cells per square centimeter were transfected with various concentrations ( 12 . 5 , 25 , and 50 nM final concentration ) of a nontargeting control siRNA , Netrin-1 siRNA , LARP1 siRNA , or EGFR siRNA ( Sigma-Aldrich ) using Lipofectamine 2000 ( Invitrogen ) , according to the manufacturer’s instructions . siRNAs sequences are listed in S3 Table . The level of infectivity of the HCV produced in cell culture was measured following the TCID50 protocol outlined by Lindenbach [73] . A human HCV antiserum at a 1/500 dilution ( initially validated against the anti-HCV capsid C7/50 clone ( Abcam ) in a double immunofluorescence assay ) and a goat anti-human Alexa 488 secondary antibody ( Invitrogen ) were used , at a concentration of 1 μg/mL . Cells were counterstained with DAPI and examined with a Nikon TE-2000E epifluorescence microscope . Titers were calculated using the Reed and Muench method [73] . RIP-Chip was performed as described in Keene et al . [74] , using control IgG and anti-LARP1 antibodies ( Novus Biologicals ) and protein G magnetic beads ( Millipore ) prior to RNA extraction and RT-qPCR . For direct or indirect protein immunoprecipitation , Huh7 . 5 cells were lysed in RIPA extraction buffer . Lysates were incubated with anti-LARP1 antibody at 4°C for 1 h . Protein G magnetic beads were then added to the antibody/lysate mixture and incubated for 45 min before immunoblotting . Caspase-3 activity assays were performed using the Caspase 3/CPP32 Colorimetric Assay Kit , according to manufacturer’s instructions ( Gentaur Biovision ) . The cell proliferation assay was performed using neutral red uptake as described by Repetto et al . [75] . Formalin-fixed , paraffin-embedded liver samples were sectioned at a thickness of 4 μm . After deparaffinisation and rehydration , tissue sections were unmasked in citrate buffer ( pH 9 ) in a 96°C water bath for 50 min . To block endogenous peroxidases , slides were incubated in 5% hydrogen peroxide in sterile water . Slides were then incubated at room temperature for 1 h with a polyclonal goat antibody recognizing human Netrin-1 ( R&D Systems ) , diluted 1/800 in an antibody diluent solution ( Dako ) . For HCV immunostaining , slides were incubated overnight at 4°C with a human anti-HCV E2 antibody [23] diluted 1/50 . After three washes in PBS , slides were incubated with a biotinylated secondary antibody bound to a streptavidin peroxidase conjugate ( Lsab+ kit , Dako ) . Netrin-1 and HCV staining were contextualized by DAB staining . Nuclei were counterstained with hemalun . HCVpp and their VSV-G and Env-negative controls were produced as described previously [44] . A construct package containing the Renilla luciferase gene under the control of the cytomegalovirus ( CMV ) promoter ( pMLVΨCMV-Luc ) was used as a reporter . Huh7 . 5 cells transduced with virions containing the luciferase transgene were lysed 3 d post-infection , and luciferase activity was monitored by using the Renilla luciferase assay system according to the manufacturer’s instructions ( Promega ) . Huh7 . 5 cells were detached with Versene buffer , washed in PBS and centrifuged at 1 200 r . p . m for 5 min . Cells were then stained with an EGFR antibody ( Calbiochem ) at a 1/1000 dilution and a mouse anti-human Alexa 488-conjugated secondary antibody ( Invitrogen ) . EGFR membrane expression was analyzed using a FACscalibur ( BD ) . Immunoblotting was performed using standard protocols with antibodies against the HA-tag ( Sigma-Aldrich ) , actin ( Sigma-Aldrich ) , total EGFR ( Millipore ) , phosphorylated ( position 1068 ) EGFR ( Cell Signaling ) , Netrin-1 ( R&D System ) , core ( Abcam ) , HSP60 ( Bd Biosciences ) , and LARP1 ( Novus ) . Cells were fixed in 4% paraformaldehyde for 10 min and permeabilized for 30 min in 0 . 15% Triton-X100 in PBS + 3% BSA . Incubation with primary antibodies was performed for 2 hrs in 0 . 15% Triton-X100 in PBS + 3% BSA at room temperature . dsRNA , EGFR , and EEA1 immunolocalization was performed using anti-dsRNA J2 ( Scicons ) [76] , anti-EEA1 , anti-Calnexin , anti-Puromycin ( BD Biosciences ) , anti-LARP1 ( Novus Biologicals ) , anti-NS5A 9E10 ( gift from C . Rice ) , and anti-EGFR ( Millipore ) antibodies ( 2 μg/mL ) . For puromycylation assays , 200 μM emetin ( Sigma ) and 90 μM puromycin ( Sigma ) were incubated on cells for 10 min before harvest . Samples were incubated with Alexa-488-goat anti-mouse and Alexa-594-goat anti-rabbit antibodies ( Invitrogen , 1 μg/mL ) for 1 h at room temperature . Nuclei were counterstained with Hoechst 33342 . Images were acquired using a Leica SP5X confocal microscope equipped with LAS AF software . Subsequent analyses were performed using the JACop ImageJ colocalization plugin ( http://rsb . info . nih . gov/ij/plugins/track/jacop . html ) and its plot profile function . Naïve Huh7 . 5 cells were transfected with Netrin-1 or VR1-expressing plasmids and infected with HCV for 4 d . The supernatant was first clarified at 8 , 000 g for 15 min , then ultracentrifugated on a 20% sucrose / 1X TNE cushion for 4 h , minimizing the carryover of soluble , non-viral material . The pelleted virions were collected in fresh culture medium and incubated with two distinct anti-Netrin-1 antibodies ( clones #2F5 from Netris Pharma and #AF1009 from R&D Systems ) or DCC-Fc ( recombinant receptor of Netrin-1 from R&D Systems ) overnight , at a concentration of 10 μg/ml . To quantify the level of viral infectivity , we performed a TCID50 assay following the TCID50 protocol . Naïve Huh7 . 5 cells were seeded in a 96-well plate at a density of 20 , 000 cells per square centimeter and infected the day after with the neutralized virions at serial dilutions of 10−1 to 10−8 ( one dilution per line ) . Hence , each biological sample was processed in the context of 12 technical replicates . Immunofluorescence staining was performed 3 d post-infection using a human HCV antiserum at a 1/500 dilution and a goat anti-human Alexa 488 secondary antibody at a concentration of 1 μg/mL . Positive wells were counted with a Nikon TE-2000E epifluorescence microscope and titers were calculated using the adapted Reed and Muench method [77] as considered by Lindenbach [73] .
|
Viruses and bacteria are implicated in 15%–20% of total cancer occurrences . Hepatitis C virus ( HCV ) infection is one of the main causative agents of liver cancer . “Dependence receptors” are a class of receptors that auto-activate and trigger apoptosis in the absence of their ligands , and “dependence receptor” ligands such as Netrin-1 are known to be overactivated in cancers , especially in inflammation-driven tumors . In this study , we show that HCV and Netrin-1 are mutual inducers—Netrin-1 expression is increased upon HCV infection and , in turn , the rise in Netrin-1 leads to an increase in the HCV particle infectivity . The effects on HCV infectivity involve the liver cancer-related epidermal growth factor receptor ( EGFR ) , which is known to be a host receptor necessary for HCV entry and which we now show is activated by Netrin-1 . Our work , therefore , illustrates a pathogenic positive feedback loop involving HCV , Netrin-1 , and EGFR , among other factors , in association with cancer development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"molecular",
"probe",
"techniques",
"gene",
"regulation",
"hepacivirus",
"pathogens",
"messenger",
"rna",
"microbiology",
"immunoblotting",
"viral",
"structure",
"viruses",
"rna",
"viruses",
"immunoprecipitation",
"molecular",
"biology",
"techniques",
"immunologic",
"techniques",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"hepatitis",
"c",
"virus",
"hepatitis",
"viruses",
"molecular",
"biology",
"virions",
"precipitation",
"techniques",
"biochemistry",
"rna",
"immunohistochemistry",
"techniques",
"immunostaining",
"nucleic",
"acids",
"flaviviruses",
"virology",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"histochemistry",
"and",
"cytochemistry",
"techniques",
"organisms"
] |
2016
|
Epidermal Growth Factor Receptor-Dependent Mutual Amplification between Netrin-1 and the Hepatitis C Virus
|
Cell adhesion molecules and downstream growth factor-dependent signaling are critical for brain development and synaptic plasticity , and they have been linked to cognitive function in adult animals . We have previously developed a mimetic peptide ( FGL ) from the neural cell adhesion molecule ( NCAM ) that enhances spatial learning and memory in rats . We have now investigated the cellular and molecular basis of this cognitive enhancement , using biochemical , morphological , electrophysiological , and behavioral analyses . We have found that FGL triggers a long-lasting enhancement of synaptic transmission in hippocampal CA1 neurons . This effect is mediated by a facilitated synaptic delivery of AMPA receptors , which is accompanied by enhanced NMDA receptor-dependent long-term potentiation ( LTP ) . Both LTP and cognitive enhancement are mediated by an initial PKC activation , which is followed by persistent CaMKII activation . These results provide a mechanistic link between facilitation of AMPA receptor synaptic delivery and improved hippocampal-dependent learning , induced by a pharmacological cognitive enhancer .
Activity-dependent synaptic changes , generally termed synaptic plasticity , underlie multiple forms of cognitive function , such as learning and memory [1] . Strong interest has accrued in understanding the molecular and cellular mechanisms underlying these changes . Additionally , it is believed that targeted manipulation of these mechanisms may help facilitate or stabilize synaptic plasticity events , with the aim of potentially improving cognitive function under pathological , or even physiological , conditions . Multiple genetic manipulations in animal models have been shown to produce cognitive enhancement , defined as improved performance in learning and memory behavioral tasks ( see , for example , [2] ) . In the vast majority of cases , cognitive-enhancing mutations are related to signaling mechanisms associated with synaptic plasticity ( reviewed in [3] ) , thus reinforcing the interpretation of synaptic plasticity as a cellular substrate for learning and memory . Nevertheless , the particular mechanism that links changes in synaptic plasticity with enhanced cognitive function is poorly defined . Additionally , for therapeutic purposes , there is great interest in developing pharmacological approaches , rather than genetic manipulations , that effectively modulate synaptic plasticity pathways in a well-defined manner . Cell adhesion molecules are well-known effectors of neuronal development and structural plasticity [4] . Some of them have also been linked to synaptic plasticity , learning , and memory [5] , particularly via their interaction with growth factor-mediated signaling [6] . From this point of view , cell adhesion molecules are being considered as potential therapeutic targets for the development of pharmacological cognitive enhancers . This is the case of neural cell adhesion molecule ( NCAM ) [5] . NCAM activity is essential for both early synaptogenesis and synaptic maturation [4] , and it influences the strength of excitatory synapses in an activity-dependent manner [7] . NCAM is a member of the immunoglobulin ( Ig ) superfamily , containing five N-terminal Ig modules followed by two fibronectin type III ( F3 ) modules . NCAM is involved in homophilic and heterophilic interactions , as well as in the activation of various signal transduction pathways [8] . Importantly , some of the functions of NCAM in cell remodeling and growth are mediated by fibroblast growth factor receptors ( FGFRs ) . In fact , NCAM interacts with extracellular domains of FGFR to modulate FGFR-dependent intracellular signaling [9] . Based on the functional interplay between NCAM and FGFR , we engineered a synthetic NCAM mimetic peptide , termed FGLoop ( FGL ) , that encompasses the interaction domain of NCAM with FGFR: F and G β-strands and the interconnecting loop of the second F3 module of NCAM ( red ribbon in Figure 1A ) . We previously showed that FGL elicits FGFR-mediated signaling [10] and induces neuritogenesis and survival in neuronal cultures [11] . Most importantly , we found that in vivo administration of FGL enhances spatial and social memory retention in rats [12] , [13] . FGL has also been shown to prevent cognitive impairment induced by stress [14] , [15] and oligomeric β-amyloid [16] , and have antidepressant-like effects in rats [17] . Therefore , FGL appears to act as a bona fide cognitive enhancer , possibly by engaging NCAM-FGFR-related signaling . However , the synaptic mechanisms by which this cognitive enhancement is produced remain unknown . Some of the signaling pathways recruited by FGFR activation [18] cross-talk with molecular mechanisms associated with synaptic plasticity , particularly long-term potentiation ( LTP ) . LTP is one of the best characterized forms of synaptic plasticity in the hippocampus and is considered to be a cellular correlate for information storage in the brain during learning and memory processes [1] . A major contributor to synaptic potentiation during LTP is the incorporation of new AMPA-type glutamate receptors ( AMPARs ) into excitatory synapses via activity-dependent trafficking [19] . Multiple signal transduction pathways have been shown to regulate AMPAR incorporation into synapses during LTP , most notably the pathways controlled by Ca2+/calmodulin-dependent protein kinase II ( CaMKII ) [20] , mitogen-activated protein kinase ( MAPK ) [21] , protein kinase C ( PKC ) [22] , and phosphoinositide-3-kinase ( PI3K ) [23] , [24] . These last three pathways ( MAPK , PKC , and PI3K ) are engaged during FGFR-dependent signaling [18] , and therefore , they are attractive candidates to mediate FGL effects on cognitive function . The present study uncovers specific synaptic mechanisms and signaling pathways responsible for the cognitive enhancement induced by the NCAM-FGFR agonist FGL . In particular , we show here that FGL enhances LTP in hippocampal slices , and it does so by facilitating AMPAR delivery at synapses upon activation of NMDA receptors ( NMDARs ) . These effects are specifically mediated by PKC activation . Furthermore , behavioral testing revealed that this PKC-dependent mechanism mediates the enhanced cognition induced by FGL . Therefore , these results delineate the intracellular signaling and molecular mechanisms that lead to enhanced synaptic plasticity and improved learning and memory caused by a pharmacological cognitive enhancer . In this manner , we have established a mechanistic link between facilitation of AMPAR synaptic delivery and enhanced cognition .
To start characterizing the signaling pathways that mediate the cognitive actions of FGL , we tested the ability of FGL to activate FGFR-dependent signaling in vitro and in vivo . We found that FGL dose-dependently induces FGFR1 phosphorylation in Trex293 cells transfected with human FGFR1 ( Figure 1B ) . Moreover , FGL triggers FGFR1 phosphorylation in vivo in the hippocampus after subcutaneous injection ( Figure 1C , black columns ) ( we have previously shown that FGL crosses the blood-brain barrier [13] ) . As a control , FGL did not induce phosphorylation of TrkB ( Figure 1C , gray columns ) , the receptor of brain-derived neurotrophic factor ( BDNF ) , which is a potent modulator of activity-dependent synaptic plasticity and shares some signaling pathways with FGFR [25] . Downstream from FGFR phosphorylation , we found in transfected Trex293 cells that FGL triggers the phosphorylation of PLCγ ( phospholipase C-γ ) and Shc ( Src homologous and collagen ) ( Figure 1D and E ) , but not FRS2 ( FGFR substrate 2 ) ( Figure 1F , black columns ) , in contrast to FGF1 ( Figure 1F , gray columns ) . Thus , FGL activates a subset of the signaling pathways triggered by FGF . After determining that FGL activates FGFR in vivo and in vitro , we evaluated its cognitive actions . FGL ( 6 . 6 mg/kg ) or control vehicle ( 0 . 9% NaCl ) were injected subcutaneously 5 and 2 d before the evaluation of cognitive function . Rats were trained to find a submerged platform in the Morris water maze ( see Text S1 ) . Spatial training was performed with the experimenter blind to the treatment of each rat . As shown in Figure 2A , FGL-treated rats outperformed their age-matched controls in this spatial learning task during the 2 d of training ( reflected by significantly shorter distances swam to find the hidden platform during the training sessions: F1 , 18 = 8 . 445 , p = 0 . 004 ) . This was particularly evident for all tested animals when comparing individual performance during the last trial of each training day ( Figure 2B ) . No significant differences were found in swimming speed between groups ( unpublished data ) , suggesting that FGL does not have peripheral effects . Therefore , we conclude that FGL triggers FGFR-dependent signaling in the hippocampus in vivo and improves hippocampal-dependent learning when injected peripherally . Most glutamatergic excitatory axons establish synapses with dendritic spines of pyramidal neurons , and changes in their density and/or shape are involved in plastic modifications associated with LTP between CA3 and CA1 pyramidal neurons [26] . Additionally , most CA1 pyramidal neurons express FGFR1 [27] and have been strongly implicated in spatial navigation and memory [28] . Thus , as a first step to evaluate the cellular substrate of FGL-induced enhanced cognition , we examined whether dendritic spines in CA1 stratum radiatum are affected by FGL administration . To this end , animals were treated with FGL as described above , and 2 d after the second FGL injection ( when cognition enhancement was observed ) , the animals were perfused with fixative , hippocampal sections were prepared , and CA1 neurons were injected with Lucifer Yellow for analysis of dendritic spine density and morphology ( Figure 3A–B ) . These analyses were performed by experimenters blind to the treatment of the animals . We found no significant differences in spine density ( Figure 3C–E ) , spine head volume ( Figure 3F–J ) , or neck length ( Figure S1 ) between FGL- and vehicle-treated animals . Furthermore , sections adjacent to those used for the intracellular injections were examined at the electron microscope level . As shown in Figure 3K–M , we did not detect changes in synaptic density or cross-sectional lengths of the synaptic junctions , quantified from electron photomicrographs using the size-frequency method [29] , [30] . Thus , the cognitive enhancement induced by FGL is not associated with detectable structural changes in CA1 stratum radiatum synapses . We then reasoned that the effect of FGL on cognition may result from functional rather than structural changes in hippocampal synapses . To test the effect of FGL on synaptic transmission , we added FGL ( 10 µg/ml ) to the medium of organotypic hippocampal slice cultures . After 24 h , the culture medium was replaced with fresh medium without FGL , and the slices were kept in culture for an additional 24 h before electrophysiological recordings ( see Materials and Methods ) . Therefore , the electrophysiological responses were evaluated 48 h after the onset of the FGL treatment . This regimen was intended to mimic the in vivo behavioral experiments , in which spatial learning is tested long after FGL has been cleared from cerebrospinal fluid ( CSF ) [13] . We placed the stimulating electrodes over Schaffer collateral fibers and recorded CA3-to-CA1 synaptic transmission . Synaptic responses were evoked at −60 mV and +40 mV holding potentials , in the presence of the γ-aminobutyric acid-A ( GABAA ) receptor antagonist picrotoxin to obtain separate AMPAR- and NMDAR-mediated responses , and we calculated the ratio between these values ( AMPA/NMDA ratio ) . As shown in Figure 4A , the AMPA/NMDA ratio of synaptic responses significantly increased after FGL treatment compared with control neurons . Similarly , we obtained AMPA/GABA ratios by evoking synaptic responses at −60 mV and 0 mV holding potentials in the presence of the NMDAR antagonist AP5 . Similar to AMPA/NMDA ratio , we found a significant increase in the AMPA/GABA ratio after FGL treatment compared with control neurons ( Figure 4B ) . Finally , we calculated NMDA/GABA ratios by recording NMDA responses at −60 mV in the absence of Mg2+ and in the presence of CNQX ( AMPA receptor antagonist ) , and GABA responses at 0 mV . As shown in Figure 4C , NMDA/GABA ratios were unaltered by the FGL treatment . Additionally , FGL did not change the presynaptic properties of excitatory transmission assessed by paired-pulse facilitation ( Figure 4D ) , or the passive membrane properties of the neuron , assessed by their input resistance and holding current ( Figure S2 ) . In conclusion , these results suggest that FGL produces a functional postsynaptic change at excitatory CA1 synapses , specifically an increase in AMPAR-mediated synaptic transmission . Enhanced AMPAR synaptic responses may be attributable to an increased number of AMPARs at synapses or a functional modification of preexisting synaptic receptors . To directly determine whether FGL induces the delivery of new AMPARs into synapses , we expressed the GluA1 subunit of AMPARs tagged with GFP ( GluAl-GFP ) in CA1 neurons in organotypic hippocampal slice cultures ( Figure 5A–B ) . We employed this subunit because it has been previously shown that newly synthesized GluA1-containing AMPARs are not spontaneously inserted at synapses , unless driven by strong synaptic stimulation or activation of specific signaling pathways associated with LTP induction [22] , [31] . In addition , overexpressed GluAl-GFP subunits form homomeric AMPARs , whose presence at synapses can be assessed from their inward rectification properties ( electrophysiological tagging [31] ) . GluA1-GFP was expressed in organotypic slice cultures for 60 h ( 24 h of FGL treatment plus 36 h in fresh medium ) , and synaptic delivery was quantified as an increase in the ratio of the evoked postsynaptic current at −60 mV relative to the current at +40 mV ( rectification index , RI = I−60/I+40 ) , in the presence of the NMDAR antagonist AP5 . We found that FGL treatment increased the rectification index in neurons that express GluA1-GFP ( Figure 5C ) . This result strongly suggests that FGL induces synaptic delivery of AMPARs . To note , FGL treatment alone did not change the rectification index in the absence of GluA1-GFP expression ( Figure 5C , “FGL uninfected” versus “Untreated control” ) , indicating that FGL does not alter the intrinsic rectification properties of endogenous AMPA receptors . Also , GluA1-GFP is not delivered spontaneously into synapses [31] . Altogether , these results indicate that FGL enhances excitatory synaptic transmission by inducing the insertion of new AMPARs at synapses . It is also important to point out that FGL is removed from the culture medium from 24 to 36 h before the electrophysiological recordings , indicating that FGL induces long-lasting changes in synaptic transmission . After establishing the specific synaptic modification produced by FGL ( enhanced AMPAR synaptic delivery ) , we identified the underlying signal transduction mechanism . Hippocampal slices were treated with FGL while blocking the three pathways downstream from FGFR activation that may modulate AMPAR trafficking ( i . e . , the MAPK , PI3K , and PKC pathways ) . As a reporter for AMPAR synaptic delivery , we monitored inward rectification of synaptic transmission in GluA1-GFP-expressing neurons , described above . Incubation of the slice culture in 25 µM PD98059 , a potent inhibitor of MAPK kinase ( MEK [32] ) , did not block the increase in the rectification index induced by FGL in GluA1-GFP-expressing neurons ( compare Figure 5C , D ) . Therefore , we conclude that FGL-induced AMPAR synaptic delivery does not require MAPK activation . Similarly , incubation in 10 µM LY294002 , a potent and selective PI3K inhibitor [33] , did not prevent GluA1 synaptic delivery ( Figure 5E ) . In contrast , 5 µM chelerythrine , a general inhibitor of PKC [34] , did block the FGL-induced increase in rectification ( Figure 5F ) ( the efficacy of chelerythrine as a specific PKC inhibitor is evidenced from the blockade of MARCKS phosphorylation and its translocation from the plasma membrane to the cytosol in response to PKC activation; see Figure S3 ) . This finding indicates that FGL enhances AMPAR synaptic delivery in a PKC-dependent manner . Moreover , incubation of slices with a selective inhibitor of classical PKC isoforms ( α , β , and γ ) ( 200 nM GF109203X [35] ) also blocked the increase in the rectification index ( Figure 5G ) , whereas a specific inhibitor of atypical PKC isoforms ( 1 µM zeta inhibitory peptide [ZIP] ) did not ( Figure 5H ) . Therefore , these findings indicated that of the multiple signaling pathways potentially triggered by FGL , the synaptic delivery of AMPARs with the consequent potentiation of synaptic transmission is mediated by PKC activation , specifically by classical PKC isoforms . These results are consistent with the failure of FGL to induce the phosphorylation of FRS2 ( Figure 1F ) , which acts as an important docking platform for the activation of PI3K and MAPK pathways [36] , [37] . The synaptic incorporation of AMPAR induced by FGL may be directly driven by PKC activation and AMPAR phosphorylation [22] or , alternatively , may be an activity-dependent process that is facilitated by FGL . This is an important distinction , because a cognitive enhancer would be expected to modulate synaptic function in a synapse-specific manner . The classic paradigm for activity-dependent synaptic delivery of AMPARs is NMDAR-dependent LTP . Therefore , we hypothesized that FGL induces an LTP-like process in response to spontaneous synaptic activity in hippocampal slices . Notable , spontaneous activity in organotypic slice cultures is typically not sufficient to induce the synaptic delivery of AMPARs [31] . To determine whether FGL-induced synaptic potentiation resembles conventional LTP processes , AMPAR delivery was examined by monitoring inward rectification ( described for Figure 5 ) after blocking NMDARs or CaMKII . We found that 100 µM AP5 , a competitive NMDAR antagonist , completely blocked the FGL-induced delivery of tagged AMPARs into synapses ( Figure 6A ) . Similarly , 20 µM KN-93 , a potent inhibitor of CaMKII catalytic activity [38] , blocked FGL-induced AMPAR synaptic delivery ( Figure 6B , left panel ) . In contrast , the inactive analog KN-92 did not block AMPAR synaptic delivery , as detected by the increase in the rectification index in FGL-treated slices ( Figure 6B , right panel ) . Therefore , FGL-triggered AMPAR delivery depends on the NMDAR and CaMKII activity . Since FGL-induced synaptic potentiation appears to mimic classic NMDAR-dependent LTP , we decided to test whether FGL affects this form of synaptic plasticity . Similar to the previous experiments , slices were pretreated with FGL for 24 h . The culture medium was then changed ( without FGL ) , and recordings were performed 24 h later . Therefore , as in the previous experiments , electrophysiological experiments start 48 h after the onset of FGL treatment . LTP was induced in CA1 neurons by pairing 3 Hz presynaptic stimulation of the Schaffer collaterals with 0 mV postsynaptic depolarization as previously described [31] . As shown in Figure 6C–E , LTP induction significantly increased AMPAR-mediated responses in both FGL-treated and untreated neurons . Nevertheless , FGL treatment dramatically enhanced LTP expression ( 3 . 5-fold potentiation with FGL versus 2-fold potentiation in control neurons; see also Figure S4 for a complete distribution of individual LTP experiments ) . To determine whether FGL-induced LTP enhancement occurs through mechanisms similar as FGL-induced GluA1 synaptic delivery , we tested the role of PKC activation in this process . Slices were incubated with 5 µM chelerythrine together with FGL . Electrophysiological recordings were then performed without chelerythrine because PKC activity is required for LTP induction [39] . Notably , the magnitude of LTP after treatment with FGL in the presence of chelerythrine ( Figure 6C–E , “FGL+Chel” ) was indistinguishable from LTP in untreated neurons , suggesting that PKC activity is required for the enhancing effect of FGL . As a control , neurons treated with chelerythrine alone had similar LTP levels as untreated neurons ( Figure 6C–E , “Chel” ) , indicating that prior PKC activity is not required for subsequent LTP induction . Finally , FGL did not have any effect on the non-potentiated ( unpaired ) pathway ( Figure 6F ) . As mentioned above , FGL ( and chelerythrine ) were washed out from the slices 24 h before performing the LTP experiments . Therefore , these data indicate that FGL produces a long-lasting , PKC-dependent enhancement of LTP . After having established the effect of FGL on synaptic potentiation , we wished to test whether FGL affects LTD , another NMDAR-dependent form of synaptic plasticity . Similar to the previous experiments , slices were pretreated with FGL for 24 h . The culture medium was then changed ( without FGL ) , and recordings were performed 24 h later . LTD was induced in CA1 neurons by pairing 1 Hz presynaptic stimulation of the Schaffer collaterals with moderate postsynaptic depolarization ( −40 mV ) , as previously described [40] . As shown in Figure S5 , LTD induction produced a similar decrease in AMPAR-mediated responses in both FGL-treated and untreated neurons . Therefore , we conclude that the enhancement of synaptic plasticity produced by FGL is specific for synaptic potentiation . It has been reported that NMDARs with different subunit composition play specific roles in different forms of synaptic plasticity , although the details of this specificity are still under debate [41] , [42] . As an initial attempt to test whether FGL may alter the subunit composition of NMDARs , we evaluated the kinetics of NMDAR-mediated responses , since NR2A- and NR2B-containing NMDARs display distinct decay time constants [43] . As shown in Figure S6 , NMDA responses from FGL-treated neurons displayed faster decay kinetics than those from untreated neurons . However , this effect was not blocked by incubation with chelerythrine to inhibit the PKC pathway ( Figure S6 , “FGL+Chel” ) . Therefore , even if FGL alters NMDAR subunit composition , this effect does not appear to be mechanistically linked to the enhancement of LTP and AMPAR synaptic delivery , which require PKC activity . Therefore , we have not pursued this effect any further . Synaptic potentiation has been shown to be accompanied by an increase in spine size and recruitment of polymerized actin into the spine head [44] . Therefore , we tested whether FGL enhances this form of structural plasticity as it enhances synaptic potentiation . To this end , we expressed GFP-actin in organotypic hippocampal slices , and then induced LTP using a pharmacological approach that allows us to maximize the number of synapses undergoing plasticity , while mimicking biochemical and electrophysiological properties of electrically induced LTP [45] . As shown in Figure S7 , LTP induction led to an increase in actin-GFP recruitment into spines , which was similar in extent between control and FGL-treated spines ( these experiments were analyzed blind with respect to the treatment the slices had received ) . Therefore , FGL treatment enhances synaptic potentiation , without altering the capacity of spines to undergo structural plasticity , at least as monitored by actin recruitment into the spine head . The experiments above indicate that the enhancing effect of FGL on synaptic plasticity does not require the continuous presence of FGL because it can last at least 24 h after FGL removal . This is also consistent with the behavioral effects of FGL treatment , which were manifested 2 d after FGL application ( Figure 2 ) . Therefore , we sought long-lasting biochemical signatures of FGL treatment by preparing whole-cell extracts from hippocampal slices at different times after adding FGL to the culture medium , and after its removal 24 h later . We then used Western blot to evaluate three key events linked to upstream FGL action and downstream LTP-related signaling: ( i ) PKC activation ( presumably triggered directly by FGL upon FGFR activation and PLCγ phosphorylation; Figure 1D ) , ( ii ) CaMKII phosphorylation ( downstream from NMDAR opening and key mediator of LTP expression [46] ) , and ( iii ) GluA1 Ser831 phosphorylation ( triggered during LTP upon CaMKII activation [47] but also catalyzed by PKC [48] ) . To evaluate global PKC activation , we monitored the phosphorylation of multiple PKC substrates with an antibody that recognizes a phospho-Ser PKC substrate motif . Phosphorylation of CaMKII at Thr286 and GluA1 at Ser831 was monitored with the corresponding phospho-specific antibodies ( see Materials and Methods ) . As shown in Figure 7A , FGL application led to rapid ( 5–20 min ) upregulation of the PKC pathway , which was expected from the activation of FGFR and PLCγ ( Figure 1D ) . Moreover , this pathway remained activated throughout the FGL treatment and 24 h after its removal ( Figure 7A , the “48 h” column represents slices treated with FGL for 24 h and then transferred to fresh medium without FGL for 24 h more ) . Interestingly , CaMKII activation ( monitored by Thr296 phosphorylation ) was transiently decreased by FGL ( 20–30 min after application ) , but it showed a gradual upregulation at late time-points , including 24 h after FGL removal ( Figure 7B , gray columns ) . Finally , GluA1 phosphorylation at Ser831 was induced early after the addition of FGL but also persisted by the end of the time course ( Figure 7C , gray columns ) . The total levels of CaMKII and GluA1 did not significantly change during or after FGL application ( Figure 7B–C , black columns ) . These results strongly suggest that FGL initiates signaling mechanisms that outlast the initial triggering events . These mechanisms may be related to a sustained enhanced synaptic activity in the slices , as a consequence of GluA1 delivery . We showed that enhanced synaptic delivery of AMPARs and LTP are induced by FGL in a PKC-dependent manner . We then hypothesized that if facilitated AMPAR synaptic incorporation and LTP contribute to the enhanced cognitive effects of FGL , then blocking the PKC pathway should also block the effects of FGL on learning . To test this point , rats were stereotactically implanted with a double-cannula into the lateral cerebral ventricles . They were then divided into four experimental groups ( n = 10–14/group ) according to treatment ( total administered volume , 5 µl ) : ( i ) vehicle ( artificial CSF [ACSF] ) , ( ii ) vehicle +20 µg FGL , ( iii ) 20 µg FGL +2 . 5 nmol chelerythrine , and ( iv ) vehicle +2 . 5 nmol chelerythrine . Drugs were injected 5 d , 3 d , and 1 d before training . The experimenter who trained the rats was blind to treatment . No side effects were observed following any of the treatments . Because performance of vehicle- and chelerythrine-treated rats was similar in this task ( F1 , 10 = 0 . 6 , p = 0 . 4 ) , data from these two control groups were pooled ( see average data in Figure S8A ) . There were significant differences in the learning curves among the remaining groups ( repeated measure ANOVA: F2 , 22 = 12 . 79 , p = 0 . 001 ) . Similar to our observations with subcutaneous FGL administration ( Figure 2 ) , FGL-treated rats ( FGL+vehicle ) outperformed their controls ( vehicle ) in the water maze throughout the training procedure ( Bonferroni's post hoc test: p<0 . 05 for trials 3 and 7; Figure 8A ) . Notably , FGL did not enhance learning when combined with chelerythrine ( FGL+chelerythrine ) , indicating that the effects of FGL on cognition depend on PKC activity ( Bonferroni's post hoc test: p<0 . 05 , for the comparison between FGL+vehicle versus FGL+chelerythrine , for trials 3 , 4 , and 7 ) . Similar blind experiments were separately carried out with the specific inhibitor of classical PKC isoforms GF109203X . Consistent with our experiments with chelerythrine , co-injection of GF109203X ( 1 nmol in 5 µl ) with FGL blocked the enhanced performance produced by the peptide alone ( Figure 8B ) ( vehicle- and GF109203X-treated rats were similar in this task , and these two control groups were pooled; see separate data in Figure S8B ) . Finally , similar experiments were attempted with MEK and PI3K inhibitors ( PD98059 and LY294002 , respectively ) , but these data could not be easily interpreted due to the intrinsic effect of these drugs on the behavioral task ( see Figure S9 ) . FGL has also been shown to enhance memory consolidation when administered immediately after training [12] . We then tested whether the cognitive effects of FGL on memory were also mediated by the same signaling pathway as the synaptic and learning effects . To this end , rats were stereotactically implanted with cannulae as in the previous experiment , but FGL ( with or without chelerythrine ) was administered immediately after the training session of each day ( see Materials and Methods ) . Spatial memory was then tested 24 h after the second training day . As shown in Figure 8C , FGL was able to enhance memory under these conditions , as previously described . Most importantly , this enhancement was blocked when FGL was co-administered with chelerythrine , indicating that PKC is also required for FGL-induced memory enhancement . In conclusion , these behavioral experiments testing the effect of FGL on both learning and memory indicated that enhanced cognition depends on the PKC pathway , similarly to enhanced AMPAR synaptic delivery and LTP . These results strongly suggest that synaptic and cognitive effects of FGL are mechanistically linked .
In this work , we have described a specific molecular mechanism and signaling pathway by which the NCAM mimetic peptide FGL produces a long-lasting enhancement of synaptic plasticity that leads to improved spatial learning and memory . We propose that the facilitation of AMPAR synaptic delivery during learning-induced plasticity events underlies the enhanced cognition produced by FGL . Interestingly , these effects were persistent , and outlasted the initial FGL stimulus . These conclusions are based on three major lines of evidence . First , we showed that FGL produces a long-lasting potentiation of synaptic transmission in hippocampal slices , based on the facilitated synaptic delivery of AMPARs upon NMDAR activation . Second , we found that synapses remain “sensitized” for further LTP induction long after FGL is cleared from the system . Third , both the synaptic and behavioral effects of FGL are based on a long-lasting increase in PKC activity , which is accompanied by a persistent activation of the CaMKII pathway . A particular novelty of this work relies on linking a direct synaptic modification ( i . e . , facilitation of AMPAR synaptic delivery ) with a high-order cognitive effect ( i . e . , enhanced spatial learning ) . Importantly , FGL is acting as a facilitator , rather than a direct trigger , for AMPAR synaptic insertion . That is , FGL-induced potentiation still remains activity-dependent , because it requires NMDAR and CaMKII activation . In fact , it appears that FGL is sensitizing the “classical” LTP pathway , because spontaneous neuronal activity is then able to trigger AMPAR delivery and stabilize synaptic potentiation . This mechanism differs significantly from other neurotrophin-related synaptic modulators , such as BDNF or tumor necrosis factor-α ( TNFα ) , which trigger AMPAR synaptic delivery while bypassing standard LTP signaling [49] , [50] . This is a significant distinction , because an effective cognitive enhancer may be expected to facilitate synaptic plasticity events , rather than provoke them in a manner unrelated to ongoing circuit activity . We also want to point out that FGL is acting as a synaptic and cognitive enhancer over physiological levels . That is , FGL treatment increases the extent of synaptic potentiation in naive hippocampal slices , and similarly , it enhances the potential for spatial learning in healthy , young adult rats . These results emphasize the notion that synaptic and cognitive mechanisms can be tuned to operate above normal physiological parameters . From a mechanistic perspective , it is interesting that FGL produces a persistent enhancement of synaptic plasticity and learning . Thus , we found that a transient activation of FGFR-NCAM signaling by FGL results in long-lasting activation of the PKC and CaMKII pathways , accompanied by enhanced LTP and spatial learning observed 24–48 h after the removal of FGL . We have also determined that the establishment of this persistent enhancement requires PKC activity during the action of FGL . In a sense , this sensitized state produced by FGL is reminiscent of what has been termed metaplasticity ( i . e . , a persistent modification that alters the ability of the synapse to undergo further plasticity events [51] ) . There is a wide variety of mechanisms that may shift the sensitivity to synaptic plasticity induction , ranging from structural alterations in the extracellular matrix [52] to changes in NMDAR subunit composition [53] or metabotropic glutamate receptor activation [54] . With regard to FGL and the subsequent activation of FGFR-NCAM signaling , we determined that the functional synaptic changes are not accompanied by detectable alterations in morphology of dendritic spines , ultrastructural synaptic organization , or presynaptic release properties ( this is in contrast with the reported effects of FGL on synaptogenesis and presynaptic function on primary neuronal cultures [12] ) . On the other hand , we established that PKC activity is required to reach this sensitized state , followed by a long-lasting increase in CaMKII activity and AMPAR phosphorylation at GluA1 Ser831 . Interestingly , these changes do not saturate ( or occlude ) further LTP expression . In fact , the potential for LTP expression is actually enhanced after FGL treatment . These considerations lead us to propose a model in which FGL-triggered FGFR-NCAM signaling acts on upstream targets that facilitate LTP induction . Subsequently , ongoing neuronal activity will be more likely to activate NMDARs and CaMKII , leading to enhanced synaptic delivery of AMPARs and potentiation of synaptic responses . This strengthening of excitatory connections may in turn facilitate the induction of further LTP-like events , resulting in the observed long-lasting increase in PKC and CaMKII activation , and AMPAR phosphorylation . These synaptic and biochemical changes will persist , as long as there is ongoing neuronal activity in the circuit . In conclusion , the present study has provided mechanistic insights into the synaptic events and molecular cascades that mediate the enhanced cognitive function produced by a pharmacological mimetic of cell adhesion-growth factor signaling .
The pentadecapeptide FGL ( EVYVVAENQQGKSKA ) , corresponding to residues E681 to A695 of the second fibronectin domain of NCAM ( Figure 1A ) , was synthesized using 9-fluorenylmethoxycarbonyl ( Fmoc ) solid phase peptide synthesis methodology . FGL was used in dimeric form by linking two monomers through their N-terminal ends with iminodiacetic acid ( N-carboxymethyl ) -glycine ( Polypeptide Laboratories , Hillerød , Denmark ) . This dimeric design binds and brings together two FGFR molecules , which is crucial for receptor phosphorylation and downstream signaling [55] . The C-terminal ends of the peptides were amidated . The peptide was at least 85% pure as estimated by high-performance liquid chromatograhy . Morris water maze spatial learning was performed under mild training conditions to assess any modulation of learning related to FGL administration . For details , see Text S1 . Protein extracts from hippocampal slices were prepared in 10 mM HEPES , 150 mM NaCl , 10 mM EDTA , 0 . 1 mM phenylmethanesulphonylfluoride ( PMSF ) , 2 µg/ml chymostatin , 2 µg/ml leupeptin , 2 µg/ml antipain , 2 µg/ml pepstatin , 10 mM NaF , 1 µM microcystin LR , 0 . 5 µM calyculin A , and 1% Triton X-100 . Western blots were developed with chemiluminescence ( SuperSignal Kit; Pierce , Rockford , IL ) and quantified using a densitometric scanning under linear exposure conditions . Rats ( n = 8/group ) were perfused with 4% paraformaldehyde ( pH 7 . 4 ) . Coronal sections ( 150 µm ) were cut on a vibratome . Cells in the hippocampus were individually injected with Lucifer Yellow . Imaging was performed on a Leica laser scanning multispectral confocal microscope ( TCS SP5 ) using an argon laser . After acquisition , the stacks were processed with a 3-dimensional blind deconvolution algorithm ( Autodeblur; Autoquant , Media Cybernetics ) to reduce the out-of-focus light ( see example in Figure 3F ) . The 3-dimensional image processing software IMARIS 5 . 0 ( Bitplane AG , Zurich , Switzerland ) was used to measure the spine head volume and neck length ( see also Text S1 ) [30] . For spine density analysis , dendrites were traced with the Neurolucida 7 . 1 computerized data collection system . Spine density was automatically calculated by dividing the number of spines on a dendrite by the dendrite length . See Text S1 for further details . Sections adjacent to those used for the intracellular injections were processed for electron microscopy . These sections were embedded in Araldite and studied using a correlative light and electron microscopic method ( described in detail in Text S1 ) . The two major morphological types of cortical synapses , namely asymmetrical and symmetrical types [29] , were clearly identified in the analyzed tissue . The synapses in which the synaptic cleft and associated membrane densities could not be visualized clearly because of the oblique plane of the section were considered uncharacterized synapses . Synaptic density per unit area ( NA ) was estimated from electron microscope samples of neuropil of the stratum radiatum of CA1 . The density of synapses per unit volume of the neuropil was calculated using the formula NV = NA/d , in which NA is the number of synaptic profiles per unit area and d is the average cross-sectional length of synaptic junctions ( for a detailed description , see [29] ) . Hippocampal slice cultures were prepared from postnatal day 5–6 rats [56] . After 4–8 d in culture , FGL was applied to the culture medium ( 10 µg/ml ) , and the medium was refreshed , without FGL , 24 h later . Electrophysiological recordings were performed 2 d after beginning the FGL treatment . Voltage-clamp whole-cell recordings were obtained from CA1 pyramidal neurons under visual guidance using fluorescence and transmitted light illumination . For details regarding the internal and external solutions , see Text S1 . Synaptic AMPAR-mediated responses were acquired at −60 mV . NMDAR responses were recorded at +40 mV at a latency at which AMPAR responses were fully decayed ( 60 ms after stimulation ) . In both cases , 100 µM picrotoxin was present in the external solution . GABAA receptor responses were recorded at 0 mV in the presence of 100 µM AP5 in the external solution . For the rectification studies , GluA1-GFP was expressed in CA1 neurons for 60 h , and AMPAR responses were recorded at −60 mV and +40 mV in the presence of 0 . 1 mM AP5 in the external solution and 0 . 1 mM spermine in the internal solution . Because only CA1 cells ( and not CA3 cells ) were infected , this configuration ensured that GluA1-GFP was always expressed exclusively in the postsynaptic cell . When specific kinase inhibitors were used , the inhibitor was added to the culture medium 1 h before the addition of FGL . The medium was refreshed 1 d later with the same inhibitor but without FGL . Electrophysiological recordings were performed 1 d later in the presence of the kinase inhibitor ( with the exception of chelerythrine , which was neither added to the medium upon refreshing nor present during the recordings ) . LTP was induced using a pairing protocol by stimulating Schaffer collateral fibers at 3 Hz for 1 . 5 min while depolarizing the postsynaptic cell at 0 mV . Whole-cell recordings were made with a Multiclamp 700A amplifier ( Axon Instruments ) . For the behavioral results , the data were analyzed with SPSS version 11 ( Chicago , IL , USA ) . Morris water maze training data were analyzed across trials with one-way analysis of variance ( ANOVA ) , with the training trial as the repeated measure , followed by Bonferroni's post hoc test when appropriate . For the morphological parameters , the data were averaged to obtain the cell mean , and the neurons from each animal were averaged for the animal mean . Normality was tested using the Kolmogorov-Smirnov test . Because both the spine density and spine morphology values had Gaussian distributions , we used a two-tail unpaired t test to assess the overall effect . When comparing mean electrophysiological values , statistical significance was determined by the Mann-Whitney test ( unless indicated differently ) if only two distributions were compared or by ANOVA followed by the Kruskal-Wallis test if multiple distributions were analyzed .
|
The human brain contains trillions of neuronal connections , called synapses , whose pattern of activity controls all our cognitive functions . These synaptic connections are dynamic and constantly changing in their strength and properties , and this process of synaptic plasticity is essential for learning and memory . Alterations in synaptic plasticity mechanisms are thought to be responsible for multiple cognitive deficits , such as autism , Alzheimer's disease , and several forms of mental retardation . In this study , we show that synapses can be made more plastic using a small protein fragment ( peptide ) derived from a neuronal protein involved in cell-to-cell communication . This peptide ( FGL ) initiates a cascade of events inside the neuron that results in the facilitation of synaptic plasticity . Specifically , we find that FGL triggers delivery of a specific type of glutamate receptor ( AMPA receptors ) to synapses in a region of the brain called the hippocampus , which is known to be involved in multiple forms of learning and memory . Importantly , when this peptide was administered to rats , their ability to learn and retain spatial information was enhanced . Therefore , this work demonstrates that cognitive function can be improved pharmacologically in adult animals by enhancing the plasticity of synaptic connections in the brain .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biology",
"neuroscience"
] |
2012
|
Facilitation of AMPA Receptor Synaptic Delivery as a Molecular Mechanism for Cognitive Enhancement
|
Hepatitis C virus ( HCV ) infection develops into chronicity in 80% of all patients , characterized by persistent low-level replication . To understand how the virus establishes its tightly controlled intracellular RNA replication cycle , we developed the first detailed mathematical model of the initial dynamic phase of the intracellular HCV RNA replication . We therefore quantitatively measured viral RNA and protein translation upon synchronous delivery of viral genomes to host cells , and thoroughly validated the model using additional , independent experiments . Model analysis was used to predict the efficacy of different classes of inhibitors and identified sensitive substeps of replication that could be targeted by current and future therapeutics . A protective replication compartment proved to be essential for sustained RNA replication , balancing translation versus replication and thus effectively limiting RNA amplification . The model predicts that host factors involved in the formation of this compartment determine cellular permissiveness to HCV replication . In gene expression profiling , we identified several key processes potentially determining cellular HCV replication efficiency .
Hepatitis C virus ( HCV ) infection is a major global health problem , with approximately 170 million chronically infected individuals worldwide and 3 to 4 million new infections occurring each year [1] . Acute infection is mostly asymptomatic , however , it develops into a chronic infection in about 80% of patients , and then is a leading cause of liver cirrhosis , hepatocellular carcinoma and subsequent liver transplantation [2] , [3] , [4] . A significant fraction of patients cannot be cured even with modern combination therapies , partially due to ab initio non-responsiveness , but also due to the emergence of drug-resistant HCV quasispecies . HCV is an enveloped plus-strand RNA virus and belongs to the Flaviviridae family . Upon entry into the host cell , its 9 . 6 kb genome is translated by a cap-independent , internal ribosomal entry site ( IRES ) mediated mechanism into a single large polyprotein . Viral and cellular proteases co- and post-translationally cleave this precursor into ten mature viral proteins , comprising three structural proteins ( core , E1 , E2 ) , the ion channel p7 as well as the six non-structural ( NS ) proteins NS2 , 3 , 4A , 4B , 5A and 5B [5] . The five “replicase” proteins NS3 to NS5B are essential and sufficient for intracellular genome replication . NS3 comprises an RNA helicase and a protease domain , the latter of which , together with the co-factor NS4A , forms the major viral protease NS3/4A , liberating itself and all other replicase proteins from the polyprotein precursor . NS4B , together with other NS proteins , induces membrane alterations , observable as convoluted , vesicular membrane structures known as the membranous web and believed to act as the sites of RNA replication [6] , [7] . The exact architecture and topology of these structures , and particularly their structure-function-relationship , is not fully understood yet . However , for Dengue virus ( DV ) , a related flavivirus , the three-dimensional makeup of the membrane rearrangements has been solved recently [8] . There , numerous small , vesicular invaginations into the rough endoplasmic reticulum ( ER ) serve as a protected environment for genome replication . NS5A is a phosphoprotein important both in RNA replication and particle assembly and/or release . NS5B , the RNA-dependent RNA polymerase ( RdRP ) , is the core enzyme of the replicase complex . In order to amplify the viral RNA , NS5B first synthesizes a complementary ( i . e . negatively oriented ) strand from the plus stranded genome , putatively resulting in a double-stranded ( ds ) intermediate [9] . From this negative strand template , NS5B then transcribes progeny plus strands . Given the ∼10-fold higher number of plus strands over minus strands within the host cell , this most likely occurs in a repetitive manner [10] . Newly synthesized plus strands are then released by an unknown mechanism from the replicative compartment and can then either be directed to encapsidation into assembling virions , or re-enter the replicative cycle by serving as templates for further translation and subsequent incorporation into a new replication complex . It is interesting to note that although HCV establishes a persistent infection , it does not have a latent phase; throughout the course of the infection , which can be decades long in many patients , there is constant production of viral RNA , proteins and infectious particles . In most viral infections , presence of non-self structures , such as dsRNA or viral proteins , is readily detected by sensors of the immune system , leading to the production of type I interferon ( IFN ) and activation of the adaptive immune response [11] . Also in case of HCV , innate as well as adaptive immune responses are elicited , however , by means of various complex interactions with cellular processes , the virus is capable to blunt these defense mechanisms and thus is able to persist [12] . This ability of HCV to maintain low profile persistence is most likely intimately linked to its tightly controlled RNA replication; for the closely related bovine diarrhea virus ( BVDV ) , which can be converted from a persistently to an acutely replicating form , a direct correlation between excessive RNA replication and the induction of cytopathic effects has been described [13] . To comprehensively study these complex and highly dynamic processes that can only inappropriately be addressed by individual experiments , an eminent approach is mathematical modeling . Consequently , a basic model of HCV infection dynamics was published almost 15 years ago [14] and has since led to the development of several related models of HCV infection and therapy dynamics [15] , [16] , [17] , [18] , [19] , [20] , [21] . However , all of these models described the long-term dynamics at the level of cell populations , organs and even organisms ( patients ) , and did not take intracellular processes such as genome translation and the actual RNA replication into account . With the development of subgenomic HCV replicons , detailed studies of intracellular RNA replication became possible [22] , [23] . A thorough quantitative analysis of persistent subgenomic replicons in Huh-7 cells [10] led to the development of a first mathematical model of intracellular steady state RNA replication [24] . Further models addressed the effect of potential drugs on viral replication [25] or included the production of virus particles [26] , [27] . However , all published models were solely based on measurements of steady state replication . In contrast , to understand how the virus on the one hand manages to efficiently ( and quickly ) establish itself within a host cell before the cell is able to mount an antiviral response , while on the other hand , it is strictly limiting its own amplification , static ( steady state ) data is not sufficient but needs to be complemented by information about the dynamic aspects of replication . Previous studies on replication kinetics in cell culture in fact point to a highly dynamic initial phase of RNA replication in the first few hours after genome transfection or infection , which then reaches a steady state within 24–72 hours [22] , [28] , [29] . Actual amplification kinetics and the absolute levels attained in the steady state vary largely between different experimental systems and are mainly determined by the permissiveness of the employed host cell [30] , [31] , [32] and by the viral isolate [30] , [32] , [33] . Therefore , in our present study we quantitatively followed the onset of intracellular RNA replication within the first couple of hours upon introduction of HCV genomes into the host cells . Based on these data we developed a comprehensive mathematical model capable of precisely describing both , the dynamic and the steady state phases of viral replication . We then used this model to study aspects of the viral replication cycle that cannot directly be accessed experimentally .
In order to comprehensively understand the observed HCV replication dynamics and its underlying molecular processes , we set up a mathematical model of the intracellular HCV RNA replication . Dahari and colleagues developed a similar model previously , upon which we could build here [24] . Briefly , our model comprises all relevant molecular species ( RNA , proteins , ribosomes , etc . ) , and describes each step in the RNA replication cycle , such as translation , protein maturation and the formation of the actual RNA replication complex , as reactions of the involved molecules using differential equations based on standard mass action kinetics . Of note , the establishment of a vesicular replication compartment ( RC ) by viral proteins ( in concert with cellular functions ) within which RNA replication takes place is reflected in the model by partitioning of the respective molecular species into distinct “cytoplasmic” and “replication compartment” pools; e . g . only cytoplasmic HCV RNA ( RPcyt ) can be translated by ribosomes , but not HCV RNA within the replication compartment ( RP ) . Model equations ( eq . ) of our final model are given in the materials and methods section and a schematic illustration is shown in figure 2C . The original model of Dahari was solely based on steady state measurements of viral RNA and protein concentrations in a stable replicon cell line [10] , and accordingly was not capable of explaining the dynamic phase during the establishing of replication as observed in our experimental data , even after re-fitting all model parameters ( high permissive cell line; total sum of squared residuals χ2 = 8 . 69 , compare supplementary figure S1 ) . From this finding it was evident , that modifications to the model were required in order to accurately capture the initial dynamics of HCV RNA replication , as it can be observed upon transfection of viral genomes into “naïve” cells . Based on biological reasoning , we extended and modified Dahari's original model at two steps of the replication cycle . For one , to account for ab initio replication in our setting ( in contrast to pre-formed steady-state replication ) , we introduced one additional RNA species Rpunp , representing the transfected “naked” replicon RNA and an according processing step ( rate k0 ) , subsuming any re-folding , association with RNA-binding proteins and other processes that might take place and be required before in vitro transcribed RNA assumes a translation-competent state ( eq . 1 and 2 ) . This processing corresponds to viral genomic RNA being released into the cytoplasm upon actual infection . We furthermore allowed RNA degradation to be different ( presumably higher ) for the “unprocessed” transfected RNA ( μpUnp ) as compared to “processed” or cell-derived RNA ( μpcyt ) . The second step that we updated to reflect the current biological understanding of the molecular processes was the initiation of minus strand RNA synthesis ( which in the model is assumed to correspond to the formation of the replicative compartment , as discussed later ) . It has been described for HCV , but also for other viruses [35] , [36] , [37] , [38] , that the formation of a productive replicase complex requires the viral polymerase ( NS5B ) and possibly other NS-proteins to be supplied in cis . This means that for reasons not yet fully elucidated , NS5B cannot initiate RNA synthesis from a free , cytosolic RNA genome , but only from the very RNA that it has been translated from . This implies a tight spatio-temporal coupling of ( poly ) protein production and initiation of RNA replication , i . e . initiation can only occur immediately after translation/polyprotein processing and therefore in close proximity to the translation complex ( TC ) . As our model does not account for spatial effects ( such as diffusion ) , we approximated this cis-process by requiring an active translation complex instead of free , non-translating RNA ( RPcyt ) for the initiation of minus strand RNA synthesis ( RIP , eq . 7 ) . This cis-triggered formation of the replicative compartment consequently is the only route for uptake of viral genomes and also NS proteins ( Ecyt ) into replication vesicles . This is a major change to Dahari's original model , in which cytosolic RNA ( Rpcyt ) and NS proteins ( Ecyt ) could freely and independently enter the compartment . This model , comprising Dahari's original model with the described extensions , we then considered our base model . We then tested , whether our base model would be capable of explaining the measured replication dynamics . We therefore fitted the model to the experimental data from the high permissive Huh7-Lunet cells . In fact , this resulted in a significantly better fit as compared to the original model ( Dahari: χ2 = 8 . 69 , base model: χ2 = 2 . 12 ) and was capable of adequately describing both , the highly dynamic initial phase as well as the ensuing steady state of viral RNA replication ( figure 2A ) . Having established a base model for HCV replication , we next wanted to assess which factors could explain differences observed between high and low permissive cell lines . In our experimental measurements for two differently permissive cell lines , Huh7-Lunet ( high permissive ) and Huh-7 lp ( low permissive ) , replication reached a steady-state within the period of observation ( 72 h ) , however , plateau levels of viral protein , plus-strand RNA and minus-strand RNA differed by approximately one order of magnitude; further , the onset of the net increase of plus-strand RNA was delayed significantly in the low permissive cells and also the minimum concentration of plus-strand RNA reached during net degradation in the first hours after transfection were significantly lower in low permissive cells ( compare figure 1B and C ) . As both , Huh7-Lunet and Huh-7 lp cells , were transfected with the same subgenomic HCV replicon , these differences must be due to differences between the host cells . In order to reflect this host influence also in our model , we tested different steps in the HCV RNA replication cycle that do or could feasibly depend on a host process: ( A ) efficiency of RNA entry or initial RNA processing; ( B ) the number of ribosomes available for RNA translation; ( C ) RNA degradation in the cytoplasm ( possibly including antiviral processes such as activation of RNaseL ) ; ( D ) polyprotein translation or maturation ( i . e . cleavage ) ; ( E ) the formation of the replicative compartment/initiation of minus-strand synthesis; ( F ) RNA synthesis or ( G ) RNA degradation inside the replication compartment; or ( H ) the export of newly synthesized RNA into the cytoplasm . To evaluate these alternatives for their capacity to explain the differences in dynamics and steady-state levels between the two cell lines , we fitted our base model simultaneously to the experimental data from both cell lines , leaving only the parameters free to differ between high and low permissive cells that , in the respective hypothesis ( A ) to ( H ) , depend on the corresponding host factor; all other parameters were constrained to be identical between the two cell lines . We found that hypotheses ( A ) , ( B ) , ( C ) , ( D ) , ( F ) , ( G ) and ( H ) could not explain the above described qualitative difference in replication dynamics: while ( C ) and ( H ) did lead to a steady-state but could not reproduce the lower plateau RNA levels in Huh-7 lp , hypotheses ( A ) , ( B ) , ( F ) and ( G ) altogether failed to establish a steady-state in low permissive cells in the course of the simulated time period of 80 h ( supplementary figure S2 ) . In order to identify the best fitting hypothesis , we also quantitatively assessed the capability of each hypothesis to fit both data sets by calculating χ2 over all data points from the two time series , as well as Akaike's information criterion ( AIC ) , which additionally takes into account the number of unconstrained parameters ( figure 2B ) . While parameter differences in the RNA synthesis inside the RC , i . e . hypothesis ( F ) , led to the lowest overall χ2 value , hypothesis ( E ) – assuming a difference in the formation of the RC and initiation of RNA synthesis– led to a slightly larger χ2 ( 5 . 84 vs . 5 . 60 ) but a significantly lower AIC ( −21 . 31 vs . −0 . 68 ) . Moreover , hypothesis ( E ) reached a steady-state within 80 h , while ( F ) did not . This comparison therefore identified the initiation of minus strand RNA synthesis ( i . e . the formation of the RC ) as the step in the model , at which alteration of a single reaction rate suffices to optimally transform replication dynamics from high permissive cells into the dynamics found in low permissive cells . Biologically , this step is highly complex and not thoroughly understood yet . After translation and polyprotein processing , reorganization of host cell endomembranes is triggered by viral NS proteins such as NS4B , which has been shown to be a key player in the formation of membrane convolutions at the rough endoplasmic reticulum . These vesicular membrane structures , dubbed the membranous web , have been reported to be the site of HCV RNA replication [7] , providing a distinct replicative compartment for the viral replicase machinery . However , the molecular mechanisms leading to the formation of productive replication vesicles are not known . Nonetheless , it is clear that host factors must be required in this complex process , for example proteins involved in membrane biogenesis and reorganization , as well as signal transducers and regulatory molecules; and also the initiation of minus strand RNA synthesis might require a cellular co-factor . It appears plausible that limited abundance of one of these factors in some cells might be responsible for their lower permissiveness for HCV replication . Therefore , we next wanted to include this host factor as an explicit species in our model , which is required for RC formation/minus strand initiation . For that purpose , we subsumed all these possible host determinants by one unspecified host factor HF ( see figure 2C ) , which we assumed to interact with viral NS proteins ( Ecyt , e . g . NS4B or NS5A ) and with actively translated HCV RNA ( TC ) to create replication vesicles and to allow for initiation of minus-strand RNA synthesis ( being part of the minus-strand initiation complex RIP , see eq . 7 and figure 2C ) . In addition , we made the ( non-crucial , see supplementary figure S3 and supplementary table S5 ) assumption that HF is only catalyzing the reaction without being consumed . With this additional modification to the mathematical description of the formation of replication compartments , and calibration of the model to the experimental data from both cell lines ( constraining parameters and initial values to biologically meaningful ranges taken from measurements or literature wherever possible ) , excellent agreement between the model and experimental data was achieved , both , for high and low permissive cells with an overall χ2 of 2 . 01 and AIC of −112 . 31 ( figure 3A and B ) . We therefore considered this our final working model , illustrated in detail in figure 2C . Briefly , the model comprises 13 molecular species in two distinct compartments , the cytoplasm and a replicative compartment ( RC ) , and is parameterized with 16 parameters , corresponding to reaction rates , as well as three non-zero initial values: the initial concentration of HCV RNA ( Rpunp ) , the initial concentration of the host factor ( HF ) , as well as the total number of ribosomes available for viral RNA translation ( Ribotot ) . The full system of differential equations and detail on the modeling procedure can be found in the Materials & Methods section; more detail on parameter optimization and analysis are given in supplementary text S1 . Interestingly , analysis of the fitted parameters showed that the concentration of the host factor was more than 10 fold higher in highly permissive Huh7-Lunet cells than in low permissive Huh-7 lp cells . This difference led to slower formation of the replication compartment in Huh-7 lp cells , which in turn resulted in the observed delay in early viral replication and in decreased steady state levels in these cells . Based on our model and computational analysis , we therefore propose that a host process is critically involved in the formation of replication vesicles and/or the initiation of minus-strand RNA synthesis , turning this into the rate-limiting step for HCV RNA replication in low permissive cells . While the model could be very well fitted to the original replication data , we then wanted to corroborate its applicability for predicting replication dynamics also under distinct conditions that were not part of the calibration process . For this purpose , we performed additional , independent experiments using mutant HCV replicons with defects at defined stages of the replication cycle . We predicted the impact of such defects on viral replication a priori using the model , and retrospectively compared the results with the experimental data in order to assess the validity of model predictions . This approach of introducing targeted mutations into the HCV genome interfering with distinct functions in the viral RNA replication cycle allows validation of individual steps in the model , thus step-wise reconfirming model assumptions and parameters . As a test of the translation phase of the model , we measured viral plus-strand RNA and protein expression using a replication deficient replicon harboring a deletion of the catalytic triad ( GDD motif ) of the NS5B polymerase . The measured RNA and protein data thus reflect only the effect of translation and degradation in the cytoplasm , with no RNA replication occurring . We predicted the impact of this intervention with our model by setting the formation rate kPin of the plus strand replication initiation complex RIp to zero ( eq . 3 , 5 and 7 ) , thus completely switching off polymerase activity at the earliest possible point , while leaving all other model parameters unchanged . Notably , our model predictions of this intervention matched the experimental data from both , Huh7-Lunet and Huh-7 lp cells , validating our model of cytoplasmic RNA degradation and translation ( Figures 4A and B ) . The fact that the experimental measurements showed almost identical RNA decay dynamics and viral protein ( luciferase ) levels in high and low permissive cells is also direct experimental confirmation of our modeling based assessment above , that differences in permissiveness cannot be related to RNA “processing” or degradation , or to ribosome availability or protein translation in the cytoplasm ( hypotheses ( A ) , ( B ) , ( C ) and ( D ) tested above ) . We next focused on validating the RNA replication steps of our model . For this purpose , we utilized chimeric replicons with heterologous 5′- or 3′-NTRs derived from a different genotype [22] . We previously showed that these chimeric replicons exhibit decreased replication efficiency due to inefficient initiation of plus-strand synthesis ( in case of the 5′-NTR exchange ) or minus-strand synthesis ( in case of the 3′-NTR exchange ) [22] . We predicted the effect of these interventions with the fitted model by decreasing the parameters kpin and k4m for the 3′-NTR exchange ( eq . 3 , 5 , 7 , 8 and 9 ) , and k5 and k4p for the 5′-NTR exchange ( eq . 7 , 8 and 9 ) , corresponding to the rates of the minus- and plus-strand initiation and synthesis , respectively ( for reference , see figure 2C ) . Comparison of our prediction with experimental measurements demonstrated that in both cases the model qualitatively agreed with the experimental data . Consequently , upon refitting of these parameters to the new data , the model was capable of quantitatively describing the perturbed replication kinetics ( figure 4C ) . Furthermore , the model correctly predicted the impact of the respective NTR-exchanges onto the ratio of plus- to minus-strand RNA at the steady state ( figure 4D ) . Predictions for both NTR-exchanges were in close quantitative agreement with our previously published experimental observations , which showed an 8 . 7∶1 ( simulation 9 . 0∶1 ) ratio between plus- and minus-strand for the wildtype , 16 . 1∶1 ( 11∶1 ) for the 3′-NTR-chimera , and 4 . 7∶1 ( 4 . 8∶1 ) for the 5′- chimera [22] . Taken together , our model was able to correctly infer the effects of targeted interventions at different steps of the replication process , including complete replication deficiency , as well as specific inhibition of plus- or minus-strand RNA synthesis , respectively . We therefore conclude that our model provides a realistic description of HCV RNA replication dynamics , and thus can be confidently used to further study such processes in silico that are difficult or impossible to address experimentally . Having such a comprehensive and accurate model at hand , we proceeded by applying it to concrete problems in the field of HCV research . The first question we addressed was which sub-steps of HCV RNA replication would be most susceptible to targeted interference . Such processes are potentially attractive targets for the design of new DAAs against HCV . To find out which step in the replication cycle has the biggest impact on the resulting RNA and protein levels , we assessed the relative sensitivity of replication towards alterations of reaction rates in the model . To account for the two clearly discernable phases of replication – the highly dynamic establishing phase and the steady-state phase – we performed a global sensitivity analysis using the extended Fourier Amplitude Sensitivity Test ( eFAST ) [39] , [40] at an early ( 4 h ) and at a late ( 72 h ) time point . We separately assessed the sensitivities of plus-strand RNA , minus-strand RNA as well as protein levels towards individual and simultaneous changes of 16 rate constants and the three initial values ( figure 5 and supplementary figure S4 ) . For the establishing phase of replication , this analysis showed that the most influential processes are the polyprotein translation ( rate k2 ) , the export rate of RNA into the cytoplasm ( rate kPout ) and the efficiency of plus- ( rate k4p ) and minus- ( rate k4m ) strand RNA synthesis inside the replication compartment , respectively ( figure 5A ) . As one would expect , alterations in k2 mainly influence the amount of viral protein ( eq . 4 and 6 ) and only to a lesser degree viral RNA ( eq . 2 and 3 ) , whereas k4m mainly affect RNA species ( eq . 7 , 8 and 9 ) . k4p and kPout in contrast strongly influence RNA and protein concentrations ( eq . 8 , 9 , 10 and 11 ) . Further important steps are the initial “processing” of the transfected RNA ( rate k0 ) , since this determines at what time and to what extent RNA is available for translation , as well as the RNA degradation rate μRC inside the replication compartment . The availability of viral RNA for rapid genome replication and the replication process inside the membranous web itself are therefore key determinants of the initial replication dynamics and thus the efficiency of infection , and consequently constitute a very attractive target for anti-viral drugs . Interestingly , the rate of polyprotein translation ( eq . 4 ) naturally has a big impact on viral protein concentration , but only a fairly restricted influence on RNA levels . Furthermore , the cleavage rate of nascent viral polyprotein ( eq . 4 and 5 , rate kc ) only very mildly impacts replication dynamics . A profoundly different pattern can be observed for the steady state phase . The single most influential parameter determining viral RNA and protein levels was found to be the degradation rate of viral RNA inside the replication vesicles μRC ( eq . 7 to 11 ) , while most other parameters showed no significant sensitivities ( figure 5B and supplementary figure S4 ) . However , it is virtually impossible to influence this parameter by cellular ( e . g . innate immune ) or pharmacological intervention ( except by physical destruction of the membranous structures ) , therefore making inhibition of viral replication particularly cumbersome once the steady state has been established . Taken together with the results from the early phase , these analyses suggest a key role of the replicative compartment for a successful establishment and maintenance of infection . In the light of the above findings , pointing to a central role of the membranous web within the RNA replication cycle , we further studied the underlying molecular functions of this compartment . For one , we assessed the importance of its protective character onto the dynamics of viral genome replication . Model fitting led to a more than 4-fold lower RNA degradation rate within the replication compartment ( μRC ) as compared to RNA degradation in the cytoplasm ( μpcyt , see table 1 ) . To simulate the effect of less stringent protection of the RNA inside the RC , we then deliberately increased its degradation rate ( μRC ) and calculated the resulting levels of plus strand RNA over time ( figure 6A ) . This analysis showed that the degradation rate inside the replicative compartment inversely correlated with the amount of RNA produced at any given time . Interestingly , this correlation was not continuous , exhibiting a threshold of productive RNA replication , constituting a “cliff” , crossing of which prevented the establishing of a ( non-zero ) steady-state and effectively killing off viral replication ( figure 6A , dark blue area , see also supplementary figure S5 ) . This highly instable region with very low ( or zero ) RNA copy numbers , strikingly , was reached once degradation inside the RC ( μRC ) was approximately equal to the degradation rate in the cytoplasm ( μpcyt ) . Our model therefore predicts that the viral RNA must be protected from active degradation during replication in order for HCV to maintain robust persistent replication . While it is virtually impossible to reproduce the above findings in a biological experiment ( i . e . increasing RNA degradation inside the replicative compartment ) , previous in vitro data actually showed that viral RNA in the cell , particularly the minus-strand , is highly resistant to nuclease treatment [10] , implying that indeed degrading enzymes cannot enter the replication vesicles . Moreover , in inhibitor studies , ongoing HCV replication was blocked by interferon or a pharmacologic NS3/4A inhibitor , leading to rather slow decrease of RNA with a half-life of 12–20 h [41] , [42] , most likely representing a slow degradation of replication vesicles . In good agreement with these studies , our model predicts a half-life for RNA inside the replicative compartment of 12 h ( rate μRC = 0 . 08 h−1 ) , whereas RNA transfected into the cytoplasm decayed with a half-life of approximately two hours in the experiments using a replication-defective replicon ( see figure 4A ) . Experimentally very hard to address , however , is the degradation rate μpcyt of cytoplasmic HCV RNA generated through replication that might exhibit a different folding or be bound by other proteins as compared to transfected RNA . Yet , it appears highly likely that this degradation rate would more closely match the rate of degradation of transfected , cytoplasmic RNA rather than that of RNA within the membranous replicative environment . In keeping with this plausible assumption , our model predicts a half-life for newly synthesized cytoplasmic RNA of approximately 165 min ( μpcyt = 0 . 363 h−1 ) . Although model estimations for both , μpcyt and μRC , exhibit a rather broad confidence interval , simultaneous modification of both parameters shows that μRC needs to be substantially lower than μpcyt in order to explain the observed kinetics ( figure 6B , dark blue area ) . In terms of viral protein , Quinkert and colleagues showed that in contrast to RNA , only a small fraction ( <5% ) of NS5B molecules is protease resistant [10] . In line with these observations , our model predicts that the vast majority of viral protein remains in the cytoplasm . Another important question , which can hardly be addressed experimentally , is the possibility of re-initiation of minus-strand synthesis inside the replication vesicle . While theoretically it is feasible that the replicative machinery re-initiates minus-strand synthesis on newly produced plus-strands inside the replication compartment ( eq . 7 , second to last term ) , the alternative hypothesis is that such an initiation event can only happen in cis upon translation in the cytoplasm ( see also section on model development above ) . In fact , when analyzing the calibrated model , we found that the rate constant for this reaction ( k3 in eq . 7 , see figure 2C for reference ) needed to be close to zero ( <10−4 h−1*molecules−1 ) to fit the experimental data , and the concentration of “active” polymerase ( E ) was severely limiting the rate of RNA synthesis during the initial dynamic phase . This resulted in an extremely low efficiency of internal re-initiation , implying that most or all of the newly synthesized viral plus-strand RNA is exported to the cytoplasm , from where it must be re-imported for further rounds of RNA replication to occur . Hence , our model supports the notion that negative-strand initiation is very different from plus-strand initiation in that it most likely depends on actively translated RNA with the required NS proteins , mainly NS5B , being supplied in cis . The observed relative shortage of active polymerase within the replication vesicles and the lack of internal re-initiation consequently prevents an exponential amplification of the viral RNA within the replicative compartment . Replication vesicles thus attenuate the rate of viral replication by limiting the availability of the factors required for minus-strand initiation . At the same time , depending on the export rate of newly synthesized plus-strand RNA from the replication vesicles ( kpout ) , they can also exert tight control over protein translation . Newly synthesized RNA can either be exported to the cytoplasm where it can be used for another round of protein translation ( or , in an actual infection setting , the assembly of new viral particles ) , or it accumulates within the replication vesicles; there , however , it cannot be used as a template for minus strand synthesis due to the above described reasons . Taken together , the development of a membranous replication compartment , by physically separating production of new protein ( translation ) and the generation of new RNA ( replication ) , therefore constitutes an important additional level of control over the virus' replication kinetics . This high degree of controllability might be one reason for the evolutionary success of membranous replicative structures , as basically formed by all positive strand RNA viruses . In case of HCV , it allows for sustained low-level replication as is required for the establishment of persistence , mainly by restricting availability of the required proteins within the replicative compartment . Particularly for a persistent virus , tight control over its own replication is essential in order to not overwhelm its host cell and thereby kill it [13] . As we have learned above , the distinct replication compartment plays a central role in this self-limitation . Consequently , we therefore studied , which processes in turn regulate the formation of replication vesicles and eventually lead to the establishment of a steady state . The host factor ( HF ) in our model has been found to be a requisite for the attainment of a steady state and its concentration was a determinant regulating plateau levels of viral RNA and protein between the two differently permissive cell lines . For that reason , we now systematically assessed the impact of different availabilities of HF onto steady-state levels of viral RNA and protein . For HCV RNA levels , this analysis showed a linear correlation with HF concentrations in the range tested: the more abundant HF was , the more RNA replication took place . Interestingly , however , polyprotein levels exhibited a saturation behavior , reaching a plateau for HF concentrations above 20 “molecules” ( note that HF is a virtual species , so “molecules” is an arbitrary unit ) ( figure 6C ) . To understand this nonlinear dependence of viral protein on HF levels , we analyzed the model under conditions of varying HF amounts and found that this saturation stems from different factors being limiting for increasing HF levels: in low permissive cells ( featuring low HF concentrations of around 4 “molecules” ) , HF availability is limiting the formation of replication vesicles ( eq . 7 ) . Therefore , overall RNA concentrations remain relatively low , leaving polyprotein production at a low but steady level; here , RNA in the cytoplasm is the rate limiting factor for protein translation . In high permissive cells ( featuring high HF levels of around 50 “molecules” ) , in contrast , rapid formation of replication vesicles occurs with an associated rapid increase in viral RNA levels . However , ribosome availability ( Ribo ) then becomes limiting for protein translation ( eq . 3 ) , explaining the plateau seen for viral protein concentrations ( figure 6C ) . Accordingly , the ratio between viral protein ( i . e . luciferase ) and plus-strand RNA is lower in the steady state in high permissive cells . This is well in line with the experimental data ( figure 1 , compare B and C ) . Interestingly , these findings suggested that the actual mechanisms governing the establishing of the steady state in low permissive and high permissive cells are different . While in low permissive cells the formation of replication vesicles is the limiting step due to a lack of host factor HF , surprisingly the host translation machinery is the bottleneck in high permissive cells . As differential abundance of the host factor ( or host process ) HF in our model sufficed to explain the observed difference in HCV replication dynamics between high and low permissive cells , it was intriguing to identify the biological nature of this factor . For that purpose , we set out to compare gene expression profiles of Huh-7 cells of different passage number or clonal origin that we had found to exhibit substantially different permissiveness for HCV RNA replication [22] , [30] ( figure 7A ) . We performed full-genomic cDNA microarray ( Affymetrix HGU133plus 2 . 0 ) analysis in eight such Huh-7 derived cell lines , including the above used Huh7-Lunet and low passage ( lp ) Huh-7 cells . Figure 7B shows a scatterplot of the normalized gene expression values for these two cell lines . Assuming a direct correlation between permissiveness and the expression of the host factor HF as suggested by the above analysis ( compare figure 6B ) , we fitted a linear model of each gene's expression level to the observed replication efficiencies in all eight cell lines . By this , we could assess how well each individual gene predicts replication efficiency over the full set of cells . On these data , we then carried out an analysis of variance ( ANOVA ) to identify genes whose expression profiles correlated significantly with replication efficiency . Figure 7C shows the resulting p-values over the degree of differential expression ( as log fold-change ) between Huh7-Lunet and Huh-7 lp ( see also supplementary table S1 ) . We could identify 355 genes , whose expression levels correlated with permissiveness ( p<0 . 2 ) and which exhibited a difference in expression levels of more than 23% ( log fold-change >0 . 3 or <−0 . 3 ) ( figure 7C and supplementary table S2 ) . We then subjected these potential HF candidates to bioinformatics analyses in order to identify host cellular processes or pathways , which are over-represented among those genes ( supplementary tables S3 and S4 ) . These analyses mainly identified metabolic processes such as lipid metabolism and cell growth and proliferation , which is in line with the notion of HCV RNA replication requiring proliferating cells for efficient replication , at least in Huh-7 cells [43] , and numerous reports on its requirement on lipid biosynthesis ( reviewed in [44] ) . While the number of potential HF candidate genes was too large to be functionally validated individually within this study , we surveyed previously published data on HCV host factors , including a manually curated database of HCV-host interactions ( VirHostNet [45] ) as well as large-scale siRNA-based screens [46] , [47] , [48] , [49] . Whereas such high-throughput approaches exhibit very high false-negative rates [50] and therefore are not suited to exclude HF candidates from our analysis , their false-positive rate is very well controlled and the identified hit genes are highly reliable . Using these data , we could in fact identify 17 of our HF candidates to be implicated with HCV ( table 1; marked in red in figure 7C ) . Six of these genes ( JAK1 , LHX2 , PIP5K1A , RPS27A , PPTC7 and COPA ) were found in siRNA-mediated approaches to directly influence HCV replication , as would be expected for a limiting host factor . Five genes ( TF , VCAN , TRIM23 , SORBS2 and MOBK1B ) were identified in a large-scale yeast-two-hybrid based interaction screening [51] to interact with at least one HCV protein ( interaction partner listed in table 1 ) . This , however , does not necessarily indicate that the interaction is essential for RNA replication . On similar lines , six further genes ( MCL1 , SERPING1 , CASP8 , PIK3CB , GAB1 and APOB ) were previously reported to interact with specific HCV proteins in individual studies . Interestingly , most of them ( MCL1 , CASP8 , PIK3CB and GAB1 ) were implicated with a modulation of apoptosis and cell survival/proliferation , supporting our above analysis , in which “cell growth and proliferation” was found to be an enriched function among the differentially expressed genes ( supplementary tables S3 and S4 ) . Based on our model prediction of a limiting host factor/process involved in the formation of functional replication compartments and utilizing our transcriptomic analysis of differently permissive cells , further studies should be devised aiming to delineate the exact nature of this factor or process . Identification of a cellular function that is essential for HCV replication but naturally limiting in certain cell lines would be very intriguing in terms of pinpointing novel targets for anti-HCV therapy . Such a factor would promise to be inhibitable without critically affecting host cell viability , while severely compromising HCV replication efficiency .
In the present study , we have developed a mathematical model of the intracellular steps of HCV replication . In contrast to previous models [24] , [25] , [26] we were not only interested in studying steady state replication in stable replicon cell lines , but specifically addressed the highly dynamic initial phase after RNA genome delivery into the host cell . We therefore performed quantitative , time-resolved measurements of viral protein translation as well as strand-specific viral RNA concentrations in two distinct Huh-7 derived cell lines , exhibiting a vastly different permissiveness for HCV RNA replication [32] . With this data , we tried to re-calibrate the most comprehensive HCV replication model available to date [24] , but found that the model was not capable of explaining the observed dynamics and ensuing steady state simultaneously . We therefore modified and extended that model by accounting for the “naked” , unprotected nature of the initially transfected in vitro transcribed RNA and by updating of the formation step of the RC and the initiation of negative strand RNA synthesis to the current biological understanding of this process . Under steady state conditions , as studied by previous models , equilibrium of the viral replication machinery with static ratios between cytosolic viral RNA and NS proteins has been achieved . Therefore , in the model by Dahari and colleagues [24] , uptake of viral RNA and protein into the replicative compartment could be described by simple first order import reactions . In our setting , however , concentrations for replication competent viral RNA and NS proteins start from zero and grow dynamically in the course of the experiment . Hence , simple first order import reactions do not suffice if the uptake depends on the abundance of more than one species , which is highly likely given biological evidence . Accounting for the above described cis-requirement for initiation of productive replication complexes [35] , [36] , [37] , which means that an RNA molecule can be used as a template for replication only by an NS5B molecule having been translated from that very RNA , we solely allowed a complex of actively translated plus-strand RNA ( i . e . translation complexes TC ) and cytosolic NS proteins ( Ecyt ) to be taken up into the RC . While these model extensions greatly enhanced the fitting quality to the data of a single cell line , we then identified that step in the model , at which an altered kinetic rate could explain the dynamics found in the second cell line as well . For this purpose , we tested a series of hypotheses , fitting the model simultaneously to the two differently permissive cell lines and allowing only those parameters to differ that would be influenced by the host cell in the respective hypothesis . By this approach , we could exclude various processes , e . g . differences in translation efficiency , altered cytoplasmic RNA degradation or different RNA synthesis rates within replication complexes . It is also biologically plausible , that these processes do not differ between the two examined Huh-7 cells lines and therefore cannot explain the observed differences in permissiveness; both , translation and RNA degradation have been shown before to be comparable across different Huh-7 cells [30] , and the polymerization rate of the HCV RdRP NS5B is unlikely to depend on host factors ( other than ribonucleotides ) . In principle , a combination of several such processes might be able to explain the observed behavior; however , following Occam's razor , we considered the simplest solution to be the most likely one . Eventually , we identified the formation process of replicative vesicles to be the best suited step in the model , altering the rate of which sufficed to fit the model to measured data from either cell line . We then introduced a host factor ( HF ) as a new species in our model , and required viral RNA ( in the form TC ) and NS protein ( Ecyt ) to form a complex with it in order to allow for the initiation of negative strand RNA synthesis and the formation of the RC . Assumption of different concentrations of this host factor then was sufficient to very accurately explain the differences in RNA replication permissiveness in the two cell lines . This final model therefore completely satisfied all experimental observations and could also correctly predict the effects of targeted perturbations during extensive validation experiments . We then used the calibrated and validated model to further study individual steps of the viral lifecycle . Sensitivity analysis was applied to pinpoint the most influential steps , perturbation of which would lead to the greatest impact on replication dynamics and yield . A very interesting first finding was that once steady state replication has been reached , the system proved to be relatively robust towards perturbation of individual sub-steps of replication . The degradation rate of RNA inside the RC was the most sensitive parameter under these conditions , and had a significantly higher influence than all other rates . This parameter , however , can hardly be influenced biologically or therapeutically . Very likely , this robustness is key to HCV's prevailing in the face of cellular stress- and innate immune responses [52] , [53] , [54] , [55] . The actual mechanistic basis of this remarkable robustness so far remains elusive . In contrast , at an early time point after introduction of HCV genomes into the cell , the system was found to be substantially more fragile with respect to the number of sensitive parameters . This suggests that therapeutic intervention with HCV replication by DAAs would be most efficient in newly infected cells , emphasizing the potential of such drugs for the prevention of reinfection upon liver transplantation . The processes found to be most sensitive during the early phase of replication were polyprotein translation as well as the RNA polymerization rate of NS5B . Of note , polyprotein cleavage by the viral NS3/4A protease was surprisingly little influential . This , however , has been described before , e . g . in a study examining the role of cyclophilin A for HCV replication [56] . In that study , viral mutations conferring resistance to the cyclophilin A inhibitor Alisporivir ( Debio-025 ) were shown to significantly affect the efficiency of polyprotein cleavage without notably affecting RNA replication of the replicon [56] . This could raise some concern about the first ( very recently ) approved direct acting antivirals for HCV , the NS3/4A inhibitors Telaprevir and Boceprevir [57]: on the one hand , they need to exhibit an extremely high potency of inhibition in order to suppress HCV RNA replication efficiently . On the other hand , there should be comparatively little restrictions to the development of escape mutations rendering NS3/4A resistant to the compounds , owing to the relatively small effect on replication dynamics even in a case where the mutation functionally lowers protease activity as it is predicted by our model . Simply put , the virus can effectively buy itself out of pharmacologic inhibition at only modest fitness costs , and in fact , at least for the first generation of protease inhibitors , this is indeed the case [58] , [59] . In contrast , according to our model analysis , HCV should be far more sensitive towards inhibition of the NS5B polymerase activity . In line with this prediction , an NS5B inhibitor ( HCV-796 ) yielded a significantly faster and stronger response when directly compared to a very potent protease inhibitor ( BILN 2061 ) , both dosed at the same multiples of their respective EC50s [60] . This difference in efficaciousness could even get potentiated when considering the development of escape mutations . Particularly for nucleoside/nucleotide analogues , which target the catalytically active center of NS5B , all so far observed resistance mutations have a negative influence on its polymerase activity [61] . Based on our model , however , lowering NS5B activity is predicted to have a pronounced impact on overall replication efficiency , thereby substantially increasing the fitness costs for such escape mutations . In fact , despite being “genetically easy” ( i . e . single nucleotide exchanges suffice ) such resistance mutations against nucleotidic inhibitors have been shown to be of negligible clinical relevance due to their extraordinarily strong impact on replication efficiency [62] . In general , we want to note that a modeling approach as ours can help in estimating and understanding the sensitivity of HCV replication upon ( e . g . pharmacologic ) inhibition of a particular step in the life-cycle . It cannot , however , generally predict the development of resistance mutations , as the actual number and position of nucleotide/amino acid exchanges required for resistance eventually determine the likelihood of their occurrence and their fitness-cost , respectively . One simplification that we accepted in developing the model is that the formation of the membranous vesicles is modeled as one step ( eq . 7 ) together with the formation of the actual replicase complexes ( i . e . the initiation of negative strand RNA synthesis ) . This is owing to a lack of an experimental handle for the discrimination of “productive” from empty or non-functional vesicles . In fact , it has been shown that the vesicular membrane structures are formed by viral NS protein also in the absence of RNA replication [6] , [63] . Therefore it seems likely that initiation of RNA synthesis will depend on the formation of membrane alterations , but still represents a distinct step in the formation of an active replication site . However , in this two-step scenario , membranous vesicles would form based on the concentration of cytosolic NS proteins ( Ecyt ) and a host factor ( HF ) , and replication complexes ( Rip ) would mainly depend on Tc ( and possibly Ecyt and HF ) and the availability of vesicles . In effect , formation of productive replicative vesicles would again depend on those three species , TC , Ecyt and HF and should in principle be compatible with our simplified one-step model . On similar lines , for reasons of simplicity , our model considers only one single , large replication compartment . This assumption is clearly not correct , as numerous sites of virus induced convoluted membrane structures have been observed in HCV replicating cells [7] and each cell holds approximately 100 negative strand RNAs ( i . e . markers for productive replication complexes ) on average [10] . However , the approximation with a single large replicative compartment should be adequate provided the real number of vesicles is large enough for formation or loss of individual vesicles not to lead to significant sudden changes of viral RNA and protein availability in the cytoplasm . As measurements of replication are technically limited to bulk assessments and cannot probe individual vesicles , for the time being this point cannot be addressed more adequately . Similarly , there might also be ( and likely is ) heterogeneity among cells in terms of kinetics and absolute numbers . Also here , probing individual cells for plus and minus strand RNA as well as for polyprotein production is almost impossible with today's technology , and consequently , our model represents an approximation of the average cellular behavior in a larger population of cells . Curiously , a central result of our study was the conclusion that the assumption of a key host factor was essential to fit our model to the dynamics of RNA replication . This factor was important to explain RNA replication in Huh-7 cells , but might not be as limiting in other HCV permissive cells , e . g . primary human hepatocytes . Moreover , in a physiological setting , restrictions in other steps of the viral life cycle , e . g . sub-threshold receptor levels during entry [64] , [65] or a limitation in the apolipoprotein system required for particle secretion [66] might play critical roles as well . Importantly , also the innate immune response ( and on a larger time-scale also the adaptive one ) poses severe restrictions on viral replication via effector genes , whose molecular identity and functions have only recently begun to be identified [67] , [68] . These influences would need to be included in a future , fully comprehensive model of HCV replication . For our present model , based on Huh-7 cells , however , we have so far neglected any impact by the innate immune system , as we could previously demonstrate that presence or absence of functional immune recognition of HCV by the ( Huh-7 derived ) host cell does not have a measurable effect on its permissiveness [32] . Still , for RNA replication in this single most important cell culture system for HCV , we found a limiting host function involved in the formation of the replication compartment to be crucial to explain the observed replication kinetics . The molecular function of this host factor is still unclear; one or more cellular proteins could be involved , taking part in the formation of the membrane alterations or in the initiation of RNA synthesis . Even a more general condition such as stress tolerance could serve as the host requirement proposed by our model . Since this host factor ( s ) /condition ( s ) HF was sufficient to model the varying RNA replication efficiencies in different Huh-7 populations , we performed gene expression profiling to identify genes potentially defining permissiveness . While our analysis identified 355 genes , whose expression correlated with the degree of permissiveness of the respective cell line , there were no single factors or well-defined pathways that stood out significantly . In order to test the limiting nature of these identified factors for HCV RNA replication , one would have to individually overexpress those genes in low permissive cells and assay for an enhancement in HCV replication . Whereas this was beyond the capacity of our current study , we made use of extensive publicly available data on cellular interaction partners of HCV ( VirHostNet [45] ) and high-throughput RNAi-based knock-down studies [46] , [47] , [48] , [49] in order to recognize genes that had been implicated with HCV before . This approach identified 17 cellular genes whose expression levels on the one hand correlated well with permissiveness for HCV replication , and that , on the other hand , were either reported to at least interact with an HCV protein , or were shown to have a direct impact on HCV replication upon knock-down ( table 2 ) . While for this small sub-set of genes a reliable functional link to HCV could therefore be established , we cannot exclude any of the remaining differentially expressed genes as potentially crucial host factors for HCV; this is true even in spite of a virtually genome-wide coverage of the published screening studies , as such approaches are characterized by extremely high false-negative rates [50] . Therefore , comprehensive future studies need to exploit the information contained in our transcriptomic analysis , systematically testing those host factors for an impact on HCV replication that most significantly correlated with permissiveness . Already during model development , but also throughout our model analyses , the formation and function of the membranous replication compartment was found to be crucial for successful viral HCV replication . Previous literature as well as our model analysis imply that membrane alterations serve at least three distinct purposes . For one , they provide a protected environment for RNA replication , shielding this very sensitive process from the host cell degradative machinery as also shown experimentally before [10] . Without this protection , the viral RNA would quickly be degraded , and replication , according to our model , would become highly vulnerable to stochastic effects due to very low molecule numbers . In fact , should cytoplasmic RNA degradation be only slightly stronger than our mean estimate for μpcyt ( but well within its confidence interval ) , e . g . upon stress or under conditions of an activated immune response , the system would cross a threshold and replication would die off inevitably . Therefore , to compensate for such a lack of protection of the replication machinery , HCV would have to develop a completely different amplification strategy , most likely involving a much higher rate of RNA synthesis in order to maintain sustained replication . This , very likely , would not be compatible with low-level , low profile replication as required for persistence [13] . Secondly , sequestration of viral replicative intermediates , such as double-stranded RNA , into membranous compartments also shields them from recognition by ubiquitous pattern recognition receptors of the intrinsic innate immunity ( which , as described above , is neglected by our current model ) . A third important aspect , however , is the fact that this strict compartmentalization allows for a tight control of viral RNA replication versus protein translation . By limiting the amount of viral and/or host protein inside , the replicative compartment not only protects , but paradoxically also attenuates RNA replication . Presumably , this serves to limit replication to levels sustainable by the cell and permitting low-level persistent replication over a long period of time with very limited detection by the immune system . At the same time , by controlling the amount of newly synthesized RNA released into the cytoplasm , the vesicles indirectly control the amount of protein translation and , in an in vivo situation , particle formation , as was also suggested by another modeling approach [26] . We provide the first comprehensive modeling of the entire RNA replication cycle of a positive strand RNA virus , from the onset of RNA replication to steady state levels . However , membranous replication sites are a hallmark of all positive strand RNA viruses with very different replication strategies . In case of HCV the membranous replication compartment seems to have a rather limiting role in virus RNA replication , probably contributing to viral persistence and chronic disease . In contrast , most positive RNA viruses replicate fast , cause acute diseases and are cleared by the immune system ( e . g . the closely related flaviviruses such as Dengue or West Nile virus ) . Interestingly , in the related group of pestiviruses , pairs of viral isolates have been found , replicating either in a non-cytopathic/persistent or in a cytopathic/acute manner [69] . Upon integration of cellular mRNA sequences into their genomes , dramatically enhancing the efficiency of viral RNA replication , these biotypes switch from well-controlled , persistent infection to an aggressively replicating , cytophatogenic phenotype [70] . Also in case of Sindbis virus , cytopathic replication can be switched to persistence by a single point mutation [71] . Both examples demonstrate a tremendous flexibility to adapt the concept of membranous replication compartments to various replication strategies . It would therefore be highly interesting to use our model as a blueprint for modeling replication kinetics of closely related positive strand RNA viruses following a lytic/acute replication strategy , e . g . Dengue virus or West-Nile-virus . Comparing the principles governing replication of such a virus to the here described strategy of HCV could offer a completely new approach to examining– and eventually comprehending– the general requirements allowing viruses to establish chronicity . Another obvious yet intriguing direction into which our presented modeling approach could be developed , is extending it to comprise the full infectious virus life cycle , including particle production and secretion , receptor binding and cell entry . In fact , two very recent publications studied RNA replication kinetics upon HCV infection [6] , [29] and found a dynamic behavior extremely reminiscent of what we describe here for subgenomic replicons: the initially present RNA is rapidly degraded early upon infection and then starts to replicate exponentially at around 6 to 8 hours post infection , which is reflected in both , plus- and minus-strand RNA signals . This similarity to the kinetics observed in our experiments is remarkable , as initial RNA concentrations are about two to three orders of magnitude less in the infection ( roughly 1–50 genomes per cell ) as compared to our transfections ( ∼4 . 000 genomes per cell ) . The single major difference to the here described situation in a replicon setting is the increasing excess of plus-strand RNA over the minus-strand for late time points ( e . g . 50-fold excess at 72 h ) which seems to be due to decreasing minus-strand levels , while plus-strand RNA basically maintains a steady-state [29] . It is intriguing to speculate that this phenomenon might reflect partitioning of the plus-strand RNA into translation/replication on the one hand , and particle assembly/genome encapsidation on the other hand . As encapsidated genomes would no longer be available for initiation of new replication complexes , minus-strand RNA levels should consequently decrease over time . In order to adapt our model to an actual infection setting , however , we will need to switch to a stochastic model to deal with extremely low copy numbers of RNA per cell . Such situations can be addressed mathematically using the Gillespie algorithm , provided appropriate single cell measurements are available . The model could then also be extended to describe the extracellular steps of the viral life cycle , up to receptor binding and cell entry , which could finally allow for very precise simulation of viral spread through a population of naïve cells . Such a comprehensive model would be highly valuable to examine and predict the effects of therapeutic intervention with viral entry or release as compared to inhibition of intracellular steps of replication . Even more importantly , it could be suited to finally link our fine-grained molecular model of HCV replication to the very interesting patient-level models of HCV infection and therapy dynamics [14] , [72] , and thereby open up new avenues to rationally designing novel therapeutic strategies , but also to understanding the effects of molecule-scale events onto the progression of a complex disease .
All cells were maintained in supplemented Dulbecco's modified Eagle medium ( DMEM ) as described previously [10] . Huh-7 low passage refers to naïve Huh-7 cells , passaged less than 30 times in our laboratory , see also Binder et al . [32] . Huh7-Lunet and Huh-7/5-2 are highly permissive clonal cell lines [32] . Huh7-Lunet NP ( unpublished ) refers to a derivative of Huh7-Lunet , which is significantly less permissive than its parental cell line . For kinetic analyses of HCV RNA replication , the genotype 2a ( JFH1 isolate ) constructs pFKi389LucNS3-3′_dg_JFH ( wild-type ) and pFKi389LucNS3-3′_dg_JFH/ΔGDD ( replication deficient ) [73] were used , as well as the NTR-chimeric constructs pFK-I341PI-Luc/NS3-3′/JFH1/5′Con ( 5′-NTR exchange ) and pFK-I341PI-Luc/NS3-3′/JFH1/XCon ( 3′-NTR exchange ) [22] . Permissiveness of cell lines was assessed using a genotype 1b ( con1 ) replicon , using the plasmid pFK-I341PI-Luc/NS3-3′/Con1/ET/∂g . In vitro transcription of HCV replicons was performed as described previously [22] , [30] . Briefly , plasmid DNA was purified by phenol/chloroform extraction and transcribed with 0 . 9 U/µl T7 RNA polymerase ( Promega ) . RNA was then purified by DNase ( Promega ) digestion , extraction with acidic phenol and chloroform and room temperature isopropanol precipitation . RNA concentration was determined spectrophotometrically and integrity was confirmed by agarose gel electrophoresis . Cells were transfected with in vitro transcribed HCV RNA by electroporation as described previously [22] . For determination of host cell permissiveness ( figure 7 ) , 5 µg of RNA were used for electroporation and cells were seeded into 6-well plates ( 1/12 electroporation per well ) . Samples were lysed at 4 , 24 , 48 and 72 h post transfection and stored at −80°C until measurement of luciferase activity . For time resolved quantitation of HCV replication , 4×106 cells were transfected with 10 µg of HCV RNA , corresponding samples were pooled and cells were seeded into 6-well plates for luciferase assays as described above or into 10 cm cell-culture dishes at a density of 4×106 cells per plate ( 2×106 cells/plate for time points 48 h and 72 h ) for RNA preparation and Northern blotting . For the 0 h RNA sample , 4×106 cells were washed twice with DMEM directly after electroporation , pelleted and lysed in guanidinium isothiocyanate . Other samples were lysed at the indicated time points ( 2 , 4 , 8 , 12 , 18 , 24 , 48 and 72 h ) and lysates were stored at −80°C until further processing . For determination of HCV replication by luciferase activity measurement , all samples of one experiment were frozen at −80°C upon harvesting and thawed simultaneously prior to luciferase detection . Measurements were performed as described in Binder et al . [22] , with all samples measured in duplicate . Luciferase activity was normalized to the input activity assessed at 2 h ( kinetic experiments ) or 4 h ( permissiveness determination ) post electroporation , to correct for transfection efficiency . RNA preparation and Northern blotting were performed according to established procedures [22] . In essence , total cellular RNA was isolated from guanidinium isothiocyanate lysates by a phenol/chloroform based single-step protocol and denatured in glyoxal . Samples were analyzed by denaturing agarose gel electrophoresis and Northern hybridization . For strand specific detection of HCV RNA , radioactively labeled riboprobes encompassing nucleotides 6273 to 9678 of the JFH1 sequence were generated by T7- ( minus-strand detection ) or T3-polymerase ( plus-strand detection ) mediated in vitro transcription of plasmid pBSK-JFH1/6273-3′ [34] . Signals were recorded by phosphorimaging using a Molecular Imager FX scanner ( BioRad , Munich , Germany ) and quantified using the QuantityOne software ( BioRad ) . To determine absolute molecule numbers , signals were quantified using serial dilutions of highly purified plus- and minus-strand in vitro transcripts of known quantity , which were loaded onto the same gel . Cross-hybridization of minus-strand probes with the plus-strand standard was observed to a low extent and corrected for . Permissiveness of eight Huh-7 derived cell-lines was assessed using a standard luciferase replication assay as described above . Total cellular RNA of untransfected cells was then isolated by Trizol extraction according to the manufacturer's protocol ( Invitrogen , Karlsruhe , Germany ) , and gene expression was measured using the Affymetrix Human Genome U133 Plus 2 . 0 platform . Data were normalized in R/Bioconductor using RMA normalization . Genes were filtered using the variance-based ( IQR ) filter in nsFilter , and log2 fold-changes between high and low permissive cells were computed . We then fitted a linear model to the data , predicting replication efficiency in the eight cell lines from the corresponding gene expression values . ANOVA was used to assess statistical significance of individual genes . Hit selection was done using a relatively low threshold of 0 . 2 on the p-value and a log fold-change of at least 0 . 3 , corresponding to a change in expression of approximately 25% . Resulting genes were intersected with published RNAi screening [46] , [47] , [48] , [49] and virus-host protein interaction [45] data as described , yielding a list of 17 host factors that are differentially expressed between the high and low permissive cells , that correlate with replication permissiveness in the eight cell lines used , and that have previously been shown to be associated with HCV infection or replication . Genes were then mapped to pathways and annotated further using DAVID version 6 . 7 [74] , [75] and IPA ( Ingenuity Systems , www . ingenuity . com ) . We developed a mathematical model using ordinary differential equations based on mass action kinetics . The model is subdivided into two compartments: 1 ) initial RNA processing , translation into the polyprotein and polyprotein processing ( cleavage ) occur in the cytoplasm , and 2 ) viral genome replication takes place inside of the replication compartment . A graphical summary of the model is shown in Figure 2C . The following set of equations was used to describe the processes in the two compartments: Cytoplasm ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) Here , Rpunp ( eq . 1 ) represents the number of plus-strand RNA molecules entering the cell upon transfection . This transfected RNA is processed into translation competent Rpcyt ( eq . 2 ) at rate k0 , describing , for example , transport and structural re-folding processes . The processed plus-strand RNA Rpcyt interacts with ribosomes Ribo at a constant rate k1 to form translation complexes Tc ( eq . 3 ) , which degrade at rate μTc . Ribosomes are recovered when translation complexes Tc degrade with rate μTc . Note that , as the total number of ribosomes in the cell ( Ribotot ) is assumed constant , the number of ribosomes available for translation is given by Ribotot – TC , and it is not necessary to introduce a separate equation for ribosomes . Unprocessed and processed RNAs Rpunp and Rpcyt degrade with rate constants μpunp and μpcyt , respectively ( eq . 1 an 2 ) . For simplicity , we assume that 10 ribosomes simultaneously translate the same HCV RNA [76] , therefore , Ribotot represents complexes consisting of 10 ribosomes . Viral polyprotein P is formed from Tc at an effective rate k2 ( eq . 4 ) . When the translation of polyprotein is complete , the translation complex dissociates into plus-strand RNA and ribosomes at rate k2 . Newly produced polyprotein is cleaved with rate kc into the mature viral nonstructural ( NS ) proteins Ecyt ( eq . 5 ) . NS proteins degrade at rate μEcyt . Eventually , plus-strand RNA and NS proteins , most notably the polymerase NS5B , interact in cis and together with NS proteins in trans ( Ecyt ) as well as a cellular factor HF to form a replication complex within the induced vesicular membrane structure . This cis interaction of Rpcyt and translated NS proteins is realized in the model by requiring active translation complexes Tc instead of free Rpcyt for the formation of replication complexes . The host factor HF catalyzes the formation of RIp , at the rate kPin . Once RIp is formed , ribosomes are freed again at rate kPin . This leads to the ternary reaction TC+ECyt+HF→RIp+RIbo , simultaneously describing formation of the replication compartments and initiation of minus strand RNA synthesis , compare also supplementary text S1 and supplementary figure S6 . In turn , HF is freed again when RIp degrades or upon completion of minus strand synthesis . As the total number of host factor molecules in the cell is assumed constant , we can replace HF by HF ( 0 ) – RIp , where HF ( 0 ) is the total number of HF molecules in the cell . Lastly , since we use a luciferase readout to measure polyprotein concentration , we furthermore include a luciferase marker L in the model , which is produced at the same rate as the polyprotein ( k2 ) , however does not require further processing and degrades with rate μL ( eq . 6 ) . Replication compartment ( 7 ) ( 8 ) ( 9 ) ( 10 ) ( 11 ) RIp is the minus-strand RNA initiation complex ( eq . 7 ) , which contains a plus-strand RNA serving as template for the synthesis of minus-strand RNA . Minus strand RNA is synthesized from RIp at rate k4m , yielding double stranded RNA Rds ( eq . 8 ) . We assume minus-strand RNA to be always bound to its complementary plus-strand in a double-stranded replicative intermediate . When the production of minus-strand RNA is complete , RIp dissociates into Rds , HF and viral NS protein E ( eq . 9 ) . Next , Rds interacts again with E to form a plus-strand RNA initiation complex , RIds ( eq . 10 ) , to initiate the synthesis of new plus-strands , Rp , with a constant rate k4p , and dissociates into Rds and E . Newly synthesized plus-strand RNA , Rp ( eq . 11 ) , then leaves the replication compartment at rate kPout to participate in translation , or interacts with the polymerase E and host factor HF to again form the minus-strand RNA initiation complex RIp at rate k3 . For simplicity , we assume that the RNA RIp , Rds , RIds and Rp , and proteins E all degrade with rate μRC . Reaction rates in the model were taken from literature as far as known , or estimated by fitting the model to the experimental data . Following Dahari et al [24] , we used a value of k2 = 100 polyproteins per hour per polysome for protein translation . RNA replication was assumed to occur at a rate of k4m = k4p = 1 . 7 viral RNA molecules per hour per replication complex , assuming plus- and minus-strand synthesis to occur at the same rate [23] , [77] , [78] . Based on an estimated half-life of Luciferase of approximately 2 hours , we estimated the corresponding degradation rate to be μL = 0 . 35 h−1 [79] , [80] . We furthermore estimated the NS protein half-life in the cytoplasm to be around 12 hours , corresponding to a rate of μEcyt = 0 . 06 h−1 [76] , [81] , [82] . We observed from model calibration that the optimization would yield values with μTc>μpcyt , violating the expectation that RNA in translation complexes should be more stable than free RNA in the cytoplasm . We hence added the constraint μTc/μpcyt = 0 . 5 , enforcing a 2-fold higher stability of RNA that is actively translated . We furthermore observed a low sensitivity of model output with respect to parameters k1 , kc , k3 and k5 , compare figure 5 , and hence fixed these parameters based on manual model analysis , for details see supplementary text S1 . Estimation of the remaining 7 model parameters , 3 initial values and a scale factor to convert luciferase measurements into polyprotein molecule numbers was done using multiple shooting , as implemented in the PARFIT package [83] , [84] , [85] . We simultaneously minimized the least squares prediction error on the high and low permissive cells in log-concentration space , using all individual measurements in the objective function . An additional scaling factor was introduced in the optimization problem to convert luciferase measurements for the viral polyprotein to molecule numbers . All model species containing viral plus-strand RNA or minus-strand RNA , respectively , were added for comparison with the experimental data , yielding Rptot = Rpunp+Rpcyt+Tc+RIp+Rds+RIds+Rp for the total plus-strand RNA and RMtot = Rds+RIds for the total negative strand RNA concentrations . Ratios of RNA as reported in literature were furthermore used to constrain the optimization [10] . As some species attain very low values , we compared results of the approximation using differential equations with a stochastic solver ( supplementary figure S7 ) . For details of the parameter estimation and objective function used see supplementary text S1 . Obtained model parameters and confidence intervals are shown in table 1 . To test our model for structural identifiability , we performed a local identifiability analysis at obtained optimal parameter values using SensSB [86] . Results of this analysis are shown in Supplementary Figure S8 . High correlation between two parameters means that a change in the model output caused by a change in one parameter can be compensated by an appropriate change in the other parameter . This then prevents the parameters from being uniquely identifiable despite the output being very sensitive to changes in individual parameters . Parameters for which values were known from literature or which were fixed were also included in this identifiability analysis , to assess their effect on results . These parameters are indicated in grey in the Figure; several of these parameters are highly correlated with other parameters , thus reiterating the importance of experimental measurements for them . Importantly , the identifiability analysis indicates that most of the parameters that had to be calibrated from data showed low correlation to other parameters only , indicating an overall satisfactory identifiability of the model and , in particular , no indication of structural non-identifiability in the model with correlation values close to ±1 . We furthermore calculated confidence intervals on estimated model parameters using the covariance matrix of the parameters , as described in supplementary text S1 . Most of the kinetic reaction rates had reasonable standard errors and confidence bands , while larger uncertainties were observed for the initial values , compare table 1 . This sloppiness is typical for models in systems biology [87] , [88] . Based on our aim to develop a predictive model and not uniquely identify individual reaction rates , our assessment was that the model is sufficiently identifiable for our purpose . Global sensitivity analysis was performed using the extended Fourier Amplitude Sensitivity Test ( eFAST ) [39] , [40] . This algorithm calculates the first and total-order sensitivity indices of each parameter , and assesses the statistical significance of these sensitivity indices by a method based on dummy parameters . For details , we refer to Saltelli et al [89] . In brief , for a given model y = f ( x ) with scalar y and input vector x = ( x1 , … , xn ) , the first order sensitivity index with respect to xi is the expected amount of variance that would be removed from the total output variance , if we knew the true value of xi , divided by the total unconditional variance:Si is a measure of the relative importance of the individual variable xi in driving the uncertainty in the output y . In contrast , the total sensitivity index with respect to a variable xi measures the residual output variance if only xi were left free to vary over its uncertainty range , and all other parameters were known:STi is a measure of how important a parameter is in determining the output variance , either singularly or in combination with other parameters . To assess the significance of obtained indices , eFast furthermore calculates the first and total order sensitivity index for a dummy parameter that is not part of the model . Indices that are not significantly larger than this dummy parameter index should not be considered different from zero [39] . Figures 6 and S4 show the resulting eFAST total order sensitivity indices of viral plus- and minus-strand RNA concentrations and viral polyprotein concentration with respect to the 16 model parameters and three initial values at two different time points , early in the viral lifecycle and after attainment of the steady state levels .
|
Hepatitis C is a severe disease and a prime cause for liver transplantation . Up to 3% of the world's population are chronically infected with its causative agent , the Hepatitis C virus ( HCV ) . This capacity to establish long ( decades ) lasting persistent infection sets HCV apart from other plus-strand RNA viruses typically causing acute , self-limiting infections . A prerequisite for its capacity to persist is HCV's complex and tightly regulated intracellular replication strategy . In this study , we therefore wanted to develop a comprehensive understanding of the molecular processes governing HCV RNA replication in order to pinpoint the most vulnerable substeps in the viral life cycle . For that purpose , we used a combination of biological experiments and mathematical modeling . Using the model to study HCV's replication strategy , we recognized diverse but crucial roles for the membraneous replication compartment of HCV in regulating RNA amplification . We further predict the existence of an essential limiting host factor ( or function ) required for establishing active RNA replication and thereby determining cellular permissiveness for HCV . Our model also proved valuable to understand and predict the effects of pharmacological inhibitors of HCV and might be a solid basis for the development of similar models for other plus-strand RNA viruses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"hepatitis",
"c",
"medicine",
"systems",
"biology",
"infectious",
"diseases",
"hepatitis",
"theoretical",
"biology",
"biology",
"computational",
"biology",
"viral",
"diseases"
] |
2013
|
Replication Vesicles are Load- and Choke-Points in the Hepatitis C Virus Lifecycle
|
Liver fibrosis was viewed as a reversible process . The activation of hepatic stellate cells ( HSCs ) is a key event in the process of liver fibrosis . The induction of senescence of HSCs would accelerate the clearance of the activated HSCs . Previously , we demonstrated that soluble egg antigens ( SEA ) of Schistosoma japonicum promoted the senescence of HSCs via STAT3/P53/P21 pathway . In this paper , our study was aimed to explore whether there are other signaling pathways in the process of SEA-induced HSCs aging and the underlying effect of SKP2/P27 signal on senescent HSCs . Human hepatic stellate cell line , LX-2 cells , were cultured and stimulated with SEA . Western blot and cellular immunofluorescence analysis were performed to determine the expression of senescence-associated protein , such as P27 , SKP2 and FoxO3a . Besides , RNA interfering was applied to knockdown the expression of related protein . The senescence of HSCs was determined by senescence-associated β-gal staining . We found that SEA increased the expression of P27 protein , whereas it inhibited the expression of SKP2 and FoxO3a . Knockdown of P27 as well as overexpression of SKP2 both suppressed the SEA-induced senescence of HSCs . In addition , the nuclear translocation of FoxO3a from the nucleus to the cytoplasm was induced by SEA stimulation . The present study demonstrates that SEA promotes HSCs senescence through the FoxO3a/SKP2/P27 pathway .
Liver fibrosis , a major health problem worldwide [1] , results from different etiologies of chronic liver injury , and eventually progresses into cirrhosis or hepatocellular carcinoma . Recently , liver fibrosis was viewed as a reversible process [2] . After years of prevention and treatment of schistosomiasis in China , the new cases of Schistosoma infection have declined significantly , but there are still thousands of patients suffering from schistosomiasis [3] . The main pathological change of schistosomiasis is the formation of granuloma around the eggs of Schistosoma japonicum ( S . japonicum ) in the liver , leading to liver fibrosis . Studies indicate that the activation of hepatic stellate cells ( HSCs ) is a key event in the process of liver fibrosis . HSCs are activated and then transform to myofibroblasts , once the liver is subjected to stimulations . Activated HSCs synthesize large amounts of extracellular matrix proteins ( ECM ) such as type I or type III collagen , laminin and fibronectin [4] . In the process of liver fibrosis induced by S . japonicum infection , HSCs gather around S . japonicum egg granuloma [5] . Activated HSCs can express a variety of inhibitors of metalloproteinases ( TIMPs ) to prevent the degradation of matrix proteins , resulting in the replacement of normal liver tissue by collagen matrix and the formation of fibrous scar . Therefore , inhibition of the HSCs activation , proliferation and accelerating the clearance of the activated HSCs are key strategies for the prevention and treatment of liver fibrosis [6] . Substantial evidences support the possibility of the reversibility of liver fibrosis [2] . Recently , studies revealed that with the development of pathologic process , the size of egg granulomas at the chronic phase ( 12 weeks ) and the advanced phase ( 24 weeks ) was smaller than that at the acute phase of S . japonicum egg-induced liver fibrosis [7] . Researches indicate that the reversion of liver fibrosis is closely related to the increase of the apoptosis of HSCs . Expression of the tissue inhibitor of metalloproteinase-1 ( TIMP-1 ) decreased , and the synthesis of metalloproteinases ( MMPs ) such as MMP-1 and MMP-13 increased , thereby inhibiting HSCs activation and proliferation , increasing the clearance of activated HSCs as well as the degradation of collagen fiber , and eventually alleviating liver fibrosis [8 , 9] . Studies showed that the induction of senescence of HSCs would accelerate the clearance of the activated HSCs as well [10] . The senescent cells usually display a cell cycle arrest in the G0 or G1 phase but maintain the metabolic activity [11] . Once senescent , senescence-associated β-gal ( SA-β-Gal ) , the specific maker of senescence , is detected in these cells . In our previous study , we demonstrated that SEA induced the HSCs senescence through the STAT3/P53/P21 pathway [12] . Besides , it has been well established that FoxO3a signaling cascade is implicated in the senescent process of multiple cells [13–15] . It has been revealed that FoxO3a inhibited the senescence of hepatocytes [16] . Additionally , S phase kinase associated protein 2 ( SKP2 ) was reported to suppress cellular senescence induced by oncogenic stimuli independent of ARF/p53 signaling . And cell cycle inhibitor P27 , the SKP2 substrate , is targeted by SKP2 for ubiquitination and degradation [17 , 18] . In this study , we investigate whether the FoxO3a/SKP2/P27 signaling participates in the SEA-induced HSCs senescence .
SEA of S . japonicum were obtained from Jiangsu Institute of Parasitic Diseases ( China ) . SEA was sterile-filtered and endotoxin was removed with Polymyxin B agarose beads ( Sigma , USA ) . Limulus amebocyte lysate assay kit ( Lonza , Switzerland ) was used to confirm the removal of endotoxins from the SEA as previously described [19] . Primary antibodies for FoxO3a , SKP2 , P27 and AKT were purchased from Santa Cruz Biotechnology ( USA , antibody dilution for Western blot of all antibodies from this company is 1:200 ) . Primary antibody for phospho-AKT was purchased from Cell Signaling Technology ( USA , antibody dilution for Western blot is 1:1000 ) . All of the secondary antibodies were obtained from Santa Cruz Biotechnology ( USA , antibody dilution is 1:2000 ) . The staining kit for SA-β-Gal was purchased from GenMed Scientifics Inc ( USA ) . LX-2 cells , the ‘immortalised’ human HSCs , were provided by Xiangya Central Experiment Laboratory ( Hunan , China ) and maintained in DMEM with 10% Fetal Bovine Serum in a humidified incubator with 5% CO2 . Culture medium was replaced every day and cells were subcultured with trypsin when they were at 80% confluence . Cells were lysed in RIPA cell lysis buffer including protease inhibitor ( 1mM ) and phosphatase inhibitors ( 1mM ) . Equal amounts of protein extract were separated by SDS-PAGE and then transferred onto polyvinylidene difluoride ( PVDF ) membranes . The membranes were blocked in 5% nonfat milk for 2 hours , incubated with the indicated primary antibodies at 4°C overnight , and then incubated with horseradish peroxidase ( HRP ) -conjugated secondary antibodies for 1 hour at room temperature . SA-β-Gal staining was performed according to the instruction of SA-β-Gal staining kit , in which cleaning solution , fixation fluid , acidic solution and staining fluid were provided as the main kit contents . Briefly , LX-2 cells were washed with cleaning solution and fixed by fixation fluid for 5 minutes at room temperature . Afterwards , cells were washed by acidic solution twice and stained with staining fluid for 16 hours at 37°C . Finally , SA-β-Gal staining positive cells were assayed using a bright field microscope . For the Immunofluorescence staining , cells were seeded in 6-well culture plates and fixed with 4% paraformaldehyde . Afterwards , cells were permeabilized with 0 . 1% Triton X-100 and then blocked in 5% BSA . After that , cells were incubated with FoxO3a antibody ( dilution is 1:50 ) and visualized with Alexa Fluor 568 conjugated secondary antibody ( Invitrogen , USA , antibody dilution is 1:200 ) under a fluorescent microscopy . pcDNA3 . 1 plasmid was digested with EcoRI and BamHI ( TaKaRa , China ) , and CDS region of SKP2 ( GenBank: NM_005983 ) was subcloned into pcDNA3 . 1 vector to generate the recombinant vector pcDNA3 . 1-SKP2 . The recombinant plasmids were verified by restriction analysis and sequencing . LX-2 cells were transfected with P27 siRNA ( GenePharma , China ) or pcDNA3 . 1-SKP2 overexpression plasmid by Lipofectamine 2000 reagents ( Invitrogen , USA ) according to the manufacturer’s instructions . After 24 hours , cells were subjected to various stimulations for indicated time . Data is expressed as mean ± SEM ( standard error of mean ) of three independent experiments . All p values were calculated using a two tailed paired Student’s t test or a one way ANOVA . p < 0 . 05 was considered as statistically significant .
Previously , we found that SEA-induced LX-2 cells senescence via the STAT3/P53/P21 pathway [11] . Since P27 , the cell cycle inhibitor , plays an important role in cellular senescence and SKP2 could cause a decrease in the level of P27 expression [13 , 20] , we next verified whether P27 signaling pathway is implicated in the progress of LX-2 senescence . As illustrated in Fig 1 , Western blot analysis showed that SEA markedly increased the expression of P27 , but decreased the SKP2 protein level . Furthermore , the expression of P-AKT , the upstream of P27 , was also significantly decreased under SEA exposure , although the total expression of AKT was not affected . Apart from the regulation of the level of P27 by SKP2 , P27 is also regulated by the FoxO3a protein at the transcriptional level [14] . Also , FoxO3a could be regulated by AKT and 14-3-3 protein [21] . Thus , we further investigated whether FoxO3a was involved in the senescence of LX-2 cells induced by SEA . The results of Western blot indicated that FoxO3a was significantly inhibited by SEA stimulation in the LX-2 cells ( Fig 2A ) . Besides , cell immunofluorescence assay confirmed that FoxO3a was transferred from the nucleus to the cytoplasm after SEA treatment ( Fig 2B and S1 Fig ) . These results suggested that FoxO3a was implicated in the SEA-induced senescence in LX-2 cells . In order to further verify the role of P27 in the SEA-induced senescence in LX-2 cells , P27 specific small interfering RNA was used to knockdown the expression level of P27 protein in LX-2 cells . As illustrated in Fig 3 , the SA-β-Gal staining showed that the senescent LX-2 cells significantly increased accompanied with the upregulated P27 upon SEA stimulation . Nevertheless , the senescence of LX-2 cells induced by SEA was reversed by the P27 siRNA . These results suggest that P27 is a key regulator in the senescence of LX-2 cells induced by SEA . Studies indicate that SKP2 plays an important role in the process of cellular senescence [17 , 20 , 22] , thus , we explored whether SKP2 is a regulator in the SEA-induced senescence in LX-2 cells . We found that SEA inhibited the expression of SKP2 ( Fig 1 ) . In order to further investigate the potential mechanism of SKP2 in the process of SEA-induced senescence , specific SKP2 over expression plasmid was constructed and transfected into LX-2 cells , and then the efficiency was confirmed by Western blot analysis . The results showed that the SKP2 protein expression in LX-2 cells was enhanced after transfection with SKP2 over expression plasmid ( Fig 4A ) , and the high expression of SKP2 could inhibit the senescence of LX-2 cells induced by SEA ( Fig 4B ) . These results suggest that SKP2 can inhibit LX-2 cells senescence mediated by SEA . To explore the mechanism of SKP2 on senescence in LX-2 cells , we also examined the expression of P27 , and we found that the expression of P27 in LX-2 cells was significantly restricted after the overexpression of SKP2 ( Fig 4A ) . On the contrary , the expression of SKP2 in LX-2 cells was not affected by the knockdown of P27 expression ( Fig 3A ) .
It has been well accepted that activation of quiescent HSCs is responsible for the excessive production of ECM in liver fibrosis [23 , 24] , and there has been increased recognition in utilizing functions of HSCs for therapeutic applications to reverse liver fibrosis [25] . Thus , preventing the activation of HSCs and increasing the clearance of activated HSCs are viewed as promising anti-fibrotic strategies [4 , 26 , 27] . Among these , induction of activated HSCs apoptosis and inhibition of activated HSCs proliferation are the common anti-fibrotic strategies to block liver fibrosis . For example , we found SEA could induce HSC apoptosis and inhibit activation of HSCs under some suitable conditions [19] . In addition , studies showed that the senescence of HSCs would block the development of liver fibrosis . Kong X et al . have demonstrated that IL-22 induced HSCs senescence and restricted the development of liver fibrosis in mice [10 , 28 , 29] . Which are different from quiescent HSCs , senescent HSCs often manifest as SA-β-Gal staining positive cells . In the previous study , our results showed that more SA-β-Gal staining positive cells could be found in SEA-treated LX-2 cells and SEA decreased the expression of α-SMA in LX-2 cells partially due to SEA-induced senescence [11] . It has been shown that P53 , tumor suppressor protein , plays a critical role in the induction of senescence . We have recently shown that SEA induced HSCs senescence through STAT3/P53/P21 pathway . SEA increased the expression of P-STAT3 , P53 and P21 . And knockdown of STAT3 or P53 inhibited the SEA-induced senescence of HSCs [12] . Besides the P53-P21 and P16-Rb signaling pathways [30–32] , there are other signaling pathways that promoting the development and progression of cellular senescence . The inactivation of retinal vascular tumor suppressor factor ( VHL ) can decrease the expression of SKP2 and increase the expression of P27 , and then induce cellular senescence [33] . Consistent with this result , the overexpression of HTLV-1 Tax protein also reduced the expression of SKP2 and accompanied with the occurrence of cellular senescence in human T cells [34] . These results suggest that the decrease of SKP2 and the induction of P27 might play direct roles in cellular senescence . SKP2 is a member of the F box protein family , and the formation of the SKP2-SCF complex exhibits the E3 ligase activity . Li Z et al . showed that SKP2 regulates cell cycle and cell proliferation by degradation of its downstream molecules such as P27 , a cell cycle inhibitor [35–37] . And recent studies have shown that inactivation of SKP2 induces cellular senescence , in which the cell cycle inhibitor P27 and P21 expression are enhanced [17] . Therefore , we suspect that SKP2 is involved in the process of SEA-induced LX-2 cell senescence . We found that SEA markedly inhibited the expression of SKP2 , but enhanced expression of P27 ( Fig 1 ) . In order to further verify the role of SKP2 and P27 in the senescence of LX-2 cells , we transfected P27 siRNA to LX-2 cells to knockdown the P27 protein expression and transfected SKP2 overexpression plasmid to upregulate the expression of SKP2 . These results further confirmed that SEA-induced cellular senescence was partially dependent of SKP2/P27 pathway ( Fig 3 and Fig 4 ) . In addition to the regulation of the post transcriptional level of P27 by SKP2 , P27 is also regulated by the FoxO3a protein at the transcriptional level [14] . The data shows that FoxO3a participates in the process of many kinds of cell senescence . Xu-Feng et al . found that FoxO3a can inhibit the senescence of cardiovascular endothelial cells by regulating the cell cycle mediated by ROS [14] . Similarly , in the experiment of Kyoung Kim H et al . , FoxO3a also exhibited an inhibition effect on human dermal fibroblast senescence . The experimental results demonstrated that knockdown of FoxO3a could promote the cell senescence [38] . Therefore , we further verify the effect of FoxO3a on the senescence of SEA-induced LX-2 cells , and our experimental results are consistent with the above phenomena . In SEA-treated LX-2 cells , FoxO3a protein expression was significantly inhibited , and FoxO3a occurred nuclear transfer from the nucleus to the cytoplasm under the role of SEA ( Fig 2B ) . Thus , FoxO3a is a key regulator in the SEA-induced senescence of LX-2 cells . To our knowledge , AKT kinases are critical players in PI3K-mediated signal transduction pathways [39] . AKT phosphorylates downstream substrates to regulate cell growth , proliferation , apoptosis , senescence , and other processes [40] . Cong Fu et al . found that P-AKT expression was down-regulated during the process of cellular senescence induced by H2O2 [41] . Studies demonstrated that AKT phosphorylated FoxO proteins , leading to the negative FoxO regulation via triggering its nuclear exclusion [21] . In addition , AKT can also promote the degradation of P27 [42] . In the present study , our results showed that the expression of P-AKT was inhibited by the SEA stimulation ( Fig 1 ) . In conclusion , SEA might slow down the progression of liver fibrosis by promoting HSCs senescence through the FoxO3a/SKP2/P27 pathway . Our previous and present findings provide evidence supporting a possible mechanism by which SEA induces senescence in LX-2 cells ( Fig 5 ) and these provide a potential target of the clinical research of liver fibrosis .
|
Activation of hepatic stellate cells ( HSCs ) is a key event of liver fibrosis . Induction of activated HSCs apoptosis and inhibition of activated HSCs proliferation are the common anti-fibrotic strategies to block liver fibrosis . The induction of senescence of HSCs is responsible for the clearance of the activated HSCs as well . Senescence of HSCs is mediated by exposure to soluble egg antigens ( SEA ) of Schistosoma japonicum via STAT3/P53/P21 pathway . In this study , we found that SEA induced the senescence of HSCs , accompanied with the increased the expression of P27 protein and the decreased expression of SKP2 and FoxO3a . Either knockdown of P27 or overexpression of SKP2 alleviates the SEA-induced senescence of HSCs . Moreover , SEA droved the nuclear translocation of FoxO3a from the nucleus to the cytoplasm . Hence , the present study demonstrates that SEA promotes HSCs senescence through the FoxO3a/SKP2/P27 pathway .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"schistosoma",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"cell",
"cycle",
"inhibitors",
"senescence",
"helminths",
"gene",
"regulation",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"animals",
"liver",
"diseases",
"plasmid",
"construction",
"physiological",
"processes",
"developmental",
"biology",
"protein",
"expression",
"organism",
"development",
"gastroenterology",
"and",
"hepatology",
"dna",
"construction",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"schistosoma",
"japonicum",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"small",
"interfering",
"rnas",
"liver",
"fibrosis",
"gene",
"expression",
"molecular",
"biology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"aging",
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"rna",
"cell",
"staining",
"cell",
"biology",
"nucleic",
"acids",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"organisms"
] |
2016
|
Soluble Egg Antigens of Schistosoma japonicum Induce Senescence of Activated Hepatic Stellate Cells by Activation of the FoxO3a/SKP2/P27 Pathway
|
Learning has been studied extensively in the context of isolated individuals . However , many organisms are social and consequently make decisions both individually and as part of a collective . Reaching consensus necessarily means that a single option is chosen by the group , even when there are dissenting opinions . This decision-making process decouples the otherwise direct relationship between animals' preferences and their experiences ( the outcomes of decisions ) . Instead , because an individual's learned preferences influence what others experience , and therefore learn about , collective decisions couple the learning processes between social organisms . This introduces a new , and previously unexplored , dynamical relationship between preference , action , experience and learning . Here we model collective learning within animal groups that make consensus decisions . We reveal how learning as part of a collective results in behavior that is fundamentally different from that learned in isolation , allowing grouping organisms to spontaneously ( and indirectly ) detect correlations between group members' observations of environmental cues , adjust strategy as a function of changing group size ( even if that group size is not known to the individual ) , and achieve a decision accuracy that is very close to that which is provably optimal , regardless of environmental contingencies . Because these properties make minimal cognitive demands on individuals , collective learning , and the capabilities it affords , may be widespread among group-living organisms . Our work emphasizes the importance and need for theoretical and experimental work that considers the mechanism and consequences of learning in a social context .
Associative learning tunes an organism's behavior to exploit statistical patterns in the environment and can improve decision-making accuracy across a wide range of scenarios [1]–[2] . In the vast majority of experiments on learning , the subject of study has been a single individual in isolation ( see [3]–[4] for reviews ) . When learning alone , there is a direct relationship between an animal's intentions and its actions: the animal observes cues in the environment and performs a behavioral response . The consequences of the behavior ( such as a reward or punishment ) may alter the animal's valuation of the environmental cues , resulting in a feedback loop that gradually tunes its behavior to its environment [3]–[7] . In contrast to this relatively simple scenario , many animals – including the majority of species commonly employed in learning experiments , such as rats , pigeons , and humans – live and forage naturally in social groups . Sociality offers many benefits to individuals , including improved sensing and decision-making [8]–[9] , decreased risk of predation [10]–[16] , improved foraging success [8] , [16]–[21] , and the capacity for thermoregulation [22] . For these and other species ( e . g . , fish [8] , [16]–[18] , [20]–[21] , [23] , birds [24]–[25] , ants [20] , honeybees [26] , cockroaches [27] primates [28]–[29] , and meerkats [30] ) , decisions are not made in isolation . Instead , in order to preserve the benefits of sociality , animal groups often must come to a consensus regarding where and when to travel or forage , despite the presence of dissenting opinions . While not universal amongst social animals , consensus decision-making is widespread in nature [15] , [31]–[34] , but does not necessarily imply that individuals are altruistic or highly cooperative . While members of some groups may be highly related ( such as ants , honeybees , and primates ) , for many other species ( such as some fish and birds ) , group members are unrelated to each other , and individuals obtain direct fitness benefits from maintaining group cohesion . These benefits provide a strong incentive for individuals to remain together , providing a platform for other emergent phenomena such as collective learning , which we explore here . A common means by which consensus is achieved in animal groups is through relatively local responses to the positions or motion of others . Thus , in many species , such as schooling fish [16] , [23] , [35]–[37] or flocking birds [25] , [38]–[39] , individuals must reconcile any personal directional preferences with their social tendency ( to avoid isolation , and to copy the movement decisions made by others ) . Spatially explicit models of collective movement ( or ‘swarming’ ) are commonly employed to describe mobile animal groups . In these models , individuals interact only with near neighbors , such as individuals within a certain Euclidian distance ( metric models ) or a set number of nearest neighbors regardless of distance ( topological models ) [38] , [40] . These neighbors may change through time due to the motion of individuals ( time-varying networks ) [23] , [35]–[36] , [39]–[45] . For this class of opinion dynamics models , groups are typically highly cohesive , and the motion of groups is well approximated by simple majority rule when collectively deciding between discrete options ( see supplemental text S1 and figures S1 , S2 ) . Effective consensus thus emerges from local interactions among individuals . Although individuals cannot explicitly ‘tally votes , ’ they nevertheless exhibit the capacity to select , collectively , the direction preferred by the majority when conflicting preferences exist [23] , [25] , [43] , even in the presence of a ‘strongly opinionated’ minority [23] . Consequently it is not necessary to simulate the full spatial dynamics to capture accurately the outcome of consensus decision-making by organisms [23] . One consequence of consensus decision-making ( regardless of the precise mechanism by which consensus is achieved ) is that it breaks the direct relationship between individual preference and action . An individual's preferences may be overridden by those of others , such that the individual experiences a part of the environment that it would not have had it been alone . This alters what individuals learn about their environment and also implies that learning by multiple grouping individuals becomes coupled; the preferences of one individual can affect what another experiences , and what one individual learns can affect the future learning of other individuals in the group . Social learning allows individuals of many social species to learn by observing the behaviors of conspecifics [46]–[52] . Individuals tend to follow the decisions of others when their personal information is unreliable [53] or costly to acquire [54] . Nonetheless , associative learning has not been investigated in a social context . Furthermore , the majority of experiments on social learning study a single test subject ( the observer ) , separated from conspecific demonstrators ( e . g . [53]–[59] ) . In a freely behaving group , however , each individual can simultaneously act as demonstrator and observer , resulting in a coupling between preferences , which potentially affects the learned behavior of all individuals in the group . The impact of these coupled dynamics on associative learning in animal groups has yet to be explored , despite the fact that associative learning ( whereby individuals learn to associate environmental cues with rewards ) , occurs in all organisms with a nervous system [60]–[61] . Since consensus decisions break the direct feedback between preference and experience , it is not clear to what degree learning is beneficial in a collective context , whether learning rules in a social context need to be more complex ( such as group size or context dependent ) in order to be effective , or how learning in isolation and subsequently pooling opinions as a group compares with learning as part of a collective . Furthermore , natural environments typically contain not one , but potentially many informative cues , and a crucial challenge for animals is to learn the appropriate relative usefulness of the cues in order to maximize decision accuracy . Optimal voting theory [62] demonstrates that the relative value of environmental cues depends on group size as well as the properties of the cues . Similar to decision-making in isolation , the reliability of a cue ( the probability that it accurately predicts a reward or punishment ) is important . Unique to collective decision-making , however , and of central importance , is the observational correlation of a cue ( the similarity between two individuals' observations ) ( figure 1 ) [63] . In nature , some cues may be subject to relatively low observational correlation , such as cryptic visual cues , where individuals exhibit a relatively independent probability of correctly observing accurate information from the cue [21] . Other cues , however , likely result in high correlation , such as loud auditory cues , strong environmental odors , or large visual landmarks that can readily be perceived by all individuals in the group . For high correlation cues , group members perceive similar observations of the cues , such that there is a high probability that they all receive true ( or false ) information ( figure 1 ) . Because correlations decrease the independence of observations made by different group members , they limit the benefits derived from aggregating observations [62]–[63] . In general , for group-living animals the optimal behavior is to rely primarily on those cues that are less correlated and those that more reliably lead to rewards [62] . Here we present a general framework for studying collective learning and consensus decision-making in animal groups ( figure 2 ) . In this framework , we simulate individual associative learning as in the existing literature , i . e . , we do not make any new assumptions regarding the mechanisms by which individuals learn to associate cues with rewards , nor do we afford additional cognitive abilities to individuals . However , we place individual learning within the context of consensus decision-making , as exhibited by many self-organized animal groups [15] , [32]–[34] . Our framework is agnostic to the mechanism by which animal groups reach consensus , and thus our conclusions are consistent with both spatial and non-spatial models of collective decision-making . This allows us to focus on the coupled dynamics between consensus decisions and associative learning , a previously unexplored aspect of animal collective behavior .
In order to gain information about the quality of the two options , individuals observe environmental cues , which could be odor , visual , auditory , or other sensory cues ( figure 2a ) . Each cue indicates to each individual that one or the other option is superior on that particular trial . For simplicity we assume two such cues . Since individuals in groups must differentiate between cues with different degrees of observational correlation in order to improve collective accuracy [62] , in our model one of the cues has low correlation ( i . e . , observations of the cue by individuals in the group are independent of each other , such that individuals may have opposing information from this cue regarding which is the superior option ) , while the other cue has high correlation ( i . e . , all individuals make the same observation of the cue and agree about which option this cue indicates is superior; figure 1 ) [63] . The reliability of a cue is the probability that it correctly predicts the superior option . These reliabilities are denoted by and for the low and high correlation cues , respectively , and can range from 0 . 5 to 1 . Effective collective learning would allow individuals in groups to give additional value to the low correlation cue ( beyond its reliability ) , due to the benefit of multiple independent observations that cue affords . Consequently , the most interesting scenario is that in which one cue has lower correlation while the other cue has higher reliability , i . e . , . Individuals translate their observations of the two environmental cues into a discrete preference , or vote , for one of the two options . Following well-established psychological models of decision-making by isolated individuals , we assume that the individual rule is to vote for an option with a probability proportional to the sum of the associative strengths ( see below ) of all of the environmental cues that indicate that option . Thus , individuals vote for option A with a probability and option B with probability , where is the current trial , and indicates a particular individual . In our model , because we have only two cues and two options , from the perspective of an individual there are only two possible scenarios: either the two cues both indicate the same option is superior , or they indicate different options . When the two cues indicate that the same option is superior , the voting rule implies that an individual always votes for that option . However , when the two cues indicate opposing options , an individual votes for the option indicated by the low or high correlation cue with a probability proportional to the associative strength of that cue . We denote the probability of choosing the option indicated by the low correlation cue as , irrespective of whether the low correlation cue indicated option A or option B . Similarly , the probability that an individual votes for the option indicated by the high correlation cue is ( figure 2b ) . Once individuals have formed an opinion about which option they consider superior , these opinions must be aggregated in order to produce a collective decision . For some species , social interactions are weak and temporary . Other species , however , are strongly social , and empirical work has shown that , despite employing different interaction rules , many animal species , including primates [28]–[29] , meerkats [30] , fish [16]–[18] , [20] , [23] , and insects [20] , [27] , typically make consensus decisions . Our model does not consider the precise mechanism by which individuals interact , since here it is the outcome of consensus decision-making , and its relationship to individual associative learning , that is important ( however , we demonstrate , in supplemental text S1 and figure S2 , that considering the specifics of interactions , such as by simulating local spatial interactions among individuals [23] , [41] , [43] , does not affect our conclusions ) . Based on experimental evidence from many types of animal groups [17]–[18] , [20]–[21] , [23] , [32]–[33] we assume that individuals can , and often do , select the option preferred by the majority ( figure 2c ) . Further empirical and theoretical work has demonstrated that the presence in the group of individuals with no preferences can even strengthen majority rule in animal groups [23] . As shown by spatially-explicit models of mobile animal groups and in experiments , when there are equal numbers of votes for each option , the group is able to avoid a deadlock and chooses an option randomly [23] , [25] , [43] ( supplemental text S1 and figure S2 ) . After the group chooses one of the options , individuals experience the outcome ( the presence or absence of a reward ) and employ an associative learning rule to update their knowledge of the environment based on this experience ( figure 2d ) . Following standard models of learning in the psychology literature , knowledge of the environment is encoded by an ‘associative strength’ for each environmental cue . Each individual stores two associative strengths , and , representing , respectively , the individual's valuation of the low and high correlation cue . Individuals in our model do not explicitly estimate the size of the group they are in , nor the observational correlation or reliabilities of the cues , which all contribute to determining the optimal voting behavior ( see below ) . It is not known to what extent animals are aware of the size of the group to which they belong , and it is likely that many animals under consideration in this model are unable to accurately estimate group size , either because of limited cognitive abilities , because the group may be large or fluctuating , making estimates of its size difficult , or because of the local nature of interactions [64] . Similarly , it is not known whether individuals can estimate the observational correlation of cues; therefore , in this model we employ a conservative approach , and assume that they are unable to do so ( also , as we will show , they need not be able to do so ) . In short , we do not make new assumptions about the process by which associative learning occurs [3]–[6] . At the start of a simulation , all individuals lack any knowledge of the two cues and therefore the associative strengths for both cues are identical and very small . Also , following standard models of learning , individuals update the associative strength ( s ) only of the cue ( s ) that indicated the option that was ultimately selected by the group . Associative strengths of cues are updated according to the following learning rule , which is similar to the well-known and experimentally-validated [3]–[6] , [65]–[69] Rescorla-Wagner rule: , where is the learning rate ( here taken to be 0 . 1 ) , is 1 if the option selected by the group was the superior option and 0 if it was not [5] , [7] , and represents the associative strength of any cue that indicated the option chosen by the group . In general , this individual learning rule increases the associative strength of cues that are consistently paired with a positive outcome ( the superior option ) and decreases those that are paired with a negative outcome ( the inferior option ) and therefore serves as a memory of past events . Because individuals observe independent and potentially different information from the low correlation cue , only a fraction of the group will update the associative strength for that cue on a given trial . This results in individuals in a group potentially learning different associative strengths for the cues despite sharing a common experience of decision outcomes . The associative strengths are related to the voting behavior in the following way: . Equivalently , individuals vote for an option proportionally to the total associative strength of the cues they perceive as indicating that option . This linear mapping between associative strengths and voting behavior is common in models of learning [7] , although we explore alternate mappings and demonstrate in the supplemental text S2 and figure S3 that this does not impact the results . During the course of repeated trials , an individual's associative strengths are modified , leading to a change in its probabilities and of voting for the two options , which are the direct determinants of the group's resulting decision accuracy . We simulate learning dynamics for a wide range of group sizes and across all combinations of cue reliabilities and in order to assess how collective learning functions under different conditions . In the model framework presented ( figure 2 ) , we have deliberately made biologically realistic but relatively simple assumptions . However our model is robust to deviations from these assumptions . For example , as we show in supplemental text S3 and figure S4 , the general conclusions we arrive at do not depend on the exact choice of the collective decision rule by which consensus decision-making is achieved , nor on the specific individual voting rule ( linear or nonlinear ) ( supplemental text S2 and figure S3 ) . In addition , though not addressed here , the model framework can readily be tailored to generate predictions about specific behavioral contexts or animal species , including species in which consensus is not strongly enforced , or in which individuals have varying degrees of influence in the group decision , due to behavioral syndromes , differing physiological needs , or dominance hierarchies . Alternate learning rules may also be studied . The core ingredients are merely that ( 1 ) individual experiences are influenced by other group members and ( 2 ) learning occurs with regard to the experienced outcome , not the individually preferred one .
In the case of non-social animals , or those in isolation , if both cues indicate that the same option is superior , maximizing reward rate requires an individual to choose that option . However , if the two cues indicate that different options are superior , then the individual should choose the option indicated by the more reliable cue: if and if ( where asterisks denote the optimal behavior ) . If we simulate such a case , we find that isolated individuals do learn to give greater weight to the option indicated by the more reliable cue , such that when and when ( figure 3a ) . This result is compatible with previous experiments on isolated animals [70]–[71] . If the collective learning process is unaffected by the observational correlation of cues or group size , we might expect the learned voting behavior of individuals in groups to be identical to that of isolated individuals . This is not what we observe . As group size increases , the learned voting behavior changes , such that individuals rely more heavily on the low correlation cue for a given environment ( figure 3b–d ) , indicative of effective collective learning . For relatively large groups , individuals rely primarily on the high correlation cue only when that cue is extremely reliable ( figure 3c–d ) . Consensus decision-making therefore results in learned individual voting behavior that is markedly different from that exhibited by isolated individuals under identical environmental conditions . We find that the coupling between the learning of group members allows individuals to incorporate observational correlations , reliabilities , and group size into their valuation ( associative strength ) of the cues in a way that allows them to make substantially more accurate consensus decisions . The above results demonstrate that grouping individuals exhibit learned voting behavior that depends not only on cue reliability , but also on the observational correlation of environmental cues , as well as group size ( without requiring them to be able to estimate any of these explicitly ) . However , it is not clear , given the environmental conditions , how close the resulting performance is to that which is optimal . To investigate this , we derived the optimal individual voting behavior that maximizes collective accuracy , for any environmental condition ( and ) and group size . In the case where both cues indicate that the same option is superior , an individual should vote for the indicated option . However , when the two cues indicate that different options are superior , the optimal behavior is to vote for the option indicated by the low correlation cue with probability when its reliability is greater than that of the high correlation cue , when the high correlation cue is very reliable , and otherwise , where , , and ( see supplemental text S4 for the complete proof ) . In short , when the two cues indicate different options are superior , the optimal voting behavior is to choose exclusively the option indicated by the high or low correlation cue if its reliability is sufficiently high , and otherwise to exhibit a mixed strategy in which individuals probabilistically choose either option ( figure 3e–h ) . We illustrate how the collective accuracy varies with the individual voting behavior for a range of environmental conditions and group sizes , and we show the optimal voting behavior ( yellow triangles ) and the learned voting behavior ( black stars ) on this landscape ( figure 4 ) . When ( black lines ) , it is always optimal to choose exclusively the option indicated by the low correlation cue regardless of group size . When ( red lines ) , individuals in isolation should value the two cues equally but , in groups , should rely exclusively on the low correlation cue . When ( blue lines ) , individuals in isolation should choose exclusively the option indicated by the high correlation cue . However , as group size increases , the optimal behavior gradually shifts towards greater reliance on the low correlation cue . In all cases , we observe that the learned behavior closely tracks the optimal behavior ( figure 4 ) . We generalize this result by showing the collective accuracy as a result of the collectively learned voting behavior for all environments and a wide range of group sizes ( supplemental figure S5 ) . We further show this accuracy as a fraction of the maximum possible accuracy , achieved by the optimal voting behavior ( figure 5a–d ) and find that across all conditions , the achieved accuracy is extremely close to the maximum possible . An implicit assumption in studies of collective intelligence is that the ‘wisdom of crowds’ accrues due to individuals pooling knowledge that was learned independently . If true , then one would expect individuals that exhibit a voting behavior learned in isolation , and whose opinions are subsequently pooled into a group decision , to also exhibit a high degree of collective intelligence . In fact , we find that such groups perform relatively poorly ( figure 5e–h ) , and do so increasingly for larger group size . Therefore , we find that it is not sufficient for individuals to learn in isolation and to subsequently pool their knowledge . Instead , it is important for individuals to incorporate observational correlation and group size into their valuation of a cue , for which collective learning is critical . In this model , and based on theoretical and empirical evidence [17]–[18] , [28] , [34] , we assumed that animal groups make decisions through simple majority rule . However , this is only one particular method of aggregating opinions . For example , an alternative consensus decision rule is to decide on an option only if a minimum proportion of the group votes for that option ( <50% or >50% , representing submajority and supermajority thresholds , respectively ) . This may occur in the context of predator detection where , because of asymmetric costs , the reaction of a small proportion of the group may cause the entire group to flee [30] , [72] . Limiting a group to simple majority rule , rather than the more general sub- and supermajority thresholds , could , in principle , constrain the accuracy that a group may attain . The provably optimal voting rule for groups in which members indicate their vote , explicitly count votes , and can adopt any type of vote aggregation rule ( capabilities unlikely to be available to most animal groups ) was found by Nitzan and Paroush [62] , in the context of human organizations such as juries , governing bodies , and medical panels . They found that the optimal strategy is to adjust the majority threshold according to the cue reliabilities and group size . We compared the efficacy of this ‘globally optimal’ group consensus decision rule to the optimal individual voting behavior with the constraint of simple majority rule that we identified and found that the two rules result in nearly identical collective accuracies ( supplemental text S5 and figure S6 ) . Therefore , the decentralized rule that many animals follow , in which a probabilistic individual behavior is employed instead of a global supermajority rule , poses very little restriction on the collective accuracy that can be achieved by groups . That the collectively learned individual voting behavior substantially outperforms the behavior learned in isolation and subsequently expressed in a group context suggests that collective learning may play an important role in group-living . However , many animals exist in ‘fission-fusion’ populations in which groups readily merge and split over a period of weeks [73] , days [74] or even minutes [75]–[81] . Thus individuals may not repeatedly learn about their environment with the same group members or in the same group size . Furthermore , natural environments are dynamic , and cues can change in their reliability in predicting the location of food or predators ( for example , food availability may be seasonal ) . In order for individuals to make accurate decisions in this setting , collective learning must be robust to the splitting and merging of groups , and to changes in cue reliability . We first suppose that individuals employ the optimal individual voting behavior for their environment and group size and consider abrupt changes in group size and environmental conditions . The resulting accuracy experienced in the new environment is compared to the accuracy that would result from using the appropriate optimal voting behavior for the new environment . Across changes in group size ( supplemental figure S7a–c ) and environmental conditions ( supplemental figure S8a–c ) , we find that subsequent to most changes in group size and environmental conditions , the collective accuracy remains close to optimal , even when learning has not occurred in the new context . This is because it is optimal to rely on the low correlation cue for most environments and group sizes ( figure 3e–h ) , so that changes within that regime do not result in substantial decreases in accuracy . We find that individuals are far from optimal only when there are large changes in group size ( particularly when many small groups are combined into a very large group , or vice-versa ) or when the reliability of environmental cues changes drastically . We selected several particularly challenging environmental transitions and subjected our collective learning model to these conditions . Individuals across all contexts are able to adaptively adjust their voting behavior subsequent to a change in group size or environment and reach an accuracy that is close to the maximum possible for the new context ( supplemental figure S7d–f , S8d–f ) . Learning in one environment does not preclude learning in a new environment , nor does the collective context impede adapting to changing environments . Thus , fission-fusion dynamics do not necessarily limit the ability of animals to locate the effective voting behavior across a wide range of group sizes .
To date , studies of associative learning have largely been informed by experiments on individuals in isolation . Under such circumstances , there is direct feedback between preference and experience that often allows individuals to accurately learn the value of cues in the environment . However , many organisms spend at least part of their lives in groups and in order to maintain the benefits of group living , often must make consensus decisions . Coming to a consensus decouples the direct relationship between individual preferences and the outcomes of decisions , and it is not clear how animals could learn an accurate valuation of environmental cues . Here we demonstrate that embedding simple associative learning in a social context fundamentally alters what individuals learn about their environment and spontaneously allows organisms to achieve close to provably optimal collective decision-making , regardless of environmental conditions . This is in contrast to individuals who learn in isolation and subsequently pool information as a group , which can result in relatively poor collective decision-making when cues have varying degrees of observational correlation . We show that the individual behavior that maximizes collective decision accuracy is a function of both group size and the properties of environmental cues ( notably their reliability and the observational correlation between individuals ) . However , when learning collectively , individuals are able to accurately value environmental cues without explicitly estimating any of these parameters . Thus , sophisticated cognitive processes are not necessary for highly effective decision-making in a wide range of environments . While our results are robust to relaxing several of the model assumptions ( see supplemental text S2 , S3 and figures S3 , S4 , S7 , S8 ) , our model framework can also be applied to other classes of collective decision-making mechanisms [82]–[90] . For example , it is plausible that learned knowledge of the environment ( encoded by the associative strengths ) may translate into influence in the group decision , whereby individuals with stronger opinions about which option is superior may have greater influence [91]–[92] . Furthermore , many groups are composed of dominance hierarchies with a small subset of individuals controlling the group decision [24] , individuals in groups may have intrinsically different leadership abilities due to behavioral syndromes [93] , and individuals may have different physiological needs [94] . These may all contribute to differential influence in the group decision and consequently alter what is learned by group members . These modifications , which may more accurately model particular animal species , are interesting avenues of future research given their potential effect on collective learning in animal groups . In our model we assumed that an individual's learning rule is similar to that found in animals learning in isolation . This assumption precluded an individual from directly detecting the observation correlation of cues or the size of the group , parameters that we showed to be important in the determination of the optimal behavior . Nonetheless , even if individuals were not afforded additional cognitive or communication abilities , they were able to learn near optimal behavior . However , it is possible that the learning rule is indeed different for animals in a collective context . Our work suggests the need for empirical work that studies how associative learning functions within animal groups . We have considered a simple , and potentially ubiquitous , form of collective learning , in which individuals' experiences of the environment is biased by the experiences of others . The same learning rules that are known to lead to effective decision-making in single individuals are shown to be equally effective in groups of any size . This affords social organisms a robust and simple mechanism for learning behaviors that lead to accurate decisions in relatively complex environments containing multiple cues that vary in reliability and observational correlation , and which may fluctuate in time . Therefore , collective learning may allow even simple group-living organisms to reliably achieve collective wisdom across diverse environmental and social contexts .
|
Learning is ubiquitous among animal species , allowing individuals to adjust their behavior in response to their environment to improve their chances of survival and reproduction . However , while many animals live and make decisions within social groups , it is not well understood how associative learning functions within a social context . We describe an empirically derived model of collective learning and compare the learned performance of animals within groups to the optimal behavior for a wide range of environmental conditions and group sizes . We find that the learning rules derived from experiments with individual animals readily generalize to a social context , and these relatively simple rules result in behavior that is close to optimal , even when individuals know neither the size of their group nor the properties of environmental cues . Individuals that learn in isolation and subsequently join together as a group make substantially worse decisions . These results demonstrate the importance of learning within a collective context and highlight the need for experimental work to investigate the role of collective learning in enhancing decision accuracy in animal groups .
|
[
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion"
] |
[
"psychology",
"ecology",
"and",
"environmental",
"sciences",
"animal",
"behavior",
"zoology",
"ecology",
"cognition",
"behavior",
"biology",
"and",
"life",
"sciences",
"sensory",
"perception",
"social",
"sciences",
"evolutionary",
"biology",
"cognitive",
"science",
"neuroscience",
"learning",
"and",
"memory",
"animal",
"cognition",
"behavioral",
"ecology"
] |
2014
|
Collective Learning and Optimal Consensus Decisions in Social Animal Groups
|
In the fission yeast Schizosaccharomyces pombe , the transcriptional-regulatory network that governs flocculation remains poorly understood . Here , we systematically screened an array of transcription factor deletion and overexpression strains for flocculation and performed microarray expression profiling and ChIP–chip analysis to identify the flocculin target genes . We identified five transcription factors that displayed novel roles in the activation or inhibition of flocculation ( Rfl1 , Adn2 , Adn3 , Sre2 , and Yox1 ) , in addition to the previously-known Mbx2 , Cbf11 , and Cbf12 regulators . Overexpression of mbx2+ and deletion of rfl1+ resulted in strong flocculation and transcriptional upregulation of gsf2+/pfl1+ and several other putative flocculin genes ( pfl2+–pfl9+ ) . Overexpression of the pfl+ genes singly was sufficient to trigger flocculation , and enhanced flocculation was observed in several combinations of double pfl+ overexpression . Among the pfl1+ genes , only loss of gsf2+ abrogated the flocculent phenotype of all the transcription factor mutants and prevented flocculation when cells were grown in inducing medium containing glycerol and ethanol as the carbon source , thereby indicating that Gsf2 is the dominant flocculin . In contrast , the mild flocculation of adn2+ or adn3+ overexpression was likely mediated by the transcriptional activation of cell wall–remodeling genes including gas2+ , psu1+ , and SPAC4H3 . 03c . We also discovered that Mbx2 and Cbf12 displayed transcriptional autoregulation , and Rfl1 repressed gsf2+ expression in an inhibitory feed-forward loop involving mbx2+ . These results reveal that flocculation in S . pombe is regulated by a complex network of multiple transcription factors and target genes encoding flocculins and cell wall–remodeling enzymes . Moreover , comparisons between the flocculation transcriptional-regulatory networks of Saccharomyces cerevisiae and S . pombe indicate substantial rewiring of transcription factors and cis-regulatory sequences .
Flocculation is an inherent characteristic of yeasts involving asexual aggregation of cells into flocs that separate rapidly from the medium ( reviewed recently in [1] , [2] ) . Individual yeast cells transition into this morphological state as an adaptation to various environmental stresses by shielding the inner cells of the flocs [3] . The flocculent trait has also proven highly beneficial in industrial yeast applications by allowing efficient and cost-effective removal of cells [4] . The ability of yeast strains to flocculate is dependent on the expression of specific cell surface glycoproteins known as flocculins . Cell-to-cell adhesion occurs via binding between the flocculin and surface carbohydrates in a calcium-dependent manner [5] . The bound carbohydrates consist of various sugars including mannose , glucose , and galactose that are specific to the type of flocculin and yeast species [6]–[8] . There has been considerable interest in elucidating the genetic control of flocculation to better understand this phenomenon and generate biotechnological advances in yeast-based industries . In Saccharomyces cerevisiae , a transcriptional-regulatory network composed of interactions between transcription factors and their flocculin gene targets is central in controlling flocculation . The primary flocculins that function in flocculation are encoded by the FLO1 , FLO5 , FLO9 , and FLO10 genes [9]–[11] . Overexpression of the individual FLO genes is sufficient to trigger flocculation [8] , [12] . However , the degree of flocculation by FLO overexpression varies from FLO1 to FLO10 exhibiting the strongest to weakest flocculation , respectively . The flocculin FLO11 also exhibits weak flocculation when overexpressed [8] , but its function is mainly in cell-to-surface adhesion [13] , diploid pseudohyphal growth [14] , and haploid invasive growth [15] . The transcription factors required for flocculation include Flo8p and Mss11p , which primarily activate FLO1 transcription [16] . The Sacc . cerevisiae laboratory strain S288C containing a nonfunctional FLO8 gene is not able to flocculate , but flocculation is restored in this strain by the overexpression of FLO8 or MSS11 [16] , [17] . In addition , Sfl1p has been shown to inhibit transcription of FLO1 in the W303-1A strain and not in S288C , likely through interactions with the Ssn6p-Tup1p global repressor and components of Mediator [18] , [19] . The control of flocculation is much less known in Schizosaccharomyces pombe . The ability of the heterothallic wild-type strains 972 h− and 975 h+ to flocculate has not been observed presumably because the inducing environmental conditions have not been identified . Phenotypic analysis of constitutive flocculent mutant strains show that flocculation is dependent on the presence of calcium , but unlike Sacc . cerevisiae , the flocculin-carbohydrate interactions involve galactose rather than mannose and glucose residues [7] . Moreover , the transcriptional-regulatory network governing flocculation in S . pombe remains poorly characterized . Only a single interaction between the Mbx2 MADS box transcription factor and the gsf2+ flocculin gene is currently known [20] , [21] . The gsf2+ gene was initially identified as highly upregulated in response to heterologous expression of FLO8 [20] . Overexpression of gsf2+ is sufficient to trigger flocculation while its deletion abrogates the flocculent phenotype of tup12Δ , lkh1Δ , and gsf1 mutants . In addition , gsf2+ displays additional roles in cell-to-surface adhesion and invasive growth [20] . The induction of gsf2+ during flocculation and invasive growth is mediated by Mbx2 [21] . Two other transcription factors implicated in flocculation have been reported . The CSL transcription factors Cbf11 and Cbf12 play opposing roles in flocculation where mutant strains lacking cbf11+ or overexpressing cbf12+ flocculate [22] . The direct targets of these transcription factors functioning in flocculation have not been identified , but could be several putative flocculin genes that show protein sequence homology to other yeast-related proteins [23] . Indeed , these putative flocculin genes , as well as gsf2+ are transcriptionally upregulated in certain Mediator mutants that flocculate indicating that these genes are likely repressed by Mediator [24] . Similar to Sacc . cerevisiae , the global transcriptional regulators Tup11 and Tup12 function in flocculation but their influence on the expression of these flocculin genes has not been addressed [25] . Importantly , it has not been directly demonstrated that these putative flocculin genes in S . pombe actually play a role in flocculation and the identity of the transcription factors that regulate them remains unknown . In this study , we have initiated an extensive characterization of the transcriptional-regulatory network of S . pombe flocculation by identifying the relevant transcription factors and their flocculin gene targets . Importantly , we have also determined that heterothallic wild-type S . pombe is able to flocculate when grown in rich medium containing ethanol and glycerol as a carbon source . A screen of transcription factor deletion and overexpression strains for flocculent phenotypes revealed five novel transcriptional regulators of flocculation ( Rfl1 , Adn2 , Adn3 , Sre2 , Yox1 ) in addition to our independent finding of Mbx2 , Cbf11 , and Cbf12 . The strongest flocculation was observed upon overexpression of mbx2+ and deletion of rfl1+ ( SPBC15D4 . 02 ) which encodes an uncharacterized fungal Zn ( 2 ) -Cys ( 6 ) transcription factor . Microarray expression profiling of the mbx2OE and rfl1Δ strains revealed good overlap in the upregulation of several flocculin genes , while ChIP-chip analysis of HA-tagged Mbx2 and Rfl1 under control of the nmt41 promoter indicated that these transcription factors bound to some of the flocculin gene promoters . Nine flocculin gene targets ( pfl1+–pfl9+ ) including gsf2+/pfl1+ were identified . The single overexpression of these genes triggered flocculation to varying degrees and cumulative effects on flocculation were observed in double overexpression experiments . Only loss of gsf2+ could abrogate the flocculent phenotype of all the transcription factor mutants indicating that gsf2+ encodes the dominant flocculin in S . pombe . Interestingly , we discovered that certain cell wall-remodeling enzymes can also function in flocculation , and some of these genes are likely regulated by the LisH transcription factors Adn2 and Adn3 . In addition to the identification of target genes within the transcriptional-regulatory network , autoregulatory and inhibitory feed-forward loops involving several transcription factors were also detected . These results provide a significant insight into the transcriptional control of flocculation in S . pombe .
Our understanding of the transcriptional-regulatory network that governs flocculation in S . pombe remains limited . To further decipher this network , we sought to systematically identify transcription factors that play a role in flocculation . A list of 101 genes encoding sequence–specific transcription factors containing a bona-fide DNA-binding domain was assembled from [26] and GeneDB [27] . From this gene list , we constructed 101 nmt1-driven overexpression strains and 92 nonessential deletions in which the entire ORF was replaced with the KanMX6/NatMX6 cassette . A detailed description of the construction and phenotypic characterization of this transcription factor mutant collection will be described elsewhere ( unpublished data ) . The transcription factor array of overexpression and deletion strains were screened for flocculation in EMM lacking thiamine and YES media , respectively . We recovered a total of eight transcription factors in which four overexpression strains ( mbx2OE , adn2OE , adn3OE and cbf12OE ) and four deletions ( rfl1Δ , sre2Δ , yox1Δ and cbf11Δ ) exhibited flocculation . These transcription factors represent positive and negative regulators of flocculation , respectively . Among these transcription factors , only the overexpression of cbf12+ and mbx2+ and deletion of cbf11+ have been reported to cause flocculation [20] , [22] . The strongest flocculation was observed in the mbx2OE and rfl1Δ strains . The flocs of the rfl1Δ strain in YES medium were larger and sedimented faster than the flocs produced in the mbx2OE strain after 48 hour induction ( Figure 1A ) . The mbx2+ gene encodes a MADS-box transcription factor which was originally isolated in a screen for genes functioning in the biosynthesis of cell surface pyruvated galactose residues [28] . Recently , Mbx2 has been shown to function in flocculation and invasive growth by regulating the flocculin gene gsf2+ [20] , [21] . The rfl1+ ( repressor of flocculation ) gene encodes an uncharacterized fungal Zn ( 2 ) -Cys ( 6 ) transcription factor . The flocculation exhibited by these overexpression and deletion transcription factor mutants recovered from our screens could be abolished with the addition of galactose , but not mannose or glucose ( data not shown ) . The amount of galactose required to completely deflocculate cells depended on the degree of flocculation . For example , mbx2OE strain could be deflocculated with 2% galactose while rfl1Δ strain required 5–10 times more galactose to completely deflocculate . Reflocculation of these strains was achieved in CaCl2 or in YES medium ( data not shown ) . The growth conditions that trigger flocculation in heterothallic wild-type S . pombe are not well known . To identify the inducing conditions , 972 h− and 975 h+ cells were tested on different carbon sources at different cell densities for flocculation . We determined that heterothallic wild-type cells were able to flocculate when cultured for five days at an initial concentration of 1×106 cells/ml in medium containing 1% yeast extract , 3% glycerol and , 4% ethanol ( referred to as flocculation-inducing medium , Figure 1B ) . The degree of flocculation was slightly enhanced in strains auxotrophic for leucine , uracil , and/or adenine indicating that nutrient status may also play a role in triggering flocculation ( data not shown ) . However , these wild-type strains flocculated significantly less in flocculation-inducing medium than the mbx2OE and rfl1Δ mutants in EMM and YES media , respectively . The weaker flocculation in these strains was more easily observed in petri-dishes incubated on an orbital rotator than in test tubes . In contrast to wild type , deletion of mbx2+ did not produce any visible flocs in the flocculation-inducing medium ( Figure 1B ) . Fungal genes that function in flocculation are usually associated with filamentous invasive growth [17] , [20] . We hypothesized that the rfl1Δ strain would exhibit hyperfilamentous invasive growth because of its strong flocculent phenotype . Indeed , the amount of cells resistant to removal from the agar by washing in the invasive assay on LNB medium with an underlayer of YE+ALU was much greater in the rfl1Δ strain than in wild type ( Figure 1C ) . Under the microscope , the filamentous growth like those detected by Dodgson et al . [29] was observed below the agar surface for both wild type and rfl1Δ strain with the latter showing much larger and more frequent formation of filamentous growth ( data not shown ) . Similarly , adn2+ and adn3+ which were previously observed to have defects in invasive growth when deleted were recovered in our screens as flocculent when overexpressed [29] . The strongest flocculation observed in the mbx2OE and rfl1Δ strains indicated that these two genes encode the major regulators of flocculation . Therefore , we initially focused on the characterization of these two transcription factors and proceeded to identify their target genes involved in flocculation . The nmt41-driven mbx2-HA strain was subjected to microarray expression profiling with a custom-designed S . pombe 8×15 K Agilent expression microarray ( Table S2 ) . The intermediate strength nmt41 promoter was sufficient for mbx2OE flocculation and was utilized in the microarray experiments in order to reduce possible secondary transcriptional effects compared to the strong nmt1 promoter . To better distinguish the direct target genes , ChIP-chip was also carried out concurrently on the same strain using the S . pombe 4×44 K Agilent Genome ChIP-on-chip microarray ( Table S3 ) . For the rfl1+ expression profiling and ChIP-chip experiments , the flocculent deletion mutant and nmt41-driven rfl1-HA strain were used , respectively ( Tables S4 and S5 ) . The highly-induced putative target genes identified by microarray expression profiling of these transcription factor mutant strains were validated by qPCR ( Table S13 ) . The list of genes that were induced at least two fold in the mbx2OE or rfl1Δ strain was subjected to gene ontology analysis using the Princeton GO Term Finder ( http://go . princeton . edu/cgi-bin/GOTermFinder ) . These induced genes were highly enriched in cell wall components with p-values of 9 . 0e-9 and 6 . 3e-6 for the mbx2OE and rfl1Δ strains , respectively . Strikingly , the most-induced genes in the mbx2OE strain encoded cell surface glycoproteins . The cell surface glycoprotein genes up-regulated above two-fold were SPAC186 . 01 , gsf2+ , SPAC977 . 07c/SPBC1348 . 08c , SPCC188 . 09c , fta5+ , SPBC947 . 04 , SPBC359 . 04c , SPBC1289 . 15 , SPAPB2C8 . 01 , SPAC1F8 . 02c , SPAPB18E9 . 04c , SPCC553 . 10 , and SPBPJ4664 . 02 , which all but gsf2+ and the last 4 genes were predicted to be pombe adhesins based on BLAST sequence analysis ( Figure 2A; [23] ) . SPAC977 . 07c and SPBC1348 . 08c are gene duplications with 100% sequence identity . To our knowledge , these genes with the exception of gsf2+ have not been characterized further . The induction of these genes in the mbx2OE strain ranged from 2 to 112-fold relative to the empty vector control ( Figure 2A , Table S13 ) . In addition , several genes ( agn2+ , psu1+ , SPAC4H3 . 03c and gas2+ ) encoding cell wall-remodeling enzymes such as glucan glucosidases and a betaglucanosyltransferase were induced up to 91-fold compared to the empty vector control when mbx2+ was overexpressed ( Figure 2A ) . In the rfl1Δ expression data , a similar set of cell surface glycoprotein genes were upregulated at a comparable level as the mbx2OE expression data except for SPAC1F8 . 02 , SPBC359 . 04c , SPAPB18E9 . 04c and SPBPJ4664 . 02 ( Figure 2A , Table S13 ) . In contrast to the mbx2OE strain , the same genes encoding the cell wall-remodeling enzymes were not highly upregulated in the rfl1Δ strain ( Figure 2A ) . Of the thirteen highly-induced cell surface glycoprotein genes in the mbx2OE expression data , nine of them were detected with ChIP-chip indicating that these genes are very likely the direct transcriptional targets of Mbx2 ( Figure 2A ) . Four of the nine highly-induced cell surface glycoprotein genes in the rfl1Δ strain were detected with ChIP-chip confirming that these genes are probably direct transcriptional targets of Rfl1 ( Figure 2A ) . For both Mbx2 and Rfl1 , gsf2+ , fta5+ and SPAPB2C8 . 01 were detected in the expression microarray and ChIP-chip experiments ( Figure 2A ) . Next , we sought further evidence that these cell surface glycoprotein genes were targets of Mbx2 and Rfl1 by epistasis studies . We decided to study a subset of these genes , which included the majority of the gene sequences analyzed by Linder and Gustafsson [23] , [24] . The mbx2+ gene was overexpressed in single deletions of these putative target genes and their degree of flocculation was determined visually in petri-dishes , as well as quantitatively ( Table S14 ) . The putative glycoprotein gene SPAPB15E9 . 01c was included in these studies , because even though the transcript was downregulated in both mbx2OE and rfl1Δ strains , ChIP-chip analysis detected Mbx2 and Rfl1 association with its promoter ( Figure 2A ) . Deletion of gsf2+ decreased mbx2OE flocculation to the greatest extent while the reduction of flocculation was less extensive in the other single deletion mutants ( Figure 2B , Table S14 ) . The degree of reduction in mbx2OE flocculation roughly corresponded to the pfl numbers , which were assigned based on the degree of flocculation when overexpressed ( see below ) . Moreover , mbx2OE flocculation was completely abrogated in the gsf2Δ pfl9Δ double mutant indicating that the reduction of mbx2OE flocculation in these mutants were additive in some cases ( Figure 2B ) . Similar experiments were performed for rfl1+ in which flocculation was assayed in the same putative target deletions in the rfl1Δ background . The flocculation exhibited in the rfl1Δ strain was completely abolished by the deletion of gsf2+ , but not by the deletion of pfl9+ ( Figure 2C ) . To further analyze the expression microarray datasets of Mbx2 and Rfl1 , the promoter regions of the differentially-expressed genes were subjected to the motif-finding algorithms RankMotif++ and MEME to identify their binding specificities [30] , [31] . Mbx2 is a member of the MEF2-MADS box transcription factor family which has been shown to bind to the consensus sequence 5′- ( C/T ) TA ( T/A ) 4TA ( G/A ) -3′ [28] , [32] , [33] . The Mbx2 binding specificity obtained by RankMotif++ closely resembled this known consensus sequence ( Figure 2D ) . Similarly , RankMotif++ generated an Rfl1 binding specificity that resembled known consensus sequences of several members of the fungal Zn ( 2 ) -Cys ( 6 ) transcription factor family ( Figure 2E ) . The binding specificity of Zn ( 2 ) -Cys ( 6 ) DNA-binding domains is composed of conserved GC-rich trinucleotides spaced by a variable sequence region differing in length among members of the transcription factor family [34] . Analyses of the Mbx2 and Rfl1 expression microarray and ChIP-chip datasets by MEME did not generate any candidate DNA motifs . Altogether , these results demonstrate that Mbx2 and Rfl1 are transcription factors responsible for regulation of flocculation in fission yeast by activating or repressing the transcription of candidate S . pombe flocculin genes , respectively . Besides gsf2+ , the other putative target genes of Mbx2 and Rfl1 that encode for cell surface glycoproteins share some amino acid sequence homology with domains found in other fungal adhesins [23] . However , the role of these glycoprotein genes in flocculation has not been demonstrated . Overexpression studies were employed to the aforementioned set of putative flocculin target genes of Mbx2 and Rfl1 to determine whether they function directly in flocculation . Each single overexpression of these flocculin genes was able to induce flocculation to varying degrees with the strongest flocculation observed in the gsf2OE strain which produced visible flocs within one day ( Figure 3A; Table S14 ) . Weaker flocculation was observed from the overexpression of the other flocculin genes after total incubation of 2–7 days in EMM minus thiamine medium with sub-culturing into fresh medium in Day 3 . The flocculation images of these overexpression strains shown in Figure 3A were captured after total of 7 days of induction . As a result of these observations , we named these genes pfl+ for Pombe Flocculins and numbered them according to their degree of flocculation when overexpressed: pfl1+/gsf2+ ( referred as gsf2+ hereafter ) , pfl2+/SPAPB15E9 . 01c , pfl3+/SPBC947 . 04 , pfl4+/SPCC188 . 09c , pfl5+/SPBC1289 . 15 , pfl6+/SPAC977 . 07c , pfl7+/SPBC359 . 04c , pfl8+/fta5+ ( referred as fta5+ hereafter ) and pfl9+/SPAC186 . 01 . Furthermore , we overexpressed some double combinations of the weaker flocculin genes to determine whether flocculation could be additive . Indeed , the pfl4+ pfl9+ , pfl6+ pfl9+ , and fta5+ pfl9+ double overexpression strains flocculated earlier and formed larger flocs than their corresponding single overexpressors , thus , demonstrating the additive effect of these flocculins ( Figure 3B , Table S14 ) . We next tested the single deletions of the pfl+ genes for their ability to flocculate in flocculation-inducing medium . No visible flocculation was observed in the gsf2Δ strain while wild type was flocculent ( Figure 1B ) . In contrast , flocculation still occurred in the pfl2Δ–pfl9Δ strains in the inducing medium indicating that gsf2+ encodes the dominant flocculin and the other flocculin genes are dispensable for flocculation ( data not shown ) . These observations revealed that the contribution in flocculation by these pfl+ genes varied and certain combinations of pfl+ were additive . The strength of flocculation by the single overexpression of pfl+ genes was directly correlated with the reduction of mbx2OE flocculation in the corresponding deletion strains ( Figure 2B and Figure 3A , Table S14 ) . For example , the pfl2OE strain which produced larger flocs than the pfl3OE–pfl9OE strains exhibited a greater inhibition of mbx2OE flocculation when deleted . Similarly , the flocculation of the rfl1Δ strain was completely abrogated by the deletion of gsf2+ , but not at all by the deletion of pfl9+ ( Figure 2C ) . Consistent with the above results , the deletion of both gsf2+ and pfl9+ led to a greater abrogation of mbx2OE flocculation compared to each deletion alone ( Figure 2B ) . In summary , we have demonstrated that these pfl+ genes encode for S . pombe flocculins and Gsf2 is the dominant flocculin . Interestingly , ChIP-chip analysis also detected binding of Mbx2 and Rfl1 to their own promoters , as well as Rfl1 binding to the mbx2+ promoter ( Figure 2A ) , indicating autoregulation and mbx2+ regulation by Rfl1 within the transcriptional-regulatory network of S . pombe flocculation . Mbx2 also appeared to be associated with the rfl1+ promoter , but this interaction was marginal as it was found just above the detection threshold for ChIP-chip ( Figure 2A ) . To investigate the autoregulation of mbx2+ , the gene was C-terminal tagged with GFP at its native locus ( mbx2-GFP ) . However , the GFP-tagged strain resulted in a hypermorphic allele that displayed constitutive flocculation and nuclear localization of Mbx2-GFP ( see below ) . We speculated that the removal of the 3′-untranslated region of mbx2+ during the C-terminal tagging may be the cause of the hypermorphic allele . To bypass this potential problem , we created an N-terminal GFP-tagged allele ( GFP-mbx2 ) with an intact 5′-untranslated region and approximately 1 kb of native promoter sequence . In contrast to the C-terminal tagged hypermorphic allele , the N-terminal tagged GFP-Mbx2 expression was comparable to background levels and the strain did not exhibit constitutive flocculation ( Figure 4A ) . Moreover , the GFP-mbx2 strain flocculated when grown in glycerol-inducing medium indicating that the tagged protein is functional ( Table S14 ) . When nmt1-driven mbx2+ expression was induced for 9 hours in the GFP-mbx2 strain , nuclear GFP-Mbx2 expression was detected , indicating that Mbx2 can activate its own expression ( Figure 4A ) . As expected , this strain was now flocculent . Longer induction of nmt1-driven mbx2+ expression resulted in greater GFP-Mbx2 expression with multi-nucleated GFP foci ( data not shown ) . The positive autoregulation of mbx2+ is likely to be direct as several putative MEF2-binding sequences ( e . g . 5′-TTAAAAATAG-3′ ) are located within 1000 bp upstream from the mbx2+ start codon ( data not shown ) . To determine whether negative autoregulation occurs with rfl1+ , a C-terminal GFP-tagged strain under native control was generated ( rfl1-GFP ) . The localization of Rfl1-GFP was nuclear in the rfl1-GFP strain ( Figure 4B ) . The induction of nmt1-driven rfl1+ expression for 18 hours in the rfl1-GFP strain led to a reduced nuclear Rfl1-GFP signal and a slightly increased cytoplasmic Rfl1-GFP signal ( Figure 4B ) . However , overall Rfl1-GFP expression in the cell was reduced when Rfl1 was overexpressed compared to the empty vector control ( Figure 4B; two-tailed t-test; p value<0 . 01 ) . In contrast to our observations with the Rfl1-GFP protein expression , we found that there was no decrease of the Rfl1-GFP transcript when rfl1+ was overexpressed ( Table S13 ) . These results indicate that although Rfl1 can bind to its own promoter , negative autoregulation appears marginal or may not be occurring . The observation that Rfl1 is associated with the mbx2+ promoter by ChIP-chip suggests that Rfl1 may oppose Mbx2 function in flocculation by repressing its expression . To test this hypothesis , we first examined the genetic interactions between mbx2+ and rfl1+ . The mbx2Δ rfl1Δ double mutant did not display flocculation indicating that mbx2+ is epistatic to rfl1+ ( Figure 5A ) . In addition , the flocculation associated with mbx2OE was abrogated by co-overexpression of rfl1+ ( Figure 5A ) . These results are consistent with mbx2+ being downstream of rfl1+ and that rfl1+ opposes mbx2+ function in flocculation . We next utilized the C-terminal and N-terminal GFP-tagged mbx2+ strains to further determine if Rfl1 represses mbx2+ expression . First , Rfl1 was overexpressed in the hypermorphic C-terminal tagged mbx2-GFP allele which shows constitutive nuclear Mbx2-GFP expression and flocculation . This resulted in the near-abolishment of both the GFP signal ( Figure 5B ) and flocculation ( data not shown ) in the hypermorphic mbx2 allele . Second , when the N-terminal tagged GFP-mbx2 strain was crossed into the rfl1Δ background , the resulting strain displayed dramatic increase in nuclear GFP-Mbx2 expression ( Figure 5C ) and flocculation strength equivalent to the rfl1Δ strain ( data not shown ) . These results support the hypothesis that mbx2+ expression is repressed by Rfl1 in non-flocculent cells . Cbf12 , a member of the CSL transcription factor family has previously been reported to trigger flocculation when overexpressed [22] . However , the target genes of Cbf12 that function in flocculation have not been identified . To further elucidate the role of cbf12+ in flocculation , we took a similar approach to identify its direct target genes by concurrent expression microarray profiling and ChIP-chip analysis of the nmt41-driven cbf12-HA strain ( Tables S6 and S7 , respectively ) . When cbf12+ was deleted and cultured in flocculation-inducing medium , flocculation was abolished ( Figure 6A ) . In contrast , overexpression of cbf12+ by the nmt1 promoter triggered flocculation ( Figure 6C ) and produced a bowling pin–shaped phenotype after 24 hours in medium lacking thiamine ( data not shown ) . Further induction of the nmt1-driven cbf12+ caused the strain to become sick and granulated , eventually leading to growth arrest ( data not shown ) . To reduce the toxic effects of cbf12+ overexpression , an nmt41-driven cbf12-HA strain was used for concurrent expression profiling and ChIP-chip analysis . Gene ontology analysis was carried out separately on the top 50 most highly-induced genes and all 160 promoter-occupied genes by Cbf12 with the Princeton GO Term Finder . Functional enrichment of genes in cell surface ( p = 1 . 8e-7 ) and plasma membrane ( p = 5 . 7e-4 ) was detected for the highly-induced and promoter-occupied genes , respectively . These genes included several flocculin genes , ( Figure 6B ) . Both gsf2+ and pfl7+ were among the five highest induced genes ( 18 . 1 and 27 . 6-fold , respectively ) in the cbf12OE strain and were also detected by ChIP-chip ( Figure 6B ) suggesting that Cbf12 directly activates the transcription of gsf2+ and pfl7+ for flocculation . The flocculation triggered by cbf12+ overexpression was completely abrogated in the gsf2Δ background , whereas deletion of pfl7+ had little effect ( Figure 6C , Table S14 ) . This was consistent with the hypothesis that gsf2+ encodes the dominant flocculin . In addition , loss of gsf2+ or pfl7+ did not alter the bowling-pin cell shape or the reduced fitness phenotypes of the cbf12OE strain indicating that these two phenotypes were not due to the upregulation of the flocculin genes ( data not shown ) . The much weaker flocculation observed in the cbf12OE strain in comparison to the mbx2OE and gsf2OE strains may be attributed to additional defects in cell and nuclear division , which would cause early growth arrest before the full flocculation potential could be reached [22] . Consistent with previous findings , C-terminal GFP-tagged Cbf12 under native control was expressed predominantly in the nucleus in stationary phase cells while expression in logarithmic cells was comparable to background ( Figure 6D; [22] ) . Compared to logarithmic growth in rich medium , Cbf12-GFP nuclear expression increased in cells grown in flocculation-inducing medium , thus supporting its role in flocculation ( Figure 6D ) . Interestingly , Cbf12 was also detected by ChIP-chip to bind to its own promoter ( Figure 6B ) . Indeed , positive autoregulation appears to occur as native Cbf12-GFP expression increased greater than three-fold when nmt1-driven cbf12+ was ectopically expressed in logarithmically growing cells ( Figure 6E ) . Recently , it was demonstrated that an N-terminal-truncated Cbf12 bound to probes containing a canonical CSL binding motif ( 5′-GTGGGAA-3′ ) by gel mobility shift assay [35] . We next searched for a similar DNA binding sequence for Cbf12 from the expression microarray and ChIP-chip cbf12OE datasets by RankMotif++ and MEME . RankMotif++ and MEME analyses of the expression microarray and ChIP-chip data , respectively , did not identify a binding specificity for Cbf12 . However , when the promoters of up-regulated genes in the cbf12OE strain belonging to the cell surface GO category were subjected to MEME analysis , a motif closely matching the canonical CSL binding motif ( 6/7 nucleotide match ) was recovered ( Figure 6F ) . These results demonstrate Cbf12 as part of the transcriptional-regulatory network of fission yeast flocculation by controlling the transcription of several flocculin genes including gsf2+ . From our transcription factor screens , the deletion of yox1+ , sre2+ , or cbf11+ also resulted in flocculation , although the size of the flocs were smaller than observed in mbx2OE , cbf12OE and rfl1Δ strains ( Figure 7A , Table S14 ) . Yox1 has been implicated in a negative autoregulatory loop to prevent inappropriate transcriptional expression of MBF gene targets , while the function of Sre2 , which shows homology to the human sterol regulatory element binding protein SREBP-1A remains largely unknown [36] , [37] . A role of Yox1 and Sre2 in flocculation has not been reported . In contrast , cbf11+ encodes a CSL transcription factor that plays a role in flocculation , but its target genes are not known [22] . To elucidate the transcriptional flocculation program of yox1+ , sre2+ and cbf11+ , expression microarray profiling was conducted on the corresponding flocculent deletion strains in rich medium ( Tables S8 , S9 , S10 ) . The expression microarray profiles of yox1Δ and sre2Δ most resembled each other compared to the other strains described in this study ( Figure 7B ) . Genes upregulated by at least two-fold in the yox1Δ and sre2Δ strains showed enrichment for ribosomal subunits ( p = 2 . 8e-31 and 7 . 4e-25 for yox1Δ and sre2Δ , respectively ) and mitochondrial membrane transporters ( p = 7 . 5e-5 and 1 . 2e-3 for yox1Δ and sre2Δ , respectively ) . These findings did not intuitively answer our questions as to how these two transcription factors might be related or associated with the flocculation pathway . We next examined whether any of the flocculin genes and their putative regulators were induced in the yox1Δ and sre2Δ strains . In the sre2Δ strain , gsf2+ , pfl3+ and fta5+ transcripts were upregulated 3 . 7 , 2 . 5 and 3 . 1-fold , respectively , indicating that the expression of these genes could be contributing to the flocculent phenotype ( Figure 7C ) . In contrast , mbx2+ and cbf12+ transcripts were downregulated approximately 2-fold suggesting that the elevated levels of gsf2+ , pfl3+ and fta5+ transcripts in the sre2Δ strain were not mediated by Mbx2 and Cbf12 ( Figure 7C ) . Similarly in the yox1Δ strain , we observed that gsf2+ and pfl3+ transcripts were upregulated although less than in the sre2Δ strain , and mbx2+ and cbf12+ were also downregulated ( Figure 7C ) . Therefore , this suggests that sre2+ and yox1+ may be involved in the repression of flocculation through a pathway independent from mbx2+ and cbf12+ . The microarray expression profile of the cbf11Δ strain revealed greater than 2-fold increase of gsf2+ and pfl3+ transcripts and a 60-fold increase of the SPAC1F8 . 02c transcript suggesting that these two flocculin genes and this uncharacterized glycoprotein gene may be responsible for the flocculent phenotype in this mutant . In contrast to the yox1Δ and sre2Δ mutants , mbx2+ did not show differential expression in the cbf11Δ strain compared to wild type . However , the cbf12+ transcript was upregulated 1 . 8-fold in the cbf11Δ strain . This suggests that cbf11+ may regulate flocculation through cbf12+ , in agreement with previous reports of the antagonistic functions of cbf11+ and cbf12+ in this process [22] . We next determined whether the flocculation caused by the deletion of yox1+ , sre2+ or cbf11+ was also dependent on gsf2+ . The absence of gsf2+ was sufficient to abolish the flocculation in yox1Δ , sre2Δ , and cbf11Δ strains , even though gsf2+ was not always the most highly-expressed flocculin gene ( Figure 7A and 7C , Table S14 ) . Taken together , these results suggest that the expression of the dominant flocculin Gsf2 is responsible for the bulk of flocculation observed in yox1Δ , sre2Δ and cbf11Δ strains . The transcription factor genes adn2+ and adn3+ are orthologous to Sacc . cerevisiae FLO8 ( http://www . pombase . org/ ) and exhibit defects in invasive growth and cell-to-surface adhesion when deleted during nitrogen starvation [29] . From our screens , we discovered that the overexpression of adn2+ and adn3+ triggered minor flocculation while loss of adn2+ and adn3+ prevented flocculation in flocculation-inducing medium ( Figure 8A and 8B , respectively ) . The flocculent phenotype of adn2OE and adn3OE strains was disrupted by the addition of galactose ( data not shown ) . To identify the target genes of Adn2 and Adn3 that are involved in flocculation , expression microarray profiling was performed on nmt1-driven adn2OE and adn3OE strains ( Tables S11 and S12 ) . Surprisingly , gsf2+ transcript levels were relatively unchanged and the majority of pfl+ genes were downregulated in both overexpression strains ( Figure 8C ) . Consistent with these results were the observations that mbx2+ and cbf12+ transcripts were downregulated greater than 2-fold in both adn2OE and adn3OE strains , whereas rfl1+ transcript levels were not differentially regulated ( Figure 8C ) . Therefore , it appeared that the flocculent phenotype of adn2+ and adn3+ overexpression could not be attributed to the pfl+ genes identified in this study . These results led us to consider that perhaps the expression of other genes besides these encoding for flocculins could be responsible for triggering flocculation in adn2OE and adn3OE strains . Interestingly , some of the aforementioned cell wall-remodeling enzymes ( gas2+ , psu1+ and SPAC4H3 . 03c ) were also highly upregulated in both adn2OE and adn3OE strains ( Figure 8C , Table S13 ) . For example , gas2+ and SPAC4H3 . 03c were the highest induced genes in the adn2OE strain ( 17 . 9 and 36 . 8-fold , respectively ) and also appeared within the top 20 most induced genes in the adn3OE strain . These genes were also induced in the mbx2OE strain except for psu1+ ( Figure 2A ) . Overexpression analysis was subsequently carried out to determine if these genes possessed some role in flocculation . Although agn2+ was not upregulated in the adn2OE and adn3OE strains , it was included in the overexpression analysis because it was the second most induced gene ( 91-fold ) , as well as detected by ChIP-chip in the mbx2OE strain . Indeed , the single overexpression of these four genes resulted in flocculation after 5-days ( including 3rd day sub-culturing into fresh medium ) in medium lacking thiamine , implicating the involvement of these cell wall-remodeling enzymes in flocculation ( Figure 8D , Table S14 ) . Since deletion of adn2+ and adn3+ results in defects of invasive growth and cell-to-surface adhesion in response to nitrogen starvation , we wanted to determine if the single overexpression of gas2+ , agn2+ , psu1+ and SPAC4H3 . 03c could cause enhancement of these processes . We discovered that the single overexpression of these four cell wall-remodeling genes increased cell-to-surface adhesion , but not invasive growth relative to wild type under the nitrogen-deprivation condition ( Figure S1 ) . Because gsf2+ encodes the dominant flocculin , we also investigated whether the flocculation caused by adn2+ and adn3+ overexpression was dependent on gsf2+ . Deletion of gsf2+ completely abrogated the flocculation in adn2OE and adn3OE strains ( Figure 8A , Table S14 ) . In addition , the adn2OE and adn3OE strains exhibited cell separation defects such as the formation of multisepta and forkhead phenotypes ( Figure 8E ) . The cell separation defect was more severe when adn3+ was overexpressed . We next determined whether the putative target genes involved in the flocculation of adn2OE and adn3OE strains also played a role in the multisepta phenotype . Overexpression of adn2+ and adn3+ in the gsf2Δ background did not alter the multisepta phenotype ( Figure 8E ) , while the overexpression of gas2+ , SPAC4H3 . 03c , psu1+ and agn2+ did not lead to formation of multisepta ( data not shown ) . These results suggest that Adn2 and Adn3 may regulate cell separation and flocculation independently through different sets of target genes . Our microarray expression data suggests that Adn2 and Adn3 may control cell separation through ace2+ , which encodes a major transcriptional activator of this process ( Alonso-Nuñez et al . , 2005 ) . Overexpression of adn2+ and adn3+ resulted in the down-regulation of ace2+ and many of its known target genes such as adg1+ , adg2+ , adg3+ , cfh4+ , agn1+ , eng1+ , and mid2+ by 1 . 5 to 3 . 4-fold ( Figure 8C ) . In summary , the regulation of flocculation by adn2+ and adn3+ is likely mediated by the induction of genes encoding the cell wall-remodeling enzymes Gas2 , SPAC3H3 . 03c , and Psu1 . The regulation of these genes is independent from Mbx2 because mbx2+ was downregulated in the adn2OE and adn3OE strains . Although gsf2+ transcript level was not significantly upregulated by adn2+ and adn3+ overexpression , it was sufficient to abrogate the flocculation when deleted . However , it is possible that other cell surface glycoprotein genes not investigated in this study but were upregulated may also play a significant role in the flocculation function of adn2+ and adn3+ .
In this study , we have deciphered a significant portion of the transcriptional-regulatory network governing flocculation in S . pombe . To date , few transcription factors and their target genes that function in flocculation have been identified . The MADS box transcription factor Mbx2 positively regulates flocculation by induction of the flocculin gene gsf2+ , while the CSL transcription factors Cbf11 and Cbf12 repress and activate flocculation , respectively , but their target genes are not known [21] , [22] . We have substantially expanded our limited knowledge of the flocculation transcriptional-regulatory network by the identification of several novel transcriptional activators ( Adn2 and Adn3 ) and repressors ( Rfl1 , Yox1 and Sre1 ) , and their putative target genes that function in flocculation . In addition , novel target genes of Mbx2 , Cbf11 and Cbf12 were identified . The putative target genes of the transcription factors implicated in flocculation encode for several cell surface glycoproteins ( gsf2+ and pfl2+–pfl9+ ) and cell wall-remodeling enzymes ( agn2+ , psu1+ , SPAC4H3 . 03c and gas2+ ) . These target genes were sufficient to trigger flocculation when overexpressed . Moreover , instances of regulation between transcription factors ( Rfl1 repression of mbx2+ ) , as well as positive ( mbx2+ and cbf12+ ) autoregulation were detected within the flocculation network . Mbx2 and Rfl1 appeared to be the major positive and negative regulators of flocculation , respectively , based on the largest flocs observed in the mbx2OE and rfl1Δ strains compared to the other flocculent mutants in this study . Our initial efforts to identify the target genes of Mbx2 and Rfl1 revealed several putative flocculin genes that were strikingly upregulated in the mbx2OE and rfl1Δ flocculent mutants . Previously , Gsf2 was the only S . pombe flocculin demonstrated to be directly involved in flocculation , and its transcription was influenced by the activity of Mbx2 [20] , [21] . Similar to these studies , we also found that overexpression of gsf2+ triggers flocculation while loss of gsf2+ abrogates the flocculent phenotype of several mutants including mbx2OE . Here , we identified an additional eight flocculin genes ( pfl2+–pfl9+ ) as putative target genes of Mbx2 . Seven of these target genes ( pfl3+–pfl9+ ) were reported to contain tandem repeats found in fungal adhesins , while pfl2+ is a sequence orphan predicted to encode a GPI-anchored protein [23] , [27] . Seven pfl+ genes ( gsf2+/pfl1+ , pfl3+ , pfl4+ and pfl6+–pfl9+ ) have been reported to be upregulated in loss-of-function flocculent mutants of Cdk8 module genes ( cdk8+/srb10+ , med12+/srb8+ ) suggesting that the transcriptional repression of these putative flocculin genes may be controlled by Mediator [24] . The transcriptional repression of flocculin genes by Mediator may not be direct , but could be through mbx2+ since its expression is highly upregulated in the cdk8 kinase-mutant and med12Δ strain ( 9 and 13-fold increase , respectively , within top 11 up-regulated genes , found in supplementary data [24] ) . This proposed role of Mediator appears conserved in Sacc . cerevisiae as FLO genes are similarly upregulated in cdk8 mutants [38] . Despite these observations , no direct evidence has been shown aside from the gsf2+ study by Matsuzawa et al . [20] that the pfl+ gene products are actually flocculins . We have shown that this is indeed the case as single and double overexpression of the pfl+ genes is sufficient to trigger flocculation and that this flocculation is galactose-specific . The degree of flocculation triggered by single overexpression of the pfl+ genes varied , with gsf2+ and pfl9+ producing the largest and smallest flocs , respectively ( the pfl numbers correspond roughly to the degree of flocculation upon overexpression ) . This result indicates that gsf2+ encodes the most dominant flocculin compared to the other pfl+ genes . In agreement are the observations that only deletion of gsf2+ and not the other pfl+ genes prevented flocculation in flocculation-inducing medium , and reduced the constitutive flocculent phenotype to the greatest extent of all the transcription factor mutants examined in this study . Moreover , the strength of the flocculins was directly correlated with the amount of reduction in mbx2OE flocculation observed in the various pfl deletion backgrounds ( Figure 2B ) . These observations are similar in Sacc . cerevisiae , where overexpression of FLO1 produces the strongest flocculation compared to FLO5 , FLO9 , FLO10 and FLO11 [8] , [12] . Furthermore , the flocculation mediated by pfl+ genes was additive as observed in our double deletion and co-overexpression experiments ( Figure 2B and Figure 3B ) . These results suggest that the varying strengths of flocculation exhibited by S . pombe strains could be attributed to the upregulation of different combinations of pfl+ genes . We identified Rfl1 , an uncharacterized Zn ( 2 ) -Cys ( 6 ) transcription factor as a novel repressor of flocculation in fission yeast . The repression of flocculation by Rfl1 appears to be primarily mediated by the inhibition of gsf2+ expression since loss of gsf2+ can abrogate the constitutive flocculent phenotype of the rfl1Δ mutant . Rfl1 represses gsf2+ either directly by association with its promoter or indirectly by inhibition of mbx2+ transcription , thereby forming an inhibitory feed-forward loop ( coherent type 2 ) within the transcriptional-regulatory network ( Figure 9 ) . These results indicate that Mbx2 and Rfl1 are opposing transcription factors , and the latter inhibits mbx2+ and gsf2+ expression under non-inducing conditions of flocculation . Aside from its role in flocculation , Rfl1 may have a role in regulating genes involved in carbohydrate metabolism such as glycolysis and gluconeogenesis . Rfl1 appeared to be associated with promoters of genes enriched in glucose catabolic and metabolic processes ( p-values = 0 . 00092 and 0 . 00269 , respectively ) including adh1+ , hxk2+ , pfk1+ , tpi1+ , adh4+ , pgi1+ , gpd3+ , tdh1+ , pgk1+ , fba1+ , eno101+ , pyr1+ , SPCC794 . 01c ( predicted glucose-6-phosphate 1-dehydrogenase ) , SPBC2G5 . 05 ( predicted transketolase ) and SPBC660 . 16 ( phosphogluconate dehydrogenase ) . Most of these genes with the exception of fba1+ , eno101+ , SPCC794 . 01 , and SPBC660 . 16 were upregulated 1 . 2 to 26-fold in the rfl1Δ strain ( Table S4 ) . From these data , we speculate that Rfl1 could serve as a negative transcriptional regulator of several enzymes involved in the glycolysis and gluconeogenesis . Because flocculation and invasive growth are associated with nutritional limitation , Rfl1 may coordinate the expression of genes involved in flocculation and carbohydrate metabolism in fission yeast . Previously , the CSL proteins Cbf11 and Cbf12 were shown to exhibit antagonistic roles in flocculation [39] . Overexpression of cbf12+ or loss of cbf11+ triggers flocculation . However , none of their target genes have been identified . We present supportive evidence that Cbf12 induces flocculation by directly activating the transcription of gsf2+ . In addition , gsf2+ expression is up-regulated approximately 2 . 4-fold in the cbf11Δ strain suggesting that the repressive flocculation function of Cbf11 may also be directly mediated through gsf2+ . The activation and repression of gsf2+ transcription by Cbf12 and Cbf11 , respectively , may occur by competitive binding to promoter sites since both transcription factors have been shown to interact with a canonical CSL consensus sequence in vitro [39] . Several putative sites with six out of seven nucleotide match to the canonical CSL consensus sequence are located within 900 base pairs of the gsf2+ promoter ( data not shown ) . Further experimentation would be required to verify this proposed mechanism of gsf2+ transcriptional regulation by Cbf11 and Cbf12 . It is likely that cbf12+ plays a lesser role in activating flocculation compared to mbx2+ since the floc size resulting from cbf12+ overexpression is considerably smaller than the mbx2OE strain . Also unlike mbx2+ , deletion of cbf12+ is not sufficient to abrogate the flocculation of the rfl1Δ strain ( data not shown ) . These data suggest that the flocculent phenotype of the cbf12Δ rfl1Δ double mutant is probably caused by the presence of mbx2+ activity . CSL transcription factors are components of the conserved Notch signaling pathways in metazoans which primarily function in cell-to-cell communication during development [40] . Although multiple fungal CSL proteins have been discovered , their exact roles remain unclear in unicellular organisms [39] . Flocculation has been described as a manifestation of social behaviour in yeast with a purpose of enhancing survival under stressful conditions [3] . Therefore , it is conceivable that CSL transcription factors originated as regulators of this primitive form of cell-to-cell communication , and later evolved into the metazoan Notch signaling pathway . We also discovered novel functions of the Yox1 and Sre2 transcription factors in the repression of flocculation . Loss of yox1+ or sre2+ results in a mild flocculent phenotype . The Yox1 homeodomain transcription factor functions as a repressor of MBF ( Mlu1 binding factor ) target genes to prevent their inappropriate expression at the end of S-phase [36] . Transcriptional repression of MBF target genes is mediated by the direct interaction of Yox1 and Nrm1 to the MBF complex [41] . Deletion of yox1+ causes a cell cycle delay and results in elevated constitutive expression of MBF gene targets [36] . Similarly , these genes ( e . g . cdc18+ , cdc22+ , cdc10+ , cdt1+ , cdt2+ , cig2+ and nrm1+ ) were also found to be upregulated 2 . 2 to 6 . 4-fold in our yox1Δ microarray expression data ( Table S8 ) . We found that the flocculent phenotype of the yox1Δ strain is also dependent upon gsf2+ . However , the pfl+ genes including gsf2+ were not highly expressed in the yox1Δ strain . One possible explanation why pfl1+ genes were not highly expressed in the yox1Δ strain is that our experiments were performed under asynchronous culturing conditions , and therefore , the upregulation of pfl+ genes including gsf2+ could have been obscured if their expressions were periodically controlled in the vegetative cell cycle . However , it is unlikely that Yox1 regulates the pfl+ genes directly because previous chIP-chip analysis did not detect binding of Yox1 to the promoters of pfl+ genes [36] . Although there has been solid evidence linking yeast morphogenesis events such as pseudohyphal and hyphal growth to cell cycle regulators [42]–[47] , the relationship between cell cycle control and flocculation remains unclear . A flocculation function for yox1+ has not been reported in other yeasts . However , disruption of YOX1 in the Sacc . cerevisiae ∑1278b strain inhibited filamentous invasive growth , a process usually associated with flocculation during nutritional limitation , while deletion of C . albicans NRM1 reduced flocculation [48] , [49] . Sre2 is an uncharacterized membrane-tethered helix-loop-helix transcription factor predicted to be an ortholog of mammalian SREBP-1a , which is responsible for the transcriptional activation of genes needed for uptake and synthesis of cholesterol , fatty acids , triglycerides , and phospholipids [50] . While sre1+ , a paralog of sre2+ has been shown to function in the transcriptional activation of sterol-biosynthetic and hypoxic-adaptation genes , there has been no direct evidence that sre2+ plays similar biological roles [37] . Loss of sre2+ results in the upregulation of gsf2+ , pfl3+ and fta5+ transcripts ( 3 . 76 , 2 . 51 and 3 . 08-fold , respectively ) ( Figure 7C , Table S9 ) which may contribute to its flocculent phenotype . The sre2Δ flocculent phenotype requires gsf2+ activity and is independent of mbx2+ and cbf12+ since these transcripts are downregulated in the deletion mutant . In addition , the microarray expression profiles of yox1Δ and sre2Δ strains displayed similar differential gene expression despite the supposedly different functions of these transcription factors ( Figure 7B ) . Mitochondrial genes were found to be highly upregulated in both deletion mutants ( Tables S8 and S9 ) . This occurrence may not be unexpected for Sre2 if it has a similar role in hypoxia as Sre1 where mitochondrial function is probably impaired [37] . It is currently not clear whether the mitochondrial genes are direct targets of Yox1 and Sre2 or induced in response to an altered physiological state in the deletion mutants . Interestingly , mitochondrial activity has been reported to be important for flocculation and invasive growth in Sacc . cerevisiae [48] , [51] . Disruption of mitochondrial activity has been shown to alter the synthesis and structure of the cell wall , possibly by interfering with the interactions of flocculins and their substrates [52] . Based on these observations , the flocculent phenotype of yox1Δ and sre2Δ strains could be partially the result of enhanced mitochondrial activity from the upregulation of mitochondrial genes . A genome-wide systematic deletion screen previously uncovered a cell-to-surface adhesion function that is sensitive to the presence of galactose for the Adn2 and Adn3 transcription factors [29] . Here , we discovered that adn2+ and adn3+ have additional functions in flocculation . Overexpression of adn2+ and adn3+ induced minor flocculation while loss of these genes prevented flocculation in inducing glycerol medium ( Figure 8A and 8B ) . However , the flocculent phenotype of the adn2OE and adn3OE strains appeared to be primarily caused by the differential regulation of genes encoding cell wall-remodeling enzymes rather than flocculins . Several genes encoding cell wall-remodeling enzymes ( gas2+ , agn2+ , psu1+ and SPAC4H3 . 03c ) were highly induced when mbx2+ , adn2+ or adn3+ was overexpressed . In the adn2OE strain , gas2+ and SPAC4H3 . 03c were the most highly induced genes ( 17 . 9 fold and 36 . 8-fold , respectively ) ( Figure 8C , Table S11 ) while in the adn3OE strain , these two genes and psu1+ appeared within the top 20 up-regulated genes ( Table S12 ) . Similarly , in the mbx2OE strain , gas2+ , agn2+ , and SPAC4H3 . 03c appeared within the top 100 up-regulated genes ( greater than 3 . 7-fold increase , Figure 2A ) . We found that the single overexpression of these four genes could trigger flocculation ( Figure 8D ) . Cell wall remodeling is an essential process for proper growth and adaptation to environmental stresses in yeast cells . Part of the cell wall-remodeling process involves the dissolution of sugar moieties in the glucan layer and elongation of glucan chains by glycoside hydrolases and glycosyltransferases , respectively . Among these four genes , three ( agn2+ , psu1+ and SPAC4H3 . 03c ) encode for glycoside hydrolases while the fourth ( gas2+ ) encodes for a glycosyltransferase . Agn2 is an endo- ( 1 , 3 ) -α-glucanase that hydrolyzes ( 1 , 3 ) -α-glucans of the ascus wall for ascospore release [53] , [54] . Although agn2+ function appears only specific for sporulation , its ectopic expression could alter the cell wall structure during vegetative growth by inappropriate hydrolysis of ( 1 , 3 ) -α-glucan . Similarly , inappropriate glucan hydrolysis of the cell wall could be occurring as a result of ectopic expression of SPAC4H3 . 03c which encodes a putative ( 1 , 4 ) -α-glucanase ( Hertz-Fowler et al . , 2004 ) . Psu1 , which exhibits close homology to the members of the SUN family in Sacc . cerevisiae and C . albicans , as well as the BglA beta glucosidase of C . wickerhamii , has an essential function in cell wall synthesis [55] . Loss of psu1+ activity conferred resistance to ( 1 , 3 ) -β-glucanase suggesting that Psu1 may influence the amount or structure of ( 1 , 3 ) -β-glucan in the cell wall [55] . In addition , the ( 1 , 3 ) -β-glucanosyltransferase Gas2 has been shown to lengthen glucan chains during cell wall assembly and its overproduction is able to suppress the cell wall defect and lethality of gas1Δ cells [56] . Then how does the overexpression of these cell wall-remodeling genes trigger flocculation in S . pombe cells ? The expression of flocculin genes during vegetative growth is not well characterized in yeasts , but studies in Sacc . cerevisiae indicate that flocculin synthesis and insertion into the cell wall initiate in early exponential phase prior to the onset of flocculation during stationary phase [57] . This suggests that the flocculins are already present in the cell wall , but cannot induce flocculation because of inaccessibility to cell surface oligosaccharides . We speculate that the restructuring of the β-glucan layer during cell wall remodeling may result in the rearrangement of flocculins that enhances galactose oligosaccharide binding , thereby promoting flocculation . Several lines of evidence in Sacc . cerevisiae support this hypothesis . First , alteration of cell wall structure by disruption of PKC1 activity results in flocculation [58] . Second , heat shock induces flocculation and regulation of cell wall-remodeling genes via the Hsf1 transcription factor [57] , [59] . Currently , we cannot rule out that agn2+ , psu1+ , SPAC4H3 . 03c and gas2+ are the only cell wall remodeling enzymes that can trigger flocculation when overexpressed . Other genes with potential functions in cell wall modification and integrity such as gas4+ , gma12+ , meu7+ , agl1+ , meu10+ and mde5+ were also detected as putative target genes of Mbx2 ( Table S2 ) . In contrast , there was little change in gsf2+ transcript levels in the adn2OE or adn3OE strains compared to the empty vector control . However , the flocculation triggered by adn2+ and adn3+ overexpression was abrogated in a gsf2Δ background indicating that gsf2+ was indispensible for this process ( Figure 8A ) . Altogether , these results suggest that Gsf2 is likely expressed in the cell wall as an inactive flocculin , and the cell wall remodeling resulting from adn2+ and adn3+ overexpression alters the arrangement of Gsf2 and possibly other flocculins that now becomes favorable for flocculation . The single overexpression of the cell wall-remodeling genes triggered flocculation to a greater extent than the adn2OE and adn3OE strains . A possible explanation for the different degrees of flocculation between the transcription factor and its target genes could be that overexpression of adn2+ and adn3+ causes reduced fitness due to toxicity effects associated with a greater misregulation of genes compared to the aberrant production of a single enzyme . Consistent with this theory is that adn2OE and adn3OE strains exhibited additional phenotypes including septation defects ( Figure 8E ) which were not observed when gas2+ , agn2+ , psu1+ and SPAC4H3 . 03c were overexpressed ( data not shown ) . Furthermore , a systematic overexpression analysis of 5280 genes in Sacc . cerevisiae revealed that genes encoding for transcription factors , signalling molecules and cell cycle regulators were more likely to cause reduced fitness [60] . In S . pombe , cell separation involves the transcriptional activation of adg1+ , adg2+ , adg3+ , agn1+ , eng1+ , cfh4+ and mid2+ by the Ace2 transcription factor , which is in turn regulated by the Sep1 forkhead transcription factor [61]–[64] . We discovered that the adn2OE or adn3OE strains displayed multisepta and forkhead phenotypes similar to loss-of-function mutations of these cell separation genes . The cell separation defect in adn2OE and adn3OE strains is likely due to the downregulation of ace2+ transcription since ace2+ and its target genes were substantially downregulated in these strains ( Figure 8C ) . However , sep1+ transcript levels remained unchanged in the adn2OE and adn3OE strains indicating that their involvement in cell separation phenotype could be either downstream of sep1+ or parallel to the sep1+ pathway . In additional to its flocculation role , Adn2 and Adn3 appear to have a separate function in cell separation perhaps by directly or indirectly repressing ace2+ transcription . Experiments are planned in the future to address these possibilities . Interestingly , we also found some evidence that supports a role of Mbx2 and Cbf12 in cell separation perhaps through repression of ace2+ activity . Overexpression of mbx2+ and cbf12+ results in significant down-regulation of all seven Ace2 target genes approximately 1 . 5 to 3 . 4-fold relative to the empty vector control . ( Tables S2 and S6 ) . The mbx2OE strain indeed showed septation defects but were slightly different in nature than the adn2OE and adn3OE strains with less multi-septation and more mislocalization of septum material ( data not shown ) . Moreover , overexpression of cbf12+ has been reported to produce multisepta phenotypes albeit at a low frequency [39] . These observations indicate the possible existence of crosstalk between flocculation and cell separation pathways mediated by the Mbx2 , Cbf12 , Adn2 and Adn3 transcription factors ( Figure 9 ) . A comparison between the flocculation network of budding and fission yeast revealed both conserved and divergent features within the transcriptional circuitry . In Sacc . cerevisiae , the positive and negative transcriptional controls of the dominant flocculin gene FLO1 by Flo8p or Mss11p , and Sfl1p , respectively , draw parallel to gsf2+ regulation by the Mbx2-Rfl1 and Cbf12-Cbf11 opposing transcription factors in S . pombe [16] , [19] . The conservation of the flocculin genes between these two yeasts is apparent among these transcription factors . Similar to Mbx2 , Mss11p and Flo8p appear to activate multiple flocculin genes ( FLO1 , FLO9 and FLO11 ) , while the latter may also regulate genes encoding cell wall enzymes ( STA1 and SGA1 which both encode glycoside hydrolases ) [9] , [16] . The putative target genes of the Sfl1 repressor have been reported to include FLO1 and FLO11 [19] , [65] . In contrast to the conservation of the flocculin genes , the types of transcription factors involved in flocculation are quite different between the two yeasts . Mbx2 belongs to the MADS box family , while the DNA-binding domains of Flo8p and Mss11p have not been defined . In addition , the Sfl1p and Rfl1 repressors contain the heat shock factor and Zn ( 2 ) -Cys ( 6 ) DNA-binding domains , respectively . Moreover , CSL transcription factors ( Cbf11/Cbf12 ) are not found in Sacc . cerevisiae . These observations would imply that the cis-regulatory elements controlling transcription of the flocculin genes have likely undergone considerable rewiring within the transcriptional-regulatory network between the two yeasts . However , it was recently demonstrated that heterologous expression of Flo8p and Mbx2 could induce gsf2+ and FLO1 transcription in fission and budding yeast , respectively [20] , [21] . Therefore , despite the divergent types of transcription factors controlling flocculin gene expression in the two yeasts , there may be some degree of conservation among the cis-regulatory sequences . Although the transcription factors regulating flocculation appear to be quite different between the two yeasts , the downstream transcriptional events involved in the repression of flocculin genes are likely to be conserved . Disruption of genes encoding the Ssn6p-Tup1p general corepressor or the Cdk8 module of Srb/Mediator complex have been shown to cause upregulation of flocculin genes and constitutive flocculation in S . pombe and Sacc . cerevisiae [65] . In the latter yeast , Sfl1p represses FLO1 and FLO11 transcription through physical interactions with Ssn6p and Srb proteins ( Srb8p , Srb9p and Srb11p ) [19] , [65] , [66] . Moreover , Sfl1p has been reported to repress FLO8 in Saccharomyces diastaticus . The observation that Sfl1p can repress FLO1 transcription directly and indirectly through FLO8 seems very similar to the inhibitory feed-forward regulation of gsf2+ by Rfl1 in S . pombe . If these connections are truly analogous , then there is a possibility that Rfl1 repression could also be mediated through physical interactions with the Cdk8 module proteins . In the srb10− mutant , gsf2+ and mbx2+ expression are upregulated suggesting that its flocculent phenotype could be caused by a failure to repress mbx2+ transcription [24] . In addition , the flocculent phenotype of tup11+/tup12+ mutants [25] and the abrogation of lkh1Δ flocculation in the absence of mbx2+ [21] supports the role of Tup11/12 corepressor in Mbx2-Rfl1-mediated flocculation . Taken together , we speculate that Srb10 and Tup11/12 activity and binding may be required for Rfl1-mediated repression of gsf2+ and mbx2+ . Future experiments focusing on the interactions between Rfl1 , Tup11/12 and Srb8-10 in relation to flocculation would provide clarification to our speculation . Our analyses of the transcription factors implicated in flocculation of S . pombe revealed the possible existence of several network motifs including positive autoregulation of mbx2+ and cbf12+ and regulation of gsf2+ by a inhibitory feed-forward loop ( coherent type 2 ) . The latter involves the Rfl1 transcriptional repression of gsf2+ directly and indirectly by inhibition of mbx2+ expression . Autoregulatory motifs have not been detected so far for FLO8 , MSS11 and SFL1 . The discovery of these network motifs in S . pombe suggests that the transcriptional inhibition of gsf2+ could occur more rapidly than its transcriptional activation . Experimental and modeling studies have proposed that positive and negative autoregulation of transcription factors generate slow and fast response times , respectively , within a transcriptional-regulatory network [67] . Under positive autoregulation , the synthesis rate of the transcription factor is initially slow at low concentrations , but increases as the concentration of the transcription factor reaches the activation threshold of the promoter , while negative autoregulation accelerates the attainment of steady state levels of the transcription factor [67] . Moreover , the inhibitory feed-forward motif of Rfl1 seems to indicate that repression of gsf2+ expression likely happens in a shorter period compared to its activation . Altogether , these data suggest that the onset of flocculation may occur gradually while repression of the flocculation pathway is a much faster process . Consistent with this speculation is the observation that it requires several days for wild-type S . pombe cells to undergo flocculation when grown in inducing medium . In summary , we have provided an initial and substantial view of the transcriptional-regulatory network governing flocculation in S . pombe . Found within this network are the master regulators Mbx2 , Cbf12 , Adn2 and Adn3 , which are able to trigger flocculation when overexpressed by the activation of their target genes encoding for flocculins and cell wall-remodeling enzymes . In addition , several repressors including Rfl1 were uncovered that play a major role in the regulation of these target genes . However , significant gaps of knowledge surrounding the transcriptional-regulatory network still remain . The environmental cues that impinge upon the activity of the positive and negative regulators , as well as the dynamics of transcription factor binding and regulation of target genes during the onset of flocculation remain to be elucidated . Also , although gsf2+ encodes the dominant flocculin , it is currently unclear whether the other flocculins have nonessential or more specialized roles during flocculation . Detailed analyses of the temporal and spatial expression of the pfl+ genes would be required to address these questions . Moreover , the exact mechanism of how other biological processes such as cell wall restructuring and mitochondrial function influence flocculation is unknown . Further studies to expand our knowledge of this transcriptional-regulatory network would provide a more comprehensive understanding of flocculation control and contribute to a valuable resource for the improvement of industrial yeast applications .
All strains used in this study are listed in Table S1 and were maintained on YES or EMM medium . Geneticin , nourseothricin , and thiamine hydrochloride were added to media at a concentration of 150 mg/L , 100 mg/L , and 15 µM , respectively . EMM medium was supplemented with amino acids when necessary at 225 mg/L each for adenine , leucine , and uracil . Matings were performed on SPAS medium . Wild type and deletion strains were assayed for flocculation in YEGlyEtOH ( flocculation-inducing ) medium containing 1% ( w/v ) yeast extract , 3% ( v/v ) glycerol , and 4% ( v/v ) ethanol . Overexpression strains containing ORFs under control of the nmt1 or nmt41 promoter were grown in EMM minus thiamine medium . Standard genetics and molecular biology techniques were performed as described in [68] . A PCR-based stitching method was utilized to construct the deletion and epitope-tagged strains . For construction of deletion strains , ∼500 bp fragments upstream and downstream of the ORF and the KanMX6 or NatMX6 cassette were PCR-amplified and gel-purified . The 3′ end of the upstream fragment and 5′ end of the downstream fragment contained ∼25 bp homology to the selectable marker cassette sequence . Approximately equimolar amounts ( ∼40 ng ) of each PCR fragment were combined and stitched together in a 20 µl PCR reaction ( 0 . 2 mM dNTPs and 0 . 4 units of Phusion HF DNA polymerase ( New England Biolabs ) , and subjected to one cycle of 98°C ( 30 sec ) , 5 cycles of 98°C ( 15 sec ) , 60°C ( 1 min ) , and 72°C ( 1–2 min ) and a final extension at 72°C ( 5 min ) . The stitched product was then amplified in a 50 µl PCR reaction by combining the entire stitched reaction with 6 nmol dNTPs , 0 . 6 units of Phusion HF DNA polymerase and 20 pmol each of the outer pair of primers and then subjected to one cycle of 98°C ( 30 sec ) , 30 cycles of 98°C ( 10 sec ) , 60°C ( 30 sec ) and 72°C ( 2 min ) , and a final extension at 72°C ( 5 min ) . The amplified product was gel-purified and transformed into the appropriate strain by lithium acetate transformation . A similar strategy was used to construct GFP-tagged transcription factors under the control of the native promoter . To tag the transcription factor with GFP at the C-terminus , ∼500 bp upstream and downstream fragments flanking the stop codon and the GFP-KanMX6 cassette ( amplified from pYM27 plasmid , [69] ) were PCR-amplified for the stitching reaction as described above . To conserve the native promoter in the N-terminal GFP fusion of Mbx2 , 1 kb upstream of the mbx2+ start codon was amplified along with four other fragments for PCR stitching: ( 1 ) ∼500 bp upstream of the aforementioned 1 kb fragment; ( 2 ) ∼500 bp downstream of the mbx2+ start codon; ( 3 ) KanMX6 cassette and; ( 4 ) the GFP ORF with its stop codon removed and a GDGAGL linker added ( adapted from [70] ) . All five fragments contained ∼25 bp overlapping homology to their respective flanking fragments and were PCR-stitched as described above . Proper gene deletion and GFP tagging were confirmed by colony PCR screen and the resulting amplicons sequenced . Genes were overexpressed with the nmt1 promoter by cloning the entire ORFs of interest into the pREP1 or pREP2 vector . For ChIP-chip experiments , C-terminal triple HA-tagged Mbx2 , Rfl1 , and Cbf12 were expressed with the nmt41 promoter by cloning the corresponding ORFs into pSLF272 [71] . All the clones were PCR-confirmed , sequenced , and transformed into appropriate strains by the lithium acetate method . Expression of the HA-tagged proteins was verified by western blotting with anti-HA F-7 antibody ( Santa Cruz Biotechnology , Santa Cruz , CA ) . Strains overexpressing the triple HA-tagged Mbx2 , Rfl1 , and Cbf12 were grown in 200 ml of EMM medium containing appropriate supplements without thiamine for 18-20 hr to induce the nmt41 promoter . The empty vector control strain was cultured concurrently to a matching cell density of ∼8×106 cells/ml prior to harvesting . The experimental culture was divided into two , each for ChIP-chip and microarray expression profiling while the control culture was only utilized in the latter . The expression profiling cultures were harvested by centrifugation ( 1800× g , 3 min , 20°C ) , followed by immediate freezing of the cell pellets in liquid nitrogen . Culturing of adn2OE and adn3OE strains were performed similarly except that these genes were driven by the nmt1 promoter and were not epitope-tagged . For transcription factor deletion strains ( rfl1Δ , cbf11Δ , sre2Δ , and yox1Δ ) , the mutant and an isogenic wild-type strain were concurrently grown in YES medium and harvested at a similar cell density as described above . Total RNA extraction , mRNA isolation , reverse transcription with aminoallyl-dUTP ( Sigma-Aldrich , Oakville , ON ) , and Cy™3/Cy™5 ( GE Healthcare , Buckinghamshire , UK ) dye coupling of cDNA were performed with dye reversal as previously described [72] . Purified Cy™3- and Cy™5-labelled cDNA ( 1 µg in total ) was hybridized onto custom-designed 8×15 K Agilent expression microarrays containing 60mer probes to all S . pombe ORFs in 2–3 times coverage per gene . The hybridization procedure was carried out according to the manufacturer's instructions ( Agilent Technology , Santa Clara , CA ) with the exception for the use of Human Cot-1 DNA . The microarrays were washed in 6× SSPE/0 . 005% sodium N-lauroylsarcosine at room temperature for 5 min followed by a second wash in pre-heated 42°C 0 . 6× SSPE for 2 min . The microarrays were scanned with a GenePix4200A scanner ( Molecular Devices , Sunnyvale , CA ) . The raw microarray data was lowess normalized [73] and the average log2 ratios with the corresponding t-test p values [74] from the dye-swap experiments were obtained using the R Bioconductor Limma package . Heat map images of the microarray expression and ChIP-chip data were constructed with Cluster 3 . 0 [75] and Java Treeview 1 . 1 . 6r2 [76] . The microarray expression data has been submitted to the NCBI Gene Expression Omnibus Database ( GSE41730 ) . Culturing of the HA-tagged transcription factor strains are described above . The culture was fixed by the addition of a final concentration of 1% formaldehyde and agitation for 30 min at room temperature . The formaldehyde was quenched by the addition of 2 . 5 M glycine to a final concentration of 125 mM and agitation for 5 min at room temperature . The cells were then centrifuged ( 800× g , 5 min , 4°C ) , washed twice in 25 ml 1× ice-cold PBS ( 137 mM NaCl , 2 . 7 mM KCl , 10 mM Na2HPO4 , 1 . 8 mM KH2PO4 pH 7 . 4 ) and washed once with 2 ml ice-cold lysis buffer ( 50 mM NaCl , 50 mM HEPES-KOH pH 7 . 5 , 0 . 1% SDS , 1% Triton X-100 , 1 mM EDTA , 0 . 1% sodium deoxycholate and 1 tablet/50 ml Protease Inhibitor Cocktail ( Roche Applied Science , Indianapolis , IN ) ) . The cell pellet was resuspended in 1 . 6 ml lysis buffer and stored at −80°C . The cell suspension was transferred to two 2 ml bead beating vials containing 800 µl of 0 . 5 mm Zirconia/Silica beads ( BioSpec Products , Bartlesville , OK ) and subjected to 3 cycles of alternating 2 min beating and 2 min incubation on ice with a Mini Beadbeater 16 ( BioSpec Products , Bartlesville , OK ) . The lysed cells were collected by puncturing the bottom of the bead-beating vial with a flame-heated inoculating needle and placing the vial on a sonication tube nested in 10 ml disposable culture tubes prior to centrifugation ( 800× g , 3 min , 4°C ) . The cell pellet was resuspended , transferred to chilled microcentrifuge tubes , centrifuged ( 16 , 000× g , 15 min , 4°C ) to remove unbound soluble proteins , and the resulting pellet resuspended in 800 µl of fresh lysis buffer in a sonication tube . Total cell lysate volume was adjusted to 2 . 2 ml with lysis buffer and subjected to 4 cycles of sonication and 1 min on ice incubation at 30% amplitude , 30 sec setting using a Sonic Dismembrator with a 1/8 tapered microtip probe ( Thermo Scientific , Waltham , MA ) . The sonicated cell lysate was centrifuged ( 4600× g , 2 min , 4°C ) and the supernatant stored at −80°C . The supernatant was tested to ensure that greater than 90% of the sonicated DNA was in the size range of 100 bp–1 kb by subjecting a sample ( ∼50 µl ) of the supernatant to overnight reverse-crosslinking at 65°C and phenol-chloroform extraction , followed by gel electrophoresis of 3–5 µg of DNA . To immunoprecipitate the chromatin-bound transcription factor , 100–200 µl of Dynabeads conjugated with sheep anti-mouse IgG ( Invitrogen Life Technologies , Carlsbad , CA ) were washed twice in 800 µl ice cold 1× PBS-BSA ( 5 mg/ml BSA , 1× PBS ) , resuspended in 800 µl cold 1× PBS-BSA with 5 µg of anti-HA F-7 antibody ( Santa Cruz Biotechnology , Santa Cruz , CA ) and shaken gently for 2 hr at 4°C on a Labquake Tube Shaker ( Thermo Scientific , Waltham , MA ) . The beads were washed twice in 1 ml cold deoxycholate buffer ( 100 mM Tris-HCl pH 8 , 1 mM EDTA , 0 . 5% ( w/v ) sodium deoxycholate , 0 . 5% ( v/v ) NP-40 , 250 mM LiCl ) and twice in 1 ml cold lysis buffer . The beads were resuspended in 200 µl 1× PBS-BSA , combined with 400 µl of sonicated cell lysate , and shaken gently for 2 hr at 4°C . Four washes of 5 minutes each were next carried out: ( 1 ) 1 . 4 ml cold lysis buffer at 4°C; ( 2 ) 1 . 4 ml cold lysis buffer with 400 mM NaCl at 4°C; ( 3 ) 1 . 4 ml deoxycholate buffer at room temperature and; ( 4 ) 1 . 4 ml TE ( pH 8 ) at room temperature . The transcription factor and bound DNA were eluted twice from the Dynabeads by incubating with 250 µl TES each ( TE pH 8 , 1% ( w/v ) SDS ) at 65°C for 6 min . Dynabead washing and the supernatant collection were performed using DynaMag™−2 ( Invitrogen , Carlsbad , CA ) . For the input DNA , 200 µl of the cell lysate was added to 300 µl TES . Both the immunoprecipitated and input cell lysates were incubated at 65°C overnight to reverse the DNA-protein cross-linking . Western blotting with anti-HA antibody was performed to confirm proper pull-down of the transcription factor . For protein removal , both immunoprecipitated and input samples were incubated with 200 µg Proteinase K ( Promega , Madison , WI ) and 20 µg glycogen ( Roche Applied Science , Indianapolis , IN ) at 56°C for 2 hr . The DNA was then extracted by phenol-chloroform extraction , ethanol-precipitated overnight , washed once with 70% EtOH , resuspended in 42 µl TE containing 0 . 1 µg DNAse-free RNaseA ( Roche Applied Science , Indianapolis , IN ) , and incubated for 30 min at 37°C . Blunt ends were generated in the entire immunoprecipitate and input DNA samples with 1 unit of T4 DNA Polymerase ( Invitrogen Life Technologies , Carlsbad , CA ) , 1× NEB Buffer #2 ( New England Biolabs , Ipswich , MA ) , 5 µg NEB BSA , and 10 nmol dNTPs in a 110 µl reaction by incubation at 12°C for 20 min , followed by phenol-chloroform extraction and ethanol precipitation with 10 µg glycogen and 1/10 volume 3 M NaOAc . The DNA pellets were washed in 70% EtOH and resuspended in 25 µl water . Approximately 1/5 of precipitated input DNA was used in the subsequent ligation reaction as input DNA concentration was >100 times greater than that of immunoprecipitated DNA . For ligation of linkers to blunt ends , the resuspended DNA was incubated with 1000 units of concentrated T4 DNA Ligase ( New England Biolabs , Ipswich , MA ) , 1× T4 DNA Ligase Buffer ( Invitrogen Life Technologies , Carlsbad , CA ) , and 200 pmol annealed linker ( 15 µM Oligo #1 5′-GCGGTGACCCGGGAGATCTGAATTC-3′ and 15 µM Oligo #2 5′-GAATTCAGATC-3′ in 250 mM Tris ) at 16°C overnight . The annealed linker and the ligation mix were kept on ice at all times prior to overnight incubation . The DNA was ethanol-precipitated , washed , and resuspended in 25 µl water as described above . The ligated DNA was PCR-amplified by adding 15 µl of labeling mix ( 2 µl aa-dUTP dNTP mix containing 5 mM each dATP , dCTP and dGTP , 3 mM dTTP , and 2 mM aminoallyl-dUTP ( Sigma-Aldrich , St . Louis , MO ) ) , 1 . 25 µl 40 µM Oligo #1 ( 5′-GCGGTGACCCGGGAGATCTGAATTC-3′ ) , 4 µl 10× ThermoPol Buffer ( New England Biolabs , Ipswich , MA ) and 7 . 75 µl water ) in a PCR cycler paused at 55°C . A 10 µl enzyme mix containing 5 units of GoTaq DNA polymerase ( Promega , Madison , WI ) , 0 . 001 units of Pfu Turbo DNA polymerase ( Stratagene , La Jolla , CA ) , and 1× ThermoPol Buffer ( New England Biolabs , Ipswich , MA ) was added and the PCR proceeded with one cycle of 55°C ( 4 min ) ; 72°C ( 5 min ) ; 95°C ( 2 min ) and 30 cycles of 95°C ( 30 sec ) ; 55°C ( 30 sec ) ; 72°C ( 1 min ) , followed by a final extension at 72°C ( 4 min ) . The PCR products were purified ( QIAGEN , Valencia , CA ) with a few modifications: ( 1 ) buffer PE was replaced with phosphate wash buffer ( 5 mM KPO4 pH 8 . 5 , 80% ethanol ) and ( 2 ) buffer EB was replaced with phosphate elution buffer ( 4 mM KPO4 pH 8 . 5 ) . A sample of the purified PCR product was run on an agarose gel to check for fragment sizes ranging between 100 bp and 1 kb . The purified PCR products were quantified , and equal amounts of immunoprecipitated samples and corresponding input samples were coupled to Cy™3 and Cy™5 dyes as described above . The labelled samples ( total amount of 3–5 µg ) were hybridized onto an Agilent 4×44 K S . pombe Genome ChIP-on-chip microarray according to the manufacturer′s instructions ( Agilent Technology , Santa Clara , CA ) except for the use of Human Cot-1 DNA . The washing and scanning of the microarrays were performed as described above . The ChIP-chip data was normalized by scaling in Limma [73] and analyzed by ChIPOTle Peak Finder Excel Macro [77] with the default setting of log2 ratio cut-off of 1 . Peaks located within 3 kb upstream of a start codon and 2 kb downstream of a start codon within a coding region or 3′-UTR , in the case of short ORFs , were assigned to the gene . ChIP-chip data sets are found in Tables S3 , S5 , and S7 . Genes with multiple peaks are noted in the data set with the peak values . The ChIP-chip data has been submitted to the NCBI Gene Expression Omnibus Database ( GSE41730 ) . The transcription factor binding specificities were determined by RankMotif++ [31] and MEME [30] . S . pombe promoter sequences 1000 bp upstream of the translational start site were used for these motif-finding algorithms . For MEME , promoter sequences of genes with various log ratio thresholds from expression microarray and ChIP-chip experiments were input into the MEME online server . RankMotif++ was applied to the entire expression microarray data since its motif-searching algorithm is threshold independent . The consensus sequences of the transcription factor binding sites were displayed by submitting the position weight matrices obtained from RankMotif++ analysis into the enoLOGOS online server [78] . Strains were grown in flasks at 30°C for the appropriate time , and 10 ml of culture was transferred to culture tubes for strains with larger floc sizes . Images were acquired immediately after vigorous shaking in glass culture tubes with a Canon G10 digital camera . For strains with mild flocculation , flocs were harder to visualize in culture tubes , and therefore , were observed in 90 mm plastic petri dishes . 10–15 ml of culture was transferred to petri dishes , followed by gentle shaking [8] on an orbital low-speed shaker ( Labnet International , Woodridge , NJ ) at maximum speed for one hour in room temperature . Floc images in petri dishes were captured using a SPImager ( S&P Robotics Inc . , Toronto , ON ) . Deflocculation of flocculent strains was performed by the addition of 2–20% D- ( + ) -galactose or 10 mM EDTA . The reflocculation of the deflocculated cells was performed by washing with water , resuspending the cells in YES or EMM medium or 100 mM CaCl2 and allowing the culture to sit for 30 min at room temperature . For the overexpression of pfl+ genes , the strains were inoculated at a concentration of 107 cells in 100 ml of EMM without thiamine and cultured for 3 days at 30°C . For the weaker flocculent strains ( pfl2+–pfl9+ ) , 5 ml of the 3-day culture was then inoculated into 100 ml of fresh EMM without thiamine and incubated for another 3–4 days at 30°C followed by the petri dish flocculation assay as described above . Fresh EMM medium was added on the third day to prevent cells from remaining in stationary phase . Flocculation assays for the more flocculent overexpression strains were similarly carried out except the induction times were less than three days and did not require refeeding with fresh EMM medium . It should be noted that the empty vector control cells also eventually flocculate after refeeding with fresh EMM medium , but the onset of flocculation and flocs were delayed for several days and less pronounced , respectively , compared to the weakest flocculent overexpression strains . Wild-type strain and deletion mutants ( mbx2Δ , gsf2Δ , cbf12Δ , adn2Δ and adn3Δ ) were induced to flocculate by inoculating cells at a concentration of 108 cells in 100 ml of YEGlycEtOH medium and culturing for 5 days at 30°C followed by the petri dish flocculation assay as described above . A patch of cells approximately 1/6 of a 90 mm petri dish was grown on YES medium for two days at 30°C and transferred as described in [29] onto a LNB plate ( 0 . 067 g/L yeast nitrogen base without amino acids ( Bacto ) , 20 g/L glucose , 20 g/L agar , salts and vitamins as for EMM ) with an underlying layer of YE + ALU ( 0 . 5% YE , 225 mg/L adenine , leucine , and uracil each ) [79] . The plates were incubated at 30°C for 2 weeks before testing for cell-to-surface adhesion by washing cells off under a gentle stream of water and for invasive growth by rubbing the remaining cells off the agar with a finger under a stream of water . For strains showing resistance to rigorous washing by finger , a small section of the agar was cut out and observed under a Zeiss AxioScope A1 tetrad microscope ( Zeiss , Thornwood , NY ) . Invasive growth was observed by the presence of elongated and branched cells remaining underneath the agar [29] , [79] . Images of GFP-tagged cells were acquired with a Zeiss AxioScope 2 microscope ( Zeiss , Thornwood , NY ) and Scion CFW Monochrome CCD Firewire Camera ( Scion Corporation , Frederick , MD ) . Fluorescence intensity was quantitated using the open source software ImageJ ( version 1 . 44 ) ( National Institutes of Health ) . First , the background signal for each image was subtracted using the “Subtract Background” function ( 50 pixel rolling ball radius ) . Individual cells were then selected as regions of interest using the freehand or polygon selection tools . Using the “Set Measurements” function both the area and integrated density were determined for each selected cell ( n ranged between 27 and 50 ) . Corrected GFP intensity was determined for each cell and was defined as the quotient of integrated density/area in background subtracted images . The averaged integrated density/area measurements for a given number cells is presented as the mean corrected GFP intensity with standard deviation . Significant differences between means were calculated by the Student t-test . To view nuclei and cell wall material , cells were methanol-fixed and stained with DAPI ( 1 µg/ml ) and calcofluor white ( 50 µg/ml ) , respectively .
|
Flocculation is a process that involves yeast cells adhering to one another to form clumps called flocs . This trait is important for industrial yeast applications as it provides a cost-effective and efficient method to remove yeast cells . The adherence between cells occurs by the binding of glycoproteins known as flocculins and carbohydrate molecules located on the cell surface . To better understand how flocculation works , the genes that encode for flocculins and the transcription factors that regulate their expression need to be identified . In the fission yeast S . pombe , many of the flocculins and transcription factors that function in flocculation are not known . To address this gap in knowledge , we have employed molecular genetics and functional genomic approaches to uncover transcription factors and their target genes that play a role in flocculation . We discover that flocculation in S . pombe is regulated by a complex network of transcription factors that activate and repress themselves , as well as multiple target genes that encode for flocculins and cell wall–remodeling enzymes . The comparison of the flocculation regulatory networks between fission and budding yeasts indicates that they mainly differ in the types of transcription factors and their binding sequences .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genomics",
"functional",
"genomics",
"model",
"organisms",
"gene",
"expression",
"genetics",
"molecular",
"genetics",
"biology",
"computational",
"biology",
"yeast",
"and",
"fungal",
"models",
"microbiology",
"genetics",
"and",
"genomics"
] |
2012
|
Deciphering the Transcriptional-Regulatory Network of Flocculation in Schizosaccharomyces pombe
|
The microsomal , membrane-bound , human cytochrome P450 ( CYP ) 2C9 is a liver-specific monooxygenase essential for drug metabolism . CYPs require electron transfer from the membrane-bound CYP reductase ( CPR ) for catalysis . The structural details and functional relevance of the CYP-membrane interaction are not understood . From multiple coarse grained molecular simulations started with arbitrary configurations of protein-membrane complexes , we found two predominant orientations of CYP2C9 in the membrane , both consistent with experiments and conserved in atomic-resolution simulations . The dynamics of membrane-bound and soluble CYP2C9 revealed correlations between opening and closing of different tunnels from the enzyme's buried active site . The membrane facilitated the opening of a tunnel leading into it by stabilizing the open state of an internal aromatic gate . Other tunnels opened selectively in the simulations of product-bound CYP2C9 . We propose that the membrane promotes binding of liposoluble substrates by stabilizing protein conformations with an open access tunnel and provide evidence for selective substrate access and product release routes in mammalian CYPs . The models derived here are suitable for extension to incorporate other CYPs for oligomerization studies or the CYP reductase for studies of the electron transfer mechanism , whereas the modeling procedure is generally applicable to study proteins anchored in the bilayer by a single transmembrane helix .
The liver-specific , mammalian cytochromes P450 ( CYPs ) are monooxygenases essential for drug metabolism [1] . CYPs accept two electrons from the CYP reductase ( CPR ) to oxidize a wide range of water-soluble and liposoluble xenobiotics . CYPs and the CPR are anchored in the membrane of the endoplasmic reticulum ( ER ) by a single transmembrane anchor with the catalytic domain facing the cytosol [2] . The phospholipid composition has been shown to modulate different steps of catalysis [3] , ligand binding [4] and the binding of CYPs to CPR [5] . The interaction of CYPs with the membrane has been suggested to permit the direct binding of liposoluble substrates to the enzyme's buried active site [6] , [7] . Therefore , it is essential to understand this interaction at atomic resolution . CYPs are anchored in the membrane by a single N-terminal transmembrane α-helix . In addition , CYPs insert a hydrophobic region of the catalytic domain in the lipid bilayer . This has been shown by experiments that probed the recognition of reconstituted microsomes by site-directed antibodies against peptides of the human CYP2B1 and the rabbit CYP2B4 [2] , [8] . Engineered CYPs with the transmembrane domain removed retain membrane-binding properties [2] , [9] . The height of the globular domain above the lipid bilayer has been estimated to be 35±9 Å [10] . The heme tilt angle with respect to the membrane ( the dihedral angle between the heme plane and the membrane plane ) has been shown to vary in different CYP isoforms from 38 to 78° [11] . The positions of different residues of CYP2C2 relative to the membrane have been inferred from tryptophan fluorescence quenching experiments [12] . These experiments provide valuable information for building models of the membrane-bound CYPs . However , not all the experiments are consistent with each other and there are multiple , different orientations of the protein in the membrane that are consistent with different experiments . Moreover , if different CYPs actually adopt different orientations in the membrane , the relevance of the experiments might be restricted to the investigated isoform . Based on the first crystal structure of a mammalian CYP , the rabbit CYP2C5 , a model with a shallow insertion of the catalytic domain in the membrane was proposed [13] . From the crystal structure of human CYP2C8 as a dimer , a model for CYP oligomerization in the membrane was suggested [14] . Follow-up experiments confirmed the dimeric state of CYP2C8 [15] . From these two models , the approximate orientation of CYPs in the membrane was proposed . The estimated values of the heme tilt angle were in the upper range of the experimentally derived values [11] . However , these models have not been investigated further by other approaches such a computer simulation . Furthermore , such models are limited by the uncertainties in the atomic resolution structures of CYPs which were resolved using engineered enzymes with the N-terminal transmembrane helix removed . In some cases , protein solubility was further increased by mutation in the region of the FG loop [7] . Computational tools such as the OPM ( Orientations of Proteins in Membranes ) database fail to predict an orientation of CYPs in the membrane that is consistent with experiments [16] . Therefore , an atomic-resolution model of the CYP-membrane interaction remains elusive . Most of the crystal structures of CYPs reveal a closed conformation of the enzymes . Several structures of the rabbit CYP2B4 show protein conformations with wide open clefts that were proposed to be relevant in the context of lipid bilayers [17] , [18] . In addition , the human CYP3A4 has been shown to adopt open conformations depending on the size of the bound substrate [19] . To what extent the membrane influences the ability of CYPs to adopt open conformations is not understood . The CYP active site is buried deep in the protein . Multiple tunnels from the active site were found in CYP crystal structures and have been classified [20] . From simulations of ligand egress , it was proposed that mammalian CYPs may use more than one tunnel for ligand passage . It was suggested that liposoluble substrates bind through tunnels leading from the lipid bilayer while soluble products are released through solvent-accessible routes [6] , [21] , [22] . How the lipid bilayer influences the opening and closing of different tunnels remains unclear . Here , we report the structure and dynamics of the full-length , membrane-bound CYP2C9 , the second most abundant CYP in the human liver which metabolizes about 20% of all xenobiotics [23] . We found two orientations of CYP2C9 in the membrane , both consistent with experiments . While the enzyme remained mostly in closed or nearly closed conformations , we found both membrane and ligand dependent correlations between opening and closing of different tunnels from the buried catalytic center . We propose that the membrane promotes binding of liposoluble substrates by stabilizing protein conformations with an open access tunnel and provide evidence for selective substrate access and product release routes in mammalian CYPs . We describe a procedure to model and simulate the membrane-bound CYPs that is not biased towards a certain orientation of the protein in the membrane and that may be extended to incorporate the CPR as well .
We applied a procedure based on multiple coarse grained simulations ( Fig . 1 , Steps 1–5 ) followed by atomic-resolution ( Fig . 1 , Steps 6–7 ) simulations to study the structure and dynamics of CYP2C9 in a 1-palmitoyl-2-oleoylphosphatidylcholine ( POPC ) bilayer . POPC was chosen because it is one of the major components of the endoplasmic reticulum membrane ( >50% phosphatidylcholine [24] ) . An initial bias towards a specific orientation of CYP2C9 in the membrane was prevented by starting the coarse grained simulations with randomized orientations of the globular domain above or slightly inserted in the membrane . We simulated models derived from two conformations of CYP2C9 ( Table 1 ) , both based on the flurbiprofen-bound crystal structure ( pdb 1R9O ) [25] . 1R9OH1 and 1R9OH2 are full-length models while 1R9O1 and 1R9O2 represent soluble forms with the transmembrane helix removed . In the FG loop , which is thought to dip into the membrane , 1R9O1 and 1R9OH1 have two short helices ( F' and G' ) while 1R9O2 and 1R9OH2 have an extended structure . The nomenclature for the CYP secondary structure is from [26] . In the coarse grained simulations , the equilibration of the systems was reached rapidly ( ∼10–20 ns ) . The root mean square deviation of the globular domain of CYP2C9 relative to the initial structure reached an equilibrium value that varied between 2 . 8 and 3 . 5 Å depending on the simulated system ( Fig . S1 ) . To specify the orientation of the protein in the membrane , we defined three parameters: the distance between the centers of mass of the protein and the membrane ( d ) , and the two angles between the z axis and the vectors v1 ( α ) and v2 ( β ) , see Fig . 2 for definition . Angles α and β are suitable descriptors of the orientation of the protein in the membrane because their fluctuations due to internal protein dynamics are negligible ( Fig . S2 ) . From the coarse grained simulations of the membrane-bound 1R9OH1 and 1R9OH2 models , we found two predominant orientations of CYP2C9 in the bilayer depending on the conformation of the FG loop . The orientations are described by the following parameters: ( i ) d = 39 . 5±2 . 5 Å , α = 100±9° , β = 123±8° , and ( ii ) d = 40 . 9±3 . 0 Å , α = 76±9° , β = 117±11° ( Fig . 3 ) . 1R9OH1 inserted slightly further into the membrane and was oriented with the end of helix I closer to the lipid with v1 pointing towards the membrane ( α>90° ) ( Fig . 2A ) , while 1R9OH2 oriented with the beginning of helix I closer to the lipid and v1 pointing away from the membrane ( α<90° ) ( Fig . 2B ) . For each membrane-bound model , one snapshot with the orientation parameters ( d , α , β ) within 1% of the peak values of the corresponding histograms ( Fig . 3 ) was selected for atomic-resolution simulations . Atomic-resolution simulations were performed for product-bound and apo forms of the membrane-bound 1R9OH1 and 1R9OH2 and of the soluble 1R9O1 and 1R9O2 models ( Table 1 ) . In addition , a substrate-bound 1R9O2 model was simulated . The systems with soluble models of CYP2C9 were equilibrated for 2 . 5 ns , whereas the systems with membrane-bound models were equilibrated for 9 ns ( Fig . S3 ) . The orientations of CYP2C9 in the membrane were generally conserved , although small deviations from the starting configurations were observed ( Fig . 3 ) . The consistency with experiments was good ( Table 2 ) . The percentage of the catalytic domain in contact with the lipid bilayer was between 24–26% ( extended FG loop ) and 32–34% ( helical FG loop ) ( Table S1 ) . The helix A , the β1 sheet , and the FG loop were in contact with the lipid bilayer in all cases . The extended FG loop inserted deeper in the membrane than the F' and G' helices . The residues at the N-terminus of the G helix were located in the proximity of the lipid head groups except for in the simulations of the ligand-free 1R9OH2 model in which they inserted in the membrane . For this model , a decrease of d was observed during the atomic-resolution simulations ( Fig . 3C , red ) . The BC loop was accessible to the solvent except for residues 101–104 which were close to the lipid head groups , independent of the FG loop conformation . In the simulations of the structures derived from 1R9OH1 , the C-terminal region was in contact with the lipid head groups; this observation accounts for the higher percentage of residues contacting the membrane in this model . As expected , and not just because of the presence of the transmembrane helix , we found that hydrophobic aminoacids penetrate deeper in the lipid bilayer whereas charged protein residues interact preferentially with the lipid head group ( Fig . S4 ) . All residues interacting with the lipids are shown in Fig . S5 . Comparing the longer atomic-resolution simulations ( with γ = 60 dyn/cm ) of the membrane-bound models with the simulations of the soluble models , we found that the lipid bilayer exerted a limited influence on protein flexibility ( Fig . 4 ) . The most flexible regions were the BC , FG , GH , and HI loops . The fluctuations in the BC and FG loops , and the β1 were reduced by the presence of the lipid bilayer , while the flexibility of the GH and HI loops was not affected . This is consistent with the positions of these loops relative to the membrane . The FG loop is inserted in the membrane and some residues in the BC loop are in contact with lipids , while the GH and HI loops are not in the proximity of the membrane . The CYP2C9 models remained in closed or nearly-closed conformations in all the simulations . This is reflected in the low root mean square deviation values of the simulation snapshots relative to a reference structure chosen for each system after the initial energy minimization ( Fig . S6A , B ) . Large opening motions such as those identified from CYP2B4 crystal structures [17] , [27] were not observed . However , we observed opening-closing motions of discrete tunnels from the enzyme's buried active site . In the models of soluble CYP2C9 , the amplitude of these motions was higher when the FG loop was extended . The largest opening-closing motions were observed in the simulation of the product-bound 1R9O2 model and were attributed mainly to an increased flexibility of the BC loop ( Fig . S6D ) , whereas the substrate-bound 1R9O2 was the least flexible of the structures with an extended FG loop . Moreover , the product-bound form was the most flexible structure among the soluble models with a helical FG loop . In the simulation of this model , the F' and G' helices unfolded and the G helix moved away from the BC loop leading to a significant increase of the root mean square deviation of the FG loop ( Fig . S6E ) . In the simulations of the membrane-bound models , the ligand-free structures were the most flexible , independent of the FG loop conformation . The stability of the F' and G' helices in the FG loop was not affected by the interaction with the lipid bilayer . In fact , the F' and G' helices were less stable in the simulations of soluble models of CYP2C9 . The association with the membrane destabilized the interactions between L102 ( BC loop ) and I213 ( F' helix ) and I223 ( G' helix ) by favoring a motion of L102 away from this hydrophobic cluster . However , this did not result in a loss of stability due to new interactions formed by L102 with L234 and L208 which were not observed in the soluble models . Residues I215 ( F' helix ) and I222 ( G' helix ) interacted with the lipid tails . In the simulations of the soluble models with an extended FG loop , L102 interacted with I222 and I223 , whereas in the simulations of the membrane-bound models , these interactions were not stable and L102 interacted either with the lipid tails or with L208 and L233 whereas I213 , I215 , and I222 were anchored in the membrane . I223 was oriented towards the active site independent of the FG loop conformation . The flexibility of the BC loop observed in the simulations of both soluble and membrane-bound models resulted from the instability of the following interactions: ( i ) hydrophobic interactions between L102 and A103 and the FG loop , ( ii ) transient hydrogen bonds between R105 and E104 and H230 , N231 ( G helix ) and D224 ( FG loop , only in models with F' and G' helices ) , ( iii ) hydrophobic interactions between A106 and L234 and V237 ( G helix ) , and ( iv ) hydrogen bonds between the backbone of A106 and N107 and/or the side chain of N207 and K241 ( G helix ) and E285 ( I helix ) . Flexibility in the β1 sheet was a result of reorientation of F69 from the active site cavity to the protein surface . The dynamics of the interactions determining the positions of the BC loop and the β1 relative to the FG loop were similar in the simulations of soluble and membrane-bound models , except for those between L102 and the FG loop . The higher flexibility of the BC loop in the simulation of the product-bound 1R9O2 model was due to the loss of the hydrogen bonds between K241 and the BC loop and of the hydrophobic interactions involving A106 . We investigated the opening and closing motions of the tunnels 2a , 2b , 2c , 2ac , 2e , 2f , and the solvent ( S ) tunnel ( the nomenclature used is from [20] ) ( Fig . 5A ) . Each of these was found open in at least one snapshot of the simulations . In addition , these tunnels were identified as ligand exit routes in a separate set of simulations of soluble CYP2C9 models ( unpublished data ) . The entrances of the tunnels were defined as described in Table 3 ( see also Fig . S7 ) . The fraction of open states of each tunnel in each simulation was computed ( Fig . 5B , C ) . We identified three factors that influence the opening and closing motions in CYP2C9: ( i ) the conformation of the FG loop , ( ii ) ligand binding , and ( iii ) association with the membrane . Tunnel 2a is significantly more open in the models with an extended FG loop . The F' and G' helices restrict the opening of 2a , while favoring the opening of 2b which is mostly closed when the FG loop is extended . Tunnels 2c and S tended to be open more in the simulations of the product-bound models . This tendency was independent of the conformation of the FG loop and the association with the membrane in the case of S whereas for 2c it was evident only in the simulations of the soluble models . 2c also opened in the shorter simulation ( with γ = 50 dyn/cm ) of the apo form of the membrane-bound 1R9OH1 model . The interaction with membrane favored the opening of S in the models with a helical FG loop and restricted it in the other models . Remarkably , the opening of 2c and S correlates very well with the closing of 2a and/or 2b ( Fig . 6 and S8 ) . In particular , we observed the closing of 2a and opening of 2c after about 18–20 ns in the simulation of the product-bound , soluble 1R9O2 model ( Fig . 6C ) . Moreover , this correlation was observed when comparing the simulations of the product-bound and ligand-free structures of the 1R9O1 model ( Fig . S8A , B ) . The lipid bilayer favored the opening of 2a regardless of the conformation of the FG loop and or whether a ligand was bound . However , the amplitude of the opening of 2a was larger in the models with an extended conformation of the FG loop . In the simulations of these models , a POPC molecule was observed to penetrate into 2a with its choline methyl groups interacting with the protein residues L208 , F69 , F100 , and W212 . The opening of 2e was in general restricted by the presence of the lipids except that it was also closed in the simulation of the soluble , product-bound 1R9O1 model . Tunnel 2f was mostly closed in all the simulations , while opening of 2ac appeared to correlate either with the opening of 2a in the models with an extended FG loop ( Fig . 6C , D ) or with the opening of 2c in the models with F' and G' helices ( Fig . S8B ) . The opening of 2a , as well as the closing of 2e , was correlated with the opening of the internal aromatic gate formed by F100 , F114 , and F476 . In addition , opening of 2a was facilitated by the loosening of the cluster of hydrophobic interactions around L102 . 2c opened when the hydrogen bonds between K241 ( G helix ) and the backbone of the BC loop and the hydrophobic interactions between A106 ( BC loop ) and L234 and V237 ( G helix ) were disrupted . As in all other simulations , R108 , which interacts directly with the substrate , kept its orientation towards the inside of the active site and the hydrogen bonding pattern between R108 , N289 , and D293 was maintained . In the 1R9O crystal structure , the aromatic gate is closed and locks the substrate in the bound position above the heme center [25] . In the simulations , dynamics of the gate regulate tunnel opening motions . To calculate the percentage of each trajectory in which the gate is open , we monitored defined distances between the phenyl rings ( Fig . 7A , B ) . We found that the association with the membrane stabilized an open conformation of the gate independent of the conformation of the FG loop ( Fig . 7C , E ) . F100 adopted two positions when the gate was open , either close to F114 blocking the entrance of 2e ( observed mostly in simulations of models with a helical FG loop ) or close to F69 lining the entrance of 2a ( observed mostly in simulations of models with an extended FG loop ) . The amplitude of the gating motions was estimated by projecting the center of the phenyl rings on the heme plane and monitoring the area covered during the simulations: the greater the area , the more mobile the phenyl ring ( Fig . S10 ) . We found that in the simulations of the models with a helical FG loop , the motion of F114 was restricted by the bound product , the motion of F476 was restricted both by the bound product and by the interaction with the membrane , and the motion of F100 was not affected ( Fig . 7D ) . These observations agree well with the position of F476 near the head group region of the lipid in these models , but not with the position of F100 in the simulation of the product-bound 1R9OH1 model ( Fig . S9A , B ) . In the simulations of the models with an extended FG loop , the interaction with the membrane restricted the motion of F100 ( Fig . 7E ) , while the motions of F114 and F476 were not affected . This is consistent with the position of F100 in the proximity of the lipid head groups in these models ( Fig . S9C , D ) .
We have described a procedure to derive models of a CYP-membrane bilayer complex in an unbiased fashion that , unlike previous approaches to model CYP-membrane complexes , does not rely on input of any experimental data on the CYP-membrane interactions . From multiple coarse grained molecular simulations started with the globular domain in random orientations above or loosely inserted in the membrane and the statistical analysis of defined parameters , two predominant orientations of CYP2C9 in the membrane were found that were both equally consistent with available experimental data . These orientations differed in the conformation of the FG loop . The two protein-membrane configurations generated from coarse grained simulations were conserved in atomic-resolution simulations with only minor adjustments in protein orientation . The main difficulty in the overall procedure for the deriving the CYP-membrane models is the generation of the initial system for the coarse grained simulations . We found that the pre-assembly of the lipid bilayer and the pre-simulation of the transmembrane helix inserted in the membrane were required . In test simulations of self-assembly [28] , [29] of CYP2C9-membrane complexes , the N-terminal helix did not assembled in a transmembrane configuration ( data not shown ) . The coarse grained force field does not maintain protein tertiary structure unless it is complemented by the application of an elastic network model to connect the Cα atoms [30] , [31] . Thus the region restrained by the elastic network must be chosen carefully . In initial simulations with all residues observed in the crystal structure of CYP2C9 included in the elastic network model , the transmembrane configuration of the N-terminal helix of CYP2C9 was not stable ( data not shown ) . Hence , we detached the unstructured region of 14 residues adjacent to the transmembrane helix from the elastic network and generated initial configurations by arbitrary modification of the backbone parameters in this region . It is difficult to achieve a full sampling of the conformations of this linker and , given the limitations of the coarse grained force field , to ensure that the conformations are plausible . Therefore , it is possible that the linker is trapped in a suboptimal local minimum after the atomic-resolution simulations . This limitation might be overcome in future applications of this procedure to generate protein-membrane complexes for other proteins with single helix membrane anchors by improvements in coarse grained force fields and the application of an iterative approach to gradually detach residues from the elastic network . We found that the CYP2C9-membrane complexes were largely consistent with experiments . However , the comparison was limited by inconsistencies among different experiments and the diversity of CYP isoforms used in the experiments . The three main inconsistencies between our simulations and the experiments were: ( i ) the β1 sheet was inserted in the lipid bilayer despite microsomes being recognized by a site-specific antibody against the corresponding region in CYP2B1 [8] , ( ii ) the coil region between the B helix and the BC loop was accessible to solvent despite experimental evidence for the association with membrane of the corresponding region in CYP2B1 [8] , ( iii ) L380 was found in the region of the lipid head groups in the simulations whereas it was inferred to insert deepest in the membrane from tryptophan fluorescence quenching experiments of CYP2C2 [12] . Major reorientation of the protein would be required to improve the consistency with these experiments . However , this would diminish the consistency with other experiments and would impair the interaction with the CPR . A putative orientation with the heme tilt angle at nearly 90° , similar to the configurations proposed based on the crystal structures of CYP2C5 and CYP2C8 [13] , [14] . would be consistent with the proposed position of L380 as the deepest in the membrane [12] . However , we did not observe such a configuration in our simulations . Furthermore , when we placed the protein in the membrane in such an orientation , we found that the model was not stable in atomic-resolution simulations ( data not shown ) . We cannot exclude that CYPs undergo large rearrangements in the membrane and adopt conformations that may show greater consistency with experiments , however it is apparent that , with the conformations sampled , it is not possible for a CYP2C9 orientation in the membrane to simultaneously satisfy all experimental data . In the models derived here the surface proposed to interact with the CPR [32] was exposed to the solvent . We predict that both orientations of CYP2C9 in the membrane are favorable for electron transfer from the CPR . CPR undergoes a major conformational rearrangement to achieve a conformation suitable for electron transfer to CYP [33] . We anticipate that the amplitude of this conformational transition in CPR is larger , and therefore more energetically costly , when the CYP is more tilted with respect to the lipid bilayer ( i . e . has a lower value of the heme tilt angle ) . This implies that the electron transfer between CPR and CYP might be more efficient when the CYP is less tilted with respect to the lipid bilayer . This proposal could be investigated by extending the simulation protocol described to include CPR in the models . We used a membrane with a size chosen to permit this extension . CYP2C9 remained in closed or nearly closed conformations in the simulations . We did not observe open conformations similar to those reported in crystal structures of other CYPs [17] , [19] , [27] or in a recent molecular dynamics study of soluble CYP2C9 [34] . Nevertheless , we identified motions that lead to opening of discrete tunnels from the enzyme's buried active site . Remarkably , we found that the opening and closing of the different tunnels were correlated and depended not only on the association with the lipid bilayer but also on ligand-binding and the conformation of the FG loop . The tunnels 2c and S are accessible from the cytosol and opened preferentially in product-bound models of CYP2C9 , suggesting that they are the preferred release tunnels for soluble products from CYP2C9 . Both have been proposed to function as ligand exit routes in other CYPs [6] , [21] , [22] . The opening of tunnel 2a leading into the membrane was facilitated by the interaction of CYP2C9 with the bilayer and by the extended conformation of the FG loop . A role for this tunnel in mammalian CYPs has not been defined . We propose that 2a is the preferred substrate access tunnel from the lipid bilayer for liposoluble compounds . This proposal is supported by the correlation observed between the opening of 2a and the closing of 2c and S . 2a was previously described as the main ligand egress tunnel in three bacterial P450s which are soluble enzymes [35] . Additional evidence for substrate access from the membrane is provided by Berka et al . ( unpublished data ) who show that the preferred position of ibuprofen , a drug metabolized by CYP2C9 , is near the lipid tails . Moreover , a charge-modifying sequence variation between CYP2C9 and CYP2C19 , a closely related CYP that does not have the specificity for acidic compounds that CYP2C9 does , is found around the entrance of 2a [36] . Despite this evidence for selective roles of different tunnels in CYP2C9 , we cannot exclude that each tunnel is used both for substrate access and product release . Tunnel 2b appears to be an alternative for 2a in the CYP2C9 models with a helical FG loop . The opening of 2b was facilitated by the F' helix blocking the entrance of 2a . However , 2b did not lead into the membrane and it remained closed whenever a larger opening of 2a was observed . 2b was also found closed in the structure of CYP2B4 in which 2a was wide open [17] . Therefore , we suggest that in the ER membrane , the opening of 2a is sufficiently large to permit ligand passage and therefore 2b does not have an obvious function . We found that 2f , an alternative tunnel pointing into the membrane , was closed in the simulations . In the case of a wider opening of CYP2C9 , 2f could merge with 2a to form a large access channel from the lipid bilayer . From the wealth of data on substrate access and product release in mammalian CYPs , it becomes apparent that there are multiple routes which are used by different CYPs depending on the ligand properties . If proven , this hypothesis may provide new strategies for the inhibition of CYPs . Mechanistically , we found that the opening of 2a was facilitated by the opening of the internal aromatic gate formed by F100 , F114 , and F476 . F114 and F476 were showed to be required for the CYP2C9 activity , whereas mutations of F100 do not affect catalysis [37] . This finding apparently contradicts the evidence from the structure of the flurbiprofen-bound complex where F100 was found to lock the substrate in the catalytically productive position above the heme . In addition , we found extensive interactions between different ligands and F100 during ligand exit simulations of soluble models of CYP2C9 ( unpublished data ) . The experiments were performed in the presence of dilauroylphosphatidylcholine ( DLPC ) ; therefore it is more appropriate to compare them with the simulations of the membrane-bound CYP2C9 . We found that the membrane stabilizes open conformations of the aromatic gate with F100 further away from the heme center . Thus , we suggest that the closed state of the gate may not be crucial for the catalytic activity of CYP2C9 in the lipid bilayer . In conclusion , we applied a procedure to model and simulate the membrane bound CYP2C9 that is not biased towards a certain orientation of the protein in the membrane . The protocol may be applied to model other mammalian CYPs , their interaction with CPR or , in general , to proteins anchored in the membrane by a single transmembrane helix . For CYP2C9 , we show that the lipid bilayer may play an important role in the selection of different tunnels for ligand exchange between the environment and the enzyme's buried active site and suggest selective roles for different tunnels .
We constructed 2 models of soluble CYP2C9 based on the flurbiprofen-bound crystal structure ( pdbid 1R9O , 1 . 9 Å resolution ) [25]: ( i ) 1R9O1 has 2 small helices in the FG loop ( F' and G' ) , and ( ii ) 1R9O2 has a small β-sheet in the mostly unstructured FG loop and is similar to an unpublished crystal structure of CYP2C9 ( Eric Johnson , personal communication ) . Details of the model building procedure are given in the Supporting Information ( Protocol S1 ) . The procedure used to model the membrane-bound CYP2C9 is outlined in Fig . 1 . We predicted ( using PSIPRED [38] and TMHMM [39] ) that residues 1 to 22 adopt a helical conformation and built the corresponding ideal α-helix with MODELLER [40] . Residues 23 to 25 connecting the last residue of the N-terminal transmembrane helix and the first residue in the crystal structure were modelled in a random-coil conformation . The 25-residue long N-terminal peptide was then converted to a coarse grained representation ( coarse grained particles were placed at the average positions of their corresponding non-hydrogen atoms ) and inserted in a pre-equilibrated coarse grained model of a bilayer composed of 608 POPC molecules ( Fig . 1 , Step 1 ) . The bilayer was chosen to be large enough to allow the insertion of CYP2C9 and the CPR . The POPC molecules that had particles within 3 Å of particles of the peptide were removed . The resulting N-terminal peptide-bilayer system was simulated for 3 µs ( in a single simulation ) as described below ( Fig . 1 , Step 2 ) . The inclination of the transmembrane helix reached an equilibrium value of 12±6° . Two complete coarse grained models of the membrane-bound CYP2C9 ( 1R9OH1 and 1R9OH2 ) were built by attaching the globular domain ( models 1R9O1 and 1R9O2 pre-converted to a coarse grained representation ) to the final structure of the simulated N-terminal peptide . The coarse grained parameters for POPC were taken from [41] , and for the protein from [30] , [42] . An elastic network model connecting the protein backbone particles closer than 7 Å with a spring constant k = 10 . 75 kcal*mol-1 Å-2 was applied to maintain the protein secondary and tertiary structure during the simulations [30] , [31] . Coarse grained parameters for the heme group are not available . Therefore , we decided to remove the heme from the protein during the coarse grained simulations . Because the heme is buried in the protein , it will not affect the CYP-membrane interaction and its role in the maintenance of the protein tertiary structure is mimicked by the elastic network model . The simulations were performed with the GROMACS 4 program [43] under periodic boundary conditions , in coarse grained water , in the NPT ensemble using the Berendsen weak coupling algorithm [44] to maintain the temperature at 300 K and the pressure at 1 atm . The time step was 20 fs . The non-bonded ( van der Waals and electrostatic ) interactions were evaluated with a cutoff of 14 Å . The dielectric constant was 20 . Semi-isotropic pressure scaling ( with relaxation time τp = 2 ps in all three space directions ) was used to account for fluctuations of the membrane size . An initial set of 2 simulations , each of 3 µs , was performed with the membrane-bound 1R9OH1 and 1R9OH2 models during which the N-terminal helix adopted an orientation parallel to the head groups at the membrane surface ( data not shown ) . The transmembrane configuration of the helix was stable only when residues 23 to 37 in the linker peptide were decoupled from the elastic network model , thus allowing different backbone configurations in the linker peptide . 7 1R9OH1 and 5 1R9OH2 structures with different orientations of the globular domain with respect to the bilayer were constructed by randomly modifying the dihedral angles in the linker peptide backbone ( Fig . 1 , Step 3 ) . Each of these 12 structures was used as the starting structure for a 1 µs simulation during which snapshots were recorded every 0 . 4 ns ( Fig . 1 , Step 4 ) . The protein orientation with respect to the lipid bilayer was evaluated by considering all the snapshots recorded in the 12 simulations except for the first 100 ns of each simulation . The conversion protocol for the membrane was similar to that described by [45] . However , we implemented it in VMD [46] and adapted it to be compatible with the generalized amber force field ( gaff ) for POPC lipids [47] which we used in the atomic resolution simulations . The rmsd between the initial coarse-grained and the converted atomic resolution POPC molecules was less than 1 . 5 Å ( Fig . S11 ) , see Protocol S2 in the Supporting Information for details of the procedure and the files necessary for its application . The globular domain ( residues 37–492 ) of CYP2C9 was converted by superimposing the atomic resolution models 1R9O1 and 1R9O2 ( taken directly after the model building procedure ) on the selected coarse grained models of the membrane-bound CYP2C9 . This ensured that the atomic resolution simulations of the membrane-bound and soluble CYP2C9 were started with the same initial models of the globular domain . The transmembrane helix ( residues 1–22 ) was converted by superimposing the ideal α-helix on the selected coarse grained models . The linker peptide ( residues 23–36 ) was converted with a simulated annealing protocol ( Protocol S1 ) in which the positions of the Cα atoms were restrained to the positions of the coarse grained backbone particles ( Fig . 1 , Step 6 ) . Atomic-resolution simulations ( Fig . 1 , Step 7 ) were performed with the following models: ( i ) 1R9O1 , ( ii ) 1R9O1 + FLO , ( iii ) 1R9O2 , ( iv ) 1R9O2 + FLU , ( v ) 1R9O2 + FLO , ( vi ) 1R9OH1 + POPC , ( vii ) 1R9OH1 + FLO + POPC , ( viii ) 1R9OH2 + POPC , ( ix ) 1R9OH2 + FLO + POPC . FLU = flurbiprofen ( substrate ) and FLO = 4-hydroxy-flurbiprofen ( product ) . The first 5 models are soluble truncates of CYP2C9 while the last 4 are atomic-resolution models of the membrane-bound CYP2C9 . In the models of the apo form of the protein , 5 water molecules were placed in positions originally occupied by the ligand . The simulations were performed with NAMD [48] in explicit solvent in 150 mM NaCl under periodic boundary conditions . The Particle Mesh Ewald method was used to evaluate the electrostatic interactions . The “ff99sb” version of the AMBER force field [49] was used for the protein and gaff was used for the lipids [47] . The ion parameters were taken from [50] . The partial atomic charges for the ligands were derived using the procedure of restrained fit to the Hartree-Fock 6-31G* electrostatic potential ( RESP ) with the RED program . The heme parameters were provided by D . Harris with the partial atomic charges derived from DFT calculations [51] . The RATTLE algorithm was used to constrain the bonds to hydrogens . The temperature was controlled with Langevin dynamics ( damping coefficient 1≤ γT ≤5 ps−1 during equilibration and γT = 0 . 5 ps−1 during production ) . Pressure was controlled with the Nose-Hoover Langevin Piston method ( oscillation period ω = 100 fs and damping coefficient γP = 50 fs−1 during equilibration and ω = γP = 1000 fs during production ) . The time step was 1 fs during equilibration and 1 . 5 fs during production . The systems were equilibrated as described in the Supporting Information ( Protocol S3 ) and then production runs were carried out as summarized in Table 1 . For snapshots at 150 ps intervals along each simulation , we calculated 10 tunnels with MOLE [52] and saved 200 segments and their radii along each tunnel . Each computed tunnel was assigned to one of the seven tunnels listed in Table 3 if at least one segment of the tunnel was less than 5 Å from the entrance of the tunnel and the entrances of the other six tunnels were further away from each segment of the tunnel . The assignments were accurate because the distance between any two tunnel entrances was larger than the maximum displacement of each entrance due to internal protein motions ( Table S2 ) . If the minimum radius of the segments along a tunnel was greater than 1 . 2 Å , the tunnel was considered open to permit the passage of at least 1 water molecule . This value was chosen to be slightly lower than the usual probe radius of water of 1 . 4 Å because the radii calculated with MOLE are smaller than the actual radii of the tunnels . The procedure was implemented in VMD [46] .
|
We describe the first atomic-detail models and simulations of a full-length , membrane-bound mammalian cytochrome P450 . To date , all the structural studies of microsomal , drug-metabolizing cytochrome P450s have been performed using engineered , solubilized forms of the enzymes and it is not yet understood how the membrane influences their structure , dynamics , and ability to bind substrates . We focused on CYP2C9 , the second most abundant cytochrome P450 in the human liver which oxidizes 20% of all marketed drugs . Here , we have derived models of CYP2C9-membrane complexes with a modeling procedure based on molecular dynamics simulations started with arbitrary configurations of the protein in the membrane and performed using both coarse grained and atomic-detail molecular representations . This procedure is expected to be generally applicable to proteins that are anchored in the membrane with a single transmembrane helix . The models and simulations provide evidence for selective substrate access and product release routes in membrane-bound CYPs . This observation may contribute to new strategies to manipulate the activity of CYPs and other enzymes with buried active sites . We anticipate that this study will bring about a paradigm shift towards studying microsomal CYPs as dynamic structures in their natural , lipid environment rather than in artificially solubilized forms .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biology",
"computational",
"biology"
] |
2011
|
Structure and Dynamics of the Membrane-Bound Cytochrome P450 2C9
|
Eradication of tuberculosis ( TB ) , the world's leading cause of death due to infectious disease , requires a highly efficacious TB vaccine . Many TB vaccine candidates are in pre-clinical and clinical development but only a few can be advanced to large-scale efficacy trials due to limited global resources . We aimed to perform a statistically rigorous comparison of the antigen-specific T cell responses induced by six novel TB vaccine candidates and the only licensed TB vaccine , Bacillus Calmette-Guérin ( BCG ) . We propose that the antigen-specific immune response induced by such vaccines provides an objective , data-driven basis for prioritisation of vaccine candidates for efficacy testing . We analyzed frequencies of antigen-specific CD4 and CD8 T cells expressing IFNγ , IL-2 , TNF and/or IL-17 from adolescents or adults , with or without Mycobacterium tuberculosis ( M . tb ) infection , who received MVA85A , AERAS-402 , H1:IC31 , H56:IC31 , M72/AS01E , ID93+GLA-SE or BCG . Two key response characteristics were analyzed , namely response magnitude and cytokine co-expression profile of the memory T cell response that persisted above the pre-vaccination response to the final study visit in each trial . All vaccines preferentially induced antigen-specific CD4 T cell responses expressing Th1 cytokines; levels of IL-17-expressing cells were low or not detected . In M . tb-uninfected and -infected individuals , M72/AS01E induced higher memory Th1 cytokine-expressing CD4 T cell responses than other novel vaccine candidates . Cytokine co-expression profiles of memory CD4 T cells induced by different novel vaccine candidates were alike . Our study suggests that the T cell response feature which most differentiated between the TB vaccine candidates was response magnitude , whilst functional profiles suggested a lack of response diversity . Since M72/AS01E induced the highest memory CD4 T cell response it demonstrated the best vaccine take . In the absence of immunological correlates of protection , the likelihood of finding a protective vaccine by empirical testing of candidates may be increased by the addition of candidates that induce distinct immune characteristics .
Tuberculosis ( TB ) is an infectious disease of major global importance . An estimated ten million people were diagnosed with TB disease in 2017 and 1 . 6 million died [1] . Current efforts to curb the TB epidemic are insufficient to achieve the 2035 targets set by the World Health Organization , of a 95% reduction in TB deaths and a 90% reduction in the TB incidence rate , compared with levels in 2015 [1] . There is widespread consensus , supported by epidemiological modeling , that a highly efficacious TB vaccine is necessary to achieve these TB control objectives [2 , 3] . Thirteen novel TB vaccine candidates were being assessed at various stages in phase 1–3 clinical trials in 2017 . However , whilst criteria have been proposed to guide advancement of vaccine candidates to efficacy testing , these are neither unanimously agreed upon nor used [4] . There is also limited stakeholder effort to harmonize or standardize clinical trial design and , as a result , different vaccine candidates are typically assessed in unrelated trials with unique design features that preclude direct comparison of results . This is also the case for assessment of immunological outcomes of TB vaccine trials . Most clinical trials measure vaccine-induced CD4 and/or CD8 cells expressing the Th1 cytokines IFNγ , TNF and IL-2 , on the basis that these cells are necessary , although not sufficient , for protective immunity against Mycobacterium tuberculosis ( M . tb ) in animal models and humans ( reviewed in [5 , 6] ) . However , the types of assays , methodologies and protocols employed by different investigators to measure Th1-cytokine expressing T cells vary considerably [7] . Direct comparison of the magnitude , character and durability of antigen-specific immune responses induced by different TB vaccine candidates is therefore highly problematic . We propose that this is an important gap in knowledge required to guide advancement of vaccine candidates through the clinical development pipeline . To facilitate rational , data-driven decisions about vaccine candidate advancement , we compared Bacillus Calmette-Guérin ( BCG ) [8] and six novel TB vaccine candidates , including MVA85A [9 , 10] , AERAS-402 [11] , H1:IC31 [12] , M72/AS01E [13 , 14] , ID93+GLA-SE [15] and H56:IC31 [16] , by their induced antigen-specific CD4 and CD8 T cell responses from data generated in human clinical trials previously completed at the South African TB Vaccine Initiative . The immune responses were measured using the same immunological assay [17 , 18] , enabling direct comparisons between vaccines . We aimed to define group ( s ) of vaccines that induced distinct immune responses . Within a group , the vaccines would induce similar immune responses , thus motivating further testing for only one vaccine per group by applying an objective , data-based criterion for vaccine prioritisation . We identified appropriate statistical approaches and performed an analysis of antigen-specific T cell responses to each antigen in the different vaccine candidates . Comparing the magnitude and cytokine co-expression profiles of vaccine-induced memory CD4 and CD8 T cell responses that persisted to the final study visit in each trial , we revealed considerable homogeneity in vaccine-induced Th1 memory response profiles , particularly in individuals with underlying M . tb infection . Our study provides a framework for interpreting immunological characteristics that may be useful for prioritization of vaccine candidates for advancement through the development pipeline .
Many individuals in countries endemic for TB , such as South Africa , are immunologically sensitized to mycobacteria due to a combination of infant BCG vaccination , exposure to environmental non-tuberculous mycobacteria and/or natural M . tb infection [20] . To investigate effects of this on pre-vaccination T cell responses , we examined frequencies of Th1-cytokine expressing T cells specific to antigens in each vaccine . CD4 T cell responses to all vaccine antigens were low in M . tb-uninfected participants , although responses to antigens in H56:IC31 , H1:IC31 and M72/AS01E were detected at frequencies significantly higher than 0 . 005% , which we defined as the positive response criterion ( Fig 1A and S1 Fig ) . By contrast , pre-vaccination CD4 responses to antigens in AERAS-402 , MVA85A and ID93+GLA-SE were not significantly larger than 0 . 005% . In M . tb-infected persons , Th1 responses to antigens in all vaccines except MVA85A were detectable at frequencies significantly higher than 0 . 005% , although the highest responses were observed for BCG and the megapool ( Fig 1A ) . The pre-vaccination responses were higher than in M . tb-uninfected individuals , but only the difference for M72/AS01E was significant ( S2 Fig ) . CD4 T cell responses to Ag85A were not detectable at frequencies significantly exceeding 0 . 005% ( S1A Fig ) . Pre-vaccination CD8 T cell responses to all vaccine antigens as well as the M . tb megapool were low and did not significantly exceed 0 . 005% for any vaccine in either M . tb-uninfected or -infected participants ( Fig 1B and S1B Fig ) . Next , we examined longitudinal frequencies of antigen-specific CD4 and CD8 T cell responses induced by each of the vaccines ( Fig 2 ) , measured at various time points in each trial ( Table 2 ) . Vaccine-induced CD4 T cell responses were higher than CD8 T cell responses . In addition , as expected and previously described [8–15 , 21] , these responses typically peaked at the measurement timepoint immediately after vaccine administration and waned thereafter . Inter-donor variability in CD4 T cell responses to M72/AS01E , H1:IC31 and BCG was high , especially during the early effector phases of the response kinetics ( Fig 2 ) . To determine which vaccines induced durable T cell responses and compare the magnitude of these responses , we examined vaccine-induced memory CD4 and CD8 T cell response frequencies . The vaccine-induced memory response is the difference between frequencies of antigen-specific Th1-cytokine producing cells at the final time point and the pre-vaccination timepoint in each trial . Among novel vaccine candidates , M72/AS01E , ID93+GLA-SE and H1:IC31 induced CD4 responses that were significantly higher than pre-vaccination levels in both M . tb-uninfected and -infected individuals , with M72/AS01E inducing greater memory responses than the other novel vaccine candidates ( Fig 3A and S3 Fig ) . H56:IC31 induced responses that persisted at levels above those observed pre-vaccination only in M . tb-uninfected individuals , and no durable response was detected for either AERAS-402 or MVA85A . BCG induced a highly variable response , which also persisted at levels above those observed pre-vaccination in M . tb-infected individuals . There was negligible statistical support for an effect of underlying M . tb infection status on the induced memory CD4 T cell response ( S4 Fig ) . We also evaluated memory CD4 T cell responses to each individual antigen in each vaccine candidate ( S5A Fig ) . Memory CD4 responses induced by H1:IC31 and H56:IC31 in M . tb-uninfected individuals primarily comprised Ag85B-specific CD4 T cells , whilst a memory response to Rv2608 was predominant in the ID93+GLA-SE-induced response ( S5A Fig ) . Vaccine-induced memory CD4 T cell responses to most other individual antigens , except for Mtb32 and Mtb39 ( which were not tested separately ) , were not detected . M72/AS01E induced low but durable memory CD8 T cell responses in M . tb-uninfected and -infected individuals . No durable memory CD8 T cell responses were detected for the other vaccines ( Fig 3B ) . However , the sample estimates of CD8 T cell responses induced by Aeras402 and BCG were higher than M72/AS01E , although these differences were not significant . Results were similar for memory CD8 T cell responses to each individual antigen in each vaccine candidate; only the combined CD8 response to Mtb32 and Mtb39 significantly exceeded the pre-vaccination response ( S5B Fig ) . Frequencies of vaccine-induced CD4 and CD8 T cells producing IL-17 were low across vaccines , and usually not significantly higher than pre-vaccination levels ( S6 Fig ) . An important feature of T cell responses to pathogens , including M . tb , is the cytokine co-expression profile of CD4 T cells , which may reflect the degree of T cell differentiation or quality of the response [6 , 22 , 23] . We analyzed cytokine co-expression profiles of the memory CD4 T cell responses induced by the vaccines . We did not include the cytokine co-expression profiles of the vaccine-induced memory CD8 T cell responses , because the magnitudes of the vaccine-induced CD8 memory response were either very low or not significant . We analyzed cytokine co-expression profiles using principal components analysis ( PCA ) biplots . In addition , 95% bivariate confidence areas of the mean response for each vaccine were computed using bootstrapping for the first two principal components of the vaccine-induced memory response . In M . tb-uninfected individuals , the responses induced by the two viral-vectored vaccines , AERAS-402 and MVA85A , were distinct . This suggests different cytokine co-expression profiles from the four sub-unit vaccines , namely H56:IC31 , H1:IC31 , M72/AS01E and ID93+GLA-SE . These grouped together regardless of their antigens ( Fig 4A ) . IFNγ+ TNF+IL-2+ , TNF+IL-2+ and IL-2+ CD4 T cells drove most variation in the first two principal components ( S7A Fig ) . The biplot axes ( Fig 4A ) show that MVA85A vaccination was characterized by higher IFNγ+TNF+IL-2+ CD4 T cell responses , the protein sub-unit vaccines by higher TNF+IL-2+ and IL-2+ CD4 T cells , and AERAS-402 by low responses for all cytokine-expressing subsets . In M . tb-infected individuals , CD4 T cell responses induced by all novel TB vaccines grouped together ( Fig 4B ) . Furthermore , the first two principal components captured 77% of the variation , indicating that most variation was due to donor response variability rather than differences in response profiles between vaccines . Variability was largest for IFNγ+TNF+IL-2+ and IFNγ+ CD4 T cell responses ( S7B Fig ) . The response to megapool stimulation , induced by M . tb infection , was characterized by a predominance of IFNγ+TNF+IL-2+ polyfunctional CD4 T cells , whereas the BCG-induced response comprised low proportions of polyfunctional CD4 T cells; responses induced by the novel TB vaccines fell in between the megapool and BCG responses . Univariate analysis of the responses by cytokine combination for each vaccine corroborated these findings ( Fig 5 and S8 Fig ) . These figures show that any differences observed in the PCA biplots corresponded to differences at a univariate level , and that the PCA biplots did not ignore cytokine combinations that discriminated between vaccines . We also assessed the effect of underlying M . tb infection on cytokine co-expression profile ( S9 Fig ) . We examined in a univariate analysis differences in scaled vaccine-induced memory responses of TNF+IL-2+ , IFNγ+ CD4 T cells and IFNγ+TNF+IL-2+ polyfunctional CD4 T cells between M . tb-uninfected and -infected persons . The results suggest that the uniformity of response profiles observed in M . tb-infected individuals was driven by the protein sub-unit vaccines inducing less TNF+IL-2+ CD4 T cells , and possibly more IFNγ+TNF+IL-2+ CD4 T cells , in M . tb-infected individuals . Taken together , these analyses show that the six novel vaccine candidates induced similar CD4 memory response profiles , exacerbated by a reduction in TNF+IL-2+ CD4 T cells in persons with underlying M . tb infection .
TB vaccine development has substantially progressed in recent years , with great advances in our understanding of vaccine platforms , antigen and adjuvant selection and correlates of risk of TB disease [4] . Animal models have been standardized to enable head-to-head comparison and selection of candidate TB vaccines for advancement to phase I clinical trials [4] . Thirteen TB vaccine candidates were assessed in clinical trials in 2017 [4] , but only one or two can be advanced to large-scale efficacy trials due to limited global resources . To provide a data-driven basis for selection of vaccine candidates for further testing in efficacy trials , we performed a comparison of antigen-specific T cell responses induced by six novel TB vaccine candidates that have been assessed in phase 1b or 2a trials at SATVI . Three major points emerged from our study: ( 1 ) Antigen-specific T cell responses induced by the candidate TB vaccines were strongly CD4 T cell biased and predominantly expressed Th1-cytokines , ( 2 ) Th1 cytokine co-expression profiles of vaccine-induced memory CD4 T cells , a feature of T cell differentiation and functional quality , demonstrated considerable homogeneity between the vaccine candidates , ( 3 ) Analysis of T cell response magnitudes showed that amongst the novel vaccine candidates , M72/AS01E induced the highest memory cytokine-expressing CD4 T cell responses . Our finding that vaccine-induced responses were strongly CD4 T cell-biased with little IL-17 production , and that Th1-cytokine expression profiles were similar across vaccine candidates , highlights a lack of diversity in immunological responses typically analysed in TB vaccine immunogenicity assessments . This result is perhaps not surprising given that most current TB vaccine candidates were designed to specifically target induction of IFNγ-expressing CD4 T cells , predicated on the well-established evidence that Th1 cells are necessary , although not sufficient , for protective immunity against M . tb , based on animal models and human studies ( reviewed in [5 , 6 , 24] ) . We acknowledge that analysis of Th1-cytokine and IL-17 expressing CD4 and CD8 T cell responses may miss important T cell functions . Further , assays that can detect alternative T cell functions or outcomes to the ones we measured , such as proliferative potential or cytotoxic function , may have revealed diversity in immunological responses that were missed by our analyses . We did not include analysis of immunogenicity data from trials performed in age groups other than adults or adolescents , or from trials performed in current or prior TB patients . We also excluded data for other vaccine candidates assessed in clinical trials at SATVI , such as MTBVAC , VPM1002 and H4:IC31 , because data from vaccinated adults or adolescents at the end of study time point were not available . The whole live mycobacterial vaccines , MTBVAC and VPM1002 , are known to induce responses by a broader range of immune cells [25 , 26] than those we observed and might add to the diversity of the immune responses induced by vaccine candidates . Further , we did not include analyses of antigen-specific antibody responses , which may also be important in immunity against M . tb [27 , 28] , largely because antibody responses were not measured in each trial assessed here . It should be noted that high-level antigen-specific IgG responses were induced by a number of these TB vaccine candidates [14 , 15] and we suggest that such responses should be measured in vaccine trials and included in future head-to-head comparisons . Our study revealed negligible evidence of an effect of underlying M . tb infection on the vaccine-induced memory response magnitude of CD4 or CD8 T cells for any vaccine , but strong evidence for an effect of M . tb infection on the vaccine-induced memory response cytokine co-expression profile . For CD4 T cells , M . tb infection was associated with a reduction in TNF+IL-2+ and possibly IL-2+ CD4 T cells , which corresponded to an increase in IFNγ+ CD4 T cells for M72/AS01E and possibly IFNγ+TNF+IL-2+ for all novel vaccine candidates . The net effect was that the response profiles induced by MVA85A and the protein sub-unit vaccines in M . tb-infected individuals were similar . This drove the response closer to that induced by M . tb infection , as detected by megapool stimulation . These data suggest that underlying M . tb infection can play a strong role in the character of the vaccine-induced T cell response , as noted in published vaccine trials [12 , 14 , 21] . Since the vaccine-induced memory Th1 cytokine co-expression profiles were similar , only the response magnitude separated vaccine candidates . M72/AS01E induced the largest antigen-specific CD4 T cell responses , with similar responses between other novel vaccine candidates . Therefore , based on CD4 T cell response magnitude , our study suggests that M72/AS01E demonstrated the best vaccine take , providing support for further clinical testing . Non-immunological differences between vaccines , such as manufacturing cost , potential production capacity and ease of logistical arrangements , are also critical factors in decisions of candidate selection for further testing . It is important to note that measures of antigen-specific T cell responses , such as the ones we analyzed , do not represent known correlates of protection against M . tb . Since immune correlates of protection against M . tb remain undefined [4 , 6 , 27] , the induced CD4 and CD8 T cell response producing Th1 cytokines can only be considered a measure of vaccine take . The recent demonstration of protection afforded by BCG re-vaccination against sustained M . tb-infection [29] presents an opportunity for elucidating immune correlates of vaccine-induced protection against M . tb . We focused our comparative analyses only on the memory T cell response that persisted to the final study visit of each clinical trial for practical reasons . This is partly because some vaccines were administered once and others twice or three times . We acknowledge that restricting our analyses to the memory response at the final study visit ignores the effector response early after vaccination and other phases of the post-vaccination response , which could have provided more heterogeneity and revealed important immune features for differentiating between the vaccine candidates . However , considering the very different study designs and the fact that the ultimate purpose of vaccination is to induce long-lasting immunological memory , we decided against analysis of earlier time points . Focusing on a single timepoint rather than a longitudinal response also strengthened the statistical analysis . It simplified interpretation and permitted standard multivariate analysis . It also increased statistical power , as it reduced the number of hypothesis tests to perform and assessed the post-vaccination timepoint with the lowest response variability . Moreover , the number of days between the final measurement and both the first vaccination and the last vaccination varied between vaccines , which confounded measurement timings with vaccine . Regardless , it is unlikely that this would have affected our interpretation that M72/AS01E induced the highest T cell response magnitude among novel vaccine candidates , since follow-up time in the M72/AS01E trial was well within the follow-up ranges for the other vaccines . Finally , the small sample sizes for some groups limited our ability to detect significant differences between vaccines . Our study provides a framework for data-driven vaccine prioritisation . To facilitate these analyses in future , we highlight important factors that relate to standardisation , statistical power and immune response measurements . The first factor to strive for is achieving as much standardization as possible across trials of vaccine candidates , particularly in terms of the immunological assay and time points at which immune responses are measured , but also of participant inclusion and exclusion criteria and the method for defining Mtb infection . Inclusion of a common stimulation antigen preparation , such as a “megapool” of M . tb peptides ( as in [19] ) , would facilitate cross-trial comparisons , although responses to antigens that are not in individual vaccines could mask vaccine-specific response differences . The second factor is sample size . Comparing vaccines entails numerous pair-wise comparisons and ensuring sufficiently large sample sizes would facilitate detecting and characterising differences . Whilst a formal calculation of the precise sample size required depends on both inherent response variability and planned hypothesis tests , our experience from this work suggests that groups with less than twenty participants handicap statistical analyses . We also recommend the following to achieve robustness in statistical comparisons . When comparing multiple outcomes or groups the false discovery rate should be stringently controlled and statistical approaches that are robust to outliers , yet still efficient ( e . g . trimmed mean ) should be used . In addition , both the significance of differences and confidence intervals should be reported . Another means of increasing power is to reduce the number of hypothesis tests to perform . This can be done by pre-defining and ranking hypotheses in order of importance [30] . One means of informing hypothesis prioritisation would be to use data from any vaccine dose groups in the same trials not used for inter-vaccine comparisons to identify likely differences between vaccines . Another would be to focus such analyses on any correlates of protection that may be identified in future [29 , 31] . The third factor is measuring relevant immune responses and transforming them appropriately . We suggest widening the scope of immune responses measured to more broadly characterise the vaccine-induced responses ( i . e . covering the widest specific immune response “real estate” ) . For analysis of antigen-specific T cells producing one of various cytokine combinations , we think that our response size and response profile measures accurately and intuitively capture two distinct and important features of the immune response . Transformations of other immune responses should also be meaningful and interpretable . In conclusion , our study suggests that the T cell response feature which most differentiated between the TB vaccine candidates was response magnitude , whilst functional profiles suggested a lack of response diversity . Since M72/AS01E induced the highest memory CD4 T cell response it demonstrated the best vaccine take . In the absence of immunological correlates of protection , the likelihood of finding a protective vaccine by empirical testing of candidates may be increased by the addition of candidates that induce distinct immune characteristics .
The dataset analyzed in this study collates vaccine-specific immune responses from different clinical trials performed at the SATVI Field Site outside Cape Town , South Africa . A summary of the different trials is presented in Table 1 . Immune responses were measured by whole blood intra-cellular staining assay ( WB-ICS ) with multiparameter flow cytometry [17 , 18] . Fresh whole blood was stimulated for 12 hours with peptide pools spanning the relevant antigens ( Table 2 ) , or whole , live BCG . The study protocol was approved in writing by the Human Research Ethics Committee of the University of Cape Town ( HREC ref: 039/2017 ) and is based on anonymized data from previously published clinical studies [8–16 , 19] . AERAS-402-vaccinated participants were from trial 003 ( South African National Clinical Trials Register; no . 1381 ) . H56:IC31-vaccinated participants were from trial C-035-456 ( clinicaltrials . gov; NCT01865487 ) . M72/AS01E-vaccinated participants were part of trials TB010 and TB012 ( clinicaltrials . gov; NCT00950612 ) . MVA85A-vaccinated participants were part of trial TB008 ( clinicaltrials . gov; NCT00460590 ) and trial TB011 ( clinicaltrials . gov; NCT00480558 ) . H1:IC31-vaccinated participants were part of trial THYB-04 ( South African National Clinical Trials Register; DoH-27-0612-3947 ) . ID93+GLA-SE-vaccinated participants were part of trial IDRI-TBVPX-114 ( clinicaltrials . gov; NCT01927159 ) . The trial registry and reference number for BCG-vaccinated participants are clinicaltrials . gov reference NCT01119521 . In some trials different vaccine doses and/or number of administrations were assessed . In these cases , to simplify interpretation of results and increase statistical power , we selected the dose and/or administration strategy that was reported as optimal in the original trial report , based on vaccine safety and tolerability as well as immunogenicity outcomes ( Table 2 ) . This was generally the dose that induced the highest T cell response magnitude . As a result of poor standardization between different clinical trials a number of important trial design features differ substantially between the different trials ( Table 2 ) , including age of vaccinees ( ranging from adolescents to adults ) , method and cut-off for diagnosing M . tb infection ( tuberculin skin test [TST] , ESAT-6/CFP-10 responses detected by IFNγ ELISpot assay or QuantiFERON Gold In-Tube [QFT] ) , number of and timing of vaccine administrations , sampling timepoints for immunological measurements and the duration of participant follow-up . All novel TB vaccine candidates , except AERAS-402 , were given to both M . tb-uninfected and -infected individuals ( Table 2 ) . Participants in the AERAS-402 trial were assessed for M . tb-infection based on a TST induration of ≥15mm [11] . In the adult MVA85A trials infection was based on a TST induration and a positive response to ESAT-6/CFP-10 peptide pool in an in-house IFNγ ELISpot assay [10] while infection was based on a TST induration of ≥15mm and a positive ELISpot response to ESAT-6/CFP-10 peptide pool in the adolescent MVA85A trial [9] . All other trials used QFT with the manufacturer’s 0 . 35IU/mL threshold . All trials included participants that had been vaccinated with BCG at birth , except for the adult MVA85A trial , where this was not an inclusion criterion [10] . It is therefore possible that some participants of the latter may not have been vaccinated with BCG at birth . However , since the adult MVA85A trial participants were all immunologically sensitized due to M . tb-infection , we propose that the BCG-vaccination status of these participants would be unlikely to play an important role in vaccine-induced immune responses . Comparisons between vaccines were confounded by differences in vaccine-administration schedule ( Table 2 ) . This study therefore compared overall vaccination strategies , rather than vaccines . Age of participants–adolescents or adults–also varied by vaccine ( Table 2 ) . Within this study , adolescents and adults were considered immunologically equivalent . Pre-vaccination is the time point at which the first vaccine was given to an individual . The memory time point for CD4 and CD8 T cell response cytokine expression profiles provides the number of samples after excluding individuals based on negligible change from the pre-vaccination timepoint . The cut-off used for exclusion was a sum of absolute changes across the different cytokine combinations from pre-vaccination levels of 0 . 02 . Antigen-specific CD4 and CD8 T cells producing IFNγ , IL-2 , TNF and/or IL-17 were measured by WB-ICS assay using flow cytometry as previously described [17 , 18] . Frequencies of T cells expressing cytokines in the unstimulated control ( background ) were subtracted from those in antigen-stimulated conditions; where the background response was greater than the stimulated response , the background-subtracted response was set to zero . When T cell responses for an individual vaccine were measured by separate peptide pools ( representing different antigens ) , background-subtracted response frequencies for the antigens were summed . We defined pre-vaccination responses as the response at day zero ( measured before the first vaccine administration ) and the memory response as the response at the final time point . To yield the vaccine-induced memory response , we subtracted the pre-vaccination response from the memory response . Where an individual lacked a pre-vaccination response measurement , the median response for its vaccine and M . tb infection status group was used . We analyzed the antigen-specific T cell response magnitude and the cytokine co-expression profile of antigen-specific T cells . To define the response magnitude , let mij be the frequencies of vaccine-induced memory CD4 or CD8 T cells for the j-th cytokine combination for the i-th individual . Then the vaccine-induced response size for the i-th individual is equal to ∑j=17mij . This is the net change from pre-vaccination in frequencies of antigen-specific CD4 or CD8 T cells producing IFNγ , IL-2 and/or TNF . The cytokine co-expression profile aimed to reflect the T cell differentiation or quality of the antigen-specific T cell response . By using mij as defined above , the profile measure for the i-th individual for the j-th cytokine combination is given by mij∑j=17|mij| . We named this the scaled response . If all the changes from pre-vaccination levels for a certain individual are positive , then the scaled response for a certain cytokine combination gives the proportion of that change that consists of CD4 T cells producing that cytokine combination . The scaled response allows us to compare the extent to which different vaccines "favour" induction of different cytokine combinations , independently of the overall magnitude of the response induced by the vaccine . For analyses of cytokine co-expression profile , we excluded individuals for which the sum of absolute changes from pre-vaccination was less than 0 . 02 , to ensure that the response profile was only analyzed in individuals with memory responses that meaningfully changed relative to pre-vaccination . Number of participants included in this study after exclusion are shown in S1 Table . Because sample sizes were often small ( Table 2 ) and the response distributions severely skewed , bootstrapping was used to both construct confidence intervals and perform hypothesis tests for univariate population statistics . Confidence intervals were constructed using the bias-corrected and accelerated method [32] based on 104 bootstrap samples . Hypothesis testing was performed using the bootstrap-t approach [33] based on 104 bootstrap samples , with standard errors calculated using the double bootstrap [34] based on 500 bootstrap samples . We used trimmed means with symmetric trimming of the smallest 20% and the largest 20% of observations , because this is more robust to outliers than non-trimmed means and reflected the typical T cell response better [35] . For one-sample hypothesis tests , the false discovery rate was controlled at 0 . 01 using the Benjamini-Hochberg procedure [36] . For groups of pair-wise comparisons , the false discovery rate was controlled at an increased 0 . 05 , due to the positive dependency of the tests . Multivariate cytokine co-expression profile data were analyzed using principal components analysis ( PCA ) [37 , 38] and biplot axes [39] were calibrated to stretch between the mean and the maximum observed value [40] . To assist interpretation of the biplot , confidence areas for each group were estimated by taking bootstrap samples of the mean , assuming normality of the bootstrap distribution , and applying the fact that the Mahalanoubis distance from the mean has a χ22 distribution . The normality of the bootstrap samples may be checked by assessing the ellipticity of the contour lines of the bootstrap sample kernel density . Axis predictivity was used to detect which variables are well represented in a biplot [41] and only variables with high axis predictivity relative to other variables have their axes displayed . This simplifies biplot interpretation , and indicates which variables primarily drove the variability in the response .
|
Tuberculosis ( TB ) causes more deaths than any other single infectious disease , and a new , improved vaccine is needed to control the epidemic . Many new TB vaccine candidates are in clinical development , but only one or two can be advanced to expensive efficacy trials . In this study , we compared magnitude and functional attributes of memory T cell responses induced in recently conducted clinical trials by six TB vaccine candidates , as well as BCG . The results suggest that these vaccines induced CD4 and CD8 T cell responses with similar functional attributes , but that one vaccine , M72/AS01E , induced the largest responses . This finding may indicate a lack of diversity in T cell responses induced by different TB vaccine candidates . A repertoire of vaccine candidates that induces more diverse immune response characteristics may increase the chances of finding a protective vaccine against TB .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"t",
"helper",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"immunology",
"tropical",
"diseases",
"vaccines",
"bacterial",
"diseases",
"developmental",
"biology",
"molecular",
"development",
"cytotoxic",
"t",
"cells",
"infectious",
"disease",
"control",
"infectious",
"diseases",
"white",
"blood",
"cells",
"memory",
"t",
"cells",
"animal",
"cells",
"tuberculosis",
"t",
"cells",
"immune",
"response",
"immune",
"system",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] |
2019
|
A comparison of antigen-specific T cell responses induced by six novel tuberculosis vaccine candidates
|
The Chicxulub bolide impact caused the end-Cretaceous mass extinction of plants , but the associated selectivity and ecological effects are poorly known . Using a unique set of North Dakota leaf fossil assemblages spanning 2 . 2 Myr across the event , we show among angiosperms a reduction of ecological strategies and selection for fast-growth strategies consistent with a hypothesized recovery from an impact winter . Leaf mass per area ( carbon investment ) decreased in both mean and variance , while vein density ( carbon assimilation rate ) increased in mean , consistent with a shift towards “fast” growth strategies . Plant extinction from the bolide impact resulted in a shift in functional trait space that likely had broad consequences for ecosystem functioning .
The Cretaceous–Paleogene boundary ( KPB ) is marked by the Chicxulub bolide impact and mass extinction [1]–[3] . In temperate North America , while the impact resulted in the extinction of more than 50% of plant species [4] , a major unresolved issue is whether this killing event was also a large-scale selection event [5] . Wolfe [6] originally proposed that the KPB selected against evergreen species . Specifically , competition in the cold and dark climates during the impact winter [7] should have selected for species with ecological strategies [8] associated with deciduousness . Because variation in leaf traits reflect ecological strategies that are coupled to whole-plant carbon and water fluxes , such a strategy shift would also have caused broad shifts in terrestrial ecosystem functioning [9] , [10] . Although some fossil data are consistent with the rise of deciduous species near the KPB , inferences on the extinction's selectivity have been based on qualitative proxies or limited occurrence data [11]–[14] . A lack of quantitative trait data has limited our understanding about the selectivity and ecological implications of this important extinction event . Here we test if the extinction event was selective for functional traits related to plant ecological strategies . We measure functional traits from fossil leaf assemblages spanning a 2 . 2 Myr interval across the KPB and assess four differing selection scenarios for functional traits: ( i ) directional selection—a shift in phenotype space , caused by novel postboundary environments replacing preboundary environments; ( ii ) stabilizing selection—a reduction in phenotype space caused by postboundary conditions making some strategies temporarily nonviable; ( iii ) diversifying selection—an increase in phenotype space due to a wider range of postboundary conditions; or ( iv ) a lack of selection , because previous adaptations were unrelated to survival after major catastrophe . We assess evidence for these selection scenarios using functional traits [15] , [16] related to the leaf economics spectrum [17] . This global spectrum details how several traits are linked to variation in plant growth and fitness . It describes a continuum between “slow-return” leaves ( low rates of carbon assimilation , long lifespans , and high tissue carbon investment ) and “fast-return” leaves ( high rates of carbon assimilation , short lifespans , and low tissue carbon investment ) . In general , evergreen leaves are “slow . ” Within angiosperms , this strategy is thought to be selected for when resource availability is less variable [18] . In contrast , “fast” deciduous angiosperm leaves are thought to be selected for when resource availability is more variable [19] . In this updated context , Wolfe's hypothesis [6] proposed that the strong variation in light levels and temperature after the bolide-caused impact winter [1] should have resulted in directional selection for “fast” strategies . Alternatively , longer-term temperature change may also have selected for certain traits across the KPB . Multiple proxies show a brief warming during the latest Cretaceous , starting approximately 300 , 000 y before the boundary near the base of Chron 29r , followed by a rapid return to cooler temperatures approximately 50 , 000 y before the boundary , which were then maintained through the earliest Paleogene [7] , [20] , [21] . This longer-term temperature change could have counteracted the hypothesized bolide-caused selection against traits that characterize “slow” strategies . Recent trait-climate theory suggests that cooling should lead to reduced evapotranspiration demand , directionally selecting for “slow” strategies ( assuming no change in atmospheric [CO2] ) [22] , [23] . This prediction appears opposite to the predictions of Wolfe's deciduousness hypothesis , but the relative strengths of both predictions have been unknown . We evaluate the causes and consequences of ecological change across the KPB by producing records of two central leaf functional traits that reflect variation in ecological strategies: leaf mass per area [LMA , g dry leaf mass/m2 leaf area ( LA ) ] , and leaf minor vein density ( VD , mm vein/mm2 LA ) . Variation in these traits reflects a tradeoff between the transport of water and carbon [24] and the carbon construction cost of the leaf [23] . LMA is a metric of leaf carbon investment and correlates negatively with deciduousness and leaf lifespan [19] , [25] . VD is a metric of the hydraulic and photosynthetic capacity of a leaf . Leaves with higher VD reflect “fast” species with higher rates of water flux and carbon assimilation [23] , [24] and are more likely to be found in warm environments [22] , [26] . Thus , by examining each trait's pre- and postboundary distribution , we can assess if the KPB extinction event was selective and which of the differing selection scenarios best matches temporal trends in leaf traits .
We measured VD and LMA for leaf fossil assemblages spanning the last ca . 1 . 4 Myr of the Cretaceous ( K ) and the first ca . 0 . 8 Myr of the Paleogene ( P ) ( Figure 1 ) . Floras come from the Hell Creek and Fort Union formations in southwestern North Dakota , United States ( paleolatitude 49°N ) [27] . Of the known leaf macrofossil assemblages spanning the KPB , these are currently the best preserved and most species-rich . We found that the mean and variance in LMA decreased across the KPB . Mean values shifted from a Cretaceous value of 86±27 s . d . g m−2 to a Paleogene value of 78±15 s . d . g m−2 ( Mann–Whitney location test , p = 0 . 04; Figures 2A and 3 ) [17] . This mean shift is small relative to the global range of LMA values ( 3–2 , 000 g m−2 [17] ) but is of comparable magnitude to shifts across modern ecosystem types ( e . g . , from tropical rain forest to tropical deciduous forest , 73 versus 83 g m−2 , respectively ) [28] . Variance in LMA fell by 67% across the KPB ( Brown–Forsythe Levene test for homogeneity of variances , p<10−3 ) . These nonparametric tests account for unequal group sample sizes ( nK = 256 and nP = 67 ) . We also found that the mean of VD increased across the KPB . Mean values shifted from a Cretaceous value of 3 . 5±0 . 6 s . d . to a Paleogene value of 4 . 6±0 . 7 s . d . mm−1 ( p<10−3; Figures 2B and 3 ) . This shift is comparable to the average difference between extant nonangiosperms and eudicots ( 1 . 8 versus 8 . 0 mm−1 , respectively ) [29] . Variance in VD did not change across the KPB ( p = 0 . 66 ) . Note that group sizes here were also unequal ( nK = 146 and nP = 24 ) . The trends towards higher VD and reduced ranges of LMA reflect large physiological and biological shifts in plant functioning . The constriction of LMA space was caused primarily by the loss of high-LMA species , many of which were abundant in the latest Cretaceous ( Figures 2A and S3 ) . The shift in VD space was caused mostly by the loss of Cretaceous-only species with low VD relative both to species that survive the KPB and to Paleogene-only species ( regression; p<10−9 , r2 = 0 . 22; Figure 2B ) . Several of these patterns may be driven by sampling biases—for example , temporal variation in facies preservation [30] . In this dataset , Paleogene sites come primarily from channel facies , whereas Cretaceous sites usually come from floodplain facies . To assess the potential impact of differential preservation , we repeated analyses after subsetting by each facies type for each trait . For channel VD , the mean effect remained significant ( p = 0 . 03; nK = 130 and nP = 4 ) , as it did for floodplain VD ( p = 0 . 004; nK = 16 and nP = 20 ) . For channel LMA , there was no longer a significant shift in mean or variance ( both p>0 . 11; nK = 222 and nP = 15 ) . Similarly no effect in mean or variance was found for floodplain LMA ( both p>0 . 29; nK = 34 and nP = 52 ) . Low Paleogene sample sizes and unavoidable facies shifts prevent further inferences . Nonetheless , together these results indicate a strong evidence for a shift in the mean of the VD distribution but yet weaker evidence for shift in the variance of the LMA distribution . Moreover , because channel ( riparian ) environments typically support more “fast-return” specialists than distal floodplain environments owing to their higher rates of physical disturbance and greater volatility in nutrient availability [31] , our facies effect should have had the opposite effect on LMA and VD than what we observed . Other data quality issues may also bias our results . First , estimates of VD may be underestimates of true values because of incomplete fossil preservation . Although this is an unavoidable problem when estimating VD for any leaf fossil , our protocol did exclude all but the best-preserved specimens ( Materials and Methods ) . Reported values are consistent with other estimates from late-Cretaceous fossils [32] . Second , estimates of LMA may be down-biased if some species were incorrectly expert-determined to be herbaceous rather than woody . We therefore repeated all LMA analyses across and within facies , with herbaceous taxa removed ( omitting n = 21 specimens ) . We found no qualitative changes in conclusions . Shifts in both functional traits were associated with ongoing climate change . The bin-mean VD and temperature were negatively correlated ( r2 = 0 . 15 , p = 0 . 01 ) , whereas bin-mean LMA and temperature were positively correlated ( r2 = 0 . 19 , p = 0 . 005 ) . Similarly at the specimen level , VD and LMA were not correlated with each other across ( p = 0 . 53 ) or within facies ( both p>0 . 45 ) , but this null result is likely due to the low number of available samples ( channel , nK = 1 and nP = 4; floodplain , nK = 31 and nP = 1 ) . Overall , these results indicate a systematic shift in trait space across the KPB whether examined at the bin-mean or specimen level .
The Chicxulub bolide impact appears to have led to the selective extinction of plant species with “slow” leaf ecological strategies . Consistent with Wolfe's hypothesis , this mass extinction was characterized by directional selection away from evergreen species [6] , as seen through both VD and LMA , as well as stabilizing selection , as seen through LMA . Our study therefore provides strong evidence that the KPB mass extinction was functionally selective for plants . The increase in VD from the late Cretaceous to the early Paleogene in our data parallels the increases in VD seen across angiosperm taxa in the early-mid Cretaceous and then again across the KPB [32] . The increase reported by Feild et al . ( 2011 ) occurred globally over an ∼70 Myr time span , with most of this increase occurring conclusively before the KPB . In contrast , our findings show a VD increase in a single region , over a much shorter time period ( ∼2 Myr ) , with the majority of the increase occurring at or after the KPB . Nevertheless , it is possible that similar atmospheric changes drove both trends . Higher VD in angiosperms and increasing angiosperm dominance across the Cretaceous could be driven by declining atmospheric [CO2] , because lower carbon dioxide availability could select for higher hydraulic capacity to maintain productivity [33] . Trait-climate theory also predicts a negative correlation between temperature and VD under low [CO2] [22] , consistent with observed data for the KPB . A hypothetical rapid [CO2] decrease across the KPB therefore could be a key driver to our findings of an observed shift in VD . One hypothetical driver of lower [CO2] could be enhanced chemical weathering , via late-Cretaceous Deccan volcanism [34] . However , extant KPB proxies do not yet have sufficient age control [35] or temporal resolution [35] , [36] to accurately reconstruct finer scale [CO2] dynamics within the time interval of interest . More detailed assessments of atmospheric composition and temperature across the KPB would be needed before this prediction could be tested . There is an important criticism of the proposed mechanism coupling between VD and [CO2] . Although selection against low hydraulic capacity in low [CO2] environments should occur for all plants , for nonangiosperms and shade species , low values of VD appear to have successfully persisted in the fossil record across the Cretaceous [37] , suggesting that low values were not necessarily selected against as originally hypothesized [33] . We therefore suggest that the increase in VD seen across the KPB is more likely to be a direct consequence of the bolide impact selecting for specific leaf economic strategies rather than of ongoing longer-term climate change . Nevertheless , the increasing dominance of deciduous angiosperms by the early Paleogene appears to have been reinforced by both bolide impact and longer-term climate change . Observed trait dynamics across the KPB likely ramify to influence ecosystem functioning . Because of the close linkage between leaf economic traits and ecosystem resource fluxes [9] , [10] , selection at the KPB should have also strongly modulated net primary productivity in terrestrial systems [28] , as well as regional hydrological cycles [38] . Our results therefore suggest that there were associated functional changes in terrestrial ecosystems in the aftermath of the Chicxulub impact . Functional trait dynamics are of wide interest when studying succession , invasion , and other dynamical questions , but contemporary time-series data are very rare ( e . g . , [39] ) Our study thus highlights the power of paleoecological functional trait data to integrate information on climate change , extinction , and species performance across time .
We analyzed fossils of nonaquatic nonmonocot angiosperm taxa from sites previously collected from the Hell Creek and Fort Union Formations , now located in southwestern North Dakota . The stratigraphy of these sites and the identification of specimens has been described previously [40] . Each specimen was previously assigned to one of 312 morphotypes and one of 208 sites corresponding to a known stratigraphic position and sedimentary facies . During the summer of 2013 , we examined all appropriate specimens at the Denver Museum of Nature and Science ( DMNH ) and Yale Peabody Museum ( YPM ) . We digitally photographed specimens with ( 1 ) intact petioles at the point of insertion with the leaf blade and ( 2 ) LAs that could be directly measured or confidently reconstructed . For each image , leaves were separated from their rock matrix using the lasso tool in Adobe Photoshop; LA and petiole width ( PW ) were then measured using ImageJ following the protocol of [41] . For woody species , LMA was then calculated using the following empirical scaling function [41]: ( 1 ) and for herbaceous species [42]: ( 2 ) The final dataset comprises 612 specimens representing 135 morphotypes and 102 sites , of which only 21 specimens are designated as herbaceous . During June 2012 , we examined all appropriate specimens at DMNH , including holomorphotypes on loan from YPM ( approximately 6 , 000 specimens total ) . We selected specimens that appeared to have complete preservation of the minor venation network in at least one region of the fossil , discarding all others ( retaining 1 , 150 specimens ) . In specimens containing more than one leaf , we made a measurement for each leaf . We captured a digital image of each leaf fossil centered on the region of interest using a dissecting microscope coupled to a digital camera ( T2i , Canon ) . Illumination was provided by ring-light . Image resolution was 243 pixels per millimeter . We then estimated VD using a line-counting program [22] developed in MATLAB . To prevent any investigator bias , images were unlabeled and analyzed in random order . We converted each image to grayscale and applied a contrast limited adaptive histogram equalization to improve quality . If vein preservation appeared incomplete ( e . g . , clear fragmentation of specimen , higher-order veins preserved in one region but not in another , dramatically fewer veins in one of two conspecific specimens ) , we discarded the specimen at this step . For acceptable specimens , we delineated a polygonal region of interest in each image ( corresponding to the area in which veins were potentially preserved ) . The program then generated a randomly oriented line segment spanning this polygon . We manually counted the number of vein-line intersections and repeated this process for 10 random line segments . We then computed the mean distance between veins as the sum of all line counts divided by the sum of all distances . This distance was then converted to VD via an equation [22] stating that VD ( mm−1 ) and intervein distance ( d , mm ) are related as: ( 3 ) This procedure yielded a final dataset of 468 specimens from 66 species and 61 sites . Each specimen was originally assigned a facies type by the original collector ( table 3 in [40] ) . Channel facies were defined as those recorded as “aban chan , ” “channel , ” “channel la , ” “channel x , ” or “Colgate channel . ” Floodplain facies were defined as those originally recorded as “ash bed , ” “carb shale , ” “carb splay , ” “floodplain soil , ” “pond , ” “pond ls , ” “pond vb , ” “splay , ” “splay/levee , ” or “volcanic ash . ” We calculated time series of VD and LMA using a binning approach . We first assigned each specimen to a 1-m depth interval defined by integer-rounding the measured stratigraphic depth . Within each bin we calculated species-at-site mean trait values , then used these to compute site means . We used these site-means to infer the distribution ( i . e . , mean , upper quartile , lower quartile ) of trait values within each bin . Because VD and LMA data were not always available within the same bin , we used piecewise linear interpolation to infer LMA values at each bin for which VD values were available ( n = 40 ) . An index of temperature ( Tc; °C ) was obtained from a leaf margin analysis of the same fossil floras using the range-through mean annual temperature values corresponding to the supplementary table 8 of [20] . These leaf-reconstructed temperatures are supported by carbonate clumped isotope paleothermometry from the same region [21] . Because vein data and margin data were not always available for the same stratigraphic bins , we estimated temperatures with a piecewise linear interpolation of range-through temperature against stratigraphic bin , choosing constant values at endpoints . We modeled the uncertainty in Tc using the error estimates provided in the original source , using the mean plus ( minus ) one standard deviation as the upper ( lower ) quartile . When no standard deviations were available , we assigned a standard deviation equivalent to the mean of all other standard deviations ( 2 . 8°C ) .
|
Sixty-six million years ago the Chicxulub bolide impacted the Earth , marking the Cretaceous–Paleogene boundary ( KPB ) . This event caused the planet's most recent mass extinction , but the selectivity and functional consequences of the extinction on terrestrial plants has been largely unknown . A key untested hypothesis has been that a subsequent impact winter would have selected against slow-growing evergreen species , a possible cause of the modern dominance of high-productivity deciduous angiosperm forests . We tested this hypothesis using fossil leaf assemblages across a 2-million-year interval spanning the KPB . We assess two key ecological strategy axes—carbon assimilation rate and carbon investment—using leaf minor vein density and leaf mass per area as proxies , respectively . We show that species that survive the KPB have fast-growth ecological strategies corresponding to high assimilation rates and low carbon investment . This finding is consistent with impact winter leading to the nonrandom loss of slow-growing evergreen species . Our study reveals a dramatic example of the effect of rapid catastrophic environmental change on biodiversity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"paleoclimatology",
"ecology",
"and",
"environmental",
"sciences",
"ecological",
"niches",
"paleobiology",
"ecology",
"ecophysiology",
"earth",
"sciences",
"paleontology",
"biology",
"and",
"life",
"sciences",
"plant",
"ecology",
"paleoecology",
"biodiversity",
"macroecology",
"paleobotany"
] |
2014
|
Plant Ecological Strategies Shift Across the Cretaceous–Paleogene Boundary
|
In addition to their protein coding function , exons can also serve as transcriptional enhancers . Mutations in these exonic-enhancers ( eExons ) could alter both protein function and transcription . However , the functional consequence of eExon mutations is not well known . Here , using massively parallel reporter assays , we dissect the enhancer activity of three liver eExons ( SORL1 exon 17 , TRAF3IP2 exon 2 , PPARG exon 6 ) at single nucleotide resolution in the mouse liver . We find that both synonymous and non-synonymous mutations have similar effects on enhancer activity and many of the deleterious mutation clusters overlap known liver-associated transcription factor binding sites . Carrying a similar massively parallel reporter assay in HeLa cells with these three eExons found differences in their mutation profiles compared to the liver , suggesting that enhancers could have distinct operating profiles in different tissues . Our results demonstrate that eExon mutations could lead to multiple phenotypes by disrupting both the protein sequence and enhancer activity and that enhancers can have distinct mutation profiles in different cell types .
Protein coding sequences have been shown to contain additional functional information such as splicing [1] , [2] , mRNA stability [3] , microRNA target sites [4] , and transcriptional enhancer activity [5]–[10] . Furthermore , by analyzing various genomic datasets , numerous exons were shown to interact with promoter and enhancer-like regions suggesting that they are involved in alternative splicing , chromatin structure and gene regulation [2] . A study that analyzed enhancer-associated ChIP-seq datasets found that , on average , 7% of peaks overlap coding exons [5] , suggesting that numerous eExons are embedded in mammalian genomes . Furthermore , functional characterization of potential eExons in zebrafish [9] and mice [5] found that over half of them are functional developmental enhancers . In addition , it was shown that eExons can regulate the gene they reside in [6]–[10] and also nearby genes [5] , suggesting that mutations in these exons could lead to phenotypes that are not due to their protein function . The functional consequence of non-synonymous mutations is well established [11] . Synonymous mutations have also been shown to have a functional effect , for example causing disease by improper splicing [12] . However , whether mutations in coding exons can lead to phenotypes by impacting enhancer function is not well known . A recent study demonstrated that human chromosomal abnormalities encompassing two eExons , DYNC1I1 exon 15 and 17 , that regulate developmental limb expression of two neighboring genes , DLX5 and DLX6 , could cause split hand and foot malformations [5] , similar to a DLX5 coding mutation [13] . Transcription factors involved in gene regulation were suggested to bind to coding protein sequences and are thought to influence codon choice and , consequently , protein evolution [14] . However , the functional consequence of point mutations in eExons has yet to be determined . Here , we identified several functional liver eExons and used massively parallel reporter assays ( MPRA ) [15]–[19] to determine the functional consequences of all possible nucleotide substitutions in three of them . We found that synonymous and non-synonymous mutations have a similar effect on enhancer activity and deleterious enhancer mutations tend to overlap liver-associated transcription factor binding sites ( TFBS ) . Using similar MPRA assays in HeLa cells , we show that their mutational profile can vary in different cell types .
In order to identify functional liver eExons for MPRA , we analyzed liver-associated ChIP-seq datasets for coding exons that have enhancer marks . We analyzed enhancer associated ChIP-seq datasets ( H3K4me1 , H3K27ac and p300 ) of human hepatocytes generated in a separate study ( Smith RP et al . , manuscript in preparation ) and in 8 week old mouse liver [20] ( see methods ) . Since these enhancer marks could also identify potential promoters , we excluded the first exons of genes that had enhancer marks from our subsequent analyses , as these exons could potentially be promoters and not necessarily enhancers ( Table S1 ) . For the H3K4me1 and H3K27ac ChIP-seq datasets , we found that 18–20% and 9–11% of all peaks overlap coding exons in human hepatocytes and adult mouse liver , respectively . Since the average peak size of these two enhancer marks is rather long ( ∼2 kb ) , we also analyzed p300 ChIP-seq datasets , which have a shorter average peak size ( ∼400 bp ) . We found that 7% and 6% of all peaks overlap coding exons in human hepatocytes and mouse liver , respectively , ( Table S1 ) . We next wanted to increase our chances of obtaining a functional eExon for our enhancer assays and subsequent MPRA . We thus chose the human hepatocyte p300 ChIP-seq dataset for all subsequent analysis ( Figure S1 ) . This was done due to its aforementioned shorter peak size , thus increasing the chances that a substantial exon-peak overlap will be indicative of function , since 80% of all human exons are <200 bp in length [21] . We selected only coding exons where at least 25% of the human hepatocyte p300 ChIP-seq peak overlapped the actual exon . Using transcription factor ( HNF4A , FOXA1 , FOXA2 ) ChIP-seq datasets from HepG2 cells [22] , our eExon candidates were further filtered for peaks that were bound by at least two of these transcription factors . From the remaining 48 eExon candidates , we selected 15 for individual enhancer assays ( Table S2 ) . To determine whether these exons function as liver enhancers , we tested the fifteen selected candidates for enhancer activity in human hepatocellular carcinoma ( HepG2 ) cells and mouse liver . The selected ChIP-seq exonic peaks were amplified from human genomic DNA and cloned into the pGL4 . 23 vector ( Promega ) that contains a minimal promoter followed by the luciferase reporter gene . The vectors were transfected into HepG2 cells and luciferase activity was measured after 24 hours . Ten out of the fifteen tested exons had significant luciferase activity versus the empty vector ( p<0 . 05; t-test ) , suggesting that they function as enhancers ( Figure 1A ) . We next tested these sequences in mice using the hydrodynamic tail vein injection [23] , [24] and found that eight out of the fifteen exons showed significant liver enhancer activity ( Figure 1B; p<0 . 05; t-test ) . Combined , seven exons functioned as enhancers in both HepG2 and mouse liver . Since eExon mutations could alter both protein coding function and enhancer activity , we were interested to assess the effect of single nucleotide variants ( SNVs ) on enhancer function at single base pair resolution . We selected three eExons with the highest enhancer activity from the mouse enhancer assay for MPRA: 1 ) exon 17 of sortilin-related receptor , L ( DLR class ) A repeats containing ( SORL1 ) gene , which encodes the YWTD domain of this receptor ( also named LR11/SorLA ) . SORL1 is thought to play a role in endocytosis and sorting and is expressed in the human central nervous system and liver [25] , but absent in the mouse liver [26] , [27]; 2 ) exon 2 of TRAF3 interacting protein 2 ( TRAF3IP2; also known as ACT1 ) which encodes the Helix-loop-helix ( HLH ) domain of this protein . TRAF3IP2 is a signaling adaptor protein involved in the regulation of adaptive immunity through the IL-17 pathway [28]–[30] and is expressed in B-cells and liver [31] , [32] . 3 ) exon 6 of the peroxisome proliferator-activated receptor gamma ( PPARG ) gene , which encodes the ligand binding domain of this protein . PPARG is a nuclear receptor that regulates fatty acid storage and glucose metabolism and is expressed in the liver [33] . Using MPRA , we systematically dissected the functional consequences of all possible SNVs for these three eExons in the mouse liver ( Figure 2 ) . By oligonucleotide synthesis and polymerase cycling assembly , we generated for each eExon a low complexity library ( ≥10 , 000 ) of enhancer mutant haplotypes that diverged from the wild type sequence by ∼2–3% ( Table S3 ) . These mutant enhancers , along with 20-bp degenerate tags , were cloned into pGL4 . 23 ( see experimental procedures ) . We performed subassembly on each library to determine the full sequence of each enhancer haplotype and its represented tag [34] . Each library was then assayed in mice using the hydrodynamic tail vein injection , livers were harvested after 24 hours and the relative activities of individual haplotypes were measured by sequencing the transcribed tags . Using linear regression analysis , the effect-size of every possible single-nucleotide change on enhancer activity was estimated ( see experimental procedures ) . We found that the functional consequence of most mutations was modest , with ∼18% affecting activity by ≥1 . 2-fold and ∼4% by ≥2-fold . The majority of SNVs with ≥2 fold change reduced enhancer activity ( 92% ) and only a few SNVs increased activity ( 8% ) ( Table S4 ) . To validate our results , we individually tested selected SNVs using the hydrodynamic tail vein assay and found a high correlation with their MPRA profiles ( R = 0 . 94; Figure S2 ) . We next examined the functional profile of each eExon . In all three eExons , we observed dispersed clusters that had high effect sizes ( Figure 2A–C ) . For example , the first ∼250 bases and the last 120 bases of SORL1 exon 17 did not have a strong effect on enhancer activity , while nucleotide substitutions at positions 259–328 , which are mostly coding sequences , had a significant effect on enhancer activity ( Figure 2A ) . For TRAF3IP exon 2 , the first ∼330 bases and the last ∼240 bases did not have a significant effect on enhancer activity , while almost all nucleotide substitutions at protein coding positions 333–409 had a significant impact ( Figure 2B ) . PPARG exon 6 had clusters of nucleotide substitutions along the entire tested sequence with a significant impact on enhancer activity ( Figure 2C ) . These results point to the existence of discrete regions in these eExons that modulate enhancer activity and also code for important protein domains . We next investigated whether the functional regions of the enhancer are subjective to the protein code or impartial to it by analyzing the effect-size of synonymous versus non-synonymous SNVs in these three eExons . We divided the functional SNVs ( Fold change ( FC ) ≥1 . 2 ) to synonymous and non-synonymous and compared the effect size of synonymous versus non-synonymous SNVs on enhancer function . We found no significant differences between synonymous and non-synonymous SNVs and their effect-sizes ( p-value = 0 . 67; Fisher test ) . A similar analysis for just the high functional SNVs ( FC≥2 ) also showed no significant differences ( p-value = 0 . 36; Fisher test ) . Thus , our results suggest that the essential functional enhancer sequences are intertwined with the protein coding sequences . We analyzed the profiles of each eExon to determine whether SNVs that significantly affect enhancer activity overlap with liver-associated TFBS . We found several positions separated by less than ∼6 nucleotides that had significantly correlated effect sizes ( P≤0 . 01 ) ( Figure S3 A–C ) , suggesting that these clusters might represent TFBS . Indeed , when we analyzed these eExons for liver-associated TFBS , we observed a striking overlap between predicted liver-associated TFBS and these clusters ( Table S5; Figure 2D , E ) . Across the three tested eExons , AP-1 and HNF4A were the most prevalent TFBS overlapping these clusters . For example , a predicted HNF4A binding site at position 259–272 ( Figure 2D ) and a predicted AP-1 binding site at position 322–327 ( Figure 2E ) in SORL1 exon 17 , overlap a cluster of SNVs that significantly affect enhancer activity . All the mutations overlapping the AP-1 binding site had a negative effect on enhancer activity , with the notable exception of a variant with a positive effect ( 325 G>C ) that likely increased the AP-1 binding affinity ( Figure 2E ) . In addition , we also observed mutations that generated novel TFBS that increased enhancer activity . For example , in TRAF3IP2 exon 2 , 373G>T , and two neighboring nuclear factor 1 ( NF-1 ) binding sites , 393A>T , 399T>A , in SORL1 exon 17 were created by nucleotide substitutions and subsequently increased enhancer activity ( Figure S3D–E ) . To assess whether the binding of AP-1 and HNF4A is affected by these SNVs , we carried out co-transfection assays in HEK293T , a human embryonic kidney cell line where these three eExons are inactive ( Figure 3A ) . Co-transfection of AP-1 and/or HNF4A ( see online methods ) along with the three enhancer vectors showed significant enhancer activity for all three eExons , suggesting that AP-1 and HNF4A regulate their activity ( Figure 3A ) . We then evaluated whether SNVs that alter AP-1 and HNF4A binding sites could change their enhancer activity . For SORL1 exon 17 , deleterious mutations in predicted HNF4A ( 260G>T , 264G>A ) and AP-1 ( 322T>G ) TFBS reduced enhancer activity when the applicable TF was transfected , while SNVs ( 393A>T , 399T>A ) coinciding with NF-1 had similar enhancer activity as the reference sequence ( Figure 3B ) . A SNV in TRAF3IP2 exon 2 ( 373G>T ) that generated a predicted AP-1 binding site led to significant enhancer activity both with and without the transfected TFs ( Figure 3B ) . Not all SNVs that altered predicted TFBSs led to the anticipated enhancer activity changes in our assay . For example , SNVs 384T>A and 390T>A in PPARG exon 6 , that alter an AP-1 TFBS , did not reduce enhancer activity in AP-1 transfected cells ( Figure 3B ) . Combined , our results demonstrate that liver expressed TFs , such as AP-1 and HNF4A , modulate the enhancer activity of these eExons and mutations in their predicted TFBSs could partially explain the effect of SNVs on enhancer activity . To determine whether the distinctive TF repertoire of different cells could modulate enhancer activity , we set out to carry a similar MPRA analysis for all three eExons in a different cell type and compare the results to the regulatory profile from mouse liver . Using available TF ChIP-seq datasets [22] , we observed that these three eExons overlap several ChIP-seq peaks in various human cell lines ( Figure 2C–E ) . Of note , PPARG exon 6 and TRAF3IP2 exon 2 specifically overlap AP-1 ChIP-seq peaks in human cervical cancer cells ( HeLa ) cells . Subsequent luciferase reporter assays showed that all three eExons are active enhancers in HeLa cells ( Figure S4A ) . We thus set out to do a similar MPRA experiment in HeLa cells for all three eExons . Using MPRA , we dissected the mutational profile of these three eExons and found that the functional consequence of most mutations in HeLa cells was low to modest , with ∼12% affecting activity by ≥1 . 2-fold and ∼1% by ≥2-fold . Comparison to mouse liver mutation profiles showed that the overall effect size of SNVs ( with ≥1 . 2 fold change ) was lower in HeLa cells . The regulatory profiles of PPARG exon 6 showed high similarity between mouse liver and HeLa cells , while discrete differences were observed for the other two exons ( Figure 4 ) . In SORL1 exon 17 for example , a deleterious mutation cluster in mouse liver that overlaps an HNF4A TFBS at positions 259–272 ( Figure 2D ) had no effect in HeLa cells ( Figure 4A , D ) , while another mutation cluster at positions 310–317 had a significant effect on enhancer activity in HeLa but not in mouse liver ( Figure S4B ) . Some functional clusters remained unchanged in both experiments ( Figure S4B–D ) . For example , SNVs at positions 322–327 that overlap an AP-1 binding site ( Figure 2E ) had a similar profile in both cells with a lower effect size in HeLa . To validate our results , individual SNVs overlapping predicted TFBSs were tested in HeLa and were compared to our previous mouse liver results ( Figure S2 ) . Overall , we observed luciferase fold-changes that correlated with their corresponding MPRA results ( R = 0 . 77; see methods ) . Combined , our results demonstrate that the regulatory profile of an enhancer can change between different cell types .
In this study , we analyzed enhancer-associated ChIP-seq datasets ( H3K4me1 , H3K27ac , p300 ) from human hepatocytes and mouse liver and observed that on average ∼6% of peaks overlap coding exons after excluding the first exon . Enhancer assays for 15 selected exons that overlap these peaks showed that 8 and 10 of them are functional enhancers in adult mice and human HepG2 cells , respectively . Our observed ∼50% success rate is in line with previous reports that tested eExons for their developmental enhancer activity [5] , [9] . However , it is worth noting that in this study we further selected for potentially positive enhancers by looking at ≥25% exon-peak overlap and additional transcription factor ChIP-seq datasets which probably increased our functional enhancer success rate . Also of note , while human SORL1 is expressed in the liver [25] , mouse Sorl1 is not expressed in this tissue [26] , [27] , and the human sequence we used showed mouse liver enhancer activity . Previous work has shown that when human sequences are tested for enhancer activity , even in zebrafish , they can portray regulatory activity even if they do not have homologous sequences in that organism [35] , [36] . Using MPRAs , we measured the distribution of effect sizes of all possible SNVs in three liver eExons in vivo . The three eExon MPRA displayed comparable functional profiles to previously reported liver enhancer MPRA [17] , and similarly show that the majority of SNVs have modest effects on transcriptional activity . Since we tested the ChIP-seq peak which includes not only the exonic sequence but also the adjacent intronic sequences , the distribution of effect sizes allowed us to determine the location of the enhancer core and whether the protein coding sequence is required for enhancer activity . For two eExons , SORL1 exon 17 and TRAF3IP2 exon 6 , the distribution of SNVs with a significant effect on enhancer activity were specifically clustered in coding sequences that encode important protein domains . In addition , non-synonymous mutations in all three eExons had deleterious effects on enhancer activity , indicating that the genetic code can contain overlapping functional information . Interestingly , the dual function of eExons is not only restricted to protein coding function , as the cluster with the highest impact on enhancer activity in SORL1 exon 17 overlaps a donor splice site ( Figure 2E ) . Alternative splicing occurs simultaneously with nascent RNA transcription [2] and these eExons could also be splicing regulators . These results suggest that mutations in eExons could lead to multiple phenotypes , both from disruption of the protein function and transcriptional regulation . Our study has several caveats . One caveat is that we are not testing the enhancer activity of these sequences in their natural setting . Similar to most standard enhancer assays , our study takes out these sequences from their natural settings and places them in front of a minimal promoter that is different than their target promoter . In addition , the transcriptional status of these exons in concert with their enhancer activity is not being assessed . Future studies that create knockin mice to selected exon coding mutations that have a deleterious effect on enhancer activity , as determined by our MPRA , could potentially assay this . We also do not know the target genes that are regulated by these eExons . The genes these exons reside in are expressed in most of the tissues and cell types that we analyzed ( Table S2 ) , but this does not necessarily establish that they are their target genes . Furthermore , analysis of available chromatin interaction Hi-C datasets [37] , [38] from different cell types ( IMR90 , hESC ) suggest that these eExons interact with their gene promoter region as well as nearby gene promoters and other intergeneic regions ( Table S7 ) . An interesting and still open question is what sort of evolutionary mechanism permitted coding exons to acquire another function as transcriptional enhancers . Two different forces could have been the drivers for their evolutionary constraint: protein function or transcription regulatory function , such as enhancer activity . Regarding protein constraint , previous studies have shown that 70% of amino acids in a protein can be altered while maintaining its structure and function [39] . This suggests that , in addition to synonymous sites , non-synonymous changes could allow for substantial flexibility , accommodating for enhancer function . Comparison between synonymous and non-synonymous SNVs in our study did not find any significant differences in their effect on enhancer activity . Supporting this is a recent study that used genomic deoxyribonuclease I footprinting analysis within synonymous and non-synonymous sequences and found that transcription factor binding could possibly impose a functional constraint of both the regulatory code and amino acid choice [14] . Analysis of synonymous constraint elements ( SCE ) [40] , which are significantly conserved synonymous regions , found that only ∼6% of them overlap exonic human hepatocytes and mouse liver ChIP-seq peaks ( Table S6 ) . Four of the 15 tested exons overlap SCE , and PPARG exon 6 was the only eExon tested by MPRA to overlap an SCE ( Table S2 ) . These results suggest that at least for our selected eExons , a tendency for synonymous constraint was not observed ( Table S2 ) . The evolutionary constraint of noncoding enhancers has frequently served as a proxy for functional constraint [41] , [42] , but studies have also shown that many noncoding enhancers evolved rapidly and that mammalian genomes contain large numbers of evolutionarily young , sometimes species-specific , enhancers [43] , [44] . Therefore , TFs that play a role in enhancer activation might not discriminate between coding and noncoding sequences and eExons might not be under additional evolutionary constraint than typical coding sequences . The activity of enhancers , both coding and noncoding , is regulated by the binding of TFs . The cell-specific TF repertoire regulates enhancer activity levels and depends on motif positioning and larger regulatory context . Techniques such as ChIP-seq can identify the TF binding repertoire of a certain enhancer , but are largely limited in their resolution , technological biases and require a priori knowledge about potential TFs that bind to the enhancer . By testing the regulatory profile of three enhancers in two cell types , we identified specific TFBS overlapping positions that can have a different impact on enhancer activity in another cell type . Moreover , functional positions that do not overlap predicted TFBS could identify novel TFs that control enhancer activity in that cell type . Enhancer sequences can encompass a wide repertoire of TFBS . Our study suggests that a collection of TFs could dictate the activity of enhancers in a specific cell type , while a different collection of TFs could dictate enhancer activity in another cell type .
In addition to raw sequencing reads available in the NCBI Short Read Archive ( SRA ) under accession SRP044727 , a full list of mutations , along with their associated effect sizes and p-values , are provided as Table S5 . Enhancer haplotypes were generated from short , doped oligonucleotides using polymerase cycling assembly ( PCA ) as previously described [17] . Sets of overlapping oligonucleotides for each eExon were designed using DNAWorks [45] . Oligonucleotides were synthesized by Integrated DNA Technologies . All positions corresponding to the enhancer region ( except for the flanking primer landing sites on either side of the enhancer ) were synthesized using a hand-mix doped at a ratio of 97∶1∶1∶1 ( that is , designated base at a frequency of 97% , and every other base at a frequency of 1% ) . The degenerate tags were first cloned to create a complex library of tagged pGL4 . 23 plasmids as described in Patwardhan et al , 2012 [17] . Briefly , the tag oligonucleotide ( TAG_OLIGO ) was made double-stranded using primer extension in a 50 µl reaction volume with 1× iProof Master Mix , 0 . 5 µg single-stranded tag oligo , 0 . 5 µg reverse primer ( TAG_EXTEND ) . The reaction was incubated at 95°C for 3 min , 61°C for 10 min and then 72°C for 5 min . The product was purified using a QIAquick column and eluted in 50 µl EB . It was further subjected to ExoI treatment in 40 µl reaction volume for 1 h at 37°C to degrade any remaining single-stranded DNA , and purified again using QIAquick columns . The resulting double-stranded tag oligo was then cloned into pGL4 . 23 at the XbaI site ( at 1 , 799 bp ) using standard InFusion ( Clontech ) protocol . The InFusion reaction was diluted to 100 µl using TE8 . 1 . 5 µl of the diluted cloning reaction was used to transform 50 µl of chemically competent FusionBlue cells ( Clontech ) using standard protocols . The enhancer haplotypes were then cloned into the tagged pGL4 . 23 plasmids using standard InFusion protocol . 2 . 5 µl of the cloning reaction was used to transform 50 µl of chemically competent cells ( Stellar ) using standard protocol . For each of the three enhancers , eight transformation reactions were pooled and grown overnight in 50 ml liquid culture at 37°C in a shaking incubator . DNA was extracted using the Invitrogen ChargeSwitch Midi Prep Kit . 1–2×105 of HepG2 , HEK293T and HeLa cells ( ATCC ) were cultured in 24 well plates overnight using standard protocols and were transfected with 500 ng of the enhancer candidate cloned into pGL4 . 23 plasmid , along with 50 ng of the Renilla vector , to correct for transfection efficiency , using X-tremeGENE HP DNA transfection reagent ( Roche ) . After 24 hours , the enhancer activity was measured using the Dual-Luciferase reporter assay ( Promega ) on a Synergy 2 microplate reader ( BioTek Instruments ) . In HEK293T , the enhancer variant vectors were co-transfected along with 100 ng of HNF4α2 and HNF4α8 expression constructs for HNF4A [46] and c-Fos and c-Jun for AP-1 , which is a heterodimer of c-Fos and c-Jun [47] . 500 ng of the luciferase reporter construct pZL . HIV . LTR . AI-4 [48] , that contains four tandemerized HNF4 response elements was used as a positive control for HNF4A transfection and 500 ng of pGL4 . 13 ( Promega ) that contains an AP-1 responsive element was used as a positive control for c-Fos and c-Jun transfections [49] . For MPRA libraries , 1–2×106 HeLa cells were cultured on 10 mm culture dishes overnight using standard protocols and transfected with 10 µg of the constructed library . After 24 hours , cells were harvested and total RNA was purified using the RNAeasy maxi prep kit ( Qiagen ) and subjected to mRNA selection ( Oligotex , Qiagen ) following the manufacturer's protocols . For the hydrodynamic tail vein assay , each tested sequence or MPRA library , cloned in the pGL4 . 23[luc2] vector , were injected ( 10 µg ) alongside 2 µg of pGL4 . 74[hRluc/TK] vector to correct for injection efficiency , into five CD1 mice ( Charles River Laboratories ) using the TransIT EE hydrodynamic gene delivery system ( Mirus Bio LLC ) according to the manufacturer's protocol . Negative ( empty pGL4 . 23[luc2] ) and positive ( ApoE liver enhancer [50] controls ( n = 5 ) were also injected in parallel at each injection date/experiment . After 24 h , the mice were euthanized , livers harvested and used to make RNA in order to sequence transcribed tags ( MPRA libraries ) and also to measure luciferase activity ( MPRA libraries and individual constructs ) . Approximately 1gr of mouse liver was used to purify total RNA using the RNAeasy maxi prep and 500 µg were used to select for mRNA . For luciferase measurements , livers were homogenized in passive lysis buffer ( Promega ) , followed by centrifugation at 4°C for 30 min at 14 , 000 rpm . Firefly and Renilla luciferase activity in the supernatant ( diluted 1∶20 ) were measured on a Synergy 2 microplate reader in replicates of six for each liver , using the Dual-Luciferase reporter assay system . The ratios for firefly luciferase:Renilla luciferase were determined and expressed as relative luciferase activity . All mouse work was approved by the UCSF Institutional Animal Care and Use Committee protocol number AN100160 . 20-bp tags were identified in liver/HeLa mRNA using previously described methods [17] . Sixteen RT-PCR reactions were performed for each of the biological duplicates , which were then multiplexed and sequenced on a GAIIx ( Illumina ) using a custom sequencing primer ( TAG_SEQ_F ) . Each run was either 20 or 36 cycles with an additional 6 cycles to read the indexing tag using the index sequencing primer ( TAG_SEQ_INDEX ) . For each aliquot , reads were filtered based on the quality scores for the first 20 bases , which correspond to the degenerate tag . The corresponding number of reads per each tag was counted and only tags that were supported by at least ten reads were used for further analysis . eExon haplotypes were associated with tags as previously described [17] . Briefly , 1 , 000 bp segments separating eExon haplotypes and tags on the pGL4 . 23 plasmid were excised by digesting with HindIII , which cuts both the 3′ of the eExon haplotype , and 5′ of the tag . The digested plasmids were purified and recircularized using intramolecular ligation ( T4 DNA Ligase from NEB ) , resulting in the tag being adjacent to the 3′ end of the eExon haplotype . The region spanning the eExon haplotype and tags was amplified from recircularized plasmids by PCR with the forward primer targeting the region immediately 5′ of the eExon haplotype ( eEXON_COMMON_F ) and the reverse primer targeting the region immediately 3′ of the tag ( TAG_PE_R ) . The amplicons were then subjected to the subassembly protocol as conceptually described in Hiatt et al . [34] . The random fragmentation step was carried out using the Nextera Tn5 transposase ( Illumina ) instead of mechanical shearing . The Nextera reaction was purified using MinElute column ( Qiagen ) and size-selected by PAGE . The size selected fragments were subjected to PCR in 25 ul reaction volume with 1× KapaHiFi Hot Start Ready Mix ( Kapa BioSystems ) , 0 . 5× SYBR Green I , 20% of the size-selected product and each of the primers , Nextera Adapter 1 and TAG_PE_R at 0 . 3 uM final concentration . Thermal cycling was carried out using BioRad Mini Opticon System using the following program: 95°C for 3 mins; and then 30 cycles of 98°C for 20 sec , 65°C for 15 sec , 72°C for 15 sec . The PCR products were purified using QiaQuick column and then sequenced on a Hi-Seq 2000 . Read1 collected 101 bp of the enhancer sequence staring at random breakpoints along the enhancer . Index read collected the 20 bp tag sequence . The reads were then grouped by tag . Reads belonging to each group were then aligned to the wild type enhancer sequence to identify the mutations on the haplotype associated with that tag using custom scripts . All linear regression analyses were done using the lm ( ) or lsfit ( ) functions available in the R Statistical Package as previously described [17] . To quantity the effect of mutation at any given position on the number of aliquots in which an enhancer haplotype was observed , a separate linear regression model was used for every position along the enhancer , with a single predictor representing whether the given position was wild-type or mutant ( univariate model ) . Similarly , a modified model was used to include the three possible nucleotide substitutions at any position that estimates effect sizes for each position being driven by specific nucleotide substitutions ( trivariate model ) . For each position in each enhancer , we constructed a linear model to assess the extent to which the presence of a mutation at that position is predictive of a change in the number of RNA aliquots in which the tag associated with an enhancer haplotype was observed . This is effectively a proxy for its impact on transcriptional activation , i . e . “effect size” . Specifically , we use the term “effect size” to describe the log2 fold change in the predicted transcriptional activity , as measured by the number of RNA aliquots in which a specific haplotype appeared , relative to the wild-type . For each of the three enhancers , we first calculated effect sizes separately on the data from each mouse ( 16 RNA aliquots per mouse ) . The effect sizes for these biological replicates were highly correlated ( Tail-vein: SORL1 exon 17: r = 0 . 88 , PPARG exon 6: r = 0 . 91 , TRAF3IP2 exon 2: r = 0 . 92; HeLa: SORL1 exon 17: r = 0 . 85 , PPARG exon 6: r = 0 . 85 , TRAF3IP2 exon 2: r = 0 . 8 ) . Based on this high reproducibility and to increase resolving power , we performed all subsequent analyses after combining data across both replicates for each enhancer . To quantity the effect of mutation at any given position on the number of aliquots in which an enhancer haplotype was observed , we built a separate linear regression model at every position along the enhancer , with a single predictor representing whether the given position was wild-type or mutant ( Table S3 ) . The predictor was thus a binary variable representing presence ( 1 ) or absence ( 0 ) of a mutation at that position . where , yi = number of aliquots in which the ith haplotype was observed ( referred to as aliquot counts ) , Xij = 1 if position j was mutant and 0 if position j was wild-type in the ith haplotype . To facilitate comparison between positions and between enhancers , we calculated the “effect size” of mutation at a position j as:All linear regression analyses were performed using the lm ( ) or lsfit ( ) functions available in the R Statistical Package . The p-value reported by the model for was used to judge whether the effect size was significant . To explore whether the estimated effect sizes for each position were being driven by specific nucleotide substitutions , we modified the model just described to include three predictors , each representing one of the three possible nucleotide substitutions at that position . The factors were set up as binary variables representing the presence ( 1 ) or absence ( 0 ) of the particular change at that position . Effect sizes were then calculated from the coefficients produced by the models as follows ( for k = 1 , 2 , 3 ) :The p-value reported by the model for was used to judge whether the effect of a given nucleotide substitution at a given position was significant . Putative TFBSs were identified using MATCH [51] to search for motifs in TRANSFAC Release 2010 . 3 [52] . MATCH was run independently on each individual sequence with default parameters . For each position with an observed nucleotide substitution , a 51 bp segment of DNA centered at the position was used for a Position Specific Scoring Matrix ( PSSM ) scan . PSSMs for a subset of liver expressed TFs ( Table S5 ) were obtained from the publicly available JASPAR and TRANSFAC databases . For each PSSM of length L , a ‘mark score’ was calculated for all subsequences of length L within the 51 bp DNA segment that overlapped the central position . A mark score for the reference subsequence ( Sr ) and the mutant subsequence ( Sm ) were calculated as:Where B is the background distribution of nucleotides in the genome . TF motifs were only considered if either Scorer or Scorem was greater than the relative entropy score of the TF . Finally , the TF with the largest absolute score change between Scorer and Scorem is listed in Table S5 . The relative entropy is defined as:where mi , j is entry in row I and column j of the PSSM .
|
Exons that code for protein can also have additional functions , such as regulating gene transcription through enhancer activity . Here , we changed every nucleotide in three different exons that also function as enhancers , and examined their enhancer activity to test whether nucleotide changes in these exons can affect both the protein sequence and enhancer function . We found that mutations with a significant effect on enhancer function can reside both in regions that change the protein sequence ( non-synonymous ) and regions that do not change it ( synonymous ) . When we conducted a similar analysis in a different cell type , we observed a difference in the nucleotide changes that cause a significant effect on enhancer activity , suggesting that the enhancer functional units can differ between tissues .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"sequencing",
"techniques",
"genome",
"expression",
"analysis",
"functional",
"genomics",
"animal",
"genetics",
"genome",
"evolution",
"gene",
"function",
"genome",
"sequencing",
"mutation",
"genome",
"analysis",
"molecular",
"genetics",
"molecular",
"biology",
"techniques",
"gene",
"expression",
"molecular",
"biology",
"transcriptome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"computational",
"biology",
"human",
"genetics"
] |
2014
|
Systematic Dissection of Coding Exons at Single Nucleotide Resolution Supports an Additional Role in Cell-Specific Transcriptional Regulation
|
Haploinsufficiency , wherein a single functional copy of a gene is insufficient to maintain normal function , is a major cause of dominant disease . Human disease studies have identified several hundred haploinsufficient ( HI ) genes . We have compiled a map of 1 , 079 haplosufficient ( HS ) genes by systematic identification of genes unambiguously and repeatedly compromised by copy number variation among 8 , 458 apparently healthy individuals and contrasted the genomic , evolutionary , functional , and network properties between these HS genes and known HI genes . We found that HI genes are typically longer and have more conserved coding sequences and promoters than HS genes . HI genes exhibit higher levels of expression during early development and greater tissue specificity . Moreover , within a probabilistic human functional interaction network HI genes have more interaction partners and greater network proximity to other known HI genes . We built a predictive model on the basis of these differences and annotated 12 , 443 genes with their predicted probability of being haploinsufficient . We validated these predictions of haploinsufficiency by demonstrating that genes with a high predicted probability of exhibiting haploinsufficiency are enriched among genes implicated in human dominant diseases and among genes causing abnormal phenotypes in heterozygous knockout mice . We have transformed these gene-based haploinsufficiency predictions into haploinsufficiency scores for genic deletions , which we demonstrate to better discriminate between pathogenic and benign deletions than consideration of the deletion size or numbers of genes deleted . These robust predictions of haploinsufficiency support clinical interpretation of novel loss-of-function variants and prioritization of variants and genes for follow-up studies .
With array-based copy number detection and the current generation of sequencing technologies , our ability to discover genetic variation is running far ahead of our ability to interpret the functional impact of that variation . Several software tools exist for predicting the phenotypic impact of mutations that change the amino acid sequence of an encoded protein [1] . These tools are essentially proteomic and genomic , rather than genetic , in perspective; no distinction is made between mutations that are dominant or recessive in action . By contrast , there is a lack of tools that predict the phenotypic impact at the organismal level of unambiguous loss-of-function mutations of an encoded protein ( e . g . truncating mutations and whole gene deletions ) . Not all loss-of-function mutations are deleterious , especially when heterozygous . It is generally considered that recessivity is the norm for diploid genomes [2] . Some loss-of-function mutations even confer selective advantages [3] . It is clear from resequenced exomes [4] and genomes [5] and CNV surveys [6] that every genome harbours tens of unambiguous loss-of-function mutations . A pressing clinical need for interpreting genetic variation is in distinguishing between pathogenic and benign copy number variants ( CNVs ) revealed by array-based profiling of patients [7] . With the current resolution of microarrays in clinical practice , these variants are typically large , rare deletions , often encompassing multiple genes . The most obvious pathogenic mechanism for heterozygous loss-of-function mutations ( such as large rare deletions ) is haploinsufficiency ( HI ) , wherein a single functional copy of a gene is insufficient to maintain normal function . Only a few hundred genes have been reported haploinsufficient so far [8] , [9] . Previous studies have shown that gene sets related to haploinsufficiency , such as genes implicated in dominant diseases and genes overlapped by CNVs , have biased evolutionary and functional properties [10]–[12] . However , such investigations often compare those gene sets to the genome average and have been descriptive rather than predictive in scope . We sought to explore further systematic biases in the properties of known HI genes , and to develop a predictive model to assess for each gene in the genome the probability that it exhibits haploinsufficiency with respect to the severe developmental disorders that are the mainstay of clinical genetic practice . As it is not known for most genes in the genome whether or not they exhibit haploinsufficiency , we maximized the power of this predictive approach by assembling a large training set of ‘haplosufficient’ ( HS ) genes that do not exhibit haploinsufficiency resulting in severe developmental anomalies . We reasoned that currently the most effective way of screening for HS genes is use robust CNV discovery to identify genes that are wholly or partially deleted among thousands of adults recruited as controls for genome-wide association studies . We take advantage of the fact that the impact of large deletions on coding sequence is more unambiguously interpretable than other types of genetic variation , such as point mutations or small insertion/deletions .
We investigated how our gene-based predictions of haploinsufficiency might be used to discriminate between benign and pathogenic genic deletions . We considered that a natural way to score the probability of a deletion causing a haploinsufficiency phenotype is to generate a LOD ( log-odds ) score comparing the probability that none of the genes covered by the deletion will cause haploinsufficiency with the probability that at least one of the genes will cause haploinsufficiency , as shown schematically in Figure 7 . This LOD score is calculated using the formula below:Higher LOD scores indicate deletions that are more likely to be pathogenic as a result of haploinsufficiency . Note that this score assumes that there is no statistical interaction between the genes . We then considered how these deletion-based haploinsufficiency scores might be used to assess whether a genic deletion observed in a patient might cause their disease . One way of framing probabilistically this intuitively simple question is to estimate the opposing probability , that the deletion is unrelated to the patient's disease status . We can relate this to the probability of drawing an individual at random from a healthy control population with a deletion at least as pathogenic as the deletion in the patient . We can estimate this probability empirically as the proportion of healthy controls with a genic deletion having the same or greater haploinsufficiency score . To test this approach , and to avoid circular reasoning , we generated a set of gene haploinsufficiency predictions using a slightly smaller set of HS genes identified in a large subset of the GWAS controls . The performance of the HI predictions using the slightly smaller HS gene training data was very similar to that of the full predictive model described in the previous section , as assessed by ten-fold cross validation ( Text S1 ) . We then identified LOF deletions in the remainder ( N = 2 , 322 ) of the GWAS controls [17] , which had not been used to train the predictive model , and determined the distribution of the maximal deletion haploinsufficiency score ( based on the new gene haploinsufficiency predictions ) observed in each control individual . We investigated whether the distribution of maximal LOD scores is significantly different between European-American ( E-A ) and African-American ( A-A ) GWAS controls . We observed that there was not a significant difference in median haploinsufficiency scores in E-A and A-A populations ( p = 0 . 71 , Mann-Whitney U test ) , although the E-A controls have a slightly longer tail of more pathogenic deletions ( e . g . a higher proportion of E-A controls have deletions with LOD scores greater than 3 , Table S4 ) . The 50% , 90% , 95% and 99% percentiles of the distribution for maximal LOD score for E-A and A-A controls cohorts are listed in Table 1 . We calculated a LOD score for each of 487 pathogenic de novo deletions submitted by clinical geneticists to the DECIPHER database [18] . We focused exclusively on deletions known to be de novo variants , as we infer that their pathogenicity has been ascribed primarily on the basis of their inheritance status . The distributions of maximal LOD scores in GWAS controls and LOD scores of pathogenic DECIPHER deletions are shown in Figure 7 . The pathogenic deletions have strikingly significantly higher LOD scores than deletions observed in GWAS controls ( p<1e-30 , Mann-Whitney U test ) . We observed that for 92% of the pathogenic deletions there was a probability of less than 5% of drawing an individual at random from our control population with a genic deletion of equal or greater LOD score , and for 83% of pathogenic deletions there was a less than 1% probability . We computed ROC curves to compare three different approaches for discriminating between pathogenic deletions and deletions seen in controls: ( i ) our LOD scores , ( ii ) the length of the deletion , and ( iii ) the number of genes in the deletion ( Figure 8 ) . These ROC curves clearly show that the haploinsufficiency LOD score is the best metric for discriminating between pathogenic deletions in patients and deletions seen in controls . We provide a script and input files to calculate LOD scores and make comparisons with control data ( Protocol S1 ) . We investigated whether the gene-based probabilities of haploinsufficiency that we have generated are of general utility across different forms of genetic variation . If this is indeed the case then we should expect that genes harbouring loss-of-function substitutions or small indels in apparently healthy individuals should not have a high p ( HI ) . We identified 349 genes as having LOF substitutions and indels in 12 recently sequenced exomes [4] , of these , we could estimate p ( HI ) for 176 that were not also in the HS training set ( and thus represent a fair set for independent comparisons ) . These genes are highly significantly enriched among genes with low probabilities of exhibiting haploinsufficiency ( p = 1 . 06e-20 when comparing to the genome , and p<1e-30 when comparing to known HI genes , Mann-Whitney U test ) . This result implies that there are not substantial differences between genes that tolerate whole gene deletions and those that tolerate smaller loss-of-function variants . Moreover , by studying the allele frequency spectrum apparent in a large gene-resequencing dataset that has been extensively studied for patterns of selective constraint [19]–[21] we observed that nonsynonymous variants in genes more likely to exhibit haploinsufficiency are highly significantly skewed towards rarer variants than nonsynonymous variants in genes less likely to exhibit haploinsufficiency , in both African Americans and European Americans ( p = 2 . 9e-7 and 4 . 0e-3 respectively , one-sided Mann-Whitney U test , see also Figure S9 ) reflecting greater , on average , selective constraint on genes we predict to exhibit haploinsufficiency ( Text S1 ) .
We have undertaken a systematic characterization of human haploinsufficient genes by contrasting them with a set of haplosufficient genes derived from non-pathogenic CNVs , developed a prediction model on the basis of the most significant differences , and assigned predicted probability of being haploinsufficient to more than half of the protein-coding genome . Our finding that functional interaction with known HI genes was the single most predictive property of HI genes probably reflects the modularity of the interaction network , suggesting certain pathways or biological processes , such as early development morphogenesis , being more sensitive to dosage changes than others . However , it is also possibly influenced by an ascertainment bias with which HI genes are discovered . The accuracy of our haploinsufficiency probabilties is limited by a number of factors , such as imperfect training data , although we have taken considerable steps to limit false positives , and missing data among predictor variables in our model . As network proximity to known HI genes is the single most predictive variable we eagerly await the construction of networks with increasing coverage and completeness , in the expectation that it should improve our prediction power . The gene coverage of our method could potentially be increased by using multiple imputation approaches to impute missing data [22] . To trial this approach , we imputed missing predictor variables using the predictive mean matching method , which allowed us to increase considerably the number of genes for which we could predict haploinsufficiency from 12 , 443 to 17 , 456 ( Text S1 , Table S5 , Dataset S2 ) . The resultant increase in size of training data also led to a slight improvement in prediction accuracy ( AUC increased from 0 . 81 to 0 . 83 , Figure S10 ) , and we observed similar levels of enrichment of known dominant genes and mouse haploinsufficient genes pre- and post-imputation ( Figure S11 and Figure S12 ) , suggesting that multiple imputation is a reliable method to increase genome coverage . Another limitation of our method is the broad phenotypic outcome being predicted . Essentially , we are contrasting HI genes that cause a broad range of developmental disorders , with HS genes for which haploidy does not majorly impair an individual's ability to participate as a control in a genome-wide association study . We note that this broad phenotype is nevertheless considerably more constrained than that being considered by prediction algorithms based on evolutionary conservation , which are essentially integrating any deleterious phenotype manifested among any of the environments encountered during millions of years of evolution , across all possible modes of inheritance and genetic backgrounds . Despite the breadth of the phenotype implicit within conservation-based predictions , this class of algorithms has been demonstrated many times to be of appreciable utility [19] , [23] . To support clinical interpretation of deletions seen in patients , we have transformed our gene-based predictions of haploinsufficiency into haploinsufficiency scores for individual deletions . Currently , clinicians typically use the length of a deletion or the number of genes deleted to assess the pathogenicity of a deletion . We have shown that our pathogenicity scores represent a superior metric to these existing approaches for classifying pathogenic deletions . We believe that the most appropriate use of these deletion-based haploinsufficiency scores is to compare deletions seen in patients with those seen in controls , and that quantifying the fraction of control individuals with a deletion at least as pathogenic as that seen in a patient provides a rational basis to classify pathogenic deletions . This fraction represents the probability of observing such a deletion by chance and thus the probability that a deletion will have been misclassified as pathogenic . A clinician can therefore set a particular haploinsufficiency score threshold to define pathogenic deletions through considering the misclassification rate with which they are comfortable . We have provided the necessary software tools to allow these haploinsufficiency scores to be calculated for any genic deletion , and automated calculation of these LOD scores and comparison with control deletions will be integrated into the forthcoming release ( v5 ) of the DECIPHER database ( personal communication: Helen Firth , Nigel Carter , Manual Corpas ) , which is used by over 170 clinical centres worldwide to interpret chromosomal rearrangements seen in patients ( the gene-based predictions ( p ( HI ) ) are already available in the current release of DECIPHER , Figure S8 ) . We only observed subtle differences in the distribution of haploinsufficiency scores seen in European-American and African-American populations ( Table S4 ) , which might reflect a higher fraction of deleterious alleles in populations with non-African ancestry . Further investigation of these differences is warranted to see whether the haploinsufficiency scores observed in a patient ought to be compared with controls from a matched population . It has recently been suggested that some developmental disorders result from the presence of two independent deletions in the same genome , the ‘two-hit’ hypothesis [24] . This hypothesis suggests a subtly different assessment of a patient's CNV data is required , through considering the question: ‘is the SET of deletions observed in my patient causal of their disease’ . Another way of viewing this important question is that it requires consideration of the genome-wide haploinsufficiency ‘burden’ rather than the haploinsufficiency scores of individual deletions . The probabilistic framework we have established for assessing pathogenicity of individual deletions naturally extends to this situation . Rather than combining the haploinsufficiency probabilities for individual genes within a deletion to calculate a haploinsufficiency score for that variant , we can combine the haploinsufficiency probabilities for all deleted genes in the genome to calculate a haploinsufficiency LOD score for that genome , and compare this genome-wide haploinsufficiency score with those observed in healthy controls to assess the probability of sampling a healthy individual with a genome with a haploinsufficiency burden at least as high as that in the patient . This approach also naturally extends to an assessment of the genome-wide haploinsufficiency burden from other classes of LOF mutation . One requirement of the haploinsufficiency score approach for assessing pathogenicity of individual deletions is that data quality between patients and controls is similar . If there are systematic differences between the sensitivity and specificity of CNV ascertainment in patients and controls then this may lead to biased comparisons of haploinsufficiency scores . This potential limitation is largely mitigated by an inherent focus on the largest deletions in a genome , which are typically long enough for many different technology platforms to have essentially complete sensitivity at very high specificities . In addition to the use of the deletion-based and genome-wide haploinsufficiency scores described above , we envisage that our gene-based predictions of haploinsufficiency have two additional applications: ( i ) prioritization of variants for follow-up studies and ( ii ) integration into association testing to increase power; and we consider each of these these in turn . First , our predictions of haploinsufficiency provide a rational basis for prioritizing heterozygous variants for follow-up genetic and/or functional studies . The burden of having to validate increasing numbers of benign variants is an appreciable barrier to the translation into clinical practice of genomic technologies of ever-increasing resolution . A method that accurately hones in on potential causal variants could alleviate this burden considerably . The prioritization of variants need not be restricted solely to unambiguous loss-of-function variants . Most rare functional variants in any given genome are heterozygous nonsynonymous substitutions , many of which result in a complete or partial loss-of-function of the encoded gene . We contend that the prediction power of popular methods for predicting the functional impact of nonsynonymous substitutions from structural information and evolutionary conservation , such as SIFT [25] and POLYPHEN [26] , is limited by an inability to discern from cross-species alignments whether purifying selection at a given site is acting in a recessive , additive or dominant manner . Combining these genotype-oblivious predictions with our predictions of haploinsufficiency , should enable rational , genotype-aware prioritization of heterozygous nonsynonymous variants . The second application of these predictions is to integrate them directly into association testing . It has been suggested that weighting variants by their probability of having a functional impact should improve power in resequencing studies to detect functional units ( e . g . genes , pathways ) enriched for functional variants [27] . As noted above , most of the rare variants considered in these studies are only observed in the heterozygous state , thus , if functional , they have to be exerting a dominant or semidominant effect , and predictions of haploinsufficiency , because haploinsufficiency is a major mechanism underlying dominance , are a highly relevant weighting factor . The framework that we have developed that integrates functional , evolutionary and genomic properties of genes , could , by judicious selection of different training datasets be easily and broadly extended to include other classes of variant ( e . g . duplications , gain of function mutations ) , different genetic models ( e . g . recessive effects ) and different , and potentially more specific , phenotypic outcomes ( e . g . disease-specific ) .
We compiled a set of CNVs from three genotyping datasets generated from Affymetrix 6 . 0 platform , 210 unrelated HapMap individuals [28] , 2 , 421 control individuals used in GWAS studies of bipolar and schizophrenia [17] and 6 , 000 individuals participating WTCCC2 as common controls , using Birdsuite [29] . All these CNVs were annotated against EnsEBML protein-coding gene annotation build 50 [30] . Genes with all transcripts satisfying one of the following criteria: deletion of over half of the coding sequence , deletion of the start codon , deletion of the first exon , deletion of splice-signal and deletion that causes frame-shift , were considered loss-of-function ( Figure S1 ) and for the gene to be included as haplosufficient , such events are required to occur in at least two apparently healthy individuals . For continuous variables , the two-tailed Mann-Whitney U test was performed to assess if positive ( haploinsufficient ) and negative ( haplosufficient ) training data have the same median value for potential predictor variables . For two-class categorical features , Fisher's exact tests were performed . Statistical tests were performed using R ( http://www . r-project . org ) . We assessed different potential sets of predictor variables for input into the predictive model using the following criteria: ( i ) they allow prediction for at least half the genes in the genome , ( ii ) the Spearman correlation between all pairs of predictor variables is less than 0 . 3 , ( iii ) they are drawn from different broad categories ( genomic , evolutionary , functional and network ) if possible , iv ) achieve best performance in model assessment ( see below ) . The sensitivity of the prediction was plotted against ( 1 - specificity ) and the area under the ROC curve ( AUC ) [44] was used as quantitative measure of the performance of the model , where sensitivity = , and specificity = . The other measure used is the Matthews correlation coefficients ( MCC ) [45] , defined as:To avoid over-fitting , the sensitivity and specificity were calculated using 10-fold cross-validation . To overcome the variability caused by random partition involved in 10-fold cross-validation , each such assessment was repeated 30 times and the mean values were reported .
|
Humans , like most complex organisms , have two copies of most genes in their genome , one from the mother and one from the father . This redundancy provides a back-up copy for most genes , should one copy be lost through mutation . For a minority of genes , one functional copy is not enough to sustain normal human function , and mutations causing the loss of function of one of the copies of such genes are a major cause of childhood developmental diseases . Over the past 20 years medical geneticists have identified over 300 such genes , but it is not known how many of the 22 , 000 genes in our genome may also be sensitive to gene loss . By comparing these ∼300 genes known to be sensitive to gene loss with over 1 , 000 genes where loss of a single copy does not result in disease , we have identified some key evolutionary and functional similarities between genes sensitive to loss of a single copy . We have used these similarities to predict for most genes in the genome , whether loss of a single copy is likely to result in disease . These predictions will help in the interpretation of mutations seen in patients .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/population",
"genetics",
"evolutionary",
"biology/genomics"
] |
2010
|
Characterising and Predicting Haploinsufficiency in the Human Genome
|
The trimeric envelope ( Env ) spike is the focus of vaccine design efforts aimed at generating broadly neutralizing antibodies ( bNAbs ) to protect against HIV-1 infection . Three recent developments have facilitated a thorough investigation of the antigenic structure of the Env trimer: 1 ) the isolation of many bNAbs against multiple different epitopes; 2 ) the generation of a soluble trimer mimic , BG505 SOSIP . 664 gp140 , that expresses most bNAb epitopes; 3 ) facile binding assays involving the oriented immobilization of tagged trimers . Using these tools , we generated an antigenic map of the trimer by antibody cross-competition . Our analysis delineates three well-defined epitope clusters ( CD4 binding site , quaternary V1V2 and Asn332-centered oligomannose patch ) and new epitopes at the gp120-gp41 interface . It also identifies the relationships among these clusters . In addition to epitope overlap , we defined three more ways in which antibodies can cross-compete: steric competition from binding to proximal but non-overlapping epitopes ( e . g . , PGT151 inhibition of 8ANC195 binding ) ; allosteric inhibition ( e . g . , PGT145 inhibition of 1NC9 , 8ANC195 , PGT151 and CD4 binding ) ; and competition by reorientation of glycans ( e . g . , PGT135 inhibition of CD4bs bNAbs , and CD4bs bNAb inhibition of 8ANC195 ) . We further demonstrate that bNAb binding can be complex , often affecting several other areas of the trimer surface beyond the epitope . This extensive analysis of the antigenic structure and the epitope interrelationships of the Env trimer should aid in design of both bNAb-based therapies and vaccines intended to induce bNAbs .
The HIV-1 envelope glycoprotein ( Env ) , a trimer comprising three gp120 and gp41 subunits , is the target of broadly neutralizing antibodies ( bNAbs ) that are known to prevent virus infection in animal models . The induction of bNAbs by vaccines is a highly desirable , but not yet achieved , goal . bNAbs isolated from HIV-1 infected individuals are templates for Env-based vaccines [1] , and may also be useful as therapeutics [2] . Around 20% of infected people generate bNAbs [1]; their emergence usually takes at least 2 years , but can sometimes occur within a year [3] . Most bNAbs recognize epitopes in four well-defined clusters . They include PG9/16 , PGT141-145 , CH01-04 and VRC26 ( gp120; quaternary structure-dependent V1V2-glycan ) , b12 , VRC01 , VRC03 , PGV04 , HJ16 , CH31 , CH103-106 , 3BNC60 , 3BNC117 , 12A12 , NIH45-46 ( gp120; CD4 binding site; CD4bs ) , PGT121-123 , PGT125-130 , PGT135-137 , 10–1074 ( gp120; Asn332-centered oligomannose patch ) , and 2F5 , 4E10 , 10E8 ( gp41; membrane-proximal external region; MPER ) [reviewed in [1]] , [4] . However , several bNAbs against new quaternary structure-dependent epitopes have now been isolated . The PGT151-158 , 35O22 and 8ANC195 bNAbs interact with diverse epitopes at the gp120-gp41 interface [5–8] . Their quaternary structural requirements mean that they bind only , or far better , to soluble trimers that adopt a native-like conformation than to gp120 monomers or uncleaved , non-native gp140 proteins [5 , 6 , 8 , 9] . The 3BC315 bNAb was originally reported to target a CD4-induced gp120 epitope [10] , but its epitope is now known to be on gp41 [11 , 12] . The precise relationships among the various bNAb epitope clusters on the Env trimer are not fully understood . An antibody cross-competition analysis helped to define the antigenicity of the gp120 subunit [13] . However , trimerization alters the conformation , surface accessibility and antigenicity of gp120 , and hence many Abs that bind well to gp120 cannot recognize the trimer ( i . e . non-neutralizing Abs; non-NAbs ) . Accordingly , we elected to conduct a comprehensive analysis of the antigenicity of bNAb epitopes on the trimer , and their relationships . The recent development of recombinant , soluble BG505 SOSIP . 664 trimers that antigenically mimic native , virion-associated Env made this study possible [12 , 14–16] . These trimers express all epitopes for bNAbs that neutralize the parental virus , except those within the MPER [4–6 , 8 , 12 , 14 , 17–20] , that are not included in the construct . Their high resolution structures in complexes with bNAbs PGV04 and PGT122 were solved by cryo-electron microscopy ( EM ) and X-ray crystallography , respectively [18 , 19] , and a higher resolution X-ray structure of a complex with PGT122 and 35O22 is now available [21 , 22] . While the BG505 SOSIP . 664 trimers are not identical to the native Env spike on viruses due to stabilizing mutations and truncation of the MPER they are excellent mimics based on extensive antigenic and structural data . Here , we describe a comprehensive antigenicity analysis of the same trimers using an antibody cross-competition ELISA and a panel of bNAbs against all known epitope clusters ( except MPER ) . Key ELISA-derived observations were corroborated and extended by surface plasmon resonance ( SPR ) . We also prepared low resolution , negative stain EM reconstructions of the Env trimers in complex with CH103 , CH106 , 1NC9 , 3BNC117 and VRC01 . The new EM data , as well as published structures , allowed us to interpret the cross-competition data in a structural context and thereby create an antigenic map of the trimer surface . Using the full set of data , we found that bNAbs could inhibit one another’s binding by four different mechanisms: 1 ) epitope overlap , 2 ) steric inhibition , 3 ) allosteric inhibition , 4 ) glycan re-orientation . These findings enhance our understanding of how HIV-1 is neutralized , how the design of vaccines intended to induce bNAbs can be improved , and how bNAbs might be used therapeutically .
The goal of our cross-competition analysis was to characterize the interrelationships among bNAb epitopes on the trimer surface . In the assay , D7324-epitope-tagged , BG505 SOSIP . 664 trimers [12] were immobilized on ELISA wells , an excess of non-biotinylated competitor bNAbs was incubated for 30 min with the immobilized trimers , and the biotinylated analyte bNAb was then added and its binding quantified ( for pilot experiments , see S1A Fig ) . We selected 24 bNAbs against all known epitope clusters ( other than the MPER ) , of which 19 could successfully be biotin-labeled . Five bNAbs ( 1NC9 , 8ANC195 , PGT151 , 35O22 , 3BC315 ) lost trimer reactivity after biotinylation presumably because key primary amines on lysine residues were modified . For those five bNAbs , we used Fabs as competitors for unlabeled IgG antibodies and detected the Fc-regions of the latter . We note that the use of Fabs might result in less steric competition compared to IgG , but , in vivo , the IgG competition is more relevant . In the resulting cross-competition matrix , we plotted the percentage of residual bNAb binding in the presence of an excess of the competitor ( Fig . 1 ) . The EC50 values for each test antibody , and hence , the molar excess , in the competition are provided in S1 Table . The epitope clusters were arranged in the matrix from top-to-bottom according to how they are approximately oriented on the trimer: the quaternary , trimer-apex cluster is at the top , the oligomannose patch and CD4bs clusters are located lower down , and finally , the gp120-gp41ECTO interface and the exclusively gp41ECTO epitopes are at the bottom . The MPER epitopes at the base of the trimer were excluded because they are not present on the SOSIP . 664 construct [12 , 23 , 24] . Note that the oligomannose patch is subdivided into V3-glycan and outer domain ( OD ) -glycan sub-clusters , based on the results of this study and reference [1] . Within each epitope cluster , the individual bNAbs were arranged according to their cross-competing properties . Self-competition , plotted along the diagonal , serves as a positive control and was generally very strong ( <25% residual binding ) . Three exceptions were 1NC9 , 8ANC195 and 35O22 , for which Fabs served as competitors for complete IgGs ( as noted , the biotin-labeled MAbs were non-reactive ) . We therefore excluded all ELISA data derived using these three Fabs , because they could not be properly interpreted . For 8ANC195 and 35O22 , we instead used SPR ( see below ) to obtain information on the gp120-gp41ECTO cluster . As expected , there were strong competitive effects between MAbs within each known epitope cluster . For example , within the V1V2-glycan trimer apex cluster , PG9 , PG16 and PGT145 were strongly and reciprocally competitive . The PGT121-123 and PGT125-128 family members all depend on the glycan at Asn332 and residues at the base of V3 . All these bNAbs competed very efficiently with each other . The 2G12 , PGT135 and PGT136 bNAbs also recognize the Asn332 glycan , but their epitopes also involve non-V3 components of the gp120 OD-glycan . 2G12 strongly competed with PGT135 and PGT136 but not vice versa , perhaps because PGT135 and PGT136 bind the trimer relatively poorly compared to 2G12 ( >10-fold higher EC50 ) and so do not compete efficiently [12] . We saw moderate but consistent and mostly reciprocal competition between bNAbs from the oligomannose patch cluster , which is consistent with these antibodies all targeting a “supersite of vulnerability” centered around the Asn332 glycan , albeit in subtly different ways [20] . Thus , different bNAbs have distinct binding motifs on different sides of this glycan , and their angle of approach can also vary [20 , 25] . Notably , competition between the OD-glycan and V3-glycan sub-cluster members was not as strong as that seen within the two individual sub-clusters . The implication is that the “Asn332 supersite” does indeed contain two bona fide sub-sites of bNAb epitopes ( Fig . 1 ) . Most bNAbs from the CD4bs epitope cluster competed strongly and reciprocally with each other . The binding of 1NC9 , originally classified as a CD4bs-targeting bNAb [26] but not characterized extensively , was inhibited by every known CD4bs bNAb . This outcome supports the assignment of 1NC9 to the CD4bs cluster , but it also has some distinctive and unusual properties ( see below ) . The PGT151 , 35O22 and 8ANC195 bNAbs recognize three new epitopes at the gp120-gp41ECTO interface [5 , 6 , 8 , 26] . As none of them retained trimer reactivity upon biotinylation , we used the Fab vs . IgG competition method . As noted above , self-competition was negligible for 8ANC195 and weak for 35O22 , probably attributable to avidity issues; we therefore excluded all 8ANC195 and 35O22 Fab competition data from the matrix and instead relied on SPR ( see below ) . In the competition ELISA , PGT151 Fab strongly inhibited 8ANC195 IgG binding to the trimer , which implies a substantial interrelationship between their epitopes . In contrast , the lack of competition between PGT151 and 35O22 suggests that these two epitopes overlap minimally or not at all . The 3BC315 bNAb did not cross-compete with any other antibody , which is consistent with its epitope being located on gp41 . However , the lack of competition between 3BC315 and 35O22 was unexpected because the two epitopes overlap slightly [11] . The generally positive inhibition pattern along the diagonal axis of the matrix confirms existing knowledge of the test antibodies and their assignment to defined epitope clusters . In contrast , the off-diagonal inhibition patterns provide new clues about the physical proximity of , or functional relationships among , the different epitope clusters , and/or about the allosteric effects that arise when bNAbs bind the trimer . The V3-glycan bNAbs PGT121 , PGT122 and PGT123 inhibited binding of the V1V2-glycan cluster bNAbs ( PG9 , PG16 , PGT145 ) relatively strongly , but in a non-reciprocal manner . PGT125 , PGT126 and PGT128 , also in the V3-glycan cluster , did not , however , block binding of the V1V2-glycan cluster bNAbs . The structure of the BG505 SOSIP . 664 trimer complex with PGT122 illustrates that PG9 , PG16 and PGT122 have slightly overlapping epitopes that all involve the Asn156 glycan [18] . We hypothesized then that steric occlusion via the constant domain of the Fabs also plays a role , providing a steric-based explanation for the cross-cluster competition . However , the non-reciprocal competition seen between , e . g . , the PGT121-123 family and PG9 , PG16 and PGT145 might be at least partially explained by the stoichiometry of binding . Thus , when three bNAbs of the PGT121-123 family ( trimer stoichiometry = 3 ) are bound to the trimer , PG9 , PG16 or PGT145 may have a very limited opportunity to gain access to their epitopes on its apex . However , when a single trimer-apex bNAb binds ( trimer stoichiometry = 1 ) , the PGT121-123 epitopes on one or two protomers could still be available for binding . Yet another possibility is that binding of PGT121-123 antibodies stabilizes the closed state of the trimer and limits the conformational flexibility required for the V1V2 apex antibodies to access their epitope on the trimer interface . Indeed , some bNAbs are now known to stabilize the closed conformation of Env [27 , 28] . We do not yet have an explanation for the strong , but non-reciprocal , competition between PG16 and PGT135 . The OD-glycan and CD4bs bNAbs competed to a moderate extent , but on a non-reciprocal basis ( Fig . 1 ) . A good example is that PGT135 and PGT136 inhibited the binding of most CD4bs bNAbs , but not vice versa ( see below ) . We also observed competitive effects between bNAbs to the V3-glycan and CD4bs clusters . Thus , PGT125 , PGT126 and PGT128 , as well as PGT135 and PGT136 , all inhibited CH31 binding; and PGT128 had the same non-reciprocal inhibitory effect on VRC01 , an outcome that was confirmed by SPR ( see below ) . The binding of 8ANC195 , which recognizes an epitope at the gp120-gp41ECTO interface , was impeded by several CD4bs bNAbs ( Fig . 1 ) , suggesting some relationship between these two sites on the trimer . The most striking off-diagonal inhibition pattern was the non-reciprocal competition between PGT145 and 1NC9 , 8ANC195 and , to a lesser extent , PGT151 and 35O22 . This outcome was surprising because these epitope clusters are not in close proximity on the trimer structure . The binding of PGT145 , which has an absolute specificity for native trimers [9 , 12 , 29] , to the apex may induce allosteric changes elsewhere in the trimer , or else stabilize the structure such that access to certain other epitopes is hampered by the reduced conformational flexibility . If so , the effect is highly specific to PGT145 as the same pattern was not seen using PG9 and PG16 , which also recognize epitopes at the trimer apex . Any changes in the trimer induced by PGT145 binding were reversible , as PGT145- and 2G12- affinity purified BG505 SOSIP . 664 trimers bound 1NC9 and PGT151 comparably ( S2 Fig ) . We found several examples of one bNAb enhancing the trimer binding of another , either reciprocally or uni-directionally . For example , the V3-glycan bNAb PGT128 bound more strongly in the presence of CD4bs bNAbs , 8ANC195 or 3BC315 . There was reciprocally enhanced binding between the CD4bs bNAb NIH45-46 and the V3-glycan bNAbs PGT123 and PGT128; and between the gp120-gp41ECTO bNAb 8ANC195 and the V3-glycan bNAb PGT126 . Non-reciprocal enhancement between NIH45-46 with PGT121 and PGT123 was seen . The CD4bs bNAbs 3BNC117 , NIH45-46 , CH103 and 1NC9 , and the gp120-gp41ECTO interface bNAb 8ANC195 increased the binding of several V1V2-glycan , V3-glycan or OD-glycan bNAbs , for reasons that are currently under investigation . When we used soluble CD4 ( sCD4 ) as the competitor ligand , we saw the expected strong inhibition of CD4bs bNAbs ( Fig . 2A ) . The V1V2-glycan bNAbs also bound markedly less well in the presence of sCD4 , which probably reflects how CD4 binding opens up the trimer apex and thereby impairs or eliminates the associated epitopes [30] . The binding of the V3-glycan bNAb PGT125 was also strongly inhibited , the family members PGT126 and PGT128 to a lesser extent . The results for PGT126 and PGT128 are consistent with observations made using sCD4 , the V3-glycan bNAbs and monomeric gp120 [31] . Again as expected , Fabs of the various CD4bs bNAb strongly competed with CD4-IgG2 for trimer binding ( Fig . 2B ) . The V1V2-glycan bNAb PGT145 also prevented CD4-IgG2 binding , consistent with its induction of long-range allosteric effects , while the V3-glycan bNAbs PGT122 , PGT125 , PGT126 and PGT128 markedly but incompletely inhibited CD4-IgG2 . Finally , the gp120-gp41ECTO interface bNAb PGT151 ( but not the other two members of this cluster ) strongly inhibited CD4-IgG2 binding . While the three bNAbs do all recognize gp120-gp41ECTO interfacial regions , their epitopes are spatially separated [8] . The structural data on the PGT151-trimer complex is consistent with its inhibitory effect on CD4-IgG2 binding [5] . To investigate whether bNAbs could induce CD4-like conformational changes , we used the CD4i antibody 17b , a non-NAb , as a probe ( Fig . 2C ) . The trimer binding of the CD4bs bNAbs VRC01 , 3BNC60 and NIH45-46W increased the subsequent binding of 17b , although only to a modest extent ( 9% for VRC01 or 3BNC60 , 16% for NIH45-46W , in a scale where sCD4 is 100% ) . Similarly to sCD4 , VRC01 induces substantial conformational changes in monomeric gp120 that better expose the 17b CD4i epitope , but it differs from sCD4 in not having that effect on the functional , virion-associated Env trimer [32 , 33] . Hence the modest induction of the 17b epitope when VRC01 binds to the BG505 SOSIP . 664 trimer ( 9% of the magnitude seen using sCD4 ) more closely reflects what is seen with native trimers than with monomeric gp120 . Of note is that the OD-glycan bNAbs PGT135 and PGT136 also enhanced 17b binding ( by 12% for PGT135 and 16% for PGT136 , again compared to 100% for sCD4 ) , but no other bNAbs had this effect . Conversely , bNAbs from the V3-glycan cluster reduced the binding of both CD4i antibodies , which is consistent with the close proximity of the relevant epitopes on the trimer [34] . We used SPR to further explore and extend selected ELISA-derived observations . His-tagged BG505 SOSIP . 664 trimers were immobilized on the chip via an anti-His Ab , as previously described [35] . In a new SPR competition format , two analytes ( NAbs ) were sequentially injected in a single cycle , and the two association curves recorded , the dissociation curve only after the second injection . For comparison , the association-dissociation curves from the individual analyte injections were superimposed on the curve for the second injection of the same analyte . The resulting plots are interpreted as follows: If the second curve for the sequential binding superimposes well on the individual binding curve there is no competition , whereas if the individual-injection curve is higher than the curve for the second injection in the sequential binding , there is competition . The conditions were adjusted such that the self-competition was nearly complete ( Fig . 3 ) . We were particularly interested in bNAbs 8ANC195 , PGT151 and 35O22 against the gp120-gp41ECTO interface because the corresponding ELISA datasets for 8ANC195 and 35O22 were incomplete . In the SPR assay , 8ANC195 , PGT151 and 35O22 all showed strong self-competition ( Fig . 3 ) . When PGT151 and 35O22 were added to the 8ANC195-trimer complex , strong competition was again seen ( Fig . 3A , middle and right panel ) . The same was true when 8ANC195 was added to the PGT151-SOSIP . 664 complex , which is consistent with the competition seen by ELISA ( Fig . 3B , left panel ) . However , 35O22 did not compete with the PGT151-SOSIP . 664 complex , which is again consistent with ELISA data ( Fig . 3B , right panel and Fig . 1 ) . There was bidirectional competition between 8ANC195 and 35O22 for trimer binding in the SPR format but , in contrast , PGT151 bound efficiently to the 35O22-trimer complex ( Fig . 3C , middle panel ) . Overall , the SPR data show that 8ANC195 cross-competes reciprocally with both PGT151 and 35O22 , whereas PGT151 and 35O22 do not cross-compete with each other . The implication is that 8ANC195 binds the trimer at a location between the PGT151 and 35O22 epitopes . A few other selected ELISA-derived observations were also checked using SPR; with only minor and generally subtle exceptions , the two datasets were in good agreement ( S3 Fig ) . To gain more insight into the relationships among bNAb epitopes in the trimer context , we used existing low resolution structures of several bNAb complexes with the BG505 SOSIP . 664 trimer [5 , 8 , 12 , 14 , 17 , 18 , 20 , 36] . We also prepared similar de novo reconstructions of these trimers in complexes with CD4bs bNAbs CH103 , CH106 , 1NC9 , 3BNC117 and VRC01 . The 2D class averages and 3D reconstructions show all five antibodies recognize the trimers in similar ways , although with important differences ( S4 , S5A Figs ) . Cross-correlation coefficient analyses of the 3D map fitting show that CH103 and CH106 have an almost identical angle of approach that is different from those taken by VRC01 , 1NC9 and 3BNC117 ( S5B Fig ) . The 3BNC117 and 1NC9 bNAbs have a higher cross-correlation coefficient with VRC01 than do CH103 or CH106 , suggesting there are subtle differences in the epitopes of these pairs of antibodies . Docking the CH103 crystal structure into the EM map indicates that gp120 has rotated relative to the ground state , and that the trimer is now in a slightly more open conformation . A study of several bNAbs using BG505 SOSIP . 664 mutant trimers revealed that the 1NC9 and 3BNC117 epitopes differ slightly from those for more typical CD4bs bNAbs ( i . e . , VRC01 , PGV04 , CH103 ) by their dependence on the Asn276 glycan ( S7 Fig ) . Combining new and published EM data allowed us to assess how bNAbs interact with the trimer , yielding a qualitative estimate that approximately 60% of the trimer surface is covered by at least one known bNAb ( Fig . 4A ) . When the various epitope clusters and the footprints of individual bNAbs were modeled onto the 3D trimer structure ( Fig . 4B , C ) , the resulting models help to explain several outcomes of the cross-competition analysis . For example , we saw nonreciprocal competition between the V1V2-glycan bNAbs ( e . g . PG9 ) and the V3-glycan bNAbs ( e . g . PGT122 ) ( Fig . 1 ) . The 3D model indicates that PG9 and PGT122 have slightly overlapping epitopes centered around Asn156 ( Fig . 5A ) [18] . The model also confirms why the competition between these two bNAbs is unidirectional ( i . e . , PGT122 impedes PG9 , but not the converse ) ; the outcome arises from the binding stoichiometry considerations referred to above . Inspection of the Asn332 supersite shows the various angles of approach to this glycan taken by the different bNAb families ( Fig . 5B ) . While PGT122 and PGT128 approach the glycan from the same side , they do so at slightly different angles , whereas PGT135 and 2G12 approach from the opposite side . As a result , almost no component other than the Asn322 glycan is shared by the PGT122/PGT128 and PGT135/2G12 epitopes . The structural data are , therefore , consistent with the variable competition between the bNAbs from the V3-glycan ( PGT122 , PGT128 ) and OD-glycan ( 2G12 , PGT135 ) sub-clusters ( Fig . 1 ) . PGT151 competed strongly and reciprocally with 8ANC195 . However , although both bNAbs target the gp120-gp41ECTO interface , their epitopes do not overlap [5 , 36] . Modeling both Abs onto the 3D image of the trimer now shows that the competition between PGT151 and 8ANC195 is rooted in steric hindrance that arises from clashes in the constant region of the Fabs; i . e . , the epitopes are sufficiently close to one another that both bNAbs cannot gain access to the trimer simultaneously ( Fig . 5C ) . One unexpected example of nonreciprocal competition involved bNAbs from the CD4bs and OD-glycan clusters . Furthermore , binding inhibition could sometimes be seen even when the competitor bNAb bound the trimer only comparatively weakly in the ELISA . Specifically , based on their EC50 values , the OD-glycan bNAbs PGT135 and PGT136 bound ~5-fold less efficiently than the CD4bs bNAbs VRC01 , 3BNC60 , 3BNC117 and NIH45-46 [12] , yet the former were still able to substantially , and unidirectionally , inhibit the binding of the latter ( Fig . 1 ) . Previous epitope mapping results from electron microscopy and x-ray crystallography studies clearly show that these two epitope clusters are independent and non-overlapping ( Fig . 5D and Fig . 6A and B ) . We hypothesized that the decreased binding of CD4bs antibodies after PGT135 attachment is not caused by steric hindrance of the Fabs themselves , but by rearrangement of either the gp120 subunit or one or more specific glycan structures . The relative positions of the Fabs also agree well with negative-stain EM maps of the same Fab-trimer complexes ( EMD-2331 [20] ) . Previous reports that the CDR H1 and H3 loops of PGT135 Fab interact closely with gp120 glycans Asn386 and Asn392 [20] are consistent with the Env trimer crystal structure ( Fig . 6C ) . After superimposing the VRC01-gp120 structure onto the Env trimer structure , it appears that , when PGT135 binds to the trimer , the penetration of its CDR H1 loop into the glycan canopy would push the Asn363 glycan , and to a lesser extent the Asn386 glycan , toward the CD4bs . This movement would then create a steric clash between the Asn363 glycan and the CDR H2 loop of VRC01 ( Fig . 6C ) [37] . In contrast , a similar analysis indicates that VRC01 binding to the trimer does not cause significant movement of the Asn386 and Asn392 glycans into positions where they could sterically impede PGT135 binding . It should be noted that the crystal structures were derived using gp120s and trimers produced in 293S cells and EndoH-treated to remove as much glycan as possible prior to crystal formation . Hence only the residual , truncated glycan components that were resolved in the crystal structures were modeled . The Asn363 , Asn386 and Asn392 glycan structures on the 293F cell derived trimers used in the competition ELISAs ( and on viruses ) will be substantially larger and the potential for clashes accordingly greater . When we performed similar docking analyses using the cryo-EM structure of 293F cell-derived trimers in complex with PGV04 , additional density could be seen to extend from Asn363 that passed over the top of the VRC01 Fab ( PGV04 and VRC01 have very similar , overlapping epitopes; EMD-5779 [19] ) . This suggests that the Asn363 glycan may have a limited range of conformations that are permissive for binding of CD4bs antibodies . The unexpected nonreciprocal competition between 8ANC195 and CD4bs bNAbs , particularly VRC01 , 3BNC60 , 3BNC117 and NIH45-46 , can also now be explained on structural grounds . The 8ANC195 epitope spans both the gp120 and gp41ECTO subunits but is dependent on the Asn276 glycan that borders the membrane-proximal side of the CD4bs ( Fig . 5E ) . Deleting this glycan increases the neutralization potency of VRC01 but , when present , the VRC01 light chain binds to it [38] . We hypothesize that when VRC01 , 3BNC60 , 3BNC117 or NIH45-46 bind the trimer , the Asn276 glycan is reoriented in a way that 8ANC195 can no longer recognize it . However , because Asn276 is not an critical component of the epitopes for some ( e . g . , VRC01 ) , but not all ( i . e . , 3BNC117 ) CD4bs bNAbs [39] ( S7B Fig ) , the competition is non-reciprocal . The enhanced binding of 3BNC117 by 8ANC195 may be related to dependency of Asn276 for 3BNC117 binding . Thus , 8ANC195 might reorient this glycan in such a way that it becomes more accessible to 3BNC117 . Other CD4bs bNAbs such as CH103 and CH106 bind to the membrane-distal side of the CD4bs and have a steeper angle of approach than VRC01; these antibodies probably do not reorient the Asn276 glycan in a way that precludes 8ANC195 binding ( Fig . 5E ) . Using a bNAb cross-competition ELISA , supported by SPR data and structural analyses , we have defined the steric and allosteric relationships among the known antigenic sites on the HIV-1 Env trimer . We have also identified four mechanisms by which bNAbs can interfere with one another’s binding ( Fig . 7 ) : 1 ) direct overlap of epitopes ( many examples ) ; 2 ) steric inhibition ( PGT151 inhibition of 8ANC195 ) ; 3 ) allosteric inhibition ( PGT145 inhibition of 1NC9 , 8ANC195 , PGT151 or CD4 ) ; 4 ) glycan reorientation ( PGT135/PGT136 inhibition of CD4bs bNAbs , and CD4bs bNAb inhibition of 8ANC195 ) . Whether competition between bNAbs is or is not reciprocal can also be influenced by both binding affinity ( competition in ELISA is likely to be influenced substantially by bNAb off-rates ) and stoichiometry ( whether 1 , 2 or 3 copies of a bNAb bind the trimer ) . Overall , based on our current observations , the binding of bNAbs to the trimer can involve local remodeling of loops or glycans , as well as triggering long-distance allosteric effects . The platform for our experiments is the soluble , recombinant BG505 SOSIP . 664 trimer , a highly stable native Env spike mimetic . While we are confident that these soluble trimers closely resemble the native spike , as shown by the strong correlation between antibody binding in ELISA and neutralization of the corresponding virus [12] , differences may need to be considered when comparing an engineered stabilized soluble protein versus the wild-type membrane-associated protein . However , we suggest that such differences are more likely to be manifested at the qualitative than the quantitative level , for example in the rates and extent of antibody-induced conformational changes to the trimer and any subsequent allosteric effects . One seemingly common belief about the trimer until recently was that it has only a few sites of vulnerability . That view emerged in a period when not many bNAbs were known [40–43] . A re-evaluation now seems to be warranted . The availability of entirely new families of bNAbs and identification of their epitopes on the structure of the trimer reveals many weaknesses in Env’s defenses against antibodies . Additional novel bNAbs are still likely to be isolated that may expose more . Thus , instead of just a few distinct sites of vulnerability there are many; and they overlap with one another . Computational modeling what is now known about the antigenic surface of the trimer allows one to take a virtual walk over the trail of partially overlapping epitopes that span the entire trimer from top to bottom . An example of the route mapped out from apex-to-base involves the following epitopes PG9-PGT122-PGT128-2G12-PGT135-CH103-1NC9-8ANC195-35O22 ( Fig . 4B and C; Fig . 5 ) . The overlap between these sites suggests that there is an almost continuous antigenic surface that can be recognized by one or more known bNAbs . Key questions now might be: does the human immune system target some vulnerable areas more frequently , or more effectively , than others ? Does the development of an antibody response against one epitope influence the development of antibodies against other sites and , if so , how [44 , 45] ? The existence of indirect interference mechanisms for bNAb binding has implications for the use of competition ELISAs , or similar assays , for mapping the NAb/Ab specificities in sera from Env-immunized animals or HIV-1-infected people . For example , if an anti-Env serum inhibited the trimer binding of a CD4bs bNAb such as VRC01 , it would not necessarily mean that VRC01-like antibodies were present in the serum as the inhibitory effect could be indirect [46] . Our detailed antigenic map of the BG505 SOSIP . 664 trimer may assist in the design of vaccines aimed at inducing bNAbs , and for mapping responses in sera from infected individuals or from recipients of Env vaccines . This information may also be useful for designing cocktails of bNAbs for therapeutic use . For example , only antibodies with mutually reinforcing and not competing properties should be used together . Finally , the competition matrix can be considered the benchmark for analysis of where newly discovered bNAbs target the Env trimer .
The design of BG505 SOSIP . 664 trimers , including the D7324-epitope tagged version , has been described extensively elsewhere , as have the methods to produce and purify them [12 , 14 , 18 , 19] . 2G12-affinity and size exclusion chromatography ( SEC ) -purified BG505 SOSIP . 664-D7324 and SOSIP . 664-His trimers were used for the ELISAs and SPR , respectively . For EM reconstructions , 2G12-affinity and SEC-purified BG505 SOSIP . 664 trimers without the D7324-epitope tag were used . In the experiments in S2 Fig , BG505 SOSIP . 664-D7324 trimers were purified via PGT145-affinity chromatography without SEC . MAbs were obtained as gifts , or purchased , or expressed from plasmids , from the following sources directly or through the AIDS Reagents Reference Program: John Mascola and Peter Kwong ( VRC01 ) ; Polymun Scientific ( 2G12 ) ; Michel Nussenzweig ( 3BNC60 , 3BNC117 , NIH45-46 , NIH45-46W , 1NC9 , 8ANC195 , 3BC315 ) ; Barton Haynes ( CH31 , CH103 , CH106 , ) ; James Robinson ( 17b ) . MAbs were biotin labelled using the Pierce EZ-Link Sulfo-NHS-Biotinylation kit ( product code #21425 ) and Fabs were produced with the Pierce Fab purification kit ( product code #44685 ) , in both cases according to the manufacturer’s instructions . The D7324-capture ELISA has been described elsewhere [12] . For the detection of biotinylated MAbs , horseradish peroxidase ( HRP ) -labeled streptavidin ( Sanquin , The Netherlands ) was used . Microlon-600 96-well , half-area plates ( Greiner Bio-One , Alphen aan den Rijn , The Netherlands ) were coated overnight with Ab D7324 ( Aalto Bioreagents , Dublin , Ireland ) at 10 μg/ml in 0 . 1 M NaHCO3 , pH 8 . 6 ( 50 μl/well ) . After washing and blocking steps , purified , D7324-tagged BG505 SOSIP . 664 trimers were added at 300 ng/ml in TBS/10% FCS for 2 h . A 25-μl aliquot of TBS ( 150 mM NaCl , 20 mM Tris ) plus 2% skimmed milk containing the competitor MAbs/Fabs ( 10 μg/ml of each in a 50 μl final volume ) was added to each well of a separate plate . Unbound Env proteins were washed away from the test plate before the competitor MAbs were added and incubated for 30 min . A 25-μl aliquot of the biotinylated MAbs/IgG , at a concentration giving 50–70% of the maximum binding signal , was then added for 2 h followed by 3 washes with TBS . HRP-labeled streptavidin ( Sanquin , The Netherlands ) or HRP-labeled Goat anti-Human IgG , Fcγ fragment specific ( Jackson Immunoresearch , Suffolk , England ) were added for 1 h at a 1:3000 dilution in TBS/2% skimmed milk , followed by 5 washes with TBS/0 . 05% Tween-20 . Colorimetric detection was performed using a solution containing 1% 3 , 3′ , 5 , 5′-tetramethylbenzidine ( Sigma-Aldrich , Zwijndrecht , The Netherlands ) , 0 . 01% H2O2 , 100 mM sodium acetate and 100 mM citric acid . Color development was stopped using 0 . 8 M H2SO4 when appropriate , and absorption was measured at 450 nm . An anti-histidine antibody ( GE Healthcare Bio-Sciences ) was immobilized onto CM5 chips by amine coupling . His-tagged BG505 SOSIP . 665 trimers were captured in amounts corresponding to 500 RU except in experiments with 8ANC195 in which 100 RU as used because of the NAb's slow association . Also PGT151 and 35O22 were used at 500nM , 8ANC195 at 1 . 5uM . SPR was performed under conditions previously described with modifications as follows [35] . In a new SPR competition format , two analytes ( NAbs ) were sequentially injected in a single cycle . The first analyte was injected and then after 200 s of the association phase , the second analyte was injected , also for 200 s , both at flow rates of 30 μl/min . After the second injection , dissociation was followed for 400 s . For comparison , each analyte was also injected on its own at the indicated concentrations . After each complete cycle , the chip surface was regenerated with a single pulse of 10mM Glycine ( pH2 . 0 ) for 120s at a flow rate of 30 μl/min . BG505 SOSIP . 664 gp140 trimers in complex with Fabs were analyzed by negative stain EM after overnight incubation at room temperature ( 6 molar excess of Fab to trimer ) . A 3 μL aliquot containing ~0 . 03 mg/mL of the Fab-trimer complex was applied for 5 s onto a carbon-coated 400 Cu mesh grid that had been glow discharged at 20 mA for 30 s , then negatively stained with 2% ( w/v ) Uranyl formate for 60 s . Data were collected using an FEI Tecnai T12 electron microscope operating at 120 keV , with an electron dose of ~25 e-/Å2 and a magnification of 52 , 000x that resulted in a pixel size of 2 . 05Å at the specimen plane . Images were acquired with a Tietz TemCam-F416 CMOS camera using a nominal defocus range of 1000 nm . Data processing methods were adapted from those used previously [9 , 12] . Particles were picked automatically using DoG Picker and put into a particle stack using the Appion software package [47] . Initial , reference-free , two-dimensional ( 2D ) class averages were calculated using particles binned by two via Iterative MSA/MRA and sorted into classes [48] . Particles corresponding to complexes were selected into a substack and binned by two before another round of reference-free alignment was carried out using the Iterative MSA/MRA and Xmipp Clustering and 2D alignment software systems [49] . For Fab-containing complexes , an ab initio common lines model was calculated from reference-free 2D class averages in EMAN2 [50] using 3-fold symmetry . EMAN [51] was used for all 3D reconstructions , and all maps were refined using 3-fold symmetry . In total , 39 , 259 particles were included in the final reconstruction for the 3D average of BG505 SOSIP . 664 trimer complex with CH103; 27 , 144 for the complex with CH106; 11 , 235 particles for the complex with 1NC9; 14 , 956 for the complex with 3BNC117; and 7 , 861 for the complex with VRC01 . To estimate the percentage of the HIV-1 Env surface covered by bnAbs , various EM models [PGT128 ( EMDB 1970 ) , PGT122 ( EMDB 5624 ) , PG9 ( EMDB PG9 ) , PGT135 ( EMDB 2331 ) , 2G12 ( EMDB 5982 ) , PGV04 ( EMDB 5779 ) , PGT151 ( EMDB 5918 ) , 8ANC195 ( EMDB 2625 ) and 35O22 ( EMDB 2672 ) ] were loaded into the Chimera software package [52] . These models were then fit into the map of full-length JR-FL Env ( i . e . , including the transmembrane region ) in complex with PGT151 ( EMDB 5918 ) after removal of the PGT151 density . The contact area between various Fab densities and the trimer was estimated and the sum divided by the total surface area of the trimer . The surface area of each bNAb footprint is an estimate based on docking Fab X-ray structures into their respective EM reconstructions , and then coloring a zone on the Env surface that is within 4 Å of the Fab atoms . The 2G12 Fab footprint is colored within 8 Å of the docked Fab structure because it binds to the tips of several high mannose glycans that are not resolved in the low-resolution model . To account for surface area at the bottom of the trimer that is inaccessible due to the cell/virus membrane , the trimer volume was segmented and the surface area for a small portion in the transmembrane region was subtracted from the total trimer surface area . In addition , the contact area between 2G12 and the trimer was doubled to allow a more accurate measurement , taking into account that a gap between the Fab and trimer densities in the EM model arises from the high-mannose glycan patch ( including the 2G12 epitope ) not being resolved in the negative-stain reconstruction .
|
The discovery of new broadly neutralizing antibodies against various epitopes on the HIV-1 envelope glycoprotein trimer and increased knowledge of its structure are guiding vaccine design . To increase our understanding of the interrelationships among the different epitopes , we generated a detailed antigenic map of the trimer using a variety of techniques . We have uncovered various mechanisms whereby antibodies can influence each other’s binding . The resulting antigenic map should further aid in design of HIV-1 vaccines to induce broadly neutralizing antibodies and in devising cocktails of such antibodies for therapeutic use .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Comprehensive Antigenic Map of a Cleaved Soluble HIV-1 Envelope Trimer
|
Despite its relatively poor efficacy , Bacillus Calmette-Guérin ( BCG ) has been used as a tuberculosis ( TB ) vaccine since its development in 1921 . BCG induces robust T helper 1 ( Th1 ) immune responses but , for many individuals , this is not sufficient for host resistance against Mycobacterium tuberculosis ( M . tb ) infection . Here we provide evidence that early secreted antigenic target protein 6 ( ESAT-6 ) , expressed by the virulent M . tb strain H37Rv but not by BCG , promotes vaccine-enhancing Th17 cell responses . These activities of ESAT-6 were dependent on TLR-2/MyD88 signalling and involved IL-6 and TGF-β production by dendritic cells . Thus , animals that were previously infected with H37Rv or recombinant BCG containing the RD1 region ( BCG::RD1 ) exhibited improved protection upon re-challenge with virulent H37Rv compared with mice previously infected with BCG or RD1-deficient H37Rv ( H37RvΔRD1 ) . However , TLR-2 knockout ( TLR-2-/- ) animals neither showed Th17 responses nor exhibited improved protection in response to immunization with H37Rv . Furthermore , H37Rv and BCG::RD1 infection had little effect on the expression of the anti-inflammatory microRNA-146a ( miR146a ) in dendritic cells ( DCs ) , whereas BCG and H37RvΔRD1 profoundly induced its expression in DCs . Consistent with these findings , ESAT-6 had no effect on miR146a expression in uninfected DCs , but dramatically inhibited its upregulation in BCG-infected or LPS-treated DCs . Collectively , our findings indicate that , in addition to Th1 immunity induced by BCG , RD1/ESAT-6-induced Th17 immune responses are essential for optimal vaccine efficacy .
Tuberculosis ( TB ) remains a major health problem , with an estimated one third of the world's population infected with Mycobacterium tuberculosis , the causative agent of TB , resulting in ∼3 million deaths annually . Bacillus Calmette-Guérin ( BCG ) , the only TB vaccine presently used in humans , has been widely used throughout the world since its inception in 1921 , and an estimated 3 billion people have received it [1] . However , its efficacy against pulmonary TB in adults is highly variable ( 0–80% ) [2] and depends on ethnicity and geographical location [3] , [4] , [5] . The antigenic component ( s ) that is absent in BCG to elicit critical protective immune responses against TB has been an area of intense research [4] , [5] . Early secreted antigenic target protein 6 ( ESAT-6 ) is one of the most prominent antigens expressed by Mycobacterium tuberculosis ( M . tb ) , but not by BCG [6] , [7] . ESAT-6-specific T cells are frequently found in TB patients as well as in infected animals [8] , [9] , [10] . Thus , ESAT-6 is being extensively studied for its potential activity as a subunit vaccine [11] . T cell receptor transgenic T cells specific for ESAT-6 exhibit significant protection against TB [12] . Consistent with this , a recombinant BCG strain that contains region of difference 1 ( RD1 ) , which includes ESAT-6 , exhibited improved protection against TB [13] . However , the basis of this improved protection remains elusive . Furthermore , the mechanism by which ESAT-6 vaccination induces protective immune responses against TB remains to be investigated . Furthermore , deletion mutants of virulent M . tb strains for RD1 or ESAT-6 ( a protein product of the RD1 region ) resemble BCG in their infectivity and attenuation [14] . Therefore , these bacterial strains provide insight into the rational selection and design of suitable candidate vaccines for M . tb infection . It is clear that vaccination with BCG produces Th1 cell-mediated immune responses , and this is moderately effective in protecting against disseminated TB and against meningitis in children [15] . However , immune responses that are critical for protection against adult pulmonary TB remain incompletely understood . Recently , it has been shown that Th17 cell responses play an important role in establishing protective immune responses against TB [16] . However , Th17 cells do not contribute to the primary immune responses in tuberculosis infection [17] . The antigen-specificity of protective Th17 cell responses in M . tb vaccination has not been reported . The differentiation of Th17 cells involves the cytokines interleukin ( IL ) -6 and TGF-β [18] , [19] . Earlier studies indicated that IL-6 production in DCs is regulated by microRNA-146a ( miR146a ) expression , which acts as a negative feedback regulator in TLR signalling by targeting IL-1R associated kinase ( IRAK ) -1 and TRAF6[20] , [21] . miR146a inhibits the expression of IRAK-1 and TRAF6 and impairs NF-κB activity , which results in suppression of IL-6 , IL-1β and TNF-α expression [21] , [22] . Recently , it has been shown that expression of miR146a is also upregulated in viral and bacterial diseases to modulate immune responses [23] , [24] . Therefore , we hypothesised that miR146a might have a key role in M . tb infection by regulating IL-6 production . Here we show that H37Rv and recombinant BCG containing the RD1 region ( BCG::RD1 ) induce improved vaccine efficacy compared with BCG and H37Rv deletion mutants for RD1 ( H37RvΔRD1 ) . The virulent strain H37Rv and BCG::RD1 induced both Th1 and Th17 cell responses , whereas BCG and H37RvΔRD1 induced only Th1 cell responses . Inhibition of IL-17 by neutralizing antibodies dramatically reduced the vaccine efficacy of H37Rv and BCG::RD1 . H37Rv and BCG::RD1 induced IL-6 and TGF-β in DCs , which generated a microenvironment conducive to the differentiation of Th17 cells . In contrast , BCG and H37RvΔRD1 induced dramatically lower levels of IL-6 and TGF-β . Interestingly , production of both IL-6 and TGF-β in DCs induced by H37Rv and BCG::RD1 was dependent on the TLR-2/MyD88 signalling pathway . Furthermore , DCs infected with H37Rv or BCG::RD1 upregulated lower levels of miR146a compared with BCG and H37RvΔRD1 , which differentially affected IL-6 production in infected DCs . Consistent with this , ESAT-6-treated DCs produced IL-6 and TGF-β in a TLR-2/MyD88-dependent manner , and facilitated the polarization of Th17 cell responses . miR146a expression in DCs was unaffected by ESAT-6 treatment and comparable to uninfected DCs , and ESAT-6 dramatically inhibited miR146a upregulation in BCG-infected or LPS-treated DCs . Therefore , these results indicate that interaction of ESAT-6 with TLR-2 generates a cytokine environment that facilitates the differentiation of Th17 cells , which in turn contributes to protection against TB .
It is well accepted that Th1 cell responses are indispensable for host protection against TB [25] . The vaccine strain BCG induces robust Th1 cell responses , yet it is not an effective vaccine against adult pulmonary TB in many individuals [3] , [4] , [5] . Therefore , additional immune response ( s ) are required for optimal vaccine efficacy . Recently , Th17 cells have been implicated in protective immunity against TB [16] . Previous studies have demonstrated that RD1 , which is absent in BCG , plays a dominant role in protective immune responses and bacterial virulence [14] . Thus , an RD1 deletion mutant of H37Rv resembles BCG in its infectivity [7] . Therefore , we tested the virulence and cytokine production by virulent strain H37Rv and vaccine strain BCG . We challenged C57BL/6 mice with a low dose ( ∼110 CFU ) of H37Rv or BCG by the aerosol route . We found that H37Rv and BCG replicated to a similar extent during the initial phase of the infection ( Fig . 1A ) . However , at later time points , growth of BCG was gradually diminished ( p<0 . 032 ) ( Fig . 1A ) , suggesting that adaptive immune responses play an important role in clearing BCG . After 21 days of infection , only a few bacilli were found in the lungs of animals infected with BCG ( Fig . 1A ) . These kinetics of BCG and H37Rv infection are in agreement with the published literature [25] , [26] , [27] , [28] . Interestingly , we observed significantly higher numbers of IL-17-producing CD4+ T cells in the lungs of animals infected with H37Rv , as compared with BCG ( p<0 . 0001 ) ( Fig . 1B ) . In sharp contrast , both H37Rv and BCG induced IFN-γ in the bronchoalveolar lavage ( BAL ) fluid ( Fig . 1B & C ) . This is further supported by the increased amounts of IL-17 produced in the BAL fluid of mice infected with H37Rv but not BCG ( p<0 . 0001 ) ( Fig . 1B & C ) . Although significantly lower , BCG-infected animals produced some IL-17 in the BAL fluid ( Fig . 1C ) . However , we were unable to detect IL-17-producing CD4+ cells in the lung of BCG-infected animals ( Fig . 1B ) , suggesting that the source of IL-17 in BCG-infected animals is not Th17 cells . It has been previously shown that γΔ T cells are the primary source of IL-17 in the lung during BCG infection [29] . Consistent with the observation that H37Rv induces higher Th17 responses in the lung , we found significant levels of IL-6 ( p<0 . 001 ) and TGF-β ( p<0 . 001 ) , two key cytokines required for the differentiation of Th17 cells [18] , [19] , in the BAL fluid of animals infected with H37Rv ( Fig . 1C ) . In contrast , both H37Rv and BCG produced similar amounts of IL-12p40 ( Fig . 1C ) , a cytokine that supports Th1 cell differentiation . These observations suggested that H37Rv creates an environment that is conducive to the differentiation of both Th1 and Th17 cells , whereas BCG promotes Th1 cell differentiation only . To investigate the molecular basis for the capacity of H37Rv to induce high levels of IL-6 , we compared induction of microRNA-146a ( miR146a ) , a negative regulator of innate immune components such as IL-6 , in infected DCs[20] , [21] . Interestingly , we found that BCG significantly upregulated miR146a in DCs as compared with H37Rv-infected DCs ( p<0 . 01 ) ( Fig . 1D ) . Furthermore , specific knock-down of miR146a expression dramatically upregulate both mRNA and protein level of IL-6 in BCG-infected DCs ( Fig . 1E & F ) . These data indicated that H37Rv promotes the differentiation of both Th1 and Th17 cell responses , whereas BCG induces Th1 responses but fails to support Th17 cell differentiation due to its induction of miR146a in infected cells . Nonetheless , a recent study indicated that BCG is unable to induce IL-17-producing cells during primary challenge but can do so after several rounds of challenge [30] . Therefore , it is highly likely that repeated immunization with BCG induces IL-17 production by innate-like cells or recruits IL-17-producing cells to the lung due to the local chemokine milieu . Compared with virulent H37Rv , BCG possesses multiple deletion mutations . These mutations are called regions of difference ( RD ) . Among these RD regions , RD1 is the most dominant and plays an important role in virulence [31] . Thus , an RD1 deletion mutant of H37Rv resembles BCG in its infectivity [14] . Within the RD1 region , the proteins ESAT-6 and CFP-10 have been shown to form a complex and participate in a type VII secretion system [6] , [32] . Therefore , we tested the virulence and cytokine production induced by H37RvΔRD1 . Consistent with the results obtained for BCG , we observed that H37RvΔRD1 initially grew to a similar extent as the parental H37Rv strain , but at later time points its growth gradually diminished ( Fig . 1A ) . Akin to BCG , H37RvΔRD1 failed to induce IL-17 production in the lung ( Fig . 1B & C ) . However , both strains induced similar quantities of IFN-γ ( Fig . 1B & C ) . Furthermore , IL-12 production was comparable between these strains , whereas IL-6 and TGF-β production was dramatically reduced as compared with H37Rv ( Fig . 1C ) . To provide further support for these findings , we performed similar experiments with the BCG recombinant strain in which the RD1 region was reintroduced ( BCG::RD1 ) . In agreement with previous reports [33] , BCG::RD1 showed a dramatically higher virulence as compared with the parental BCG strain , but was comparable to H37Rv ( Fig . 1A ) . Consistent with this finding , BCG::RD1 induced both IFN-γ and IL-17 ( Fig . 1 B & C ) , as well as the Th1- and Th17-differentiating cytokines IL-12p40 , IL-6 , and TGF-β ( Fig . 1C ) in the lungs . Additionally , H37RvΔRD1 induced significantly higher levels of miR146a than the parental H37Rv strain in infected DCs ( p<0 . 02 ) ( Fig . 1D ) , and knock-down of miR146a significantly improved IL-6 production both at mRNA transcript and protein level ( Fig . 1E & F ) . To provide further support for these data , we performed similar experiments with the BCG recombinant strain containing RD1 ( BCG::RD1 ) . Consistently , BCG::RD1 induced both IFN-γ and IL-17 ( Fig . 1 B & C ) , and the Th1- and Th17-differentiating cytokines IL-12p40 , IL-6 , and TGF-β ( Fig . 1C ) in the lungs , but failed to induce miR146a in infected DCs ( Fig . 1D ) . These observations suggested that the RD1 region is responsible for the induction of Th17 cell responses . Previously , it has been shown that IL-17 plays an important role in the secondary immune response following vaccination with virulent H37Rv [16] . Thus , we examined whether IL-17 induced by virulent strains contributes to improved vaccine efficacy . For this purpose , we infected animals with H37Rv , BCG , H37RvΔRD1 , or BCG::RD1 for 30 days . These animals were subsequently treated with antibiotics for four weeks , and then rested for an additional month . We could not find any detectable M . tb organisms in these animals . These mice were then challenged with H37Rv through aerosol infection . We found that animals previously infected with BCG or H37RvΔRD1 exhibited robust protective immunity compared with primary infection ( p<0 . 01 ) ( Fig . 2A ) . Interestingly , we found that animals that were previously infected with H37Rv produced dramatically enhanced protective immune responses ( Fig . 2A ) . This is in agreement with previous reports suggesting that virulent strains of M . tb H37Rv induce superior protective immune responses [34] , [35] . However , the kinetics of host protective responses in our hand are somewhat different from these studies , which may be due to differential environmental factors in different geographical regions . In fact , it has been well documented that the efficacy of BCG in human vaccine trials dramatically varies depending on the geographical location ( [3] , [4] , [5] ) . Nevertheless , we tested M . tb antigen-specific responses induced by randomly selected animals from our colony . We challenged splenocytes from sixteen animals with M . tb-derived complete soluble antigen ( CSA ) or the unrelated antigen ovalbumin ( OVA ) and measured lymphoproliferation . We observed that animals from our colony responded weakly to CSA , whereas no response was detected against OVA . Therefore , these animals were likely exposed to environmental organism ( s ) that share antigenic similarities with M . tb . As a positive control , we used spleen cells from CSA immunized mice , which showed dramatic proliferation against in vitro rechallenge with CSA ( Fig 2B ) . These observations might also be relevant to the variable vaccine response of BCG . In either case , our findings suggested that , while BCG and H37RvΔRD1 induced significant protective immunity against TB , this was not sufficient to confer complete protection against disease pathology . In contrast , H37Rv and BCG::RD1 induced improved protective immune responses . Collectively , these observations suggested that the RD1 region enhances protective immune responses . Importantly , we found that animals that were previously infected with H37Rv or BCG::RD1 induced significantly higher numbers of Th17 cells in the lungs than animals infected with BCG or H37RvΔRD1 ( p<0 . 001 ) ( Fig . 2C ) . Therefore , we tested whether IL-17 was responsible for the improved vaccine efficacy of H37Rv or BCG::RD1 . For this purpose , we injected animals with anti-IL-17 or control mouse IgG antibodies every 72 hours during re-challenge with H37Rv or BCG::RD1 . Treatment with anti-IL-17 abrogated the observed enhancement in protective immune responses induced by H37Rv or BCG::RD1 ( Fig . 2D ) . Therefore , these observations suggested that H37Rv induced Th17 cell responses , which complemented Th1 cell responses for improved protection against TB . Our in vivo experiments demonstrated that H37Rv induces Th17 cell differentiation . Therefore , to provide insight into the mechanism whereby H37Rv promotes Th1 and Th17 cell differentiation , we compared the cytokines induced by DCs ( characterized with CD11c , CD11b , CD80 , CD86 , and MHC Class II markers Fig . 3A ) infected with H37Rv , BCG::RD1 , BCG , and H37RvΔRD1 . We found that H37Rv- or BCG::RD1-infected DCs produced substantial amounts of IL-12p40 , IL-6 , and TGF-β ( Fig . 3B ) . However , BCG and H37RvΔRD1 induced dramatically reduced amounts of IL-6 ( p<0 . 001 ) and TGF-β ( p<0 . 001 ) by DCs than H37Rv and BCG::RD1 ( Fig . 3B ) . Nevertheless , IL-12p40 was induced at comparable levels by all bacterial strains . Interestingly , we found that IL-6 and TGF-β production was dependent on TLR-2 and MyD88 , whereas IL-12p40 production was independent of TLR-2 but required MyD88 ( Fig . 3B ) . To determine whether these cytokines play a role in Th cell differentiation , we co-cultured ovalbumin ( OVA ) -specific CD4+ T cells from OT-II T cell receptor ( TCR ) transgenic ( Tg ) animals with infected DCs in the presence of OVA peptide and collected supernatant to determine the production of IFN-γ and IL-17 ( Fig . 3C ) . The results , which were supported by intracellular cytokine staining , indicated that H37Rv-infected DCs directed the differentiation of both IL-17- and IFN-γ-producing cells ( Fig . 3C & 4A ) . In sharp contrast , DCs infected with BCG or H37RvΔRD1 supported only Th1 cell differentiation . While the levels of IFN-γ production were similar for DCs infected with H37Rv , BCG , H37RvΔRD1 or BCG::RD , production of IL-17 was significantly higher in DCs infected with H37Rv or BCG::RD1 as compared with cells infected with BCG or H37RvΔRD1 ( p<0 . 001 ) . Furthermore , it is known that IL-22 is also secreted by IL-17 producing Th cells and recent study suggested that IL-22 was upregulated during M . tb infection in rhesus macaques and protective in function[36] . Therefore , we have also checked the IL-22 mRNA transcript level in our DC-T cells co-culture experiments and found that IL-22 mRNA transcript was 5–8 fold upregulated in H37RV or BCG::RD virulent strains compared to BCG and H37RvΔRD1 ( Fig . 4B ) . Therefore , these data indicated that the RD1 locus plays an important role in directing Th17 cell responses during M . tb infection . ESAT-6-reactive T cells are prevalent in TB patients and in animals infected with M . tb [8] , [9] , [10] . Furthermore , it has been shown that ESAT-6-specific T cells provide substantial protection against TB [12] . Therefore , it has been assumed that ESAT-6 is a good candidate for development of a TB vaccine [37] . From the preceding section , it was clear that the RD1 region plays an important role in directing Th17 cell differentiation , which in turn contributes to protective immune responses against TB . Differentiation of Th17 cells requires IL-6 and TGF-β simultaneously [18] , [19] . Therefore , we determined whether ESAT-6 induces these cytokines in DCs . We found that DCs treated with ESAT-6 produced both IL-6 and TGF-β ( Fig . 5A ) . However , ESAT-6 did not induce IL-12 in DCs ( Fig . 5A ) . CD4+ T cells from OT-II TCR Tg mice co-cultured with DCs in the presence of ESAT-6 and OVA differentiated into IL-17-producing cells ( Fig . 5B ) . Our findings indicated that ESAT-6 induces IL-6 and TGF-β production in DCs , which drives Th17 cell differentiation . Previous reports have suggested that ESAT-6 binds to TLR-2 [38] . Therefore , we tested whether TLR-2 is required for the capacity of ESAT-6 to induce IL-6 and TGF-β production . For this purpose , we compared cytokine production by DCs derived from wild type and TLR-2-/- mice . We found that DCs from TLR-2-/- mice were unable to produce IL-6 and TGF-β ( Fig . 5A ) . To confirm that innate immune signalling is required for the capacity of ESAT-6 to induce IL-6 and TGF-β production in DCs , we performed experiments with DCs isolated from MyD88-/- mice . As expected , DCs from MyD88-/- mice were also unable to produce IL-6 and TGF-β in response to ESAT-6 stimulation ( Fig . 5A ) . Interestingly , we observed that ESAT-6 dramatically inhibited miR146a expression in both LPS- and BCG-treated DCs ( Fig . 5C ) . Thus , ESAT-6 allows IL-6 production in DCs by inhibiting the induction of miR146a . Taken together , our findings indicated that IL-6 and TGF-β induced by ESAT-6 in DCs generate an environment that promotes the differentiation of Th17 cells . Next , we determined the capacity of DCs treated with H37RvΔRD1 to produce IL-6 and TGF-β . Our results clearly showed that neither H37RvΔRD1 nor BCG were able to induce IL-6 and TGF-β in DCs derived from either wild type , TLR2-/- , or MyD88-/- mice ( Fig . 3B ) . Interestingly , we found that the parental strain H37Rv , BCG::RD1 , H37RvΔRD1 and BCG induced IL-12p40 production in DCs isolated from both wild type and TLR-2-/- mice . However , none of these strains induced IL-12p40 production in DCs derived from MyD88-/- mice . Therefore , taken together , these observations suggested that ESAT-6 induces IL-6 and TGF-β production in a TLR-2-dependent manner . In contrast , production of IL-12p40 by DCs following infection with mycobacteria is independent of ESAT-6 and TLR-2 expression . However , IL-12p40 production induced by mycobacteria is dependent on MyD88 signalling . From the preceding section it is clear that interaction of ESAT-6 with TLR-2 creates an environment that is conducive to the differentiation of Th17 cells , which in turn results in protective immunity against TB . Therefore , to confirm that TLR-2 signalling is important for the observed Th17 cell responses and improved vaccine efficacy , we performed vaccination experiments in TLR-2-/- animals . These animals were infected with H37Rv and subsequently treated with antibiotics as described in the materials and methods . These animals were then challenged with virulent strain H37Rv . H37Rv-immunized TLR-2-/- mice generated protective immunity against H37Rv at a level similar to wild-type mice immunized with BCG during primary challenge ( Fig . 6A ) . However , TLR-2-/- mice exhibited reduced protective immune responses as compared with wild type littermates ( Fig . 6A ) . H37RvΔRD1 and BCG::RD1 also showed reduced protection in TLR-2-/- mice ( Fig . 6A ) similar to BCG and H37Rv . Other mycobacterial components , such as LAM and lipoproteins , can also activate TLR-2 [39] . Therefore , the observed differences in protective immune responses in TLR-2-/- animals could be caused by multiple TLR-2-dependent agonists . However , most of the M . tb-derived TLR-2 ligands induce only suppressive immune responses [39] . Therefore , the observed protective responses are most likely contributed by RD1-derived proteins . Furthermore , these differences are comparable with responses induced by H37Rv versus BCG in wild-type animals . Next , we analyzed effector T cells in the lungs . We found that TLR-2-/- animals generated IFN-γ-producing cells comparable to wild type littermates . However , these animals produced significantly fewer numbers of IL-17-producing cells in their lungs ( p<0 . 001 ) ( Fig . 6B ) . This finding is further strengthened by a recently published report , suggesting that TLR-2 is indispensible for the generation of Th17 responses during M . tb infection [40] . Therefore , these observations suggested that TLR-2 plays an important role in mounting Th17 cell responses to H37Rv , which in turn confers protective immunity to TB .
It is well accepted that Th1 cells play a central role for protection against TB [41] . Therefore , animals that are deficient in IFN-γ , IFN-γ receptor , Stat-4 , T-bet , or IL-12 exhibit increased susceptibility to M . tb infection [25] , [42] , [43] . BCG induces a robust Th1 response , but this does not appear to be sufficient for optimal protection against challenge with virulent M . tb [44] . Abundant Th1 cell responses have been found in TB patients as well as M . tb infected animals [45] , [46] . Thus , Th1 cells alone are not sufficient for protection against TB . Therefore , in addition to Th1 cell responses , a vaccine needs to induce additional Th cell response ( s ) to provide optimal protection against TB . Recently , it has been shown that Th17 cells play an important role in the secondary immune response against TB [16] . However , Th17 cells do not appear to participate in the primary immune response against TB [17] . Our findings clearly demonstrated that BCG and H37RvΔRD1 are unable to induce Th17 cell responses in the lung . Therefore , we considered the possibility that BCG lacks an antigen that drives Th17 cell differentiation and that such a response is required for optimal protection against TB . We observed that the virulent M . tb strain H37Rv and the engineered BCG::RD1 strain induced Th17 cell responses , which correlated with improved protection against re-infection and , thus , improved vaccine efficacy . This is in agreement with a previous report indicating that immunization with virulent M . tb H37Rv induces superior protective memory T cell responses [34] . However , unlike our results these authors found only 2–3 log differences in CFUs upon re-challenge with virulent H37Rv . This apparent discordance could be due to several differences in the experimental procedures employed . Jung et al . ( 2005 ) employed an extended ( 100 days ) antibiotic treatment protocol , which may have influenced immune responses , and increased the age of the mice at re-challenge . Furthermore , these investigators rechallenged mice with a higher ( two-fold ) dose of bacteria . In addition , it is also possible that differences in the microflora in different animal facilities might contribute to these apparent differences . Indeed , our results demonstrated that unimmunized animals from our facility were able to respond to M . tb antigens , albeit weakly , suggesting that exposure to environmental organisms that share antigenic properties with M . tb might contribute to improved protection . This may explain the differential vaccine responses against BCG that have been observed in different geographical locations and in subjects from different ethnicities . Previous reports have suggested that ESAT-6 , a protein within the RD1 region that is absent in BCG , is a promising vaccine candidate [8] . Furthermore , deletion mutants of H37Rv for RD1 resembled BCG in many aspects [8] . Interestingly , we found that the RD1 mutant of virulent H37Rv was unable to induce Th17 cell responses . BCG and H37RvΔRD1 were unable to induce persistent infection and , hence , the bacterial load declined much more rapidly [14] over time as compared with H37Rv or BCG::RD1 . Therefore , the observed differential Th1 and Th17 cell responses could be related to bacterial replication and antigenic load . However , previous reports indicated that BCG inhibits Th17 cell responses in lung and other organs [47] , [48] . Furthermore , primary infection with a high dose ( 2 . 5×105 CFU ) of BCG by intratracheal injection was unable to induce IL-17 in the lung , until re-challenge with PPD-coated beads , and the cellular source of IL-17 produced in this situation is not known [49] . Therefore , BCG alone , even at a high dose , is unable to induce Th17 cell responses . Nonetheless , a recent report indicated that BCG was unable to induce IL-17-producing cells in a primary challenge , however it did so upon repeated re-challenge [30] . Although our study indicated that the RD1 recombinant strain exhibited an improved vaccine efficacy compared with the parental BCG strain , it does not exclude a role for other RD regions in inducing improved host protective immune responses . We found that H37Rv and BCG::RD1 induce both Th1 and Th17 cell responses that contribute to improved vaccine efficacy as compared with BCG , which selectively induces Th1 cell responses . Th17 cell responses are directed by IL-6 and TGF-β , derived from antigen presenting cells ( APCs ) . Therefore , pathogen-associated molecular patterns ( PAMP ) encoded within the RD1 region are likely responsible for inducing these two cytokines . Considering that ESAT-6 induces protective immune responses and that the RD1 mutant was unable to induce Th17 cell responses , we considered the possibility that ESAT-6 induces Th17 cell-polarizing cytokines in APCs . Therefore , we tested IL-6 and TGF-β production in DCs infected with the wild-type and RD1 mutant H37Rv strain . These studies revealed that the mutant strain induced significantly lower levels of IL-6 and TGF-β in DCs . Furthermore , our studies with miR146a , a negative regulator of innate immune components such as IL-6 in infected DCs [21] , suggested that BCG and H37RvΔRD1 significantly upregulated miR146a in DCs as compared with H37Rv- or BCG::RD1-infected DCs or uninfected DCs . Consistent with these findings , knock-down of miR146a expression dramatically upregulated IL-6 production in BCG-infected DCs . Of note , however , our studies cannot exclude the possibility that , in addition to ESAT-6 , other factors encoded within the RD1 region contribute to the enhanced vaccine efficacy of BCG::RD1 . Previous studies have provided evidence that ESAT-6 directly binds to toll-like receptor-2 ( TLR-2 ) , a pattern recognition receptor ( PRR ) [38] . This led us to investigate the capacity of DCs from TLR-2-/- animals to produce Th17 cell-polarizing cytokines in response to ESAT-6 treatment . Indeed , we found that production of IL-6 and TGF-β was dependent on TLR-2 . To further substantiate this observation , we performed experiments with DCs from MyD88-/- animals . Consistent with the results obtained with TLR-2-/- mice , DCs from MyD88-/- mice were unable to produce IL-6 and TGF-β in response to ESAT-6 treatment . However , a previous report suggested that engagement of TLR-2 on macrophages by ESAT-6 inhibits LPS-induced cytokine production , especially IL-6 and IL-12p40 [38] . Therefore , we revisited this issue for DCs . In our hands , we did not observe any influence of ESAT-6 on IL-12 production by DCs . However , ESAT-6 augmented IL-6 and TGF-β production by LPS-stimulated DCs . While the reasons for these discordant findings remain unknown , we speculate that the cell line used in the experiments by Pathak et al . [38] might have already been tolerized by LPS , so that secondary stimulation with ESAT-6 resulted in even lower production of IL-6 due to higher steady-state miR146a levels . [22] . It has been well-established that ESAT-6-specific T cells are prevalent in TB patients and in animal models of TB [8] , [9] , [10] . Furthermore , ESAT-6-reactive TCR Tg cells confer substantial protection against TB in an animal model [12] . In fact , it has been shown that ESAT-6 recombinant BCG provides protection against TB [13] . However , the mechanism of ESAT-6-mediated protection has been unclear until now . One difference between mouse models and human infection with M . tb is that wild-type M . tb does not offer significant protection against reinfection in humans , despite containing ESAT-6 and other major antigens . This could be due to various reasons . First , there might be genetic differences between mice and humans that cause altered immune responses . Second , environmental exposures may alter protective immunity . Third , different individuals might respond differently to drugs used to treat TB and , thus , it is difficult to determine whether treatment was complete , while the remaining bacteria may cause secondary infection . Fourth , M . tb evolved many different types of immune evasion mechanisms [50] . For example , we have recently shown that bacteria that are within granuloma-like structures are sequestered from host protective immune responses by mesenchymal stem cells [51] . In summary , our findings indicate that , in addition to Th1 cells , Th17 cells play a critical role in conferring optimal protection against TB . The ESAT-6 protein , which is present in H37Rv and BCG::RD1 but not in BCG and H37RvΔRD1 , directs Th17 cell differentiation by inducing IL-6 and TGF-β in DCs in a TLR-2- and MyD88-dependent manner . Therefore , ESAT-6 can contribute to vaccine preparations by promoting Th17 cell responses .
Animal experiments were performed according to the guidelines approved by the Institutional Animals Ethics Committee meeting held on 22nd November 2007 at ICGEB ( approval ID; ICGEB/IAEC/IMM-13/2007 ) , New Delhi , India and Department of Biotechnology ( DBT ) guidelines , Government of India . All mice used for experiments were ethically sacrificed by asphyxiation in carbon dioxide according to institutional and DBT regulations . C57BL/6 and OT-II TCR transgenic mice ( 6–8 wks of age ) were initially purchased from The Jackson Laboratories , USA . TLR-2 and MyD88 knock-out mice ( 6–8 weeks of age ) , both on a C57BL/6 background , were the kind gift of Prof . Ruslan Medzhitov , Yale University , New Haven , USA . All animals were subsequently bred and maintained in the animal facility of the International Centre for Genetic Engineering and Biotechnology ( ICGEB ) , New Delhi , India . Mycobacterium tuberculosis strain H37Rv was a kind gift from the Colorado State University repository . H37RvΔRD1 and BCG were a kind gift from Prof . David Sherman , ( SBRI , Seattle , WA , USA ) . The integrative cosmid vector pYUB412 ( control vector ) and the recombinant cosmid vector RD1-2F9 harboring RD1 locus of M . tuberculosis [33] were kind gifts from Prof . Stewart Cole of the École Polytechnique Fédérale de Lausanne ( EPFL ) , Switzerland . The control and recombinant cosmids were electroporated individually into electrocompetent cells of BCG ( Danish ) to obtain BCG::YUB412 and BCG::RD1 strains , essentially as described previously [33] . Briefly , 100 ml of bacilli suspension ( OD600nm , 0 . 4 ) from a 7-day-old Middlebrook 7H9 ( Difco ) culture , supplemented with albumin-dextrose-catalase ( ADC; Difco ) , was pelleted by centrifugation at 2500 g for 15 min at 16°C , washed twice with 10% glycerol and finally resuspended in 3 ml of 10% glycerol . Two hundred microliters of the electrocompetent bacilli were mixed with 5 microliter of the control vector pYUB412 ( 85 ng µl−1 ) or recombinant vector RD1-2F9 ( 100 ng µl−1 ) and electroporated using the Gene Pulser Xcell Electroporation System ( Bio-Rad Pacific , Hong Kong ) with settings of 2 . 5 kV , 25 µF and 1000 Ω . After electroporation , cells were resuspended in 5 ml of 7H9 medium supplemented with ADC , and kept overnight at 37°C . The cells were then pelleted by centrifugation , resuspended in 100 µl of 7H9 medium , and plated on Middlebrook 7H11 medium supplemented with oleic acid-albumin-dextrose-catalase ( OADC , Difco ) , hygromycin ( 200 µg ml−1 ) and ampicillin ( 100 µg ml−1 ) . After three to four weeks of incubation at 37°C , hygromycin-resistant clones were selected . BCG::RD1-2F9 ( BCG::RD1 ) clones were characterized for secretion of ESAT-6 by immunoblotting using mouse anti-ESAT6 antibody ( unpublished data ) . Detailed procedures for preparation and characterization of recombinant ESAT6 have been described in our earlier publication [52] . Briefly , E . coli BL21 ( plysS ) transformed with pET23b+ vector ( Novagen ) carrying esat6 gene of M . tuberculosis was grown to mid-log phase , induced with IPTG ( 0 . 4 mM final conc . ) for 4 hrs , and the recombinant ESAT6 protein was extracted from the inclusion bodies in 8 M urea . The recombinant ESAT6 protein was then purified by Nickel -nitrilotriacetic acid ( Ni-NTA ) chromatography , checked for LPS contamination by LAL ( limulus amebocyte lysate ) tests , and characterized for purity by SDS-PAGE , immunoblotting and N-terminal amino acid sequencing as described previously [52] . The purified and LPS-free recombinant ESAT6 protein was aliquoted and kept at -80°C until further use . All mycobacterial strains were grown in 7H9 ( Middlebrook , Difco , USA ) medium supplemented with 10% ADC and with 0 . 05% Tween 80 and 0 . 2% glycerol , and cultures were grown to mid-log phase . Aliquots of the cultures in 20% glycerol were preserved at −80°C and these cryo-preserved stocks were used for infections . Mice were infected with various mycobacterial strains ( namely H37Rv , H37RvΔRD1 , BCG , or BCG::RD1 ) via the aerosol route using a Madison aerosol chamber ( University of Wisconsin , Madison , WI ) with its nebulizer pre calibrated to deposit a total of ∼110 to the lungs of each mouse as previously described [51] , [53] . Briefly , mycobacterial stocks recovered from a −80°C freezer were quickly thawed and subjected to light ultra-sonication to obtain a single cell suspension . Fifteen ml of the bacterial cell suspension ( 10×106 cells per ml ) was placed in the nebulizer of the Madison aerosol chamber pre-calibrated to deliver via aerosol route the desired number of CFUs to the lungs of animals placed inside the chamber . A day after the aerosol exposure procedure , three randomly selected mice were sacrificed at various time points and organs were harvested , homogenised in 0 . 2 µm filtered PBS containing 0 . 05% Tween 80 and plated onto 7H11 Middlebrook ( Difco USA ) plates containing 10% oleic acid , albumin , dextrose and catalase ( OADC ) ( Difco USA ) . Undiluted , ten-fold diluted and one hundred-fold diluted lung and spleen cell homogenates were plated in duplicate on the above 7H11 plates and incubated at 37°C for 21–28 days . Colonies were counted and CFU were estimated . Mice from various groups were euthanized at the indicated time points in various experiments; their organs were harvested for obtaining CFU counts and/or immune cell subpopulations for immunological studies as described under other sub-sections . Luminex kits were purchased from Millipore and Bio-Rad . GM-CSF and IL-4 were obtained from R&D Biosystems , USA . Purified or fluorescently-conjugated monoclonal antibodies against mouse CD11c ( N418 ) , CD11b ( M1/70 ) , CD80 ( 16-10A1 ) , CD86 ( GL1 ) , and MHC-II ( NIMR-4 ) were purchased from eBioscience , USA , and fluorescently-conjugated anti-mouse IgG2a ( R19-15 ) was purchased from BD Pharmingen . LPS was obtained from Sigma-Aldrich ( L-2654 ) . Mice were euthanized and the femurs were isolated . Bone marrow was flushed out with RPMI-1640 medium using a 2 . 0 ml syringe ( 26 . 5G ) . The cells were washed twice with PBS and then cultured in complete RPMI-1640 ( Gibco , UK ) medium supplemented with GM-CSF ( 40 ng/ml ) and IL-4 ( 10 ng/ml ) on 24-well plates ( 1 million/ml ) . On the third day , 75% of the medium was replaced with fresh DC culture medium . On day 5 , the suspended cells were removed and the loosely adherent cells were collected as immature DCs ( CD11c-positive cells were >90% ) . Flowcytometric analysis by using anti-CD11c , -CD11b , -CD80 , -CD86 , -MHC class II , and - IgG2a ( isotype control ) antibodies suggested that >95% of the cells were conventional DCs . BM cells were isolated from different mouse strains ( C57BL/6 , TLR-2-/- and MyD88-/- ) and differentiated into immature DCs as described above and cultured in 24-well plates ( 1×106 cells per well ) . Cells were infected with H37Rv , H37RvΔRD1 , BCG or BCG::RD1 ( MOI of 1∶10 ) . Similarly , 1×106 DCs were cultured in 24-well plates in the presence or absence of LPS at 1 µg/mL and co-stimulated with ESAT-6 protein at a final concentration of 5 µg/mL , with PBS as the negative control . Supernatants from cells were collected at 24 , 48 and 72 hrs for cytokine profiling . For Th1 and Th17 cell differentiation , CD4+ T cells ( 1×106 ) were purified by MACS method ( CD4+ T cell isolation beads kit; Miltenyi Biotech , Germany ) from OT-II TCR transgenic mice and cultured with immature DCs ( 1×106 ) infected with H37Rv , H37RvΔRD1 , BCG or BCG::RD1 ( MOI of 1∶10 ) in the presence of ovalbumin ( 10 µg/ml ) peptide ( Thermo Scientific , USA ) for 72 hours . Then , CD4+ T cells were harvested and subjected to intracellular staining for IL-17 and IFN-γ expression . Thirty days post infection , groups of mice were treated with 0 . 1 g/L rifampicin and 0 . 1 g/L isoniazid ( Sigma-Aldrich , St . Louis , MO , USA ) administered in the drinking water ( changed daily ) for 4 weeks . M . tuberculosis-infected control mice received plain drinking water . A control group of infected mice was sacrificed at the start of treatment ( early control group ) . A second group of infected but untreated mice was sacrificed 4 weeks after therapy was initiated ( late control group ) . Lungs from infected or uninfected animals were harvested and washed by swirling in PBS . They were opened up by cutting longitudinally and then cut into ∼0 . 5-cm pieces . These lung pieces were agitated in 25 ml of extraction buffer ( PBS , 3% FCS , 1 mM dithiothreitol , 1 mM EDTA ) for 30 min at 37°C . This slurry was passed through a loosely packed nylon wool column to remove the aggregates . The filtrate was layered on a discontinuous Percoll gradient ( Amersham Pharmacia Biotech , USA ) . This gradient was then centrifuged at 900 × g for 20 minutes . Cells at the interface of the 40/70% layer were collected and washed in staining buffer ( PBS , 3% FCS ) . Cells were cultured for intracellular staining as described below . Bronchoalveolar lavage ( BAL ) fluid was collected from lungs by intratrachial infusion of PBS and cell-free BAL was used for cytokine assay [54] . Spleens were isolated from sixteen randomly selected C57BL/6 mice of our colony or CSA-immunized ( 100 µg/ml in 200 µl incomplete Freund's adjuvant ) mice . Spleens were macerated by frosted slides in complete RPMI 1640 ( Gibco , Invitrogen , UK ) and made into a single cell suspension . Red blood cells ( RBCs ) were lysed with RBC cell lysis buffer and washed with complete RPMI 1640 . Splenocytes were counted and plated at 0 . 1×106 cells/well in 96-well plates and stimulated with different concentrations of M . tb complete soluble antigen ( CSA ) . Cells were cultured for 48 hours and then pulsed with tritiated thymidine ( 3H-TdR , 1 . 0 µCi per well; Amersham Biosciences UK ) before measuring incorporation of 3H-TdR by means of a cell harvester and liquid scintillation counter 16 hours later ( Wallac Trilux , Perkin Elmer , UK ) . For intracellular cytokine staining , cells were treated with 50 ng/ml phorbol myristate acetate ( PMA ) and 500 ng/ml ionomycin in the presence of 10 µg/ml brefeldin A ( Sigma-Aldrich or eBiosciences , USA ) added during the last 6 h of culture . Cells were washed twice with PBS and resuspended in a permeabilization buffer ( Cytofix/Cytoperm kit; BD ) , and stained with the following fluorescently conjugated monoclonal antibodies: anti-CD4 ( clone: GK1 . 5 ) -APC , anti-IFN-γ ( clone: XMG1 . 2 ) -FITC , and anti-IL-17 ( clone: 17B7 ) -PE ( all from eBiosciences , USA ) . Fluorescence intensity was measured by flow cytometry ( FACS Calibur; BD ) and data were analysed with FlowJo ( Tree star , USA ) . Bone marrow derived DC were isolated and infected with different bacterial strains ( H37Rv , H37RvΔRD1 , BCG and BCG::RD1 ) as described above and cultured for 24 hours for RNA isolation . Total RNA , including miRNAs was isolated by miRNAeasy isolation kit ( QIAGEN , Germany ) according to the manufacturer's instructions . Real-time quantitative RT-PCR analysis was performed using BioRad Real-Time thermal cycler ( BioRad , USA ) and miRCURY LNA universal reverse transcriptase microRNA PCR SYBR green master mix ( EXIQON , Denmark ) for miRNA amplification and IQ BioRad SYBER green master mix ( BioRad , USA ) for IL-6 expression , respectively . cDNA was synthesized by the miRCURY LNA universal reverse transcriptase microRNA cDNA synthesis kit ( EXIQON ) and the reaction was set up according to the manufacturer's protocol . For amplification of miR146a , LNA PCR miR146a and reference 5S rRNA primer sets were used and the reaction was set up as recommended by EXIQON . The relative expression level of miRNAs was normalized to that of internal control 5S rRNA by using 2-ΔΔCt cycle threshold method . Furthermore , for amplification of IL-6 or IL-22 , cDNA was synthesized by Omniscript RT kit ( QIAGEN ) using oligodT primers ( Fermentas , Maryland , USA ) . For IL-6 mRNA expression analysis primer sequences were IL-6F , 5′-TGGAGTCACAGAAGGAGTGGCTAAG-3′ and IL-6R , 5′- TCTGACCACAGTGAGGAATGTCCAC-3′ , and control GAPDH-F , 5′- CGTCCCGTAGACAAAATGGT-3′ and GAPDH-R , 5′- TTGATGGCAACAATCTCCAC-3′ and for IL-22 mRNA expression analysis primer sequences were IL-22F 5′-GTGACGACCAGAACATCCAG-3′ and IL22R 5′-ATCTCTCCGCTCTCTCCAAG-3′ . Data were normalized by the level of GAPDH expression in samples as described above . For transfection of anti-miR146a and scramble control ( EXIQON ) into DCs , cells were transfected at day 4 of culture in antibiotic free media using Lipofectamine ( Invitrogen , UK ) reagents and at day 5 cells were infected with H37Rv , H37RvΔRD1 , BCG or BCG::RD1 . After 24 hours of bacterial infection , cells were harvested for RNA preparation and analyzed for miR146a and IL-6 expression by quantitative real-time PCR as described above . Cytokines in the culture supernatant of DCs were assayed by a Luminex microbead- based multiplexed assay using commercially available kits according to the manufacturer's protocol ( Milliplex kit , Millipore and BioPlex , Bio-Rad , USA ) . All data were derived from at least three independent experiments . Statistical analyses were conducted using SPSS software and values were presented as mean±SD . Significant differences between the groups were determined by ANOVA followed by Tukey's multiple comparison test ( SPSS software ) . A value of p<0 . 05 was accepted as an indication of statistical significance . CFP-10: 10 kDa culture filtrate protein [Mycobacterium tuberculosis H37Rv] . ACCESSION ACH88465 . ESAT-6: early secreted antigenic target 6 kDa [Mycobacterium tuberculosis H37Rv] . ACCESSION AAC83446 .
|
Tuberculosis is a global health problem , with one-third of the global population infected with tubercle bacteria . Numerous studies have shown that Th1 cell responses are indispensable for protective immunity against TB . However , while the vaccine strain BCG induces sufficient Th1 cell response , this response does not appear to be sufficient for immune protection in many individuals . Here , we provide evidence for the first time that Th17 cell responses in the lung play a critical role for enhanced protection against TB . Surprisingly , the virulent M . tb strain H37Rv induced Th17 cell responses in the lung . Consequently , antibiotic-treated animals that were previously infected with H37Rv , as compared with similarly treated BCG-infected mice , generated improved protective immune responses against infection with virulent M . tb . We also provide evidence that the ESAT-6 protein , which is absent in BCG but present in H37Rv , induces IL-6 and TGF-β in dendritic cells in a TLR-2 and MyD88-dependent manner , which generates an environment that is conducive for the differentiation of Th17 cells in the lung . Our findings indicate that , in addition to Th1 cells , Th17 cells play a critical role in conferring optimal protection against TB .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"biology"
] |
2011
|
Early Secreted Antigen ESAT-6 of Mycobacterium tuberculosis Promotes Protective T Helper 17 Cell Responses in a Toll-Like Receptor-2-dependent Manner
|
In contrast to existing estimates of approximately 200 murine imprinted genes , recent work based on transcriptome sequencing uncovered parent-of-origin allelic effects at more than 1 , 300 loci in the developing brain and two adult brain regions , including hundreds present in only males or females . Our independent replication of the embryonic brain stage , where the majority of novel imprinted genes were discovered and the majority of previously known imprinted genes confirmed , resulted in only 12 . 9% concordance among the novel imprinted loci . Further analysis and pyrosequencing-based validation revealed that the vast majority of the novel reported imprinted loci are false-positives explained by technical and biological variation of the experimental approach . We show that allele-specific expression ( ASE ) measured with RNA–Seq is not accurately modeled with statistical methods that assume random independent sampling and that systematic error must be accounted for to enable accurate identification of imprinted expression . Application of a robust approach that accounts for these effects revealed 50 candidate genes where allelic bias was predicted to be parent-of-origin–dependent . However , 11 independent validation attempts through a range of allelic expression biases confirmed only 6 of these novel cases . The results emphasize the importance of independent validation and suggest that the number of imprinted genes is much closer to the initial estimates .
Why diploid organisms would forgo the safety-net of a redundant genome and preferentially express one allele in a parent-of-origin dependent manner has been a matter of debate since the discovery of imprinted transcription . Our understanding of this issue as well as the range of processes affected by imprinting is dependent on our catalog of the identity , function , and spatial-temporal specificity of imprinted genes . Imprinting was initially characterized with genetics . Regions where uniparental disomy is not tolerated were mapped by intercrossing reciprocal translocations and comparing viability when both copies of a genomic segment were inherited from one parent to viability when inherited from the other [1] . Continued use of complementation tests in function-based screens provided the most dramatic examples of parent-of-origin effects on development , implicating imprinted transcription in neonatal behavior as well as pre-natal and post-natal growth [2] . Imprinting was then formally demonstrated using nuclear transfer experiments , where aberrant development was observed following transfer of two pronuclei from parents of the same sex [3] . Subsequently , the first individual imprinted transcripts were mapped [4]–[7] . After screening many translocations , genome-wide estimates stood at 100–200 imprinted transcripts [8] . Although based on what we now know was an overestimate for the total number of genes ( 60 , 000–100 , 000 ) and an underestimate of the number of known imprinted clusters [8] , [9] , the ∼20 year-old estimate has endured all screening methods applied , including those that do not depend on overt phenotypes . Some of these screens revealed that imprinting occurs outside of the growth axis and affects transcription unrelated to viability or growth [10]–[13] . By Dec 2008 , genetic and molecular screening efforts combined yielded 128 confirmed imprinted genes in the mouse . Transcriptome sequencing of F1 mouse hybrids provides an unbiased alternative for discovering imprinted transcription in wild-type animals [14] , [15] . The approach is based on detecting allelic expression with RNA sequencing reads that map over heterozygous SNPs , where the identity of the base is used to distinguish allelic origin and a reciprocal cross is used to discriminate parent-of-origin from strain-specific biases . The first two applications of this approach yielded a small number of novel imprinted transcripts each [14] , [15] . However , two recent studies used this approach to identify more than 1 , 300 imprinted loci , including 484 noncoding RNAs and 347 genes that were sex-specific [16] , [17] . These 1 , 300 loci are an aggregate of the discoveries from E15 brains , adult medial prefrontal cortex ( PFC ) and adult preoptic area ( POA ) , and represent a ten-fold increase over previously known imprinted genes . The authors suggest that improved sensitivity from increased sequencing depth and improved resolution from sequencing the parents for de novo identification of SNPs enabled these advances . To investigate the biological robustness of these novel imprinted loci , we repeated the embryonic brain screen . Despite faithful technical reproduction of the experimental design , library construction , sequencing , and analysis , we could not reproduce the majority of novel imprinted genes . In this study we demonstrate that biological variation in the approach and technical variation of the assay introduce considerably more noise than was appreciated previously . We develop methods to account for this variation and demonstrate their utility through reanalysis of the published data mentioned above as well as new embryonic brain data .
To distinguish known from novel imprinted loci we compiled a catalog of genomic coordinates for all 128 known mouse imprinted genes that we were able to recover from the literature [13]–[15] , [18]–[23] ( accessible from GEO [24] under accession GSE27016 ) . All 128 were published prior to the recent papers reporting many more imprinted loci [16] , [17] . E15 brains yielded considerably more novel imprinted genes than either the adult PFC or POA ( 553 vs 153 and 256 respectively ) , providing the richest opportunity to test reproducibility [17] . Aside from an inexact match in developmental time points ( E17 . 5 vs E15 ) , our experiment was a faithful reproduction of the approach used by Gregg et al . We both used brains from reciprocally crossed C57Bl/6J ( B ) and CAST/EiJ ( C ) F1s ( from here on BxC will be used to describe F1s derived from B mother and C father; CxB will denote the reciprocal ) . We both constructed sequencing libraries using the standard Illumina RNA-Seq protocol and sequenced them to 36 bp ( single-end ) on an Illumina platform . We both used Novoalign ( www . novocraft . com ) to map reads to UCSC mouse transcripts and noncoding RNAs from the functional RNA database [25] . We used the same set of SNPs [17] and the same criteria for identifying imprinted transcripts ( i . e . containing at least one SNP with 10 or more reads with reciprocally biased expression , p<0 . 05; chi-square test ) . We observed 100% agreement on known imprinted gene calls in E15 brain , POA , and PFC [17] , confirming that our analyses were consistent . We detected 38 and 42 known imprinted genes in E17 . 5 and E15 data respectively . 32/42 ( 76 . 2% ) were detected in both samples ( 0 . 1 transcripts expected by chance; Figure 1 ) . This was in sharp contrast to 51/396 ( 12 . 9% , 24 expected by chance ) novel imprinted genes that confirmed in our screen ( Figure 1 ) . This discrepancy is not inconsistent with the experimental validation carried out on novel imprinted genes by Gregg et al . : included in these 51 replicating genes were 2/2 with no previous evidence of imprinting ( Eif2c2 and DOKist ) that were discovered and further validated in E15 brain [17] . To investigate the benefits of sequencing parents to identify heterozygous SNPs , we repeated this analysis using publicly available SNPs . Perlegen [26] used microarrays to resequence the CAST genome in 2007 and the Wellcome Trust Sanger Institute sequenced 17 mouse strains and released ∼19 million C57Bl/CAST SNPs in 2009 [27] . We converted SNP transcript coordinates published by Gregg et al . [17] to genomic coordinates ( July 2007 NCBI 37/mm9 ) using coordinates of UCSC Known Genes [28] . 99 . 96% SNPs [17] mapped successfully to 136 , 532 unique positions . 88 . 9% of these were among the 19 . 6 million SNPs identified by Perlegen [26] and/or Sanger [27] , of which 98 . 9% agreed on base identity . The transition∶transversion ratio of SNPs that agree with [26] , [27] was 3 . 00 , and 2 . 06 for the remaining 11% novel SNPs , suggesting that the novel SNPs reported by Gregg et al . [17] have a higher proportion of false-positives . False-positive SNPs cannot lead to an imprinting call since there would be no reads supporting the non-reference CAST allele . Reanalysis of the data using only the 88 . 9% SNPs that also exist in the public domain yielded nearly identical results ( Figure 1; see numbers in parentheses ) with no reduction in sensitivity for known imprinted genes and less than 3% sensitivity reduction in novel regions . This demonstrates that sequencing biological parents when SNPs are publicly available from sequenced parental strains provides little added benefit [26] , [27] . Statistical modeling of allele-specific expression measured by transcriptome sequencing is an unresolved challenge [29] , [30] . Gregg et al . [16] , [17] used a chi-square metric that assumes no experimental biases are introduced during library construction , sequencing , genomic alignment , and that each sequencing read is independent of all other reads . Unfortunately these assumptions are often violated [30]–[34] , and systematic error in quantifying allele-specific expression by RNA-Seq is just now becoming apparent [29] , [35]–[37] . To begin to understand the underlying cause of inconsistent imprinting calls we investigated the accuracy of ASE quantification with RNA-Seq . It has previously been shown that ASE measured at the same SNP is highly reproducible across technical and biological replicates [17] , [38] . However , this comparison is immune to systematic error such as priming , fragmentation , and PCR biases that arise during library construction , sequencing chemistry [36] , and read alignment [31] . Since most sources of systematic error are sequence dependent , a more informative test would compare concordance in ASE within the same sample , but between independently sampled sites where the level of ASE is the same . We reasoned that SNPs within the same coding exon should satisfy this requirement since there are few known biological phenomena that could disrupt this relationship . Allele-specific premature transcriptional termination is possible , but has to our knowledge not been documented . Furthermore , premature termination codons are relatively infrequent [39] and those introduced by alternative splicing lead to immediate degradation by nonsense-mediated decay [40] . We did not use exons containing UTRs since transcription start sites vary significantly [41] and 3′UTRs are extensively processed [42] . We also restricted our analysis to RefSeq coding exons , which are based on experimentally validated transcripts to maximize the accuracy of same-exon SNP associations . We estimated the accuracy of ASE quantification as the frequency of SNP pairs within the same exon that agree on direction of bias . We considered all SNP pairs separated by more than 40 bp , which is the longest used sequencing read length [17] , to ensure independent sampling . At a p-value threshold of 0 . 0001 ( i . e . both p-values in each SNP pair are less than 0 . 0001 ) we observed 1 , 388 SNP pairs in our E17 . 5 BxC data , of which 278 ( 20 . 03% ) disagreed on direction of bias ( Figure 2A ) . If ASE measured with RNA-Seq was adequately modeled by random sampling , we would expect less than 0 . 1% to be discordant at this level of significance ( Figure 2B and see Materials and Methods ) . We observed exceedingly higher levels of discordance than expected across all levels of significance ( Figure 2B ) . Nonetheless , p-values computed from a basic chi-square test are predictive since higher thresholds result in lower rates of discordance and do not reach an asymptote ( within extent of available data ) . This indicates that: 1 ) our test is valid and not confounded by biological differences in ASE between SNPs in the same coding exon , and 2 ) counting statistics can be a reliable approach for identifying ASE , although at face-value lead to a major over-estimate of significance . For discovery of imprinted genes , a simple negative control that accounts for systematic error , technical variation , and biological variation is to ask how many SNPs/genes exceed significance in a mock reciprocal cross ( i . e . comparing samples with the same parental background as though they were from reciprocal crosses ) . In such a comparison any reciprocally biased expression cannot be caused by genomic imprinting and is a measure of the technical and biological variation of the experimental approach . Data from two animals of opposite sex was available for PFC and POA and enabled two mock-cross analyses ( Figure 3A ) . Strikingly , in the mock reciprocal cross , nearly as many imprinted gene calls exceeded the significance threshold used by Gregg et al . as in the reciprocal cross ( Figure 3B ) . Similar to our comparison in embryonic brains , the majority of the known imprinted gene calls from reciprocal crosses were the same , but novel calls were not ( Figure 3C ) . Comparing males to females in the reciprocal analysis controlled for sex-specific expression biases . We confirmed that mock-reciprocal hits are not caused by differences in sex by generating additional sequencing data from a male E17 . 5 brain sample . This enabled a comparison based on true ( sex-matched ) biological replicates and revealed that male vs male and male vs female BxC mock comparisons produced equivalent numbers of false positive measurements ( Figure 3D ) . While we used approximately half of the data to generate calls ( Figure 3B ) , randomly removing mapped reads from aggregated data revealed that sensitivity is not markedly different at 50% vs 100% of input reads ( Figure 3E ) . Furthermore , the estimated proportion of novel imprinted genes that are false-positives is not impacted by further down-sampling ( the slope of the line is consistent when sufficient data exists to overcome noise; Figure 3F ) and an overestimate on account of reduced sequencing depth is thus unlikely . We note that an aggregate mock comparison ( 1+4 vs 2+3 ) is not an informative negative control since potential sex-specific imprinted genes would not be balanced in this scheme and the output would be a mixture of true sex-specific imprinted genes and false-positives where the contribution of each is not clear . A high false-discovery rate may also explain the large number of sex-specific imprinted genes reported by Gregg et al . [16] since these , by definition [16] , only reach significance in a comparison of one reciprocal cross ( e . g . between males ) and not the other ( e . g . between females ) . If this were a reliable assay for identifying sex-specific imprinted genes then nothing should meet significance by comparing opposite sexes within each cross , since expression at a sex-specifically imprinted locus would always be biallelic in one animal . 51 imprinted genes in reciprocally crossed male PFC samples reached significance that did not reach significance in reciprocally crossed females at the threshold used by Gregg et al . ( but had sufficient coverage to make a call; 36 agreed with Gregg et al . [16] , p<1e-63 ) . However , a similar number reached significance in negative controls ( 63 genes in mixed-sex and 39 in mock reciprocals; Figure 4 ) . We obtained similar ratios for female-specific PFC imprinted genes and all POA comparisons ( Figure S1 ) , demonstrating that this approach is not sufficiently powered at the selected threshold of statistical significance . To estimate the total number of imprinted genes we first asked how many are detected in the four available datasets ( E15 brain , E17 . 5 brain , adult PFC and POA ) . Aggregating allele-specific reads across SNPs in the same gene improved our sensitivity in known imprinted regions ( data not shown ) and we thus applied this approach genome-wide . We also took advantage of all publicly available SNPs [17] , [26] , [27] and expanded our alignment reference to include the whole genome ( see Materials and Methods ) . Using the mock/reciprocal approach to estimate false-discovery ( Figure 5A ) , we proceeded with p = 1e-4 ( FDR<0 . 05 ) as a threshold of significance for calling a gene imprinted ( in addition to the standard reciprocal bias toward sex of parent and ≥10 reads in each cross ) . We selected p = 1e-4 as a threshold ( e . g . as opposed to 1e-3 ) since candidates with scores between 1e-3 and 1e-4 did not validate by pyrosequencing ( see below ) and sensitivity is negligibly impacted ( Figure 5A ) . We identified a total of 53 putative imprinted genes in at least one sample ( Figure S2 ) . 5/53 occurred in all 4 samples and 3 ( Eif2c2 , Cdh15 , and DOKist4 ) were validated by Gregg et al [17] . We also detected 56 genes that were previously known to be imprinted ( 27 in all 4 samples ) . Of the putative novel genes , 4 are probable extensions of known imprinted genes based on EST or transcription evidence derived from this data ( 3 of the 5 putative novel imprinted genes which recur in all 4 samples ) , 7 others are associated with known imprinted clusters ( within 1 Mb ) and 42 are completely novel ( Table S1 ) . From manual inspection we identified three distinguishing features among the known imprinted genes that we detected and reasoned these may be useful for predicting novel imprinted regions . These include: 1 ) reciprocal allelic bias and high sequencing depth ( reflected in conjunction as the imprinting score ) , 2 ) agreement on imprinted expression among neighboring SNPs , and 3 ) recurrence of signal across biological replicates and/or tissues . To further investigate the predictive potential of these features we identified 37 candidate imprinted loci that represent a range of values for each feature and tested these by pyrosequencing . In addition to 5 positive controls , we tested 17 candidate loci selected at random from the list of imprinted SNPs reported by Gregg et al [17] in E15 , 4 candidates with marginal imprinting scores ( between 1e-3 and 1e-4 ) , 2 candidates detected in adult but not embryonic brain , 7 candidates detected only in embryonic brain , and 5 candidates detected in at least two samples . All 17 candidates reported Gregg et al . are from the ‘complex’ category where the imprint does not agree with other SNPs in the same gene ( this category accounts for 94 . 8% of the novel imprinted genes reported [17] ) . Pyrosequencing validation suggests that all three features are predictive ( Figure 5B; Table S2; Figure 6 ) . To establish a level of technical and biological variance in our pyrosequencing assays , we first measured ASE for the 37 loci in biological replicates ( two BxC E17 . 5 brain samples ) and observed excellent agreement ( Figure S3 ) . This also enabled us to establish a meaningful threshold for detecting differences in ASE ( see Materials and Methods ) . In agreement with our results suggesting that the majority of the Gregg et al imprinting calls are false-positives as well our own imprinting calls on these 17 loci ( 16/17 predicted to be negative ) , none validated by pyrosequencing ( Figure 5B; Table S2 ) . Of 11/16 predicted parent-of-origin effects in embryonic brain ( incl . 5 positive controls ) that validated , all contain more than one SNP where parental bias was observed , suggesting that consensus SNP calls may be a valuable predictor . The imprinting score was also predictive; the average ( absolute ) value of the 6/11 novel parent-of-origin effects that validated was 28 . 4 vs 6 . 7 for the 5/11 predictions that did not validate and none of the four tested predictions in the 3–4 range validated ( Figure 5B; Table S2 ) . 5/7 predictions detected in only one of the embryonic samples as well as the two detected only in POA did not validate , suggesting that recurring detection in more than one related sample is also informative .
A recurring conclusion throughout this study is that there is no evidence for mammals having an order of magnitude more imprinted loci than was previously appreciated , as two recent papers claim [16] , [17] . Independent replication of this work , reanalysis that included negative controls to estimate false-discovery , and follow-up validation using an independent assay unilaterally suggest that the vast majority of the reported imprinted genes are false-positives explained by variation in the assay and experimental approach . This is in agreement with long-standing genetic estimates and other high-throughput screens , some of which used the same global approach [14] , [15] , that reported at most a few novel imprinted genes each . Furthermore , novel loci that typically validate ( this study incl . ) are close to known imprinted regions and are likely uncharacterized extensions of known imprinted transcripts and/or rely on an imprinting mechanism of an established region . Nearly all the novel candidates emerging from RNA-Seq screens thus far ( this study incl . ) are marginally significant relative to known expressed imprinted genes that typically have a strong signal . Is it possible then that hundreds to thousands of undiscovered imprinted genes exist ? Yes , but there is no evidence for it . These would need to occur in developmental stages or tissues not yet assayed or represent transcripts that are invisible to standard RNA-Seq . Examples of imprinted expression that evade RNA-Seq are non-polyadenylated transcripts , antisense transcripts imprinted/expressed to similar degrees and in opposite directions such that the signal cancels out , and biases that occur to such a minor extent that they are detected below the threshold of noise . Because RNA-Seq cannot be modeled with counting statistics that assume each read is randomly and independently sampled and free of systematic bias , it is imperative that this threshold is firmly defined . Without accounting for background , SNPs with very small parent-of-origin biases resulting from assay variance may appear imprinted with high statistical significance , particularly if the SNP is highly expressed . Even if some of these were real , our pyrosequencing efforts did not validate any . One could still argue that the effect is below the threshold of pyrosequencing detection ( e . g . less than a 5% difference in this study ) , but if this were the case then the effect on transcriptional load would be small and likely without functional consequence . In any case , the current state of RNA-Seq and analysis cannot detect these minor effects , even if they do exist . A more fruitful application of RNA-Seq to imprinting discovery in the near term may involve screening additional developmental time-points , tissues , and species . RNA-Seq methods that do not require polyA+ selection and that retain strand-specificity may also uncover novel transcripts of the noncoding/antisense variety , some of which have already been shown to be important for establishing and maintaining imprinting states [43] . Finally , integration with complementary global datasets , such as genome-wide allele-specific methylation maps , will improve both specificity of imprinting discovery as well as insight into the underlying mechanisms . Overall , 6/11 novel parent-of-origin effects validated by pyrosequencing , with the average allelic bias measured by pyrosequencing indicated in parentheses: U80893 ( 44 . 4% ) , Ccdc40 ( 5 . 5% ) , Bcl2l1 ( 12 . 5% ) , Mapt ( 3 . 6% ) , Adam23 ( 9 . 3% ) and Wars ( 3 . 9% ) . As with U80893 , Adam23 and Wars , Bcl2l1 is located in close proximity to known imprinted transcript ( all <150 kB from a known imprinted transcript ) . U80893 is a trinucleotide-repeat ( CAG ) containing gene of unknown function [44] . Ccdc40 is essential for left-right patterning in both mice and humans , full range cilia motility and a causal mutation for a variant of primary ciliary dyskinesia in humans [45] . Bcl2l1 inhibits apoptosis [46] . Mapt is primarily known as a major component of the prevalent neurofibrillary tangle pathophysiology in Alzheimer's disease [47] , suggesting that imprinted expression of Mapt in humans may have relevance to the inheritance and progression of Alzheimer's . Adam23 mutants are smaller than wild type littermates and exhibit tremors and ataxia [48] and Wars is uncharacterized . Three of these genes , Wars , Ccdc40 , and Mapt , were marginally biased in their parent-of-origin expression ( Figure 5B , Figure 6 , Dataset S1 ) . This may be due to a mechanism that causes incomplete silencing or due to allele-specific expression in only a subset of cell types/tissues comprising the organ assayed . Using more homogeneous samples provides an obvious path forward , but revealing the functional consequences to these types of imprinting cases is the real challenge since many novel imprinted genes awaiting discovery will likely fall into this category . An estimate for the total number of imprinted genes must account for several variables , including the increasing difficulty in validating novel candidates . Of the 6/11 candidates that validated in this study , only 2 of the 6 represent parent-of-origin effects that are clearly not associated with previously known imprinted regions ( Figure 5B and Figure S2 ) . This severely limits our power to establish a firm number and an estimate should be interpreted with caution . Nonetheless , assuming a confirmation rate of 37 . 5% ( 3/8 ) for the remaining untested embryonic brain candidates and a detection sensitivity of 56/128 ( known imprinted genes ) extrapolates to an estimate of 37 imprinted genes awaiting discovery and validation , yielding an estimate for a total number of ∼175 imprinted genes . Although the statistically-derived imprinting score is the strongest predictor of imprinting , it cannot be interpreted as a direct measure of probability . As others have noted , RNA-Seq is not free of systematic error [29] , [31] , [35] , [36] which may impact measurement of ASE . We show that allele-specific read counts are substantially less accurate for quantifying ASE than expected from a random sampling approach . Although our metric for quantifying error in measuring ASE is concordance in direction of bias , this measure becomes synonymous with false-positive rate in identifying ASE as sampling iterations become large and the full range of significance thresholds is represented . Our model could thus be extended to serve as a correction for the significance of ASE to yield an empirical measure of false-discovery when identifying regions with ASE . Since detection of imprinting requires demonstration of reciprocal bias , most of the effect introduced by systematic error ( which is sequence-dependent ) is likely eliminated . Our analysis confirms this expectation: FDR estimated from a mock-reciprocal analysis ( Figure 5A ) reaches manageable levels at more relaxed thresholds of significance than modeling ASE ( Figure 2B ) . Second , recurrence of imprinted expression across related samples increases the likelihood that novel candidates validate . Although exceptions exist [13] , if expression of a given gene is imprinted in adulthood , its expression is generally also imprinted in the developmental precursor of that tissue . Data in Wamidex [13] , for example , suggests ∼90% concordance between imprinted expression of an adult tissue with its precursor tissue . 20 . 0% and 75 . 0% of the putative novel imprinted genes in the POA and PFC respectively were also imprinted in the E15 and/or E17 . 5 brain ( Figure S4 ) . These numbers increase to 68 . 5% and 93 . 3% respectively , if known imprinted genes are included . The rate of independent confirmation amongst putative imprinted transcripts was higher for those reaching the threshold in multiple samples , suggesting that these figures are likely an underestimate . Other high probability candidates that were not assayed independently but fit this profile include: AK039650/AK044369 , DQ715667 , Klhdc10 , and C230091D08Rik , while Herc3 is likely to confirm in the adult POA . Third , consensus in parent-of-origin allelic bias among neighboring SNPs provides additional predictive potential . Read aggregation across SNPs in the same gene enabled detection of concordance ( the same gene exhibiting imprinted expression in more than one sample ) in several genes that appeared biallelic using p<0 . 05 without read aggregation [17] . Examples include Wars ( POA ) , Adam23 ( POA ) , Klhdc10 ( E15 , POA ) , and Cdh15 ( E15 , PFC ) . Fourth , proximity of novel candidates to known imprinted regions or to each other is also predictive . Most known imprinted genes occur in clusters that can span hundreds of kilobases [14] and typically share regulatory mechanisms . 7 ( of 8 ) putative novel imprinted genes that were identified in at least two samples are associated with known imprinted regions ( <1 Mb ) , as well as three detected in only one sample ( Figure S2 and Table S1 ) . Adam23 is ∼100 kB from Zdbf2 , C230091D08Rik is ∼1 kB from Ube3a , U80893 may be part of the contiguous transcription extending from Zfp127 [14] , Aceview [49] suggests DQ715667 is part of Eif2c2 which is 500 kB from Kcnk9 , AK039650/AK044369 appears to be an extension of Kcnk9 , Klhdc10 is ∼250 kB from Mest , and Wars is ∼150 kB from Begain . Furthermore , 6 putative novel imprinted genes form 3 clusters of 2 genes within 200 kB of each other ( Table S1 ) . Two of these , Wars and Slc25a29 , are <600 kB from the Dlk1 locus . In conclusion , until we can accurately model ASE measured with RNA-Seq , estimating FDR of imprinted gene discovery will ideally be done empirically . The additional criteria mentioned above can be used to further rank novel candidates , but since no combination of criteria was absolutely predictive among the novel imprinted genes identified in this study , we assert that independent validation is essential for making definitive claims about imprinting of any gene .
We downloaded sequencing data and SNP calls from GEO ( GSE22131 ) and used the same strategy as Gregg et al . [16] , [17] to score imprinted expression . We aligned the data to UCSC known genes ( transcripts ) and noncoding RNAs downloaded from the Functional RNA Database [50] . We called a SNP ‘imprinted’ if greater than 50% of the reads in each sample mapped to alleles from the same parental sex ( p<0 . 05 , chi-square test ) in both samples . We compiled a non-redundant set of gene models from UCSC known genes [28] by taking the union of transcribed bases when multiple isoforms overlap ( i . e . collapse isoforms into one gene model that contains all exons ) , yielding 26 , 214 models , of which 19 , 867 were coding ( available under GSE27016 ) . 13 , 604 gene models had at least one SNP identified by Gregg et al [17] . We called a gene imprinted if it contained one or more imprinted SNP ( s ) . Sex-specific calls were made as described [16] . The raw fastq data was realigned against RefSeq transcripts and the mouse genome ( mm9 ) using Novoalign . 19 . 6 million C57Bl/CAST SNPs , representing the union of Perlegen [26] and Sanger [27] , were masked prior to alignment to reduce alignment biases caused by more sequence mismatches between CAST and C57Bl reference genome than C57Bl reads [31] . Uniquely mapping reads ( in genomic space; ∼80% of all reads on average ) were retained for further analysis . Non-redundant RefSeq coding exons were identified to avoid multiple sampling bias arising from the same SNP present in multiple isoforms . In cases where more than one coding exon overlapped in the genome , the longest coding exon was selected . All pairs of SNPs within each exon were used for comparisons ( e . g . three comparisons were done if three SNPs were present ) , but SNP pairs separated by 40 bp or less ( the longest sequencing read length [17] ) were not considered to ensure independent sampling . Significance of ASE was computed using a chi-square test as described above , and the less significant ( i . e . higher ) p-value was used to set the significance threshold for that comparison ( x-axis in Figure 2B ) . For the simulated analysis , the more significant SNP in each pair was replaced with expected counts and adjusted for variance with random sampling from the chi-square distribution with one degree of freedom . For example , if one SNP had 30 and 70 C57 and CAST reads respectively ( expect 50 and 50 , chi-square p = 6 . 3e-5 , df = 1 ) and the other had 30 and 20 ( expect 25 and 25 , chi-square p = 0 . 157 , df = 1 ) , the allelic counts of the first SNP would be replaced by 60 and 40 ( ±variance ) . If a random sampling of the chi-square distribution yields , for example , χ2 = 1 , rearranging Pearson's formula for the χ2 statistic and solving the quadratic reveals simulated counts of 65 or 55 and 35 or 45 for C57 and CAST respectively . A schematic summary of the approach is provided ( Figure S5 ) . Replacing the more significant SNP ( as opposed to the less significant SNP ) was done to ensure direct comparison with the observed data where the less significant p-value defines the threshold of significance for both SNPs . Replacing the less significant pair led to a further reduction in expected discordance ( data not shown ) . This process was repeated for each SNP pair and concordance was computed as described above . Imprinting calls were made using criteria described above . To minimize sampling bias reads aligned to chromosome X and the mitochondrial chromosome were not considered in this analysis . Furthermore , all samples were normalized to have the same number of aligned input reads . For example , for the PFC comparison , the BxC male sample had the fewest aligned reads so reads were randomly removed from the three remaining samples ( BxC female , CxB male , and CxB female ) to exactly match BxC male . All animals were housed and treated in accordance with the Institutional and Governmental Animal Care Committee guidelines of the University of Toronto . Whole brains were dissected from 2 male BxC E17 . 5 brains , 1 female BxC E17 . 5 brain and 1 male CxB E17 . 5 brain , snap chilled in liquid nitrogen and immediately homogenized in Trizol . Total RNA was extracted with Trizol according to the manufacturer's recommendations ( Invitrogen ) and integrity was confirmed on an Agilent Bioanalyzer ( RIN>9 for both samples ) . PolyA+ RNA was isolated using a Dynabead mRNA Purification Kit according to the manufacturer's instructions ( Invitrogen ) . Double stranded cDNA libraries were made using the Illumina mRNA-Seq kit according to the manufacturer's recommendations ( Illumina ) and converted into libraries ( adapter ligation , PCR , cleanup ) using a NEBNext kit according to the manufacturer's recommendations ( NEB ) . We retained strand-specificity by using dUTP during second-strand synthesis and an UNG treatment prior to the final amplification as described previously [51] . Libraries were verified on an Agilent Bioanalyzer and by qRT-PCR , and each sequenced to 36 bp on an Illumina HiSeq 2000 platform , yielding on average 80 . 1 million reads . Raw data and alignments are available at GEO under accession GSE27016 . Brains were processed independently such that each library is derived from a unique sample without pooling . For global discovery of novel imprinted genes we aligned to gene models as described above , all possible splice-junction sequences ( 100 bases; 50 bases from each flanking exon ) representing up to two exon-skipping events in these gene models , as well as the genome . We converted all alignments to genomic coordinates and retained uniquely mapping reads for further analysis . 79 . 9% of reads aligned uniquely ( on average 80 . 1% per sample ) , of which 7 . 7% spanned splice-junctions ( 17 . 6% of reads that overlapped coding exons also spanned splice junctions ) and 0 . 12% represented exon-skipping events . Imprinting was assessed at each SNP as described above . We identified SNPs amenable to pyrosequencing ( non-consecutive nucleotides separated from other SNPs ) within transcripts of interest and designed assays using Pyromark Assay Design 2 . 0 requiring assay scores >87 . The sequence ‘CGCCAGGGTTTTCCCAGTCACGAC’ was added to the 5′ end of all primers designated for biotinylation to enable biotin incorporation during PCR following an approach reported previously [52] with some modifications . Specifically , biotinylated amplicons were generated directly from RNA using the Pyromark OneStep RT-PCR kit ( Qiagen ) according to the manufacturer's instructions with the addition of a third HPLC purified universal biotinylated primer ( biotin-CGCCAGGGTTTTCCCAGTCACGAC ) added at 9/10 the recommended molarity to supplement the RT-PCR primer designated for biotinylation , which was added at 1/10 the recommended molarity . Other combinations ranging from 5∶5 to 9∶10 did not result in a noticeable difference in product yield and size ( as assayed by an Agilent Bioanalyzer ) or pyrosequencing performance . We performed sequencing on a Pyromark Q96 MD ( Qiagen ) according to the manufacturer's instructions and quantified allelic bias using the AE quantification software included with the instrument ( all traces available in Dataset S1 ) . To maximize sensitivity of detecting small allelic changes we defined the null ratio by sequencing DNA as described previously [53] . Transcripts were called imprinted if the difference in allelic bias between the reciprocal samples were >5 . 02% and where each ratio was reciprocally biased ( i . e . in opposite directions ) relative to ratio obtained with DNA . 5 . 02% represents 2 standard deviations of variance determined from comparing allelic ratios between biological replicates ( Figure S3 ) and corresponds to a significance of p<0 . 022 . Both biological replicates had to meet these criteria and DNA ratios were averaged between the two DNA samples prior to normalization . A negative pyrosequencing call includes cases that do not meet significance as well as cases where reciprocal bias was not observed . The majority of negatives fall into the latter category and relaxing the threshold of significance to p<0 . 2 did not change the outcome of any calls . 5/42 assays attempted were removed due to technical failure ( two because of a high skew in the DNA ratio and three where the Pyromark AE quantification software could not accurately identify peaks ) leaving 37 assays successfully executed across all 5 samples .
|
Typically both copies of mammalian genes are expressed , but in some cases , “imprinting” restricts expression to the maternal or paternal copy . Having two copies of each gene is considered advantageous since in enables compensation when one does not function properly . Why imprinting evolved and its utility to each sex is widely debated , and having a complete catalog of imprinted genes and their functions is essential for fully characterizing this phenomenon . 25 years of screening has revealed about 130 imprinted genes , and the slowing rate of discovery suggests that we are reaching saturation . Two recent studies based on high-throughput sequencing of RNA reported more than 1 , 300 imprinted genes . To understand the basis of this paradigm shift , we first attempted to reproduce these results . Unable to do so , we performed additional analyses that show that most of these discoveries are due to noise in the experimental approach and assay . We remedy this with new methods that account for this noise and applied them to identify 50 novel putative imprinted genes . These methods will be useful for identifying genuine novel cases of imprinted expression as this type of screening approach becomes broadly utilized .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"dementia",
"genomics",
"neurological",
"disorders",
"neurology",
"genetics",
"biology",
"computational",
"biology",
"genetics",
"and",
"genomics"
] |
2012
|
Critical Evaluation of Imprinted Gene Expression by RNA–Seq: A New Perspective
|
To investigate the clinical features , clinical course of granuloma , serologic findings , treatment outcome , and probable infection sources in adult patients with ocular toxocariasis ( OT ) . In this retrospective cohort study , we examined 101 adult patients diagnosed clinically and serologically with OT . Serial fundus photographs and spectral domain optical coherence tomography images of all the patients were reviewed . A clinic-based case-control study on pet ownership , occupation , and raw meat ingestion history was performed to investigate the possible infection sources . Among the patients diagnosed clinically and serologically with OT , 69 . 6% showed elevated immunoglobulin E ( IgE ) levels . Granuloma in OT involved all retinal layers and several vitreoretinal comorbidities were noted depending on the location of granuloma: posterior pole granuloma was associated with epiretinal membrane and retinal nerve fiber layer defects , whereas peripheral granuloma was associated with vitreous opacity . Intraocular migration of granuloma was observed in 15 of 93 patients ( 16 . 1% ) . Treatment with albendazole ( 400 mg twice a day for 2 weeks ) and corticosteroids ( oral prednisolone; 0 . 5–1 mg/kg/day ) resulted in comparable outcomes to patients on corticosteroid monotherapy; however , the 6-month recurrence rate in patients treated with combined therapy ( 17 . 4% ) was significantly lower than that in patients treated with corticosteroid monotherapy ( 54 . 5% , P = 0 . 045 ) . Ingestion of raw cow liver ( 80 . 8% ) or meat ( 71 . 2% ) was significantly more common in OT patients than healthy controls . Our study discusses the diagnosis , treatment , and prevention strategies for OT . Evaluation of total IgE , in addition to anti-toxocara antibody , can assist in the serologic diagnosis of OT . Combined albendazole and corticosteroid therapy may reduce intraocular inflammation and recurrence . Migrating feature of granuloma is clinically important and may further suggest the diagnosis of OT . Clinicians need to carefully examine comorbid conditions for OT . OT may be associated with ingestion of uncooked meat , especially raw cow liver , in adult patients .
Toxocariasis is a globally prevalent illness caused by infestation of the parasite Toxocara canis or Toxocara cati larvae , which is the most ubiquitous gastrointestinal helminth in dogs and cats [1] , [2] . Human beings generally become infected through ingestion of embryonated eggs from contaminated sources such as soil or improperly cooked paratenic hosts [1] , [2] . In addition , pet owners can sometimes be accidentally infected by their dogs or cats . After a human ingests the eggs , the eggs hatch in the small intestine and release parasitic larvae . These larvae then penetrate the intestinal wall , enter the circulation , and migrate to organs where they induce inflammatory reactions . Symptoms of the infection vary , depending on the involved organs [1] , [2] . Larvae that migrate to the eye cause ocular toxocariasis ( OT ) , which is relatively uncommon and occurs primarily in children . They are most commonly infected through playground and sandbox where contaminated dirt and/or sand may be ingested because of playing habits and poor hygiene [1] , [3] , [4] . Several OT case series have addressed the demographics , clinical features , and causes of vision loss [3] , [5]–[12] . These reports primarily describe young patients , who were under 16 years of age [3] , [5]–[11] . However , little is known about the epidemiologic , demographic , and clinical features of OT in adult patients . Most studies had a cross-sectional design , and the clinical course of OT has not been studied extensively . Furthermore , although the mainstay of OT treatment involves steroid use to reduce inflammatory responses [13] , the treatment regimen for OT has not been standardized . In particular , the efficacy of combining steroid therapy with anthelmintics has not been determined . In the present study , we aimed to elucidate the clinical features and course of OT in 101 adult patients with OT , in whom a Toxocara infection was confirmed with ELISA serum testing for IgGantibody to the Toxocara larva crude antigen [2] , [13] , [14] . In addition to the ELISA titers , complete blood count ( CBC ) and serum immunoglobulin E ( IgE ) levels were obtained in each patient to identify the hematologic/immunologic indicators of OT . Furthermore , to determine the potential sources of Toxocara exposure , history of pet ownership , occupation , and raw meat ingestion of the patients were investigated and compared to those of healthy controls . In addition , spectral domain optical coherence tomography ( SD-OCT ) was performed to investigate OT-related pathologic retinal changes .
A retrospective cohort study was conducted in all consecutive adult ( >20 years old ) patients diagnosed with OT at 3 institutions ( Seoul National University Hospital , Seoul National University Bundang Hospital , and Seoul Metropolitan Government Seoul National University Boramae Medical Center ) between January 2009 and June 2013 . A clinical diagnosis of OT was made based on ( 1 ) typical clinical features of OT [1] , [3] , [4] , ( 2 ) positive results by serologic testing , and ( 3 ) exclusion of other possible causes of granuloma such as ocular toxoplasmosis ( absence of Toxoplasma-specific IgG and IgM ) , sarcoidosis ( absence of hilar adenopathy or upper lobe disease on chest radiography , absence of skin lesions suggesting sarcoidosis , absence of hypercalcemia or nephrocalcinosis , and normal levels of angiotensin-converting enzyme ) , tuberculosis ( negative results on interferon gamma release assays , absence of serpiginous choroiditis or retinal vasculitis suggesting ocular tuberculosis , and clinical response to topical/systemic steroid without anti-TB medication ) , and fungal infection ( absence of disseminated fungal diseases , no history of penetrating ocular trauma or surgery within a 6-month period , absence of retinal hemorrhage , which is often observed in eyes with fungal infection but seldom observed in eyes with OT , and clinical response to topical/systemic steroid without anti-fungal agents ) . The typical clinical features of OT included the presence of a peripheral granuloma ( focal , white peripheral nodule with pigmentary scarring or traction retinal detachment ) , posterior pole granuloma ( focal , white nodule with or without posterior pole variable pigmentation ) , or nematode endophthalmitis ( diffuse intraocular inflammation and serology results only positive for Toxocara ) [1] , [3] , [4] . Among the patients with clinical OT , specific IgG antibody titers were measured by indirect ELISA , based on the T . canis larva crude antigen [14] . The mean titer of 2 ELISA tests was used in analyses . An ELISA titer of ≥0 . 250 was considered serologically positive since a previous study to determine the sensitivity and specificity of ELISA testing in patients with toxocariasis showed that a cut-off optical density of 0 . 250 has a sensitivity and specificity of 92 . 2% and 86 . 6% , respectively [14] . The ELISA test was performed on serum in all the patients and on a 1-ml undiluted vitreous sample ( obtained during vitreous surgery ) in 9 patients who were treated with vitreoretinal surgery . Additionally , a CBC was performed and serum total immunoglobulin-E ( IgE ) was examined to evaluate any serologic/immunologic abnormalities . The results of abdominal computed tomography ( CT ) and chest CT were analyzed in this study , if they were performed within 6 months of OT diagnosis . From these images , the prevalence of granulomas or abscesses in other organs such as the lung or liver , as determined by a trained radiologist , were determined . Best-corrected visual acuity ( BCVA ) , intraocular pressure ( IOP ) , slit-lamp biomicroscopy findings , and dilated fundus examination findings were reviewed in all the patients . Inflammation in the anterior chamber and vitreous chamber was graded based on the number of cells in a 1×1 mm slit beam under maximal light intensity and magnification [15] . Briefly , grade 0 indicated <1 cell; grade 0 . 5+ , 1 to 5 cells; grade 1+ , 6 to 15 cells; grade 2+ , 16 to 25 cells; grade 3+ , 26 to 50 cells; and grade 4+ , >50 cells . Funduscopic findings were photographed with a Kowa VX-10 fundus camera ( Kowa Co Ltd , Tokyo , Japan ) . Changes in granuloma size and location were evaluated using photographs from each follow-up visit . SD-OCT ( Spectralis , Heidelberg engineering , Heidelberg , Germany ) was performed to sectionally image the retina and view pathologic changes in eyes with granuloma and other vitreoretinal complications . Patients with OT were treated with drugs or surgery based on symptom severity , inflammation , and retinal comorbidities . Drug therapy involved corticosteroids when intraocular inflammation was present . Systemic ( oral prednisolone; 0 . 5–1 mg/kg/day loading dose and tapering ) and topical ( prednisolone acetate 1% four times a day ) corticosteroids were used depending on the site of inflammation . Patients with eosinophilia or elevated serum IgE level were treated with albendazole ( 400 mg twice a day for two weeks ) . For patients with retinal comorbidities requiring surgery such as visually significant epiretinal membrane , vitreous opacity obscuring visual axis , and tractional or rhegmatogenous retinal detachment , pars plana vitrectomy was performed . Patients were separated into 4 groups , based on the medical treatment they received ( i . e . , combined steroid and albendazole , albendazole only , steroids only , or no treatment ) . For patients with a ≥3-month follow-up period , treatment response was evaluated based on clinical characteristics observed at the follow-up visits . These included BCVA , intraocular inflammation grades , symptom improvements , and recurrence rates . Recurrence was defined as returning intraocular inflammation or new granuloma development . Treatment outcomes were assessed in each treatment group . For investigation of the probable infection sources of OT , we conducted a standardized interview and ensured complete responses through improved understanding of the participants to enhance the validity of the interview . During a face-to-face interview , a trained interviewer ( medical doctor ) used standardized interview procedures to collect data concerning history of eating raw animal tissues and contact with animals and soil during the period between January 2011 and June 2013 . Both patients and controls were asked the same set of questions , including puppy/kitten exposure; history of ingestion of raw animal liver , raw meat , and raw animal blood; and occupation-associated contact with animals or soil . The data obtained also included the time of ingestion and the species of animals . For the interview , the questions and uniform nonverbal signals were presented in exactly the same way by one trained interviewer to avoid introducing biases into the responses . Patients who visited our clinics during January 2011 and June 2013 and showed no abnormal ocular findings on complete ophthalmic examination were selected as controls . Among 59 control subjects who responded to our interview , 50 were matched for age ( within 3 years ) and sex with the 52 patients who responded to our interview . Between the included controls ( n = 50 ) and the others ( n = 9 ) , there were no significant differences in demographic features and probable infection sources , except age ( 51 . 0±11 . 4 in the included controls and 28 . 3±19 . 6 in the excluded , P<0 . 001 ) . Descriptive statistical analyses were performed on the demographic data , clinical features , funduscopic and OCT findings , serologic markers , systemic involvement , and granuloma clinical course . Snellen BCVA measurements were converted into logarithmic minimum angle of resolution ( logMAR ) equivalent values for statistical analysis . The association between funduscopic and OCT findings and granuloma location was assessed by a chi-square test , which compared the frequencies of funduscopic findings in patients with posterior pole granulomas to those with peripheral granulomas . Treatment outcomes were compared between and within groups using the Wilcoxon signed-rank test and Mann–Whitney test , respectively . Continuous and interval data are reported as mean ± standard deviation . Statistical analyses were performed using SPSS for Windows ( Ver . 18 . 0 , Statistical Package for the Social Sciences , SPSS Inc . , Chicago , IL ) , and a P value<0 . 05 was considered statistically significant . The Institutional Review Board approved this study ( Approval #: B-1101/120-102 ) and all patient data were anonymized for the analysis . The study adhered to the tenets of the Declaration of Helsinki .
The demographic and clinical features of the patients are presented in Table 1 . Most OT patients were men ( 76 of 101 , 75 . 2% ) , and the mean presentation age was 51 . 7±12 . 6 years ( range: 21–77 years ) . Three of the 101 patients ( 3 . 0% ) were known to have been infected with Toxocara before they were diagnosed with OT . In the other 98 patients , OT was the first symptom of toxocariasis . Systemic involvement of the Toxocara granulomas is also summarized in Table 1 . CT images revealed liver ( 17 . 6% ) or lung ( 42 . 9% ) granulomas in a significant proportion of our patients . Ninety-three of 101 patients ( 92 . 1% ) were diagnosed with OT based on the presence of a retinal granuloma . Eight patients ( 7 . 9% ) were diagnosed based on diffuse intraocular inflammation and a positive Toxocara antibody ( ELISA ) . Of the 8 patients with nematode endophthalmitis , OT was confirmed in 2 patients , based on the presence of a peripheral granuloma , which became visible only after vitreous opacities had been cleared by vitrectomy . Depending on the location of the granuloma , Toxocara granuloma was classified as posterior pole ( 47 eyes , 50 . 5% ) , peripheral ( 41 eyes , 44 . 1% ) , or combined ( both posterior pole and peripheral granulomas , 5 eyes , 5 . 4% ) . Intraocular inflammation was observed in 78 eyes ( 77 . 2% ) , most of which had intermediate uveitis ( 53 eyes , 67 . 9% ) . All OT cases were unilateral . Mean BCVA changed from 0 . 51 ( 20/64 Snellen equivalent ) ±0 . 65 ( range: no light perception [NLP] to 30/20 ) at baseline to 0 . 45 ( 20/56 Snellen equivalent ) ±0 . 65 ( range: NLP to 30/20 ) at the final visit . Seventeen ( 16 . 8% ) and 14 ( 13 . 9% ) patients had severe vision loss ( BCVA<20/200 Snellen equivalent ) at baseline and the final visit , respectively . Possible causes of vision loss in patients with OT include retinal damage caused by granuloma itself , comorbidities of OT ( Table 1 ) , and intraocular inflammation . Eosinophilia ( >500 eosinophils/µl peripheral blood or ≥10% of total white blood cell count [16] ) at the time of diagnosis was noted in 10 of 86 patients ( 11 . 6% ) in whom CBC results were available . Increased serum IgE level ( >100 unit/ml ) was noted in 39 of 56 patients ( 69 . 6% ) . Mean ELISA titer for serum Toxocara IgG was 0 . 398±0 . 115 ( range: 0 . 254–0 . 737 ) . A few photographs demonstrating retinal and vitreous findings are shown in Figure 1 . In addition to retinal granuloma , patients with OT showed retinal nerve fiber layer ( RNFL ) defect ( Figure 1A ) in 32 of 101 eyes ( 31 . 7% ) , epiretinal membrane ( ERM , Figure 2B and 3C ) in 27 eyes ( 26 . 7% ) , vitreous opacity ( Figure 1F ) in 22 eyes ( 21 . 8% ) , retinal detachment ( Figure 1E ) in 13 eyes ( 12 . 9% ) , macular edema in 4 eyes ( 4 . 0% ) , and macular hole in 2 eyes ( 2 . 0% ) , as summarized in Table 1 . Table 2 shows the association between these vitreoretinal comorbidities and the location of granuloma ( posterior pole or peripheral retina ) . For example , eyes with a posterior pole granuloma had more frequent RNFL defects ( 53 . 2% vs . 7 . 3% , P<0 . 001 ) and ERMs ( 40 . 4% vs . 14 . 6% , P = 0 . 007 ) than eyes with a peripheral granuloma . In addition , vitreous opacity was observed more often in eyes with a peripheral granuloma than in those with a posterior pole granuloma ( 31 . 7% vs . 12 . 8% , P = 0 . 031 ) . Pathologic retinal changes were visible on SD-OCT images , which showed a moderately hyper-reflective round mass that sometimes had posterior shadowing ( Figure 2 ) . Granulomas were observed in almost all retinal layers . Secondary ERM was also commonly seen , and Figure 3 shows the course of retinal damage that leads to vision loss from granuloma and other OT-associated retinal pathologies . Three patterns were noted in the clinical course of Toxocara granuloma: complete/partial granuloma resolution , persistent granuloma , and granuloma migration ( Table 3 ) . Granulomas completely or partially resolved in 36 of 93 patients ( 38 . 7% ) , with approximately half of these resulting in pigmentary scarring ( Figure 4 ) . Forty-two patients ( 45 . 2% ) showed no significant changes in the size , number , or location of granulomas . Continuous or discontinuous granuloma migration within the eye was observed in 15 patients ( 16 . 1%; Figure 5 ) . In continuous migration ( 12 eyes , 12 . 9% ) , the Toxocara granuloma migrated but remained adjacent to the originally observed location . However , in discontinuous migration ( 4 eyes , 4 . 3% ) , the granulomas moved discontinuously ( relocated far from the originally observed location ) , increasing the total number of granulomas . One patient showed both types of intraocular migration . Table 4 shows treatment responses to the various OT medical therapies . Four types of treatments—combined corticosteroid and albendazole use , corticosteroid use only , albendazole use only , and no treatment—were performed for our patients . The use of corticosteroids significantly decreased the degree of intraocular inflammation ( Table 4 ) , but there was no significant improvement in BCVA 3 months after the initiation of drug therapy in any group . In patients with active intraocular inflammation , there was no significant difference in changes in inflammation grade ( P = 0 . 619 ) , BCVA ( P = 0 . 445 ) , or symptomatic improvement ( P = 0 . 274 ) between patients treated with only corticosteroids and those receiving a combination of corticosteroid and albendazole ( Figure S1 ) . In those without active intraocular inflammation , no significant difference was noted in the changes in inflammation grade ( P = 1 . 00 ) , BCVA ( P = 0 . 855 ) , and symptoms ( P = 0 . 206 ) between patients treated with ( Albendazole only group ) and without albendazole ( No treatment group ) . However , the 6-month rate of recurrence was significantly lower in the combination treatment group ( 17 . 4% ) than in the steroid only group ( 54 . 5% , P = 0 . 045 ) . In eyes without active inflammation , however , no recurrences were observed in both the Albendazole only and No treatment groups . Thirty-two of 101 patients ( 31 . 7% ) were surgically treated due to ERM ( n = 19 ) , vitreous opacity ( n = 9 ) , and/or retinal detachment ( n = 2 ) . The surgical outcomes ( i . e . , BCVA , anatomic success , symptomatic improvement , and recurrence ) in each surgical indication are summarized in Table S1 . Anatomic success , defined as complete removal of the ERM , vitreous opacity , or retinal reattachment , was achieved in 13 ( 68 . 4% ) , 8 ( 88 . 9% ) , and 2 ( 50% ) patients with ERM , vitreous opacity , and retinal detachment , respectively . Table 5 lists the probable sources of infections in adult patients with OT . In demographic features , no significant differences were observed between the patient and control groups in the mean age ( 51 . 6±13 . 0 in the patient group and 51 . 0±11 . 4 in the control group , P = 0 . 81 ) and sex distribution ( men∶women = 38∶12 and 39∶13 in the patient and control groups , respectively , P = 0 . 92 ) . Compared to healthy controls , there were no significant differences in the proportion of patients who had ingested raw animal blood , were exposed to puppies or kittens , or had occupation-associated contact with animals and/or soil . However , raw animal meat ( 71 . 2% vs . 52% , odds ratio [OR] = 2 . 28 , P = 0 . 047 ) or cow liver ( 80 . 8% vs . 22 . 0% , OR = 14 . 9 , P<0 . 001 ) was ingested significantly more often in OT patients than in normal controls .
Despite being the most prevalent human helminth infection in industrialized countries [17] , toxocariasis remains relatively unknown to the public [2] . This study describes the pathologic changes caused by OT-associated retinal granulomas using SD-OCT images . The clinical course of granuloma was also examined , and we showed that intraocular granuloma migration is an important and distinguishing clinical feature of OT . In addition , our study showed an association between OT and ingestion of raw cow liver or uncooked meat . This information may also be helpful in diagnosing OT in adult patients . Systemic and ocular manifestations of toxocariasis have rarely been reported in the same group of patients , and only a few such cases have been described in the literature [18] . In our study , both ocular larva migrans ( OLM ) and visceral larva migrans ( VLM ) were assessed in the same group of patients who had undergone ocular examination and liver or chest CT , although it was not proven that granulomas on CT images were caused by Toxocara infection . A significant proportion of our patients had liver or lung granulomas on CT images , and further study is needed to better understand the association between VLM and OLM . Given that the vast majority of patients ( 98 of 101 patients , 97 . 0% ) with toxocariasis were diagnosed first with OT , a thorough ophthalmologic examination is important for the detection of human toxocariasis . The present study identified several diagnostic serologic markers of Toxocara infection . The standard , current test for diagnosing human toxocariasis is detection of serum anti-toxocara IgG using an indirect ELISA based on the Toxocara larva antigen [2] , [13] . Testing of intraocular fluids for Toxocara antibodies has been helpful in diagnosing toxocariasis in a few previous studies [19]–[21] and in some patients of our study . However , the low positive rates ( 33% ) obtained by using the same cut-off value with serum antibodies ( 0 . 250 ) for vitreous antibodies , may not be acceptable for OT detection . Because there is no consensus on the cut-off titers for vitreous antibodies , further research on the diagnostic capabilities of vitreous ELISA is needed . Our study also revealed that serum IgE level is elevated in about 70% of OT patients . Eosinophilia was not as helpful as serum anti-toxocara IgG ( ELISA ) or total IgE evaluations , although it may indicate the presence of VLM , as shown in previous reports [2] , [13] . Toxocara-associated RNFL defects , ERMs , and vitreous opacities are common comorbidities of OT . These features were associated with granuloma location , which suggests that careful examination for these associated complications is necessary in patients with retinal granulomas . Vitreous opacities and ERMs were common causes of vision loss in our OT patients , for which surgical treatment was needed . Secondary ERMs in eyes with OT progressed rapidly , resulting in severe retinal distortion and vision deterioration . SD-OCT revealed retinal granulomas and secondary complications , including ERM and tractional membranes , and thus played a critical role in the clinician's decision to perform surgery . Therefore , in patients with OT , SD-OCT may be an important imaging modality for diagnosis and decision making in clinical settings . The migration of Toxocara larvae from the circulatory system into the posterior segment of the eye causes OT . Although migration is a key feature of Toxocara larvae , migration within the eye and its clinical significance have not been studied . In this report , we demonstrated a continuous and a discontinuous pattern of intraocular migration . These patterns have been individually reported in different case reports [22] , [23] , but their incidences have not been determined in longitudinal studies with a large number of patients . Our study showed that 12 . 9% and 4 . 3% of patients had continuous and discontinuous migration , respectively . In addition , our findings show clinical implications of the intraocular migration of Toxocara larvae . Continuous migration widened RNFL defects and discontinuous migration increased the number of RNFL defects since localized RNFL defects followed granuloma formation . In one case , discontinuous migration resulted in a macular granuloma , which resulted in significant macular destruction and subsequent vision loss , as well as a secondary ERM ( Figure 3 ) . Although the mainstay treatment for OT is the use of corticosteroids to reduce ocular inflammation , this treatment has not yet been standardized . In our study , compared to steroid monotherapy , a combination of albendazole and corticosteroids resulted in a lower rate of recurrence over 6 months . The efficacy of this combination therapy , i . e . , no ocular inflammation recurrence and visual acuity improvement , has been shown previously in one case [24] . Although our results showed no significant difference in BCVA before and after treatment , the reduced risk of OT recurrence favors the use of corticosteroids with albendazole in patients with severe OT . Therefore , we recommend using both corticosteroids and albendazole to minimize severe , recurrent inflammation and associated retinal damage . Our demographic analyses revealed that OT predominantly occurred in men , and disease transmission from pets was less frequent than that reported previously [3] . Male predominance has been previously reported in Japanese [25] ( male∶female ratio = 2 . 5∶1 ) and Korean [26] ( 4∶1 ) populations . In these studies , the mean age was >30 years , which is much higher than that that in studies conducted in Western countries , in which patients were generally <20 years old . Together , these results suggest that East Asian men may have a toxocariasis-related behavior , for example , the ingestion of raw cow liver , which is served in some restaurants in East Asian countries and is believed to be nutritionally beneficial for the middle-aged . Indeed , most reported cases from East Asia involved middle-aged men with a history of ingesting uncooked meat from infected animals [19] , [25]–[28] . Contact with a puppy or kitten , which was reported in 82% of patients in a previous OT study [3] , has been considered a major infection source , but was only reported in approximately 20% of our OT patients . This indicates that the infection source may differ based on geographic and behavioral differences [2] , and clinicians should consider the local cultural context ( e . g . , food habits ) when identifying the probable infection source in adult patients with OT . Increasing public awareness about toxocariasis is the first step in reducing human exposure to Toxocara . Proper handwashing , limiting children's outdoor activity in sandboxes , appropriate disposal of dog and cat feces , and controlling infections in dogs and cats through deworming are the recommended prevention strategies , especially in children [29] . We suggest that OT may be prevented in many adults by avoiding uncooked meat ingestion , especially raw cow liver . Educating the targeted population ( middle-aged , East Asian men ) as well as clinicians regarding this disease may be effective in preventing the disease [30] . The Japanese government recently banned restaurants from serving raw cow liver . Although intended to prevent infection with a virulent strain of Escherichia coli , the action may also reduce the incidence of toxocariasis , especially if applied in East Asia . Some limitations of our study should be considered . First , although our study included a relatively large cohort of patients compared to previous studies , it is a retrospective study , with intrinsic drawbacks that may introduce bias . The patients presented at >20 years of age , but it does not necessarily mean adult presentation . Additionally , our results on adult infection source may not necessarily extrapolate to the rest of the world as regional food habits vary widely . The serum IgE level was measured only in 55% of the included patients since we started our investigation of immunologic indicators of OT in October 2010 . Although the decision regarding whether a patient with OT would be tested for IgE or not was not made by the clinician , selection bias may not be neglected . Additionally , comparability issues exist in the comparison of treatment outcomes . The method of treatment was determined based on clinical presentation . In OT patients with active intraocular inflammation but without eosinophilia or elevated IgE levels , the current standard treatment for OT , i . e . , corticosteroid , was administered . However , in the OT patients with eosinophilia or elevated IgE levels , anthelmintic treatment combined with corticosteroid was preferred due to the possibility of VLM . In this setting , it may not be feasible to compare treatment outcomes between combination therapy and steroid monotherapy groups; similarly , in patients without active inflammation , it may not be feasible to compare albendazole monotherapy and no treatment groups . However , as there were no differences in the baseline ocular characteristics between the two treatment groups , the ocular treatment outcomes could be compared despite the limitation . Further prospective randomized trials are required to compare treatment outcomes between the groups . Despite these limitations , our analyses may be valuable since an optimal treatment for OT has not yet been determined and the role of anthelmintics in OT has not been evaluated . Another limitation of our study is related to the method of interview used for the investigation of probable infection sources of OT . We tried to avoid introducing biases in the interview procedures between patients and controls by using a standardized interview protocol . However , clinical diagnosis of the participants was not completely blind to the interviewer , possibly leading to bias . In conclusion , the present study showed that intraocular granuloma migration is an important and distinguishing clinical feature of OT . Among the serologic markers , total IgE , in addition to anti-toxocara IgG antibody ( ELISA ) , may be useful for the diagnosis of OT . In cases showing severe inflammation , combined corticosteroid and albendazole therapy may reduce inflammation and recurrence of OT . The risk of OT in adults may be reduced by avoiding ingestion of uncooked meat , particularly cow liver; however , further studies are required to validate this suggestion .
|
Toxocariasis is one of America's most common neglected infections of poverty and a helminthiasis of global importance . Little is known about the epidemiologic , demographic , and clinical features of ocular toxocariasis ( OT ) in adult patients , and the treatment regimen for OT has not been standardized . We conducted a retrospective cohort study examining the clinical features , serologic markers , clinical course of granuloma , probable infection sources , and treatment outcome in 101 adult patients diagnosed clinically and serologically with OT . All the patients had unilateral involvement . Ninety-three ( 92 . 1% ) and 78 ( 77 . 2% ) of 101 adult patients had retinal granuloma and intraocular inflammation , respectively . In addition to retinal granuloma , retinal nerve fiber layer defect , epiretinal membrane , vitreous opacity , retinal detachment , macular edema , and macular hole were observed in the eyes with OT . Granuloma in OT can involve all retinal layers , and its intraocular migration was observed in 15 patients ( 16 . 1% ) . Among the 101 patients , 69 . 6% and 11 . 6% showed elevated immunoglobulin E levels and eosinophilia , respectively . We believe that OT may be associated with ingestion of uncooked meat , especially cow liver , in adult patients . Furthermore , we suggest that combined albendazole and corticosteroid therapy may reduce intraocular inflammation and recurrence .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"helminth",
"infections",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"toxocariasis",
"larva",
"migrans",
"neglected",
"tropical",
"diseases",
"tropical",
"diseases",
"parasitic",
"diseases"
] |
2014
|
Clinical Features and Course of Ocular Toxocariasis in Adults
|
All types of small RNAs in plants , piwi-interacting RNAs ( piRNAs ) in animals and a subset of siRNAs in Drosophila and C . elegans are subject to HEN1 mediated 3’ terminal 2’-O-methylation . This modification plays a pivotal role in protecting small RNAs from 3’ uridylation , trimming and degradation . In Arabidopsis , HESO1 is a major enzyme that uridylates small RNAs to trigger their degradation . However , U-tail is still present in null hen1 heso1 mutants , suggesting the existence of ( an ) enzymatic activities redundant with HESO1 . Here , we report that UTP: RNA uridylyltransferase ( URT1 ) is a functional paralog of HESO1 . URT1 interacts with AGO1 and plays a predominant role in miRNA uridylation when HESO1 is absent . Uridylation of miRNA is globally abolished in a hen1 heso1 urt1 triple mutant , accompanied by an extensive increase of 3’-to-5’ trimming . In contrast , disruption of URT1 appears not to affect the heterochromatic siRNA uridylation . This indicates the involvement of additional nucleotidyl transferases in the siRNA pathway . Analysis of miRNA tailings in the hen1 heso1 urt1 triple mutant also reveals the existence of previously unknown enzymatic activities that can add non-uridine nucleotides . Importantly , we show HESO1 may also act redundantly with URT1 in miRNA uridylation when HEN1 is fully competent . Taken together , our data not only reveal a synergistic action of HESO1 and URT1 in the 3’ uridylation of miRNAs , but also independent activities of multiple terminal nucleotidyl transferases in the 3’ tailing of small RNAs and an antagonistic relationship between uridylation and trimming . Our results may provide further insight into the mechanisms of small RNA 3’ end modification and stability control .
MicroRNAs ( miRNAs ) , a class of small non-coding RNAs with 20–24 nt in size , are master regulators of gene expression at post-transcriptional levels in both plants and animals [1 , 2] . They impact various biological processes such as development , metabolism and response to different biotic and abiotic stresses [1] . In Arabidopsis , miRNAs are typically derived from their precursor RNAs called primary miRNA transcripts ( pri-miRNAs ) through a sequential cleavage , from either loop proximal or loop distal , by DICER-LIKE 1 ( DCL1 ) and its associated proteins [2–5] . After generation , the 3’ terminal riboses of miRNA/miRNA* duplexes are 2’-O-methylated by the RNA methyltransferase HUA ENHANCER1 ( HEN1 ) [6] . One strand of the mature , methylated miRNA/miRNA* duplexes is selectively incorporated into their effector protein , mainly ARGONAUTE1 ( AGO1 ) , to form a silencing complex [7] . In the hen1 mutants , miRNAs accumulate at lower levels and become heterogeneous in size , due to different levels of trimming and non-templated nucleotides addition ( tailing , also commonly referred to as uridylation because uridine is preferentially added ) at their 3’ ends [6 , 8] . Similar process also occurs to plant small interfering RNAs ( siRNAs ) , animal PIWI-interacting RNAs ( piRNAs ) and a subset of animal siRNAs [9–16] . These results reveal an evolutionary conserved role of 3’ end methylation in protecting small RNAs from 3’ end tailing , trimming and degradation . HEN1 SUPPRESSOR1 ( HESO1 ) is a major enzyme responsible for the small RNA uridylation in Arabidopsis [17 , 18] . Loss-of-function mutations in HESO1 result in reduced miRNA tailing length , increased miRNA abundance and partially restored morphological phenotypes of hen1 . In contrast , over-expression of HESO1 in hen1 causes reduced miRNA accumulation and more severe morphological phenotypes . Collectively , these results demonstrate that uridylation triggers miRNA degradation [17 , 18] . Nevertheless , A substantial amount of miRNA uridylation is present in the null hen1 heso1 mutants , suggesting the existence of additional small RNA uridylyltransferase ( s ) [17 , 18] . Besides HESO1 , the Arabidopsis genome encodes nine additional terminal nucleotidyl transferases ( TNTases ) [17 , 18] . However , their involvement , if any , in miRNA uridylation is probably masked by HESO1 because none of their single mutants has any visible effect in terms of the restoration of hen1 morphological phenotypes [18] . To identify additional miRNA uridylyltransferase ( s ) , we screened for mutants with further recovered fertility in the hen1-2 heso1-2 background . In this study , we show that a Proline-to-Leucine substitution at amino acid 618 ( P618L ) in URT1 increases the silique length and enhances the miRNA function of hen1-2 heso1-2 . URT1 was previously identified as an enzyme that uridylates some mRNAs with short poly ( A ) tails and exhibits robust U-tailing activity in vitro [19] . We find that URT1 interacts with AGO1 and is responsible for miRNA , but not heterochromatic siRNA , uridylation in hen1-2 heso1-2 . miRNA uridylation is globally abolished in the hen1-2 heso1-2 urt1-3 mutant , accompanied by an extensive increase of trimming . These results demonstrate that HESO1 and URT1 act synergetically in miRNA uridylation and that trimming and tailing antagonize each other for the occupancy of miRNA 3’ ends . Moreover , a close investigation of the tailing status in hen1-2 heso1-2 urt1-3 also reveals the addition of non-uridine nucleotides by yet uncharacterized TNTase ( s ) , suggesting the complexity of miRNA tailing . Finally , we show HESO1 and URT1 may redundantly uridylate some miRNAs in the HEN1 competent background .
To identify novel components in miRNA stability control , we performed another genetic screen in hen1-2 heso1-2 ( note that hen1-2 is a weak allele and heso1-2 is a null allele ) . Fertility , as reflected by the silique length , has been proved to be an effective indicator of miRNA activity in this system [17 , 20] . A mutant ( m37-6 ) with markedly longer siliques than hen1-2 heso1-2 was isolated from the M2 population of ethylmethanesulfonate ( EMS ) mutagenized hen1-2 heso1-2 . The average silique length of m37-6 was ~1 . 0 cm , which was about 1 . 3 and 2 . 9 fold of those of hen1-2 heso1-2 and hen1-2 , respectively ( Fig 1A and 1B ) . Backcross analyses revealed that the increased silique length phenotype was caused by a single recessive mutation . The mutation was roughly mapped to the marker nga168 on Chr . II , where HESO1 and two other TNTases ( At2g40520 and At2g45620/URT1 ) were nearby this marker . Sequencing of these candidate genes revealed a single C-to-T nucleotide substitution , at the position +1853 relative to the translation start site ( ATG ) of URT1 ( the mutation was hereafter referred to as urt1-3; urt1-1 ( Salk_087647C ) and urt1-2 ( WISCDSLOXHS208_08D ) were previously characterized to be defective in the uridylation of mRNAs with shortened poly ( A ) tails . [19] ) , which resulted in a Proline-to-Leucine change at amino acid 618 ( P618L ) ( Fig 1C and 1D ) . P618 is located in an uncharacterized region that links the PAP domain and the PAP-associated domain . Blast analysis suggested that P618 is rather conserved , especially among those well-characterized TNTases ( Fig 1D ) . Introduction of a genomic fragment of URT1 fused with a GFP ( pURT1-URT1-GFP ) ( S1 Fig ) at its C-terminus into hen1-2 heso1-2 urt1-3 restored its silique length to the hen1-2 heso1-2 level , suggesting that the urt1-3 mutation is responsible for the increased fertility of hen1-2heso1-2 . Introgression of the urt1-3 mutation into hen1-1 heso1-2 also increased its fertility ( S2 Fig ) . Since hen1-1 is a null allele of hen1 , this observation suggested that the increase of hen1 heso1-2 fertility by the urt1-3 mutation doesn’t rely on the residual HEN1 activity . Morphological defects in hen1 are mainly due to a reduction of miRNA activities and the null heso1-2 mutation only partially restores the miRNA activities in hen1 [17 , 18] . To determine whether the miRNA activities in hen1-1 heso1-2 were further recovered by urt1-3 , we compared the expression levels of five miRNA targets in Ler ( wild type , Wt ) , hen1-1 , hen1-1 heso1-2 and hen1-1 heso1-2 urt1-3 by quantitative real-time RT-PCR ( qPCR ) . Confirming our previous findings [17] , the expression levels of all tested targets were increased in hen1-1 as compared to those in Wt and their levels were partially restored in hen1-1 heso1-2 ( Fig 1E ) . Notably , the expression of these transcripts in hen1-1 heso1-2 urt1-3 was further reduced , to the levels comparable to those in Wt ( Fig 1E ) . Therefore , we conclude that the urt1-3 mutation further enhances the miRNA function of hen1-1 heso1-2 . We next examined the expression patterns of several miRNAs in Wt , hen1-2 , hen1-2 heso1-2 and hen1-2 heso1-2 urt1-3 by the small RNA northern blot assay . We found that miRNA species with sizes longer than their annotated ones were barely detected in hen1-2 heso1-2 urt1-3 for all the tested miRNAs ( Fig 2 ) . Instead , miRNAs in hen1-2 heso1-2 urt1-3 became severely shortened in size as compared with those in hen1-2 and hen1-2 heso1-2 ( Fig 2 ) . The miRNA expression patterns of the genomic complementation line ( hen1-2 heso1-2 urt1-3 + pURT1-URT1-GFP ) resembled those of hen1-2 heso1-2 , demonstrating that the urt1-3 mutation causes the changes of miRNA profiles in hen1-2 heso1-2 ( Fig 2 ) . Since unmethylated miRNAs can be frequently trimmed before uridylated [21] , the size information given by the small RNA northern blot assay may only reflect the overall combination effect of trimming and tailing . In order to analyze the degree of uridylation reduction and trimming enhancement in detail , small RNA libraries with two biological replicates for each genotype were prepared from Wt , hen1-2 , hen1-2 heso1-2 and hen1-2 heso1-2 urt1-3 and subject to the Illumina deep sequencing . Small RNA 3’ end modification analysis was performed according to a previous report with minor modifications ( for details , see Materials and Methods ) [18] . Based on the algorithm , small RNA reads were split into a 5’ genome matched component ( 5GMC ) part and a 3’ tailed part , which reflects the degree of trimming and tailing , respectively . Only reads with 5GMC longer than 12nt were included in our analysis . We first compared the overall length distribution of miRNAs among these four genotypes . As expected , about 84% of the miRNAs were 21nt in size in Wt . The size of miRNAs became heterogeneous in hen1-2 , ranging from 18nt to 25nt ( reads cutoff = 0 . 5% ) with peaks at 20nt and 21nt ( Fig 3A and S3A Fig ) . An obvious shift of this size range towards shorter ones was observed in hen1-2 heso1-2 and became more prominent in hen1-2 heso1-2 urt1-3 ( Fig 3A and S3A Fig ) , which is consistent with our northern blot results ( Fig 2 ) . Notably , only less than 2 . 5% of the miRNAs were 22nt or longer in hen1-2 heso1-2 urt1-3 , compared with ~9 . 1% of those in hen1-2 heso1-2 ( p<0 . 0001 , F-test ) ( Fig 3A and S3A Fig ) . Moreover , the 5GMC pattern of hen1-2 heso1-2 urt1-3 resembled the miRNA distribution pattern ( Fig 3A and 3B and S3A and S3B Fig ) , indicating a drastic loss of tailing activity by the loss of functions in HESO1 and URT1 . The distribution of 5GMC also revealed that a reduction of uridylation was accompanied by an increase of trimming ( Fig 3B and S3B Fig ) . Indeed , both the levels and extents ( more extensive truncation ) of trimming were gradually increased in hen1-2 heso1-2 and hen1-2 heso1-2 urt1-3 when we quantified the degrees of tailing and trimming separately ( Fig 3C and S3C Fig ) . Notably , a small peak of 5GMC at 17nt was observed in hen1-2 ( Fig 3B and S3B Fig ) , indicating a proportion of miRNAs favor to be trimmed to 17nt before they are further tailed . However , such a peak was not observed in either hen1-2 heso1-2 or hen1-2 heso1-2 urt1-3 , implicating a dynamic correlation between tailing and trimming . Analyses of individual miRNAs further support our northern blot results and global miRNA profiles ( Fig 2 and Fig 3 ) . For most of the miRNAs analyzed , a gradual loss of uridylation activity was always accompanied by an increase of trimmed species and/or normal sized ones , albeit the degree of which varied among different miRNAs ( Fig 4 and S4 Fig ) . Notably , long-tailed species could be readily detected in hen1-2heso1-2 for some miRNAs ( e . g . miR163 and miR164a ) and the tails were almost abolished in hen1-2heso1-2urt1-3 ( Fig 4 and S4 Fig ) . More intriguingly , we found that several miRNAs ( e . g . miR390 and miR398b ) were resistant to both tailing and trimming modifications ( Fig 4 and S4 Fig ) . The reason for this resistance is currently unknown . Consistent with the overall length distribution ( Fig 3A and S3A Fig ) , 3’ tailing was almost abolished in hen1-2 heso1-2 urt1-3 ( Fig 3C , S3C Fig and S1 Table ) . Intriguingly , we noticed a remarkably reduced preference for uridine ( 39 . 1% ) when we calculated the nucleotide composition of the residual tails in hen1-2 heso1-2 urt1-3 . As controls , the percentage of uridine addition in tails of Wt , hen1-2 and hen1-2 heso1-2 was 76 . 7% , 88 . 4% and 90 . 3% , respectively ( Fig 3C , S3C Fig and S1 Table ) . Further analysis revealed that the raw counts of the other three types of nucleotides additions ( i . e . adenosine , cytidine and guanosine ) were comparable among hen1-2 , hen1-2 heso1-2 and hen1-2 heso1-2 urt1-3 , although their relative proportions were increased in the triple mutant , due to an almost complete loss of uridylation ( S1 Table ) . Based on our definition of tailing , a tiny proportion of tailed-only miRNAs might be alternatively introduced by the misprocessing , alternative processing and/or processing of pre-tailed pre-miRNAs [22 , 23] . We thus investigated miRNA species with 1nt trimmed and 1nt tailed , ensuring that the tails we analyzed were added after processing . We found that uridylation , but not other types of nucleotides addition were affected upon loss of function in HESO1 and URT1 ( S2 Table ) . Taken together , these data unambiguously suggested that HESO1 and URT1 predominantly catalyze the uridylation of miRNAs and other TNTase ( s ) must exist to catalyze the non-uridine nucleotides addition with low activities . We also tested whether URT1 , as like HESO1 , plays a role in siRNA uridylation . A reduction in 3’ uridylation and/or an increase in 3’ end trimming were observed in several ta-siRNAs , suggesting that these ta-siRNAs are also recognized by URT1 ( S4 Fig ) . Unexpectedly , the 3’ end modification pattern was largely unaffected in all three tested heterochromatic siRNAs ( hc-siRNAs ) in hen1-2 heso1-2 urt1-3 as compared with those in hen1-2 heso1-2 ( Fig 4 and S4 Fig ) , which were shown to be substrates of HESO1 [18] . The enzyme ( s ) responsible for siRNA uridylation in hen1heso1-2 awaits future characterization . In addition , our results may also suggest a differentiation of terminal uridyl transferases in substrate recognition . A small proportion of miRNAs in HEN1 competent background are unmethylated and these miRNAs are extensively uridylated and/or trimmed ( Fig 5 and S5B Fig ) [8 , 18 , 21] . We thus examined whether HESO1 and URT1 are catalyzers of miRNA uridylation in HEN1+ plants . To test this , hen1-2 heso1-2 urt1-3 was crossed to heso1-2 [17] and heso1-2 urt1-3 was obtained from the F2 progeny by genotyping . Like heso1-2 , heso1-2 urt1-3 didn't show any morphological defects as compared with Wt under normal growth condition ( S5A Fig ) [17 , 24] . Small RNA deep sequencing reads ending with annotated miRNA sites were removed and the rest of miRNA variants were renormalized to 100% and subject to analysis as in Fig 4 . We found the patterns for the loss of uridylation and the gain of trimming in Ler , heso1-2 and heso1-2 urt1-3 resembled those observed in hen1-2 , hen1-2 heso11-2 and hen1-2 heso1-2 urt1-3 ( Fig 5 and S5B Fig ) , suggesting that HESO1 may act redundantly with URT1 in miRNA uridylation when HEN1 is fully competent . Although the uridylyltransferase activity of URT1 on single stranded RNAs in vitro has been well documented [19 , 25] , the effect of P618L substitution on its activity has not been determined and thus , how URT1P618L increases the fertility in hen1 heso1-2 and alters the miRNA profiles is unclear . To test this , we expressed both GST-URT1 and GST-URT1P618L in E . coli BL21 and purified the recombinant proteins by the Glutathione Sepharose 4B beads ( Fig 6A ) . Consistent with previous reports [19 , 25] , GST-URT1 but not GST alone possessed a robust activity in incorporating UTP to the 3’ end of its RNA substrate ( Fig 6B ) . In addition , GST-URT1 was also able to add a few adenosines ( A ) or cytidines ( C ) but only one to two guanosines ( G ) to its RNA substrate ( Fig 6B ) . In contrast , only residual uridylyltransferase activities were detected for GST-URT1P618L ( Fig 6B ) , indicating the P618L single amino acid substitution severely impairs the URT1 activity without affecting its UTP preference . We next examined the localization of URT1 by transient expression in Nicotiana Benthamiana . URT1-GFP was localized to the cytoplasmic foci , a pattern similar to those observed in HESO1 and AGO1 ( Fig 6C ) [24 , 26 , 27] . Indeed , we found that URT1-GFP colocalized with HESO1-RFP at the cytoplasmic foci ( Fig 6D ) , supporting a role of URT1 in miRNA uridylation . While HESO1 was also targeted to the nucleus ( 100% , n = 43 ) ( Fig 6C ) [17] , URT1-GFP signals could only be detected around the nuclear envelope but not in the nucleoplasm ( 100% , n = 53 ) ( Fig 6C ) . Interestingly , URT1P618L-GFP signals were readily detected in the nucleoplasm ( 100% , n = 50 ) and the number of cytoplasmic foci was greatly reduced ( n = 50 ) ( Fig 6C ) , suggesting a change of localization of URT1 by the P618L single amino acid substitution . Western blot analysis suggested that the change of localization of URT1P618L-GFP might be not caused by passive diffusion of truncated URT1P618L-GFP proteins ( S6C Fig ) . Bioinformatics prediction also failed to detect a loss of nuclear export signal ( NES ) /a gain of nuclear localization signal ( NLS ) by P618L . Our recent study suggests that HESO1 uridylates AGO1-associated miRNAs through its interaction with AGO1 [24] . We thus asked whether URT1 also interacts with AGO1 and if so , whether URT1P618L affects its interaction with AGO1 . To test this , we transiently co-expressed URT1-GFP/ URT1P618L-GFP and 10xMYC-AGO1 in N . benthamiana . We were able to detect URT1-GFP in 10xMYC-AGO1 immunoprecipitates and vice versa , suggesting an interaction between URT1 and AGO1 ( Fig 6E and 6F and S6 Fig ) . In contrast , GFP and 10xMYC-AGO1 could not co-immunoprecipitate with each other ( Fig 6E and 6F and S6 Fig ) . Moreover , 10xMYC-DCL3 ( aa1-280 ) failed to co-immunoprecipitate with URT1-GFP . These data suggest a direct interaction between URT1 and AGO1 ( S6D Fig ) . The interaction between URT1-GFP and 10xMYC-AGO1 was not affected by the treatment of RNAse A ( 50 μg/ml ) , suggesting their association may be RNA-independent ( S6D Fig ) . We also detect a positive , albeit weaker , interaction between URT1P618L-GFP and 10xMYC-AGO1 ( Fig 6E and 6F and S6 Fig ) . However , we could not rule out the possibility that the diminution of protein-protein interaction was caused by the change of sub-cellular localization .
We have previously shown that HESO1 is a major enzyme responsible for the uridylation of all types of small silencing RNAs in hen1 in Arabidopsis and also reveals the redundancy of small RNA uridylation [17 , 18] . Besides HESO1 , the Arabidopsis genome encodes additional nine TNTases [17 , 18] . However , despite the conservation of PAP and/or PAP-associated domains , other sequences among these TNTases are highly divergent , which hinders the prediction of their exact function . Moreover , individually knock out of these nine TNTases in hen1-8 fails to rescue the morphological phenotype of hen1-8 [18] . In this study , we successfully identified URT1 as a functional homologue of HESO1 in triggering miRNA uridylation . In the accompanied paper , Tu et al also show that lack of URT1 , but not other TNTases , results in an only minor reduction of uridylation of several miRNAs in hen1-8 [28] . Taken together , these observations suggest that HESO1 and URT1 can compensate for each other to some extent and function synergistically . A single point mutation in URT1 results in the substitution of Proline by Leucine ( P618L ) in a yet uncharacterized region ( Linker region ) that bridges the poly ( A ) polymerase ( PAP ) domain and the PAP-associated domain ( Fig 1C and 1D ) . Although P618L is rather conserved among several well-characterized TNTases ( Fig 1D ) , sequence analysis reveals that only HESO1 and URT1 possess a complete “PAP-Linker-PAPA” structure among all ten TNTases in Arabidopsis ( S3 Table ) , suggesting the importance of this structure for their proper function . As P618L diminishes URT1 activity without affecting its UTP preference in vitro ( Fig 6B ) , it is likely that urt1-3 is not a null allele . In agreement with this , we found that miRNA uridylation was not completely abolished in the hen1-2 heso1-2 urt1-3 triple mutant ( Fig 3C , S3C Fig and S1 Table ) . However , we cannot exclude the possibility that the residual miRNA uridylation is catalyzed by other uridydyl transferase ( s ) and urt1-3 is null in vivo . URT1 co-localizes with HESO1 and both enzymes interact with AGO1 , consistent with their roles in the uridylation of AGO1-associated miRNAs [24 , 28] . However , unlike HESO1 , URT1 is barely detectable in the nucleoplasm and may not affect heterochromatic siRNA uridylation in the hen1heso1 background ( Fig 6C , Fig 4 and S4 Fig ) . Moreover , URT1 appears to be less active than HESO1 in vivo since loss-of-function in URT1 alone has limited effects on the miRNA profile and growth of hen1 [17 , 18 , 28] . However , we found that HESO1 and URT1 share very similar catalytic properties in vitro in terms of the UTP preference ( Fig 6B ) [17] . It shall be noted that an accurate comparison of their activities will require more strict conditions such as the removal of fusion tags . Moreover , our results also suggest that URT1 may be responsible for adding long tails to the 3’ ends of at least some miRNAs ( e . g . miR163 and miR164a ) in vivo . Alternatively , URT1 may trigger mono- or di-uridylation , which is required for additional tailing by yet uncharacterized TNTase ( s ) . Future studies on their expression regulation and catalytic dynamics will help to understand their differences in the functional output . Consistent with previous notion [17] , the reduced levels of miRNA uridylation are accompanied by the increased levels and extents ( more extensive truncation ) of trimming , demonstrating that uridylation antagonizes trimming for the occupancy of 3’ ends of miRNAs ( Fig 3B and S3B Fig ) . More interestingly , other types of nucleotides addition ( C , A , G ) become evident when uridylation was drastically reduced in the hen1-2 heso1-2 urt1-3 triple mutant ( Fig 3C , S3C Fig , and S1 and S2 Table ) , suggesting the involvement of additional TNTases in the miRNA tailing process . Indeed , these non-uridine nucleotides additions are largely unaffected by the loss of functions in HESO1 and URT1 , albeit they occur at very low frequencies among different genotypes ( S1 and S2 Table ) . Thus , the landscape of TNTases in small RNA 3’ end tailing is much more complicated than previously appreciated , although the biological relevance of non-uridine tailing modifications remains obscure . What is the function of HESO1 and URT1 in Wt background ? Although no visible morphological defects are detected in either heso1-2 or heso1-2 urt1-3 under normal growth condition , we show that uridylation of unmethylated miRNAs is greatly reduced in heso1-2 and is almost abolished in heso1-2 urt1-3 ( Fig 5 and S5 Fig ) . Since uridylation triggers degradation of unmethylated miRNAs , it is possible that HESO1 and URT1 may play a role in the fine adjustment of miRNA abundance in Wt plants . In fact , it has been hypothesized that uridylation may act coordinately with SMALL RNA DEGRADING NUCLEASEs ( SDNs ) , a class of 3’-to-5’ exonucleases that can remove methylated nucleotides , in the active turnover of miRNAs [29 , 30] . In addition , HESO1 and URT1 are likely involved in the clearance of 5’ cleavage products of miRNA guided slicing of target mRNAs . In fact , HESO1 together with one or more other TUTases can uridylates 5’ cleavage products and promotes their degradation [24] . Based on the action mode of URT1 in miRNA uridylation , it is plausible to speculate that URT1 is one of other enzyme ( s ) that catalyzes the uridylation of 5’ cleavage products . Moreover , URT1 has been shown to uridylate some mRNAs with short poly ( A ) tails [19] , suggesting HESO1 and URT1 may have a broad spectrum of substrate RNAs in vivo . While the biological significance of the clearance or tentatively stabilization of their substrate RNAs through uridylation modification is unknown , one can simply envision that heso1-2 and heso1-2 urt1-3 plants are under “sub-healthy” conditions . Examining the effect of HESO1 and URT1 under various stress conditions will provide new clues for their functions .
All the Arabidopsis ( Arabidopsis thaliana ) strains used in this study are in the Landsberg erecta ( Ler ) background unless otherwise indicated . hen1-1 , hen1-2 , heso1-2 , hen1-1 heso1-2 and hen1-2 heso1-2 are previously described [17 , 18 , 31] . Ethyl methanesulfonate ( EMS ) mutagenesis in hen1-2heso1-2 was conducted according to the Arabidopsis manual [32] . F2 population obtained from a cross between m37-6 and hen1-8 heso1-1 ( in the Columbia-0 background ) [18] was used for the bulk segregation analysis . To obtain the hen1-1 heso1-2 urt1-3 and heso1-2 urt1-3 mutants , hen1-2 heso1-2 urt1-3 was crossed with hen1-1 heso1-2 and heso1-2 , respectively . Individuals containing respective genotypes were identified in the F2 population . hen1-1 and hen1-2 were genotyped according to [31] . urt1-3 was identified by digestion of the URT1dCAPSF/R PCR product with DdeI , which cut urt1-3 but not Wt . For the genomic complementation assay , an approximately 4 . 2-kb DNA fragment containing the URT1 coding region as well as 1 . 4 kb promoter region was amplified from Ler genomic DNA using PrimeSTAR ( Takara , R045A ) with primers URT1gGWF/URT1gGWR and sub-cloned into an Entry vector pENTR-D-topo to produce pENTR-URT1g . The entry clone was transferred into the destination vector pMDC204 [33] by LR recombination to produce pURT1:URT1-GFP . The resulting plasmid was transformed into hen1-2 heso1-2 urt1-3 via floral dipping [34] and hygromycin resistance was used for the transgenic plants selection . qPCR analysis and small RNA northern blotting were performed as previously described [17 , 35] . Total RNA extracted from inflorescence tissue of indicated genotypes was used for small RNA library construction . Small RNA libraries were prepared as indexed ( i . e . individually bar-coded ) and sequenced at 51bp single-end on an Illumina Hiseq2000 following the standard protocol . The small RNA sequence data were deposited to Gene Expression Omnibus ( GEO ) under the accession number GSE60826 . Adaptor sequence was removed using Perl scripts . Analysis of small RNA 3’ end modification was performed according to [18] . The reference miRNA sequences were obtained from miRbase v17 . 0 ( http://www . mirbase . org/ ) . For the determination of 5GMC , the 5’ end of miRNA was locked ( i . e . miRNAs with 5’ end different from the reference were removed from analysis ) and no mismatch was permitted during the alignment . Nucleotides between the annotated 3’ end and 5GMC were considered as trimmed nucleotides , while all nucleotides downstream of 5GMC , regardless of templated or not , were considered as tailed . URT1 CDS was amplified by PCR with primers GST-URT1F/GST-URT1R and cloned into pGEX-4T-1 to generate a Glutathione-S-transferase ( GST ) tagged plasmid ( GST-URT1 ) . The in vitro enzymatic assay was conducted as previously described [17] . pH7-WGY-AGO1 ( pAGO1:YFP-AGO1 ) [26] , 35S:HESO1-YFP [17]and 35S:HESO1-RFP [24] were previously described . pURT1:URT1P618L-GFP was generated the same wasy as pURT1:URT1-GFP except that a hen1-2heso1-2urt1-3 genomic DNA was used as a template . Respective constructs were expressed in N . benthamiana leaves for 40~48hr and subjected to imaging using a Nikon A1 confocal mounted on an Eclipse 90i Nikon compound microscope . The full-length URT1 coding sequence ( CDS ) was amplified from a Ler cDNA sample by PCR with primers URT1cGWF/URT1cGWR and sub-cloned into an Entry vector pENTR-D-topo to produce pENTR-URT1c . The entry clone was transferred into destination vectors pMDC83 [33] by LR recombination to produce 35S-URT1-GFP . 35S-URT1P618L-GFP was constructed the same way from a hen1-2heso1-2urt1-3 cDNA sample . URT1-GFP/URT1P618L-GFP and 10xMYC tagged AGO1 ( in pGWB521 backbone ) [24 , 36] were co-expressed in the presence of P19 in N . benthamiana leaves for 72hr and harvested . Protein extraction and Co-immunoprecipitation analysis were performed as previously described [35] . Anti-GFP ( Clontech , 632459 ) antibody pre-coupled to the protein G-agarose beads ( Santa Cruz , sc-2002 ) and anti-c-MYC ( Sigma , A7470 ) agarose beads were used to immunoprecipitate the GFP tagged and 10xMYC tagged proteins , respectively . Anti-GFP ( Covance , MMS-118R ) and Anti-MYC ( Abgent , AM1007a ) were used for the western blot assay . Online softwares cNLS Mapper ( http://nls-mapper . iab . keio . ac . jp/cgi-bin/NLS_Mapper_form . cgi ) , NucPred ( http://www . sbc . su . se/~maccallr/nucpred/cgi-bin/single . cgi ) and NLStradamus ( http://www . moseslab . csb . utoronto . ca/NLStradamus/ ) were used to predict the NLS peptide and NetNES ( http://www . cbs . dtu . dk/services/NetNES/ ) was used to predict the NES peptide using URT1 and URT1P618L as bait sequences . All the primers , LNA probes and other RNA/DNA oligos used in this study are listed in S4 Table .
|
Small silencing RNAs are key regulators of gene expression in both plants and animals . HEN1-mediated 3’ terminal 2’-O-methylation plays a crucial role in small RNA stability control . In the absence of HEN1 , several types of small RNAs become frequently uridylated ( non-templated uridine addition ) and trimmed , a phenomenon that is conserved across species . However , the underlying molecular mechanism is barely understood . In this study , we have discovered UTP: RNA uridylyltransferase ( URT1 ) that acts synergistically with HESO1 in miRNA uridylation , in addition to its role in oligo-adenylated mRNA uridylation . Analyzing the miRNA profiles also reveals the existence of multiple terminal nucleotidyl transferases in the miRNA tailing process and an antagonistic action between uridylation and trimming . We believe this study will shed light on our understanding of how various terminal nucleotidyl transferases recognize their substrates and function coordinately .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Synergistic and Independent Actions of Multiple Terminal Nucleotidyl Transferases in the 3’ Tailing of Small RNAs in Arabidopsis
|
The Plasmodium vivax vaccine candidate Duffy Binding Protein ( DBP ) is a protein necessary for P . vivax invasion of reticulocytes . The polymorphic nature of DBP induces strain-specific immune responses that pose unique challenges for vaccine development . DEKnull is a synthetic DBP based antigen that has been engineered through mutation to enhance induction of blocking inhibitory antibodies . We determined the x-ray crystal structure of DEKnull to identify if any conformational changes had occurred upon mutation . Computational and experimental analyses assessed immunogenicity differences between DBP and DEKnull epitopes . Functional binding assays with monoclonal antibodies were used to interrogate the available epitopes in DEKnull . We demonstrate that DEKnull is structurally similar to the parental Sal1 DBP . The DEKnull mutations do not cause peptide backbone shifts within the polymorphic loop , or at either the DBP dimerization interface or DARC receptor binding pockets , two important structurally conserved protective epitope motifs . All B-cell epitopes , except for the mutated DEK motif , are conserved between DEKnull and DBP . The DEKnull protein retains binding to conformationally dependent inhibitory antibodies . DEKnull is an iterative improvement of DBP as a vaccine candidate . DEKnull has reduced immunogenicity to polymorphic regions responsible for strain-specific immunity while retaining conserved protein folds necessary for induction of strain-transcending blocking inhibitory antibodies .
Plasmodium vivax is a causative agent of malaria , inflicting significant morbidity and impeding economic growth in highly endemic areas [1 , 2] . Increasing evidence indicates the severity of disease , economic impact , and burden of P . vivax has been severely underestimated [1 , 2] . Among the proposed methods for disease control , vaccines are appealing for a multitude of reasons . Vaccines are cost-effective , efficient , and have been historically successful in combating infectious diseases especially in resource poor environments [3] . Individuals living in regions with P . vivax develop naturally acquired protective immunity and antibodies isolated from those naturally immune have anti-DBP inhibitory effects that correlate with results from in vitro functional assays [4–6] . Establishment of a successful host infection necessitates specific receptor-ligand interactions between host red blood cells and Plasmodium parasites [7] . For P . vivax , the critical interaction is that between the merozoite Duffy binding protein ( DBP ) and the Duffy antigen receptor for chemokines ( DARC ) on reticulocytes . DARC-negative individuals are resistant to clinical P . vivax infection , and naturally immune individuals can possess anti-DBP antibodies that inhibit the DBP-DARC interaction and prevent parasite growth [6 , 8–12] . Additionally , polyclonal antibodies elicited by recombinant DBP exhibit similar protective and inhibitive effects to naturally acquired antibodies [6 , 11 , 13 , 14] . Certain isolates of P . vivax have been reported to invade Duffy-negative cells [15] . However , sequencing of these isolates identified a gene encoding a DBP paralog suggesting the increased copy number and/or expression of DBP may enable invasion into Duffy-negative cells [16] . Together , this highlights the central importance of the DBP-DARC interaction in P . vivax infection and presents DBP as a crucial parasite protein that can be developed as a vaccine target . DBP is a member of the Duffy binding-like erythrocyte binding protein ( DBL-EBP ) family , and binds DARC through a conserved cysteine-rich DBL domain known as region II ( DBP-II ) [17–22] . DBP-II engages DARC through a multimeric assembly mechanism where two DBP-II domains initially bind one DARC to form a heterotrimer that rapidly recruits a second DARC to form a heterotetramer [23–26] . DBP-II amino acids F261-T266 , L270-K289 , and Q356-K367 form critical contacts with the DARC ectodomain during this process [23] . This receptor-induced ligand-dimerization model is conserved amongst other members of the DBL-EBP family and provides spatial orientation for DBL domains at the parasite-RBC membrane interface [24–30] . Residues that mediate multimeric assembly are important targets of protective immunity as the epitopes of naturally acquired anti-DBP-II antibodies that disrupt the DBP-DARC interaction localize to residues at the dimerization interface , DARC binding pockets , and the RBC proximal face of DBP-II [10] . However , clusters of highly polymorphic residues flank these protective epitopes , which is a pattern seen in pathogens undergoing selective pressure that results in an immune evasion where allelic variants can escape immunity elicited by a previous infection [10 , 21 , 26 , 31–37] . Therefore , polymorphic residues of DBP appear to have a high potential to be the basis of strain specific immune responses that misdirects immune responses away from conserved targets of broadly neutralizing protection . Although strain specific immunity can be protective these seemingly more immunogenic epitopes offer limited value because of the strain-limited nature of the immunity . Genetic analysis of DBP-II alleles reveal a high dN/dS ratio often seen when selection pressure drives allelic diversity as a mechanism for immune evasion [38–42] . In order to proceed with DBP as a P . vivax vaccine target , it is therefore critical to address the challenges presented by polymorphism and immune misdirection inherent in this allelic diversity . Immunization with DBP-II elicits weakly reactive and allele specific immune responses , a far cry from the end objective of inducing strain-transcending protection [38] . The poor protectivity appears to be due in part to polymorphic non-functional residues diverting the immune response away from the more conserved , less immunogenic , critical receptor binding residues [10 , 38 , 43–45] . Consistent with this view , the most polymorphic region , identified as the DEK epitope , is positioned immediately adjacent to the conserved DARC-binding groove ( Fig . 1 ) [10 , 23] . Antibodies to the DEK epitope can disrupt DBP function , but inhibition is strain limited . Therefore , we refer to DEK as a decoy epitope that distracts the immune response away for more conserved functional epitopes that could serve as basis of a broadly neutralizing protective immunity . To overcome this inherent deficiency of DBP as an immunogen , a novel synthetic DBP-II antigen termed DEKnull was engineered where the polymorphic residues that comprise the DEK epitope were mutated to amino acids not usually present ( Fig . 1 , S1 Fig ) [38] . These proof of principle studies demonstrated the feasibility of redirecting the immune response to conserved , critical residues by eliminating polymorphic epitopes with the goal to create a vaccine that induces a greater percentage of protective antibodies to more conserved , less immunogenic epitopes . Indeed , anti-DEKnull sera lost reactivity towards the polymorphic patch as predicted , but still retained the ability to generate inhibitory antibodies , including epitopes reactive to naturally-occurring immune antibodies of persons infected with P . vivax [38 , 46] . DEKnull also induced strong anamnestic responses that were protective and cross-reactive against a panel of different DBP-II alleles [5] . Furthermore , DEKnull produced a more consistent inhibitory profile across variants [46] . However , mutation can alter the three-dimensional structure of a protein that in turn would alter the available epitopes presented in a synthetic antigen . This study presents the structure of a synthetic Plasmodium antigen and its implications for the future of vaccine design in targeting malaria . We determined the structure of DEKnull to identify if any shifts in fold and secondary structure or sub-domain rearrangements had occurred , and whether these changes affect DEKnull’s potential as a vaccine surrogate for native alleles [26] . The effects of mutating the DEK polymorphic patch on conserved protective epitopes was identified by comparison with the pre-existing Sal1 structure [26] . We examined and compared the epitope profile of DEKnull to DBP-II using computational approaches as well as through interrogation with a panel of DBP monoclonal antibodies [47] . Together these studies inform future efforts to guide the rational design of the next iteration of a synthetic DBP-II antigen to improve its immunogenicity and ability to mount a thoroughly protective response .
DEKnull was obtained by oxidative refolding . Inclusion bodies expressed in E . coli were solublized in 6 M guanidinium hydrochloride and refolded via rapid dilution in 400 mM L-arginine , 50 mM Tris pH 8 . 0 , 10 mM EDTA , 0 . 1 mM PMSF , 2 mM reduced glutathione , and 0 . 2 mM oxidized glutathione . Refolded protein was captured on SP Sepharose Fast Flow resin ( GE Healthcare ) , eluted with 50 mM MES pH 6 . 0 , 700 mM NaCl , and dialyzed overnight in 50 mM MES pH 6 . 0 , 100 mM NaCl . The protein was subsequently purified by sequential size exclusion chromatography ( GF200 ) and ion exchange chromatography ( HiTrapS ) . Protein was finally buffer exchanged into 10 mM HEPES pH 7 . 4 , 100 mM NaCl with size exclusion chromatography . Sal1 DBP-II was purified similarly as DEKnull , but without overnight dialysis . DEKnull crystals were grown by hanging-drop vapor diffusion . First , 1 μL of protein solution at 3–9 mg/mL was mixed with 1 μL of reservoir containing 0 . 2 M di-sodium tartrate , 20% PEG 3350 to create needle clusters . Crystals were shattered and microseeded into a mix of 1 μL of protein solution at 4 mg/mL and 1 μL of reservoir containing 0 . 2 M lithium chloride , 20% PEG 3350 . Large needle rods of DEKnull grew within a week and were flash frozen in liquid nitrogen . Data was collected to a resolution of 2 . 1 Å at beamline 4 . 2 . 2 of the Advanced light Source , Lawrence Berkeley National Laboratory and processed with XDS [48] . The DEKnull structure was solved by molecular replacement in PHASER [49] using a single Sal1 DBP-II domain from 3RRC as a starting model . Manual rebuilding in COOT [50] and refinement in PHENIX led to a final refined model with final R-factor/R-free of 21 . 77%/25 . 88% with good geometry as reported by MOLPROBITY [50–52] . The MOLPROBITY score of 0 . 81 places this structure in the top 100th percentile of structures 1 . 85–2 . 35 Å . 98 . 22% of residues lie in favored , 1 . 78% of residues lie in additionally allowed , and 0% lie in disallowed regions of the Ramachandran plot . Atomic coordinates and structure factors have been deposited into the Protein Data Bank with accession code 4YFS . The ELISAs were performed as previously described [28] . Briefly , BSA , Sal1 DBP-II , and DEKnull were coated on the plate overnight at 4° . The plates were washed with PBS/Tween-20 and then blocked with 2% BSA in PBS/Tween-20 for one hour at room temperature . The plates were washed with PBS/Tween-20 and then incubated with anti-DBP antibodies ( 2C6 , 2D10 , 2H2 , 3C9 , 2F12 , 3D10 ) individually for one hour at room temperature . The plates were again washed with PBS/Tween-20 and then incubated with an anti-mouse secondary antibody conjugated to Alexafluro-488 for 30 minutes at room temperature . After a final wash step , the fluorescence was measured using a POLARstar Omega ( BMG Labtech ) plate reader .
We obtained the crystal structure of the DEKnull antigen to a resolution of 2 . 1 Å ( Table 1 ) . DEKnull maintains the overall fold and conserved disulfide bonding patterns of a DBL domain similar to that found in P . vivax DBP Sal1 , from which DEKnull is derived [23 , 26] . The DBL fold is a conserved structural feature in other important Plasmodium adhesion proteins , including the P . falciparum EBA-175 and EBA-140 , P . knowlesi α-DBP protein , and the NTS-DBL1α1 , DBL6ε , and DBL3x domains of PfEMP-1 ( Fig . 2A ) [26 , 27 , 53–58] . DEKnull also retains the characteristic three sub-domain architecture of DBL domains with critical intra-domain disulfide bonding patterns ( Fig . 2B ) . Sub-domain 1 ( S1 ) contains residues K215 to L253 with two disulfide bonds , C217-C246 and C230-C237 . Sub-domain 2 ( S2 ) contains residues H262 to E386 and has a single disulfide bond C300-C377 . Sub-domain 3 ( S3 ) contains residues P387 to S508 and has three disulfide bonds: C415-C432 , C427-C507 , and C436-C505 . All cysteines in DEKnull are involved in disulfide bonding and are structurally conserved with Sal1 DBP-II [26] . Alignment of DEKnull and Sal1 DBP-II structures shows minimal differences with an overall root-mean-square ( r . m . s . ) deviation of 0 . 435 Å ( Fig . 2C ) , indicating there is minimal differences overall between the native and engineered domains . S1 alignment has a r . m . s . deviation of 0 . 308 Å and is not significantly different ( Fig . 2D ) . S2 alignment has a r . m . s . deviation of 0 . 288 Å , and the only change is the region comprising K366 to I376 , which is now structured in DEKnull as compared to Sal1 DBP-II ( Fig . 2D ) . S3 alignment has a r . m . s . deviation of 0 . 310 Å and show shifts in loops G417 to D423 and K465 to T473 , changes that can be attributed to solvent exposed flexible loops ( Fig . 2D ) . Strikingly , the DEKAQQRRKQ polymorphic stretch within S2 overlaps well between DEKnull and Sal1 DBP-II . Alteration of these amino acids to ASTAATSRTS had no affect on the secondary structure nor do they shift peptide backbone Cαs ( Fig . 3A , 3B ) . The dimer interface and DARC binding residues play important roles in host-receptor binding [23 , 26] . These functional regions are recognized by naturally acquired antibodies that block the DBP-DARC interaction [10 , 23] . Any DBP-II based synthetic antigen must accurately replicate the three-dimensional conformation of these regions for antibody generation and epitope recognition . We therefore examined if the changes in DEKnull altered these important functional regions . The dimerization and DARC binding surfaces overlap well with the parental Sal1 DBP-II; there is no allosteric change to secondary structure or peptide backbone Cαs , retaining the conformational shape of protective targets ( Fig . 3C , 3D ) . Furthermore , Define Secondary Structure of Proteins ( DSSP ) analysis assigns identical secondary structure elements between that of Sal1 DBP-II and DEKnull [59 , 60] . Together , these structural data demonstrate that the DEKnull conformation is not significantly different from that of the naturally occurring allele , except for the polymorphic DEK epitope , and supports the development of DEKnull as a DBP vaccine . B-cell epitopes fall within two classes: linear and conformational . Linear epitopes are continuous stretches of amino acids in which the primary structure alone is responsible for immunogenicity and antibody recognition . Conformational epitopes can be continuous or discontinuous , but require a fold for immunogenicity and antibody binding . Ablation of the fold through the use of denaturants eliminates antibody recognition of conformational epitopes . While vaccines are able to induce either class , natively folded antigens tend to have a bias towards inducing conformational-dependent antibodies that are protective [61 , 62] . As a result , it is important to identify and characterize inhibitory and non-inhibitory epitopes on Sal1 DBP-II . Bioinformatic B-cell epitope prediction methods for conformational epitopes are a powerful tool that can aid in the rational design and analysis of vaccine targets . DiscoTope is a widely used web-based computational algorithm that focuses on identifying potential discontinuous conformational epitopes based on available crystal structures [63] . DiscoTope analysis of Sal1 DBP-II identifies several distinct epitopes with the strongest signal located at the DEKAQQRRKQ polymorphic patch that is altered within DEKnull ( Fig . 4A ) . The predicted residues are all solvent exposed and are spread across the entire surface of the protein , with no discernible predilection for certain sub-domains ( Fig . 4B ) . DEKnull is predicted to have similar patches of epitopes , but lacks the signal at the DEK location induced by the mutational changes ( Fig . 4A , 4C ) . Comparisons between the Sal1 DBP-II and DEKnull prediction results demonstrate only the DEKAQQRRKQ region is significantly different ( Fig . 4A ) . An important concern of removing decoy-epitopes through mutation is the possibility of introducing novel epitopes caused by the amino acid changes . DiscoTope analysis determines that no new epitopes specific to DEKnull are introduced further demonstrating that DEKnull is a suitable surrogate antigen from native alleles of DBP-II . The structural and computation approaches indicate that there are no signification changes to epitopes in DEKnull with the exception of the mutated DEKAQQRRKQ epitope . We sought to independently assess the DEKnull antigen retained recognizable epitopes by interrogation with a panel of conformationally dependent anti-Sal1 DBP-II antibodies [47] in ELISA assays . Two non-inhibitory and four inhibitory antibodies were probed; all six antibodies showed no difference in antigen recognition between that of Sal1 DBP-II and DEKnull ( Fig . 5 ) . This provides evidence that the DEKnull mutations have minimal effect on the overall structural fold of the protein , and are consistent with the antigenicity results seen in the DiscoTope B-cell epitope prediction ( Fig . 4 ) . It is interesting to note that two non-inhibitory antibodies , 3D10 and 2F12 , bound to both DBP-II Sal1 and DEKnull equally well ( Fig . 5 ) . This suggests that DEKnull still retains at least one other immunogenic region that may continue to function in immune evasion , necessitating further development of DEKnull as a vaccine candidate .
The central role of P . vivax DBP and the necessity of DARC recognition in parasite invasion during the asexual red blood stage makes it an ideal vaccine target [8] . Anti-DBP antibodies isolated from naturally immune individuals and those generated through immunization are able to prevent DBP-DARC interactions and inhibit parasite growth [6] . However , the inherent polymorphic nature of DBP poses challenges that must be overcome in order to maximize its effectiveness as a vaccine [39 , 40] . Polymorphic immunodominant epitopes divert the immune system away from weakly immunogenic protective epitopes that are conserved across alleles , resulting in strain-specific responses as opposed to strain-transcending protection [43 , 45] . This is seen not only with DBP , but is an inherent problem observed with other Plasmodium vaccine candidates wherein single allele vaccinations often provide strain-specific inhibition but are yet susceptible to alternate alleles [4 , 10 , 64–72] . Currently two parallel strategies exist to enhance DBP as a vaccine candidate and to bypass the issue of polymorphism—a multi-allele vaccine composed of variants found in endemic areas , and a modified vaccine that directs immune responses towards conserved epitopes in order to impart broad protection [46 , 66 , 73 , 74] . The synthetic antigen DEKnull is the brainchild of the latter , an antigen in which a dominant variant B-cell epitope is mutated from the parent Sal1 allele [38] . Vaccination studies with DEKnull demonstrate early proof-of-concept success in manipulating the immune system towards protective responses [5 , 46] . Further iterations in design are expected to improve immunogenicity , protectivity , and cross-reactivity [5 , 46] . Here , we present the first structure of DEKnull , a synthetic Plasmodium vaccine candidate . These results demonstrate that the DEKnull antigen has insignificant structural changes relative to the parent Sal1 structure [26] . There are virtually no differences in overall DBL fold , orientations of sub-domains 1–3 , disulfide bonding , or within the secondary structure and backbone of the mutated region itself ( Fig . 2 , Fig . 3 ) . The conservation of DBL fold in DEKnull is confirmed with immunological assays examining reactivity against a panel of conformational dependent α-DBP-II ( Sal1 ) antibodies [47] . Of the six antibodies tested , none had significant binding differences between Sal1 and DEKnull ( Fig . 5 ) . The structure of DEKnull additionally allowed us to perform state-of-the-art bioinformatic B-cell epitope analysis through the use of DiscoTope [63] . The prediction results are significant for several reasons . First , the strong signal of the DEK polymorphic patch on the DBP-II Sal1 allele supports that it is strongly immunogenic and can divert immune responses away from conserved protective epitopes . Second , the loss of DEK antigenicity in DEKnull compared to Sal1 further reflects a success in synthetic antigen design in achieving the desired manipulation of epitopes . Third , the DEK mutation did not confound the design of the synthetic antigen by introducing novel epitopes . And finally , the conservation of the remaining epitopes between Sal1 and DEKnull indicates that the mutation does not change the protein’s overall epitope profile suggesting protective epitopes have been retained . The results first and foremost reflect a success in the strategy of using a modified antigen to bypass DBP allele polymorphisms and poor-protectivity induced by strain-specific epitopes . This study demonstrate that antigen engineering to focus the immune response to conserved functional regions , such as the DARC binding residues and/or DBP dimer interface , is a viable and practical approach . The predicted dominant variant B-cell epitope was eliminated without affecting immunogenicity of the remaining epitopes . Furthermore , the results presented here build upon previous works to establish that protein engineering is a viable approach towards problematic multi-allelic vaccine targets and should guide future vaccine design in other pathogens [5 , 38] . It has been shown that preliminary immunogenicity studies with DEKnull elicited an immune response comparable to Sal1 DBP-II [5] . A next key step in evaluating DEKnull as a vaccine construct is to demonstrate that DEKnull is able to generate highly potent antibodies that are broadly protective across multiple strains . As a corollary , and one that is predicted in silico by DiscoTope results presented here ( Fig . 4 ) , DEKnull must also not generate DEKnull-specific antibodies that would be useless against natural alleles . The ELISA data presented show that DEKnull still possess non-inhibitory epitopes ( Mab 3D10 and Mab 2F12 , Fig . 5 ) . Characterizing these antibodies will give insight towards the design of future versions of DEKnull . A continual process of eliminating non-protective epitopes from this synthetic antigen will better focus immune responses towards protective targets . Future studies will examine further iterations of DEKnull to improve upon its overall immunogenicity , broad-spectrum inhibitory profile across different P . vivax DBP alleles , as well as to address the antigenicity of remaining non-protective epitopes .
|
Plasmodium vivax is an oft neglected causative agent of human malaria . It inflicts tremendous burdens on public health infrastructures and causes significant detrimental effects on socio-economic growth throughout the world . P . vivax Duffy Binding Protein ( DBP ) is a surface protein that the parasite uses to invade host red blood cells and is a leading vaccine candidate . The variable nature of DBP poses unique challenges in creating an all-encompassing generalized vaccine . One method to circumvent this problem is to synthetically engineer a single artificial protein antigen that has reduced variability while maintaining conserved protective motifs to elicit strain-transcending protection . This synthetic antigen is termed DEKnull . Here , we provide structural and biochemical evidence that DEKnull was successfully engineered to eliminate polymorphic epitopes while retaining the overall fold of the protein , including conserved conformational protective epitopes . Our work presents validation for an improved iteration of the DBP P . vivax vaccine candidate , and provides evidence that protein engineering is successful in countering DBP polymorphisms . In doing so , we also lay down the foundation that engineering synthetic antigens is a viable approach and should be considered in future vaccine designs for pathogens .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Structural Analysis of the Synthetic Duffy Binding Protein (DBP) Antigen DEKnull Relevant for Plasmodium vivax Malaria Vaccine Design
|
Osteoporosis is a major public health problem . It is mainly characterized by low bone mineral density ( BMD ) and/or low-trauma osteoporotic fractures ( OF ) , both of which have strong genetic determination . The specific genes influencing these phenotypic traits , however , are largely unknown . Using the Affymetrix 500K array set , we performed a case-control genome-wide association study ( GWAS ) in 700 elderly Chinese Han subjects ( 350 with hip OF and 350 healthy matched controls ) . A follow-up replication study was conducted to validate our major GWAS findings in an independent Chinese sample containing 390 cases with hip OF and 516 controls . We found that a SNP , rs13182402 within the ALDH7A1 gene on chromosome 5q31 , was strongly associated with OF with evidence combined GWAS and replication studies ( P = 2 . 08×10−9 , odds ratio = 2 . 25 ) . In order to explore the target risk factors and potential mechanism underlying hip OF risk , we further examined this candidate SNP's relevance to hip BMD both in Chinese and Caucasian populations involving 9 , 962 additional subjects . This SNP was confirmed as consistently associated with hip BMD even across ethnic boundaries , in both Chinese and Caucasians ( combined P = 6 . 39×10−6 ) , further attesting to its potential effect on osteoporosis . ALDH7A1 degrades and detoxifies acetaldehyde , which inhibits osteoblast proliferation and results in decreased bone formation . Our findings may provide new insights into the pathogenesis of osteoporosis .
Osteoporosis , characterized primarily by low bone mineral density ( BMD ) , is a major public health problem because it increases susceptibility to low-trauma osteoporotic fractures ( OF ) . Hip fractures , which are the most common and severe form of OF , are associated with high morbidity and mortality , as well as tremendous health care expenditures [1] . Due to an aging population , the annual incidence of hip fractures worldwide is predicted to be ∼6 . 27 million by the year 2050 , with an estimated cost of ∼$131 . 5 billion [1] . The ultimate goal of osteoporosis research is to reduce the incidence and prevalence of OF . Genetic factors play an important role in susceptibility to osteoporosis . Both BMD and OF have high genetic determinations [2] , [3] , [4] , [5] . BMD has been identified as the major risk factor for susceptibility to OF and is currently the predominant study phenotype for osteoporosis . Variations in BMD account for ∼50–70% of the variation in total bone strength [6] and risk of OF [7] . Additional risk factors , including those not readily quantifiable ( e . g . bone microstructure [8] and cartilage organization [9] ) , also contribute to the risk of OF . Most of genetic studies of osteoporosis have focused primarily on the surrogate phenotype BMD , whereas little effort has been expended on the study of OF per se as a focal phenotype or on the relevance of genes associated with BMD on OF [3] . The major obstacle to this approach has been assembling a homogeneous sample with a homogenously defined OF type . Genetic factors associated with variations in BMD and risk of OF overlap , to some extent , but are not all identical [2] . The ultimate goal of osteoporosis research is to reduce the incidence and prevalence of OF . Therefore , it is useful to conduct genetic studies of OF per se , in conjunction with other intermediate phenotypes ( e . g . BMD ) that influence the risk of OF . This approach can be used to identify quantifiable measures for early prevention and intervention before the adverse clinical outcome , OF , actually occurs . So far , several specific genes contributing to osteoporosis ( i . e . those impacting BMD or risk of OF ) have been identified , such as ESR1 with OF risk , COL1A1 and VDR with BMD and vertebral fracture risk , OPG and LRP5 with BMD [10] , [11] , [12] , [13] , [14] , [15] . However , the majority of genetic variants that influence osteoporosis remain unknown . With current high throughput SNP genotyping platforms and our knowledge about the distribution and correlation of SNPs in the human genome ( e . g . , haplotype structure ) , genome-wide association study ( GWAS ) has proven itself to be a feasible , powerful and effective approach for identifying novel genes associated with complex phenotypes . Four recent GWAS's [12] , [13] , [16] , [17] have identified several specific genes for osteoporosis . In the current investigation , based on significant heritability of ∼50% for OF [2] , [4] , we utilized a GWAS to identify genetic variants underlying susceptibility to osteoporosis that are directly relevant to the risk of OF . Using the Affymetrix 500K array set , we successfully genotyped a study population of 700 elderly Chinese Han subjects consisting of 350 cases with homogeneous hip OF and 350 healthy matched controls . A follow-up replication study was performed in an independent Chinese sample consisting of 390 cases with hip OF and 516 controls . For SNPs that were identified for OF , we further examined their relationships with hip BMD in two ethnic groups ( Chinese and Caucasians ) , involving additional 9 , 962 subjects , in order to determine whether the genetic basis for their contribution to the risk of OF might also be , at least partially , attributable to their effects on variation in BMD .
The study design included an initial exploratory stage in a Chinese Han sample of moderate size and follow-up replication and validation studies with much larger sample sizes in independent Chinese Han and Caucasian samples . Table 1 details the basic characteristics of the respective samples . In the GWAS discovery stage , a total of 281 , 533 SNPs passed our quality control criteria for GWAS analyses . A quantile-quantile ( QQ ) plot is presented in Figure 1 . The χ2 distributions for the association tests across the SNPs tested showed little evidence of overall systematic bias ( genomic inflation factor λ = 1 . 02 ) . The highest χ2 was consistent with the presence of true association . We further performed the principal component analysis implemented in EIGENSTRAT to guard against possible population stratification . The first two principal components were not associated with case status ( P values>0 . 05 ) , further indicating that it is very unlikely that positive associations in this study would be attributable to confounding due to population structure . The association analyses by EIGENSTRAT confirmed , qualitatively , our main results and consequently , the results of the EIGENSTRAT analyses are not detailed here . Table 2 lists the most promising results from GWAS analyses . We identified five SNPs with P values<5×10−6 by allelic association analyses . After applying the Bonferroni correction for multiple testing , a single SNP , rs13182402 , reached a genome-wide significance level ( P<1 . 78×10−7 ) . SNP rs13182402 achieved a P value of 8 . 53×10−9 in the allelic test ( Bonferroni corrected P = 2 . 40×10−3 ) . The odds ratio ( OR ) was 2 . 94 ( 95% confidence interval ( CI ) : 2 . 02–4 . 30 ) for minor allele G . The frequency of the G allele was 0 . 162 in cases , and 0 . 061 in controls . When all covariates were considered simultaneously in a multivariate logistic regression model , this SNP remained a significant predictor of OF risk , independent of age , sex , height , and weight ( P = 2 . 21×10−8 ) . Given the significant evidence for rs13182402 , we imputed the genotypes of SNPs located surrounding this SNP based on our GWAS data and Asian HapMap data , and presented a regional association plot in Figure 2 . The most significantly associated SNP , rs13182402 ( GWAS: P = 8 . 53×10−9 ) , is located 394 bp downstream from exon 5 of the ALDH7A1 gene ( aldehyde dehydrogenase 7 family , member A1 ) on chromosome 5q31 . According to the FASTSNP program ( http://fastsnp . ibms . sinica . edu . tw ) , a change of “A→G” at rs13182402 may lead to removal of binding sites for transcription factors RORalp and CdxA .
In this study , we first performed a GWAS and follow-up replication on OF and identified a novel susceptibility gene ( ALDH7A1 ) that significantly impacts the risk for OF per se . Next , we examined this gene's relationship with hip BMD both in Chinese and Caucasian populations , and this gene was consistently associated with hip BMD even across ethnic boundaries . The effect size on BMD was modest and lower than the effect size on OF risk . One interpretation of this differential effect would be that BMD is not the only risk factor for OF; other risk factors also contribute to the risk of OF . It is consistent with and supports our statement in the introduction . It might also be caused by the differences in power between the relatively small hip OF samples compared to the large BMD samples . In addition , because we didn't have BMD measurements for the hip OF cases , we couldn't adjust the OR for BMD to see if the risk would be attenuated by the adjustment . However , regardless of this differential effect , the significant association results we identified both for BMD and OF risk strongly support the potential contribution of ALDH7A1 to the pathogenesis of osteoporosis . The ALDH7A1 gene encodes an enzyme of the acetaldehyde dehydrogenase superfamily , which degrades and detoxifies acetaldehyde generated by alcohol metabolism . Acetaldehyde has been shown to inhibit osteoblast proliferation and to decrease bone formation [19] . In addition , previous studies have identified that polymorphisms of the ALDH2 gene , another member of the acetaldehyde dehydrogenase family , are significantly associated with osteoporosis [20] . Our findings , combined with the above lines of evidence , suggest that ALDH7A1 might be a novel and potential candidate gene contributing to the risk of osteoporosis . Using the genotyped and imputed genotypes in our GWAS sample of 700 Chinese , we examined the associations between hip OF and the key SNPs identified in previous GWAS on osteoporosis [12] , [13] , [16] . Table 4 summarizes the major results . Only two SNPs in RANKL were confirmed to be associated with hip OF in our sample , including rs9594759 ( P = 0 . 020 ) and rs9594738 ( P = 0 . 045 ) . The data provided may serve as a reference for other investigators searching for replication for their GWAS results . An apparent advantage of this study is that our GWAS sample came from a homogenous population with well defined homogeneous phenotype . The genomic control factor was quite close to 1 . 0 ( λ = 1 . 02 ) ( expected under no population stratification ) and , analyses by EIGENSTRAT showed qualitatively supportive results . Thus , our association results are unlikely to be plagued by spurious associations due to population stratification . In particular , since the significant associations with BMD are shown in both Caucasian and Chinese samples , the results are even less likely to be due to population stratification/admixture . A potential limitation of our study is the relatively small size of the GWAS sample and the replication sample , which might lead to over estimation of the effect size for the significant SNPs identified . However , hip fractures are the most severe OFs followed by high mortality rates , making subjects recruitment difficult . It took us several years to accumulate such a homogeneous hip OF sample . This study represents the best we can do under current conditions to identify genes for OF . Meanwhile , we are keeping the recruitment of hip OF subjects . As a future direction , a new GWAS needs to be implemented on a larger sample to identify more comprehensively novel genes for OF . In addition , since genetic and environmental backgrounds vary for different populations , replication across a wide range of populations is necessary to determine the generality of our findings to the broader population , or to specific ethnic groups or populations [21] . In summary , using data from over 11 , 500 individuals , we have identified and validated ALDH7A1 as a novel susceptibility candidate gene for osteoporosis . Further studies are warranted to explore the generality of our findings for ALDH7A1 identified by GWAS to other populations , and to determine the mechanisms by which this gene and its products contribute to the pathogenesis of osteoporosis .
The study was approved by the local institutional review boards or the office of research administration of all participating institutions . After signing an informed consent , all subjects received assistance completing a structured questionnaire including anthropometric variables , lifestyles , and medical history . The study was initially performed with a GWAS discovery stage for SNPs of potential significance for OF in a Chinese Han , case-control sample . The significance of the SNPs identified in the discovery stage was subsequently confirmed through replication study in another independent Chinese case-control sample . For SNPs identified for OF , we further examined their relationships with hip BMD within/across ethnic groups in a Chinese unrelated BMD sample and three independent Caucasian samples . Table 1 details the basic characteristics of the respective samples , with additional descriptions below . The sample for the initial GWAS consisted of 350 patients with osteoporotic ( low trauma ) hip fractures ( including 124 males and 226 females ) and 350 elderly controls ( including 173 males and 177 females ) . Since fractures at different skeletal sites may have different underlying pathological mechanisms , we focused exclusively on hip fractures in order to minimize potential clinical and genetic heterogeneity of the study phenotype . All the subjects were unrelated northern Chinese Han adults living in the city of Xi'an and its neighboring areas . Affected individuals with low trauma hip fractures were recruited from the affiliated hospitals and their associated clinics of Xi'an Jiaotong University . Inclusion criteria for cases were ( i ) onset age of hip OF>55 years , to make sure all female subjects were postmenopausal , and the onset of OF was largely due to decreased BMD; ( ii ) age<80 years to minimize the effect due to age , since previous studies showed that approximately half of females aged 80 years or older have fractures [22]; ( iii ) fractures occurred with minimal or no trauma , usually due to falls from standing height or less; ( iv ) the fracture sites were at the femoral neck or inter-trochanter regions; ( v ) hip fracture was identified/confirmed through diagnosis of orthopedic surgeons/radiologists according to radiological reports and x-rays . Patients with pathological fractures and high-impact fractures ( such as due to motor vehicle accidents ) were excluded . Patients with chronic diseases before the onset of HF were also excluded . Healthy control subjects were enrolled by use of local advertisements . They were geography- and age-matched to the cases . Inclusion/exclusion criteria for controls were: ( i ) age at exam must be >55 years , without any fracture histories ( all female controls were postmenopausal ) ; ( ii ) subjects with chronic diseases and conditions that might potentially affect bone mass , structure , or metabolism were excluded . Diseases/conditions resulting in exclusion included chronic disorders involving vital organs ( heart , lung , liver , kidney , brain ) , serious metabolic diseases ( diabetes , hypo- and hyper-parathyroidism , hyperthyroidism , etc . ) , other skeletal diseases ( Paget's disease , osteogenesis imperfecta , rheumatoid arthritis , etc . ) , chronic use of drugs affecting bone metabolism ( e . g . , hormone replacement therapy , corticosteroid therapy , anti-convulsant drugs ) , and malnutrition conditions ( such as chronic diarrhea , chronic ulcerative colitis ) ; ( iii ) subjects taking anti-bone-resorptive or bone anabolic agents/drugs , such as bisphosphonates were excluded . For replication of our GWAS findings for hip OF , we used an independent Chinese sample containing 906 unrelated Han subjects ( 390 cases with hip OF and 516 controls ) . All subjects were drawn from the same geographic area as the above GWAS discovery sample , and the sample inclusion and exclusion criteria for cases and controls were the same as those adopted in the recruitment of the above GWAS sample . For SNPs that were identified for OF , we further performed validation analyses to evaluate their relevance with hip BMD ( with targeted experimental genotyping of candidate SNPs discovered in the initial GWAS ) in two ethnic groups , including a Chinese sample and two US Mid-West Caucasian samples . We finally performed in silico validation to compare the association signals of our most promising GWAS results with those achieved in the Framingham Heart Study ( FHS ) [23] . The Chinese BMD sample contained 2 , 955 unrelated ethnic Han adults . This sample came from Changsha , Hunan province , which is more than 1 , 000 km from Xi'an where the sample for the GWAS was recruited . The subjects were randomly selected from an established and expanding database with BMD measurements . The exclusion criteria were the same as those adopted in the recruitment of healthy control subjects in the GWAS sample , and have been detailed in our earlier publication [2] . The US-MidWest BMD samples with a total of 4 , 054 subjects consisted of two independent sample sets , including one sample of unrelated subjects and the other sample of nuclear families , which were all US Caucasians of Northern European origin living in Omaha , Nebraska , and its surrounding regions in Midwestern USA . They were normal healthy subjects defined by the same exclusion criteria as above in Chinese samples . The unrelated sample contained 1 , 725 subjects . The related sample contained 2 , 329 subjects from 593 nuclear families . All hip BMD measurements for the above BMD samples were obtained with dual-energy X-ray absorptiometry using the same type of machine ( Hologic 4500 ) under the same protocol defined by the manufacturer ( Hologic Inc . , Bedford , MA , USA ) . The machines were calibrated daily . The coefficients of variation ( CV ) of the hip BMD measurements were 1 . 33% for Chinese and 1 . 40% for Caucasians , respectively . The US-Framingham BMD sample is from the Framingham Osteoporosis Study , an ancillary study of the Framingham SNP Health Association Resource ( SHARe ) data sets [23] . Details and descriptions about the Framingham Osteoporosis Study have been previously reported [24] . Both genotype and phenotype data were downloaded from dbGaP database ( http://www . ncbi . nlm . nih . gov/sites/entrez ? db=gap ) . Data download and usage was authorized by SHARe data access committee ( phs000007 . v3 . p2 , phs000078 . v3 . p2 ) . We have the data on 2 , 953 phenotyped Caucasian subjects , 448 from the Original cohort ( 160 men and 288 women ) and 2 , 505 from the Offspring cohort ( 1 , 114 men and 1 , 391 women ) . The Original Cohort participants had BMD measures by dual x-ray absorptiometry machine ( Lunar DPX-L ) at the hip performed at exam 24 . The Offspring Cohort participants were scanned with the same machine at exam 6/7 . As reported before [24] , the CV was 1 . 7% for femoral neck . Genomic DNA was extracted from peripheral blood leukocytes using standard protocols . The genome-wide scan was performed using the Affymetrix Human Mapping 500K array set ( Affymetrix , Santa Clara , CA , USA ) according to the Affymetrix protocol . Data management and analyses were conducted using the Affymetrix GeneChip Operating System . Genotyping calls were determined from the fluorescent intensities using the DM algorithm with a 0 . 33 P-value setting [25] as well as the B-RLMM algorithm [26] . Quality control procedures were as follows . First , only samples with a minimum call rate of 95% were included . Due to efforts of repeat experiments , all samples ( n = 700 ) met this criteria and the final mean BRLMM call rate reached a high level of 99 . 02% . Second , out of the initial full-set of 500 , 568 SNPs , we discarded: 1 ) SNPs with a call rate <90% in the total sample ( n = 54 , 845 ) ; 2 ) those deviating from Hardy-Weinberg equilibrium ( HWE ) in controls ( P<0 . 001 , n = 22 , 002 ) ; 3 ) those having a minor allele frequency ( MAF ) <0 . 05 in the total sample ( n = 142 , 188 ) . Therefore , 281 , 533 SNPs were available for subsequent analyses . Based on the initial GWAS results , we selected the 10 most promising SNPs for subsequent genotyping in the Chinese hip OF replication sample based on the following inclusion criteria: ( i ) P values≤5×10−6 in the GWAS allelic association analyses ( 5 SNPs ) ; ( ii ) P values between 5×10−6 and 5×10−5 , with neighboring SNPs having P values≤10−4 showing a consistent trend of association ( 5 SNPs ) . Genotypes were obtained using MALDI-TOF mass spectrometry on a Sequenom system ( Sequenom , Inc . , San Diego , CA ) with iPLEX assay [27] . Primers were designed using MassARRAY Assay Design 3 . 1 software . Genotyping quality control procedures leading to SNP exclusion were call rate <90% , MAF<0 . 05 in the total sample and P<0 . 001 for deviations from HWE in controls . The average call rate was 98 . 7% for the Sequenom system and the corresponding consistency of genotyping ( replication or concordance rates ) , as obtained by duplication samples , was 99 . 8% . Nine of the ten genotyped SNPs were qualified for subsequent association analyses . The Chinese BMD sample was also genotyped using the Sequenom system , which was the same as that used for the OF replication sample . Genotyping of the two US-MidWest samples was performed by a service company KBioscience ( Herts , UK ) using a technology of competitive allele specific PCR ( KASPar ) , which is detailed at the website ( www . kbioscience . co . uk ) . The US-Framingham sample was genotyped using approximately 550 , 000 SNPs ( Affymetrix 500K mapping array plus Affymetrix 50K supplemental array ) . For the GWAS analyses for OF risk , single-marker allelic association analysis was performed by comparing SNP allele counts among cases and controls with a χ2 test . ORs with the corresponding 95% CIs were also computed . For the interesting SNPs identified by allelic tests , we also used a multivariate logistic regression model to examine associations with OF risk , taking into account potential covariates such as age , sex , height , and weight . For the sex chromosome analyses , the affymetrix platform does not assay the Y chromosome . The X chromosome needs to be treated differently from the autosomes since males have only one copy of the X chromosome . As most loci on the X chromosome are subject to X chromosome inactivation , it is reasonable to treat males as if they were homozygous females , and then the assumption was the same as tests for autosomes . All association statistical analyses were carried out using HelixTree 5 . 3 . 1 software ( Golden Helix , Bozeman , MT , USA ) . We adjusted for multiple testing by adopting the conservative Bonferroni correction . The genome-wide significance threshold was set at a P value of less than 1 . 78×10−7 ( 0 . 05/281 , 533 SNPs that passed our quality control check ) . To correct for potential population stratification that may lead to spurious association results , we estimated the inflation factor ( λ ) for the GWAS sample using a method of genomic control [28] . λ was calculated as the median of the observed χ2 statistics divided by the median of the expected χ2 statistics for the genome-wide SNP set . This led to an λ of 1 . 02 . Results presented in this study were based on adjusting χ2 statistics by dividing each of them by 1 . 02 . The data were also analyzed by the principal component analyses implemented in EIGENSTRAT [29] for cross-checking the association results while controlling for admixture . For OF replication analyses , the same allelic association analysis was performed by χ2 tests . For BMD validation analyses , significant parameters ( P<0 . 05 ) such as age , sex , height and weight were used as covariates to adjust for the raw BMD values . For the unrelated samples , ANOVA was conducted to achieve the association tests . ANOVA is a model free test and more robust than assuming any genetic models . The independent variable was the genotype , which was divided into three levels corresponding to the three genotypes observed for each SNP ( 1 , 1; 1 , 2; 2 , 2 ) . Since ANOVA can't give the effect size , we estimated the effect size of significant SNPs using the linear regression assuming the additive model in SAS ( SAS Institute Inc . , Cary , NC ) . For the related samples , we used the BMD residuals after adjustment to conduct family-based association tests under an additive model using FBAT program [30] . FBAT is a powerful approach to handle family sample , which tests for differences in probability of transmission of a genotype from parents to offspring based on phenotype . FBAT examines association within families , which is not affected by population stratification bias . To quantify the overall evidence of associations , we performed meta-analyses by using the Mantel-Haenszel method to calculate the P values and OR for the combined OF samples . For the BMD samples , we used a weighted Z-score method to calculate the combined P values . The individual Z-score ( a standard normal deviate , the statistic associated with a P value ) was weighted by the square root of the sample size of each sample . We added the individual weighted Z-score together and divided by the square root of the total sample size to obtain a combined Z-score and an associated combined P value ( Stouffer method ) [31] . If an individual result was nonsignificant and gave no other useful data for calculation of a Z-score , we set it as 0 to calculate the combined probability . For the interested genomic regions , IMPUTE program [32] was utilized to impute the genotypes of all SNPs located in the regions based on Asian HapMap data . SNPTEST [32] was used to test for associations between the imputed SNPs and OF . SNAP was used to depict the regional association plot [33] .
|
Osteoporosis is a major health concern worldwide . It is a highly heritable disease characterized mainly by low bone mineral density ( BMD ) and/or osteoporotic fractures . However , the specific genetic variants determining risk for low BMD or OF are largely unknown . Here , taking advantage of recent technological advances in human genetics , we performed a genome-wide association study and follow-up validation studies to identify genetic variants for osteoporosis . By examining a total of 11 , 568 individuals from Chinese and Caucasian populations , we discovered a susceptibility gene , ALDH7A1 , which is associated with hip osteoporotic fracture and BMD . ALDH7A1 might inhibit osteoblast proliferation and decrease bone formation . Our finding opens a new avenue for exploring the pathophysiology of osteoporosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/complex",
"traits"
] |
2010
|
Genome-Wide Association Study Identifies ALDH7A1 as a Novel Susceptibility Gene for Osteoporosis
|
Between 2014 and 2016 more than 3 , 800 imported human cases of chikungunya fever in Florida highlight the high risk for local transmission . To examine the potential for sustained local transmission of chikungunya virus ( CHIKV ) in Florida we tested whether local populations of Aedes aegypti and Aedes albopictus show differences in susceptibility to infection and transmission to two emergent lineages of CHIKV , Indian Ocean ( IOC ) and Asian genotypes ( AC ) in laboratory experiments . All examined populations of Ae . aegypti and Ae . albopictus mosquitoes displayed susceptibility to infection , rapid viral dissemination into the hemocoel , and transmission for both emergent lineages of CHIKV . Aedes albopictus had higher disseminated infection and transmission of IOC sooner after ingesting CHIKV infected blood than Ae . aegypti . Aedes aegypti had higher disseminated infection and transmission later during infection with AC than Ae . albopictus . Viral dissemination and transmission of AC declined during the extrinsic incubation period , suggesting that transmission risk declines with length of infection . Interestingly , the reduction in transmission of AC was less in Ae . aegypti than Ae . albopictus , suggesting that older Ae . aegypti females are relatively more competent vectors than similar aged Ae . albopictus females . Aedes aegypti originating from the Dominican Republic had viral dissemination and transmission rates for IOC and AC strains that were lower than for Florida vectors . We identified small-scale geographic variation in vector competence among Ae . aegypti and Ae . albopictus that may contribute to regional differences in risk of CHIKV transmission in Florida .
Native to Africa , chikungunya virus ( CHIKV ) emerged to produce intermittent outbreaks from the 1950s in Southeast Asia ( Asian CHIKV lineage ) and regional outbreaks in India in the 1960s and 1970s [1 , 2] . Chikungunya also emerged in Kenya in 2004 ( Eastern/Central/Southern African , ECSA , CHIKV lineage ) , followed by an outbreak of chikungunya fever on the island of La Réunion in 2005–2006 involving the Indian Ocean CHIKV ( IOC ) lineage , a descendent of the ECSA CHIKV lineage [2 , 3] . In 2013 an Asian lineage of CHIKV ( AC ) was detected and transmitted locally on St . Martin Island , a French collectivity in the Caribbean [4 , 5] , followed by spread throughout much of the Americas by 2015 [6] . The Old-World outbreaks of CHIKV in Kenya in 2004 [3] and islands of the Indian Ocean in 2005 subsequently spread to India and Europe including Italy and France [7 , 8] involving more than one million cases . The outbreaks in Europe were one of the first demonstrations that CHIKV could extend its tropical/subtropical distribution into temperate regions using the Asian tiger mosquito vector Aedes albopictus ( Skuse ) . Due to its ability to tolerate lower temperatures [9] Ae . albopictus occurs at more northern latitudes than Aedes aegypti ( L . ) , which is usually considered the primary vector of CHIKV . The Old-World outbreaks were caused by the Indian Ocean strain of CHIKV [5 , 10] . Asia is part of the invasive range of CHIKV where Ae . aegypti is the primary vector [11 , 12] . The virus is native and endemic to Africa , where arboreal mosquitoes are part of its sylvan cycle , including members of the Ae . furcifer-taylori group [11 , 13] . Chikungunya virus can cause widespread epidemics with infection rates exceeding 25% in some locations ( e . g . , La Réunion , Americas ) [6 , 14] . It is estimated that more than 4 million cases have occurred worldwide in the past 12 years [12] . Human CHIKV infection causes high fever , rash , headache , joint swelling , and joint pain [15] . Additionally , chronic musculoskeletal diseases may last for months to years following infection [16] . The widespread , invasive mosquito Ae . albopictus was the vector of CHIKV on La Réunion during 2005–2006 [17 , 18] and likely the primary vector of an outbreak in 2007 in Gabon [19 , 20] . A single mutation in the E1 protein of CHIKV enhanced infection and transmission in Ae . albopictus [21] , a species that was considered to be secondary in importance to the primary vector Ae . aegypti . There is now evidence showing that there have been multiple independent events of CHIKV exposure to Ae . albopictus populations followed by development of this adaptive mutation [21] . This observation suggests the potential for outbreaks involving Ae . albopictus as the vector in regions where Ae . aegypti is rare or absent . The expansion of CHIKV in the Americas that began in 2013 increases the burden of disease in a region of the world recently invaded by Zika virus and where dengue is endemic . Local transmission in the U . S . is a major public health risk especially in Texas and Florida where both potential mosquito vector species reside , the environmental conditions promote vector abundance throughout much of the year , and there is a high potential for virus introduction [22] . The first documentation of locally-acquired cases of CHIKV in the continental U . S . states occurred in 2014 involving 11 cases in Florida [23] . During 2014 , the Dominican Republic was one of the countries with the most numerous suspected CHIKV cases in the Americas [6 , 24] . Although it remains unclear what accounts for the observed high number of cases , this country is highly urbanized and many people live in densely populated marginal barrios . These environmental conditions together with domestic storage of water and irregular trash collections foster conditions favorable for the proliferation of Ae . aegypti which is present in most major cities in the Caribbean Basin . Similar environmental conditions likely exist in other parts of the Americas that experienced a high number of CHIKV cases . Additionally , the Dominican Republic is a top destination for tourism , which may have facilitated introduction of CHIKV and subsequent local transmission . Variation in vector competence and the extrinsic incubation period is another possible explanation; Ae . aegypti originating from the Dominican Republic may be highly competent for CHIKV , and the high infection rates in humans are a product of an efficient vector population . Studies to date have demonstrated distinct differences in vector competence of Ae . aegypti and Ae . albopictus depending on geographic origin of the mosquitoes and CHIKV lineage [25 , 26] . Aedes aegypti tends to be a more efficient transmitter of the Asian and ancestral East/Central/South Africa ( ECSA ) lineages , whereas Ae . albopictus is a more competent vector of the Indian Ocean strain of CHIKV [21 , 25 , 27] . Although a few studies have assessed infection and viral dissemination in Ae . aegypti and Ae . albopictus from Florida [28–31] , little is known about the ability of these Aedes to transmit CHIKV ( i . e . , transmission efficiency ) . [30] showed that 0–20% of Ae . aegypti and 7–21% of Ae . albopictus were capable of transmitting the Indian Ocean strain of CHIKV ( LR2006-OPY1 ) . However , that study used Florida-derived Ae . aegypti and Ae . albopictus that had been maintained as laboratory colonies for > 50 generations and may not be representative of field populations . Furthermore , the vector competences of Ae . aegypti and Ae . albopictus populations differ among the three CHIKV lineages ( ECSA , West Africa , Asian ) [25 , 27] , and transmission of Caribbean CHIKV from the Asian lineage is likely to be affected by genetic differences in vector competence of Florida Aedes vectors . An assessment of the vector competence of 35 populations of American Ae . aegypti and Ae . albopictus for two strains ( i . e . , individual isolates ) in the ECSA lineage and one strain in the Asian lineage of CHIKV revealed that viral dissemination was high for all mosquito populations [32] . However , transmission rates differed vastly among American populations ( 11–97% ) suggesting that salivary gland infection/escape barriers affect vector competence among these two-vector species and their potential to transmit CHIKV . For the Indian Ocean and ancestral ECSA genotypes ( both in the ECSA lineage ) transmission efficiencies among F1 generation Ae . aegypti and Ae . albopictus from Vero Beach , FL were lower ( <30% ) than most other American populations of these species . However , the vector competence of these Aedes mosquitoes from Vero Beach was not tested for the Asian lineage of CHIKV . [32] provided preliminary information that Florida Aedes may differ from other American populations of these species in their response to CHIKV attributable to differences in salivary gland infection/escape barriers . Little is known about the vector competence of Aedes mosquitoes for the Asian lineage of CHIKV responsible for the outbreak in the Americas , including Florida . Chikungunya virus in the Americas belongs to the Asian lineage , suggesting that Ae . albopictus will transmit at a lower rate than Ae . aegypti due to an adaptive constraint [33] . However , other studies have suggested that Ae . albopictus were as competent as Ae . aegypti for transmission of an isolate of CHIKV from Saint Martin Island belonging to the Asian lineage [34] . With the potential for a major CHIKV epidemic in Florida , there is a need to appraise the relative risk and emergence of Chikungunya fever in Florida . An assessment using meteorological driven models to inform baseline risk for local Zika virus transmission in the U . S . , and presumably other viruses including CHIKV and dengue viruses , showed that cities in southern Florida and south Texas were highly suitable for Ae . aegypti and imported viral cases [22] . In this paper , we examine CHIKV disseminated infection and transmission in Florida mosquitoes for two putative vector species , Ae . aegypti and Ae . albopictus . We tested whether local populations of Ae . aegypti and Ae . albopictus show regional differences in susceptibility to infection and transmission to two emergent lineages of CHIKV , Indian Ocean ( IOC ) and Asian genotypes ( British Virgin Islands , AC ) . As a baseline comparison , we compare susceptibility to infection and transmission Aedes vectors from Florida to Ae . aegypti from the Dominican Republic , one of the countries associated with the most numerous cases of CHIKV during the American outbreak in 2014 . Although Ae . albopictus is present in the Dominican Republic , it is often found at far lower abundances than Ae . aegypti and so we focused on the latter species [35] .
We chose collection sites ( Fig 1 ) based on distributions of these Aedes species across the state of Florida and areas where local arbovirus transmission ( chikungunya , dengue , and Zika viruses ) has been detected in areas where these vector species are present . Larval Ae . aegypti and Ae . albopictus were collected in 2014 from cemeteries or tire/salvage yards across Florida where these species are present alone or coexist [36] . Collection sites for Ae . aegypti included Manatee ( Bradenton ) , Monroe ( Key West ) , and Indian River/St . Lucie ( Vero Beach or White City ) Counties . We initially made collections from separate sites in Indian River Co . and St . Lucie Co . but later decided to combine these to augment sample size given their proximity to one another ( 23 km between sites ) . Collection sites for Ae . albopictus included Alachua ( Gainesville ) , Manatee ( Bradenton ) , and Indian River/St . Lucie ( Vero Beach or White City ) Counties . So , we included collections from distinct regions of Florida ( East , West , North , and South ) for our assessment of regional differences in susceptibility to infection and transmission of two emergent lineages of chikungunya virus among Ae . aegypti and Ae . albopictus . We also included a laboratory colony of Ae . aegypti originally collected in Orlando , FL ( Orange Co . ) and maintained in colony since 1952 . Although no strong geographic genetic differentiation among Florida populations of Ae . aegypti has been reported , there is some evidence of genetic isolation of Florida Keys Ae . aegypti from mainland Florida [37] . We were provided with eggs of Ae . aegypti collected in 2014 from La Romana , Dominican Republic by the University of Texas Medical Branch which were propagated at the Florida Medial Entomology Laboratory for the CHIKV infection study . The inclusion of the Dominican Republic strain of Ae . aegypti enabled us to compare Florida Aedes vectors to a separate vector population involved in outbreaks with the most numerous cases of CHIKV in the Americas in 2014 [6] . Field-collected mosquitoes were reared to adulthood on a diet of equal parts of brewer’s yeast and liver powder larval food at 26–28°C . Pupae were collected daily and placed in vials with a cotton seal and upon emergence identified to species . Adults were provided with 10% sucrose solution and allowed to feed through hog casing membranes on commercially purchased defibrinated bovine blood ( Hemostat Laboratories , Dixon , CA ) once per week to propagate eggs . Larvae were reared at an approximate density of 150 larvae/L water in plastic photo trays ( 25cm width , 30cm length , 5cm height; Richard MFG Co . Fernandina Beach , FL , U . S . A ) with 900 mL of water and 0 . 4 g larval food at hatching and supplemented again with the same amount 3–4 days later . Larvae developed to the pupal stage between 5–7 days after egg hatch . Adult males and females were held together for nine days in a cage ( 0 . 33 m ) in a climate controlled room ( 26–28°C , photoperiod of 14:10 light:dark ) and provided with 10% sucrose solution . Females were placed in cylindrical cages ( height x diameter: 10 cm by 10 cm , 50 females/cage ) with mesh screening one day before exposure to CHIKV infected blood and deprived of sucrose but not water . The F1-3 generation progeny of field-collected Ae . aegypti and Ae . albopictus , including Ae . aegypti from the Dominican Republic , were used for the CHIKV infection studies in the biosafety level-3 virology facility at the FMEL in Vero Beach , FL . A strain of CHIKV from the British Virgin Islands ( BVI ) ( Asian lineage , GenBank accession: KJ451624 ) , which is responsible for outbreaks that began in St . Martin in 2013 , was obtained in December 2013 from an infected human . The Indian Ocean genotype ( IOC ) of CHIKV ( LR2006-OPY1 , GenBank accession: KT449801 ) , responsible for the outbreak in the Indian Ocean region and parts of Europe [2] , was isolated from a febrile patient in France who had been infected in La Réunion [38] . The virus isolates were obtained from the Centers for Disease Control and Prevention and the University of Texas Medical Branch in Galveston , TX . These CHIKV strains ( passaged twice ) were propagated in culture using African green monkey ( Vero ) cells , in which viral titer was determined by plaque assay [28] . Ten to thirteen-day old adult females were provided with CHIKV infected defibrinated bovine blood ( Hemostat , Dixon , CA ) using an artificial membrane feeding system ( Hemotek , Lancashire , United Kingdom ) as described previously [39] . Aliquots of blood were stored at -80°C for later determination of virus titer . Briefly , to prepare fresh virus for mosquito infection , monolayers of Vero cells in T-175 cm2 flasks were inoculated with 500 μl of diluted stock CHIKV ( multiplicity of infection , 0 . 1 ) and incubated for 1 hr at 37°C and 5% CO2 atmosphere , after which 24 mL media ( M199 medium supplemented with 10% fetal bovine serum , penicillin/streptomycin and mycostatin ) were added to each flask and incubated for an additional 47-hours . Mosquitoes were fed either a low dose ( 5 . 8 log10 pfu/ml ) or high dose of CHIKV infected blood ( 8 log10 pfu/ml ) . After the feeding trials , fully engorged females were held in cylindrical cages and maintained at a 14:10 hour light:dark photoperiod and 30°C . To assess ability to transmit CHIKV mosquitoes were transferred to 37-mL plastic tubes ( height x diameter: 8 by 3 cm ) along with an oviposition substrate . Each tube held one mosquito and was fitted with a removable screen lid . Mosquitoes were deprived of sucrose for one day before the transmission trial started . Only Ae . aegypti from St . Lucie Co . were fed the low dose of the BVIC strain of CHIKV and held at 25°C or 30°C and tested for infection and salivary infection six days after ingesting CHIKV infected blood . Cohorts of mosquitoes were tested for transmission of CHIKV at 2 , 5–6 , and 12–13 days after feeding on a high dose of infected blood . Each tube containing a mosquito was presented with a honey-soaked filter paper ( ≈1 cm diameter ) fastened to the inside of the lid . The honey was dyed with blue food coloring ( McCormick ) which provided a visual marker indicating that a mosquito fed on the honey and presumably deposited saliva during feeding . A similar system using FTA® cards ( Flinders Technology Associates filter paper ) instead of filter paper has been used successfully as a surveillance system to detect arboviruses that exploits the fact that female mosquitoes expectorate virus in their saliva during feeding on sugar sources [40] . An initial assessment using FTA cards to test for ability to transmit CHIKV suggested a toxic effect on mosquitoes ( early death ) and so we switched to using filter paper as the substrate to collect mosquito saliva . Here we use this methodology as a proxy for potential to transmit CHIKV . Mosquitoes were examined with a flashlight for blue coloring in their crop after 24 and 48-hours during the transmission assay . Mosquitoes and filter paper were collected upon first detection of blue in the crop and frozen at -80°C and later analyzed for expectorated virus using quantitative ( q ) RT-PCR [28] , so that CHIKV remained on the filter paper no longer than 24-hours before being frozen . Mosquitoes that did not feed on blue honey were not tested for CHIKV transmission . Cohorts of mosquitoes were tested for transmission of CHIKV at 2 , 5–6 , and 12–13 days after feeding on infected blood . Additionally , saliva was collected from these same mosquitoes in capillary tubes with immersion oil as described previously [41] after they had fed on blue honey for the second and third time points only . Additional studies ( Alto et al . , in preparation ) indicate that the blue honey method is equivalent or slightly underestimates virus in saliva compared to capillary tube methods . All mosquitoes were immediately killed and stored at -80°C upon completion of each transmission assay . Mosquitoes were individually dissected and the bodies and legs were tested separately for the presence of CHIKV RNA by qRT-PCR using methods of [28] . Primers were designed to target a nonstructural polyprotein gene common to both lineages ( accession ID of transcript , KU365292 . 1 ) with the following sequences: forward , 5'-GTACGGAAGGTAAACTGGTATGG-3': reverse , 5'-TCCACCTCCCACTCCTTAAT-3' . The probe sequence was: 5'-/56-FAM/TGCAGAACCCACCGAAAGGAAACT/3BHQ_1/-3' ( Integrated DNA Technologies , Coralville , IA ) . Disseminated infection was calculated as the percent of infected legs from the total number engorged with blood . Transmission was calculated as the percent of saliva infected mosquitoes from the total number of mosquitoes with infected legs . The legs and filter paper were homogenized separately in 1 . 0 mL of 199 media . The saliva from mosquitoes collected in capillary tubes was combined with 300 μL of media . A 140 μL sample of mosquito legs , filter paper , and saliva homogenate was used for RNA isolation using the QIAamp viral RNA mini kit ( Qiagen , Valencia , CA ) and eluted in 50 μL of buffer per the manufacturer’s protocol . CHIKV RNA was detected using the Superscript III One-Step qRT-PCR with Platinum Taq kit by Invitrogen ( Invitrogen , Carlsbad , CA ) as described previously [28] . Quantitative RT-PCR was performed with the CFX96 Real-Time PCR Detection System ( Bio-Rad Laboratories , Hercules , CA ) using primers and probes specific to the Asian and Indian Ocean lineages of CHIKV . The program for qRT-PCR was as follows; 50°C for 30 minutes , 94°C for 2 minutes , 39 cycles at 94°C for 10 seconds and 60°C for 1 minute , and lastly 50°C for 30 seconds . A standard curve method was used to express the titer of CHIKV of mosquito samples by comparing cDNA synthesis for a range of serial dilutions of CHIKV in parallel with plaque assays of the same dilutions of virus , expressed as plaque forming unit equivalents ( pfue ) /ml [42] . Mosquito species , time , location , and species by time interaction effects on transmission were analyzed using maximum likelihood categorical analyses of contingency tables ( PROC CATMOD , SAS 2002 ) based on the number of mosquitoes categorized for the presence or absence of CHIKV on the filter paper ( first time point ) and in capillary tubes ( second and third time points ) . When significant treatment effects were found , follow-up analyses included pairwise comparisons of treatments , correcting for multiple comparisons using the sequential Bonferroni method . We chose this analysis for consistency and improved comparison to other CHIKV studies [28 , 32 , 34 , 43 , 44] . Separate analyses of transmission were performed for the first time point and combined for the second and third time points because different methods were used to collect saliva . Also , separate analyses were performed for each of the CHIKV lineages since these experiments were performed at different times . Maximum likelihood categorical analyses of contingency tables were used to test for treatment effects ( see salivary infection methods ) on viral dissemination efficiencies to gauge barriers to transmission ( midgut escape barrier ) . Each infection experiment with Ae . aegypti and Ae . albopictus and CHIKV was conducted only once . Individual mosquitoes are the unit of replication and we analyzed infection responses by analysis of frequency distribution [45] . Analysis of variance was used to test for differences in virus titers in the legs and saliva of the individual mosquitoes . Significant effects were followed by Tukey-Kramer multiple comparisons among treatment least-squares means for pairwise comparisons .
Preliminary studies using Ae . aegypti mosquitoes from St . Lucie Co . and AC identified baseline infection and disseminated infection rates at constant 25°C and 30°C six days after ingesting CHIKV infected blood . Although not significant ( χ2 = 2 . 55 , df = 1 , p = 0 . 11 ) , we observed 3-fold differences in susceptibility to infection between 30°C ( 10 . 5% , 67 mosquitoes tested ) and 25°C ( 3 . 6% , 83 mosquitoes tested ) . Viral dissemination ( χ2 = 0 . 05 , df = 1 , p = 0 . 83 , 8 mosquitoes ) and transmission ( χ2 = 0 . 44 , df = 1 , p = 0 . 51 , 8 mosquitoes ) did not differ between the two temperatures . Viral dissemination rates of 100% and 40% were observed at 25°C and 30°C , respectively . Transmission of 33 . 3% and 0% was observed at 25°C and 30°C , respectively , using only the capillary tube method of collection of saliva . There was an effect of species , origin of Ae . aegypti population , and the species x time since exposure interaction on disseminated infections of IOC ( Table 1 ) . There were significantly more Ae . albopictus with disseminated infections at days 2 and 12 , but not day 5 , after IOC exposure than Ae . aegypti ( day 2 albo ( 92 . 3% ) > aeg ( 73 . 3% ) , χ2 = 13 . 6 , df = 1 , p = 0 . 0002; day 5 albo ( 88 . 0% ) = aeg ( 89 . 5% ) , χ2 = 0 . 11 , df = 1 , p = 0 . 73; day 12 albo ( 94 . 7% ) > aeg ( 75 . 6% ) , χ2 = 7 . 9 , df = 1 , p = 0 . 0049; Percentages combine over geographic populations in Table 2 ) . Aedes aegypti from the Dominican Republic and Manatee Co . , FL had lower or similar virus dissemination than Ae . aegypti from other locations in Florida ( Fig 2 ) . There was an effect of species , origin of Ae . albopictus population , and origin of Ae . aegypti on transmission of IOC two days following ingestion of infected blood ( Table 1 ) . Aedes albopictus transmission was higher than Ae . aegypti on day 2 after IOC exposure ( albo = 38 . 6% , aeg = 21 . 7% , χ2 = 20 . 7 , df = 1 , p<0 . 0001; Percentages combine over geographic populations in Table 2 ) . Aedes albopictus from Indian River/St . Lucie Co . , FL had higher transmission than other geographic locations ( Fig 3 ) . Ae . aegypti from Monroe Co . had significantly higher transmission than from Manatee Co . ( χ2 = 15 . 0 , df = 1 , p = 0 . 0001 ) and Indian River/St . Lucie Co . , FL ( χ2 = 8 . 0 , df = 1 , p = 0 . 0047 ) ( Fig 4 ) . Origin of Ae . albopictus affected transmission of IOC five to twelve days following ingestion of infected blood ( Table 1 ) . Aedes albopictus from Indian River/St . Lucie Co . and Manatee Co . had higher transmission than from Alachua Co . , FL ( Fig 5 ) . There was an effect of species , origin of Ae . albopictus population , and time since exposure on disseminated infection of AC ( Table 3 ) . Higher rates of disseminated infection were observed in Ae . aegypti ( 80 . 9% ) than Ae . albopictus ( 71 . 2% ) ( χ2 = 4 . 3 , df = 1 , p = 0 . 03; Percentages combine over geographic populations and days post infection in Table 4 ) . Disseminated infections increased from day 2 ( 55 . 3% ) to day 5 ( 96 . 2% ) , then subsequently decreased again on day 12 ( 82 . 8% ) following exposure . Disseminated infection was significantly different for each of the three time points measured ( All χ2>21 . 2 and p<0 . 0001 ) . Aedes albopictus from Alachua Co . , FL had higher disseminated infection than other locations , followed by Indian River/St . Lucie Co . and Manatee Co . ( Fig 6 ) . Origin of Ae . aegypti population significantly affected transmission of AC two days following ingestion of infected blood ( Tables 3 and 4 ) . Aedes aegypti from Monroe Co . , FL had higher transmission rates than those from the Dominican Republic ( χ2 = 6 . 71 , p = 0 . 009 ) . However , after correcting alpha for multiple comparisons this difference was only marginally significant . Geographic origin of Ae . aegypti , time , and the species x time since exposure interaction significantly affected transmission of AC five to twelve days following ingestion of infected blood ( Table 3 ) . Aedes aegypti from the Dominican Republic and the laboratory colony ( Orlando ) had similar or lower transmission rates than from all other locations in Florida ( Fig 7 ) . Transmission rates decreased during the infection between days five and twelve , but this effect differed between Ae . aegypti and Ae . albopictus . Aedes aegypti transmission was higher than Ae . albopictus at twelve days , but not five days , following ingestion of AC infected blood ( Day 5 aeg vs . albo , χ2 = 3 . 76 , df = 1 , p = 0 . 05; Day 12 aeg vs . albo , χ2 = 6 . 22 , df = 1 , p = 0 . 01 ) . There were significant temporal differences in the leg titer equivalents of individuals following exposure to both emergent CHIKV lineages ( Table 5 ) . For IOC , viral titer equivalents were significantly higher on days 5 and 12 than day 2 ( Fig 8A ) . For AC , viral titer equivalents were significantly higher on day 5 than days 2 and 12 ( Fig 8B ) . There were no significant treatment effects of species , location , or species by time interaction on leg titer equivalents for either AC or IOC ( Table 5 ) . Our ability to detect temporal patterns in saliva viral titer equivalents was limited because we used two different analyses for day two versus days five and twelve . There were no significant differences in viral titer in saliva from the species , location , time , or species by time interaction for either AC or IOC ( Table 5 ) . The one exception was that there was a location effect for Ae . albopictus so that individuals originating from Alachua Co . FL had significantly higher IOC viral load in the saliva than individuals from Indian River/St . Lucie Co . and Manatee Co . FL ( Table 5 ) .
Preliminary studies using Ae . aegypti mosquitoes from St . Lucie Co . , FL and AC identified baseline infection and disseminated infection rates at constant 25°C and 30°C . We observed that Ae . aegypti had a low susceptibility to infection for AC , but a relatively permissive midgut escape barrier after ingesting a low dose of CHIKV infected blood . A midgut infection barrier refers to the inhibition of ingested arboviruses from entering or replicating in midgut cells . A midgut escape barrier refers to the inhibition of arboviruses from spreading beyond the basal lamina of the midgut cells to the hemocoel . The low infection rates were attributed to a relatively low dose of CHIKV in blood meals . Lack of significant differences among dissemination and transmission rates is most likely attributable to low sample sizes . Our main study demonstrated that the midgut infection barriers can be surpassed by high virus titers [47] . All populations of Ae . aegypti and Ae . albopictus mosquitoes displayed susceptibility to infection and transmission for the two emergent lineages of CHIKV at high titers . Viral dissemination to the hemocoel for Ae . aegypti and Ae . albopictus mosquitoes was rapid and co-occurred with infection of the saliva , with substantial transmission occurring by day 2 dpi ( Tables 2 and 4 ) . This observation has important implications for CHIKV epidemiology because both Ae . aegypti and Ae . albopictus exhibit gonotrophic discordance whereby mosquitoes will blood feed more than once in a single gonotrophic cycle [48–51] , allowing for the possibility of transmission during each feeding event . Florida vectors are highly competent , especially given the short extrinsic incubation period of CHIKV [44] which strongly contributes to vectorial capacity as an exponential function [52] . Viral disseminated infection indicated rapid propagation in the midgut and spread to other mosquito tissues , with rates being higher for IOC than AC . Viral dissemination within Ae . aegypti and Ae . albopictus occurred for most individuals following five days of extrinsic incubation , suggesting a lack of substantial midgut escape barriers for IOC and AC [32] . For IOC , Ae . albopictus had higher disseminated infection than Ae . aegypti in most instances , suggesting that Ae . albopictus is more permissive to infection by this strain than Ae . aegypti . Similarly , Ae . albopictus had more efficient transmission of the IOC lineage sooner after ingesting CHIKV infected blood than Ae . aegypti . Specifically , Ae . albopictus had a 44% greater proportion of transmission than Ae . aegypti . However , later in the infection process Ae . aegypti and Ae . albopictus had similar transmission . These observations are consistent with other studies showing more efficient viral dissemination for IOC into mosquito secondary organs and transmission in Ae . albopictus than Ae . aegypti [27 , 32 , 53] . We can infer from our observations that the duration of the extrinsic incubation period of IOC is shorter in Ae . albopictus than Ae . aegypti [43] . Differences in vector competence between Ae . albopictus and Ae . aegypti contribute as a linear function , and so relatively weakly , to vectorial capacity whereas changes in the extrinsic incubation period contribute as an exponential function and thus more strongly [52] . A short incubation period in Ae . albopictus probably contributed to its role as the vector in the chikungunya outbreaks in the Indian Ocean in 2005–2007 . Other contributing factors to this outbreak include an increased infectivity of Ae . albopictus to this strain by 100-fold and that Ae . aegypti was relatively rarer and non-anthropophilic on Reunion Island . Tsetsarkin et al . [21] tested the hypothesis , using viral infectious clones of CHIKV , that a mutation in the envelope protein gene ( E1-A226V of IOC ) influenced viral fitness for different vector species . Their study demonstrated that the E1-A226V mutation was directly responsible for increased infectivity and more efficient viral dissemination into mosquito secondary organs and transmission for Ae . albopictus compared to Ae . aegypti [21] . This adaptive mutation has been selected for on multiple independent occasions , evidence for convergent evolution and the ability of IOC to adapt locally to vectors [54] . Viral dissemination , but not transmission , of AC were lower than IOC two days following ingestion of infectious blood by both Ae . aegypti and Ae . albopictus . However , transmission was similar or higher for both mosquito species after five days of extrinsic incubation to AC than to IOC ( Tables 2 and 4 ) . Viral dissemination and transmission was higher in Ae . aegypti than Ae . albopictus which is consistent with other studies comparing the vector competence of these two species for the Asian lineage of CHIKV [21 , 27 , 32] . Although viral dissemination rates of AC and IOC were high for most mosquito populations , transmission was lower , suggesting substantial salivary gland infection or escape barriers [32] . Viral dissemination and transmission of AC decreased from 5 to 12 days of extrinsic incubation , suggesting that transmission risk declines with length of infection . Interestingly , the decline in transmission of AC was less in Ae . aegypti than Ae . albopictus , suggesting that older Ae . aegypti females are relatively more competent vectors than similar aged Ae . albopictus females . Older mosquitoes represent a greater epidemiological threat , because they are more likely to have ingested a virus-infected blood meal and completed the extrinsic incubation period and are more likely capable of transmitting the virus during these advanced stages of infection [55] . Our observations that transmission efficiency declines with length of infection are likely to mitigate age-enhanced transmission potential . Indices of risk of transmission , such as vectorial capacity , do not consider that different species or genotypes of pathogens are affected differently by time since infection , irrespective of daily vector survivorship . Existing models of vector-borne arboviruses could be used to determine whether the pattern of transmission is altered by the addition of time-dependent transmission efficiency during the lifespan of the vector , or by species-specific variation in EIP , transmission or mosquito life history traits . Preliminary studies using a model to investigate the likelihood of CHIKV epidemics in FL after introductions showed differences in the outcome when species-specific variation in vector competence for CHIKV , mosquito mortality and human biting frequency were considered . Ongoing research is investigating effects of the variation in EIP observed in this study and will be published elsewhere . One plausible mechanism that would account for decline in transmission efficiency with length of infection is virus modulation of the infection by the mosquito [56] . This hypothesis predicts that reduced viral titer in mosquito tissues would be observed on day 12 compared to day 5 because of the observed decline in transmission at the later time point . We would thus expect to see changes in leg viral titers in association with changes in saliva infection over time . The observed reduction in leg viral titer , in association with reduction in transmission , for AC supports this hypothesis . Further support comes from the observation that saliva infection did not decline over time in IOC , and neither did leg viral titer . Salazar et al . [57] showed that after ingestion of dengue-2 virus , peak virus titer in Ae . aegypti was observed after 7–10 days of extrinsic incubation followed by a steady decline later during infection . Although the molecular mechanism responsible for reduced virus load was unclear , the authors suggested several potential means including physiologically compromised epithelial cells , post-transcriptional or post-translational repression , or an antiviral response [57] . Similar observations were made with Ae . aegypti infected with dengue-2 by Sánchez-Vargas et al . [58] in which they showed that dengue infection and viral titer in Ae . aegypti were modulated by the RNAi defense system . Additional experiments are needed to identify the mechanism ( s ) responsible for modulation of infection observed in the current study . Large scale geographic variation in vector competence of Ae . aegypti and Ae . albopictus have been observed among lineages of CHIKV [32 , 59] as well as for other arboviruses ( Zika , yellow fever , and dengue-2 viruses ) [60–64] . In the current study , we have identified variation in vector competence on a smaller spatial scale than previously recognized . Regional differences in mosquito-virus interactions , especially as they might relate to the EIP , may have important implications for risk of disease transmission . However , the geographic variation wasn’t consistent between CHIKV lineages and mosquito species and the differences were often relatively small . Aedes aegypti from Manatee Co . , Florida and Dominican Republic had lower or similar viral dissemination of IOC from one or more other locations in Florida ( Fig 2 ) . Along the same lines , transmission was observed to be lower for Ae . aegypti from Manatee Co . and from Indian River/St . Lucie Co . than Monroe Co ( Fig 4 ) . A study on the phylogeography of Ae . aegypti in Florida did not find strong genetic differentiation among Florida populations of Ae . aegypti from East and West coasts , but there was some evidence of genetic isolation of Florida Keys Ae . aegypti from mainland [37] which may , in part , explain our observation . Similar studies characterizing genetic differentiation among Florida populations of Ae . albopictus have not yet been published . Studies of Ae . aegypti from single nucleotide polymorphisms and sequenced nuclear genes have demonstrated differences in populations from the Caribbean compared to mainland U . S . [65] , which is consistent with large scale difference in quantitative genetics of vector competence [66] . For Ae . albopictus , although IOC disseminated infection was high and homogenous between collection sites , transmission rate was much lower and varied by origin , suggesting distinct barriers to transmission that may be operating at small geographic scales ( Figs 3 and 5 ) . Specifically , Ae . albopictus from Alachua Co . had lower IOC transmission than individuals originating from either Manatee Co . or Indian River/St . Lucie Co . , FL , with the latter location resulting in the higher transmission potential . In contrast , Ae . albopictus from Alachua Co . had higher BVIC viral dissemination than individuals from Manatee or Indian River/St . Lucie Counties ( Fig 6 ) . However , this effect did not correspond to similar changes in transmission . Aedes aegypti from the Dominican Republic and the laboratory strain had lower transmission potential than recent colonies from Florida ( Fig 7 ) . Taken together , these observations suggest complex interactions between mosquito and CHIKV genotypes . Aedes aegypti originating from the Dominican Republic had viral dissemination and transmission potential rates for IOC and AC that were lower than Florida vectors ( Figs 2 and 7 ) . The fact that a large outbreak of CHIKV occurred in the Dominican Republic indicates that these lower rates are still sufficiently high to sustain transmission in nature and suggests that other factors largely contribute to transmission [60 , 67 , 68] , such as biting rates of humans by the vectors . To our knowledge there are no entomological surveys in Florida during the 2014 outbreak that would incriminate potential mosquito species as being infected with CHIKV . However , local infections in 2014 occurred in the ranges of both potential vector species in Florida . Our experiments demonstrated variation in dissemination and transmission among mosquito populations and virus strains , however , in some instances sample sizes were low limiting our ability to detect differences . Logistic constraints limited the number of time points we were able to sample , affecting precision in estimating minimum EIP and viral dynamics in the mosquitoes . With these constraints , however , we were still able to show geographic variation in vector competence . As similar variation may occur in other mosquito species , our results highlight the need for detailed investigations of vector competence across species and populations in different regions . Although geographic differences in the vector competence described in the current study may modulate local risks of infection and transmission , other components of vectorial capacity , such as vector survivorship and human biting rate , are likely to be more important determinants of the potentials for epidemic or endemic transmission [52] . Additionally , biting rate is probably enhanced by the abundance of breeding sites associated with water storage and irregular trash collection , suggesting that source reduction may play an important role in reducing transmission risk .
|
The emergence of mosquito-borne chikungunya virus in the Americas starting in 2013 has been associated with geographically widespread outbreaks of human illness . Transmission of chikungunya virus in the U . S . is a major public health risk , especially in Florida where the environmental conditions are favorable for the two main mosquitoes involved in transmission . We measured susceptibility to infection and transmission for Florida Aedes aegypti and Aedes albopictus mosquitoes for two emergent strains of chikungunya virus ( Indian Ocean and Asian strains ) . Both mosquito species showed high susceptibility to infection and rapid spread of the virus throughout the body of the mosquito , including the saliva for both emergent strains of chikungunya virus . Aedes albopictus had higher body infection and transmission of the Indian Ocean strain sooner after feeding on chikungunya virus infected blood than Ae . aegypti . Aedes aegypti had higher body infection and saliva infection later during infection with the Asian strain of chikungunya virus than Ae . albopictus . We also observed declines in body infection and transmission over time , suggesting that transmission risk declines with length of infection . The information here will be useful as parameters in models of risk of chikungunya virus transmission .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"dominican",
"republic",
"united",
"states",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"legs",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"chikungunya",
"infection",
"pathogens",
"tropical",
"diseases",
"microbiology",
"geographical",
"locations",
"limbs",
"(anatomy)",
"animals",
"alphaviruses",
"viruses",
"north",
"america",
"chikungunya",
"virus",
"rna",
"viruses",
"neglected",
"tropical",
"diseases",
"caribbean",
"insect",
"vectors",
"florida",
"infectious",
"diseases",
"musculoskeletal",
"system",
"aedes",
"aegypti",
"medical",
"microbiology",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"arthropoda",
"people",
"and",
"places",
"mosquitoes",
"blood",
"anatomy",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"species",
"interactions",
"organisms"
] |
2017
|
Transmission risk of two chikungunya lineages by invasive mosquito vectors from Florida and the Dominican Republic
|
To get beyond the “low-hanging fruits” so far identified by genome-wide association ( GWA ) studies , new methods must be developed in order to discover the numerous remaining genes that estimates of heritability indicate should be contributing to complex human phenotypes , such as obesity . Here we describe a novel integrative method for complex disease gene identification utilizing both genome-wide transcript profiling of adipose tissue samples and consequent analysis of genome-wide association data generated in large SNP scans . We infer causality of genes with obesity by employing a unique set of monozygotic twin pairs discordant for BMI ( n = 13 pairs , age 24–28 years , 15 . 4 kg mean weight difference ) and contrast the transcript profiles with those from a larger sample of non-related adult individuals ( N = 77 ) . Using this approach , we were able to identify 27 genes with possibly causal roles in determining the degree of human adiposity . Testing for association of SNP variants in these 27 genes in the population samples of the large ENGAGE consortium ( N = 21 , 000 ) revealed a significant deviation of P-values from the expected ( P = 4×10−4 ) . A total of 13 genes contained SNPs nominally associated with BMI . The top finding was blood coagulation factor F13A1 identified as a novel obesity gene also replicated in a second GWA set of ∼2 , 000 individuals . This study presents a new approach to utilizing gene expression studies for informing choice of candidate genes for complex human phenotypes , such as obesity .
In the current age of genome-wide association ( GWA ) studies using hundreds of thousands of single nucleotide polymorphisms ( SNPs ) , obesity has been a popular phenotype to investigate . Obesity , often measured as the body mass index ( BMI ) and defined as BMI≥30 kg/m2 , is important not only because of high and increasing prevalence , its strong association to many clinical complications such as type 2 diabetes , but also for the relative ease and availability of phenotypic measurement in most studies of human health . The heritability of BMI in twin and adoption studies range from 45%–85% [1] and still substantial , though somewhat lower in family studies . There are rare Mendelian syndromes with associated obesity and insights from rodent models and many linkage studies of obesity have identified chromosomal regions possibly harboring obesity genes on all chromosomes except the Y chromosome [2] . However , the known mutations account for only a small fraction of severe , early onset obesity and virtually none of the variability in adult overweight and obesity , while linkage studies and candidate gene studies have been plagued by a lack of replication [2] . Even meta-analyses of linkage have failed to reveal new genes . The advent of GWA studies of BMI has lead to novel , robust findings . However , the genes identified so far have been able to explain only a small fraction of the variation in BMI; FTO , the first gene unequivocally associated with BMI was estimated to explain ∼1% of the total variance in the discovery sample set [3] . That study involved the genotyping of more than 30 , 000 individuals and the identification of the next most convincing , common BMI associated gene MC4R required the combined analysis of more than 77 , 000 individuals [4] . The most recent results from the GIANT Consortium with more than 91 , 000 analyzed DNA samples were able to identify an additional six new loci for BMI , together explaining 0 . 40% of the total variance in BMI [5] . Significantly , these six new loci in conjunction with the known associations at FTO and MC4R only accounted for 0 . 84% of the total variance [5] , an observation that highlights the “winner's curse” phenomenon whereby effect sizes of associated variants are often exaggerated in the initial study relative to follow-up studies [6] . A simultaneous publication reported similar findings in a study consisting largely of the Icelandic DeCode data set . There , four loci identified in the GIANT consortium study and seven new loci influencing BMI were uncovered ( each with explanatory power comparable to the variants from the GIANT consortium study ) [7] . Even if only a fraction of the observed heritability of BMI consists of genes with universal effects in most , if not all populations , the remaining obesity predisposing genes surely number in the hundreds , if not thousands . However , their individual contributions are expected to be increasingly more modest and as such , more difficult to identify by ( now ) conventional means of GWA . The uncovering of these remaining genes will either call for unrealistically large sample sizes , or rather , novel methods for nominating potential candidate genes likely to be causally involved with the trait of interest . The integration of genome-wide expression data from relevant tissues with SNP scans is a promising approach for identifying novel genes involved in the development of complex disease traits such as obesity . Microarray data and SNP scans from individuals have been combined in order to map variants ( eQTLs ) that control the expression of nearby genes in segregating populations [8] , [9] . While these studies have identified large numbers of cis-acting variants and a smaller number of trans-acting variants , they are not powered to link the changes in gene expression with the development of complex diseases and their trait components . Correlation of gene expression with a phenotype such as BMI however does not imply causation and the difficulty comes in discerning the transcripts that merely react to the disease state in target tissues from those that are actually related to causal processes . This is an especially challenging task in human studies , where the system cannot be perturbed in the same sense as is possible with animal models or cell lines . In order to circumvent some of these challenges and to infer causality , we studied adipose tissue gene expression in a unique collection of young monozygotic twin pairs without other co-morbidities discordant for BMI [n = 13 pairs , age 24–28 years , 15 . 4 kg mean weight difference , with no significant height differences ( <3 cm ) ] as well as in a set of 77 unrelated individuals in order to establish a set of genes most likely to be related to causal processes in controlling the degree of human adiposity . Given that monozygotic twins are identical at the level of DNA sequence ( precluding somatic mutations ) , differences in the expression of the genes encoded therein are the result of regulatory and/or epigenetic changes in response to lifestyle and the environment . By this logic , the genes with significant expression differences in the adipose tissue between the lean and obese monozygotic co-twins in discordant pairs can be designated as reactive to the obesity . In contrast , genes whose transcript levels in adipose tissue correlate significantly with BMI in a large sample of non-related individuals will be a mixture of reactive genes as well as genes related to causal processes . While it would be expected that there would be a significant degree of overlap in the list of genes identified in the two different sample sets , the subset of genes that correlate significantly with BMI only in the non-related sample would present as candidates for primarily being related to causal processes . Genes identified as possibly “causal” to obesity can subsequently be interrogated for association with BMI in a large GWA study cohort in order to establish whether the genes harbor sequence variants that effect their expression in cis . This study design is presented in Figure 1 .
The precise effects of acquired obesity have been difficult to investigate due to the complex interaction of genes and the environment that are thought to be involved in its development . The correlation of BMI as measured between pairs of genetically identical twins representative of the population is very high , with a correlation coefficient of 0 . 79 in the FinnTwin16 study from which the twins here analyzed were recruited [10] . It is the rare cases of significantly discordant twin pairs that allows for the most detailed analysis of the environmental effects , in this case of acquired obesity . With the MZ co-twin control design applied here , we were able to control for gender , age , cohort effects as well as other exposures and experiences of childhood that are shared by siblings in a family . Thus , we carefully assess the effects that obesity has on the transcription of genes in adipose tissue . A thorough description of these 13 pairs of MZ twins ( age 24–28 years , eight male pairs and six female pairs ) most discordant for BMI has been published earlier by our group [11] . Briefly , the selected pairs represent 5% extreme ends in divergence of phenotypes and the obese co-twins were on average 15 . 4 kg ( 22% ) heavier than the lean co-twin . They exhibited a significantly larger fat mass in every depot measured , including subcutaneous , intra-abdominal and liver fat ( P<0 . 002 ) . Consequently the obese twins were also much more insulin resistant , as evidenced by the lower M-value and higher fasting serum insulin levels . Obesity in the heavier co-twins had developed after post-pubertal adolescence [12] and thus the metabolic abnormalities represent consequences of early stages of excess adiposity ( Table S1 ) . The sample of unrelated individuals used consisted of 77 individuals from the same Finnish population as the MZ twins . Subjects had been recruited as a part of a large study on familial dyslipidemias and consisted of 41 women and 36 men . Phenotypic measures ( ±1S . D . ) were as follows: BMI 26 . 6±4 . 1 kg/m2 ( range 20 . 1–37 . 8 kg/m2 ) , age at adipose tissue biopsy 51±11 . 6 years ( range 27 . 4–68 . 9 years ) . They thus exhibit a wide and representative range of BMI . First , in order to establish the list of genes most likely to be “reactive” to the obese state , samples of subcutaneous adipose tissue from the discordant MZ twins were analyzed . Following co-twin normalization of the expression data , we identified 2 , 674 transcripts differentially regulated in the adipose tissue of the obese twins compared to that of the lean co-twin ( non-parametric Welch t-test , using a liberal test value , i . e . P<0 . 05 ) . Next , in a data set consisting of adipose tissue samples from 77 unrelated individuals , using stringent criteria ( Pearson correlation P<10−4 ) we identified 84 transcripts that correlated with BMI . Of these , 56 ( or 2/3 ) were shared with those identified in MZ study samples and designated as “reactive” . The remaining 28 transcripts encoded by 27 unique genes that , while strongly correlated with BMI , were not differentially regulated in the discordant MZ twins and as such were candidates for being related to causal processes ( Table 1 ) . The P-value for the Pearson correlation between BMI and gene expression for these genes in the unrelated sample ranged from 2 . 5×10−6 for HADHA , to 9 . 99×10−5 for SAP30BP . By definition , these genes were not differentially expressed between the obese and lean co-twins ( P-values 0 . 35 and 0 . 56 respectively ) . The P-values for correlation of expression level with BMI among the twins ranged from 0 . 051 to 0 . 849 . A number of the known obesity associated genes were not included in the analyses , as they were filtered out either as reactive ( Leptin ) , showed no significant correlation with BMI in either sample ( FTO ) , or were not expressed in fat ( MC4R ) . In order to assess , whether any of the genes identified as putatively causative in influencing obesity also harbored SNP variants associated with the trait , we employed the large ENGAGE consortium study sample consisting of more than 21 , 000 GWA scans from well characterized European population cohorts , not ascertained for any specific trait . An analysis of the 197 SNPs , -representing the 27 putatively “causative” genes revealed a significant excess of small P-values , differing from the expected uniform distribution ( X2 P-value 4×10−4 ) . By contrast , the 498 SNPs representing the 48 genes “reactive” to obesity did no better than expected by chance ( X2 P-value 0 . 59 ) , as witnessed by the Q-Q plots with very different profiles for the two gene sets ( Figure 2 ) . It is possible for such Q-Q plots to be inflated due to a large number of associating SNPs in one or a few genes ( with high local LD ) , but re-drawing these plots while retaining only the best single SNP per gene did not significantly change their profiles ( Figure S1 ) . Out of the 27 tested putatively causative genes , 13 genes harbored a total of 23 SNPs that were nominally associated with BMI in our sample ( P<0 . 05 ) ( Table 2 ) . The top hit was SNP rs2274393 ( P = 0 . 003 ) located in an intron of the 177 kb gene encoding the A1 subunit of the coagulation factor XIII ( F13A1 ) , a gene previously associated with the risk of venous thromboembolism [13] . Altogether 7 SNPs in the F13A1 gene associated with BMI in the 21 , 000 ENGAGE samples ( Figure S2 ) . Among the other genes for which positive evidence of association ( P<0 . 05 ) could be shown were the sialyltransferase ST3GAL6 , ribonuclease 4 ( RNASE4 ) , peroxiredoxin 6 ( PRDX6 ) , SH3 domain binding glutamic acid-rich protein ( SH3BGR ) , the influenza virus NS1A binding protein ( IVNS2ABP ) , the apoptosis antagonizing transcription factor ( AATF ) , transmembrane protein 101 ( TMEM101 ) as well as some novel transcripts with no known functions . While the potential “reactive” genes did not deviate significantly from the expected uniform distribution , the top 4 SNPs were of some interest , as they were all located in the URB gene ( upregulated in bombesin receptor subtype 3 knockout mouse ) , recently implicated as having a role in human obesity [14] . The best P-value was obtained for intronic SNP rs9870432 ( P = 0 . 001 ) ( Table 3 ) . The association signal for URB was coming solely from the women in the analysis , suggesting a possible gender-specific role for this gene in influencing obesity and BMI . In order to assess whether any of the associations observed in the large ENGAGE sample were replicable , we utilized a smaller available GWA study sample , namely the GenMets study - a case/control study with ∼2 , 000 individuals designed to investigate the genetics of metabolic syndrome . The top-hit among the putatively causative genes was again a SNP in the F13A1 gene ( rs714408 P = 0 . 002 ) ( Table S2 ) . A meta-analysis of the two combined study samples ( N∼23 , 000 ) increased the significance of this F13A1 SNP to P = 0 . 0004 ) . In this replication dataset the effect size of this rs714408 variant on BMI was −0 . 2519 ( ±0 . 1472 ) BMI units per G allele carried . This equates to 0 . 14% of the variance in BMI . No other genes implicated in the ENGAGE study replicated in this smaller GenMets study sample , but the Q-Q plots bore the same distinct differences as in Figure 2 with the putatively causal genes outperforming the reactive genes ( Figure S3 ) . Among the reactive genes , the URB variants associated with BMI in the ENGAGE sample did not replicate in the GenMets study ( best P = 0 . 46 ) .
While BMI is a highly heritable and much investigated trait in humans , the success in identifying common predisposing genetic variants has been relatively modest with until very recently , only two solidly established gene findings to date , namely the FTO [3] and MC4R [4] genes , each explaining only a small fraction of the variance . The newest study identifying 6 additional obesity predisposing variants in combination with FTO and MC4R explained only 0 . 84% of the variance in BMI . While highlighting the complex genetic and environmental background of obesity , this realization also calls for novel designs in order to identify the remaining genes , each arguably with increasingly more modest effect sizes , or having characteristics that are not captured by the current generation of SNP-chips . Utilizing a unique experiment of nature , namely monozygotic twin pairs discordant for obesity and contrasting the obese/lean co-twin differences in expression profiles in adipose tissue with those from a collection of unrelated individuals , we were able to circumvent some of the difficulties experienced by other studies and infer putative causality of the observed expression changes in relation to obesity . Testing for association in a large collection of GWA data from European population cohorts ( N = 21 , 000 ) we were able to illustrate the power of this data mining scheme in uncovering novel genetic variants associated with complex human traits like BMI . In addition to genetic variation , the expression levels of genes , as measured by appropriate microarrays , can be affected by several factors including environmental variation , epigenetic modifications as well as by experimental conditions of in vitro cell lines . Using monozygotic twin pairs discordant for obesity in our experiment allowed for controlling for many of these factors , like gender , age , early developmental environment as well as cohort-effects . Being able to designate the differentially expressed transcripts as reactive to the obesity allowed for the identification in the unrelated sample set of transcripts with a putatively causative role in effecting human obesity . SNP variants in the genes designated as “causative” based on the MZ data performed considerably better in the test for association with BMI than the variants in the reactive genes , as evident in Figure 2 . A total of seven BMI associated SNP variants , including the top SNP of the study were located in the F13A1 gene . Multiple F13A1 SNPs were also associated with BMI in the replication GWA cohort , with the strongest evidence for association in the final meta-analysis for SNP rs714408 ( P = 0 . 0004 ) . In our data set , F13A1 transcript levels were also significantly associated with BMI ( Pearson correlation 0 . 41 , P-value 9 . 6×10−6 ) among the samples from unrelated individuals . F13A1 , coded for by a 177 kb gene in chromosome 6p25 . 1 is a subunit of coagulation factor XIII , the last zymogen to become activated in the blood coagulation cascade and responsible for forming the crosslinks between fibrin molecules that stabilize a clot [15] . A common V34L ( rs5985 ) variant in F13A1 has previously been associated with venous thromboembolism , and supported by a recent large meta-analysis [13] . The V34L variant was genotyped in this study , but failed to associate with BMI directly in the ENGAGE set ( P = 0 . 70 ) . In light of this , it does not seem like the V34L polymorphisms is the one underlying the BMI association observed here , neither do we believe to have captured the functional F13A1 variant itself . The linkage disequilibrium between the V34L polymorphism and the rs714408 variant here identified is R2 = 0 . 142 . The polymorphism ( s ) may influence F13A1 in several ways by either modulating its expression in the cells , by effecting its stability , or by changing its activity in the circulation , as has been shown to be the mode of action of the V34L polymorphism [16] . Further genotyping , sequencing and functional analyses are required to elucidate the functionality of the polymorphisms here identified . Aside from a report describing a correlation of F13A1 expression with liver fat in humans [17] , this is to our knowledge the first instance where F13A1 has been implicated in obesity . Given that obesity is well known to predispose to the development of deep vein thromboses and it is considered a chronic prothrombic state [16] , these observations may open new avenues for studying the link between obesity and thrombosis . This may also help account for the increased risk of cardiovascular disease in obese subjects . Circumstantial evidence from a number of other studies lends support for the possible role in obesity of several other genes identified in the present study . The second best association with BMI was for SNP rs865474 located in the gene ST3GAL6 , coding for a sialyltransferase . While no known link with obesity exists for this gene , a previous study found its expression to be highly variable and much more highly correlated between related individuals in lymphoblastoid cell lines , suggesting that the expression of this gene is under strong genetic control [18] . Our results in the MZ twins agree with this finding and given the strong correlation of the transcript with obesity in the unrelated individuals ( Pearson correlation −0 . 47 , P = 2 . 48×10−5 ) , this makes ST3GAL6 a potentially interesting new candidate . The third associating gene RNASE4 is a ribonuclease previously been shown to be upregulated in adipose tissue of women following a low calorie diet [19] and to be markedly increased after a 24 hour treatment with cortisol [20] . In mice RNASE4 has been shown to be regulated in the liver by dietary fat [21] . PRDX6 , an antioxidant enzyme with Ca-independent phospholipase A2 activity [22] , was identified in a recent study as a possible candidate gene underlying a novel obesity locus on chromosome 1q24 in an isolated population of Cilento [23] . The expression of SH3BGR was previously reported to be downregulated in adipose tissue following weight reduction [24] , a finding consistent with the positive correlation of transcript levels with BMI identified here ( Pearson correlation 0 . 32 , P = 5 . 0×10−5 ) . An interesting candidate for obesity is IVNS1ABP ( influenza virus NS1A binding protein ) . While it has no known function with a relation to obesity , a recent study of Oceanian population genetics utilizing genome-wide SNP analyses identified the IVNS1ABP gene as having undergone a selective sweep in the Oceanian population [25] . Among other such genes identified in the study were the VLDL-receptor and other lipid metabolism related genes , suggesting that IVNS1ABP may represent just the kind of obesity predisposing thrifty genes that have been hypothesized to be enriched in the Oceanian populations with a well known current epidemic of obesity and type 2 diabetes in conjunction with the westernization of their diets and lifestyle . Another very plausible candidate identified is the neuropeptide Y receptor 1 ( NPY1R ) , a receptor for a neuropeptide exhibiting a diverse range of physiologic activities including effects on food intake [26] . While a whole number of genes had been filtered out from the analyses as being clearly reacting to the obesity , there is no reason to believe that they too could not contain variants that effect the development of obesity . Genes that are designated as “reactive” , but still associate with obesity may be ones whose transcript levels are affected by the developing obesity but that concurrently harbour variants predisposing an individual to weight gain . In fact , the top result ( top 4 SNPs ) among the genes identified as reactive , was URB , a gene linked with obesity in mice [27] and recently implicated as having a role in obesity in humans [14] . Our results here support a gender specific role for SNP variants in URB in the development of obesity . A few of the transcripts designated as “causative” also approach nominal significance in regards to correlation with BMI ( Table 1 ) , supporting the notion that some genes could eventually be classified as both “causative” and “reactive” . The power of this study lies in the availability of the obesity discordant monozygotic twins that enable the dissection of cause and effect relationships between gene expression and obesity . A similar approach could be utilized for many other complex traits where discordant MZ twins and relevant tissue biopsies are available , for example muscle biopsies from pairs discordant for physical activity . Such collections of twins are admittedly rare , but other approaches ( animal clones exposed to different environments , human subjects that have gained or lost significant amounts of weight ) can approximate the experimental set up applied here . The most recent GWA studies on obesity give considerable hope that additional variants , each explaining in the range of 0 . 1% of the variance will be identified in new , large scale GWA studies utilizing tens of thousands of samples [5] , [7] . However , novel integrative methods as described here are called for in order to uncover the remaining variants . Such alternative designs , combined with comprehensive sequencing of candidate genes and loci , should give us a clearer picture of the allelic architecture and help to discern the relative contributions of rare and common variants to human obesity .
Adipose tissue biopsies from two study samples originating from the Finnish population were used . Both studies were performed according to the principles of the Helsinki Declaration . The twins from discordant pairs were recruited from a population-based longitudinal study of five consecutive birth cohorts ( 1975–1979 ) of twins , their siblings and parents ( n = 2 , 453 families ) , identified through the national population registry of Finland [28] . The twins selected for this study represented the top 5% most obesity-discordant MZ twin pairs ( one co-twin not obese [BMI∼25 kg/m2] , and the other one obese [BMI∼30 kg/m2] ) , with no significant height differences ( <3 cm ) or chronic illness . Of the 18 pairs thus identified , 13 pairs participated in the adipose tissue biopsy . The study subjects have been previously thoroughly described [11] . The protocol was approved by the Ethical Committee of the Helsinki University Central Hospital . The unrelated study subjects were recruited as a part of the European Multicenter Study on Familial Dyslipidemias in patients with Premature Coronary Heart Disease ( EUFAM ) and consisted of both healthy and affected individuals . The collection of subjects has been described earlier [29] , [30] . The Ethical Committee of the Department of Medicine , Helsinki University Central Hospital approved the study . All fat biopsies were collected under local anesthetic by a needle aspiration of periumbilical subcutaneous fat , immediately frozen in liquid nitrogen and stored at −80°C for later extraction of total RNA using the RNeasy Lipid Mini Kit ( Qiagen ) according to manufacturer's instructions . Quality of RNA was analyzed using the 2100 Bioanalyzer platform ( Agilent Technologies ) . Two micrograms of total RNA were treated according to conventional Affymetrix eukaryotic RNA labeling protocols ( Affymetrix , Santa Clara , CA ) . Fifteen micrograms of biotin labeled cRNA was fragmented according to Affymetrix eukaryotic sample protocol . Hybridization , staining and washing of the Affymetrix U133 Plus 2 . 0 chips were performed using the Affymetrics Fluidics Station 450 and Hybridization oven 640 under standard conditions . Prior to analysis , raw expression data were normalized using the GCRMA algorithm , separately for the MZ twin and non-related individuals . Analysis of the adipose tissue expression data was done using the GeneSpring GX 7 . 3 software ( Agilent Technologies ) . In order to include in the expression analyses only those probes with reliably high signal , the expression data was first filtered according to expression level so that only probes with expression value≥50 in at least ½ of the samples were included . Of the ∼54 , 000 probesets available on the array 17 , 026 ( ∼31% ) passed the filtering and were considered for analysis . Given the unique set-up involving MZ twins , following standard GC-RMA normalization of the expression signals , we next performed a “co-twin normalization” procedure which involved dividing the obese twin's expression values with those of the non-obese co-twin in order to correct for the identical genetic background ( and by definition also for gender and age ) . Using a liberal threshold for significance ( non-parametric Welch t-test P<0 . 05 ) we identified 2 , 674 transcripts differentially regulated in the adipose tissue of the obese twins . Next , in a data set consisting of adipose tissue samples from 77 unrelated individuals , using stringent criteria ( Pearson correlation P<10−4 ) we identified 84 BMI correlated transcripts . Of these , 56 were shared with those identified in MZ study samples and designated as “reactive” . The remaining 28 transcripts encoded 27 unique genes that , while strongly correlated with BMI , were not differentially regulated in the discordant MZ twins and as such were candidates for being related to causal processes . For the twin samples a liberal threshold was chosen in order to avoid unintentionally designating them as “causative” . This , contrasting with the stringent criteria for significance applied for the sample of unrelated individuals was done in order to avoid false-positives in the last stage of filtering . For the genes identified through the expression profiling as being either “causative” or “reactive” , we retrieved all the genotyped SNPs located in the coding sequence positions ±1 kb , as annotated in the UCSC genome browser ( March 2006 build ) . Altogether 197 SNPs in these genes were tested for association with BMI in the large set of ENGAGE cross-sectional population cohorts from populations of European origin ( N = 21 , 000 ) . All available SNPs were similarly retrieved for the 49 genes ( represented by 56 Affymetrix probes ) that were designated as “reactive” . One of the genes was X-chromosomal for which no genotypes were available . This resulting set of 48 genes was used as the control gene set against which we compared the association findings with the genes designated putatively causative . Within each of the 16 ENGAGE cohorts logarithmized BMI measures were stratified by sex and in each of the stratum outcomes were adjusted for age using linear regression model . The residuals resulting from these models were then standardized and used as outcome measures in the association analyses . Association analyses were performed with linear regression assuming additive mode of inheritance . These cohort specific results were combined into a fixed effects meta-analysis with reciprocal weighting on the square of standard errors of the effect size estimates . As a replication study sample we utilized the GenMets study collected from the Finnish population for an investigation of the genetics of metabolic syndrome . GenMets individuals were sampled from a Finnish population-based Health 2000 study . Details of the Health 2000 study and phenotype measurements have been reported earlier [31] . The sample consisted of 919 individuals with non-diabetic metabolic syndrome , and 932 individuals without the metabolic syndrome , as defined by the International Diabetes Federation [32] . The individuals were genotyped using Illumina HumanHap 610 K chip and imputed using the Hapmap 1+2 CEU reference data . GCRMA normalized gene expression values in the sample of unrelated individuals were evaluated for correlation with BMI using the Pearson correlation as implemented in SPSS 11 . 0 for Windows ( SPSS Inc . , Chicago , IL , USA ) . A P-value of <10−4 was considered significant . Following co-twin normalization , differential expression of transcripts between the BMI discordant MZ twin pairs was evaluated in Gene Spring GX 7 . 3 software ( Agilent Technologies ) by the non-parametric Welch t-test with P<0 . 05 considered significant for further analysis . The Pearson X2 test of categorized P-values was used to test for deviation of the distribution of observed P-values from the uniform distribution . In GenMets GWA analyses logarithmized BMI measures were adjusted for sex , age and metabolic syndrome affection status using linear regression model Metabolic syndrome status was used as a covariate in the linear model to minimize the effect of the study design . The main component behind the metabolic syndrome definition used in GenMets is waist circumference , although correlated with BMI , it may have a different genetic background from BMI . However , by adjusting we may run into over-adjusting and losing some statistical power . The residuals resulting from these models were standardized and used as outcomes as in ENGAGE analyses , Association analysis was conducted using PLINK [33] option linear .
|
Obesity has a strong genetic component and an estimated 45%–85% of the variation in adult relative weight is genetically determined . Many genes have recently been identified in genome-wide association studies . The individual effects of the identified genes , however , have been very modest , and their identification required very large sample sizes . New approaches are therefore needed to uncover further genetic variants that contribute to the development of obesity and related conditions . Much can be learned from studying the expression of genes in adipose tissue of obese and non-obese subjects , but it is very difficult to distinguish which genes' expression differences represent reactions to obesity from those related to causal processes . We studied monozygotic twin pairs discordant for obesity and contrasted the gene expression profiles of obese and lean co-twins ( controlling for genetic variation ) to those from unrelated individuals to try to discern the cause-and-effect relationships of the identified changes in gene expression in fat . Testing the identified genes in 21 , 000 individuals identified numerous new genes with possible roles in the development of obesity . Among the top findings was a gene involved in blood coagulation ( Factor XIIIA1 ) , possibly linking obesity with known complications including deep vein thrombosis , heart attack , and stroke .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"diabetes",
"and",
"endocrinology/obesity",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/gene",
"expression"
] |
2010
|
Use of Genome-Wide Expression Data to Mine the “Gray Zone” of GWA Studies Leads to Novel Candidate Obesity Genes
|
Why most of the in vivo experiments do not find the 30-nm chromatin fiber , well studied in vitro , is a puzzle . Two basic physical inputs that are crucial for understanding the structure of the 30-nm fiber are the stiffness of the linker DNA and the relative orientations of the DNA entering/exiting nucleosomes . Based on these inputs we simulate chromatin structure and show that the presence of non-histone proteins , which bind and locally bend linker DNA , destroys any regular higher order structures ( e . g . , zig-zag ) . Accounting for the bending geometry of proteins like nhp6 and HMG-B , our theory predicts phase-diagram for the chromatin structure as a function of DNA-bending non-histone protein density and mean linker DNA length . For a wide range of linker lengths , we show that as we vary one parameter , that is , the fraction of bent linker region due to non-histone proteins , the steady-state structure will show a transition from zig-zag to an irregular structure—a structure that is reminiscent of what is observed in experiments recently . Our theory can explain the recent in vivo observation of irregular chromatin having co-existence of finite fraction of the next-neighbor ( i + 2 ) and neighbor ( i + 1 ) nucleosome interactions .
In cells , DNA , with the help of a large number of proteins , is packaged into a higher order structure known as chromatin . In eukaryotes , in its simplest level of packaging , DNA is wrapped around histone proteins to form a “beads on a string” form of chromatin having a width ≈ 10 nm . The ‘bead’ here is a nucleosome—147 bp of DNA wrapped around an octamer-complex of histone proteins [1] . When the cells are preparing to divide , one observes that the DNA , with the help of many proteins , assumes a highly condensed structure . It is being thought that the 10-nm string of chromatin is further packaged in a hierarchical manner to produce this highly compact mitotic chromosome [1] . What this hierarchy of structures , if any , is and exactly how a DNA chain gets packaged into this highly compact form are interesting open questions [2–4] . Based on theoretical arguments and a few ( mostly in vitro ) experiments , it was proposed that the 10-nm chromatin string will fold itself into a higher order structure having an approximate width of 30-nm [5–7] . Some of the earliest theoretical models argued that the 30-nm chromatin should form a solenoid structure by coiling the known 10-nm fiber [5 , 6] . However , later studies argued that the 10-nm chromatin should rather be arranged in a zig-zag fashion [8–12] , which was subsequently supported by high resolution X-ray imaging of crystals made of tetra nucleosomes [13] and other in vitro biochemical studies of nucleosome arrays [14] . Since the zig-zag model is compatible with chromatin having a variety of linker DNA lengths , it became the most popular model for higher order structure , in the literature . However , contrary to all theoretical expectations and in vitro observations , experimental investigations could not prove the existence of 30-nm fibers in vivo [15–19] . Even though some experiments on chicken erythrocyte and starfish sperm chromatin , where non-histone proteins are very low , reported short-range zigzag structure [17 , 20] , detailed studies of chromatin in other nuclei , using Cryo-EM and synchrotron X-ray scattering methods , argued that the 30-nm structure is non-existent in vivo/in situ [21 , 22] . A set of small angle X-ray scattering experiments concluded that 30-nm fibers are non-existent in vivo , and they could not find any particular lengthscale beyond 10 nm [22] . Neither in the interphase nor in the mitotic phase could experiments find any of the expected higher order packaging [21–24] . Why the theoretically predicted , and in vitro-observed , 30-nm fiber is elusive in vivo is a puzzle in the field of chromatin biology . In this work , we address this puzzle of the 30-nm structure , based on basic biophysics principles involved in the packaging of chromatin . However , to get into the details , we need to understand the basic logic behind the proposed zig-zag structure of chromatin . We know a few facts: ( i ) DNA wraps around histone-octamer in nearly two turns such that the entry and exit segments of the DNA make a fairly restricted angle [25 , 26] ( see Fig 1 ) . ( ii ) Linker DNA lengths ( Ll ) are in the range 20–60bp . ( iii ) DNA is a highly stiff elastic chain in the length scale below persistence length ( Lp = 150bp ) [10 , 11 , 27] . Since typically Ll < Lp , the linker DNA will be rigid and straight . Given that nearly 2 turns around histone-octamer would bring back the DNA to approximately the same side of the nucleosome where it entered , the combination of the two factors—entry/exit angle and stiff linker region—would give rise to a preferred structure like a zig-zag ( see Fig 1b ) . There have been experimental studies that have indicated the occurrence of straight linker regions [10 , 11] and restricted entry/exit angles [25 , 26] in zig-zag chromatin . All other constituents ( ions , low-salt buffer conditions , linker histones ) can be thought of as agents influencing these two factors [28–33] . Even though the entry/exit angle and DNA rigidity have been the primary argument supporting zig-zag structure , this neglects a couple of facts: ( i ) in vivo , non-histone proteins can bind on the linker region and distort the orientation . For example , highly abundant high mobility group ( HMG ) proteins ( e . g . , nhp6 in yeast ) , and many other architectural transcription factors are known to bend DNA [34–38] . This bending by non-histone proteins can potentially affect the structure of chromatin . ( ii ) Since DNA has a nonlinear bending elasticity , it can have flexible hinges [39–43] . Given that DNA bending due to non-histone proteins is a certainty , it is important to consider them in any model that investigates the chromatin structure in vivo . Existing models primarily predict zig-zag or solenoid-like structures [44–46] . Recent models also predict polymorphic structures based on the variation in nucleosome repeat lengths [32 , 47] . However , none of the existing papers , to the best of our knowledge , have accounted for the effect of local DNA curvatures resulting from non-histone protein binding and/or the possibility of highly flexible regions due to other factors . Neither have these studies systematically investigated how the chromatin structure would appear when a large number of nucleosomes ( beyond 100 nucleosomes ) are present in the presence of DNA-bending proteins . In this work we study the higher order folding of nucleosome-bound DNA taking into account the possibility of non-histone proteins binding along the linker region , bending the DNA locally . Performing extensive simulations with two different models for chromatin , we show that ( i ) in the absence of non-histone protein binding , the nucleosome-bound DNA folds into a regular zig-zag structure; ( ii ) after the introduction of the non-histone proteins that bend linker DNA , the regular zig-zag structure starts disappearing . We compute the chromatin structure in the presence of specific proteins such as nhp6 and HMG . As a function of the density of bound non-histone proteins , we find that there is a transition from a zig-zag structure to an irregular higher order structure . We also investigate the influence of linker length on the formation of irregular structures .
In the first model ( Fig 1 ( c ) ) , DNA is considered as a polymer chain of N beads of type-1 ( small , yellow beads ) connected by linear springs of stiffness ks . Bending elasticity of DNA is introduced using the worm like chain model ( see details in the following text ) . The system we simulate has M nucleosomes where each nucleosome is created by wrapping 14 beads of DNA ( =147 bp ) in ≈ 1 . 7 turns ( 8 beads per wrap ) around a second type of bead ( histone core particle bead , blue , Fig 1 ( c ) ) . The DNA-nucleosome interaction is accounted for by connecting the 14 beads via another set of linear springs having stiffness kn . Linker histones ( H1 ) are modeled as another type of spring connecting DNA segments that enter/exit nucleosomes . Two nucleosomes , when they are closer than a prescribed cut-off distance , can interact with each other; in the model , this inter-nucleosome interactions are also approximated as spring forces between two histone core particles . An important addition in the model is the presence of DNA-bending non-histone proteins—we assume that non-histone proteins can bind in the linker regions and bend the DNA locally . To mimic this , we consider non-histone protein as a spring that will connect two DNA-beads in the middle of the linker region and bring them together ( see Fig 1 ( c ) , red ) ; the net result is bending of the linker DNA . As mentioned earlier , the stretching energy of DNA ( Us ) , the DNA histone interaction energy ( Un ) , the linker histone interaction energy ( Ul ) , the interaction energy of the non-histone protein ( Up ) , and the inter-nucleosome interaction energy ( Uh ) are approximated using a quadratic potential with different stiffnesses ( kα ) and equilibrium extensions ( rα ) as: U α = k α 2 ∑ i , j ( | r i ( β ) - r j ( γ ) | - r α ) 2 ( 1 ) where α ∈ {s , n , l , p , h} , β and γ are symbols indicating whether the bead is a DNA-bead ( r i ( 1 ) ) or a nucleosome bead ( r i ( 2 ) ) . ( i ) For DNA chain α = s , β = γ = 1 , j = i+1 and i ranges from 1 to N − 1 . ( ii ) For nucleosome core-DNA interaction α = n , β = 2 , γ = 1 , i ranges from 1 to M and j takes 14 values , for each i , representing the identity of the correspoding histone-bound DNA beads . Since nucleosomes are not dynamic in our study ( no sliding or disassembly of nucleosomes ) , we have also prepared the system , equivalently , replacing the nucleosome as an effective bead with U n = k n eff ∑ i 1 - cos ( α i - α n ) which constraints the DNA entering and exiting the nucleosome at an angle at αn ( see Fig 1 ( d ) ) [48] . This saves computational time and allows us to run BD simulations for a longer chromatin . ( iii ) For linker histone interaction α = l , β and γ represent the entry and exit DNA beads , respectively , j = i and i ranges from 1 to M − 1 . ( iv ) For non-histone protein binding at the linker region α = p , β and γ represent the identity of the location where non-histone proteins bind; i goes from 1 to ν , j = i + 3 , where ν is the total number non-histone proteins bound . Different non-histone proteins will have different rp values representing the bending angles they induce , as we discuss later in the manuscript ( also see Supporting Information ( SI ) S1 Text and S1 Fig ) . ( v ) For inter-nucleosome interaction α = h , β = γ = 2; at every instant i and j are computed such that the ith nucleosome can interact with the jth neighbor positioned below a certain cut-off distance = 9a . When two nucleosomes are beyond this cut off distance , the inter-nucleosome interaction energy of the pair is zero . Since inter-nucleosome interactions are dominated by histone tails , and since nucleosomes have only finite number of tails , we assume that each nucleosome interacts only with two nearby nucleosomes ( j ≤ 2 ) ( We have also done simulations accounting for more than two tails ) . To mimic the steric hindrance of DNA beads , the repulsive part of the standard Lennard-Jones ( LJ ) potential ( ULJ ) is used ( when | r i ( 1 ) - r j ( 1 ) | < 2 a ) such that U L J = ϵ ∑ i = 1 j > i N − 1 [ ( 2 a ) 12 ( | r i ( 1 ) − r j ( 1 ) | ) 12 − 2 ( 2 a ) 6 ( | r i ( 1 ) − r j ( 1 ) | ) 6 + 1 ] ( 2 ) where ϵ is potential well depth . The energy is zero when | r i ( 1 ) - r j ( 1 ) | ≥ 2 a The bending energy ( Ub ) of the DNA chain is given by U b = k b 2 a ∑ i = 1 N - 1 ( 1 - cos θ i ) ( 3 ) where kb is the bending stiffness of DNA and θi is the angle between two nearby bonds in the bead-spring model . The total energy of the chromatin in this model is given by Utot = Us + ULJ + Ub + Ul + Un + Up + Uh . Since expanding any potential energy close to its stable minimum gives us a quadratic function , we assumed a harmonic nature for most of our interactions . However , we have also done simulations with other functional forms ( e . g , Morse potential ) to ensure that the results are robust ( see S1 Text ) . The system , accounting for all potentials , is simulated using BD by solving the equation given by dri ( γ ) dt=−μ0 ( γ ) ∇ri ( γ ) Utot ( t ) +ξi ( γ ) ( t ) ( 4 ) where γ = 1 or 2 represents the identity of the beads , namely the DNA beads and nucleosome core particles , respectively . μ0 is the mobility , and ξi is the thermal noise experienced by each particle such that 〈ξi〉 = 0 and 〈ξi ( t ) ⋅ ξj ( t′ ) 〉 = 6kB Tμ0 δij δ ( t − t′ ) ( see [50] for details ) . Given an initial condition , we solve the equation taking a set of parameters as discussed in the following text . We have also done simulations with standard simulator tool LAMMPS to ensure robustness of our results [51] ( see S1 Text ) . From the simulations , we obtain the positions of all beads as a function of time ( r ( t ) ) ; when the system equilibrates we sample shapes of the chromatin and compute different quantities as discussed in the Results section . The basic length scale is taken as one helical repeat of DNA 10 . 5 bp , which is also the diameter of the DNA bead ( 2a ) and the equilibrium distance of the DNA spring ( rs ) . The bending stiffness of DNA is kb = 50 kBT nm where T = 300 K , and kB is the Boltzmann constant [32] . Since , within this model , we want to hold the interacting DNA nearest-neighbor beads , DNA-nucleosome beads , and DNA-protein spring unstretchable , we took the corresponding stiffness values to be very high: k s = k n = k p = 100 k B T / r s 2 . We have taken the linker histone interaction constant kl in the range 30 to 100 k B T / r s 2 and k n eff = 50 k B T . For the purpose of this work , the exact value of these stiffness parameters are irrelevant as long as they are high enough to reduce large fluctuations ( see S1 Table ) . Based on the literature and various geometrical constraints in the problem , we take rn = 8rs/π , rl = 2 . 5rs , and rh = 4 . 2rs [32] . We consider effects different specific non-histone proteins by varying value of rp . Since highly abundant proteins like nhp6 ( yeast ) and HMG-B are known to bend DNA making bend angles of ≈ 90° or ≈ 120° [52 , 53] , we chose values rp = 2 . 4rs and rp = 2rs such that the appropriate angles are formed ( see S1 Text for more information ) . The timescales and mobility parameters are taken as Δt = 0 . 04ns , μ 0 ( 1 ) = 2 × 10 - 4 r s / k B T Δ t , and μ 0 ( 2 ) = 1 . 5 × 10 - 4 r s / k B T Δ t respectively ( see S2 Table for details on non-dimensionalization of some of these parameters ) . We run the simulations for a time ( 0 . 04s ) which is much longer than the time it takes for the system to reach a steady state ( milliseconds ) having constant mean energy and radius of gyration . To understand how chromatin is structured in vivo , one needs to simulate a long chromatin having a large number of nucleosomes ( thousands of nucleosomes ) in a large parameter space . Since it is computationally intensive to do so using BD simulations ( typical BD simulations simulate <100 nucleosomes ) , a freely rotating chain model for the chromatin is employed . In the FRC model , we consider chromatin as a long 3D chain made of N tangent vectors . Orientations of the vectors are chosen such that the angle between the vector i and i − 1 is θi ( see Fig 1 ( e ) ) . If the ith vector is a DNA vector , it makes a relative angle of θ i = 4 a / L p with its neighbor [54] . As per the definition of FRC , the rotation angle ϕ is chosen randomly [54] . From a structural point of view , it can be argued that binding of the histone-complex ( or any protein ) constrains DNA subunits entering and exiting the histone octamer ( protein ) [26] . Therefore , if the ith bead in the FRC is a nucleosome , a relative angle of θ i = 2 π 3 is assigned between the two adjacent tangents [25] . The corresponding ϕi angle is chosen in a range of −10° to 10° randomly so as to satisfy the known zig-zag structure seen in earlier simulations . We also introduced the DNA-bending non-histone proteins in the FRC as follows . While constructing the FRC , at every linker region , one neighboring vector pair in the middle is chosen with a given probability ρ , and imposed a certain angle θp representing the binding activity of non-histone proteins . First we consider bending due to specific proteins like nhp6 and HMG-B ( θp ≈ 90° and 120° ) [52 , 53] and then we do a calculations considering a distribution of angles around these typical values . The angle ϕ here is chosen randomly between −10° and 10° . In this manner , we constructed a large number of chromatin configurations ( many realisations ) , each having 2000 nucleosomes . To quantify the structure of these configurations we calculated I ( k ) which is the probability that any nucleosome is in “contact” with its kth neighbor [32] ( see S2 Text for details ) . Two nucleosomes are defined to be in contact when the distance between them is below a cut-off distance of 16 nm ( 9a ) , which is the estimated mean length of histone tail interactions .
We investigated the role of specific DNA-bending proteins in determining the structure of chromatin using the above BD simulations . DNA-bending protein in yeast nhp6 is reported to bend DNA making angles of θp ≈ 120° [52 , 56] or ≈ 90° [52 , 57] . The HMG-B protein is reported to bend DNA making θp = 90° [53 , 58] . The proteins are typically known to have a size ( footprint on DNA ) of ≈ 20bp ( HMG-B , nhp6; bend involving 2 bonds ) or ≈ 30bp ( some proteins in the HMG family; bend involving 3 bonds ) [58 , 59] . In Fig 3 we present our results in the presence of nhp6 and HMG-B proteins bending DNA with angles θp = 90° and 120° . In the absence of any protein ( θp = 0 ) , I ( k ) peaks at k = 2 indicating zig-zag ( Fig 3 ( a ) ) . When we introduce nhp6 or HMG-B having DNA-bending angle of 90° , size = 20bp , we get a peak at k = 1 ( green curve ) . The peak at k = 1 becomes more dominant if the bending angle increases to 120° ( blue curve ) . For larger proteins ( size 30bp ) too we see that the zig-zag structure is destroyed ( Fig 3 ( b ) , peak is not at I ( 2 ) ) . All the above results are for a protein density of ρ = 0 . 5 per linker region , which is close to the known abundance of the DNA-bending non-histone proteins [35] . see S6 , S7 and S8 Figs for more angles and parameters . We also performed simulations with different types of inter-nucleosome interactions by varying the strength of the interactions and the nature of the potential ( see S3 , S4 and S9 Figs ) . To get an intuitive understanding of the above results , we performed simple 2D simulations using the FRC model as discussed in the text earlier . We simulated the structure of a chromatin having 24 nucleosomes , with uniform linker length of 42 bp , and the results ( XY position nucleosomes and linker DNA ) are plotted in Fig 4 . In the absence of non-histone proteins , the structure is a clear zig-zag ( Fig 4a ) . When we add non-histone proteins that bind along the linker region and bend the linker DNA , beyond a critical density of bound non-histone proteins , we completely lose the zig-zag nature and obtain a chromatin with irregular structure ( Fig 4b , ρ = 0 . 24 , θp = 120° with random orientations ) . This simple numerical study supports the hypothesis that the role of non-histone proteins is crucial in deciding the higher order structure of chromatin . So far we did simulations with a small number of nucleosomes ( M ≈ 20 , approximate length scale of a single gene ) . However , what will be the chromatin organisation in the longer length-scale ( length scale of many genes ) accounting for a large number of nucleosomes is an important question . Since performing BD for large systems is computationally expensive , and since both BD and FRC simulations give similar results , we implemented the 3D FRC model to probe large-scale organization of chromatin . Using the FRC model , we generated a large number of ( ≈ 107 ) equilibrium configurations of chromatin , each having 2000 nucleosomes , both with and without non-histone proteins . Systematically varying protein density ( ρ ) , we investigated the amount of non-histone proteins required for the appearance of an irregular structure . To compare the 3D FRC model with and without proteins and to quantify the resulting nucleosome organization in space , here too we computed I ( k ) ( Fig 5a ) . In the absence of non-histone proteins ( Fig 5a , blue ) , the peak is at k = 2 implying a zig-zag structure . As protein density ( ρ ) increases , the k = 2 becomes less probable , and the probability of finding k ≠ 2 increases . Here the bending angle of non-histone protein is chosen as θp = 90° representing proteins like HMG-B or nhp6 [52 , 53] . Results presented in Fig 5a suggest that , as a function of one parameter—non-histone protein density—there exists a transition from a zig-zag structure ( dominant peak is at k = 2 ) to an irregular structure ( dominant peak at k ≠ 2 ) . The plot of peak position ( the value of k at which I ( k ) peaks ) as a function of protein density ( Fig 5a , inset ) captures this transition and shows that the transition happens when the protein density is ≈0 . 48 . This is also the parameter regime where I ( 1 ) is comparable to I ( 2 ) —this is important since recent experiments have indicated that chromatin has an irregular structure , and at the same time there exist both i/i + 1 and i/i + 2 contacts [60] . Given that , in reality , there are proteins of different types binding in a heterogeneous manner ( e . g . LEF1: 107°-127° , nhp6: 120° ) we repeated the calculation for different set of bending angles—angles randomly chosen from a range between 90° and 135° [56 , 59] . The results are shown in Fig 5b . Here too the inset shows a transition from zig-zag to irregular structure as a function of protein density . In both the figures here , the chromatin becomes irregular when ρ < 0 . 5 ( at ρ = 0 . 48 in Fig 5a and ρ = 0 . 28 in Fig 5b ) . This is equivalent to having one non-histone protein for every four nucleosomes ( ≈ 1/0 . 28 ) or two nucleosomes ( ≈ 1/0 . 48 ) , consistent with abundance of non-histone proteins in the cell [35] . It is known that there exists roughly one HMG protein for every two nucleosomes [35] and hence such irregular structures can be expected for chromatin in which nearly all available HMGB proteins bind at linker regions . Even though the typical chromatin linker length ( Ll ) is rarely larger than the persistence length ( Lp ) , it has been argued that the linker length variation may affect the structure of chromatin [32 , 47 , 61 , 62] . We did FRC simulations for chromatin having different Ll values ( see S10 Fig ) . For large Ll , I ( 2 ) decreases considerably; however given that the largest biologically relevant Ll values ( ≈ 70 bp ) are much smaller than the Lp of DNA , I ( 2 ) still remains dominant . We find that as Ll increases the ρ required to create irregular structures becomes smaller . We systematically investigated how the two variables—non-histone protein density and linker length—decide the structure of the chromatin . The result is summarized in a phase diagram shown in Fig 5c . For every value of linker length , we have a “critical” protein density beyond which the chromatin structure is irregular . This prediction can be tested in vitro by varying nucleosome density and protein density appropriately . We also did simulations of chromatin having non-uniform linker lengths ( Ll taken from a distribution ) , and it also gave similar results ( see S3 Text and S11 Fig ) . We then computed how different types of proteins , having different bending angles , will affect the phase diagram ( see Fig 5d ) . We varied the bending angles in the range of 90°–135° ( see S12 Fig ) , since this is the range in which most of the relevant non-histone proteins bend DNA . As expected , the more the bending angle , the less the density of protein that is required to make the chromatin structure irregular . We studied a few other aspects as well: ( i ) Varied the angle induced by the nucleosomes ( αn ) and examined its effect ( see S13 Fig ) . ( ii ) Examined whether the zig-zag or non-histone protein-bound chromatin have a fractal nature [22 , 63 , 64] , and how non-histone proteins would affect packing ratio ( Rg ) ( see S4 Text and S14 Fig ) . ( iii ) Varied the histone tail interactions ( the number of interacting tails and strength ) and found how nucleosome contact map might alter in highly interacting regions ( heterochromatin ) vs loosely interacting regions ( euchromatin ) ( see S2 , S3 and S4 Figs ) .
Our predictions may be tested in a few different ways . One may perform an in vitro experiment reconstituting chromatin in the presence of DNA bending proteins [13 , 14 , 66 , 67] . The DNA-bending proteins are expected to destroy any regular zig-zag structure otherwise seen in the absence of non-histone proteins . Another way would be to perform Micro-C experiments ( like Hseig et al [60] ) by mutating DNA-bending proteins like nhp6 . The prediction is that , when the dominant DNA-bending proteins are mutated , the next neighbour contact probability value will increase . One may also map the positions of DNA-bending proteins such as HMG and correlate them with Micro-C ( Hi-C ) data so that one obtains a better picture of how the non-histone proteins influence the 3D chromatin structure . To conclude , in this work , we indicated the importance of non-histone proteins in determining the higher order structure and quantified the contribution of DNA-bending in generating irregular structures . We show that non-histone protein binding with biologically relevant proportions can create chromatin structures with equal fraction of next-neighbor ( i + 2 ) and neighbor ( i + 1 ) nucleosome interactions , as observed in recent experiments . Our work provides insights into understanding local chromatin structure ( in the length scale of a few genes ) , which is crucial for studying phenomena like histone-modification spreading , accessibility of regulatory regions , and gene regulation in general . We must also acknowledge the role of other factors: in reality , the structure of chromatin will be determined not just by the non-histone proteins but by an interplay between different factors such as crowding , linker length variations , concentration of various ions ( e . g Mg++ ) , DNA-bending proteins , and other constituents [30–32] . We hope that this work will trigger more computational and experimental studies in future to probe this aspect more quantitatively , and delineate the relative contributions of all these factors in great detail .
|
The fate of a cell is not just decided by the genetic code but also by the nature of the 3D organization of the protein-bound DNA , known as chromatin . Chromatin packaging is believed to be in a hierarchical manner , and one of the crucial stages in the packaging is argued to be having a zig-zag structure with specific width of 30 nm . However , most of the recent experiments failed to find any zig-zag-like ordered arrangement of chromatin in living cells . In this work , we address this puzzle , and argue that any regular , ordered , packaging of chromatin is unviable given that certain types of proteins can bind and bend the chromosomal DNA .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"protein",
"interactions",
"dna-binding",
"proteins",
"protein",
"structure",
"epigenetics",
"dna",
"dna",
"structure",
"chromatin",
"chromosome",
"biology",
"proteins",
"gene",
"expression",
"histones",
"molecular",
"biology",
"nucleosomes",
"biochemistry",
"biochemical",
"simulations",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"macromolecular",
"structure",
"analysis"
] |
2017
|
Binding of DNA-bending non-histone proteins destabilizes regular 30-nm chromatin structure
|
Mycobacterium ulcerans , the causative agent of Buruli ulcer in humans , is unique among the members of Mycobacterium genus due to the presence of the virulence determinant megaplasmid pMUM001 . This plasmid encodes multiple virulence-associated genes , including mup011 , which is an uncharacterized Ser/Thr protein kinase ( STPK ) PknQ . In this study , we have characterized PknQ and explored its interaction with MupFHA ( Mup018c ) , a FHA domain containing protein also encoded by pMUM001 . MupFHA was found to interact with PknQ and suppress its autophosphorylation . Subsequent protein-protein docking and molecular dynamic simulation analyses showed that this interaction involves the FHA domain of MupFHA and PknQ activation loop residues Ser170 and Thr174 . FHA domains are known to recognize phosphothreonine residues , and therefore , MupFHA may be acting as one of the few unusual FHA-domain having overlapping specificity . Additionally , we elucidated the PknQ-dependent regulation of MupDivIVA ( Mup012c ) , which is a DivIVA domain containing protein encoded by pMUM001 . MupDivIVA interacts with MupFHA and this interaction may also involve phospho-threonine/serine residues of MupDivIVA . Together , these results describe novel signaling mechanisms in M . ulcerans and show a three-way regulation of PknQ , MupFHA , and MupDivIVA . FHA domains have been considered to be only pThr specific and our results indicate a novel mechanism of pSer as well as pThr interaction exhibited by MupFHA . These results signify the need of further re-evaluating the FHA domain –pThr/pSer interaction model . MupFHA may serve as the ideal candidate for structural studies on this unique class of modular enzymes .
Buruli ulcer is a disease of skin and soft tissues caused by the bacteria Mycobacterium ulcerans [1] . It is the third most important mycobacterial disease after tuberculosis and leprosy [2] , and the prevalence continues to increase in tropical and sub-tropical countries [3] . M . ulcerans evolved from an Mycobacterium marinum ancestor through reductive evolution and acquired a large virulence determinant plasmid ( pMUM001 ) [4] . This plasmid encodes genes for mycolactone synthesis that are required to circumvent the host immune response , as a strain lacking this plasmid is avirulent . Therefore , the pMUM001 plasmid is considered to be a key determinant of M . ulcerans pathogenesis [5] , [6] . Pathogenic species of mycobacteria require stringent control on cell division for survival in their host , and thus are likely to acquire specialized mechanisms through evolution to achieve this control . Signaling proteins that sense environmental changes and mediate cell response are important for regulating cell division . For instance , bacterial Serine/Threonine Protein Kinases ( STPKs ) are known to regulate cell division by sensing and responding to specific signals in the host environment [7] . Moreover , according to phospho-proteome analysis , numerous Ser/Thr phosphorylated proteins have been identified in Mycobacterium tuberculosis , suggesting that STPKs may regulate multiple cellular processes [8]–[12] . Indeed , eleven STPKs have been identified in M . tuberculosis and the majority of them have been shown to be involved in pathogenesis and drug resistance [13]–[15] . ForkHead-Associated ( FHA ) domain containing proteins are the key interacting partners of STPKs that mediate the signals inside the cells emanating from the cognate kinases [16]–[18] . FHA domains are highly conserved modules , known to bind phosphorylated residues within the proteins involved in diverse processes in bacteria , such as protein secretion , antibiotic resistance , transcription , peptidoglycan synthesis , metabolism and virulence [17]–[20] . Most of the FHA domain containing proteins in M . tuberculosis are phosphorylated by STPKs [18] , and these proteins have been shown to recruit several other proteins in addition to being STPK substrates [17] , [18] , [21] . Based on the importance of STPKs in M . tuberculosis physiology and virulence , we explored the molecular transducers and their associated FHA domains in the M . ulcerans genome . We identified 13 STPK-encoding genes in silico and focused on a novel STPK- PknQ , encoded by the virulence-associated plasmid pMUM001 . We identified MupFHA ( Mup018c ) and MupDivIVA ( Mup012c ) as the substrates of PknQ and characterized the interaction of PknQ and MupDivIVA with MupFHA . MupFHA contains one FHA domain that interacts with PknQ , as well as with phosphorylated MupDivIVA . Importantly , we found that the interactions of these three M . ulcerans proteins encoded by the virulence-associated plasmid are dependent on phosphorylation of serine and threonine residues .
BLAST search was performed using the NCBI-BLASTp ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) with proteomes of M . ulcerans and M . marinum as target and desired protein sequence as query . The sequences of M . ulcerans STPKs were extracted using NCBI . These sequences were used for multiple sequence alignments , performed using ClustalW ( http://www . ebi . ac . uk/Tools/msa/clustalw2 ) . Phylogenetic analysis was performed using the sequences of STPKs in PHYLIP and phylogenetic tree was constructed [22] . The presence of conserved domains was detected by NCBI- conserved domain database ( http://www . ncbi . nlm . nih . gov/Structure/cdd/cdd . shtml ) . To generate the structure of PknQ , intracellular kinase domain of M . tuberculosis PknB ( PDB ID: 1MRU ) that shows 42% identity with the PknQ sequence , was chosen as template . We focused on the predicted catalytic domain of PknQ ( 1 to 280 aa ) . The three-dimensional models for PknQ were generated using MODELLER version 9v1 . General features were evaluated on the basis of the MODELLER's energy and DOPE scores while detailed reliability indices were obtained by the PROCHECK program . The PROSA Z-score was calculated using PROSA-II . Best models were chosen and further refinements were carried out . Disordered activation loop region was identified and loop refinement procedure was applied using the automatic loop refinement method provided in MODELLER 9v1 auto-model class protocol . In order to model the loop reliably , 500 loop models were sampled followed by model validation using DOPE scores and Verify 3D . PTM-Viena server was used to modify the protein and phosphate group was added at 2 different positions- Ser170 and Thr174 . Ab initio loop modeling was further used to minimize the post-translation modification effects . For this purpose , loop modeling protocol of standalone version of Rosetta 3 . 4 modeling suite was used and 50 outputs were taken [23] . On the basis of energy score and visual inspection of the loop we identified top 3 best loop models and used for further studies . Similarly , structural modeling of the FHA domain was performed with MODELLER 9v1 . MupFHA protein sequence spanning the residues 10 to 100 was modeled ( containing the FHA domain and the interaction motif ) . In order to use the maximum benefits of sequence identities and to model the maximum part of sequence contacting FHA domain , multi-template approach provided by MODELLER 9v1 was utilized and the two templates of Rv0020c FHA domain ( PDB ID: 2LC1 , 3PO8 ) were used to predict the three dimensional homology model of MupFHA . The 3D models for native FHA domain and mutants were generated , on the basis of previously generated structures of Rv0020c and pThr peptide [24] , [25] . The templates were pair-wise structurally aligned . Further models were built and validated as described for PknQ homology modeling . Protein-protein dockings were performed using HEX 6 . 3 molecular docking program , correlation type were chosen as shape and 2000 solutions were chosen for final evaluation . Other parameters were chosen as default . Structures of the wild type FHA domain and mutants were used for docking with PknQ containing pSer170 and pThr174 . First 10 solutions were analyzed for each docking and best docked poses were chosen . A short minimization was performed with Gromacs to minimize the complexes . Intermolecular interactions and docking scores were analyzed . Three complexes were prepared using molecular dynamics simulation- ( 1 ) a double phosphorylated PknQ-FHA complex , ( 2 ) mutant complex having single phosphorylation on Ser170 and mutation at pThr174 to alanine ( PknQ-pSer170/Thr174Ala ) , ( 3 ) phosphorylation at Thr174 and mutation at pSer170 to alanine ( PknQ-Ser170Ala/pThr174 ) . All the simulation studies were performed on GROMACS 4 . 5 . 5 using ffG43a1p force field provided by Gromacs official website ( http://www . gromacs . org/Downloads/User_contributions/Force_fields ) and uploaded by Graham Smith , which included extended parameters for modified residues [26] . The protein complex was dissolved at the center of a cubic box , solvated with single point charge water molecules . Solute-box distances of 1 . 0 nm were specified . To simulate the protein-protein complex system and in order to solve the issue of surface effects , periodic boundary conditions were applied . To neutralize the net charge on the protein , Na+ ions were added . A short energy minimization step was performed on the solvated system with steepest descent algorithm for 50000 steps . Two phases of equilibration were conducted for 100 ps each- under constant Number of particles , Volume and Temperature ( NVT ) followed by constant Number of particles , Pressure and Temperature ( NPT ) . Positional restraints were applied on the protein to allow solvent molecules to relax around the structure . In the second stage , positional restraints were lifted and the system was coupled to a heat bath at 300 K using the Berendsen thermostat and allowed to equilibrate for 100 ps [27] . In an NPT ensemble , the Parrinello-Rahman barostat was used for controlling the pressure [28] . Time constants for controlling the temperature and pressure were set to 0 . 1 ps and 2 ps , respectively . The production run was performed with suitable parameters for 10 ns at a temperature and pressure of 300 K and 1 atm , respectively . Coordinate sampling was performed at every 2 fs time interval . Bond-lengths were constrained using the Linear Constraint Solver ( LINCS ) algorithm [29] . Various utilities of GROMACS-4 . 5 . 5 package were used for detailed analysis of MD trajectories . All of the Gromacs MD simulations were run in the HPC-Supercomputer facility ( CSIR-IGIB , India ) on 32 Cores at 11 . 4 ns/day maximum performance speed . Escherichia coli strain DH5α ( Novagen ) was used for cloning and BL21 ( DE3 ) ( Stratagene ) for the expression of recombinant proteins . E . coli cells were grown and maintained with constant shaking ( 200 rpm ) at 37°C in LB broth supplemented with appropriate antibiotic ( 100 µg/ml ampicillin and/or 12 . 5 µg/ml chloramphenicol ) , when needed . For cloning of pknQfl ( mup011 , 1–660 amino acids ) , its catalytic kinase domain pknQkd ( 1–344 aa ) , mup018c ( 1–362 aa ) and mup012c ( 1–87 aa ) , the respective genes were amplified by PCR from M . ulcerans Agy99 plasmid pMUM001 using specific primers ( Table S1 ) . The resulting PCR products were cloned into the selected vectors ( pProEx-HTc , pGEX-5X-3 , pMAL-c2x and/or pACYCDuet-1 ) . Rv0020c was PCR amplified from M . tuberculosis H37Rv genomic DNA and cloned in pProEx-HTc vector . The plasmid derivatives were confirmed with restriction digestion and DNA sequencing ( Invitrogen ) ( Table S1 ) . M . tuberculosis PstP ( Rv0018c ) and Bacillus anthracis PrkD/PrkDS162A were cloned as described earlier [30] , [31] . To generate the specific site mutants ( Table S1 ) of PknQ , MupFHA and MupDivIVA , site directed mutagenesis was carried out using QuikChange XL Site-Directed Mutagenesis Kit ( Stratagene ) according to the manufacturer's protocol . The mutants were confirmed by DNA sequencing . The plasmids were transformed and proteins were over-expressed in E . coli BL21 ( DE3 ) . The recombinant GST-tagged fusion proteins were affinity purified with glutathione sepharose column . The His6-tagged proteins were purified by Ni2+-NTA affinity chromatography . For both purifications , similar procedures were followed as described before [12] , [32] . MBP-tagged MupDivIVA was purified using amylose resin as described before [9] . The purified proteins were resolved by SDS-PAGE and analyzed after staining with coomassie brilliant blue R250 . The concentration of purified proteins was estimated by Bradford assay ( Bio-Rad ) . In vitro kinase assays of PknQ or PknQK41M ( 1 µg each ) were carried out in kinase buffer ( 20 mM HEPES [pH 7 . 2] , 10 mM MgCl2 and 10 mM MnCl2 ) containing 2 µCi [γ-32P]ATP ( BRIT , Hyderabad , India ) followed by incubation at 25°C for 0–30 minutes , as described previously [30] , [33] . In all the reactions , kinase domain of PknQ was used ( PknQkd ) , unless specified . Myelin Basic protein ( MyBP ) ( 5 µg ) was used as an artificial substrate for PknQ in a time-dependent in vitro kinase assay . Substrates MupFHA and MupDivIVA ( 5 µg each ) were added with PknQ ( 1 µg ) for phosphotransfer reactions carried out in kinase buffer containing 2 µCi [γ-32P]ATP followed by incubation at 25°C for 30 minutes . To determine the ionic requirements of PknQ , in vitro kinase assays were performed in the kinase buffer containing 20 mM HEPES [pH 7 . 2] with various concentrations of divalent cations ( MnCl2 , MgCl2 , ZnCl2 , FeCl2 and ammonium [iron-III] citrate ) were included additionally , as indicated . Inhibition assays with ammonium [iron-III] citrate and hemin ( Sigma ) were performed in the presence of 10 mM Mn2+ and Mg2+ each . Reactions were terminated by 5× Laemmli sample buffer followed by boiling at 100°C for 5 minutes . Proteins were resolved by 10% SDS-PAGE and analyzed by Personal Molecular Imager ( PMI , Bio-Rad ) . Quantification of radioactive bands was done by Quantity One 1-D Analysis Software ( Bio-Rad ) . Autophosphorylated 32P-PknQkd was separated by SDS-PAGE after in vitro phosphorylation and electroblotted onto Immobilon PVDF membrane ( Millipore ) . PAA analysis by two-dimensional thin layer electrophoresis ( 2D-TLE ) was performed as described earlier [34] . Substrates 32P-MupFHA and 32P-MupDivIVA ( phosphorylated by PknQkd ) were analyzed similarly . In order to identify the sites of autophosphorylation in PknQ , His6-PknQkd ( 5 µg ) was autophosphorylated in vitro using 1 mM cold ATP , in the presence or absence of MupFHA ( 20 µg ) . For identification of phosphosites in MupFHA and MupDivIVA , kinase assays were performed with GST-MupFHA ( 5 µg ) and MBP-MupDivIVA ( 5 µg ) in the presence of 1 mM cold ATP and PknQkd ( 2 µg ) . Samples were resolved on 10% SDS-PAGE and gels were stained with coomassie brilliant blue stain . The stained bands corresponding to desired protein sizes were cut-out and used for mass-spectrometry as described earlier [35] . Proteins were resolved by SDS-PAGE and transferred onto nitrocellulose membrane . Membrane was then blocked with 3% bovine serum albumin ( Sigma ) in phosphate-buffered saline containing 0 . 1% Tween-20 ( PBST ) overnight at 4°C . After blocking , the blot was washed thrice with PBST followed by incubation with primary antibodies at 1∶10 , 000 dilution for 1 h at room temperature . Subsequently , after five washes , the blot was incubated in secondary antibodies ( Bangalore Genei ) at 1∶10 , 000 dilution for 1 h at room temperature . The blots were developed using SuperSignal West Pico Chemiluminescent Substrate kit ( Pierce Protein Research Products ) according to manufacturer's instructions . For affinity pull-down assays , Pierce Pull-Down PolyHis Protein∶Protein Interaction Kit was used ( Pierce , Thermo Scientific ) . Briefly , His6-PknQ was over-expressed in E . coli BL-21 ( DE3 ) cells and whole cell lysate was allowed to bind to the resin . The resin was washed to remove the non-specific unbound proteins . The lysates of E . coli BL-21 ( DE3 ) cells were prepared separately containing over-expressed GST-tagged MupFHAwt ( wild type ) , MupFHAR41A and MupFHAS55A . These three lysates were incubated with immobilized His6-PknQ for 1 h at 4°C . The resin was washed again followed by elution . The eluted fractions were analyzed by SDS-PAGE and immunoblotted with anti-GST antibodies ( Abcam ) . mup018c cloned in pGEX-5X-3 or mup012c cloned in pMAL-c2x were co-expressed in E . coli BL-21 ( DE3 ) cells with pACYC-PknQ . PknQK41M was used as a negative control to generate unphosphorylated proteins . MupDivIVA or MupFHA were thus purified as MBP-tagged and GST-tagged proteins , respectively . Phosphorylation status of these proteins was analyzed by Pro-Q Diamond phospho-protein gel staining of SDS-PAGE gels followed by SYPRO Ruby protein gel stain , as described before [30] . These proteins were utilized for subsequent assays . Sandwich ELISA was performed as described earlier [18] . Briefly , His6-tagged kinase ( PknQ and its mutants or PrkD/PrkDS162A mutant ) or MBP-MupDivIVA ( and its mutants ) were dissolved in coating buffer ( carbonate-bicarbonate buffer [pH 9 . 2] ) at a concentration of 10 µg/ml and adsorbed ( 1 µg/well ) on the surface of a 96-well ELISA plate ( Maxisorb , Nunc ) for 2 h at room temperature . After rinsing the wells five times with PBST , the reactive sites were blocked ( 3% BSA in PBST ) overnight at 4°C . The adsorbed proteins were challenged with varying concentrations of soluble GST-tagged proteins ( MupFHA and its mutants or Rv0020c ) dissolved in blocking buffer for 1 h at room temperature . After five washes with PBST , the wells were treated with HRP-conjugated monoclonal antibody against GST ( Abcam ) at 1∶10 , 000 dilution for 1 h at room temperature . Followed by five washes , the chromogenic substrate o-phenylenediamine dihydrochloride ( 0 . 4 mg/ml OPD in 0 . 1 M phosphate/citrate buffer , pH 5 . 0 ) and H2O2 were used to measure the interaction . After addition of stop solution ( 2 . 5 M H2SO4 ) the absorbance was read at 490 nm . The experiments were performed 3 times with freshly purified proteins along with their mutants . For interaction study with peptides , GST-tagged MupFHA or MupFHAS55A ( 150 nM each ) was adsorbed on the 96-well ELISA plate . After washing , 100 nM of pThr ( KRpTIRR , Millipore ) , pSer ( RRApSVA , Millipore ) or a random unphosphorylated peptide ( DRRRRGSRPSGAERRRRRAAAA , [36] ) was allowed to interact with MupFHA . The wells were washed with PBST and His6-tagged PknQ ( 500 nM ) was added . The interaction was measured as described above except the use of anti-His antibody ( Abcam ) at 1∶10 , 000 dilution . The resulting values were normalized to MupFHAS55A interaction , which was used as a negative control .
To identify the total number of STPKs present in M . ulcerans , we performed a BLASTp search using the sequence of the catalytic domain of M . tuberculosis PknB ( 1–331 aa ) , corresponding to the most conserved mycobacterial STPK [37] . Using this approach we identified 13 distinct ORFs encoding for STPKs , of which twelve are encoded on the chromosome and one ( Mup011 or PknQ ) is encoded on the virulence-associated plasmid pMUM001 ( Fig . 1A ) . Analysis of the M . ulcerans STPKs showed that they possess unique kinase modules that are divergent from its close homolog , M . marinum . Out of the 24 putative STPK encoding genes present in M . marinum , only 13 are retained in M . ulcerans . This finding suggests that although M . ulcerans has evolved from M . marinum , it has retained only those STPKs that are necessary for its survival in humans while excluding those that may confer adaptation to M . marinum in fish . STPKs are broadly classified on the basis of conserved catalytic Arg/Asp ( RD ) residues present between subdomains VIa and VIb [30] , [38] . Twelve STPKs in M . ulcerans belong to the RD kinase family , with PknG being the only non-RD kinase . The domain architecture of these kinases show a modular organization in which the kinase domain is located at the N-terminus ( Fig . S1 ) . SMART domain analysis revealed that 9 out of 13 kinases possess a transmembrane region that divides the N-terminal kinase domain from the C-terminal residues ( Fig . S1 ) . Multiple sequence alignments and phylogenetic analyses of M . ulcerans STPK sequences revealed that they belong to diverse origins and form distinct clades with strong conservation patterns in the catalytic domains , similar to M . tuberculosis [39] ( Fig . 1B ) . The most striking feature of M . ulcerans signaling is the presence of PknQ on the virulence-associated plasmid pMUM001 , which was most likely acquired for adaptability . Therefore , we decided to characterize structural features and biochemical properties of PknQ . PknQ contains 660 amino acids with an estimated isoelectric point of 6 . 46 . The domain analysis indicates that the cytosolic N-terminal region possesses the characteristic Ser/Thr kinase domain ( kd ) , which is separated from the extracellular C-terminal domain harboring a FepB-like iron transporter module through a transmembrane region ( Fig . S1 ) . PknQkd ( 1–344 aa ) harbors all 12 conserved Hank's subdomains present in eukaryotic STPK counterparts [40] ( Fig . S2 ) . Homology modeling of PknQkd and subsequent structural analysis revealed that the kinase domain consists of two lobes joined by a hinge segment . Catalysis occurs at the interface of the two lobes , where the catalytic amino acid residues interact with both ATP and the protein substrate ( Fig . 2A ) . Based on our analysis of homology modeling and sequence similarities , we identified the residues Lys41 and Asp134 important for the phosphorylation reaction . Therefore , in order to characterize PknQ and decipher its regulation , the gene coding for PknQkd ( 1–344 aa ) was cloned , over-expressed and the His6-tag fusion protein was purified from E . coli ( Fig . 2B ) . The kinase activity of purified protein was assessed by in vitro kinase assay , in a time-dependent manner ( Fig . 2C ) . As shown , the maximum activity was achieved in 30 minutes under given conditions . In order to confirm PknQ phosphorylation , the conserved Lys41 residue was mutated to methionine as a control . The kinase and its Lys41 mutant were then assessed for autophosphorylation and phosphorylation of the universal substrate , myelin basic protein ( MyBP ) . As shown in Fig . 2D , PknQ was able to phosphorylate MyBP , while PknQK41M was inactive . The phospho-transfer potential of PknQ was also assessed in a time dependent manner and phosphorylation of MyBP was quantified ( Fig . 2E ) . This confirmed the time-dependent increase of the phospho-transfer potential of PknQ , with the saturation of signal observed after 30 min . The domain organization of PknQ includes a C-terminal ion transporter module , indicating that cofactors may play an important role in regulating its activity . To analyze the PknQ ionic requirements , in vitro kinase assays were performed with [γ-32P]ATP in the presence of different divalent cations known to regulate the activity of STPKs [32] , [33] , [41] . PknQ kinase activity was found to be dependent on the presence of Mn2+ , although slight activation was also observed in the presence of Mg2+ , Fe2+ , and Zn2+ ions ( Fig . 3A ) . However , no activation was observed in the presence of Fe3+ ions , rather they inhibited the activity of PknQ in a concentration-dependent manner , even in the presence of Mn2+ and Mg2+ ( Fig . 3B ) . To confirm these results we performed an in vitro kinase assay with PknQkd and hemin , which contains integrated Fe3+ ions . Since the C-terminal FepB domain of PknQ is known to release iron from heme [42] , the kinase activity of PknQfl was also checked . Hemin indeed inhibited the activity of PknQfl as well as PknQkd ( Fig . 3C ) . We identified the phosphorylated amino acid ( s ) on PknQkd using 2D-TLE . We found that both serine and threonine residues were phosphorylated , while no phosphorylation was observed on tyrosine residues ( Fig . 4A ) . Mass spectrometric analysis of PknQkd identified multiple phosphorylated Ser/Thr residues ( Table 1 , Fig . 4B ) . Sequence analysis of residues proximal to the phosphosites identified proline residues in close proximity to seven phosphosites . The importance of proline in proximity to phospho-acceptor residues has been well established in eukaryotic STPKs [43] , [44] . Thus , these residues may help in PknQ autophosphorylation and would be useful for identifying PknQ-specific phosphorylation motifs . We next generated the two phospho-ablative mutants of Thr164 and Thr166 sites , since their homologous residues are known to be the most conserved phosphorylated residues in the activation loop of STPKs [33] , [45] , [46] . Equal amounts of PknQ and its mutants were used in the in vitro kinase assays with [γ-32P]ATP , resolved on SDS-PAGE , and analyzed by autoradiography . In the quantitative analysis , there was no significant loss in phosphorylation observed in the PknQT164A single mutant , while PknQT166A and PknQT164/166A showed significant loss in signal intensity ( Fig . 4C ) . The Ser170 and Thr174 residues were also present in the activation loop of the PknQ catalytic domain ( Fig . 4B ) . In order to determine the impact of phosphorylation of these residues together with the residues of the juxtamembrane region , site-directed mutagenesis was performed to generate mutations at these sites . We observed that the PknQS170A mutant had a marked reduction in phosphorylation activity , suggesting that this site is critical for PknQ activation ( Fig . 4C ) . PknQT174A and PknQT260A mutants also exhibited ∼80% loss in phosphorylation signal . Comparison of PknQ , PknQK41M and PknQT174A phosphorylation indicated that PknQT174A is partially active and is distinct from PknQK41M derivative , which is completely inactive ( Fig . S3A ) . No loss in phosphorylation levels was observed with other single site mutants ( Fig . S3B ) , indicating that phosphorylation of these residues ( including the juxtamembrane region ) does not seem to play a major role in PknQ autokinase activity . Hence , the most important phosphorylation sites in the PknQ are Ser170 , Thr174 , Thr166 and Thr260 . In most bacterial STPKs , activity is regulated by threonine residues in the activation loop and not by serine [11] , [30] . Thus , it was surprising that Ser170 plays a major role in PknQ activation and indicates a novel feature of PknQ in the mycobacterial kinome . In order to regulate and amplify signals , protein kinases associate and interact with a number of proteins within the cell . Most of the bacterial protein kinases are known to exhibit synteny with their substrates . In M . ulcerans , analysis of pknQ genetic loci revealed a possible operon between mup012c and mup018c . These genes code for proteins whose homologs are known as key kinase substrates in M . tuberculosis and many other bacteria . The search for kinase interacting domain took us to the FHA domain that is present in Mup018c ( renamed as MupFHA ) . In M . tuberculosis , several STPK-FHA domain containing protein partners are broadly conserved at the same genetic loci , such as PknB-Rv0019c , PknF-Rv1747 or PknH-EmbR [18] , [47] , [48] . FHA domains are comprised of approximately 55–75 amino acids with three highly conserved blocks- GR , SXXH , and NG- separated by divergent spacer regions [49] . Structurally , the FHA domain contains an 11-stranded β-sandwich with small helical insertions at the loops connecting the strands [49] . Recent reports suggest that the FHA domain-containing proteins interact with and recruit other phospho-proteins [17] , [18] , [21] . In eukaryotes , STPKs associate with multiple signaling domains , such as the BRCT , 14-3-3 , Polo box , SH2 , WW , WD40 , and FHA domains [18] , [50] , while in bacteria , only FHA domains have been identified as the conserved STPK interacting domains . In addition , their role has been studied in important cellular processes [49] . In order to determine the status of FHA domain containing proteins in M . ulcerans , we performed a BLASTp search using Rv0020c of M . tuberculosis as a query [51] and found six additional FHA domain containing proteins ( compared to M . marinum , which has ten FHA domain containing proteins , including the one encoded by the pMUM003 plasmid in M . marinum DL240490 ) . A domain analysis and homology search showed that all of the proteins have functional homologs in M . tuberculosis except MupFHA ( Table 2 ) . To further explain the relationships of MupFHA with other FHA domain homologs in other bacteria that carry a megaplasmid , for example non-pathogenic bacteria Mesorhizobium cicero , we performed BLASTp search with MupFHA sequence in the M . cicero database . There are two FHA domain containing proteins encoded by the M . cicero megaplasmid . However , the two proteins named Mesci_6382 and Mesci_6368 are annotated as type-VI secretion system FHA domain proteins and thus , are different from MupFHA . Similarly , in another pathogenic bacterium , Yersinia pestis , the virulence-associated plasmid pLB1 encodes an FHA domain protein YscD that is a part of the type-III secretion apparatus . Thus , it is possible that MupFHA , YscD , Mesci_6382 and Mesci_6368 are related evolutionarily . However , MupFHA does not contain any such secretory domain indicating divergence at some point . Thus , these analyses indicate that the MupFHA is a unique FHA domain containing protein present in the virulence-associated plasmid of M . ulcerans . mup018c ( mupFHA ) is present in the vicinity of pknQ ( mup011 ) , suggesting that the two proteins encoded by these genes may interact with each other . In order to confirm this hypothesis , we first performed multiple sequence alignment of MupFHA and other known FHA-domain containing proteins . The alignment shows that the five most conserved residues of FHA-domain ( Gly40 , Arg41 , Ser55 , His58 and Asn76 ) are also present in MupFHA ( Fig . S4 ) . Next , we generated a structural model of the FHA domain of MupFHA using M . tuberculosis Rv0020c as a template . MupFHA was found to contain 11 β-strands as is true for all the FHA-domain containing proteins [24] , [49] . PknQkd and MupFHA models were then used for the docking studies and to characterize the key interacting residues . Amongst the major phosphorylation sites of PknQ ( Thr166 , Thr174 , Ser170 and Thr260 ) , Thr166 and Thr260 do not make close contact with MupFHA ( Fig . S5 ) . Thus , the docking was further performed using PknQ phosphorylated at Ser170 and Thr174 . PknQ was found to interact with the residues present in the loops β3–β4 , β4–β5 and β6–β7 of MupFHA , as described previously for other FHA domains [24] , [52] , [53] . Figure 5 shows the various docked complexes emphasizing the role of activation loop residues PknQ-pSer170 ( Sep170 ) and PknQ-pThr174 ( Tpo174 ) in stabilizing the complex formed with MupFHA ( Fig . 5A and 5B ) . PknQ-Thr174 exhibits canonical binding with the FHA-domain residues as observed in various previous studies [24] , [52] , [53]; and forms H-bonds with Ser55 , Ser75 and Arg53 residues of MupFHA ( Fig . 5B and 5C ) . PknQ-pThr174 interacts with MupFHA-Ser55 , one of the five most conserved FHA-domain residues , exactly in the same manner as Human Rad53-FHA1 and M . tuberculosis Rv0020c bind to their phospho-peptides [24] , [53] . Rv0020c contains a Thr at position corresponding to MupFHA-Ser75 ( Fig . S4 ) that also forms H-bond with pThr of the phospho-peptide [24] . The MupFHA model differs from the known Human Rad53 structure with respect to the binding of pThr174 with the other absolutely conserved residue MupFHA-Arg41 . MupFHA-Arg41 was not found to interact with pThr174; instead it forms H-bond with PknQ-pSer170 . MupFHA-Arg41 holds PknQ-pSer170 like a clip with the help of MupFHA-Arg56 ( Fig . 5B and 5D ) , reminiscent of pSer binding by Human PNK-FHA and M . tuberculosis Rv0020c [24] , [52] . Presence of negatively charged pSer residue in proximity to these positively charged arginine residues have been proposed to provide favorable interaction [52] . MupFHA-Arg56 is not among the five absolutely conserved residues but is found in most of the FHA domains ( Fig . S4 ) . Rad53-FHA1 contains an asparagine at this site which also forms H-bond with pThr [53] . This analysis indicates that residues Ser170 and Thr174 of PknQ are important for its interaction with MupFHA . To further apprehend the interaction between these two proteins , we performed docking of the FHA domain mutants of the two absolutely conserved residues- Arg41 and Ser55 with the wild-type PknQkd . Significant loss of H-bonds within PknQ activation loop was observed in both the cases ( Fig . 5E and 5F ) . Serine to alanine substitution at position 55 of MupFHA led to the loss of H-bonding with PknQ-Thr174 ( Fig . 5E ) . Similarly , MupFHAR41A mutation abolished the formation of H-bonds with PknQ-Ser170 that were essential for holding the complex ( Fig . 5F ) . Together , these analyses identified critical amino acid residues involved in the PknQ and MupFHA interaction . Molecular dynamics simulation was performed to probe the conformational relaxation of the PknQ-MupFHA complex and to evaluate the dynamic interaction of phosphorylated residues in PknQ activation loop ( pSer170 and pThr174 ) with the FHA domain ( Supplementary movie files S1 , S2 and S3 ) . Apart from the dually phosphorylated PknQ-pSer170/pThr174 ( Fig . 6A , left panel ) , the molecular dynamics simulation was also performed with PknQ-pSer170/Thr174Ala ( Fig . 6A , middle panel ) or PknQ-Ser170Ala/pThr174 ( Fig . 6A , right panel ) . The snapshots of complexes at different time points suggest that the dually phosphorylated form of PknQ makes the most stable complex with MupFHA . Both the mutations- T174A and S170A , lead to destabilization and fluctuations in the complex ( Fig . 6A ) . The analysis of overall H-bonding in the complex again shows that the dually phosphorylated PknQ forms a stable complex with MupFHA . The number of H-bonds decreases in the case of complex formed with PknQ mutants . Analysis of charge distribution in the complex suggests that in the dually phosphorylated model , Ser55 and Ser75 of MupFHA play an important role in the recognition and form 3–4 H-bonds during interaction with PknQ-pThr174 . PknQ-pSer170 interacts with the Arg41 and Arg56 residues of MupFHA . These interactions are maintained even when there is a Thr174 to alanine substitution in PknQ . When Ser170 is substituted to alanine , MupFHA-Arg56 becomes available for H-bonding . There is a relative movement of Arg56 so that it now interacts with PknQ-pThr174 . The analogous interaction of MupFHA-Arg56 and PknQ-pThr174 has also been observed previously in other studies [24] , [52] , but is not seen when there is a pSer170 residue in PknQ . Thus , MupFHA-Arg56 moves towards PknQ-pThr174 only when pSer170 is absent . In this case , when H-bonds between PknQ-Thr174 and MupFHA-Arg56 were analyzed , it indicated that before 4 ns there is no significant interaction , but after 4 ns , 2–3 H-bonds are formed between these two residues ( Fig . 6B ) . The RMSD values were then calculated and plotted for all the three complexes ( Fig . 6C ) . The graph shows an initial structural rearrangement ( 1–2 ns ) contributing to higher fluctuations in RMSD values for all the three protein complexes ( Fig . 6C ) . The complexes formed with single phosphorylated residue of PknQ exhibit higher structural rearrangements and higher RMSD fluctuations compared to the double phosphorylated protein complex . These results indicate that Thr174 and Ser170 of PknQ are important mediators of the interaction with MupFHA . Phosphorylation of Ser170 may regulate interaction of MupFHA with PknQ by making MupFHA-Arg56 and Arg41 unavailable for pThr174 . To validate the interactions between MupFHA and PknQ observed through docking analysis , we performed affinity pull-down assays . Mup018c was cloned in an E . coli expression vector and the corresponding protein MupFHA was purified as a GST-tagged fusion protein . Site-specific mutants of MupFHA were also generated and used for pull-down assays with PknQ to validate the docking studies . Pull-down assays showed that PknQ strongly interacts with wild-type MupFHA and that the interaction was reduced with the MupFHAR41A and MupFHAS55A mutants ( Fig . 7A ) . To further validate the interaction of the FHA domain with PknQ , we studied specific protein-protein interactions through sandwich ELISA . His6-PknQ was adsorbed on a 96-well plate and allowed to interact with equimolar amounts of GST-tagged substrates . Significant interaction was observed between His6-PknQ and GST-MupFHA ( Fig . 7B ) . Interaction assays of PknQ were also performed with the GST-MupFHAR41A and GST-MupFHAS55A mutants . The interactions were severely disrupted by mutating the two residues of MupFHA ( Fig . 7B ) . Thus , our results confirmed that the interaction of PknQ∶MupFHA occurred via the FHA domain and requires Arg41 and Ser55 . Docking studies and MD analysis indicated the critical role of the activation loop residue Ser170 in the PknQ∶MupFHA interaction along with the pThr174 . FHA domains are primarily pThr binding domains . In fact , such pThr specificity helps in decreasing the potential interaction sites of FHA domains , as 90% of all Ser/Thr kinase activity in eukaryotes is directed towards serine phosphorylation [54] . However , in prokaryotes , and even more so in Mycobacterium , STPKs more often act upon threonine residues [11] . However , PknQ is an exception , as it requires Ser170 in addition to threonine residues . Therefore , to investigate the role of Ser170 in PknQ∶MupFHA interaction , we performed an ELISA with MupFHA and PknQS170A . Our results showed a loss of interaction between MupFHA and PknQS170A compared to wild-type PknQ and PknQT166A ( Fig . 7C ) . The interaction of PknQS170A was similar to the kinase inactive mutant PknQK41M , which is the completely unphosphorylated form of the kinase . We also compared this interaction with interaction of other key phospho-residue mutants ( Thr174 and Thr260 ) ( Fig . 7D ) . Apart from Ser170 , the only residue important for interaction is Thr174 confirming the results obtained from structural modeling ( Fig . 5A ) . Furthermore , MupFHA-PknQ interaction was probed in competition ELISA-based assays using synthetic peptides phosphorylated either on threonine or serine residue . Immobilized MupFHA was first incubated with pSer , pThr or random unphosphorylated peptide and competitively replaced by PknQ . In this assay , the MupFHA-PknQ interaction was proportionally dependent on MupFHA∶phospho-peptide binding . We observed that both pThr and pSer peptides showed affinity to MupFHA but not the unphosphorylated random peptide ( Fig . S6 ) . Notably , the binding of MupFHA with both phospho-peptides was not very strong , which probably indicates the role of neighbouring residues ( of phospho-acceptor site ) in these interactions . To further probe the pSer/pThr interaction , we also utilized Bacillus anthracis kinase PrkD that autophosphorylates on Ser162 residue , in addition to several threonine and tyrosine residues [30] . MupFHA was allowed to bind with PrkD and PrkDS162A where the interaction was found to be reduced after mutation of Ser162 ( Fig . S7 ) . Together , these results indicate the affinity of MupFHA with both pThr and pSer . FHA domain containing proteins are known to be phosphorylated by their neighboring kinases [18] , [20] , [48] . To explore this possibility in M . ulcerans , we performed kinase assays with PknQ and MupFHA and found that PknQ phosphorylated MupFHA , while no phosphorylation was observed with PknQK41M ( Fig . 8A ) . To further validate MupFHA phosphorylation by PknQ , we co-expressed the two proteins in E . coli using compatible vectors , one expressing either PknQ or PknQK41M ( pACYCDuet-1 ) and the second expressing MupFHA ( pGEX-5X-3 ) . Affinity purified MupFHA that was co-expressed with PknQ ( renamed as MupFHA-P ) showed an intense phosphorylation signal , while no phosphorylation was found in MupFHA co-expressed with PknQK41M ( MupFHA-UP ) ( Fig . 8B ) . We also assessed the role of active site residues of PknQ on its ability to phosphorylate MupFHA . Considerable loss of phosphotransfer was observed with the mutants- PknQT166A , PknQS170A and PknQT174A compared to the wild-type and PknQT164A ( Fig . 8C ) . Therefore , in addition to autophosphorylation , the residues Ser170 , Thr166 , and Thr174 are also critical for regulating the phosphotransfer . To identify the MupFHA residues phosphorylated by PknQ , PAA analysis was performed . PknQ phosphorylated MupFHA on threonine residue ( s ) , while no signal was observed on the spots corresponding to pSer and pTyr ( Fig . 9A ) . We subsequently identified the phosphorylation sites in MupFHA by mass spectrometry . PknQ phosphorylated MupFHA on four threonine residues ( Thr8 , Thr123 , Thr210 , and Thr214 ) , which confirmed the results obtained from the PAA analysis ( Fig . 9B , Table 3 ) . Interestingly , none of these residues were present in the FHA domain . To verify these phosphorylation sites , we generated single and multiple phospho-ablative mutants of MupFHA and compared the level of phosphorylation in all the mutants with wild-type MupFHA . Surprisingly , loss of only Thr210 resulted in approximately a 50% reduction in phosphorylation signal , while ∼20% loss of signal intensity was observed with the double mutant MupFHAT8A/T214A ( Fig . 9C ) . No loss was observed for the MupFHAT123A mutant . Since the FHA domain of MupFHA is involved in its interaction with PknQ , it was essential to study the role of the FHA domain in its phosphorylation . Therefore , the two conserved FHA domain residues , Arg41 and Ser55 , were mutated to alanine and the effect on PknQ-mediated phosphorylation of MupFHA was subsequently analyzed . We found that the MupFHAR41A and MupFHAS55A mutants had an approximately 40% and 80% loss of phosphorylation signal compared to wild-type protein , respectively ( Fig . 9D ) . These results indicate that FHA-mediated interaction of MupFHA is necessary for its phosphorylation by PknQ . This cooperation between FHA domain residues and other phosphorylated residues most likely helps in enhancing and regulating the interaction as well as phosphorylation of MupFHA . Importantly , these results are also in agreement with the docking studies where the Arg41 and Ser55 mutations weakened the interaction between MupFHA and PknQ ( Fig . 5E and 5F ) . In the M . tuberculosis CDC1551 strain , the FHA domain containing protein EmbR2 ( a structural homologue of EmbR ) affects the kinase activity of PknH . Although EmbR2 is not phosphorylated by PknH , it inhibits the kinase autophosphorylation and inactivates the protein [55] . Similarly , Rv0020c , a FHA domain containing protein of M . tuberculosis , regulates the cell wall regulator pseudokinase MviN [17] . In our study , we found that MupFHA specifically inhibited PknQ activity in the in vitro kinase assay , indicating that their interaction may have a negative impact on the kinase activity ( Fig . 8A ) . To further investigate the effect of the MupFHA interaction on PknQ activity , we used mass spectrometry to assess changes in phosphorylation patterns . PknQ was allowed to autophosphorylate in the presence of MupFHA and then analyzed by mass spectroscopy to determine the phosphorylation sites . Our results showed that PknQ only autophosphorylates on eight sites in the presence of MupFHA ( Table 4 ) as compared to 20 sites in the absence of MupFHA ( Table 1 ) . These results confirm that MupFHA acts as a negative regulator of PknQ kinase activity . Interestingly , the phosphorylation of Thr166 , Ser170 and Thr174 residues was impervious to MupFHA-based inhibition , underscoring the critical requirement of activation loop residues in the activity of PknQ . The kinase activation process involves initial phosphorylation on activation loop residues and subsequently other sites are phosphorylated to generate its active conformation . Inhibition by MupFHA may reduce/abrogate these conformational changes and therefore the kinase may only be able to reach its fully active conformation either in the absence of MupFHA or by any other ligand that might abolish this interaction . Interestingly , Ser/Thr phosphatases are the only known negative regulators of STPK-mediated signaling and there is no such phosphatase encoding gene present on the pMUM001 plasmid . Therefore , MupFHA may help in regulating the activity of PknQ by limiting phosphorylation to specific substrate ( s ) and act as a principal controller of this signaling scheme . However , MupFHA may only act as an additional regulator of the kinase activity , while Ser/Thr phosphatase ( Mul_0022 ) encoded in M . ulcerans genome may control the dephosphorylation as we found that MupFHA and PknQ get dephosphorylated by M . tuberculosis Ser/Thr phosphatase PstP ( which is ∼94% similar to M . ulcerans PstP Mul_0022 , Fig . S8 ) . STPKs regulate peptidoglycan synthesis and other cell wall processes in diverse bacteria [56]–[58] . In M . tuberculosis , PknA and PknB regulate Wag31 , which is a DivIVA domain containing protein that regulates growth , morphology , polar cell wall synthesis , and peptidoglycan synthesis [59] , [60] . In pMUM001 , the kinase gene mup011 ( pknQ ) and mup012c ( encoding the DivIVA domain-containing protein MupDivIVA , Fig . S9 ) are adjacent . In earlier reports of M . tuberculosis STPKs , the proteins encoded by the neighboring genes of kinases were found to be specific substrates of those kinases [18] , [39] , [61] . To validate this hypothesis in the case of M . ulcerans , we determined if MupDivIVA was a PknQ substrate . The mup012c was cloned into pMAL-c2x and MupDivIVA was purified as a MBP-tagged fusion protein . We found that PknQ efficiently phosphorylated MupDivIVA in the in vitro kinase assay while there was no phosphorylation on MupDivIVA with PknQK41M ( Fig . 10A ) . To test the authenticity of PknQ-dependent phosphorylation of MupDivIVA , we co-expressed PknQ or PknQK41M with MupDivIVA in the surrogate host E . coli using compatible expression vectors ( pMAL-c2x-MupDivIVA and pACYCDuet1-PknQ ) . Phosphorylation-specific ProQ Diamond staining revealed that MupDivIVA was phosphorylated only when co-expressed with the catalytically active kinase ( MupDivIVA-P ) and not with the kinase-inactive mutant ( MupDivIVA-UP ) ( Fig . 10B ) . PAA analysis showed that phosphorylation was located at serine and threonine residues of MupDivIVA ( Fig . 10C ) . Mass spectrometry analysis of MupDivIVA identified six phosphorylation sites ( Table 5 ) . Incidentally , all phosphorylation sites were found to be localized within DivIVA core domain ( 42–83 aa ) ( Fig . 10D ) . Comparison of MupDivIVA phosphorylation sites with only phosphorylated threonine residue identified in M . tuberculosis Wag31 revealed that the three serine phosphorylation sites are present in the same DivIVA region in both the proteins ( Fig . S9 ) . Such strikingly similar phosphorylation patterns may indicate their conserved role . The three serine phosphorylation sites were mutagenized to alanine and were compared for their phosphorylation levels . The triple mutant MupDivIVAS43/45/49A showed ∼70% loss of phosphorylation , indicating the importance of serine phosphorylation ( Fig . 10E ) . Since the FHA domain recognizes phosphorylated proteins , we hypothesized that MupFHA may interact with the phosphorylated MupDivIVA through a three-way regulatory process with PknQ . To evaluate this hypothesis , the MBP-tagged proteins MupDivIVA-P ( phosphorylated ) or MupDivIVA-UP ( unphosphorylated ) and GST-tagged MupFHA were used in a sandwich ELISA . We found that MupDivIVA phosphorylation increased its affinity for MupFHA ( Fig . 11A ) . Furthermore , there was a considerable loss in interaction with MupDivIVA when the conserved FHA domain residues of MupFHA were mutated ( Fig . 11B ) . These results confirm that the FHA domain mediates the interaction of MupFHA with phosphorylated MupDivIVA . Therefore , MupFHA interacts with MupDivIVA through a phosphorylation dependent manner , which is regulated by PknQ . Further , we used MupDivIVA to probe pSer binding affinity of MupFHA . We compared the interaction of MupDivIVA phosphorylation site mutants with MupFHA . M . tuberculosis Rv0020c was used to compare pThr specificity ( Fig . 11C ) . This analysis revealed role of phosphorylated serine/threonine residues in MupFHA∶MupDivIVA interaction while only pThr residues regulate Rv0020c∶MupDivIVA interaction .
In this study we analyzed the STPK-mediated signaling system of M . ulcerans , which is the causative agent of Buruli ulcer . M . ulcerans is a slow growing bacterium ( slower than M . marinum and M . tuberculosis ) [62] , [63] and this slow growth together with restrictive temperature requirements are the major reasons for our limited understanding about this important human pathogen and its signaling systems . Using in silico analysis , we identified 13 STPKs in the M . ulcerans genome that are distinct from its close relative M . marinum that has 24 STPKs [39] . STPKs of M . tuberculosis have been classified in five clades [39] , and phylogenetic analysis reveals that M . ulcerans also has individual STPKs related to all five clades with an over-representation of the PknF/PknI/PknJ clade . Analyses of STPKs and FHA domain encoding genes confirmed that M . ulcerans underwent reductive evolution compared to M . marinum . The presence of PknQ on the virulence-associated plasmid pMUM001 makes the M . ulcerans kinome exclusive than M . tuberculosis and other characterized bacterial kinomes [64] . Three such plasmids have been identified in the Mycobacterium species , including pMUM001 ( M . ulcerans Agy 99 ) , pMUM002 ( M . liflandii 128FXT ) , and pMUM003 ( M . marinum DL240490 ) [65] . These three plasmids are involved in mycolactone synthesis and most likely in pathogenesis . In addition , all of the plasmids encode a homolog of STPK ( MUP011 , MULP_022 , and MUDP_075 , respectively ) , although a frameshift mutation suggests that MUDP_075 is a pseudogene [65] . Notably , these plasmids have a conserved kinase locus ( from the STPK gene [mup011] to the FHA domain-containing protein [mup018c] ) , although pMUM003 is slightly different , most likely due to a frameshift mutation [65] . Large plasmids , such as pMUM001 , which encode proteins involved in adaptation to new environments , represent regular theme among many other bacterial pathogens , such as B . anthracis , Y . pestis , and Shigella [64] , [66]–[68] . These species share a nearly identical genome structure and sequence with other species in their genera , but due to the acquisition of virulence-associated plasmids , these bacteria tend to acclimatize better to new conditions . In the plague causing bacteria Y . pestis , the virulence-associated plasmid pLB1 encodes a STPK YpkA that acts as a direct inducer of cell death by promoting apoptosis and actin depolymerization during infection [66] , [69] . Interestingly , other than YpkA , pLB1 encodes several other antigens , and one of them , the type III secretion apparatus protein YscD also possesses an FHA domain [70] . The role of a virulence-associated plasmid in the pathogenesis of M . ulcerans has previously been described [4] , [71] , but its significance beyond mycolactone biosynthesis has not been appreciated until now . Our results demonstrated that multiple proteins encoded on pMUM001 are regulated by phosphorylation , and therefore this plasmid may be an important component of signaling cascades in M . ulcerans . To understand the significance of PknQ in M . ulcerans , we elucidated the biochemical characteristics of this kinase . Our biochemical characterization identified both common and novel features . PknQ kinase activity is dependent on cofactors such as Fe2+ , Mn2+ , Mg2+ , and Zn2+ . Interestingly , the kinase is activated by Fe2+ , while its activity is inhibited by Fe3+ . This inhibition occurs in the presence of hemin , indicating that PknQ activity is regulated by iron and its redox state in the cellular milieu . The role of iron in the regulation of its kinase activity is justified by the presence of a FepB-like iron transporter domain at the C-terminus of PknQ . However , further studies are required to validate this aspect of PknQ signaling . To understand the activation mechanism of PknQ , several mutants were generated . We found that the autophosphorylation of serine and threonine residues in the activation loop region regulates PknQ activity . In most M . tuberculosis STPKs , activity is regulated by two threonine residues present in the activation loop . Analogous residues ( Thr164 and Thr166 ) are also phosphorylated in PknQ , but only Thr166 is critical for autophosphorylation activity . Furthermore , a novel serine residue , Ser170 , was found to regulate the autokinase and substrate phosphorylation activities of PknQ . In fact , phosphorylation of activation loop residues is known to be required for stabilization of kinases as it can induce specific conformational changes important in substrate binding [72] . Thus , further structural studies on PknQ are needed to understand the role of each phosphoresidue and how this serine phosphorylation is mechanistically different from threonine phosphorylation . We also identified the FHA-domain containing protein MupFHA as an interacting partner of PknQ . To understand how MupFHA and PknQ interact , we generated structural models to establish the basis of MupFHA and PknQ interaction . Interestingly , the PknQ and MupFHA interaction highlighted several novel aspects . The structural analyses suggest that PknQ-MupFHA interaction follows the canonical binding mode of pThr interaction with specific arginine and serine residues of FHA-domain . However , MupFHA additionally interacts with a pSer residue present in the activation loop of PknQ . The structural models were validated by experimental analysis using ELISA and affinity pull-down assays , thus establishing the structure-activity relationship for PknQ . Together , these findings highlighted the unconventional molecular recognition patterns of kinase∶FHA interactions . The second unique aspect of the MupFHA∶PknQ interaction is the role of MupFHA as a negative regulator of PknQ . To the best of our knowledge , this is only the second report of any bacterial FHA-mediated regulation of the cognate kinase activity . It has been previously reported that EmbR2 inhibits PknH phosphorylation [55] , but this is restricted to the M . tuberculosis CDC1551 strain . Nevertheless , MupFHA signaling is unique , as PknH does not phosphorylate EmbR2 , while MupFHA is a substrate of PknQ . This regulation of kinase activity by FHA domains could have important implications in the spatio-temporal regulation of cellular signaling . It is important to note that the critical activation loop residues were phosphorylated even in the presence of MupFHA , thus indicating that MupFHA may only regulate the secondary phosphorylation sites and substrate binding of PknQ . To understand this inhibition further , we applied mass spectrometry and found a significant reduction in number of phosphorylation sites . However , the mass spectrometry analysis did not quantitate the phosphorylation stoichiometry of each site and we cannot rule out the possibility that the phosphorylation of key activation loop residues is also inhibited . The third aspect of MupFHA signaling is the ability of both PknQ and MupFHA to interact with another phosphorylated protein , MupDivIVA . Our results indicate that PknQ phosphorylates MupDivIVA , which is a homolog of M . tuberculosis Wag31 , another DivIVA domain-containing protein [60] . Wag31 in M . tuberculosis is already known to be phosphorylated by PknB and PknA and this phosphorylation is critical for cell growth and peptidoglycan synthesis [59] . Upon phosphorylation , MupDivIVA also interacts with MupFHA in a phosphorylation-dependent manner . Therefore , this study provides insights into three-way regulation involving dynamic signaling between PknQ , MupFHA , and MupDivIVA . In conclusion , our study is the first analysis of signaling pathways in M . ulcerans and has revealed many novel aspects of signaling systems among Mycobacterium species . Our results indicate that PknQ could be an important sensor of extracellular cues in M . ulcerans and can propagate the signals to MupFHA and MupDivIVA . Moreover , iron and MupFHA may act as quenchers in this phosphorylation cascade . Taken together , our data underscore the importance of structure-activity studies in unraveling the PknQ-MupFHA signaling axis in M . ulcerans and provide an interesting starting point to work towards understanding this pathogen .
|
Mycobacterium ulcerans is a slow growing pathogen , which is prevalent in many tropical and sub-tropical countries . M . ulcerans possesses unique signaling pathways with only 13 STPK containing genes . This is strikingly different from its closest homolog Mycobacterium marinum and surprisingly closer to the human pathogen , Mycobacterium tuberculosis . PknQ , MupFHA and MupDivIVA are regulatory proteins encoded by the virulence determining plasmid pMUM001 of M . ulcerans . In addition to characterizing the STPK , we focused on deciphering the basis of interaction between the three partner proteins leading to the identification of critical residues . Present study describes the newly identified phosphoserine-based interactions , which is unique amongst the FHA-domain containing proteins . We confirmed our results using structural analysis via specific mutants and their interaction profiles . Importantly , these data highlight the significance of FHA domains and their role in understanding cellular signaling . This work will encourage further studies to elucidate role of M . ulcerans signaling systems . It will also raise questions like how less studied tropical bacterial pathogens acquire eukaryotic-like Ser/Thr protein kinase and exhibit unusual mechanisms to interact with its partner domains .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"protein",
"interactions",
"enzymes",
"regulatory",
"proteins",
"enzymology",
"microbiology",
"emerging",
"infectious",
"diseases",
"protein",
"structure",
"enzyme",
"chemistry",
"bacterial",
"pathogens",
"infectious",
"diseases",
"proteins",
"medical",
"microbiology",
"enzyme",
"regulation",
"microbial",
"pathogens",
"recombinant",
"proteins",
"proteomics",
"biochemistry",
"enzyme",
"structure",
"biology",
"and",
"life",
"sciences"
] |
2014
|
Identification of Ser/Thr kinase and Forkhead Associated Domains in Mycobacterium ulcerans: Characterization of Novel Association between Protein Kinase Q and MupFHA
|
Lentiviruses can infect non-dividing cells , and various cellular transport proteins provide crucial functions for lentiviral nuclear entry and integration . We previously showed that the viral capsid ( CA ) protein mediated the dependency on cellular nucleoporin ( NUP ) 153 during HIV-1 infection , and now demonstrate a direct interaction between the CA N-terminal domain and the phenylalanine-glycine ( FG ) -repeat enriched NUP153 C-terminal domain ( NUP153C ) . NUP153C fused to the effector domains of the rhesus Trim5α restriction factor ( Trim-NUP153C ) potently restricted HIV-1 , providing an intracellular readout for the NUP153C-CA interaction during retroviral infection . Primate lentiviruses and equine infectious anemia virus ( EIAV ) bound NUP153C under these conditions , results that correlated with direct binding between purified proteins in vitro . These binding phenotypes moreover correlated with the requirement for endogenous NUP153 protein during virus infection . Mutagenesis experiments concordantly identified NUP153C and CA residues important for binding and lentiviral infectivity . Different FG motifs within NUP153C mediated binding to HIV-1 versus EIAV capsids . HIV-1 CA binding mapped to residues that line the common alpha helix 3/4 hydrophobic pocket that also mediates binding to the small molecule PF-3450074 ( PF74 ) inhibitor and cleavage and polyadenylation specific factor 6 ( CPSF6 ) protein , with Asn57 ( Asp58 in EIAV ) playing a particularly important role . PF74 and CPSF6 accordingly each competed with NUP153C for binding to the HIV-1 CA pocket , and significantly higher concentrations of PF74 were needed to inhibit HIV-1 infection in the face of Trim-NUP153C expression or NUP153 knockdown . Correlation between CA mutant viral cell cycle and NUP153 dependencies moreover indicates that the NUP153C-CA interaction underlies the ability of HIV-1 to infect non-dividing cells . Our results highlight similar mechanisms of binding for disparate host factors to the same region of HIV-1 CA during viral ingress . We conclude that a subset of lentiviral CA proteins directly engage FG-motifs present on NUP153 to affect viral nuclear import .
Retroviruses integrate their reverse transcribed genomes into host cell chromosomes to provide a permanent vantage from which to amplify themselves for subsequent transmission . As the nuclear envelope physically separates the host chromosomes from the cytoplasm during interphase , retroviruses have evolved mechanisms to bypass this natural barrier to the nuclear compartment . The γ-retrovirus Moloney murine leukemia virus ( MLV ) is believed to await the dissolution of the nuclear envelope during mitosis , a mechanism that limits infection by this virus to actively dividing target cells [1]–[3] . Lentiviruses such as HIV-1 infect post-mitotic cell subtypes during the establishment of host systemic infection , and correspondingly harbor mechanisms to infect cells during interphase , likely circumventing the nuclear envelope by passing through the channel present in the nuclear pore complex ( NPC ) [4] , [5] . The vertebrate NPC is a large ∼120 MDa macrostructure , composed of ∼30 different proteins called nucleoporins ( NUPs ) that stack in rings of eight-fold symmetry to form the tubular pore as well as the attached cytoplasmic filaments and nuclear basket substructures [6] , [7] . Approximately one-third of the NUPs harbor domains rich in phenylalanine-glycine ( FG ) motifs , commonly observed as FxF , FxFG , or GLFG patterns [8] . These FG-rich domains line the central channel of the NPC , as well as the cytoplasmic and nuclear openings [9] , and dictate the selective passage of macromolecules through the pore; small molecules are able to passively diffuse , while molecules greater than ∼9 nm in diameter need to be ferried by specialized carrier proteins capable of interacting with the FG-based permeability barrier [10] . The HIV-1 nucleoprotein substrate for proviral integration , called the pre-integration complex ( PIC ) , is estimated at ∼56 nm in diameter [11] , and thus requires active translocation into the nucleus . While initial studies suggested that HIV-1 integrase ( IN ) , matrix , and Vpr proteins , as well as a triple-stranded DNA structure of the reverse transcribed genome called the DNA flap , were key viral elements required for PIC nuclear import , subsequent studies found none of these factors to be essential [12] . Contrastingly , the viral capsid ( CA ) protein was shown to be the major viral determinant for infecting non-dividing cells [13] , [14] . Various host proteins have also been shown to participate in HIV-1 nuclear import , with perhaps the most promising candidates emerging from a series of genome-wide RNA interference ( RNAi ) screens; factors identified in more than one of these screens include transportin-3 ( TNPO3 or TRN-SR2 ) , NUP358 ( RANBP2 ) , and NUP153 [15]–[17] . We have been particularly interested in NUP153 , which plays an important CA-dependent role in HIV-1 PIC nuclear import [18] , [19] . NUP153 is a FG nucleoporin that predominantly locates to the nuclear side of the NPC and exchanges dynamically with a nucleoplasmic population [20] . While NUP153 is anchored to the nuclear rim of the NPC through its N-terminal domain [21] , its C-terminal FG enriched domain ( referred to as NUP153C herein ) is natively unfolded and highly flexible [22] . The ∼200 nm long NUP153C potentially reaches through to the cytoplasmic side of the NPC channel [23] , shifting in spatial distribution in a transport-dependent manner [24] , [25] . Human NUP153C contains 29 FG motifs ( FxF , FG , and FxFG patterns ) , which provide a vital role in NUP153-mediated nucleocytoplasmic transport [26]–[28] . While numerous studies have demonstrated the functional significance of CA for HIV-1 nuclear import and integration , the mechanistic details for these connections are incompletely understood . Retroviral CA proteins are composed of two α-helical domains , the N-terminal domain ( NTD ) and C-terminal domain ( CTD ) , separated by a short flexible linker . CA multimerizes into hexameric arrays during particle maturation , while twelve interspersed pentamers dictate the overall shape of the condensed viral core [29]–[32] . While relatively intact cores enter the cell upon viral-cell membrane fusion , little if any CA remains associated with the PIC within the nucleus [33]–[37] . The precise location and mechanism of CA core disassembly remains controversial: while initial steps of core uncoating are tied to reverse transcription [38] , subsequent events may involve binding to host proteins . This may involve cyclophilin A ( CypA ) and the NUP358 cyclophilin homologous domain ( CHD ) , both of which bind the cyclophilin binding loop protruding from the top of the CA NTD [39] , [40] , or cleavage and polyadenylation specific factor 6 ( CPSF6 ) , which binds a hydrophobic pocket [41] located between α-helices 3 and 4 within the NTD . The small molecule PF-3450074 ( PF74 ) , which inhibits HIV-1 infection by destabilizing incoming CA cores , also engages this same pocket [42] . CA-containing protein complexes have been observed alongside the nuclear envelope [43] , suggesting that the ultimate steps of core uncoating may occur at the nuclear periphery and/or during PIC nuclear transport . Here , we find that the CA proteins from numerous lentiviruses , including HIV-1 and equine infectious anemia virus ( EIAV ) , directly bind NUP153C , with subsequent mapping demonstrating the importance of individual FG motifs themselves . A panel of HIV-1 CA mutants highlights the importance of side-chains lining the CA NTD helix 3/4 hydrophobic pocket , and competition with both CPSF6 and PF74 support this as the site of NUP153C binding . Correlation between NUP153 binding and dependence on endogenous NUP153 expression additionally support the relevance of this interaction during infection . HIV-1 CA mutant viruses N57A and N57D were defective for NUP153C binding and acutely sensitive to the arrest of the cell-division cycle , with a significant correlation between cell cycle and NUP153 dependencies observed among an expanded set of CA mutant viruses . Our data support a model whereby partially uncoated cores directly engage NUP153 FG motifs within the NPC to affect HIV-1 PIC nuclear import .
As we previously found CA to be the dominant viral determinant of the requirement for NUP153 during HIV-1 infection [19] , we tested whether a physical interaction between NUP153 and HIV-1 CA exists . Our initial assay utilized a recombinant viral fusion protein consisting of HIV-1 CA and nucleocapsid ( NC ) proteins , which when assembled in vitro in the presence of high salt and single stranded nucleic acid forms large tube-like structures that readily pellet through cushions of sucrose [29] . In this way , CA-interacting proteins can co-sediment with the tube structures [18] , [44] . Full length or various fragments of HA-tagged NUP153 expression constructs were transfected into 293T cells , and the resulting proteins were tested for their ability to co-sediment with CA-NC assemblies . Full-length NUP153 ( residues 1–1475 ) pelleted through the sucrose cushion in a CA-NC dependent manner ( Figure 1A and 1B ) . The NUP153 N-terminal domain ( residues 1–650 ) failed to bind CA-NC under conditions that supported efficient NUP153C ( residues 896–1475 ) binding . The C-terminal NUP153 deletion mutant comprised of residues 1–1198 failed to bind , confirming the importance of the NUP153 FG-repeat domain in binding , and mapping the interaction to residues 1199–1475 of the full length protein . We addressed whether the NUP153–CA-NC interaction was the result of direct protein binding through the use of purified , recombinant NUP153 protein . We attempted to express full-length NUP153 fused to glutathione S-transferase ( GST ) in bacteria , but despite extensive effort , were unable to define conditions that yielded usable quantities of GST-NUP153 protein . Based on our preliminary binding data ( Figure 1B ) , we instead expressed and affinity purified GST-NUP153C . NUP153C was liberated from the GST tag by site-specific proteolysis , with the remaining CA binding studies utilizing tag-free NUP153C protein . Approximately 40% of the input recombinant NUP153C protein was recovered during co-sedimentation under conditions where binding of a negative control GST protein was undetected ( Figure 1C ) . To test whether NUP153C binds CA in the absence of NC and nucleic acid , his-tagged HIV-1 CA expressed and purified from Escherichia coli was utilized in Ni-nitrilotriacetic acid ( NTA ) pulldown assays . Approximately 30% of input NUP153C was pulled down by his-tagged HIV-1 CA protein . Notably , this interaction is likely independent of CA oligomerization , as double mutant W184A/M185A CA , which is unable to dimerize and form higher-ordered assemblies [45] , pulled-down comparable amounts of NUP153C ( Figure 1D ) . The isolated CA NTD ( CAN ) was expressed as a his-tagged protein and purified to next probe the binding region within HIV-1 CA; CAN pulled down ∼30% of input NUP153C protein ( Figure 1E ) . Although these data do not quantitatively address potential CA oligomerization-based affects on NUP153C binding , the relatively robust interaction with CAN suggests that NUP153 may efficiently engage monomeric CA during HIV-1 infection . The preceding results established a direct interaction between NUP153 and HIV-1 CA proteins in vitro . We next examined whether an assay could be constructed to visualize the interaction in the context of HIV-1 infection . We scored for potential intracellular interaction by relying upon the potent capability of rhesus Trim5α ( rhTrim5α ) to inhibit HIV-1 infection . RhTrim5α is a cytoplasmically localized restriction factor , capable of blocking HIV-1 infection at an early post-entry step [46] . While the C-terminal B30 . 2 ( SPRY ) domain recognizes patterns present on the surface of retroviral CA cores [47] , [48] , the N-terminal RING , B-box 2 , and coiled coil ( RBCC ) effector domains block infection by eliciting a combination of inhibitory activities , including premature disassembly of the viral core [44] , proteasomal targeting [49] , and triggering of innate immune signaling [50] . Both naturally occurring , as well as artificially engineered variants of Trim5 have been discovered wherein the SPRY domain is replaced by heterologous coding sequences , retaining viral restriction while changing the method by which the viral core is recognized [40] , [51] , [52] . In this vein we tested for intracellular recognition between NUP153C and HIV-1 CA by replacing the SPRY domain of rhTrim5α with NUP153C , concomitantly introducing either an internal- or C-terminal HA epitope tag to enable detection of the fusion proteins by western blotting ( Figure 2A ) . These constructs , as well as control constructs encoding only the epitope-tagged rhTrim5 RBCC or NUP153C , were stably introduced into human osteosarcoma ( HOS ) cells ( Figure 2B ) . While a single species of C-terminally HA tagged Trim-NUP153C of the expected molecular weight was detected by western blot , the internally tagged construct revealed the protein susceptible to degradation , with the full-length protein representing only a minority of the expressed products at steady state ( Figure 2B and 2C ) . Regardless , Trim-NUP153C expressing cells potently restricted HIV-1 infection , yielding consistent 5–10 fold reductions in viral titer ( Figure 2D ) . The combination of both rhTrim5 RBCC and NUP153C domains was necessary , as neither domain expressed alone inhibited HIV-1 infection . Knockdown of endogenous NUP153 acutely attenuates HIV-1 infection with little or no effect on MLV [19] . Importantly , the observed attenuation of HIV-1 infection by Trim-NUP153C expression was specific , as infection by an MLV reporter virus was unaffected ( Figure 2D ) . Similar to parental rhTrim5α , Trim-NUP153C located to the cell cytoplasm ( Figure S1A and S1B ) and prevented HIV-1 from completing reverse transcription ( Figure S1C–F ) , suggesting that it likely recognizes the HIV-1 CA core in the cytoplasm shortly after viral entry . We conclude that although NUP153C in the context of the Trim5 protein likely engages HIV-1 CA earlier than endogenous NUP153 protein , the novel fusion nonetheless affords the analysis of the NUP153-CA interaction in the context of HIV-1 infection . Due to the marginally greater level of restriction imparted by the internally tagged construct , the Trim-HA-NUP153C variant was used in subsequent experiments . Subjecting a panel of divergent retroviral reporter viruses to Trim-NUP153C inhibition further validated the readout for intracellular CA core recognition . Primate lentiviruses SIVmac , SIVagmSab , SIVagmTan , and HIV-2 were similarly sensitive to Trim-NUP153C inhibition ( Figure 3A ) . Though EIAV was also sensitive , not all lentiviruses were: neither bovine immunodeficiency virus ( BIV ) nor feline immunodeficiency virus ( FIV ) was inhibited by Trim-NUP153C . The more distantly related α-retrovirus Rous sarcoma virus ( RSV ) was also unresponsive . To correlate the results of Trim-mediated restriction of virus infection to direct protein binding , a subset of the sensitive ( EIAV ) and nonresponsive ( MLV and FIV ) CAN proteins was purified following their expression in bacteria . EIAV CAN bound NUP153C as efficiently as HIV-1 CAN , whereas binding to either MLV or FIV CAN was significantly less efficient ( P<0 . 01 ) ( Figure 3B and 3C ) . Reliance on NUP153 during retroviral infection was compared with CA-NUP153C binding ( Figure 3A ) by correlating percent infectivity in the face of NUP153 knockdown [19] ( repeated here using HOS cells; Figure 3D ) . The resulting Spearman rank coefficient of 0 . 673 was statistically significant ( P = 0 . 039 ) ( Figure 3E ) . Mutations within NUP153C were made to decipher the components of NUP153 critical for binding . As the HIV-1 restriction assay was higher throughput than the expression and purification of separate NUP153C proteins , we first engineered mutations within the Trim-NUP153C fusion construct . Since the starting fusion construct contained the entire ∼580 amino acid NUP153C , we generated cell lines stably expressing Trim fusion proteins with roughly quarter-size deletions of NUP153C , and determined the extent to which these constructs inhibited HIV-1 and EIAV infection , using MLV and FIV as negative controls ( Figure 4A and 4B ) . Relative levels of HIV-1 and EIAV infection were compared to ease the interpretation of results to Trim-NUP153C mediated restriction; parental Trim-NUP153C yielded an HIV-1 to EIAV infectivity ratio of ∼0 . 41 ( Figure 4B ) . Deletion of residues 896 to 1045 at the N-terminus of NUP153C resulted in a construct that potently inhibited HIV-1 infection to a level ∼8 fold greater than the full-length construct , yet lost the ability to inhibit EIAV , yielding an HIV-1 to EIAV infectivity ratio of ∼0 . 01 ( Figure 4B ) . Contrastingly , deletion of C-terminal residues 1350 to 1475 resulted in a protein still capable of inhibiting EIAV infection to a level comparable to the full-length construct , yet incapable of inhibiting HIV-1 infection beyond the level of the control viruses , resulting in an infectivity ratio of ∼4 . 70 . These effects were specific to sequences deleted in the preceding constructs , as neither internal deletion noticeably perturbed the original Trim-NUP153C restriction pattern; both constructs displayed the same slight advantage to inhibit HIV-1 infection over EIAV , with HIV-1 to EIAV infectivity ratios similar to the full-length construct . Western blotting confirmed that each deletion construct was expressed at roughly similar levels ( Figure 4C ) . The mapping of the HIV-1 binding determinant on NUP153C to residues 1350–1475 by Trim-mediated restriction notably coincides with our preliminary identification of the region C-terminal to residue 1198 using CA-NC tubes and HA-tagged NUP153 deletion constructs ( Figure 1B ) . We next focused on the initial quarter of NUP153C for its importance in mediating restriction of EIAV infection . Stable cell lines expressing only the first quarter of NUP153C fused to the Trim RBCC , as well as smaller derivatives of the NUP153C sequence , were generated ( Figure 5A ) . Residues 896–949 , which yielded the smallest construct capable of restricting EIAV infection ( Figure 5B ) , harbored only two of the 29 FG motifs present within NUP153C . The importance of these FG motifs in mediating EIAV restriction was tested by substituting four consecutive alanine residues for each corresponding FKFG sequence . The combination octa-alanine 903A/924A Trim-NUP153C mutant construct lost its ability to inhibit EIAV infection despite being expressed at a level equal to or greater than unmodified Trim-NUP153C ( Figure 5B and 5C ) . The 903A/924A mutant moreover retained potent HIV-1 restriction . Separate mutation of each motif revealed 924-FKFG-927 as the dominant FG sequence for mediating EIAV restriction . Sequence components of NUP153C that mediated restriction of HIV-1 infection were investigated next . Attempts to recover cells expressing the responsible C-terminal quarter of NUP153C ( residues 1350–1475 ) fused to Trim RBCC were unsuccessful . We instead undertook the alternative strategy to internally delete segments of residues 1350–1475 from the full-length Trim-NUP153C construct ( Figure 5D ) . Deletion of residues 1410–1447 selectively diminished inhibition of HIV-1 without affecting EIAV , yielding an increased HIV-to-EIAV infectivity ratio of 2 . 7 , while deletion of residues 1350–1410 did not drastically alter the ratio from that observed with the full length construct ( Figure 5E and 5F ) . As residues 1410–1447 contained only one FxFG and one FxF motif , these were mutated to alanine residues , initially in the context of the Δ1350–1410 construct . Combinatorial alteration of both tetra- and tri-peptides reduced restriction of HIV-1 without significantly affecting EIAV restriction ( HIV-1/EIAV infectivity ratio = 2 . 29 ) . Separate mutation showed this effect was largely , if not entirely due to 1415-FTFG-1418 , and the 1415A mutation largely prevented restriction of HIV-1 in the full-length construct as well ( infectivity ratio = 2 . 17 ) . Combined , these results highlight the importance of FG motifs for Trim-NUP153C mediated restriction of HIV-1 and EIAV infection . Moreover , different FG motifs appear to selectively recognize HIV-1 versus EIAV CA proteins . We subsequently tested for Trim-NUP153C FG motif recognition of EIAV and HIV-1 CA proteins in vitro . HA-tagged NUP153C or analogous quarter deleted fragments expressed in 293T cells were used as bait for pull-down by various his-tagged retroviral CAN proteins ( Figure 6A ) . The construct lacking residues 1045–1198 was expressed far less than the other constructs , and was not interpreted . As expected ( Figure 3B ) , none of the constructs bound MLV CAN to levels greater than those observed with beads alone . Consistent with the results from Trim-NUP153C restriction , the protein lacking residues 1350–1475 was selectively bound less well by HIV-1 CAN . Contrastingly , EIAV CAN bound all of the fragments tested , including the fragment that lacked residues 896–1045 . We further tested whether HIV-1 CAN binding was traceable to specific FG motifs . HA-NUP153C containing the 1415-FTFG-1418 tetra-alanine replacement bound HIV-1 CAN essentially as well as the unmutated fragment ( Figure 6B ) . Since we observed strongly diminished binding when the last quarter of HA-NUP153C was deleted ( Figure 6A ) , we next mutated all 7 of the FG motifs within this segment to alanines ( HA-NUP153C7×FG/A ) . The combination of these mutations selectively abrogated binding of HA-NUP153C to HIV-1 CAN; importantly , effective binding of the mutant protein to EIAV CAN was retained ( Figure 6B ) . Decreased Δ1350–1475 and 7×FG/A mutant binding to HIV-1 CAN was also observed with purified NUP153C proteins . HIV-1 CAN bound purified NUP153C ( 0 . 5 µM ) in a dose-dependent manner , revealing a corresponding Kd of ∼28 . 3 µM at half-maximal saturation ( Figure 6C ) . Although CAN displayed some affinity for NUP153CΔ1350–1475 and NUP153C-7×FG/A , the shapes of these linear response curves were notably different from the unmutated protein , and half-maximal saturation was not reached under these assay conditions . We hypothesized that differing states of CA multimerization might contribute to the partially overlapping specificities observed in the CAN pull-down ( Figure 6A and 6B ) versus Trim-NUP153C restriction ( Figure 5E ) assays . To test this , assembled CA-NC tubes were substituted for monomeric CAN protein . Under these conditions , the 1415A mutant protein displayed significantly diminished binding , similar to the effects observed with the 7×FG/A and Δ1350–1475 mutant proteins ( P<0 . 01 ) ( Figure 6D ) . These findings seemingly agree with the results of the Trim-NUP153C mediated restriction assays ( Figure 4B and 5E ) . We and others previously observed that various CA mutant viruses exhibit altered sensitivity to NUP153 knockdown [18] , [19] . We next characterized an expanded set of CA mutant viruses for altered sensitivity to Trim-NUP153C restriction . Mutants were selected based on prior descriptions of pre-integrative defects during HIV-1 infection . Alteration of CA residue ( s ) Pro38 , Glu45 , Thr54/Asn57 , or Gln63/Gln67 can effect core stability [13] , [38] , [53] , [54] , whereas Thr54 , Asn57 , Lys70 , Asn74 , Gly89 , Pro90 , Ala92 , Gly94 , and Thr107 mutants can alter dependencies on various host proteins , including CPSF6 , TNPO3 , NUP358 , CypA , or NUP153 [18] , [19] , [40] , [41] , [55]–[58] . As a number of these mutants exhibit drastically diminished overall levels of infectivity ( Figure 7A , top ) , an unrelated IN mutant virus ( D167K ) , which infects cells at ∼8% of the level of wild-type ( WT ) HIV-1 [59] , was included to control for our ability to reproducibly measure restriction at reduced viral titers . While the IN mutant virus was as sensitive as the WT virus to Trim-NUP153C restriction , a number of CA mutant viruses exhibited significantly reduced susceptibility ( P<0 . 001 ) ( Figure 7A , bottom ) . Included among these were CypA and NUP358 CHD binding mutants G89V and P90A [40] , [55] , as well as mutants E45A , T54A/N57A , N57A , N57D , Q63A/Q67A , Q67A , K70R , and N74A . As these CA mutant viruses could resist Trim-NUP153C restriction for any number of reasons , we tested for direct binding defects by pulling down NUP153C with correspondingly purified CAN mutant proteins . Residue Asn57 was critical for binding , as mutant proteins T54A/N57A , N57A , and N57D were strongly diminished in their abilities to pull down NUP153C ( Figure 7B ) . Although not critical for binding , both Lys70 and Asn74 appeared to participate: mutation of Lys70 to arginine diminished binding while mutation to alanine enhanced binding; contrastingly , mutation of Asn74 to alanine diminished binding , while mutation to aspartic acid enhanced binding to NUP153C . The Q63A/Q67A mutation marginally diminished binding by ∼1 . 3 fold . This binding hierarchy was also observed for HA-NUP153C protein expressed in mammalian cells , with Asn57 again proving key for the interaction , and mutants K70A and N74D yielding hyper-binding activity ( Figure S2 ) . Overall , CA mutant viral sensitivities to Trim-NUP153C restriction correlated well with CAN mutant binding to NUP153C protein in vitro ( Figure 7C ) . As we predict that mutant viruses that require NUP153 for infection also bind NUP153C , we compared the sensitivities of CA mutant viruses to NUP153 knockdown with their susceptibility to Trim-NUP153C mediated restriction . We observed that CA mutant viruses that require endogenous NUP153 for infection were also sensitive to Trim-NUP153C mediated restriction . A strong correlation supported this relationship across the entire panel of CA mutant viruses ( Figure 7D ) . This included NUP153C loss-of-binding mutants T54A/N57A , N57A and N57D , which retained approximately 85% , 102% and 58% of their infectivity , respectively , upon NUP153 knockdown . Residues Asn57 , Lys70 , and Asn74 , highlighted in our binding assays , surround a hydrophobic pocket within CAN formed by α helices 3 and 4 , and this pocket has been shown to be the binding site of the small molecule inhibitor PF74 [42] ( Figure 8A ) . To probe potentially similar binding modes , we tested whether PF74 could compete for HA-NUP153C binding to CAN ( Figure 8B ) . PF74 indeed competed for binding to CAN in a dose-dependent manner , with an IC50 of ∼13 . 6 µM . While PF74 binds WT and N74D CAN proteins similarly [41] , the small molecule was less effective at competing for HA-NUP153C binding to N74D CAN , yielding an IC50 of 145 . 3 µM , perhaps due to the increased binding observed between NUP153C and N74D CAN ( Figure 7B and S2 ) . PF74 does not bind K70A mutant CAN [41] , and accordingly did not compete for HA-NUP153C binding to this mutant protein ( Figure 8B ) . This same pocket also engages the mRNA splicing cofactor CPSF6 [18] , [41] , [51] , which was first implicated in HIV-1 biology by the ability for an exogenously expressed C-terminal truncation mutant CPSF6358 to restrict PIC nuclear import [18] . Though vastly differing molecules , co-crystal structures of PF74-CAN and CPSF6 ( residues 313–327 ) -CAN complexes revealed that each exhibit nearly identical insertions of methyl benzyl residues ( Phe321 in the case of CPSF6 ) within the helix 3/4 pocket , in both cases forming two hydrogen bonds with the carboxamide side-chain of CA residue Asn57 ( Figure 8A and 8C ) . Based on these observations , we tested whether purified NUP153C could compete with full-length CPSF6 protein for binding to CA . HA-tagged CPSF6 expressed in 239T cells was incubated with HIV-1 CA-NC tubes prior to centrifugation through a 20% sucrose cushion . CPSF6 pelleted only in the presence of CA-NC ( Figure 8D ) . This interaction indeed required binding to the CAN hydrophobic pocket , as excess PF74 counteracted it . We additionally observed that co-incubation with purified NUP153C significantly diminished CPSF6 binding ( P<0 . 0001 ) by ∼7 fold as compared to the level observed in the absence of competing factors . This competition was specific , as NUP153C mutants Δ1350–1475 and 7×FG/A , both of which exhibit greatly diminished binding to CA-NC ( Figure 6 ) , were significantly less effective at competing for CPSF6 binding ( P<0 . 05 ) ( Figure 8D ) . CA residues that mediate binding to NUP153C and CPSF6 were further analyzed by assessing CA mutant sensitivities to restriction by the artificial restriction factor Trim-CPSF6358 ( Figure 8E ) , a larger derivation of the Trim-CPSF6 fusion proteins previously tested [51] . Though conferring similar levels of restriction , far fewer of the CA mutant viruses were able to resist Trim-CPSF6358 inhibition ( Figure 8F , red data points ) compared to Trim-NUP153C ( black points ) . CypA binding mutants G89V and P90A were partially resistant to Trim-CPSF6358 restriction , whereas N57A , N74A , and N74D in large part conveyed full resistance . The N57A and N74D changes were notably previously shown to prevent binding of CAN to the CPSF6 peptide [41] . Interestingly , changes at Asn57 and Asn74 conferred distinguishable resistance profiles to Trim-NUP153C versus Trim-CPSF6358: both conservative N74D and non-conservative N74A changes rendered HIV-1 resistant to Trim-CPSF6358 , while only N74A rendered the virus partially resistant to Trim-NUP153C ( Figure 8F ) . Contrastingly , both conservative and non-conservative Asn57 changes prevented Trim-NUP153C recognition , while the conservative N57D mutant remained as sensitive to Trim-CPSF6358 restriction as the WT virus . The breadth of CA mutants restricted by Trim-CPSF6358 in HOS cells appeared to contrast with prior results of CPSF6358-mediated restriction of HIV-1 in Hela cells , where many of the same CA mutations conferred resistance to inhibition [60] . We confirmed these phenotypes in HOS cells , where we observed that many additional CA mutant viruses resist CPSF6358-mediated restriction ( Figure S3 ) . Many of the CA mutant viruses selectively resistant to CPSF6358 over Trim-CPSF6358 restriction were also insensitive to endogenous NUP153 knockdown , resulting in a moderate correlation between CA mutant sensitivities to CPSF6358 restriction and NUP153 knockdown ( Figure 8G ) . PF74 destabilizes the structure of purified CA cores and can inhibit reverse transcription , which likely accounts for at least part of its antiviral activity [61] . We assessed whether PF74 could additionally antagonize NUP153C engagement by CA in the context of HIV-1 infection , given the caveat that we could not unambiguously correlate data from protein binding assays ( Figure 8B ) with effects from PF74-induced capsid destabilization in cells . PF74 exhibited dose-dependent inhibition of WT HIV-1 and N74D CA mutant viral infection , but had no effect on CA mutant T54A/N57A , which lacks the critical Asn57 side-chain necessary for PF74 binding [41] ( Figure 9A , upper panel; results replotted below to reveal EC90 values under conditions of Trim-NUP153C restriction ) . WT virus was noticeably less sensitive to PF74 in Trim-NUP153C expressing cells , with an EC90 of 5 . 65 µM as opposed to 0 . 65 µM in control cells ( Figure 9A ) . The competing effect of PF74 on Trim-NUP153C inhibition seemingly occurred between the concentrations of 0 . 1 and 1 µM ( light green shading in Figure 9 ) , as the inhibition curves within the two cell lines were nearly superimposable outside of these concentrations . N74D CA mutant virus also exhibited a shift in the PF74 EC90 concentration in Trim-NUP153C cells , though this occurred at higher PF74 concentrations than with the WT virus . Interestingly , an almost identical effect was observed with WT virus when PF74 was titrated onto NUP153 knockdown cells; the EC90 shifted from 0 . 54 µM to 5 . 41 µM , with the same window of concentrations likely accounting for the discrepancy in inhibition curves ( Figure 9B ) . While the exact mechanism of NUP153 antagonism – direct , or indirect through the alteration of the state of CA multimerization – is difficult to discriminate , the nearly superimposable interference profiles of PF74 in Trim-NUP153C expressing and NUP153 knockdown cells support the relevance of the Trim-NUP153C restriction assay as a surrogate readout for the engagement of endogenous NUP153 protein by the virus . Retroviral CAN proteins exhibit remarkable similarity in secondary and tertiary structure despite marked differences in primary sequence [62] , [63] . With the exception of HIV-1 residue Gln67 , the previously described polar residues flanking the helix 3/4 hydrophobic pocket ( Asn57 , Lys70 , and Asn74 in HIV-1 ) exhibit variability across divergent retroviruses ( Figure 10A , yellow boxes ) . While HIV-2 and SIVmac only differ at these positions with Arg69 in place of HIV-1 Lys70 , EIAV exhibits greater difference: Leu71 corresponds to HIV-1 Lys70 , and EIAV Asp58 and Asp75 correspond to HIV-1 Asn57 and Asn74 , respectively ( Figure 10B ) . These differences may account for the resistance of EIAV to inhibition by PF74 ( Figure 10C ) and CPSF6358 [51] , which we confirmed using HOS cells expressing Trim-CPSF6358 ( Figure 10D ) . As Asp58 exhibits similar physiochemical properties as its HIV-1 Asn57 counterpart , we mutated this as well as residue Asp75 to test their contributions to NUP153C binding . Similar to HIV-1 mutant N57A , EIAV CA mutant D58A was poorly infectious ( Figure 10E ) , and the corresponding CAN protein was unable to pull down appreciable levels of NUP153C protein ( Figure 10F ) . Contrastingly , EIAV CA mutant D75A behaved similar to WT EIAV ( Figure 10E and 10F ) . The Trim-NUP153C sensitivities of these viruses corresponded with the binding profiles of their CAN proteins: D58A was completely insensitive to Trim-NUP153C mediated restriction , while D75A remained as sensitive as WT EIAV ( Figure 10G ) . Changes at Asn57 in HIV-1 CA have previously been associated with cell cycle dependence: T54A/N57A infection was attenuated in both chemically arrested cell lines and non-dividing primary macrophages [13] , [64] , and the N57A mutant virus was recently shown to lose infectivity upon chemical arrest of Hela cells [40] . We confirmed the importance of Asn57 , as well as other previously observed cell cycle dependent phenotypes , with our panel of CA mutant viruses; alanine substitution of residue Glu45 , Thr54 , Asn57 , or Gln67 rendered the virus significantly sensitive to growth arrest ( Figure 11A and 11B ) . Notably , we found even the conservative N57D substitution rendered the virus as , if not more sensitive , than these mutants to growth arrest . A handful of CA mutant viruses have been described to be sensitive to cell cycle arrest in Hela cells in a CypA-dependent manner [64] , [65] . We found N57A and N57D CA mutant viruses to remain highly cell cycle dependent when the interaction with CypA was blocked by the addition of cyclosporine during infection ( Figure 11C ) . Based on the coincident NUP153-insensitive and cell cycle dependent phenotypes of Asn57 mutant viruses , we tested the association between NUP153 requirement and cell cycle dependency in the context of our expanded panel of mutant viruses . We observed a moderately strong inverse correlation between requirement for NUP153 and cell cycle dependence during infection ( Figure 11D ) . Notably , of the viruses tested in our panel , all of the ones that were cell cycle dependent were NUP153 independent . The correlation however was not absolute , as N74D , G89V , P90A , and T107N mutant viruses did not require NUP153 for infection yet remained cell cycle independent . There was a moderate correlation between cell cycle dependence and Trim-NUP153C resistance ( Figure 11E ) . We observed a moderate to low correlation between cell cycle dependence and CPSF6358 mediated restriction , and no correlation with Trim-CPSF6358 mediated restriction ( Figure 11F and 11G ) . These results reveal that cell cycle dependence is associated with NUP153 independence , and that this relationship likely depends on CA-NUP153 binding .
Green fluorescent protein ( GFP ) -tagged NUP153 expressed in animal cell lysate was recently shown to co-sediment with HIV-1 CA-NC tubes in vitro [66] . We confirmed this observation for HA-tagged protein , and extended it by using purified recombinant protein to demonstrate direct binding between the FG-enriched NUP153C and the HIV-1 CA NTD . Mutation of CA residue Asn57 , Lys70 , or Asn74 , which each flank the hydrophobic pocket between CA α-helices 3 and 4 , perturb binding of NUP153C protein to HIV-1 CAN . Furthermore , NUP153C competes with PF74 and CPSF6 for binding , both of which engage the same pocket . Notably , co-crystal structures between HIV-1 CAN and the latter two molecules exhibit an almost identically situated benzyl ring within the hydrophobic cavity , with the amide nitrogen and carbonyl oxygens of this phenylalanine moiety each forming a hydrogen bond with the side chain of Asn57 [41] ( Figure 8 ) . This observation , in conjunction with our finding that FG motifs within NUP153C strongly contribute to binding with CAN , suggest that the phenylalanine moieties of specific FG motifs found in NUP153C likely take on a similar conformation during binding . We accordingly speculate that hydrogen bonding with Asn57 underlies the FG motif interaction , as both N57A and N57D mutations abrogated binding . While originally described to support CPSF6 binding [41] , the high degree of amino acid conservation within this region of CA amongst primate lentiviruses likely also reflects the requirement for binding to NUP153 during virus infection [19] . Supporting the relevance of the NUP153-CA interaction , both a divergent set of retroviruses and a targeted set of CA missense mutants exhibited significant correlations between CA binding to NUP153C – either tested in vitro or inferred through Trim-NUP153C recognition – and requirement for endogenous NUP153 protein during infection ( Figures 3 and 7 ) . Notably , loss-of-binding CA mutant viruses T54A/N57A , N57A , and N57D infected cells independent of endogenous NUP153 expression . The relationship between NUP153 binding and host factor requirement was consistent with PF74 sensitivity as well; while potentially mediated through an indirect effect on uncoating , PF74 interfered with Trim-NUP153C restriction at the same concentrations that it antagonized the inhibition of infection caused by NUP153 knockdown ( Figure 9 ) . Woodward and colleagues reported that ectopically-expressed NUP153C protein imparted an approximate twofold defect on HIV-1 infection [67] , a result we did not reproduce despite efficient NUP153C expression ( Figure 2 ) . By contrast , appending NUP153C to the RBCC domains of rhTrim5α resulted in potent HIV-1 restriction , allowing us to infer the results of NUP153C binding to the CA shell during virus infection . NUP153 has been shown to bind HIV-1 IN [67] , and though we observed minimal binding ( ≤1% of input IN recovered by GST-NUP153C pull-down; Figure S4 ) , it was comparably weaker than our findings with HIV-1 CA ( 30–40% of input NUP153C recovered ) , and was less correlative with lentiviral requirement for endogenous NUP153 ( Figure 3D ) as FIV IN bound more robustly than HIV-1 IN to NUP153C in our hands ( Figure S4 ) . Thus , while NUP153 may bind more than one HIV-1 determinant , our results are consistent with a direct interaction between NUP153C and viral CAN underlying the requirement for NUP153 during HIV-1 infection . Different FG motifs within Trim-NUP153C mediated restriction of EIAV versus HIV-1 infection ( Figure 5 ) . Contrastingly , correspondence to protein binding in vitro was less strict: NUP153CΔ896–1045 effectively bound EIAV CAN , though this deletion variant could not inhibit EIAV as a Trim-fusion . The 1415-FTFG-1418 tetra-alanine mutant , which lost the ability to inhibit HIV-1 as a Trim-fusion , was little if at all reduced for pull-down by HIV-1 CAN , though alteration of all seven FG motifs in the last quarter of NUP153C yielded a protein greatly deficient for binding to HIV-1 CAN ( Figure 6 ) . Because the tetra-alanine 1415-FTFG-1418 NUP153C mutant protein was significantly defective for binding assembled CA-NC tubes , we infer that this specific FG motif is particularly important for NUP153C binding to multimerized CA . We believe our results reflect the nature of the NUP153C-CA interaction during HIV-1 infection . Unlike a bimolecular interaction between two well-folded domains , each with a single binding site , NUP153C exhibits no appreciable secondary structure and is highly repetitive in its primary sequence , particularly for phenylalanine-based FG motifs . As FG sequences appear to dictate NUP153C binding to CAN , each of the 29 motifs may possess some affinity for CAN . Residues adjacent to the phenylalanine , such as glycine , may allow proper flexibility to fit into the helix 3/4 pocket for Asn57 engagement . We envision that residues peripheral to the motif may also contribute intra- and inter-molecular interactions . This interpretation is consistent with the mode of CPSF6 binding: the CPSF6 FG dipeptide ( residues Phe321 and Gly322 ) is critical for CPSF6358 mediated restriction [51] , while upstream residues Val314 and Leu315 fulfill important secondary roles through engaging additional hydrophobic patches located between CAN helices 4 and 5 . CPSF6 backbone functional groups also interact to varying degrees with the side-chains of CA residues Asn74 , Thr107 , Lys70 , and Gln67 [41] ( Figure 8C ) . Given this model , we hypothesize that differential accessibility of the CAN helix 3/4 pocket might factor into the contrasting binding specificities observed between monomeric and oligomerized CA: while the pocket is likely exposed as a soluble NTD fragment in the pull-down assay , it may be less available within the context of a multimeric CA array . The CTD of the adjacent CA subunit covers the bottom edge of the cavity ( Figure S5A ) , and the interacting NUP153C peptide would need to reach into the crevice between CA subunits , past the cyclophilin-binding loop , and under helix 5 to reach the pocket ( Figure S5B ) . These steric requirements likely limit the number of NUP153C FG motifs capable of forming energetically favorable interactions with the oligomerized CA array present on the viral core . Accordingly , alterations in the rate or extent of CA core uncoating may alter engagement of NUP153 during infection . Though both Trim-NUP153C and Trim-CPSF6358 presumably encounter CA cores shortly after entry ( Figure S1 ) [51] , Trim-CPSF6358 restricted the hyperstable CA mutant viruses E45A and Q63A/Q67A [13] , [38] , [53] , [54] , [68] as efficiently as WT cores , while Trim-NUP153C was less effective at restricting either of these mutants ( Figure 8F ) . Both mutant CAN proteins in large part retained NUP153C binding in vitro ( Figure 7B ) , suggesting that some CA disassembly may be needed for interaction with NUP153C within cells . These hyperstable CA mutant viruses acutely depend on the cycling state of the cell . Comparison between cell cycle dependence and NUP153 reliance resulted in a strong negative correlation within the panel of CA mutant viruses ( Figure 11D ) . This correlation was stronger than the relationship between CPSF6358 sensitivity and cell cycle dependence ( Figure 11F ) , suggesting a more direct association with NUP153 engagement . Consistent with this , the CPSF6 binding mutant N74D was cell cycle independent , while N57A and N57D mutant viruses , both of which are also defective for NUP153 binding , were sensitive . While the direct cause of cell cycle dependence is not clear , we suspect that defective NUP153 binding is a key contributor , and that hyper-stable CA cores may phenotypically mimic this effect . The HIV-1 CA side-chains involved in NUP153C binding overlap those identified to interact with CPSF6 . Accordingly , we found recombinant NUP153C able to compete with CPSF6 for binding to HIV-1 CAN in vitro ( Figure 8D ) . The overlapping binding sites suggest these proteins may take interdependent or even antagonistic roles during infection . While the role of endogenous CPSF6 protein in HIV-1 infection is unknown , the cytoplasmic CPSF6358 truncation variant potently restricts HIV-1 [18] , [41] , [51] , [60] , [69] . Like Trim-NUP153C , CPSF6358 may interact with the viral core shortly after entry; both a Trim-fusion protein containing the CPSF6 binding domain [51] , and the cytoplasmically expressed CPSF6375 isoform [70] , prevent the completion of reverse transcription . Interestingly , CPSF6358 does not inhibit reverse transcription , but instead blocks HIV-1 nuclear import . Additionally , CPSF6358 appears to inhibit only a subset of CA mutant cores that it is able to bind [60] ( Figure 8F and S3 ) . This may reflect an incomplete understanding of the mechanism of CPSF6358 restriction , which could involve antagonism of the CA-NUP153 interaction ( Figure 8G ) . While CPSF6358-mediated stabilization of the CA core [60] , [69] may contribute to the nuclear import defect , it seems possible that direct competition for NUP153 binding may also be at play . Small molecules that bind the helix 3/4 pocket in CA may also preclude NUP153 binding during HIV-1 infection . At least part of the PF74 antiviral mechanism occurs before nuclear entry , as it can inhibit HIV-1 reverse transcription [61] . Yet , its altered dose-response curve in NUP153 depleted cells suggests that it antagonizes CA engagement of NUP153 as well ( Figure 9 ) . Notably , recently identified pyrrolopyrazolone small molecules BI-1 and BI-2 bind the same pocket , yet inhibit HIV-1 nuclear import [71] . As both PF74 and the pyrrolopyrazolone compounds bind CAN with similar affinity [41] , [71] , we speculate that the contrasting phenotypes observed with these small molecules is due to their similar abilities to directly compete with host factors that bind the helix 3/4 pocket juxtaposed with their differential affects on CA core stability: PF74 destabilizes incoming capsids [61] , whereas BI-1 and BI-2 can stabilize capsid structures in vitro [71] . TNPO3 depletion is proposed to mis-localize endogenous CPSF6 into the cytoplasm , recreating the phenotypes conferred by CPSF6358 expression [60] . Resembling our observations with NUP153 knockdown cells , infection of TNPO3 depleted cells exhibited a similar profile of reduced sensitivity to PF74 [72] . A number of HIV-1 CA mutant viruses beyond the above noted hyperstable mutants exhibited resistance to NUP153 knockdown while maintaining near WT levels of protein binding . Many of these are defective for binding to other HIV-1 CA interacting host factors [40] , [41] . Interestingly , the N74D mutant is defective for CPSF6358 binding [18] , [41] but is competent to bind both NUP153 ( Figure 7B ) and NUP358 [40] , yet requires neither of these nucleoporins for infection . Similarly , CA mutant virus N57A is NUP153 and CPSF6 binding-defective , detectably binds the NUP358 CHD , yet does not require NUP358 expression for infection [40] . The CypA and NUP358 CHD binding mutants G89V and P90A are also comparably less sensitive to NUP153 depletion and CPSF6358 restriction [19] , [40] , [60] . Though the mechanistic reasons for these relationships are unclear , loss of binding to one of these factors appears to render HIV-1 independent of the others . Current data and the known biology of this protein suggest NUP153 is likely important for trafficking the HIV-1 PIC through the nuclear pore and into the nucleus [16] , [19] , [73] ( Figure 12 ) . The viral nucleoprotein complex is likely to initially dock to the NPC by engaging NUP358 [73] through its CHD [40] , though additional determinants of NUP358 engagement [74] , such as FG motifs , may also participate . While intact HIV-1 cores are too large to enter the central channel , CA cores in various stages of disassembly may enter far enough for remaining CA to be accessed by the FG domains present in NUP153C . Bypassing NUP153 affects downstream steps of infection , including integration , likely by altering the chromosomal environment encountered by the PIC; NUP153 depletion or the N57A CA mutation shifts integration events away from gene-dense regions of chromatin [40] , [66] , [75] , similar to the effects observed from NUP358 or TNPO3 knockdown , or the N74D CA mutation [40] , [76] . CA binding with NUP153C may serve two distinct roles during infection . Firstly , NUP153 may be responsible for physically translocating the PIC by engaging CA molecules that may associate with it . The relatively short half-life of NUP153 at the NPC may contribute to the release of the PIC into the nucleoplasm [20] . Secondly , as even a partially disassembled core could remain too large to efficiently pass through the NPC channel , CA interaction with NUP153 may be required to fully uncoat the viral core at the NPC and prime the PIC for nuclear import . Indeed , CA cores have been shown to dock to NPCs for several hours before PIC nuclear translocation [43] . CA oligomers may interact with a limited subset of NUP153C FG motifs , while increased CA pocket accessibility from progressive core disassembly may expose monomeric CA to an expanded number of NUP153C FG repeats . While CA mutant viruses such as N74D may uncoat differently and circumvent this mechanism without penalty in various transformed cell lines , they apparently incur steep costs to infectivity in other cell types , such as primary macrophages [40] , [77] . Divergent viruses have adapted to use NUP153 for their own devices . Our results suggest EIAV , which presents different amino acid residues flanking the CAN hydrophobic pocket , may have either retained , or convergently evolved NUP153 binding . Hepatitis B virus ( HBV ) has also been reported to bind NUP153 during its nuclear transport; though the HBV core is sufficiently small to traverse the NPC channel , NUP153 binding is believed to be important for HBV core conformational change and genome release within the nuclear basket [78] . This interaction may also require binding to NUP153 FG motifs , as both of the broadly defined regions mapped for HBV capsid binding overlapped parts of NUP153C . The S . pombe homolog of NUP153 , Nup124p , is important for Tf1 retrotransposition and binds the Tf1 Gag protein , though binding did not necessarily appear to map to Nup124p FG motifs [79] , [80] . Perhaps akin to effects caused by differential HIV-1 uncoating , the requirement for Nup124p appears to be related to the state of Tf1 Gag multimerization [81] . It remains to be determined whether FG motifs found on additional nucleoporins may bind HIV-1 and aid its infection . While the effects of NUP98 depletion on HIV-1 infection are relatively modest [18] , [66] , [82] , this protein can also co-sediment with HIV-1 CA-NC tubes in vitro [66] . Similarly , the GLFG-motif enriched domain of S . cerevisiae NUP100 , predicted to be orthologous to vertebrate NUP98 , binds Ty3 Gag protein [83] . Alternatively , while CA binding with the CHD is proposed to determine the requirement for NUP358 [40] , it remains to be seen whether its own FG domains may bind CA and contribute to its function during infection . While numerous FG nucleoporins exist , it is likely that certain characteristics specific to NUP153 , including its length , flexibility , and its relatively high dissociation rate from the NPC , along with its spatial location around the nuclear rim of the NPC , makes this protein particularly important for lentiviral passage through the nuclear pore .
Infection assays utilized single-round viruses carrying either GFP or luciferase reporter genes . GFP-based constructs included HIV-1 , EIAV , BIV , RSV , FIV , MLV , HIV-2 strain ROD , simian immunodeficiency viruses from Macaca mulatta clone 239 ( SIVmac ) , Chlorocebus sabaeus ( SIVagmSab ) , and Chlorocebus tantalus ( SIVagmTan ) , all described previously [19] , [84] . HIV-1 CA mutations were generated through site-directed mutagenesis of the HIV-1NL4-3-based pHP-dI-N/A packaging plasmid [85] ( AIDS Research and Reference Reagent Program [ARRRP] ) , which were co-transfected with either pHI-vec2 . GFP or pHI-Luc transfer vectors [19] . Human NUP153 ( accession number NM_005124 . 3 ) , or deletion mutants thereof , fused to N-terminal HA tags were expressed from the pIRES-dsRed Express HA-NUP153 expression vector [19] . Trim-fusion constructs , which were built within pLPCX-rhTrim5α-HA [46] , were created by engineering a BamHI site at nucleotides corresponding to residues 301 and 302 of rhesus Trim5α , and ligating the digested vector with sequences encoding HA-NUP153C , NUP153C-HA , or CPSF6358-HA [18] . Truncated Trim-HA was engineered by modifying the Trim-HA-NUP153C vector to encode two stop codons at the nucleotides corresponding to the first two residues of NUP153C . All deletion and missense mutations within animal-cell expressed NUP153C were engineered by site-directed-mutagenesis of plasmids pLPCX-Trim-HA-NUP153C or pLPCX-HA-NUP153C . HIV-1NL4-3 CA carrying C-terminal his and FLAG tags was expressed from the pET11a-HIV1-CA-his-flag bacterial expression vector . The vector encoding tagged HIV-1 CA NTD ( pET11a-HIV1-CAN-his-flag ) was constructed by removing nucleotides corresponding to CA residues 147–231 from the full-length expression vector . Bacterial expression vectors for FIV CA were generated by amplifying DNA encoding full-length FIV CA ( residues 1–223; pET11a-FIV-CA-his ) or NTD only ( residues 1–140; pET11a-FIV-CAN-his ) from pFP93 [86] with a primer encoding a C-terminal his-tag , and ligating with digested pET11a DNA . pET22b-based bacterial expression vectors encoding C-terminally his-tagged N-tropic MLV ( pET22b-NMLV-CA-his ) and EIAV ( pET22b-EIAV-CA-his ) were obtained from the laboratory of Dr . Joseph Sodroski , and CA NTDs were engineered from full-length his-tagged constructs by removing nucleotides corresponding to residues 133–263 of N-MLV , and residues 149–231 of EIAV , by site-directed mutagenesis . The construct pGEX2T-GST-NUP153C , which encoded GST fused to NUP153C , was created by deleting sequences that encoded residues 1–895 from the full-length human protein pGEX2T-hNUP153 bacterial expression vector [87] . Plasmid pGEX2T-his-GST-pp-NUP153C , which was utilized to obtain tag-free NUP153C protein , was derived from pGEX2T-GST-NUP153C by sequentially engineering a PreScission protease site between GST and NUP153C , and then appending a his-tag N-terminal to GST . A stop codon was introduced at the nucleotides corresponding to residue 1350 to generate the Δ1350–1475 truncation mutant . The 7×FG/A mutant NUP153C was engineered for bacterial expression by swapping the WT sequence present in pGEX2T-his-GST-pp-NUP153C with a fragment encoding NUP153 residues 1178–1475 amplified from pLPCX-HA-NUP153C-7×FG/A . All coding sequences were verified through DNA sequencing . 293T and HOS cells were cultured in Dulbecco's modified Eagle's medium ( DMEM ) ( Invitrogen ) supplemented with 10% fetal bovine serum ( FBS ) , 100 U/ml penicillin , and 0 . 1 mg/ml streptomycin . HOS cells stably transduced with MLV-derived LPCX transfer vectors were subsequently selected and maintained with 2 µg/ml puromycin . Approximately 25 , 000 HOS cells seeded per well of a 24-well plate were transfected the next day with a final concentration of 40 nM siNUP153#1 ( GGACTTGTTAGATCTAGTT ) or a mismatch control of siNUP153#1 , referred to as siControl ( GGTCTTATTGGAGCTAATT ) ( Dharmacon ) [19] , using RNAiMax ( Invitrogen ) according to the manufacturer's instructions . Dividing or cell cycle arrested cells were collected at the time of infection , fixed in 70% ethanol , and incubated for 30 min at room temperature in staining solution [0 . 1% Triton X-100 , 0 . 2 mg/ml RNAse A ( Invitrogen ) , and 20 µg/ml propidium iodide in phosphate-buffered saline ( PBS ) ] . The cells were washed , and cellular DNA content was assessed with a FACSCanto flow cytometer ( Becton , Dickenson and Company ) equipped with FACSDIVA software . Viral vector particles were produced by transfecting 293T cells in 10-cm plates with 10 µg total of various ratios of the aforementioned virus production plasmids using CaPO4 . The cells were washed 16 h after transfection , and supernatants collected from 24 to 72 h thereafter were clarified at 300× g , filtered through 0 . 45 µm filters ( Nalgene ) , and either allotted and frozen or concentrated by ultracentrifugation using an SW32Ti rotor at 50 , 000× g for 2 h at 4°C before freezing . Concentrations of HIV-1 and EIAV CA mutant viral stocks were determined alongside concomitantly produced WT viruses using an exogenous 32P-based assay for RT activity [88] . HOS cells ( 10 , 000 or 2 , 500 ) seeded onto 48-well or 96-well plates , respectively , were infected with various reporter viruses . Percentages of GFP-positive cells were determined 48 h post-infection ( hpi ) using a FACSCanto flow cytometer equipped with FACSDIVA software . GFP reporter experiments comparing retroviral genera were performed with virus inoculates adjusted to yield ∼40% GFP-positive cells in control samples . HIV-1 or EIAV CA mutant viruses ( 2×105 RTcpm ) were used to infect 96-well and 48-well plates of cells , respectively . HIV to EIAV infectivity ratios were calculated after initially normalizing to the average of MLV and FIV negative control viruses to account for slight differences in overall infectivities between stable cell lines . Cyclosporine ( 5 µM , Sigma ) was introduced to cells at the time of infection . Cell cycle arrest experiments were performed by plating 2 , 500 control or 5 , 000 experimental cells treated with 5 µM Etoposide-phosphate ( Calbiochem ) the day before infection . Quantitative PCR for the accumulation of viral late reverse transcripts and 2-long terminal repeat ( LTR ) -containing circles were performed as previously described [19] . The quantitation of early reverse transcripts was performed using primers AE989 and AE990 and Taqman probe AE995 [89] . Cells stably expressing Trim-fusion proteins were lysed in Buffer A [25 mM Tris-HCl pH 7 . 5 , 200 mM NaCl , 1 mM DTT , 1 mM EDTA , Complete protease inhibitor ( Roche ) ] and sonicated for 30 s total with a misonix sonicator . Protein concentration of the bulk lysate was determined by Bradford assay ( Bio-rad ) , and 75 µg of each sample were electrophoresed through Tris-glycine polyacrylamide gels , and transferred onto polyvinylidene fluoride membranes . Transiently expressed HA-tagged proteins were either extracted with buffer H [10 mM Tris-HCl pH 8 . 0 , 10 mM KCl , 1 . 5 mM MgCl2] followed by repeated freeze-thaws , or Triton buffer [50 mM Triethanolamine , 250 mM NaCl , 0 . 5% Triton X-100] , and pelleted in a microcentrifuge for 20 min at 21 , 000× g at 4°C . Stably expressing cells were also fractionated by initial lysis in Buffer F1 [20 mM Tris-HCl pH 7 . 5 , 10 mM NaCl , 1 . 5 mM MgCl2 , 0 . 25% Triton X-100 , and Complete Protease Inhibitor] , followed by centrifugation at 6 , 000× g . The supernatant was removed as Fraction 1 , and the process was repeated , with the resulting supernatant combined with the previous fraction . The subsequent pellet was resuspended in Buffer F2 [Buffer F1 lacking Triton X-100 , but with 0 . 5% sodium deoxycholate and 1% Tween-40] , and pelleted at 21 , 000× g for 15 min . The supernatant was removed as Fraction 2 , and pellet was resuspended in 1× Turbo DNase buffer and treated with 40 U/ml Turbo DNase ( Ambion ) for 10 min at 37° C . Two parts fraction 1 , one part fraction 2 , and one part of the remaining fraction ( Fraction 3 ) were each mixed with sample loading buffer and separated on Tris-glycine polyacrylamide gels . Exogenously expressed HA-tagged proteins were detected using a 1∶4 , 000 dilution of HRP-conjugated 3F10 antibody ( Roche ) or 1∶4 , 000 dilution of mouse 16b12 antibody ( Covance ) and developed with ECL prime ( GE Healthcare ) or Femto ( Thermo Scientific ) detection reagents . NUP153 , NUP62 , and NUP358 were detected with a 1∶4 , 000 dilution of mouse monoclonal antibody mab414 ( Abcam ) . HRP-conjugated mouse anti-β-actin antibody or mouse anti-α-tubulin antibody ( Abcam ) were used at 1∶10 , 000 dilutions to confirm equal lysate loading across samples . His-tagged HIV-1 CA was detected with 1∶15 , 000 α-his HRP ( Clontech ) . CA-NC protein was detected with 1∶5 , 000 mouse anti-p24 antibody ab9071 ( Abcam ) . Histone H3 was detected with 1∶2 , 000 rabbit histone H3 antibody #9715 ( Cell Signaling Technology ) . All mouse and rabbit primary antibodies were detected using 1∶10 , 000 dilutions of anti-mouse or anti-rabbit HRP secondary antibodies ( Dako ) . Cells transduced with empty LPCX vector or stably expressing HA-epitope tagged rhTrim5α , NUP153C , or fusion proteins thereof , were cultured on eight-well chamber slides . After 24 h , the cells were fixed with 4% paraformaldehyde for 10 min , washed , and permeabilized with PBS containing 0 . 5% Triton X-100 . The permeabilized cells were blocked with PBS containing 10% FBS for 1 h , and stained with a 1∶100 dilution of anti-HA antibody 16b12 . After a 30 min wash with PBS , the cells were incubated for 1 h with a 1∶1 , 000 dilution of an Alexa Fluor 555 conjugated goat anti-mouse IgG antibody ( Invitrogen ) , as well as Hoechst 33342 ( Invitrogen ) diluted to a concentration of 0 . 2 µg/ml . After an additional 30 min wash with PBS , the samples were covered with mounting medium [150 mM NaCl , 25 mM Tris pH 8 . 0 , 0 . 5% N-propyl gallate , and 90% glycerol] . The processed samples were analyzed on a Nikon Eclipse spinning disk confocal microscope . GST-NUP153C was expressed in BL21-CodonPlus ( DE3 ) -RILP E . coli ( Agilent ) grown in 2× YT media and induced at an optical density of 0 . 8 at 600 nm ( OD600 ) with 1 mM isopropyl β-D-1-thiogalactopyranoside ( IPTG ) for 1 h at 18°C . Cells were pelleted at 6 , 000× g , and sonicated for 5 min in buffer A . The lysate was centrifuged for 30 min at 35 , 000× g , and the pellet was resuspended in buffer B [1 M NaCl , 25 mM Tris-HCl pH 7 . 5 , 1 mM DTT , 1 mM EDTA , Complete protease inhibitor] with a dounce homogenizer . The lysate was again spun at 35 , 000× g , and the pellet was resuspended in Buffer C [2 M Urea , 200 mM NaCl , 25 mM Tris-HCl pH 7 . 5 , 1 mM DTT , 1 mM EDTA , Complete protease inhibitor] with a dounce homogenizer . After a last centrifugation at 35 , 000× g , the supernatant was collected and incubated with glutathione-sepharose beads ( GE Healthcare ) overnight at 4°C . The beads were washed with buffer D [200 mM NaCl , 25 mM Tris-HCl pH 8 . 0 , 1 mM DTT , 1 mM EDTA , Complete protease inhibitor] , and the protein was eluted with buffer D containing 20 mM glutathione . Eluted protein was dialyzed against buffer D to remove excess glutathione , spin concentrated by ultrafiltration through a 10 , 000 nominal molecular weight limit ( NMWL ) Amicon filter ( Millipore ) , and flash frozen in liquid nitrogen for storage at −80°C . BL21-CodonPlus ( DE3 ) -RILP E . coli transformed with pGEX2T-his-GST-pp-NUP153C was grown to an OD600 of 0 . 8 , followed by induction with 1 mM IPTG for 1 h at 18°C . Cells were pelleted at 6 , 000× g , and sonicated for 5 min in buffer A . The lysate was then centrifuged for 30 min at 35 , 000× g , and the pellet was resuspended in buffer E [6 M Urea , 200 mM NaCl , 25 mM Tris-HCl pH 7 . 5 , 1 mM DTT , 1 mM EDTA , Complete protease inhibitor] with a dounce homogenizer . The lysate was then centrifuged at 40 , 000× g for 1 h , and the resulting supernatant was incubated with Ni-NTA conjugated agarose beads ( Qiagen ) overnight . The beads were then initially washed with buffer E , and then progressive dilutions of buffer E into cleavage buffer [150 mM NaCl , 50 mM Tris-HCl pH 7 , 1 mM DTT , 1 mM EDTA] ( 3∶1 , 1∶1 , 1∶3 ) , with a final wash in cleavage buffer only , each supplemented with 7 . 5 mM imidazole . The beads were incubated with 5 U of PreScission protease ( GE Healthcare ) for 48 h . The supernatant , which was cleared with 0 . 1 volumes of Ni-NTA beads and glutathione-sepharose beads each at 4°C to remove uncleaved protein and residual PreScission protease , was centrifuged at 21 , 000× g for 15 min at 4°C . The resulting supernatants were quantitated following fractionation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) and staining with SYPRO Ruby ( Invitrogen ) or Coomassie blue , as compared to a standard curve of bovine serum albumin ( BSA ) , using ChemiDoc MP imager ( Bio-Rad ) with Image Lab software . Cleaved full length NUP153C was recovered at ∼50% purity , with the predominant contaminants degradation products of the full-length protein , as inferred through comparison with western blots using mab414 antibody . Recombinant HIV-1 CA-NC was expressed in E . coli , purified , and assembled into CA-NC complexes as previously described [29] . Expression constructs encoding full-length HA-NUP153 or fragments thereof were transiently transfected into 293T cells using X-tremeGENE 9 DNA transfection reagent ( Roche ) . Cells were collected after 48 h , lysed with successive freeze thaws in buffer H , and clarified by centrifugation at 21 , 000× g at 4°C . CA-NC complexes were incubated with clarified lysates for 1 h at room temperature before ultracentrifugation for 30 min at 100 , 000× g through a 50% sucrose cushion prepared in PBS . The resulting pellet was resuspended in 1× sample loading buffer , and fractionated by SDS-PAGE . Experiments with purified proteins were stained with Coomassie blue or SYPRO Ruby , while experiments using a lysate component were developed by western blot . Quantification was performed with a ChemiDoc MP imager using Image Lab software . His-tagged HIV-1 , MLV , EIAV , and FIV capsid proteins , either full length or NTD only , were expressed in BL21-CodonPlus ( DE3 ) -RILP E . coli , grown to an OD600 of 0 . 6 , and induced for 4 h with 1 mM IPTG . Bacteria pelleted by centrifugation were resuspended in Buffer A , sonicated , and centrifuged at 30 , 000× g for 30 min . The supernatants were incubated overnight with Ni-NTA-sepharose beads , eluted with 20 mM Tris-HCl pH 8 . 0 , 200 mM imidazole elution buffer , and dialyzed into Tris Buffer ( 20 mM Tris-HCl pH 8 . 0 ) . Dialyzed protein was concentrated by ultrafiltration through a 10 , 000 NMWL filter , centrifuged at 21 , 000× g , and the resulting soluble protein was quantitated by spectrophotometer . Pull-down assays with full-length CA or CAN proteins were performed by mixing 20 µl reactions with the following final concentrations: 0 . 02 µl packed volume Ni-NTA beads per µl ( 0 . 4 µl total ) , 20 µM CA , 25 mM Tris-HCl pH 8 . 0 , and either 0 . 5 µM purified NUP153C with 0 . 1% NP-40 and 150 mM NaCl , or 100 µg 293T lysate overexpressing HA-tagged NUP153 with 0 . 25% Triton X-100 and 200 mM NaCl . Mixtures were left rocking at room temperature for 1 h after which the samples were washed twice in buffer M [25 mM Tris-HCl pH 8 . 0 , 150 mM NaCl , and 0 . 1% NP-40] , allowing the beads to settle by gravity , and finally resuspended in 1× sample loading buffer . Saturation curves were achieved by mixing 3 µl packed volume Ni-NTA beads with 0 . 5 µM purified WT or mutant NUP153C , 150 mM NaCl , 25 mM Tris-HCl pH 8 . 0 , and 0 . 1% NP-40 , with half-log increments of HIV-1 CAN from 2 µM to 200 µM . Both bead-bound and supernatant fractions were separated by SDS-PAGE and stained with SYPRO Ruby , with the percent of NUP153C protein bound calculated at each concentration . The Kd of NUP153C binding was calculated by subtracting nonspecific binding to beads and fitting the resulting data-points with a one-site saturation binding nonlinear regression using Prism6 software ( GraphPad ) . CPSF6 competition experiments were performed through modification of the CA-NC protocol . Assembled CA-NC was diluted to a final concentration of 0 . 8 µM in the reaction mixture . WT or mutant NUP153C was added to a final concentration of 4 µM , along with 10 µg total 293T extract expressing C-terminally HA-tagged CPSF6 , resulting in final concentrations of 170 mM NaCl , 75 mM Tris-HCl pH 8 . 0 , and 0 . 025% Triton X-100 . Mixtures ( 20 µl ) were incubated at room temperature for 20 min , after which they were spun over a 30 µl 20% sucrose cushion in a microcentrifuge at 21 , 000× g for 20 min at 4°C . The resulting pellet was resuspended in sample loading buffer and separated by SDS-PAGE . Western blotting with p24 antibody indicated ∼35% of input CA-NC was recovered in the pellet . CA-NC co-sedimentation assays with WT or FG mutant NUP153C were performed similarly , but were instead centrifuged over a 25% sucrose cushion . His-tagged HIV-1 and FIV IN [84] and GST [90] were expressed and purified as previously described . Pull-down of soluble IN was performed as previously described for GST-LEDGF326–530 [59] , with 0 . 8 µM of his-tagged HIV-1 or FIV IN incubated with 0 . 47 µM GST-NUP153C or control GST pre-bound to glutathione-sepharose beads in PD buffer [150 mM NaCl , 25 mM Tris-HCl pH 7 . 4 , 5 mM MgCl2 , 5 mM DTT , 0 . 1% NP-40] . BSA ( 5 µg ) was included as an additional specificity control . The reaction was incubated for 2 h at 4°C , after which the beads were washed 4 times with PD buffer , and settled each time for 20 min in the absence of centrifugation . Recovered samples were resolved by SDS-PAGE , and stained with Coomassie blue and western blotted with anti-his antibody . Dependencies between variables were assessed by Spearman rank correlation using Prism6 software . The significances of pair-wise differences were calculated by Student's t-test ( two-tailed ) using Prism6 software .
|
Lentiviruses such as HIV-1 possess mechanisms to bypass the nuclear envelope and reach the nuclear interior for viral DNA integration . Numerous nuclear transport proteins are important for HIV-1 infection , suggesting the viral nucleoprotein complex enters the nucleus by passing through nuclear pore complexes . HIV-1 was previously found to utilize cellular nucleoporin ( NUP ) 153 protein in a manner determined by the viral capsid protein . Here , we show HIV-1 capsid directly binds NUP153 in a phenylalanine-glycine motif-dependent manner; such motifs form the general selectivity barrier that restricts transport through the nuclear pore . We find that NUP153 binds a hydrophobic pocket found on capsid proteins from both primate and equine lentiviruses , suggesting an evolutionary predilection for this interaction . The pocket on HIV-1 capsid also binds phenylalanine moieties present in a small molecule inhibitor of HIV-1 infection , as well as a separate host factor implicated in the nuclear import pathway . We found that these molecules compete for NUP153 binding , providing insight into their mechanisms of action during HIV-1 infection . These results demonstrate a previously unknown interaction important for HIV-1 nuclear trafficking , and posit direct binding of viral capsids with phenylalanine-glycine motifs as a novel example of viral hijacking of a fundamental cellular process .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Nucleoporin NUP153 Phenylalanine-Glycine Motifs Engage a Common Binding Pocket within the HIV-1 Capsid Protein to Mediate Lentiviral Infectivity
|
Natural genetic transformation is widely distributed in bacteria and generally occurs during a genetically programmed differentiated state called competence . This process promotes genome plasticity and adaptability in Gram-negative and Gram-positive bacteria . Transformation requires the binding and internalization of exogenous DNA , the mechanisms of which are unclear . Here , we report the discovery of a transformation pilus at the surface of competent Streptococcus pneumoniae cells . This Type IV-like pilus , which is primarily composed of the ComGC pilin , is required for transformation . We provide evidence that it directly binds DNA and propose that the transformation pilus is the primary DNA receptor on the bacterial cell during transformation in S . pneumoniae . Being a central component of the transformation apparatus , the transformation pilus enables S . pneumoniae , a major Gram-positive human pathogen , to acquire resistance to antibiotics and to escape vaccines through the binding and incorporation of new genetic material .
Natural transformation , first discovered in Streptococcus pneumoniae [1] , is observed in many Gram-negative and Gram-positive bacteria [2] . It increases bacterial adaptability by promoting genome plasticity through intra- and inter-species genetic exchange [3] . In S . pneumoniae , a major human pathogen responsible for severe diseases such as pneumonia , meningitis and septicemia , transformation is presumably responsible for capsular serotype switching and could therefore reduce the efficiency of capsule-based vaccines after a short period [4] . In this species , it occurs during a genetically programmed and differentiated state called competence that is briefly induced at the beginning of exponential growth . During this competent state , pneumococci secrete a peptide pheromone called Competence-Stimulating-Peptide ( CSP ) [5] , which spreads competence in the pneumococcal population . Interestingly , in S . pneumoniae , some antibiotics and DNA-damaging agents induce competence , which would act as an alternative SOS response and ultimately increases bacterial resistance to external stresses [6] . During transformation , environmental DNA is bound at the surface of competent cells and transported through the cell envelope to the cytosolic compartment . This process has been mostly studied in the Gram-positive bacterium Bacillus subtilis with additional information coming from studies in S . pneumoniae [7] , [8] . In both species , a DNA translocation apparatus mediates the transfer of DNA through the cellular membrane . In S . pneumoniae , it is composed of ComEA , EndA , ComEC and ComFA . Incoming double-stranded DNA would bind the membrane receptor ComEA . One DNA strand crosses the membrane through ComEC while the endonuclease EndA degrades the other strand . On the cytoplasmic side , ComFA , an ATPase that contains a helicase-like domain , would facilitate DNA internalization through ComEC . Once inside the bacterium , single-stranded DNA is either integrated into the chromosome by RecA-mediated homologous recombination or entirely degraded . Strikingly , all transformable Gram-positive bacteria also carry a comG operon that resembles operons encoding Type IV pili and Type II secretion pseudopili in Gram-negative bacteria , as well as a gene encoding a prepilin peptidase homolog , pilD [7] . In B . subtilis and S . pneumoniae , comG and pilD genes are exclusively expressed in competent cells and are essential for transformation [9] , [10] , [11] . In S . pneumoniae , the comG operon encodes a putative ATPase ( ComGA ) , a polytopic membrane protein ( ComGB ) and five prepilin candidates named ComGC , ComGD , ComGE , ComGF and ComGG ( Figure 1A and B and table S1 ) . By homology with Type IV pili , it is generally proposed that these proteins could be involved in the assembly of a transformation pseudo-pilus at the surface of competent cells [7] , [8] , [12] . So far , two studies show that a large macromolecular complex containing ComGC can be found at the surface of competent B . subtilis cells [9] , [12] . In this complex , ComGC subunits appear to be linked together by disulfide bridges [9] . All the other ComG proteins and the PilD homolog , ComC , are necessary for the formation of this complex [9] . It was proposed that this complex could correspond to the transformation pseudo-pilus . Despite these first clues , no transformation appendage could be directly visualized at the surface of any competent Gram-positive bacterium . In addition , the function of the ComG proteins during transformation remains unclear . Mutations in the cytosolic ComGA protein abolish DNA binding at the surface of both B . subtilis and S . pneumoniae [13] , [14] , [15] . This strongly suggests that the ComGC-containing macromolecular complex detected at the surface of competent B . subtilis cells could bind DNA . However , it was recently shown that ComGA is the only ComG protein essential to the initial DNA binding at the surface of competent B . subtilis cells [14] . This protein would interact with an unknown DNA receptor at the surface of competent cells while the other ComG proteins would only be required at a later stage during transformation . In this study , we provide the first direct evidence for the existence of a transformation pilus in a Gram-positive bacterium . We discovered a new appendage at the surface of competent pneumococci that we could visualize using immuno-fluorescence and electron microscopy . Competent cells harbor one or a few appendages that are morphologically similar to Type IV pili found in Gram-negative bacteria . We were able to purify this pilus and showed that it is essentially composed of the ComGC pilin . We also demonstrate that pilus assembly is required for transformation . As we provide direct evidence that the transformation pilus binds extracellular DNA , we propose it is the primary DNA receptor at the surface of competent pneumococci .
Mechanical shearing is frequently used to release bacterial surface appendages and to study their protein composition [16] . To see if ComGC was part of a macromolecular complex at the surface of competent S . pneumoniae , we adapted the method to this bacterium and raised antibodies against the purified soluble domain of ComGC . Using this antibody , we showed that ComGC could be detected by immunoblotting in the sheared fraction of competent bacteria ( Figure 2A ) . While ComGC level in the cell fraction was not affected , no ComGC could be found in the sheared fraction in a comGA knockout mutant ( Figure 2A ) . These data strongly suggest that ComGC is part of an extra-cellular appendage and that ComGA is necessary to its assembly . We inserted a FLAG tag at the C-terminus of ComGC to directly visualize the competence-induced appendages by immuno-fluorescence . It was not possible to insert the sequence encoding the tag at the comGC locus on the chromosome because comGC and comGD genes overlap in the comG operon . Therefore , a copy of comGC encoding a C-terminally FLAG-tagged ComGC ( ComGC-FLAG ) was integrated ectopically into the chromosome of S . pneumoniae under the control of a competence-induced promoter [17] . The transformation efficiency was not affected in this strain ( Figure 2B ) . Using anti-FLAG antibodies , we could show by immuno-fluorescence that almost all the cells appeared to harbour one or a few ComGC foci or distinct fluorescent appendages ( Figure 3A and B; Figure S1A ) . Due to sample preparation , many broken appendages were also found in the background . No preferential location of the foci/appendages at the cell surface was observed . They are absent in comGA knockout cells ( Figure 3A ) . Note that anti-ComGC antibodies were not able to label the competent cells . They probably recognize epitopes that are masked when ComGC is included in the appendages . Using electron microscopy , we observed filaments attached to the cell surface of negatively stained competent pneumococci ( Figure 4A and B ) . These flexible filaments are 5–6 nm in diameter . Their length could reach up to 2–3 micrometers ( Figure 4A ) . A maximum of 2–3 filaments per cell could be observed . Their average length was difficult to assess because they break easily into smaller fragments during sample preparation . Using the ComGC-FLAG expressing strain , we confirmed by immunogold-labelling that they contain ComGC ( Figure 4C ) . Appendages were then purified using anti-FLAG affinity chromatography after mechanical shearing . Appendage fragments of between 50 and 500 nm in length were observed by electron microscopy ( Figure 5A ) , showing that these filamentous structures do not disassemble during purification . SDS-PAGE analysis of the purified fraction showed that ComGC is the major component of the appendages ( Figure 5B ) . Using whole protein mass profiling by high-resolution mass spectrometry [18] , we could only detect ComGC and ComGC-FLAG in the purified material ( Figure 5C ) , confirming that ComGC is the major constituent of these appendages . Monoisotopic mass measurements of intact proteins and top-down fragmentation using a variety of activation techniques confirmed that the ComGC prepilin is cleaved after the alanine residue in position 15 and that the first amino acid of the mature protein is methylated , presumably by PilD ( Figure S2 ) . Indeed , PilD homologs in Gram-negative bacteria catalyze this post-translational modification of the Type IV pilins [19] . No other post-translational modification was detected in ComGC . Other proteins , including other ComG proteins , were not detected in the purified material by the methods used in this study . This suggests that these proteins are either absent , present in very low amount within the appendage or weakly bound to it and lost during sample preparation . These morphological and biochemical features are typical of Gram-negative Type IV pili . Therefore , we propose that the competence-induced appendage observed in S . pneumoniae belongs to the Type IV pilus family . It was important to determine whether these competence-induced pili were involved in transformation . Indeed , it was previously shown that S . pneumoniae and B . subtilis comGA knockout could not be transformed ( Figure 2B ) [9] [13] . In this study , we were able to show in S . pneumoniae that comGA mutant cells lack pili ( Figure 2A and 3A ) . It was enticing to conclude that competence-induced pili assembly is essential for transformation . However , it was recently shown that a comGA mutation could have a pleiotropic effect on transformation in B . subtilis [14] . Therefore , we generated a comGC mutant in S . pneumoniae in which the conserved glutamic acid in position 5 was substituted by an alanine ( Figure 1B ) . Such a substitution was shown to impair Type IV pilus assembly in Gram-negative bacteria [20] . ComGC cellular level was not affected by this point mutation ( Figure 2A ) . Our results show that this mutant strain could not assemble any pilus and that it was defective for transformation ( Figure 2A and B ) . Therefore we conclude from the analysis of both comGA and comGC ( E5A ) mutants that the assembly of the competence-induced pilus is required for transformation . The nature of the primary DNA receptor at the surface of transformable Gram-positive bacteria is not known . It is generally proposed that the transformation pseudopilus would bind extracellular DNA at the surface of competent Gram-positive bacteria [8] , [21] . However , this hypothesis has never been confirmed experimentally . Using affinity purification , we show that DNA naturally released in the culture medium co-fractionates with the purified pili . No DNA could be found in the purified fraction in absence of the pilus ( Figure 6A ) . These data were a first hint suggesting that DNA present in the environment could bind to the transformation pilus . However , it was not clear if this binding was related to the transformation process or fortuitous . By using specific electron microscopy methods [22] , we visualized DNA directly bound to the transformation appendage after adding linear double stranded DNA ( dsDNA ) to competent bacteria . Long stretches of dsDNA interacting with the transformation pilus were observed with clearly visible multiple contact points ( Figure 6 B–E ) . Interestingly , it was extremely difficult to see DNA bound on the pilus in the reference bacteria ( R1501 strain ) , which are known to internalize exogenous DNA quickly [23] . On the other hand , in ComEC and comFA mutants , we could easily observe bound DNA on transformation pili . These strains are defective for DNA uptake and accumulate bound DNA at their surface [13] . Given that the dsDNA was added in large excess , no difference between the reference and mutant strains should be observed if DNA binding on the pili was a coincidental event . The fact that the uncoupling of DNA binding and uptake processes facilitates the observation of the DNA/pilus interaction is a strong indication that DNA binding on the transformation pilus is related to the transformation process .
The pneumococcal transformation pilus represents a newly discovered pneumococcal surface structure . For a long time , no external appendage could be found at the surface of S . pneumoniae cells while many electron microscopy images were published in the literature . Recently , sortase-mediated pili have been discovered in some pathogenic S . pneumoniae strains [24] . To our knowledge , no specific ultrastructural study of competent S . pneumoniae has ever been described . Here , we analysed a laboratory strain that is commonly used to study the transformation process in S . pneumoniae [10] [13] . In this strain , competence can be induced in a rapid and synchronous manner upon addition of synthetic CSP in the medium of an exponentially growing culture [5] , [25] . To make sure that the appearance of the transformation pilus is a common feature of competent pneumococci and not a mere one-off property of our reference strain , we observed negatively stained G54 and CP strains by electron microscopy . The G54 strain is a wild-type clinical strain . The CP strain is a laboratory strain that has a different genetic background than our reference strain [26] . In both cases , transformation pili were observed at the surface of competent cells ( Figure S3 ) . Therefore , we think that transformation pili are found at the surface of most , if not all , pneumoccocal strains , including clinical strains . The pneumococcal transformation pilus is morphologically very similar to Type IV pili found in many Gram-negative bacteria . Its major component , the ComGC pilin , is cleaved and probably methylated by a PilD homolog . We therefore propose that the transformation pilus is a bona fide Type IV pilus . Since its length can reach up to 2–3 µm , we think that the “pseudo-pilus” appellation does not apply to the pneumococcal transformation appendage . By comparison , the type II secretion pseudo-pilus is just 50–100 nm long [27] . The transformation pilus is the first Type IV pilus clearly observed in a Gram-positive bacterium . So far , Type IV pilus-dependent gliding motility had been described in Clostridium species [28] . However , no clear picture of this pilus was provided . A recent genomic study show the existence of numerous and diverse Type IV pilus-like operons in a wide range of Gram-positive bacteria [29] . This suggests that many other Type IV-like pili remain to be discovered in these bacteria . The conservation of comG operons argues in favor of the presence of a transformation pilus in all naturally transformable Gram-positive bacteria . However , species-specific variations in pilus length can be anticipated because of variations in thickness of the capsule and/or the cell wall .
Cells were grown at 37°C under anaerobic condition , without agitation , in a Casamino Acid Tryptone medium ( CAT ) up to OD600 = 0 . 3 for stock cultures [40] . After addition of 15% glycerol , stocks were kept frozen at −80°C . For competence induction , cells were grown in CAT supplemented with BSA ( 2 g/L ) , calcium chloride ( 1 mM ) and adjusted to pH = 7 . 8 . Competence was triggered by incubating cells with the Competence Stimulating Peptide ( CSP ) at OD600 = 0 . 1 for 12 min as described previously [40] . For transformation , DNA was then added and transformants were selected on CAT agar plates [17] . Competence was induced following the same protocol in G54 and TCP1251 strains . For transformation efficiency assays , 100 µL of competent bacteria were transformed by the addition of 100 ng of S . pneumoniae R304 genomic DNA ( contains the streptomycin resistance gene str41 ) . Bacteria were plated in presence and absence of streptomycin ( 100 µg/mL final concentration ) and incubated at 37°C overnight before colony counting . The annotated names of the comG genes in different strains of S . pneumoniae are listed in Table S1 . The S . pneumoniae strains used derived from the non-capsulated R6 strain and are listed in Table S2 . The comGC-FLAG gene was cloned by PCR using genomic DNA of pneumococcal R6 strain ( ATCC BAA-255 ) as template . The resulting fragment was digested with NcoI and BamHI and inserted into the same sites of the pCEPx vector [17] . RL001 strain was constructed by transformation of R1501 cells with the pCEPx plasmid containing comGC-FLAG , followed by selection with kanamycin ( Kan ) . RL002 was obtained by transformation of RL001 with R1062 chromosomal DNA and selection with spectinomycin ( Spc ) . For RL003 , a 2 kb genomic fragment of R6 genome containing comGC in the middle was amplified , and the codon 20 was changed from GAG to GTG by cross-over PCR . R1501 was transformed with this modified genomic fragment , and clones were screened by sequencing the comGC gene . Chemically competent Escherichia coli BL21 Star ( Life Technologies ) were used for heterologous production of ComGC soluble domain . The corresponding DNA sequence was amplified from genomic DNA of strain R800 and cloned into pET15b expression vector ( Novagen ) , using NdeI/XhoI . The protein was purified from the soluble fraction using IMAC affinity and gel filtration in 50 mM Tris/HCl pH = 8 , 200 mM NaCl . The anti-ComGC were raised against the purified protein ( Eurogentec ) . Shearing experiments were adapted from Sauvonnet et al . [41] . Competence was induced exactly as described above in a 50 mL culture . Cells were harvested by centrifugation 15 min at 4 , 500 g , 4°C . The pellet was suspended in 1 mL LB and immediately vortexed for 1 min to apply mechanical pressure . The suspension was then centrifuged twice at 13 , 000 g for 5 min to separate the bacteria ( pellet fraction ) from the pilus-enriched supernatant ( sheared fraction ) . The supernatant was then precipitated with 10% trichloroacetic acid for 30 min on ice . Both fractions were loaded on SDS 15% polyacrylamide gels and subjected to electrophoresis and immunoblot with rabbit polyclonal antibodies raised against ComGC soluble domain ( 38–108 ) or anti-FLAG M2 antibody ( Sigma-Aldrich F1804 ) . The pili containing ComGC-FLAG were purified from the sheared fraction of a 1 L culture . Shearing was performed in 2 mL Tris Buffered Saline ( TBS , Tris pH 7 . 6 0 . 05 M , NaCl 0 . 15 M , protease inhibitor cocktail Roche 11873580001 ) and incubated overnight on a rotating wheel at 4°C with ANTI-FLAG M2 affinity resin ( Sigma-Aldrich A2220 ) . After washing with TBS , the pili were eluted by adding 3×FLAG-peptide at 100 µg/mL ( Sigma Aldrich F4799 ) 30 min at room temperature under agitation . To prevent DNA aspecific binding on the ANTI-FLAG M2 affinity resin , the resin was saturated 2 h at 4°C with a 1 . 5 kb PCR fragment ( 20 ng/µL ) . For DNA detection , 20 µL of the eluted pili were run on a 1% agarose gel and stained with SYBR safe ( Life technologies S33102 ) . Competence was induced exactly as described above in a 10 mL culture . Cells were harvested by centrifugation 15 min at 4 , 500 g , 4°C . The pellet was suspended in 60 µL phosphate-buffered saline ( PBS ) ( Sigma-Aldrich P4417 ) . A drop of this suspension was placed on a glow discharged carbon coated grid ( EMS , USA ) for 1 min . The grid was then placed on a drop of PBS-3% formaldehyde , 0 . 2% glutaraldehyde for 10 min , and washed on drops of distilled water . The grids were then treated with 2% uranyl acetate in water . Specimens were examined using a Philips CM12 transmission electron microscope operated at 120 kV . Pictures were recorded using a camera KeenView ( SIS , Germany ) and ITEM software . For immunogold labelling , additional steps were applied after fixation: 3 washes with PBS , PBS–50 mM NH4Cl ( 10 min ) , 3 washes with PBS , PBS with 1% BSA ( 5 min ) , 1 hour incubation with ANTI-FLAG M2 antibody ( Sigma-Aldrich F1804 ) diluted 1/100 in PBS with 1% BSA , 3 washes with PBS-BSA 1% ( 5 min ) , 1 hour incubation with goat anti-mouse antibody ( 5 nm gold particles , BritishBioCell , UK ) diluted 1/25 in PBS containing 1% BSA . S . pneumoniae cells were grown in the same conditions as above for visualization by electron microscopy . Cells were harvested by centrifugation for 15 min at 4 , 500 g , 4°C . The pellet was suspended in 500 µL PBS and directly immobilized on poly-L-lysine-coated coverslips . Samples were fixed for 30 min with 3 . 7% formaldehyde , washed 3 times with PBS containing 1% BSA and incubated on a 100 µL drop of anti-FLAG antibodies ( 1∶300 ) and secondary Alexa Fluor 488- coupled anti-mouse IgG ( Invitrogen ) . Samples were examined with an Axio Imager . A2 microscope ( Zeiss ) . Images were taken with AxioVision ( Zeiss ) and processed in ImageJ [42] . Protein samples were desalted and eluted directly into a 10 µL spray solution of methanol∶water∶formic acid ( 75∶25∶3 ) . Approximately 4 µL was loaded into a coated , medium sized , nano-ESI capillary ( Proxeon ) and introduced into an Orbitrap Velos mass spectrometer , equipped with ETD module ( Thermo Fisher Scientific , Bremen , Germany ) using the off-line nanospray source in positive ion mode . A full set of automated positive ion calibrations was performed immediately prior to mass measurement . The transfer capillary temperature was lowered to 100°C , sheath and axillary gasses switched off and source transfer parameters optimised using the auto tune feature . Helium was used as the collision gas in the linear ion trap . For MSn experiments , ions were selected with a 3 Da window and both CID and HCD were performed at normalised collision energies of 15–25% , with the appropriate HCD charge state set and other activation parameters left as default . For ETD the reagent gas was fluoranthene and the interaction time 10 ms . Supplemental activation was used as noted . The FT automatic gain control ( AGC ) was set at 1×106 for MS and 2×105 for MSn experiments . Spectra were acquired in the FTMS over several minutes with one microscan and a resolution of 60 , 000 @ m/z 400 before being summed using Qualbrowser in Thermo Xcalibur 2 . 1 . Summed spectra were then deconvoluted using Xtract and a , b , c−1 , y , z , z+1 ions assigned using in house software at a tolerance of 5 ppm . N-terminal ions were verified manually . Five microliters of bacterial culture ( wild-type , ΔcomFA or ΔcomEC ) were diluted in 45 µL of Tris 10 mM , pH 8 , NaCl 150 mM . Bacteriophage lambda DNA ( 0 , 1 mg/ml final ) was then added to bacteria . Five µL of mix were immediately adsorbed onto a 600 mesh copper grid coated with a thin carbon film , activated by glow-discharge . After 1 min , grids were washed with 0 , 02% ( w/vol ) uranyl acetate solution ( Merck , France ) and then dried with filter paper . TEM observations were carried out with a Zeiss 912AB transmission electron microscope in filtered crystallographic dark field mode . Electron micrographs were obtained using a ProScan 1024 HSC digital camera and Soft Imaging Software system .
|
Natural genetic transformation , first discovered in Streptococcus pneumoniae by Griffith in 1928 , is observed in many Gram-negative and Gram-positive bacteria . This process promotes genome plasticity and adaptability . In particular , it enables many human pathogens such as Streptococcus pneumoniae , Staphylococcus aureus or Neisseria gonorrhoeae to acquire resistance to antibiotics and/or to escape vaccines through the binding and incorporation of new genetic material . While it is well established that this process requires the binding and internalization of external DNA , the molecular details of these steps are unknown . In this study , we discovered a new appendage at the surface of S . pneumoniae cells . We show that this appendage is similar in morphology and composition to appendages called Type IV pili commonly found in Gram-negative bacteria . We demonstrate that this new pneumococcal pilus is essential for transformation and that it directly binds DNA . We propose that the transformation pilus is an essential piece of the transformation apparatus by capturing exogenous DNA at the bacterial cell surface .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"microbial",
"pathogens",
"biology",
"microbiology",
"bacterial",
"pathogens"
] |
2013
|
A Type IV Pilus Mediates DNA Binding during Natural Transformation in Streptococcus pneumoniae
|
Recent evidence from serum metabolomics indicates that specific metabolic disturbances precede β-cell autoimmunity in humans and can be used to identify those children who subsequently progress to type 1 diabetes . The mechanisms behind these disturbances are unknown . Here we show the specificity of the pre-autoimmune metabolic changes , as indicated by their conservation in a murine model of type 1 diabetes . We performed a study in non-obese prediabetic ( NOD ) mice which recapitulated the design of the human study and derived the metabolic states from longitudinal lipidomics data . We show that female NOD mice who later progress to autoimmune diabetes exhibit the same lipidomic pattern as prediabetic children . These metabolic changes are accompanied by enhanced glucose-stimulated insulin secretion , normoglycemia , upregulation of insulinotropic amino acids in islets , elevated plasma leptin and adiponectin , and diminished gut microbial diversity of the Clostridium leptum group . Together , the findings indicate that autoimmune diabetes is preceded by a state of increased metabolic demands on the islets resulting in elevated insulin secretion and suggest alternative metabolic related pathways as therapeutic targets to prevent diabetes .
Type 1 diabetes ( T1D ) is an autoimmune disease that results from the selective destruction of insulin-producing β-cells in pancreatic islets . The diagnosis of T1D is commonly preceded by a long prodromal period which includes seroconversion to islet autoantibody positivity [1] and subtle metabolic disturbances [2] . The incidence of T1D among children and adolescents has increased markedly in the Western countries during the recent decades [3] and is presently increasing at a faster rate than ever before [4] , [5] . This suggests an important role of environment and gene-environment interactions in T1D pathogenesis . Metabolome is sensitive to both genetic and early environmental factors influencing later susceptibility to chronic diseases [6] . Recent evidence from serum metabolomics suggests that metabolic disturbances precede early β-cell autoimmunity markers in children who subsequently progress to T1D [2] . However , the environmental causes and tissue-specific mechanisms leading to these disturbances are unknown . Given its relatively low disease incidence in the general population and even among subjects at genetic risk [1] , studies on early phenomena of T1D pathogenesis in humans are a huge undertaking as they require long and frequent follow-up of large numbers of subjects [2] , [7] , [8] to be able to go “back to the origins” of the disease once a sufficient number of subjects in the follow-up have progressed to overt disease . In order to effectively prevent this disease it is thus fundamental to identify suitable experimental models that recapitulate findings from such large-scale clinical studies while being amenable to mechanistic studies at the systems level . The non-obese diabetic ( NOD ) mouse is a well characterized model of autoimmune disease [9] which has been widely used in studies of T1D . It is clear that the NOD experimental model does not completely mimic the immune system and T1D pathogenesis in man [10] . Only a fraction of NOD mice progress to disease , with the incidence of spontaneous diabetes being 60%–80% in females and 20%–30% in males [9] . There is thus a stochastic component to diabetes pathogenesis in NOD mice , believed to be due to random generation of islet-specific T cells [11] . The disease incidence does seem to depend on the environment and there is evidence indicating that it is the highest in a relatively germ-free environment [12] and that gut microbiota may affect disease incidence via the modulation of the host innate immune system [13] . Herein we performed a murine study in NOD mice that recapitulated the protocol used in human studies [2] and applied a reverse-translational approach ( Figure 1 ) to ( 1 ) map the lipidomic profiles of T1D progressors in human studies to lipidomic profiles in NOD mice and derive a surrogate marker to stratify mice according to risk of developing autoimmune diabetes , then ( 2 ) perform multiple follow-up studies in NOD mice where metabolic phenotypes , tissue-specific metabolome and transcriptome as well as gut microbiota are characterized in the context of early disease pathogenesis .
Our first objective was to validate whether the NOD mouse was a good model of T1D able to recapitulate the lipidomic-based metabolic phenotypes observed in the longitudinal study of children who later progressed to T1D ( Type 1 Diabetes Prediction and Prevention project; DIPP ) [2] , [8] . Hence we performed a murine study using NOD mice and using a similar protocol as applied in human studies ( Study 1 ) . A total of 70 NOD/Bom mice ( 26 female ) were monitored weekly with serum collection from age 3 weeks until either ( a ) the development of diabetes ( progressor group ) , or ( b ) followed until 36 weeks of age in females and 40 weeks in males in the absence of a diabetic phenotype ( non-progressor group ) ( Figure 2A ) . Similarly as in the DIPP study [2] , we were primarily interested in early pre-diabetic differences of lipidomic profiles , in mice of the same colony , between diabetes progressors and non-progressors . Lipidomic analysis using established methodology based on Ultra Performance Liquid Chromatography ( UPLC ) coupled to mass spectrometry ( MS ) [14] was performed on a complete sample series from 26 female mice ( 12 progressors , 14 non-progressors ) and 13 male ( 7 progressors , 6 non-progressors ) mice , comprising a total of 1172 samples or 30 samples/mouse on an average ( 733 samples from female and 439 from male mice ) , with 154 lipids measured in each sample . When comparing the lipid concentrations of diabetes progressors and non-progressors , the first weeks of life ( 3–10 weeks ) were characterized by an overall lipid-lowering trend among the female progressors , while the period close to the disease onset ( 15 week and older ) was characterized by elevated triglycerides and phospholipids ( Figure 2B ) . No such changes were observed in male mice ( Figure S1 ) . The NOD female progressors had similar levels of glycemia ( Figure 2C ) than the non-progressors but to our surprise the progressors exhibited higher fasting as well as glucose-stimulated plasma insulin levels ( 2-way ANOVA p = 0 . 025 for diabetes progression ) ( Figure 2D ) despite the fact that no body weight differences were evident between progressors and non-progressors at 10 weeks of age ( Figure 2E ) . To account for multiple comparisons , false discovery rates among significantly differing lipids were estimated using q-values [15] , [16] . Together , these results imply that the mice who later progress to diabetes are characterized by enhanced glucose-stimulated insulin secretion ( GSIS ) at an early age or that they are inappropriately insulin resistant for their degree of body weight . In fact this increased GSIS associated to early evolutive stages towards T1D is consistent with our earlier findings indicating that the children who later progress to diabetes are characterized by low serum ketoleucine and elevated levels of the more insulinotropic aminoacid leucine prior to seroconversion to insulin autoantibody ( IAA ) positivity [2] , [17] . In order to systematically investigate similarities between early metabolic phenotypes of autoimmune diabetes progressors in mice and men , we proceeded with comparative analysis of longitudinal lipidomic profiles from female NOD mice and DIPP study children [2] . One inherent challenge in the studies of early disease pathophysiology is variable disease penetration . The metabolic profiles may individually change at different paces , and it is not obvious how they should be compared between individuals or species in the context of the disease process . We recently introduced a concept that the maturation of metabolic profiles with age , such as during normal development or early disease pathogenesis , can be described in terms of metabolic states derived using the Hidden Markov Model ( HMM ) methodology [18] . Instead of observing progression of average lipidomic profiles ( Figure 2B ) , our approach allows for each individual's lipidomic profiles to mature at their own pace . Such individual profiles are captured into a set of progressive HMM states ( described by mean lipid profiles ) using an underlying statistical model . Firstly , we proceeded with the analysis of previously reported longitudinal lipidomic profile data from the DIPP study children [2] . The nested case-control study included 56 T1D progressors and 73 matched non-progressors , comprising a total of 1196 samples or 9 . 3 samples per child on average between birth and the diagnosis of T1D ( in progressors ) . We applied the HMM methodology to study the longitudinal lipidomic profiles in DIPP children and identified a three-state HMM , developed separately for T1D progressors and non-progressors , to describe the progression of metabolic states at early ages ( up to 3 years ) ( Figure 3A , B ) . As expected based on the earlier report [2] , the first state corresponding to the first year of life was characterized by low triglycerides and specific phospholipids in T1D progressors ( Figure 3C ) . In both progressors and non-progressors the states followed a similar time course ( Figure 3B ) , but the first and third states , corresponding to the first and third years of life , respectively , were qualitatively different between the two groups . On the other hand , there were no such clear differences in the second state , corresponding to the second year of life in average . Such multi-stage progression of lipidomic profiles during the first 3 years of life was not detected when examining them cross-sectionally in different age cohorts . We then applied the HMM methodology to study the longitudinal lipidomic profiles of female NOD mice and identified a three-state HMM to describe the progression of metabolic states at early ages ( 3–10 weeks ) ( Figure 4A , B ) . The first two states , corresponding to mean ages of approximately 4 weeks and 6 weeks , respectively , were similar to the first state in DIPP children ( Figure 3C ) and characterized by decreased phospholipids and triglycerides among the progressors ( Figure 4B ) . In the third state , corresponding to approximately 7 weeks of age when a large fraction of the NOD mice already seroconvert to islet autoantibodies [9] , the differences observed in the first two states have disappeared . Instead , the levels of proinflammatory lysophosphatidylcholines ( lysoPCs ) were increased in diabetes progressors ( 1%–10% confidence interval for progressors having higher concentrations , see Figure 4B ) . The similarity of state progression in children ( Figure 3 ) and female NOD mice ( Figure 4B ) presenting with diabetes suggests that the early disease stages as reflected in the lipidomes share similar metabolic perturbations . However , it is always a challenge to compare species exhibiting differences in systemic lipid metabolism as well as diet-related effects on the lipidomic profiles . Consequently the mapping of molecular lipids between mouse and man may not be trivial and our results should be considered qualitative . In order to compare progression of mouse and human lipidomic profiles we applied a mapping algorithm [19] that captures their dependencies across the two species . By using this strategy it is possible to compare lipidomic profiles across species , and we therefore sought for the disease effect by a two-way analysis on progressors/non-progressors vs . men/mice . By this approach , we uncovered associations of functionally and structurally related lipids between the species ( Figure 4C ) and confirmed strong association of diminished phospholipids with the development of the disease at an early age ( HMM state 1 ) . We can thus conclude that the lipid changes seen in children prior to the first seroconversion to autoantibodies are also characteristic of the early changes in female NOD mice progressors . Seroconversion to islet autoantibody positivity is associated with transiently elevated lysoPC serum levels in children who subsequently progress to T1D [2] . Here we measured the IAA levels in NOD mice at 8 weeks of age and similarly confirmed that the IAA-negative ( IAA− ) progressor female NOD mice had elevated lysoPC as compared to IAA- non-progressors at a marginal significance level ( p = 0 . 091 , see Figure 2F ) . Intriguingly , IAA positivity had the opposite association with diabetes progression since the IAA-positive ( IAA+ ) mice with high lysoPC were protected from diabetes ( Figure 2F ) . It can be speculated that due to their opposite association with disease progression IAA measurement in combination with lysoPC may help stratify the NOD mice according to their risk of developing diabetes . We derived a surrogate marker by combining autoantibody positivity and lysoPC concentration , which reasonably well discriminated between progressors and non-progressors ( χ2 = 5 . 75 , Pχ2 = 0 . 0044; Figure S2 ) , with the NOD mice in the assigned “High-risk” group being at 4 . 3-fold higher risk ( 95% lower tolerance bound = 2 . 6 , as calculated from 1000-fold resampling ) of developing autoimmune diabetes as compared to the mice in the “Low-risk” group . In an independent experiment normoglycemic female NOD mice from the same colony as in the first experiment were sacrificed at 8 ( n = 57 ) or 19 ( n = 14 ) weeks of age and blood , liver and pancreas samples were collected ( Study 2 ) . We selected sixteen 8-week-old mice ( seven were IAA+ ) and thirteen 19-week-old mice ( six were IAA+ ) for UPLC-MS based serum lipidomics analysis for subsequent risk stratification using the algorithm described above . Mice at high risk of developing diabetes showed a tendency towards more severe insulitis ( Figure 5A ) , therefore providing an independent validation of the surrogate marker . In parallel liver and islet transcriptomics was performed in 19-week-old mice . When comparing high- and low-risk mice , independent of IAA level , the pathway analysis of islet gene expression data using Gene Set Enrichment Analysis ( GSEA ) [20] expectedly revealed upregulation of several apoptotic and immunoregulatory pathways in the high-risk group ( Table 1 and Table S1 ) . These pathways were associated with the autoimmune status , as they were also upregulated when comparing IAA+ and IAA− mice independent of diabetes risk . In support of our findings from pre-diabetic mice , some of the upregulated gene products of these pathways are in fact known to be implicated in progression to autoimmune diabetes , including CD3 from the CTLA4 pathway [21] , pro-inflammatory chemokine CCL5 ( or RANTES ) from the toll like receptor signalling pathway [22] , [23] , [24] and the IL-7 pathway [25] ( Table S2 ) . Several upregulated pathways in high-risk mice were not associated with the IAA titer . These pathways associated with high risk of developing diabetes were mainly metabolic pathways and included upregulated genes from TCA cycle and glycolysis/gluconeogenesis ( Table 1 ) . In order to directly measure the metabolic products of these pathways , we performed metabolomic analysis of islets using two-dimensional gas chromatography coupled to time-of-flight mass spectrometry ( GC×GC-TOFMS ) [26] . Metabolomics confirmed dysregulation of energy and amino acid metabolism in the islets of high-risk mice ( Figure 5B ) , as several key metabolites of these pathways were found upregulated , including glutamic and aspartic acids , as well as at a marginal significance level all three branched chain amino acids ( BCAAs ) . These elevated amino acids are known insulin secretagogues in β cells [27] . In agreement with this , the insulin signaling pathway was upregulated in the livers of high-risk mice ( Table S3 ) . The top ranking gene in this pathway , Glucose-6-phosphatase , catalytic , 2 ( G6PC2; fold change high- vs . low-risk group +11% , P = 0 . 0034 ) , controls the release of glucose from liver into the bloodstream . However , the animals included in this study , as in the earlier longitudinal study , were normoglycemic and there were no differences in body weight between the two groups . The metabolic changes in β cells and liver can thus explain the observed elevated GSIS in mice at high risk for developing autoimmune diabetes ( Figures 2C–E ) . We recently observed that the serum metabolome of germ-free mice is similar to pre-autoimmune metabolomes of children who later progress to T1D [28] , thus implying that gut microbiota of T1D progressors may be devoid of important constituents or has an impaired function that predisposes the children to T1D . Given the observed similarities of metabolomes of diabetes progressors in mice and men ( Figure 4 ) , we hypothesized that the observed metabolic differences between the high- and low-risk mice may be reflected in differences in their gut microbial composition . We characterized the microbial composition of caecum samples from high- and low-risk mice from Study 2 using predominant bacterial as well as five different bacterial group-specific ( namely Eubacterium rectale – Blautia coccoides group , Clostridium leptum group , Bacteroides spp . , bifidobacteria , and lactobacilli ) denaturing gradient gel electrophoresis ( DGGE ) methods as previously described [29] , [30] . With such an approach to profile microbiota it is possible to detect the phylotypes that constitute over 1% of the specific group in question [29] , [31] . Analysis of the composite dataset , which included all the different group-specific DGGE results , showed that the total bacterial composition did not markedly differ between the groups but was slightly more coherent in low-risk mice than in high-risk mice ( see the deviation bars in Figure S3 ) . In addition , the high-risk mice had significantly diminished diversity of the Clostridium leptum group of the Firmicutes phylum ( Figure 5C ) . There is evidence from clinical studies that insulin resistance is a risk factor for progression to T1D [32] , [33] . It is also known that the NOD genetic background may predispose the mice to insulin resistance [34] . To test for insulin resistance as a potential explanation for the observed metabolic phenotype of high-risk mice , we performed two independent studies in another NOD colony where ( Study 3 ) 36 female NOD/MrkTac mice were tested for GSIS , glucose and insulin tolerance , and plasma leptin between 8 and 11 weeks of age; and ( Study 4 ) 42 female NOD/MrkTac were sacrificed at 10 weeks of age and tested for insulitis , plasma leptin and adiponectin . As before , serum lipidomics and IAA assays were performed to stratify the mice into high- and low-diabetes-risk groups . We confirmed the elevated GSIS in high risk mice ( Figure 6A ) but found no significant difference in glucose responses to intraperitoneal glucose or insulin between the groups ( Figures 6B , C ) , in the Homeostatic model assessment of insulin resistance ( HOMA-IR ) index or GLUT4 expression in white adipose tissue and muscle ( Figure 6D , F ) . In agreement with the results from older mice ( Figure 5A ) and in further support of the surrogate marker applied to stratify the mice according to disease risk , the 10-week old female NOD mice at higher risk of developing diabetes have already signs of more insulitis than their low-risk counterparts , although the average degree of insulitis is mild in both groups ( p<0 . 05 , see Figure 6G ) . Surprisingly , the adipose tissue derived hormones leptin ( p<0 . 05 , see Figure 6H ) and adiponectin ( p<0 . 05 , see Figure 6I ) were both elevated in plasma of high-risk mice despite no significant differences in weight or adiposity ( p>0 . 05 , see Figure 7A–C ) . However , both adiponectin and leptin correlated with the gonadal adipose tissue mass ( Figure 7D , E ) . Given that the metabolic profile is normalized in children following seroconversion to autoantibody positivity [2] , we proposed earlier that the generation of autoantibodies may be a physiological response to early metabolic disturbances . In the present study ( mice from Study 2 ) , we investigated the pathways in the IAA+ low-risk female mice and compared them to all other groups . The IAA+ low-risk mice were characterized by several elevated signaling pathways in the islets , including the IL-4 and IL-6 pathways ( Table 1 ) . IL-4 is known to be protective from diabetes in NOD mouse [35] . Conversely IAA+ low-risk mice had reduced expression of pathways mainly related to mitochondrial function and TCA cycle , BCAA catabolism , beta oxidation and oxidative phosphorylation . It is unclear how downregulation of these pathways may protect against T1D . However downregulation of these pathways will lead to a state of reduced production of reactive oxygen species ( ROS ) [36] which may explain at least in part the conserved β cell functionality . This would offer a potential protective mechanism linking decreased ROS production to the prevention of β cell apoptosis in IAA+ mice which do not progress to diabetes . Our results stress the need for similar studies in terms of protection from diabetes in individuals who seroconverted but did not progress to overt disease .
This study emphasizes the translatability of our previous findings from the large-scale clinical study [2] into the tissue-specific context . Also , our study highlights that specific metabolic disturbances are identifiable early on during the evolutive stages and could potentially be linked to pathogenic mechanisms implicated in the progression to autoimmune diabetes . Although the specific causes , likely to be diverse amongst humans and between the NOD mouse and humans , of these early metabolic disturbances remain to be established , our findings pave the way for studies focused on how the metabolism and the immune system interact in early stages of the disease . The lipidomic profiles associated with progression to T1D in children [2] were similar to early lipidomic profiles in female but not male NOD mice that later progressed to autoimmune diabetes . It is known that female NOD mice are more likely to progress to autoimmune diabetes [9] although the reasons for this are poorly understood . Notably , in humans the T1D incidence is nearly 2-fold higher in men than women [37] . In the present study , we have not pursued the reasons for the gender-specific metabolic differences in NOD mice and have instead focused on studies in female mice since they displayed the similar metabolic patters as observed in human studies . Both in man and mouse , the metabolic states as determined by HMM followed the similar progression in disease progressors and non-progressors ( Figures 3B and 4A ) , but were qualitatively different between these two groups ( Figures 3C and 4B ) . However , notably no major qualitative differences were observed in the second state in the human study and in the third state of the mouse study . These two states correspond to the ages when the first diabetes-associated autoantibodies have appeared in many of the human T1D progressors [2] and NOD mice [9] , respectively . In the DIPP study we have previously shown that the seroconversion to autoantibody positivity appears to normalize the metabolic profiles , suggesting that the immune system may be involved in the metabolic regulation and vice versa . In fact , the metabolic demands of T cells are extraordinary , rivaling that of cancer cells [38] , [39] . For example , differentiating T cells consume 10-fold more glutamine than other cells in the body [39] , and we in fact found that glutamine is diminished in children within a period of months prior to seroconversion [2] . Concentration changes of circulating metabolites as detected in our studies may thus have a direct effect on T-cell function . We therefore hypothesize that the second metabolic state in human progressors , and similarly the third state in NOD mice , reflect the period following the seroconversion when the metabolic profiles have been restored to normal levels via interaction with the immune system . In the NOD mice , this apparent interaction between the immune system and metabolic status is underlined by the opposite association of the IAA and lysoPC ( Figure 2F ) at 8 weeks of age , i . e . , in the age group corresponding to the third state in the HMM model ( Figure 4B ) . Based on this observation , a surrogate marker was derived combining information on IAA positivity and lysoPC concentration ( Figure S2 ) , which was utilized to stratify mice according to risk of developing autoimmune diabetes in subsequent studies . The so-classified high-risk mice had higher degree of insulitis , a histopathological hallmark of progression to diabetes in NOD mice [9] , as assessed in two independent studies ( Figures 5A and 6G ) . While the association of the surrogate marker with established characteristics of progression to T1D validated our approach in the present study , it also suggests that prediction of autoimmune diabetes in NOD mice using combined metabolic and immune markers may be feasible . However , further prospective studies are needed in different NOD strains , similar in design as our Study 1 , to determine and validate such markers . As already demonstrated in our study , the use of such markers sensitive to disease risk may facilitate investigations of early pathophysiological phenomena at a tissue-specific level prior to any symptoms of the disease . Our results indicate that early stages of progression to T1D are characterized by acute increased response to high glucose-stimulated insulin secretion . Furthermore , this response is accompanied by increased concentration of insulinotropic amino acids and other markers of energy metabolism in the islets and more specifically of insulin signaling pathways in the liver . Together with human data [2] , our study provides compelling evidence that increased GSIS is an event that heralds diabetes progression already in pre-autoimmune stages of the disease pathogenesis . In NOD mice , elevated GSIS at young age may be an initial response associated with early insulitis . Our data suggest that this response might reflect a state of insulin resistance; however our insulin tolerance tests do not support this insulin-resistant component in diabetes progressors . One potential link may be increased circulating insulin concentrations-suppressing leptin [40] and insulin-sensitizing adiponectin [41] . Adiponectin is known to be elevated in patients with T1D , but very limited data exist on its levels during the pre-diabetic period [42] . Leptin , however , is known as an important immune regulator [43] . Leptin is a negative regulator of CD4+CD25+ regulatory T cells [44] and promoter of Th1 immune responses [45] , [46] . In fact administration of leptin accelerates autoimmune diabetes in female NOD mice [47] . Of interest , endoplasmic reticulum stress is known to induce leptin resistance [48] . Together , our findings from the studies of female NOD mice at high- and low-risk of T1D within the same colony suggest that elevated leptin in high-risk mice is a consequence of early metabolic stress , and that leptin may play a role in mobilization of deleterious Th1 immune responses characteristic of T1D [49] . This would offer an explanation for the epidemiological findings that obesity [50] and decreased insulin sensitivity [51] are risk factors of T1D . Given the global rise of obesity and related metabolic complications among children [52] , our study thus suggests that improving insulin sensitivity while avoiding harmful immune responses in genetically susceptible individuals may be an important new strategy for early T1D prevention . Our study also implicates that early metabolic disturbances in progression to autoimmune diabetes associate with diminished diversity of specific bacterial groups such as C . leptum group . This is in agreement with a recent pilot study in the DIPP cohort where phylum Firmicutes was found decreased in the four children who later progressed to diabetes [53] . Microbial communities are sensitive to disturbances and may subsequently not return to the their original state [54] . Interestingly , diminished diversity of the anti-inflammatory commensal bacterium Faecalibacterium prausnitzii from the C . leptum group characterizes also Crohn's disease [55] . Our study thus revealed a candidate microbial group which may be further considered in the context of diabetes prevention . The fact that the diabetes-associated differences in microbial composition were observed among the mice of the same colony suggests that the observed diminished microbial diversity is rather a consequence than a primary cause of immunological or metabolic responses . The mechanisms by which the gut microbial community is modulated by specific metabolic and immune factors associated with progression to T1D are at present unclear and demand further investigation . However , these findings may still be important by having a role in early disease pathogenesis . In fact , recent study revealed that microbes from C . leptum group induce regulatory T cells in colonic mucosa [56] . Diminishment of C . leptum diversity along with elevated leptin may therefore be two mechanisms which promote negative regulation of CD4+CD25+ regulatory T cells , and therefore also promote the autoimmune response [57] . Our study uncovered multiple factors which may contribute in parallel to progression towards autoimmune diabetes . It is unlikely that any of them is a primary cause to initiate the disease process . Instead , as an early mathematical model of T1D describing changes in numbers of β-cells , macrophages , and Th-lymphocytes concluded , the “onset of type 1 diabetes is due to a collective , dynamical instability , rather than being caused by a single etiological factor” [58] . In this context , understanding the spatial and temporal balance of different disease-contributing factors is important [59] . The study design such as ours may help identify the early factors contributing to the disease as well as their mutual dependencies . Finally , the metabolic phenotypes described here could be relevant as end points for studies investigating T1D pathogenesis and/or responses to interventions . By proceeding from a clinical study via metabolomics and modeling to an experimental model using a similar study design , then evolving further to tissue-specific studies , we hereby present a conceptually novel approach to reversed translation ( Figure 1 ) that may be useful in future therapeutic studies in the context of prevention and treatment of T1D as well as of other diseases .
All experimental procedures were approved by the Committee for Laboratory Animal Welfare , University of Turku . The mice were kept in an animal room maintained at 21±1°C with a fixed 12∶12 hr light-dark cycle . Standard rodent chow ( Special Diet Services , Witham , UK ) and water were available ad libitum . The colonies of NOD/Bom mice used were bred and maintained in the animal facilities of University of Turku and originated from mice purchased from Taconic Europe ( Ry , Denmark ) . 26 female and 44 male NOD mice ( Study 1 ) underwent weekly blood sampling by venopuncture from the tail vein starting at 3 weeks of age until the mice developed diabetes ( blood glucose ≥14 . 0 mmol/in two consecutive weeks ) or until female mice reached 36 weeks and male mice 40 weeks of age . Serum was separated and quickly frozen in −70°C for metabolomic analysis . Blood samples for detection of insulin autoantibodies ( IAA ) were collected from tail vein at the age of 8 weeks . Plasma samples for insulin were collected between noon and 2 PM after 4 hr fast and two days later 5 minutes after intraperitoneal glucose ( 1 g/kg ) administration at the age of 10 weeks . Another set of euglycemic NOD/Bom female mice ( Study 2 ) was sacrificed with decapitation under CO2 anesthesia at the age of 8 weeks ( n = 57 ) or 19 weeks ( n = 14 ) , and blood , liver and pancreas samples were collected . Two separate batches ( n = 36 and 42 , Studies 3 and 4 ) of female NOD/MrcTac were delivered from Taconic USA ( Hudson , NY , USA ) at 5 weeks of age . In Study 3 , intraperitoneal glucose tolerance test was performed after 4 hr fast at 8 weeks of age by administering glucose ( 10% [wt/vol] , 1 g/kg body weight ) and measuring tail vein blood glucose and serum insulin . Serum samples for lipidomics and IAA were collected from tail vein at 10 weeks of age . Intraperitoneal insulin tolerance test was performed after 1 hr fast at 11 weeks of age by administering human insulin ( 1 . 0 IU/kg body weight , Protaphane , Novo Nordisk , Bagsvaerd , Denmark ) . In Study 4 , mice were sacrificed at 10 weeks of age after 4 hr fast by cardiac puncture under anesthesia . Gonadal white adipose tissue ( WAT ) depot was carefully dissected and weighted , and was used as a marker of adiposity . Serum samples for IAA , lipidomics and adipokine panel assays , gonadal WAT , gastrocnemius muscle and pancreas samples were collected , and stored at −70°C until analyses . HOMA-IR , an estimate of insulin resistance , was calculated as fasting insulin ( µIU/ml ) ×fasting glucose ( mmol/l ) /22 . 5 . Statistical significances were analyzed with Student's t-test or two-way ANOVA using GraphPad Prism 4 . Blood glucose was measured with Precision Xtra™ Glucose Monitoring Device ( Abbott Diabetes Care , IL ) . Plasma insulin was analyzed with Mouse Ultrasensitive ELISA kit ( Mercodia , Uppsala , Sweden ) or together with leptin with Milliplex Mouse Adipokine Panel ( Millipore , Billerica , MA , USA ) . Plasma adiponectin was measured with Mouse Adiponectin ELISA kit from Millipore . Pancreatic islets were isolated using Ficoll 400 ( Sigma-Aldrich , St Louis , MI , USA ) gradient method [60] . In brief , the pancreata were incubated with Collagenase P ( 0 . 5 mg/ml , Roche Diagnostics , Mannheim , Germany ) in HBSS containing 10 mM HEPES , 1 mM MgCl2 , 5 mM Glucose , pH 7 . 4 for 17 min . After two rounds of washing , the pellet was resuspended in Ficoll 25% , and the densities 23% , 20% and 11% were layered on top . After centrifugation , the islet layer between densities 23% and 20% was collected and washed twice before snap freezing the pellet for metabolomic analysis or homogenization in lysis buffer for RNA extraction . Samples were stored in −70°C until analyses . Pancreata from euglycemic NOD mice were cryosectioned . 5 µm sections with >20 µm intervals were stained with hematoxylin & eosin and graded for insulitis as follows: 0 , no visible infiltration , I peri-insulitis , II insulitis with <50% and III insulitis with >50% islet infiltration . Total 678 islets from eight female 10-week-old low-risk mice ( 60–123 islets/each ) and 633 islets from eight high-risk mice ( 59–102 islets/each ) , and 52 islets from four female 19-week-old low-risk mice ( 11–17 islets/each ) and 28 islets from three high-risk mice ( 7–10 islets/each ) were graded . Statistical significance was analyzed with Student's t-test or Chi Square test using GraphPad Prism 4 . Murine IAA were measured by a radiobinding microassay ( RIA ) with minor modifications to that previously described for human IAA [61] . Mouse sera ( 2 . 5 µl ) and serial dilutions of standard samples ( 5 µl ) of a serum pool obtained from persons with a high IAA titer were incubated for 72 h with 15 , 000 cpm mono125I- ( TyrA14 ) -insulin ( Amersham , GE Healthcare , Buckinghamshire , UK ) in the presence or absence of an excess of unlabeled human recombinant insulin ( Roche Diagnostics , Mannheim , Germany ) . Antibody complexes were precipitated by adding 50 µl TBT buffer ( 50 mM Tris , pH 8 , 0 , 0 , 1% Tween 20 ) containing 8 µl Protein A and 4 µl Protein G Sepharose ( Amersham ) . After repeated washings the bound radioactivity was measured with a liquid scintillation detector ( 1450 Microbeta Trilux , Perkin Elmer Life Sciences Wallac , Turku , Finland ) . The specific binding was calculated by subtracting the non-specific binding ( excess unlabeled insulin ) from total binding and expressed in relative units ( RU ) based on standard curves run on each plate . The cut-off value for mouse IAA positivity was set at the mean+3SDS in 29 BALB-mice , i . e . 0 . 90 relative units ( RU ) . Serum samples ( 10 µl ) in Eppendorf tubes were spiked with a standard mixture containing 10 lipid compounds at a concentration level of 0 . 2 µg/sample , and mixed with 10 µl of 0 . 9% sodium chloride and 100 µl of chloroform∶methanol ( 2∶1 ) . After 2 min vortexing and 1 hr standing the samples were centrifuged at 10000 rpm for 3 min and 60 µl of the lower organic phase was taken to a vial insert and spiked with 20 µl of three labelled lipid standards at a concentration level of 0 . 2 µg/sample . The lipidomics runs were performed on a Waters Q-Tof Premier mass spectrometer combined with an Acquity Ultra Performance LC™ ( UPLC; Milford MA ) . The solvent system consisted of 1 ) water with 1% 1 M NH4Ac and 0 . 1% HCOOH and 2 ) LC/MS grade acetonitrile/isopropanol ( 5∶2 ) with 1% 1 M NH4Ac , 0 . 1% HCOOH . The gradient run from 65% A/35% B to 100% B took 6 min and the total run time including a 5 min re-equilibration step was 18 min . The column ( at 50°C ) was an Acquity UPLC™ BEH C18 ( 1×50 mm , 1 . 7 µm particles ) and the flow rate was 0 . 200 ml/min . The lipids were profiled using ESI+ mode and the data collected at a mass range of m/z 300–1200 . The data was processed by using MZmine software ( version 0 . 60 ) [62] , [63] and the lipid identification was based on an internal spectral library [64] . Data was normalized using the appropriate internal standards as previously described [14] , [65] . Depending on the protein concentrations of PBS buffered cell solutions , 20–40 µl samples were taken for islet metabolomic analysis . 10 µL of an internal standard labeled palmitic acid-16 , 16 , 16-d3 ( 250 mg/l ) and 400 µl of methanol solvent were added to the sample . After vortexing for 2 min and incubating for 30 min at room temperature , the supernatant was separated by centrifugation at 10 , 000 rpm for 5 min . The sample was dried under constant flow of nitrogen gas and derivatized with 25 µl of MOX ( 1 h , 45°C ) and MSTFA ( 1 hr , 45°C ) . 5 µl of retention index standard mixture with five alkanes ( 125 ppm ) was added to the metabolite mixture . Islet samples were analyzed by two-dimensional gas chromatography coupled to time of flight mass spectrometry ( GC×GC-TOFMS ) . The instrument used was a Leco Pegasus 4D ( Leco Inc . , St . Joseph , MI ) , equipped with an Agilent GC 6890N from Agilent Technologies ( Santa Clara , CA ) and a CombiPAL autosampler from CTC Analytics AG ( Zwingen , Switzerland ) . The modulator , secondary oven and time-of-flight mass spectrometer were from Leco Inc . The GC was operated in split mode with a 1∶20 ratio . Helium with a constant pressure of 39 . 6 psig was used as carrier gas . The first dimension GC column was a non-polar RTX-5 column , 10 m×0 . 18 mm×0 . 20 µm ( Restek Corp . , Bellefonte , PA ) , coupled to a polar BPX-50 column , 1 . 50 m×0 . 10 mm×0 . 10 µm ( SGE Analytical Science , Ringwood , Australia ) . The temperature program was as follows: initial temperature 50°C , 1 min→295°C , 7°C/min , 3 min . The secondary oven was set to 20°C above the oven temperature . Inlet and transfer line temperatures were set to 260°C . The second dimension separation time was set to 5 s . The mass range used was 45–700 amu and the data collection speed was 100 spectra/second . Raw data were processed using Leco ChromaTOF software , followed by alignment using Guineu software ( version 0 . 7 ) [66] . The metabolites were identified by using an in-house reference compound library together with The Palisade Complete Mass Spectral Library , 600K Edition ( Palisade Mass Spectrometry , Ithaca , NY ) . RNA extraction from islets was carried out with Rneasy minikit ( QIAGEN GmbH , Hilden , Germany ) and from liver , skeletal muscle ( m . gastrocnemius ) and gonadal white adipose tissue with Trizol reagent ( Invitrogen , Carlsbad , CA ) followed by RNase-free DNase I treatment ( QIAGEN GmbH ) and purification with Rneasy minikit . Pancreatic islets and liver for microarray analysis were collected from 19-week-old euglycemic female NOD/Bom mice . Skeletal muscle and adipose tissue for GLUT4 mRNA expression were collected from 10-week-old female NOD/MrkTac mice . GLUT4 mRNA expression in skeletal muscle and gonadal white adipose tissue was measured by quantitative real-time PCR . CDNA synthesis was performed with High Capacity RNA-to-cDNA Kit according to manufacturer's protocol . Real-time PCR was performed with 7300 Real Time PCR system , pre-designed TaqMan® Gene Expression Assay for GLUT4 and TaqMan® Endogenous Control Assay for β-actin . The 20 µl PCR reactions contained 8 µl cDNA , 8 µl TaqMan® Gene Expression Master Mix , 1 µl GLUT4 TaqMan Gene Expression Assay , 1 µl b-actin TaqMan Endogenous control Assay and 2 µl depc water . Cycling parameters for real-time RT-PCR were as follows: 50°C for 2 min , 95°C for 10 min followed by 40 cycles of 95°C for 15 seconds and 60°C for one minute . GLUT4 mRNA levels were expressed relative to β-actin , which was used as a housekeeping gene . Relative gene expression was calculated using the comparative CT method and RQ = 2−ΔΔCT formula . All reagents were from Applied Biosystems ( Foster City , CA , USA ) . RNA amplification was performed from 300 ng total RNA with Ambion's ( Austin , TX ) Illumina RNA TotalPrep Amplification kit ( cat no AMIL1791 ) . IVT reaction overnight ( 14 hr ) , during it cRNA was biotinylated . Both before and after the amplifications the RNA/cRNA concentrations where checked with Nanodrop ND-1000 ( Wilmington , DE ) and RNA/cRNA quality was controlled by BioRad's Experion electrophoresis station ( Hercules , CA ) . The samples were hybridized in the Finnish DNA Microarray Centre , at the Turku Centre for Biotechnology . 1 . 50 µg each sample was hybridized to Illumina's MouseWG-6 Expression BeadChips , version 2 ( BD-201-0602 ) at 58°C overnight ( 18 hr ) according to Illumina® Whole-Genome Gene Expression Direct Hybridization protocol , revision A . Hybridization was detected with 1 µg/ml Cyanine3-streptavidine , GE Healthcare Limited ( Chalfont , UK ) ( cat no PA43001 ) . Chips were scanned with Illumina BeadArray Reader , BeadScan software version 3 . 5 . The numerical results were extracted with Illumina's GenomeStudio software v . 1 . 0 without any normalization . Bead Summary data , exported from Illumina's GenomeStudio software , was preprocessed using beadarray package [67] of R/Bioconductor [68] as follows . Data was transformed to logarithm ( base 2 ) , and normalized using quantile method [69] , which equalizes the distribution of probe intensities across a set of microarrays . Gene Set Enrichment Analysis ( GSEA ) [20] , a commonly used pathway analysis technique for microarray gene expression data analysis , uses a Kolmogorov-Smirnov like statistic to test whether selected gene sets are enriched among the most up or down regulated genes . Multiple hypothesis testing was addressed by computing the false discovery rate q-values based on random permutation of membership of genes across gene sets as implemented in the GSEA software [20] . Linear Models for Microarray Data ( LIMMA ) approach [70] identifies differentially expressed genes by fitting a linear model to the expression data of each gene , and computing moderated t-statistic using posterior residual standard deviations to account for the gene-specific variability of expression values . Here , we used the R/Bioconductor package [68] and LIMMA [70] for testing differential expression of genes . We then performed pre-ranked GSEA analysis using the moderated t-statistic for ranking the gene list , to test for enrichment of gene sets from a variety of pathway databases such as Gene Ontology ( GO ) [71] , GenMAPP [72] , BioCarta ( http://www . biocarta . com ) , Signal Transduction Knowledge Environment ( STKE ) ( http://stke . sciencemag . org/ ) , and KEGG [73] curated in Molecular Signatures Database ( MSigDB ) [20] . Leading edge genes of an enriched pathway are the genes that account for the enrichment signal [20] . For selected pathways that are found statistically significant by GSEA , the pathway profiles are calculated as average expression of all leading edge genes . This matrix of pathway profiles of selected pathways was then augmented with selected metabolite profiles . Then the numerical values in this matrix were normalized with the autoantibody-negative low-risk group ( IAA− & LR ) , i . e . , each numerical value of a variable is divided by the average values from IAA− & LR samples , and transformed to logarithmic ( base 2 ) scale . Then the variables were scaled for unit variance . Finally , hierarchical clustering was applied using Euclidean metric and complete linkage method [74] for computing inter-cluster distances . An R package called gplots ( http://www . r-project . org/ ) was used for the clustering and displaying the numerical values as a heat map . DNA was extracted from 200 mg of fecal sample from caecum using FastDNA Spin Kit for Soil ( QBIOgene , Carlsbad , CA , ) with modifications to the manufacturer's instructions [29] . PCR-DGGEs of predominant bacterial PCR-DGGE and five different group specific PCR-DGGEs ( bifidobacteria , Lactobacillus-group , Eubacterium rectale – Blautia coccoides clostridial group ( Erec-group ) , Clostridium leptum clostridial group ( Clept group ) , and genus Bacteroides ) were performed as described previously [30] . The comparison of the profiles and the quantification of the amplicons were performed using BioNumerics software version 5 . 1 ( Applied Maths NV , Sint-Martens-Latem , Belgium ) . The statistical analysis of amplicon numbers was performed with the Student's t-test with unequal variances . Clustering was performed with Pearson correlation from each bacterial group separately besides using composite datasets ( included predominant bacterial DGGE and five group specific DGGEs ) in which amplicons with the total surface area of at least 1% were included in the similarity analysis . Principal component analysis was performed with the BioNumerics software . R statistical software ( http://www . r-project . org/ ) was used for data analyses and visualization . The concentrations were compared using the Wilcoxon rank-sum test , with p-values <0 . 05 considered statistically significant . Due to the large number of tests , one test for each of the 154 lipids , for the differences in mean concentrations between the progressor and non-progressor groups some p-values may be small due to chance . In order to quantify the number of such false significant findings we estimated the false discovery rates using q-values [15] , [16] . A q-value is associated for each lipid with the interpretation that among those lipids that have p-value less than or equal to the p-value of the lipid a fraction q are falsely stated significant . To account for multiple comparisons , false discovery rates among significantly differing lipids were estimated using q-values [15] , [16] . False discovery rates were computed using the R package q-value . The fold difference was calculated by dividing the median concentration in high-risk group by the median concentration in low-diabetes-risk group and taking the base-2-log of the resulting value . This makes interpretation easy as values greater/smaller than zero correspond to up/down-regulated lipids in the high-risk group . In clustering we applied a customized correlation based distance metricWhere and denote the concentrations of lipids an in the sample set . Ward's method was then applied in hierarchical clustering using this distance measure [75] . Metabolic state development in diabetes progressors and non-progressors was modeled by separate Hidden Markov Models [18] , making it possible to align individuals based on metabolic states instead of age , and to compare the metabolic states in progressors and non-progressors . The modeling assumptions under which the models are fitted to data are that individuals share a similar developmental progression but the timing of the states may vary , and that metabolite profiles in each state may be different for progressors and non-progressors . Model fitting was done by the standard Baum-Welch algorithm using the MATLAB toolbox by Kevin Murphy ( http://www . cs . ubc . ca/~murphyk/Software/HMM/hmm . html ) . The model structure was validated by the bootstrap in the same way as in our earlier studies [18] , and confidence intervals were estimated with non-parametric bootstrap ( 5000 samples ) . Let and be two data matrices with and samples , , and dimensions and , respectively . The task is to find a permutation of samples in such that each sample in is matched with in . We assume a one-to-one matching of samples between the two data matrices . Since the data matrices do not lie in the same data space , it is not possible to use distance as the matching criterion . We have recently introduced a methodology based on statistical dependencies between the data sets to solve this problem [19] . The idea is to compute from the data features or statistical descriptors that maximize statistical dependencies , and do the matching based on the descriptors . In practice , we project the data onto a lower-dimensional subspace such that the statistical dependencies between the datasets are maximized , and find a matching of samples in this comparable subspace . In order to find disease effects shared by NOD mice and humans in the DIPP study , we first paired metabolites of the two organisms , then estimated the metabolic states of progressor and non-progressor men and mice by HMMs , and finally did a bootstrap-based two-way analysis on progressors/non-progressors vs . men/mice to identify disease and organism effects and their interactions . The data-driven pairing or the metabolites and the four HMMs were computed as described above . The two-way analysis of disease effect was done by first removing the organism effect , represented with a single mean parameter estimated by least squares , and then computing bootstrap confidence intervals for the disease effect of pooled men and mice . Organism and cross effects were estimated analogously .
|
We have recently found that distinct metabolic disturbances precede β-cell autoimmunity in children who later progress to type 1 diabetes ( T1D ) . Here we performed a murine study using non-obese diabetic ( NOD ) mice that recapitulated the protocol used in human , followed up by independent studies where NOD mice were studied in relation to risk of diabetes progression . We found that young female NOD mice who later progress to autoimmune diabetes exhibit the same lipidomic pattern as prediabetic children . These metabolic changes are accompanied by enhanced glucose-stimulated insulin secretion , upregulation of insulinotropic amino acids in islets , elevated plasma leptin and adiponectin , and diminished gut microbial diversity of the Clostridium leptum subgroup . The metabolic phenotypes observed in our study could be relevant as end points for studies investigating T1D pathogenesis and/or responses to interventions . By proceeding from a clinical study via metabolomics and modeling to an experimental model using a similar study design , then evolving further to tissue-specific studies , we hereby also present a conceptually novel approach to reversed translation that may be useful in future therapeutic studies in the context of prevention and treatment of T1D as well as of other diseases characterized by long prodromal periods .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"systems",
"biology",
"medicine",
"clinical",
"immunology",
"autoimmune",
"diseases",
"diabetes",
"mellitus",
"type",
"1",
"biology",
"computational",
"biology",
"metabolic",
"networks",
"immunology"
] |
2011
|
Metabolic Regulation in Progression to Autoimmune Diabetes
|
Infection with the helminth Schistosoma ( S . ) mansoni drives the development of interleukin ( IL ) -10-producing regulatory B ( Breg ) cells in mice and man , which have the capacity to reduce experimental allergic airway inflammation and are thus of high therapeutic interest . However , both the involved antigen and cellular mechanisms that drive Breg cell development remain to be elucidated . Therefore , we investigated whether S . mansoni soluble egg antigens ( SEA ) directly interact with B cells to enhance their regulatory potential , or act indirectly on B cells via SEA-modulated macrophage subsets . Intraperitoneal injections of S . mansoni eggs or SEA significantly upregulated IL-10 and CD86 expression by marginal zone B cells . Both B cells as well as macrophages of the splenic marginal zone efficiently bound SEA in vivo , but macrophages were dispensable for Breg cell induction as shown by macrophage depletion with clodronate liposomes . SEA was internalized into acidic cell compartments of B cells and induced a 3-fold increase of IL-10 , which was dependent on endosomal acidification and was further enhanced by CD40 ligation . IPSE/alpha-1 , one of the major antigens in SEA , was also capable of inducing IL-10 in naïve B cells , which was reproduced by tobacco plant-derived recombinant IPSE . Other major schistosomal antigens , omega-1 and kappa-5 , had no effect . SEA depleted of IPSE/alpha-1 was still able to induce Breg cells indicating that SEA contains more Breg cell-inducing components . Importantly , SEA- and IPSE-induced Breg cells triggered regulatory T cell development in vitro . SEA and recombinant IPSE/alpha-1 also induced IL-10 production in human CD1d+ B cells . In conclusion , the mechanism of S . mansoni-induced Breg cell development involves a direct targeting of B cells by SEA components such as the secretory glycoprotein IPSE/alpha-1 .
Helminths can persist for up to decades in the human host . This is hypothesized to be , at least in part , because of their evolutionarily adapted relationship with the host [1] . Helminths are well-known for their strong capacity to promote the regulatory arm of the hosts immune system , thereby prolonging their survival within the host [2] . As a bystander effect , helminths can also suppress immune responses to other antigens , such as allergens and auto-antigens , and other pathogens . This bystander effect seems to be so pronounced that it may prevent the development of inflammatory diseases . Indeed , both epidemiological studies and mouse models show a clear protective role of helminths against various forms of auto-immunity , allergic airway inflammation , colitis etc . [3 , 4 , 5 , 6 , 7] . The formation of a network of regulatory immune cells plays a crucial role for the protective effect . Helminth infection , and in particular infection with schistosomes such as Schistosoma ( S . ) mansoni are well-known to induce regulatory B ( Breg ) cells [8–15] , a relatively new member in the network of regulatory immune cells . Breg cells have gained considerable attention due to their ability to down-modulate inflammation in a variety of conditions ranging from autoimmune disorders such as experimental autoimmune encephalomyelitis ( EAE ) , collagen-induced arthritis ( CIA ) , lupus and chronic colitis to anaphylactic and allergic airway inflammation [8 , 10–12 , 16–23] . Regulatory B cells suppress pro-inflammatory immune responses via several mechanisms , of which the ones best described are the expression of the regulatory cytokine interleukin-10 ( IL-10 ) and induction of regulatory T ( Treg ) cells [24] . We previously reported the induction of Breg cells by schistosome infection in both mouse and human , and found the most potent IL-10-producing Breg cells within the human CD1d+ B cell subset . This corresponds to the CD1d+CD23lowCD21+ marginal zone ( MZ ) B cell subset in mice , which efficiently reduced experimental allergic airway inflammation in our model [12] . The cellular mechanisms to achieve Breg cell induction as well as the nature of the B cell-activating S . mansoni antigens however remain largely unknown . The identification of relevant stimulatory molecules and optimal Breg cell-inducing conditions is a critical step in enhancing the activity of Breg cells for use as a new therapeutic tool against inflammatory diseases . Both an indirect induction of a regulatory phenotype in B cells by activation of accessory cell types , as well as a direct binding and interaction between S . mansoni antigens and B cells via pattern recognition receptors ( PRRs ) such as Toll-like receptors ( TLRs ) expressed on B cells [25] are plausible options . In the splenic MZ , located at the border of white and red pulp , MZ B cells are nested between SIGN-R1+ MZ macrophages and Siglec-1+ metallophilic macrophages [26] . MZ macrophages not only fulfill a main function in sensing blood-borne pathogens , but also perform functional interactions with MZ B cells . These interactions have important implications for the maintenance of the MZ itself and the function of MZ B cells and macrophages [27–29] . Hence , it is therefore tempting to speculate that MZ macrophages are a prime candidate as Breg cell induction partner . On the other hand , the direct ligation of various TLRs on B cells , including TLR2 , TLR4 , TLR7 and TLR9 , has been described to induce IL-10 production [30 , 31] . In addition , BCR and CD40 engagement were described to be involved in IL-10-dependent regulatory B cell function in models of EAE , CIA , and contact hypersensitivity [17–19 , 32] . In the current study , we tested the hypothesis that eggs and/or egg-derived excretory-secretory molecules from S . mansoni , without the context of natural infection , are sufficient to drive Breg cell development by activating splenic MZ B cells . In addition , we investigated whether Breg cells are induced indirectly by activation of accessory cell types in the MZ , or by direct binding and interaction via PRRs on B cells . We found that egg antigens drive Breg cell development in vivo and in vitro by direct interaction with splenic B cells , which after binding and internalization of egg antigens secrete elevated levels of IL-10 and are capable of driving Treg cell development . The egg antigen-induced Breg cell development was independent of macrophages of the marginal zone but was enhanced by CD40 ligation . Most importantly , we identified the egg glycoprotein IPSE/alpha-1 as a single molecule from S . mansoni that is capable to induce Breg cells both in mice and man . This knowledge will assist to further define helminth-specific conditions for the generation of Breg cells to be used in therapeutic approaches .
To elucidate the mechanism by which S . mansoni can drive the development of Breg cells , we first investigated whether schistosome eggs or their soluble antigens were sufficient to drive Breg cell development in vivo , without the context of natural infection . Intraperitoneal treatment of C57BL/6 mice with two doses of 5000 S . mansoni eggs or 100 μg of soluble egg antigens ( SEA ) one week apart efficiently induced IL-10 protein expression in splenic CD19+ B cells one week after the last injection . IL-10 protein secreted during 2 days ex vivo restimulation with SEA was 3 to 4-fold increased compared to the amount secreted by restimulated B cells from control-treated animals ( Fig 1A ) , while IL-6 was unchanged . This indicates a typical cytokine expression pattern characteristic for Breg cells . The frequencies of B cells expressing intracellular IL-10 protein were likewise significantly 2-fold increased ( Fig 1B and 1C ) . Also the surface activation marker CD86 , often upregulated on activated B cells and Breg cells [33–35] , was increased on splenic B cells by egg or SEA treatment , while CD40 expression was not significantly changed ( Fig 1D ) . To verify that the observed effects are specific and exclude a general influence of protein solutions on B cell IL-10 production and activation , we treated mice with human serum albumin ( HSA ) as infection-unrelated control protein . We did not observe an increased IL-10 secretion ( S1A Fig ) or CD86 expression ( S1B Fig ) by B cells compared to the PBS group , and concluded that PBS is a suitable control for subsequent experiments . Furthermore , injection of eggs or egg antigens was as efficient in increasing the frequency of IL-10-expressing B cells as was chronic infection with S . mansoni ( S1C Fig ) , and egg-injected mice continued to have an elevated B cell IL-10 production until at least 4 weeks after the last egg injection ( S1D Fig ) , indicating that this phenotype is persisting over longer periods . SEA was purified from liver eggs and can contain LPS to variable extent . Since the TLR4 ligand LPS is known for its capacity to drive B cell IL-10 expression and Breg cell development [33 , 36 , 37] , it is crucial to exclude that the Breg driving capacity by SEA in vivo was due to LPS contamination of the schistosome antigen preparations . Therefore , the same experiment was repeated in TLR4-deficient animals and compared to wild-type . Upon SEA treatment , IL-10 secretion of B cells as well as intracellular IL-10 and surface CD86 expression was overall comparable in both groups ( S2 Fig ) , indicating that the Breg-inducing capacity by SEA was largely not attributable to a putative LPS contamination . To confirm the regulatory function , splenic B cells from the various groups were tested for their capacity to drive Treg cell development , an acquired phenotype previously described for splenic B cells during natural schistosome infections [12] . Indeed , splenic B cells from egg- or SEA-injected , but not control-treated , mice induced the development of CD25+Foxp3+ Treg cells during 4 day co-culture with CD25-depleted CD4 T cells ( Fig 1E and 1F ) , which confirms the regulatory capacity of egg antigen-activated B cells ex vivo . As expected , IL-10 protein concentration in co-culture supernatants was only increased in presence of egg antigen-activated but not control B cells ( Fig 1G ) . Collectively , these data show that schistosome eggs or their soluble antigens on their own are sufficient to induce IL-10-producing Breg cells in vivo , without the context of natural infection , and that these B cells are bona fide Breg cells that can drive Treg cell development . Different B cell subsets have been described to give rise to Breg cells , especially in spleen where subsets differ e . g . in tissue localization and pathogen recognition receptor expression [25 , 38] . We and others had previously identified CD23lowCD21+ marginal zone B cells as the major IL-10-producing splenic B cell subset during chronic Schistosoma infection and mediating protection in a mouse model of airway inflammation [8 , 12] . To test whether soluble egg antigens act on the same splenic subset , we sorted splenic CD23lowCD21+ marginal zone B cells from egg- , SEA-treated , or control mice for subsequent ex vivo restimulation and cytokine analysis , and compared this with the major splenic B cell subset , CD23hiCD21- follicular B cells ( Fig 2A ) . Only marginal zone B cells but not follicular B cells showed significantly increased IL-10 secretion as was measured in culture supernatants after 2 day restimulation with SEA ( Fig 2B ) . As for total B cells ( Fig 1A ) , also for the individual subsets IL-6 expression was not increased ( Fig 2C ) . Intracellular IL-10 expression and CD86 expression was likewise significantly upregulated in marginal zone B cells of egg- or SEA-treated mice compared to control-treated mice . SEA seemed to be more potent than egg injection in activating follicular B cells , as SEA-injection also significantly increased intracellular IL-10 and CD86 expression in this subset , although expression levels remained significantly lower compared to marginal zone B cells ( Fig 2D and 2E ) . For analysis of B cell activation we generally restimulated cells ex vivo with SEA . This increased the baseline expression of CD86 in all groups compared to medium ( average MFI of follicular B cells: 1013–1632±62 . 2; for marginal zone B cells: 2640–4118±176 . 9 , for all groups and without significant differences between groups ) , but was required for detection of B cell cytokines as a result of the in vivo antigen exposure . Without SEA restimulation , we found a trend of increased IL-10 production by B cells which only reached significance upon additional restimulation ( S1E Fig ) , indicating that renewed exposure to antigen is required to achieve detectable B cell activity and cytokine production . This is also supported by experiments in IL-10 GFP reporter mice , in which IL-10 ( GFP ) accumulates in B cells during the entire in vivo treatment period . Here , increased IL-10 ( GFP ) expression , without ex vivo restimulation , was only detectable in B cells of 14 weeks chronically infected mice , but not in B cells of mice that received two injections of eggs within a relatively short period of 2 weeks ( S1F and S1G Fig ) . Altogether , these data indicate that the result of in vivo development of IL-10 producing B cells is in principle detectable without restimulation ( S1E and S1G Fig ) , but the data from egg-injected IL-10 reporter mice also suggest that an ex vivo SEA restimulation is required to visualize its full IL-10 potential , something which will happen in vivo during a natural infection due to the constant production of eggs and the high levels of circulating antigens . Taken together , these data support the notion that B cells , and in particular marginal zone B cells , are responsive to in vivo schistosome antigen stimulation thus supporting the findings in natural schistosome infections . Molecules secreted by schistosome eggs are highly glycosylated and known to bind to C-type lectin receptors [39 , 40] . Since B cells show a very restricted expression of those receptors [41] , we hypothesized that other C-type lectin receptor-expressing cell types in the splenic marginal zone , such as macrophages or dendritic cells , bind SEA and provide additional signals to the marginal zone B cells to support Breg cell development . Among the accessory cell types , macrophages of the splenic marginal zone were of particular interest because of their known interactions with marginal zone B cells [28 , 42] and schistosome antigens [43 , 44] . However , it was unknown whether marginal zone macrophages can capture SEA in vivo and are important for B cell IL-10 expression . To evaluate this , fluorescently labeled SEA was injected i . v . and 30 minutes to 24 hours later various splenic cell types were analyzed for bound SEA using fluorescence microscopy and flow cytometry . Already after 30 minutes of injection , SEA clustered along the marginal zone area of the spleen as detected by fluorescence microscopy of splenic tissue sections ( Fig 3A ) . SEA localized predominantly within two specialized macrophage subsets of the marginal zone: SIGN-R1-expressing MZ macrophages and Siglec-1-expressing marginal metallophilic macrophages ( Fig 3B ) . In contrast , after injection of labeled ovalbumin ( OVA ) as non-schistosomal control protein no fluorescence signal was detected in the spleen ( S3 Fig ) . Flow cytometry confirmed that 83% of metallophilic macrophages were positive for SEA only 30 minutes after injection , and still 53% of cells after 24 hours ( Fig 3C ) . In addition , F4/80-expressing red pulp macrophages and Ly6Chi monocytes efficiently bound SEA ( 81 . 7 and 82 . 6% , respectively ) , while only a small fraction of dendritic cells and neutrophils did bind SEA ( Fig 3C , S4A–S4C Fig for gating scheme of cell types ) . Metallophilic macrophages not only bound SEA abundantly , they also significantly upregulated typical surface activation markers such as CD11c and CD86 , but not MHCII , at 30 minutes and 2 hours after SEA injection , respectively ( Fig 3D ) . Because macrophages of the marginal zone were most potent in binding SEA , we next addressed their role for SEA-mediated marginal zone Breg cell induction . To this end , macrophages were depleted in vivo by i . p . injection of clodronate-containing liposomes [45] prior to injection of schistosome eggs . Eggs were injected 3 and 4 weeks after clodronate treatment , at time-points when only macrophages , including metallophilic and MZ subsets , were significantly reduced in spleens , but no other cell types ( S4D Fig and [46] ) . Successful and specific depletion of splenic macrophages was also confirmed by fluorescence microscopy of tissue sections ( Fig 3E ) and flow cytometry ( S4E Fig ) at 7 days after the last egg injection , when B cell activity was analyzed . Unexpectedly , IL-10 secretion of splenic B cells from macrophage-depleted mice was equal to that from control liposome-treated mice ( Fig 3F ) . Also the upregulation of intracellular IL-10 ( Fig 3G ) and CD86 expression in B cells , as well as the induction of Foxp3+CD25+ and IL-10+CD25+ CD4 T cells ( S4F and S4G Fig ) following SEA injection was not affected by the absence of macrophages . In conclusion , splenic macrophages are not essential for schistosome antigen-induced Breg cell development , despite the high binding of SEA by different macrophage subsets . To test whether B cells directly bind and interact with schistosome antigens without the help of surrounding accessory cell types , fluorescently labeled SEA was injected i . v . and its co-localization with B cells analyzed by fluorescence microscopy of splenic tissue sections . Indeed , egg antigens were found to co-localize with some splenic B220+ B cells ( Fig 4A ) . Flow cytometry , a more sensitive method compared to fluorescence microscopy , showed that only MZ B cells but not follicular B cells did bind SEA in vivo , with a maximum of 56 . 4% of cells being positive at 2 hours after SEA injection . SEA was still detectable on 11 . 5% of marginal zone B cells at 24 hours ( Fig 4B ) . Marginal zone B cells also showed an increased CD86 surface expression following SEA injection , which was significant at 6 hours after SEA injection ( Fig 4C ) . Interestingly , marginal zone B cells that bound SEA showed a higher CD86 expression compared to cells that were found to be negative for SEA , i . e . approximately 3-fold ( SEA positive ) versus only 1 . 5-fold ( SEA negative ) compared to B cells from untreated animals ( Fig 4C ) . This suggests that marginal zone B cells not only efficiently bind egg antigens in vivo , but also become ( more ) activated as a consequence of this interaction . Binding of SEA to B cells was confirmed in vitro by culturing splenic B cells with fluorescently labeled SEA for 60 minutes , after which up to 16% were positive for SEA as measured by flow cytometry ( Fig 4D ) , which is a similar percentage as found for total B cells after in vivo application of labeled SEA ( Fig 4B ) . By using SEA labeled with the pH-sensitive dye pHrodo , it was shown that egg antigens were not only bound to the surface but were also internalized by B cells into acidic cellular compartments ( Fig 4D ) . As in vivo , also in vitro the marginal zone B cell subset bound SEA more efficiently than the follicular B cell subset , with in average 15 . 9% versus 6 . 9% of cells being positive for SEA ( Fig 4E ) . Collectively , these data show that B cells , and in particular MZ B cells , are capable of directly interacting with schistosome antigens by binding and internalization of those antigens , both in vivo and in vitro . Next , we investigated whether the observed direct interaction of B cells with SEA can drive IL-10 expression and induction of regulatory B cell function in vitro . To this end , CD19+ splenic B cells from naïve mice isolated using MicroBeads were cultured for 3 days with SEA . SEA-stimulated B cells secreted significantly more IL-10 , but not IL-6 , compared to non-stimulated B cells ( Fig 5A ) , showing a typical cytokine pattern characteristic for schistosome-induced Breg cells . Separate cultures of sorted MZ and follicular B cells showed once more that the MZ B cell subset reacted more potently to SEA , e . g . with a significantly higher IL-10 secretion compared to the follicular subset ( Fig 5B ) . In addition , frequencies of B cells expressing intracellular IL-10 were likewise increased ( Fig 5C ) . To ensure that the IL-10 phenotype is not reliant on artificially high levels of stimulation with PMA/ionomycin which is added to facilitate detection of intracellular IL-10 , we repeated the assay using cells from IL-10-GFP reporter ( TIGER ) mice with similar results ( S5 Fig ) . CD40 and CD86 expression were upregulated compared to cultures in medium alone ( Fig 5D ) . Importantly , in vitro SEA-activated B cells were also capable of driving Treg cell development during a 4 day co-culture with CD25-depleted CD4 T cells ( Fig 5E ) , thus providing further evidence for a bona fide regulatory function of the in vitro induced Breg cells . Because we had seen internalization of egg antigens into acidic compartments ( Fig 4D ) , we wondered whether lysosomal processing is necessary for induction of IL-10 expression . Addition of chloroquine , an inhibitor of endosomal acidification [47] , significantly reduced the IL-10 secretion and frequency of IL-10+ B cells induced by SEA and CpG ( ligand for endosomal TLR9 ) , but not by Pam3Cys ( ligand for surface TLR2 ) ( Fig 5F and 5G ) . This suggests that internalization and endosomal processing of SEA is required for B cell IL-10 induction . The type of receptor involved in direct activation of Breg cells by SEA remains unknown . Egg antigens are abundantly glycosylated and known to bind to C-type lectin receptors [39 , 40] . Lex-motifs , one of the most abundant glycan structures present in SEA , bind to the C-type lectin receptor SIGN-R1 . However , when treating SIGN-R1-deficient mice with SEA , IL-10 expression was equally well induced compared to wild-type mice ( S6A Fig ) , suggesting no involvement of the Lex-motifs . Furthermore , stimulation of various TLRs on B cells , including TLR2 , TLR4 , TLR7 and TLR9 , has been described to induce IL-10 production [30 , 31] . Because SEA has been reported to contain TLR2 activity [48 , 49] , we compared SEA-induced Breg cell responses in wild-type and TLR2-deficient B cells . We did however not observe any difference in IL-10 secretion between the two strains ( S6B Fig ) , excluding a role of TLR2-triggering SEA components in SEA-induced B cell IL-10 production . This is further supported by the fact that TLR2-mediated B cell activation was independent of endosomal processing while SEA-mediated activation was dependent on it ( Fig 5F and 5G ) . Collectively , these data demonstrate that Breg cells can be generated in vitro by culture with schistosome antigens , that endosomal processing is involved in this process , and that these Breg cells are functional in the sense that they support Treg cell development . Previous studies highlighted a role for CD40 ligation during in vitro Breg cell induction [50 , 51] . We therefore tested whether CD40 ligation could increase the SEA-mediated effect on Breg cell development . Addition of anti-CD40 stimulatory antibody ( Ab ) to the 3 day SEA culture increased IL-10 secretion of splenic B cells by 1 . 7-fold compared to SEA alone . A similar enhancing effect was observed for IL-6 secretion ( Fig 6A ) and CD86 expression ( Fig 6B ) . In contrast to anti-CD40 Ab , addition of anti-IgM Ab did not significantly enhance the SEA-mediated B cell activation ( S7 Fig ) . To exclude stimulatory effects from LPS in vitro , B cells from TLR4-deficient mice were stimulated with SEA with or without addition of anti-CD40 Ab . The fold increase of IL-10 secretion compared to B cells cultured in medium was similar for wild-type and TLR4-deficient cells , thus excluding a major effect of the TLR4 ligand LPS ( S6C Fig ) . Finally , co-culture of CD25-depleted CD4 T cells with anti-CD40-stimulated B cells increased the frequency of CD25+Foxp3+ Treg cells , which was further increased if B cells had been stimulated with SEA plus anti-CD40 Ab ( Fig 6C ) . Thus , SEA stimulation plus CD40 ligation of B cells further enhanced the capacity to drive the development of IL-10-producing Breg cells in vitro . Previous reports only addressed the role of CD40 ligation in vitro , but the relevance for in vivo Breg cell induction was not investigated . We therefore blocked CD40 ligand in vivo by i . p . injection of a hamster anti-mouse CD40 ligand blocking mAb ( 200 μg; every 4 days starting at day -1 prior to 1st SEA injection ) during SEA treatment of mice and analyzed the effect on Breg cell activation . In hamster IgG-injected control mice , SEA treatment increased the amount of B cell-derived IL-10 secretion by 3 . 3-fold compared to PBS treatment . Upon anti-CD40 ligand administration this increase was only 1 . 6-fold and thereby significantly lower ( Fig 6D ) . Importantly , the SEA-mediated upregulation of intracellular IL-10 and CD86 expression by B cells was even fully abolished when CD40 ligand was blocked ( Fig 6E and 6F ) . Taken together , CD40 ligation enhances the Breg cell-inducing effect of SEA both in vitro and in vivo . SEA is a complex mixture of several different antigens . In the next step , we therefore aimed to identify specific antigens in SEA that are relevant for Breg cell induction . We focused on three major antigens that provoke an antibody response in nearly all infected patients: omega-1 , kappa-5 and IPSE/alpha-1 [52 , 53] . B cells were able to bind fluorescently labelled natural IPSE/alpha-1 ( nIPSE ) , a secreted egg antigen we purified from egg extracts , in a dose-dependent manner during 60 minutes in vitro culture ( Fig 7A ) . During 3 days culture however , nIPSE induced significantly elevated IL-10 but not IL-6 secretion by B cells in a concentration dependent manner ( Fig 7B and S8A Fig ) . Importantly , recombinant IPSE/alpha-1 expressed in tobacco plants ( pIPSE ) , which behaves as nIPSE in terms of protein dimerization and human basophil activation ( S9 Fig ) , had similar effects to the natural molecule on B cell IL-10 and IL-6 secretion ( Fig 7C ) . Both nIPSE and pIPSE-stimulated B cells were capable of driving Treg cell development during B cell-T cell co-culture ( Fig 7D ) . Interestingly , SEA depleted of IPSE/alpha-1 ( SEAΔIPSE ) was as efficient as total SEA in inducing IL-10 secretion and CD86 expression by B cells , which suggests that also SEA antigens other than IPSE/alpha-1 can activate B cells ( Fig 7B ) . However , as opposed to IPSE/alpha-1 , other major components in SEA , such as omega-1 and kappa-5 , did not increase IL-10 in any of the tested concentrations ( 1–20 μg/ml ) ( S8B Fig ) , thus excluding a role for these antigens in SEA-mediated Breg cell induction . Notably , omega-1 is toxic to B cells at concentrations of 5 μg/ml and above , and was therefore only tested at 1 μg/ml . For better comparability , we determined the following average relative amounts of IPSE/alpha-1 , omega-1 and kappa-5 within SEA , based on the yields of several purifications of these molecules from SEA: 1 . 2% IPSE/alpha-1 , 0 . 6% omega-1 and 1 . 8% kappa-5 . Furthermore , to proof that the Breg cell-inducing effect is specific for molecules in SEA , we stimulated B cells in vitro with adult worm antigen ( AWA ) as control of an S . mansoni-derived antigen mixture not containing IPSE . AWA was unable to induce IL-10 secretion when tested in the same concentration as used for SEA ( S8C Fig ) . For a possible future therapeutic application of antigen-activated Breg cells against e . g . allergic diseases , it is crucial to confirm the IL-10-inducing effect in human B cells . After 3 days in vitro stimulation with SEA and anti-CD40 , we found a significant increase in the percentage of total human IL-10+ CD19+ B cells compared to cells cultured with anti-CD40 alone . Comparing different B cell subsets , which have previously been attributed with regulatory properties [54 , 55] , we found the increase in IL-10+ B cells after SEA stimulation to be most pronounced among CD1d+ B cells rather than CD24+CD27+ and CD24+CD38+ B cells . Both nIPSE and pIPSE significantly increased the fraction of IL-10-expressing cells among CD1d+ B cells , whereas neither had an effect on the other two subsets investigated . As CD1d+ B cells only comprise a very small fraction of all B cells ( CD24+CD27+ B cells >> CD24+CD38+ B cells > CD1d+ B cells ) , the effect of nIPSE and pIPSE does not translate into an increase in the percentage of IL-10+ cell in the total B cell pool in contrast to SEA ( Fig 8 ) . This also suggests that additional molecules in SEA may have an IL-10-inducing effect . Collectively , we demonstrated that SEA was bound to and internalized by B cells , and that this direct interaction drives the development of Breg cells . Furthermore , we identified IPSE/alpha-1 as a single molecule of SEA that induces Breg cells in mice and humans , both as a natural and a recombinant molecule .
In this study , we sought to identify the molecules and mechanisms involved in the induction of IL-10-producing B cells by the helminth S . mansoni . We found that soluble antigens derived from schistosome eggs , amongst which the secretory antigen IPSE/alpha-1 , directly interacted with B cells . This led to the development of Breg cells characterized by IL-10 secretion and Treg cell-inducing capacity . Next to the potentially therapeutic relevance of how to generate regulatory , anti-inflammatory cells , this study also provides mechanistic insight into how schistosomes interact with the host immune system , expanding the regulatory arm of immunity and thereby prolonging its survival in the host . While we and others have shown IL-10 expression in splenic Breg cells during natural infection with S . mansoni [8 , 11 , 12] , the contribution of S . mansoni-derived egg antigens was not yet studied . SEA is highly immunostimulatory and well-known to promote Th2 as well as Treg cell responses in the host [56] . Here , we found that S . mansoni egg antigens are also able to induce IL-10-producing B cells in vivo , without the context of natural infection . Because SEA is a complex mixture of several different antigens , it is not unexpected that different types of immune cells and different qualities of immune responses are induced . The use of Breg cells as a therapeutic tool against inflammatory diseases is especially attractive because Breg cells in turn can induce Treg cell development [57] , which would thus amplify the beneficial regulatory effect . It was therefore important to investigate whether SEA-induced IL-10-producing B cells have the capacity to trigger Treg cell development . This was indeed the case as we could show by in vitro co-cultures of T cells with egg antigen- or IPSE-activated B cells . Similarly , Breg cells isolated from naturally schistosome-infected mice and humans were previously shown to drive Treg cell development in vitro [8 , 12 , 58] , thus indicating a common feature of schistosome-induced Breg cells . Our finding that egg antigens could induce splenic Breg cell development in vivo raised the question whether those antigens can directly interact with B cells in the spleen . In in vivo binding studies using fluorescently labeled SEA and fluorescence microscopy , egg antigens were indeed found to directly bind to splenic B cells . Although various egg antigens are abundantly glycosylated , the restricted and low expression of C-type lectin receptors by B cells [41] argues rather for the involvement of non-C-type lectin PRRs expressed by B cells in the direct binding of SEA components . Indeed , preliminary experiments with B cells of SIGN-R1-deficient mice , showed an equal IL-10 expression in response to SEA compared to wild-type littermates , suggesting no involvement of the Lex-motifs , one of the most abundant glycan structures present in SEA and known to bind SIGN-R1 . Instead , we found that within a mixture of schistosome antigens , at least the egg glycoprotein IPSE/alpha-1 was capable of driving Breg cell development in vitro by directly interacting with B cells , equipping them with Treg cell-inducing capacity . Two independent experimental approaches suggested that egg antigens are taken up and processed in acidic lysosomes . Previous reports on in vitro induction of Breg cells by helminth antigens did not use highly sort-purified B cells as in our study , but total splenocyte preparations [8 , 10] or merely B cell-enriched cultures [13] . Hence , it was impossible to exclude indirect , accessory cell type-mediated B cell stimulation or IL-10 production by other cell types [59 , 60] . Other reports addressed B cell activation by IL-10 production , but did not study the regulatory activity of schistosome antigen-exposed B cells compared to unstimulated B cells [14] . It must be emphasized that the sole demonstration of upregulated IL-10 expression is not sufficient to characterize B cells as Breg cells , as IL-10 can fulfill other roles in B cell biology independent from a regulatory function . We thus present the first report on direct induction of functional Breg cells with in vitro regulatory activity by helminth antigens . With respect to the development of therapeutic applications , it would be interesting to see whether SEA-induced Breg cells are more potent than Breg cells induced by other compounds , like TLR7 or TLR9 ligands . Opposite to SEA , stimulants like R848 and CpG also induce substantial amounts of B cell proliferation and pro-inflammatory cytokines like IL-6 in addition to high levels of IL-10 . It is currently unknown which side-effects would result in a therapeutic application , but it is tempting to speculate that compounds that selectively induce IL-10 are more preferable . We tried to compare IL-10-producing B cells induced in vitro by different agents , including SEA , for their capacity to inhibit allergic airway inflammation in vivo . We were however not able to confirm a suppressive capacity for any of the conditions despite the usage of a published model [31] . In the past , we have successfully applied adoptive transfers of in vivo , schistosome-induced Breg cells ( generated during a natural infection ) in allergic airway inflammation models [12] . Therefore , we assume that underlying differences between in vivo and in vitro stimulation of Breg cells may be crucial for the activity in a disease model . This may be related to issues like a differential homing or to the strength and kinetics of activation and cytokine production which determine the suppressive capacity of Breg cells on bystander immune activation in the host . Knowing that schistosome antigens can directly induce Breg cell development , we next addressed signals that regulate or enhance antigen-induced B cell IL-10 expression . In previous studies , CD40 engagement was described to induce B cell IL-10 expression [18 , 19 , 50 , 51 , 59] . This is in line with our results showing that SEA-induced IL-10 expression was significantly increased by addition of agonistic CD40 Ab . This points to the potential of a combined therapy for inflammatory diseases using helminth antigens together with anti-CD40 Ab treatment . Indeed , a report by the group of Mauri et al . provided evidence that experimental therapy with an agonistic Ab against CD40 can ameliorate autoimmune disease [61] , as did cellular therapy with Breg cells [50] , although a combined treatment was not yet tested . The groups of Fillatreau and Mauri proposed a two-step model for the acquisition of regulatory properties by B cells , with exposure to innate stimuli—such as TLR ligands—as one step and CD40 or BCR engagement as second step to establish Breg cell function [36 , 50] . We found a similar dependency for in vivo Breg induction by helminth antigen , for which CD40 ligation was crucial . Our data also show that , although Breg cell induction in vitro can be achieved alone without additional stimuli , engagement of CD40 further enhances this effect . Several cell types including T cells , B cells , DCs , basophils , NK cells , mast cells and macrophages express CD40 ligand ( CD40L , CD154 ) [62] and could in principle serve as interaction partner ligating CD40 on B cells . It is however tempting to speculate that neutrophils play a role as they have been reported to express CD40L and activate MZ B cells for immunoglobulin production in a contact-dependent manner [63] . As we found MZ B cells to be the main IL-10-producing B cell subset , it was tempting to speculate that accessory cell types of the splenic marginal zone interact with schistosome antigens , and subsequently drive the development of MZ Breg cells . Macrophages of the splenic MZ were of particular interest because of their known interactions with MZ B cells during steady state [28 , 42] and their expression of SIGN-R1 . This C-type lectin receptor was found to bind schistosome antigens in vitro by using SIGN-R1-overexpressing fibroblasts [43 , 44] . However , it was unknown whether MZ macrophages can capture SEA in vivo and are important for B cell IL-10 expression . As hypothesized , we found macrophages of the MZ to efficiently bind SEA upon in vivo administration . However , Breg cell induction was not affected upon in vivo depletion of macrophages , thus excluding a major role of macrophages in this process . Indirectly , also Mangan et al . [10] showed that macrophages were dispensable for Breg cell induction during schistosomiasis , as macrophage depletion did not affect the B cell-mediated control of anaphylaxis . The immunological role of SEA-binding MZ macrophage subsets and the identity of the binding receptor remain to be determined . A limited number of reports is available that used specific helminth antigens to induce B cell IL-10 expression , namely the filarial antigen ES-62 [64] and the oligosaccharide lacto-N-fucopentaose III ( LNFP III ) that contains the LeX trisaccharide antigen present on various schistosome glycoproteins [13] . Both antigens were either applied in vivo or used in vitro for stimulation of B cell-enriched cultures that still contained other cells , which means it remains unclear whether those antigens can directly bind to and interact with B cells , without indirect support from other cell types . Our study therefore identified IPSE/alpha-1 as the first helminth molecule with direct Breg cell-inducing capacity in mice and humans . Importantly , this capacity was resembled by recombinant IPSE , which is an important prerequisite for a possible therapeutic use . The use of helminth molecules for therapeutic purposes has gained renewed interest as controlled human infections with helminths showed disappointing effects in recent phase II and phase III trials ( reviewed in [65] ) . More studies are required to define the optimal antigen , dose , time point and length of treatment as well as the suitability to treat specific inflammatory diseases [66 , 67] . Therefore , the identification of helminth-specific Breg-inducing antigens is warranted , even more so as the availability of active recombinant forms will ultimately allow its production under GMP conditions . IPSE/alpha-1 was originally described as basophil IL-4-inducing principle of Schistosoma eggs [53] and was shown to induce a mixed Th1/Th2 type of immune response in spleen upon in vivo administration [68] . In our assays , IPSE/alpha-1 directly interacted with murine B cells via a still unknown receptor , which led to activation , induction of B cell IL-10 secretion and Treg cell induction in vitro . Although IPSE/alpha-1 is a highly glycosylated protein , we consider a role of IPSE-related glycans as unlikely because both , pIPSE and nIPSE were capable to induce B cell IL-10 expression despite differences in glycosylation ( native IPSE contains Lex motifs [69] , while plants per definition cannot make Lex motifs [70] . In addition , natural omega-1 and IPSE share a similar glycosylation ( Lex related ) but have opposing activities both in B cells ( here ) and on DCs [71] . IPSE/alpha-1 has been shown to not only bind to IgE but also to IgG , both to Fc and Fab fragments [53] . We therefore hypothesize that IPSE could bind to B cells via the B cell receptor or surface-exposed IgG . Particularly important for a possible therapeutic use is our finding that both natural and plant-derived IPSE/alpha-1 induced IL-10 expression also in human CD1d+ Breg cells . This is even more intriguing as the CD1d+ B cell subset has been previously described to be increased in number and activity both in experimental infections in mice and in people living in endemic areas [12 , 58] . We propose a mechanism of schistosome-induced Breg cell induction in which B cells directly interact with schistosome egg antigens by binding and internalizing antigen , lysosomal processing and subsequent up-regulation of CD86 and IL-10 expression . The MZ B cell subset appeared to be particularly responsive , and CD40 engagement further enhanced Breg cell activity . Furthermore , we have successfully identified the secreted egg antigen IPSE/alpha-1 as one of the Breg-inducing antigens . These egg antigen-induced Breg cells were potent in driving Treg cell development , allowing for induction of two potent regulatory responses by the same antigen . To our knowledge , our study provides the first description of a helminth-specific molecule that interacts with and induces Breg cells , and a mechanistic insight into how schistosomes interact with their host , influence its regulatory immunity and thereby promoting their prolonged survival in the host .
Female C57BL/6OlaHsd mice from Harlan , TLR4-deficient mice ( on C57BL/6 genetic background , kindly provided by Dr . S . Akira , Osaka , Japan ) , TLR2-deficient mice ( on C57BL/6 genetic background , kindly provided by the group of Dr . K . Willems van Dijk ) , SIGN-R1-deficient mice ( on C57BL/6 genetic background , kindly provided by the group of Dr . W . Unger ) , DEREG ( DEpletion of REGulatory T cells ) mice ( on C57BL/6 genetic background , kindly provided by Dr . T . Sparwasser ) and IL-10-GFP reporter ( TIGER ) mice ( on C57BL/6 genetic background , kindly provided by Dr . R . A . Flavell ) were housed under SPF conditions in the animal facilities of the Leiden University Medical Center in Leiden , The Netherlands , and used for experiments at 8–14 weeks of age . Percutaneous infection of mice with S . mansoni was performed as described elsewhere [12] , and mice sacrificed in the chronic phase ( 14–15 weeks ) post infection . Freshly isolated S . mansoni eggs from trypsinized livers of hamsters infected for 50 days were washed in RPMI medium with 300 U/ml penicillin , 300 μg/ml streptomycin ( both Sigma-Aldrich , Zwijndrecht , The Netherlands ) and 500 μg/ml amphotericin B ( Thermo Fisher Scientific , Breda , The Netherlands ) and then kept at -80°C . SEA , AWA , omega-1 , kappa-5 and IPSE/alpha-1 were prepared and isolated as described previously [53 , 71 , 72 , 73] . The purity of the antigen preparations was checked by SDS-PAGE and silver staining , and protein concentrations determined using the BCA procedure . The antigen preparations had an endotoxin content of less than 150 ng/mg protein ( SEA ) or 3 ng/mg protein ( purified molecules ) as tested by Limulus Amoebocyte Lysate ( LAL ) test and TLR4-transfected HEK-reporter cell lines ( kindly provided by Prof . Golenbock , University of Massachusetts Medical School , Boston , USA ) . Recombinant IPSE was produced by transient expression in Nicotiana benthamiana and purified according to the methods described in [70] . In short , the complete sequence encoding the 134 AA mature Schistosoma mansoni IPSE ( Smp_112110 ) was codon optimized and preceded by a signal peptide from the Arabidopsis thaliana chitinase gene ( cSP ) and a N-terminal 6x histidine-FLAG tag ( H6F ) was included . The full sequence was synthetically constructed at GeneArt and cloned into a pHYG expression vector . In all experiments the silencing suppressor p19 from tomato bushy stunt virus in pBIN61 was co-infiltrated to enhance expression . For gene expression the two youngest fully expanded leaves of 5–6 weeks old N . benthamiana plants were infiltrated by injecting Agrobacterium tumefaciens containing the IPSE expression plasmid . N . benthamiana plants were maintained in a controlled greenhouse compartment ( UNIFARM , Wageningen ) and infiltrated leaves were harvested at 5–6 days post infiltration . Plant-produced recombinant IPSE was obtained by applying leaf apoplast fluid containing IPSE to Ni-NTA Sepharose ( IBA Life Sciences ) in 50 mM phosphate buffered saline ( pH 8 ) containing 100 mM NaCl . Bound IPSE was eluted with phosphate buffered saline ( pH 8 ) containing 0 . 5M imidazole . Total soluble apoplast proteins and purified IPSE were separated under reducing/non-reducing conditions by SDS-PAGE on a 12% Bis-Tris gel ( Invitrogen ) and subsequently stained with Coomassie brilliant blue staining . Single cell suspensions of murine spleens were prepared by dispersion through a 70 μm cell strainer ( BD Biosciences , Breda , The Netherlands ) , and erythrocytes depleted by lysis . For analysis of splenic myeloid cell populations , spleens were digested for 1 hour at 37°C by incubation with collagenase D ( 2 mg/ml; Roche , Woerden , The Netherlands ) and DNase I ( 2000 U/ml; Sigma-Aldrich ) before dispersion . B cells were purified from spleens by using anti-CD19 MicroBeads ( Miltenyi Biotec , Leiden , The Netherlands ) following the manufacturer’s protocol . Purity was routinely ~95–98% . After a typical CD19 MACS sort , circa 83% of all contaminating cells were CD3+ T cells , 7% CD11b+CD11c+ cells , 2% CD11b+ CD11c- cells , and 8% other cells . To determine cytokine secretion of splenic B cell subsets , CD19+ B cells were subsequently sorted by flow cytometry for follicular B cells ( CD23+CD21low ) and marginal zone B cells ( CD23-CD21hi ) which resulted in purities of > 98% . CD4+ T cells were purified from spleens by negative selection and depleted of CD25-expressing cells using anti-CD25 MicroBeads ( Miltenyi Biotec ) . Peripheral blood mononuclear cells ( PBMCs ) were isolated from heparinized blood of healthy volunteers by Ficoll gradient centrifugation , and B cells were purified from PBMCs by using anti-CD19 MicroBeads ( Miltenyi Biotec , Leiden , The Netherlands ) following the manufacturer’s protocol . Mouse splenic CD19+ B cells ( 1 . 5x106/ml ) were cultured in medium ( RPMI 1640 glutamax; Thermo Fisher Scientific ) , containing 5% heat-inactivated Fetal Bovine Serum ( FBS; Greiner Bio-One , Alphen aan den Rijn , The Netherlands ) , 5 × 10−5 M 2-Mercaptoethanol ( Sigma-Aldrich ) and antibiotics ( 100 U/mL penicillin and 100 μg/mL streptomycin; Sigma-Aldrich ) . Human B cells ( 1 . 5 x 106/ml ) were cultured in medium ( RPMI 1640; Thermo Fisher Scientific ) , containing 10% heat-inactivated Fetal Bovine Serum , pyruvate ( 1 mM ) , glutamate ( 2 mM ) and antibiotics ( 100 U/mL penicillin and 100 μg/mL streptomycin; all Sigma-Aldrich ) . The following stimuli were added as indicated in the figure legends: SEA ( 20 μg/ml ) , SEA depleted of IPSE ( SEAΔIPSE , 20 μg/ml ) , natural ( 1 , 5 , 10 , 20 μg/ml ) or plant-derived IPSE ( 10 μg/ml ) , omega-1 ( 1 μg/ml ) , kappa-5 ( 1 , 5 , 10 , 20 μg/ml ) , AWA ( 5 , 10 , 20 μg/ml ) . For some conditions , co-stimulatory rat anti-mouse CD40 antibody ( 2 μg/ml; clone 1C10; BioLegend , Uithoorn , The Netherlands ) or goat anti-mouse IgM ( 0 . 5 , 1 , 2 μg/ml; Jackson ImmunoResearch , Suffolk , UK ) was added to the culture . After 3 days culture at 37°C , supernatants were collected for cytokine analysis by ELISA . Cells were restimulated with PMA ( 100 ng/ml ) and ionomycin ( 1 μg/ml ) for 4 hours in the presence of Brefeldin A ( 10 μg/ml; all Sigma-Aldrich ) for flow cytometric analysis of intracellular IL-10 . In experiments addressing involvement of lysosomal acidification , the inhibitor chloroquine ( 5 μM; Sigma ) was added at the start of a two days culture and refreshed after 24 hours , and the TLR ligands CpG ODN 1826 ( 5 μg/ml; Invivogen ) or Pam3Cys ( 10 μg/ml; Invivogen ) used as control stimuli next to SEA . For in vivo stimulation of B cells , mice were i . p . injected with two doses of 5000 eggs or 100 μg SEA in PBS , determined as optimal doses where B cell IL-10 production plateaued in prior dose-titration experiments , and PBS or 100 μg human serum albumin ( HSA ) in PBS as control 7 days apart . At day 14 after the first injection , splenic B cells were harvested and cultured in medium at 1 . 5 x 106 cells/ml or restimulated for 2 days with SEA ( 20 μg/ml ) to allow detection of cytokines , as established for in vivo schistosome-exposed B cells before [12] . Supernatants were collected to determine cytokine concentration by ELISA . Cells were cultured for additional 4 hours with Brefeldin A ( 10 μg/ml; Sigma-Aldrich ) to detect intracellular IL-10 by flow cytometry . In some experiments , mice were treated i . p . with 200 μg of hamster anti-mouse CD40 ligand blocking antibody ( clone: MR1 ) or 200 μg hamster IgG as control ( Jackson ImmunoResearch ) 4 times , every 4 days starting one day before SEA or PBS treatment . For in vivo depletion of macrophages , mice were i . p . injected with 200 μl clodronate-containing liposomes and control mice with 200 μl of PBS liposomes ( ClodronateLiposomes . com , Amsterdam , The Netherlands ) [45] three weeks prior to egg antigen treatment . Successful and specific depletion of splenic macrophage subsets was confirmed by fluorescence microscopy and flow cytometry . In vitro or in vivo SEA-stimulated CD19+ B cells were co-cultured with MACS-sorted CD4+CD25- T cells at 1:1 ratio ( each 1 x 106/ml ) to test for in vitro Treg cell induction . After 4 days , Treg cell frequencies were determined by flow cytometry by gating for Foxp3+CD25+ cells in the CD3+CD4+ T cell population , and culture supernatants collected for subsequent ELISA . SEA , IPSE/alpha-1 and ovalbumin ( OVA ) were fluorescently labeled with PF-488 or PF-647 using the PromoFluor labeling kits ( PromoCell , Heidelberg , Germany ) according to the manufacturer’s protocol . For some experiments , SEA was co-labeled with the pH-sensitive pHrodo Red dye ( Thermo Fisher Scientific ) . After protein labeling , non-reacted dye was removed using Zeba desalt spin columns ( Thermo Fisher Scientific ) . For analysis of binding in vitro , CD19+ splenic B cells were cultured for 60 minutes at 37°C with 20 μg/ml fluorescently labeled SEA or 1–10 μg/ml of IPSE antigen , then washed in ice-cold PBS before analysis by flow cytometry . For analysis of in vivo binding , mice were i . v . injected with 200 μg of fluorescently labeled SEA or OVA as non-schistosomal control protein and spleens snap-frozen 30 minutes to 24 hours later . Binding of SEA to B cells was analyzed by confocal fluorescence microscopy of tissue sections and by flow cytometry . Basophils were purified from 250 ml of peripheral blood of healthy human donors to a mean purity of 99% by a three-step protocol consisting of a density gradient centrifugation via Ficoll/Percoll ( 100/6 , density 1 . 080 g/l ) , followed by enrichment of the basophils via counter flow elutriation and final purification by magnetic cell sorting using the basophil isolation kit II for negative selection of basophils ( Miltenyi-Biotech ) . Purified basophils were cultured in Iscove's Modified Dulbecco's Media ( IMDM; PAA ) containing 2 mM glutamine ( PAA ) , 5 μg/ml insulin ( Gibco ) , 50 μg/ml apo-transferrin ( Sigma-Aldrich ) , 100 μg/ml Pen/Strep ( PAA ) , 10% heat-inactivated Fetal Calf Serum ( FCS-Gold; PAA ) and 2 . 5 ng/ml IL-3 ( kind gift of Kirin Brewery , Japan ) . Basophils were pre-incubated for regeneration for 30 min at 37°C , 6% CO2 , and then stimulated at a concentration of 0 . 025 x 106 basophils /ml in 96well flat-bottom culture plates in 100 μl at 37°C , 6% CO2 . Concentration of stimuli was as indicated . Culture supernatants were collected after 18h and stored at -20°C . Flow cytometric analysis of murine B cells was performed by staining with fluorochrome-labeled antibodies against CD19 , CD21 ( both BD Biosciences ) , CD23 , CD40 , CD86 or IL-10 ( all eBioscience ) after fixation with 1 . 9% paraformaldehyde and permeabilization with 0 . 5% saponin ( Sigma-Aldrich ) . Human B cells were stained for CD19 , CD38 ( both BD Biosciences ) , CD24 , CD27 ( both eBioscience ) , CD1d , IL-10 , TNF ( all Biolegend ) , and CD39 ( Sony Biotechnology , San Jose , USA ) . Splenic myeloid cell subsets were discriminated using fluorochrome-labeled antibodies against CD11b , CD11c ( both eBioscience ) , CD8 , Ly6C ( both Biolegend ) , F4/80 ( AbD Serotec , Puchheim , Germany ) , Gr-1 ( BD Biosciences ) , and Siglec-1 ( Dr . J . den Haan , VUMC , Amsterdam , The Netherlands ) . Treg cells were fixed and permeabilized with the eBioscience Foxp3 fixation/permeabilization kit and stained using fluorochrome-labeled antibodies against CD3 , CD4 , Foxp3 ( all eBioscience ) and CD25 ( BD Biosciences ) . All cells were stained with Aqua dye ( Thermo Fisher Scientific ) prior to fixation to discriminate dead cells . For all flow cytometric stainings , FcγR-binding inhibitor ( 2 . 4G2 ) was added and FMOs were used for gate setting . Flow cytometry was performed using a FACSCanto or Fortessa ( BD Biosciences ) . The concentration of murine IL-6 and IL-10 as well as human IL-4 present in culture supernatants was quantified by commercial ELISA kits according to the manufacturer’s instructions ( BD Biosciences or Eli-Pair , Diaclone ) . Spleens were snap-frozen in O . C . T . medium ( Tissue-Tek; Sakura , Alphen aan den Rijn , The Netherlands ) . Cryosections ( 10 μm ) were fixed in ice cold acetone for 10 minutes , air-dried , and blocked in 1% BSA plus 20% FBS in PBS before staining with Abs at room temperature . Cryosections were incubated with rat anti-mouse Siglec-1 ( clone SER-4; provided by Dr . J . den Haan , VUMC , Amsterdam , The Netherlands ) followed by Alexa555-conjugated goat anti-rat IgG ( Invitrogen ) , anti-SIGN-R1 Alexa647 ( clone 22D1; Dr . J . den Haan ) and anti-B220 eFluor450 ( eBioscience ) . Images were acquired using a Zeiss LSM 710 confocal laser scanning microscope with Zen software ( Carl Zeiss Microimaging , Jena , Germany ) . All data are presented as mean ± standard error of the mean ( SEM ) . Statistical analysis was performed with GraphPad Prism version 6 . 00 for Windows ( GraphPad Software , La Jolla , CA , USA ) using nonparametric Mann-Whitney U test to compare different groups and Wilcoxon paired test to compare B cell subsets . One-sample t-test of log-transformed data was applied to calculate significant changes for data which are expressed as fold increase . All p-values < 0 . 05 were considered significant . All animal studies were performed in accordance with the Animal Experiments Ethical Committee of the Leiden University Medical Center ( DEC-12204 ) . The Dutch Experiments on Animals Act is established under European Guidelines ( EU directive no . 86/609/EEC regarding the Protection of Animals used for Experimental and Other Scientific Purposes ) . For the isolation of B cells from PBMCs , human subjects were recruited within the framework of the study P09 . 170 , which was approved by the Medical Ethical Committee of the Leiden University Medical Center . For the isolation of basophils , donors were recruited under approval by the Ethics Committee of the University of Luebeck ( AZ-12-202A ) . Studies were performed according to the declaration of Helsinki and all participants were adults and have given written informed consent .
|
Infection with helminth parasites is known to be inversely associated with hyper-inflammatory disorders . While Schistosoma ( S . ) mansoni has been described to exert its down-modulatory effects on inflammation by inducing a network of regulatory immune cells such as regulatory B ( Breg ) , the mechanisms of Breg cell induction remain unclear . Here , we use in vivo and in vitro approaches to show that antigens from S . mansoni eggs , among which the major glycoprotein IPSE/alpha-1 , directly interact with splenic marginal zone B cells of mice which triggers them to produce the anti-inflammatory cytokine IL-10 and their capacity to induce regulatory T ( Treg ) cells . We also found that IPSE/alpha-1 induces IL-10 in human CD1d+ B cells , and that both natural and recombinant IPSE/alpha-1 are equally effective in driving murine and human Breg cells . Our study thus provides insight into the mechanisms of Breg cell induction by schistosomes , and an important step towards the development of helminth-based treatment strategies against hyper-inflammatory diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"helminths",
"immunology",
"light",
"microscopy",
"animals",
"developmental",
"biology",
"microscopy",
"molecular",
"development",
"research",
"and",
"analysis",
"methods",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"animal",
"cells",
"fluorescence",
"microscopy",
"t",
"cells",
"spectrophotometry",
"immune",
"system",
"antibody-producing",
"cells",
"cytophotometry",
"cell",
"biology",
"b",
"cells",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"regulatory",
"t",
"cells",
"macrophages",
"organisms"
] |
2017
|
Schistosome egg antigens, including the glycoprotein IPSE/alpha-1, trigger the development of regulatory B cells
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.