HIV-1 transmitted drug resistance-associated mutations and mutation co-variation in HIV-1 treatment-naïve MSM from 2011 to 2013 in Beijing, China

A total of 262 HIV-1 positive individuals were randomly recruited from 2011 to 2013 at voluntary counseling and testing sites (VCT) in Beijing Chaoyang District Center for Disease Control and Prevention, following three criteria: having had history of MSM sex, being ART-naïve and newly diagnosed. This study was approved by the Institutional Research Ethics Community, China Chaoyang CDC, and all subjects signed informed consent forms prior to blood collection. Epidemiological data was collected by trained interviewers. HIV-1 infection status was determined by an enzyme immunoassay (ELISA, Wantai, China) and confirmed by Western blot assay (HIV BLOT 2.2, MP Diagnostics, Singapore). Blood plasma was separated and stored at −70°C prior to genetic analysis.

Viral RNA was extracted from 200 μl EDTA-anticoagulated plasma using a QIAamp viral RNA kit (Qiagen Inc., Germany) according to the manufacturer’s instructions. The HIV-1 pol gene (1,197 bp length), containing the full-length protease (PR) gene and the first 300 codons of the reverse transcriptase (RT) gene, were amplified and sequenced, using an in-house drug resistance genotyping method as previously described [18]. The target sequence was amplified with One Step Reverse Transcription PCR reagents (Qiagen Inc., Germany) using primers listed in Table 1. Amplification steps were as follows: reverse transcription at 50°C for 30 min, pre-denaturation at 94°C for 5 min, 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, extension at 72°C for 2.5 min, and an additional extension at 72°C for 10 min. Nested PCR was performed using Taq PCRmaster mix (Qiagen Inc., Germany) with primers in Table 1. The cycling conditions were: pre-denaturation at 94°C for 5 min, 30 cycles of denaturation at 94°C for 30 s, annealing at 63°C for 30 s, extension at 72°C 2.5 min, and an additional extension at 72°C for 10 min. PCR products were visualized by 1% agarose gel electrophoresis and sequenced using ABI 3730xl Automated DNA Analyzer (Applied Biosystems, Foster City, CA). Each step was carried out with negative controls.

All assembled sequences were submitted to the Los Alamos National Laboratory HIV Sequence Database (http://​www.​hiv.​lanl.​gov/​content/​index) to determine HIV genotype, which were further confirmed by phylogenetic analysis using standard reference sequences representing subtypes A–D, F–H, J, K, CRF01_AE, CRF07_BC, and CRF08_BC (www.​hiv.​lanl.​gov). DNA alignment was performed by the Clustal W method using MEGA5 [19], followed by manual adjustment. Phylogenetic analysis was also conducted with MEGA5 using neighbor-joining trees under a Kimura 2-parameter model and tested by the bootstrap method with 1,000 replicates.

Sample pol gene sequences were compared to a consensus sequence using HIV db software (Stanford HIV Drug Resistance Database, http://​hivdb.​stanford.​edu, version 7.0) to detect drug resistance mutations, including major and minor protease inhibitor (PI) resistance mutations, nucleoside reverse transcriptase inhibitor (NRTI), and non-nucleoside reverse transcriptase inhibitor (NNRTI) resistance mutations.

We analyzed co-variation between TDR-associated mutations and positive selected mutation using the CorMut package [20]. Briefly, the procedure was: positively selected mutations were identified using selection pressure (Ka/Ks ratio) based method [21],[22], in which a Ka/Ks value of >1 indicates a positive selection. Log odds (LOD) confidence score was used to measure the significance of selection pressure (cut off > = 2). The 05GX001 strain (subtype CRF01_AE) was used as a reference when performing the computation. The Jaccard similarity coefficient was used to measure the co-variation between TDR mutations and positively selected mutations. Fisher’s exact test was used to check the significance of co-variation. False discovery rate was controlled using the Benjamini and Hochberg procedure with a 0.2 cut-off. P adjusted value < 0.2 was considered statistical significance.

Among 262 HIV-1 positive samples, the pol genes of 223 samples (85.11%) were successfully amplified and sequenced. The mean age of the 223 patients was 30.3 (range: 17–64). 75.3% of subjects were never married, 21.5% were married, and 3.1% were divorced or widowed. More than two-thirds of participants (69.5%) had received college-level or higher education degree. The basic demographic characteristics are shown in Table 2.

Phylogenetic analysis of the amplified pol gene regions (1197 bp) showed that the samples were generally tightly clustered within their respective subtypes (Figure 1). Their genotype distribution was as follows: 135 cases (60.5%) were subtype CRF01_AE, 62 cases (27.8%) were subtype CRF07_BC, 22 cases (9.9%) were subtype B, two were CRF01B, one was CRF55_01B, and one was CRF08_BC (Table 2). There was no significant difference in HIV-1 subtype distributions between each year.

Amplified gene regions were assessed for TDR-associated mutations through the Stanford HIV Drug Resistance Database. Mutations were classified and summarized according to their ability of conferring resistance to PI, NRTI, or NNRTI drug classes. Among our samples, 25.6% (57/223) had TDR-associated mutations, including 16.1% (36/223) for PI mutations, 0.9% (2/223) for NRTI mutations, and 9.0% (20/223) for NNRTI mutations. PI mutations included L10I/V (7.2%, 16/223) and A71L/T/V (6.3%, 14/223), with relatively high frequency; NRTI mutations, included L74I (0.45%, 1/223) and V75L (0.45%, 1/223); and the most frequent NNRTI mutation was V179D/E (5.4%, 12/223). Among these detected mutations, only L74I, M46L, G190E and E138G may result in drug resistance directly. Most of these mutations only conferred potential drug resistance. The detailed frequency of the mutations was shown in Table 3. Only six (2.7%) samples carried mutations conferring known levels of drug resistance, with 1.35% (3/223) against PIs (nelfinavir), 0.45% (1/223) against NRTIs (abacavir, didanosine), and 0.90% (2/223) against NNRTIs (efavirenz, etravirine, nevirapine, riplivirine).

The type and frequency of TDR-associated mutations were different among different HIV-1 subtypes. The proportion of TDR-associated mutations was 23.7% (32/135) among CRF01_AE recombinant strains, the most frequent mutations being L10I/V and V179D/E. Among CRF07_BC recombinant strains, 21.0% (13/62) had TDR-associated mutations, A71T/V being the most frequent. Among subtype B, 54.5% (12/22) had TDR-associated mutations, A71T/V and V106I being the most frequent. Detailed information is summarized in Table 3.

Of note, the distribution of TDR-associated mutations differed by sampling year (Table 3). In 2011, 14.2% (3/21) had TDR-associated mutations, while the latter two years saw rates that were more than two times as high, with 29.4% (37/126) in 2012 and 22.4% (17/76) in 2013.

Co-variation analysis was performed to determine mutations or polymorphisms that were positively selected in association with TDR-associated mutations (Table 4). Eight mutation pairs with significant co-variation were identified in the RT region for CRF01_AE strains between the V179D/E TDR-associated mutation and seven overlapping polymorphisms. No significant mutation pair was identified for the PR region and for other subtypes.