我国HIV-1流行株CRF07_BC和B’耐药进化研究

【作者】 李臻鹏；

【导师】 马丽英；

【作者基本信息】 中国疾病预防控制中心 ， 病原生物学， 2014， 博士

【摘要】 HIV-1CRF07_BC和HIV-1B亚型是我国艾滋病的主要流行病毒。尽管抗病毒治疗能够延长HIV-1感染者的生存寿命、有效的减少艾滋病的传播,但HIV耐药性的产生不仅导致了抗病毒治疗失败,而且增加了耐药病毒株传播的风险。由于HIV-1在药物和体内免疫的双重压力下产生的耐药性及其病毒进化的高度复杂性,使得现有的病毒耐药进化分析受到一定的局限,特别是针对我国流行的HIV-1非B亚型的研究较为有限。为此,本研究从病毒的整体分子变异进化的角度,首先开发了基于选择压力的相关突变分析软件,对HIV-1CRF07_BC的逆转录酶区相关突变模式、耐药进化通路和我国既往献血HIV-1感染人群病毒基因变异对长期抗病毒治疗的影响进行了系统的分析。该项研究有助于理解HIV-1在抗病毒治疗下特异的进化模式和多态性位点在耐药发生发展中的作用,为改进药物组合的治疗策略提供重要的科学依据。1.基于选择压力的相关突变分析软件的开发相关突变(Correlated Mutation)是进化生物学中的一个基本概念,在一个可编码蛋白质的基因上,氨基酸突变会受到蛋白质功能的限制。我们开发了一个基于R/Bioconductor的软件包CorMut来发现阳性选择位点之间的相关突变。软件包组合了Ka/Ks比值和相关突变分析。CorMut提供了计算单个位点或特异氨基酸Ka/Ks并发现他们之间相关突变的函数。CorMut提供了三种方法来发现相关突变,包括条件选择压力、互信息和Jaccard系数。计算主要由两个步骤组成：第一,发现阳性选择位点；第二,计算阳性选择位点之间的突变关联。值得注意的是第一个步骤是可选择的。同时,CorMut可以通过构建相关突变网络方便的比较两种条件下的相关突变关系。该软件包已发布在国际的生物信息学平台Bioconductor网站,至今已被他人下载2000余次。(http://www.bioconductor.org/packages/devel/bioc/html/CorMut.html)。2.我国HTV-1CRF07_BC流行株逆转录酶区相关突变模式分析为了解HIV-1CRF07_BC的耐药进化,本研究选择了552例未治疗的HIV-1感染者和261例接受抗病毒治疗的HIV-1感染病人[zidothymidine (AZT)/lamivudine (3TC)/nevirapine (NVP)或AZT/3TC/efavirenz (EFV)],分析了HIV-1逆转录酶区和蛋白酶区突变共变异的改变,采用分层网络的方法展示突变的共变异。首先,识别了逆转录酶区和蛋白酶区的三种类型的特征突变：治疗相关突变、治疗拮抗突变和重叠多态性位点。在6个治疗相关突变(K103N,M184V, Q197K, G190A,Y181C和M230L)和5个重叠多态性位点(A36E,R135I,R277K,L283I和D291E)之间发现了10对显著的相关突变。同时,在逆转录酶区的治疗相关突变(I132L)与与蛋白酶区的重叠多态性位点(L10I)之间发现了一对相关突变。最后,发现逆转录酶区与蛋白酶区的重叠多态性位点在整体上与治疗相关突变之间存在显著的相关性,说明多态性位点对耐药突变位点产生一定的影响。与HIV-1欧美B亚型相比,CRF07_BC表现出独特的突变共变异模式,这种存在于HIV-1CRF07_BC的逆转录酶区和蛋白酶区的一些多态性位点在耐药发生过程中可能发挥着重要作用。该研究对于揭示HIV-1CRF07_BC在抗病毒治疗下特异的进化模式,为今后我国药物组合的改进提供重要的科学依据。3.我国HIV-1CRF07_BC流行株逆转录酶区耐药进化通路分析该部分探讨HIV-1CRF07_BC重组病毒在药物压力下的进化通路,并与B亚型进行比较。选择上述未治疗和治疗的HIV感染者样本病毒的逆转录酶区序列,采用选择压力的方法识别耐药相关位点,使用贝叶斯网络法构建CRF07BC在抗病毒治疗下的耐药进化通路,同时利用国际HIV耐药数据库中的数据构建B亚型在相同治疗方案下耐药进化通路,并与CRF07_BC进行了比较。预测的CRF07_BC主要耐药位点为K103N、Q197K、V179D和Y188L。预测的B亚型主要耐药位点为M184V、K103、Y181C、T69N、G190A, K238T、Y188H和P225H。两者重叠较小,但经典的TAM1(41L,210W和215Y)和TAM2(67N,70R和219E/Q)通路均存在于两个通路,与HIV-1欧美B亚型不同的是预测的CRF07_BC的主要耐药位点中不包含TAM相关突变,并且在核苷类药物相关突变与非核苷类药物相关突变之间存在较强的依赖关系。CRF07_BC形成了以K103N、Q197K、V179D和Y188L为主要耐药位点的独特耐药进化通路,并与B亚型的耐药进化通路相比有较大差异。4.我国既往献血HIV-1感染人群病毒基因变异对长期抗病毒治疗的影响为了研究HIV-1在抗病毒治疗下的耐药进化,尤其是其在长期抗病毒治疗情况下的动态变异以及与抗病毒治疗效果,本研究利用我国最初在河南安徽开展建立的抗病毒治疗回顾性队列,首先选择每个病人接受抗病毒治疗最近时间点和未治疗时的各一条逆转录酶区序列,得到一组病人组成的治疗前后的横断面数据,比较治疗前后的HIV-1逆转录酶区序列,然后识别横断面数据的B’耐药新突变及突变关联性,从而分析了HIV-1的进化特征。从队列的角度探索这些新突变及突变关联的动态变化及其与抗病毒治疗效果的关系。在基于横断面的分析中,我们预测了治疗组的新耐药相关突变位点,共得到42个耐药相关的位点。进一步建立了HIV-1三类治疗特征性突变(治疗相关突变、治疗拮抗突变和重叠多态性位点)在治疗前后的相关突变网络。在治疗相关突变与重叠多态性位点之间发现了17对相关突变,这一结果进一步提示治疗前后阳性选择的重叠多态性位点对于耐药相关位点产生的重要作用。这些相关突变对分为上下两大类,逐渐升高和先升高后降低。研究结果提示HIV-1多态性位点数目与病毒学失败时间呈显著的负相关关系,而多态性位点数目与耐药位点出现的时间负相关并不显著。进一步从队列的角度,观察这些多态性位点是否对与之相关的耐药位点的出现时间产生影响,我们筛选得到了16对多态性位点与治疗相关位点相关突变对对耐药相关位点出现的时间存在显著的影响,进一步说明HIV-1逆转录酶区多态性位点变异对于耐药、病毒学失败的产生一定的影响,这一结果提示在设计抗病毒治疗方案时应考虑HIV多态性位点,从而避免某些耐药性位点的产生,进而选择有效的抗病毒药物。更多还原

【Abstract】 HIV-1CRF07_BC and B’ are the main epidemic strains in china. Anti-viral therapy has effectively reduce the transmission of aids, prolong the survival of life, however, antiviral treatment failure correspondingly increased owing to drug resistance, which has greatly affected the effects of treatment and increased the risk of drug-resistance strains transmission. The drug resistance and the complex evolution of HIV-1under the drug and immune pressure have posed the change to the traditional drug resistance analysis, especially for the non-B subtypes circulating in china. So, this study was carried out systematically from the view of viral genetic variations. Firstly, this study developed a software package CorMut for correlated mutation analysis, and then we did the analysis for mutation covariation and drug resistance pathway of HIV-1CRF07_BC reverse transcriptase during antiretroviral therapy, as well as the influence of HIV-1variation to treatment effect in former plasma/blood donors. This study will help us understand the distinct drug resistance evolution patterns and the roles of polymorphisms in drug resistance development, related studies will provide crucial support to design multiple drug combination for therapy.1. The development and application of software for computing correlatedmutations based on selection pressureCorrelated mutations constitute a fundamental idea in evolutionary biology, and understanding correlated mutations will, in turn, facilitate the understanding of the genetic mechanisms governing evolution. We developed an R/Bioconductor package to detect the correlated mutations among positive selection sites by combining Ka/Ks ratio and correlated mutations analysis. CorMut is an R package designed to compute correlated mutations in the unit of codon or amino acid mutation.CorMut incorporates three classical methods to detect correlated mutations, including conditional selection pressure, mutual information and Jaccard index. The computation for correlated mutations consists of two steps:First, the positive selection sites are detected using the selection pressure-based method. Second, the mutation correlations are computed among the positive selection sites using the three methods described above. CorMut also enables the comparison of correlated mutations between two different evolutionary conditions. CorMut is released under the GNU General Public License within Bioconductor project, and freely available at following website http://bioconductor.org/packages/release/bioc/html/CorMut.html, and now CorMut has been downloaded more than2000.2. Mutation covariation of HFV-1CRF07_BC reverse transcriptase during antiretroviral therapyAs an epidemic recombinant subtype in China, the variation of HTV-1CRF07_BC under antiretroviral therapy (ART) has not been completely understood. The changes of mutation covariation in the reverse transcriptase (RT) and protease (PR) of HIV-1CRF07_BC were analyzed by comparing the552treatment-naive patients and261treatment patients under ART with zidothymidine (AZT)/lamivudine (3TC)/nevirapine (NVP) or AZT/3TC/efavirenz (EFV). Meanwhile, the stratified networks were used to display the mutation covariation.At first, three types of featured mutations for RT and PR were identified. These included treatment-associated mutations, treatment-agonistic mutations and overlapping polymorphisms. Ten pairs of significant correlated mutations were found between6treatment-associated mutations (K103N, M184V, Q197K, G190A, Y181C and M230L) and5overlapping polymorphisms (A36E, R135I, R277K, L283I and D291E). Meanwhile, a pair of correlated mutation between treatment-associated mutations (I132L) and overlapping polymorphisms (L101) for PR was also detected. Finally, overlapping polymorphisms for RT and PR were both found to have significant correlations with treatment-associated mutations, indicating the possible association between polymorphisms and drug resistance. The mutation covariations for RT and PR of HIV-1subtype B under the same regimens were also analyzed, and we found that CRF07_BC showed a distinct pattern of mutation covariation compared with subtype B.Some polymorphisms may play crucial roles in the development of drug resistance in HIV-1CRF07_BC. The analysis could help reveal the specific evolution of HIV-1CRF07_BC under ART, and the information might be useful for improving the efficacy of drug combinations.3. Resistance evolutionary pathway analysis of HTV-1CRF07_BC reverse transcriptaseWe studied the resistance evolution pathway of HIV-1CRF07_BC, a major circulating recombinant form in China, under drug selection pressure, and made a comparison with B subtype under the same regimens. Based on the reverse transcriptase region of CRF07_BC HIV-1from588treatment-naive and274treatment patients, we have used selection pressure based method to select resistance-associated mutations, and Bayesian network to construct the resistance evolutionary pathway under antiretroviral therapy. Meanwhile, we also construct the resistance evolutionary pathway for B subtype with the same regimens using the data from HIV resistance database, and made a comparison with CRF07_BC. We have identified the major resistance mutations for CRF07_BC, including K103N, Q197K, V179D and Y188L. While for B subtype, the major resistance mutations include M184V, K103N, Y181C, T69N, G190A, K238T, Y188H and P225H. Much difference was observed between these two classes. However, the classical TMA1(41L,210W and215Y) and TMA2(67N,70R and219E/Q) pathways exist in both pathways. As different from B subtype, the predicted major drug resistance mutations for CRF07_BC did not contain TAM-related mutations, and the relationship between nucleoside reverse transcriptase inhibitor-related mutations and non-nucleoside reverse transcriptase inhibitor-related mutations showed strong dependence. HIV-1CRF07_BC showed distinctive resistance evolutionary pathway, the mutations K103N, Q197K, V179D and Y188L were the major resistance mutations, and different resistance evolutionary pathways were observed between HTV-1CRF07_BC and B subtype.4. The influence of HIV-1genome variation to therapy effect in plasma donors infected with HIV-1As main epidemic subtype, the drug resistance of HTV-1B’ was rarely reported, the resistance evolution under ART, especially the dynamic genetic variation and its relation to long-term therapy effect need further investigate. We first transform the cohort data, sourcing from Henan and Anhui, into the cross-sectional data, then compare HIV-1RT region between the treatment-naive patients and treatment patients under ART, then identified42resistance-related mutations and mutation correlations. After that we investigate the dynamics of resistance-related mutations and mutation correlations and its relation to therapy effect in cohort view. At first, three types of featured mutations for RT were identified. These included treatment-associated mutations, treatment-agonistic mutations and overlapping polymorphisms, and then we constructed the correlated mutation network based on these features mutation.17pairs of significant correlated mutations were found between treatment-associated mutations and overlapping polymorphisms. This evidence further indicated that some polymorphisms may play crucial roles in the development of drug resistance. These correlated mutations can be basically divided into two parts after cluster analysis:those gradually increased and those increased first and then decreased.The polymorphisms in RT region may influence the resistance and virological failure, our results indicated there exist significant negative correlation between the number of polymorphisms and the time to virological failure, while the correlation between the number of polymorphisms and the time to resistance was not significant. In addition, we observed the influence of these polymorphisms to the appearance of resistance mutations. We screened16such polymorphisms and resistance-related mutations pairs. This result indicated the influence of polymorphisms in RT region to HIV-1drug resistance and virological failure, and polymorphisms may be considered before designing therapy regimens to avoid the emerging of some resistance mutations.更多还原