Genetic diversity and drug resistance among newly diagnosed and antiretroviral treatment-naive HIV-infected individuals in western Yunnan: a hot area of viral recombination in China

Blood samples were collected from 320 individuals newly diagnosed with HIV-1 infection in Longchuang and Luxi counties of Dehong prefecture between January 2009 and December 2010. HIV-1 infection was confirmed by Western Blot assay (HIV BLOT 2.2, MP Diagnostics, Singapore). All study participants were ART-naive and provided informed consent. This study was approved by the institutional review board of Yunnan Center for Disease Control and Prevention.

Viral RNA was extracted from 200 μl plasma using an automated extractor (MagNA Pure LC system, Roche, Branchburg, NJ) according to the manufacturer’s instructions. For the amplification of the pol regions, an in-house nested reverse transcription PCR method was used. The target sequence was amplified with One Step Reverse Transcription PCR reagents procured from Takara (Dalian, China) using primers MAW26 (5^′-TTGGAAATGTGGAAAGGAAGGAC-3^′) and RT21 (5^′-CTGTATTTCTG CTATTAAGTCTTTTGATGGG-3^′) in a 25 μl reaction for 30 cycles. Cycling conditions were 50°C for 30 min, 94°C for 5 min, 94°C for 30 s,55°C for 30 s,72°C for 2.5 min, followed with an extension at 72°C for 10 min. The nested PCR was performed using Taq PCR master mix (Tiangen, Beijing, China), using primers PRO-1 (5^′-CAGAGCCAACAGCCCCACCA-3^′) and RT20 (5^′-CTGCCAGTTCTAGCTCTGCTTC-3^′) in a 50 μl reaction for 30 cycles and the cycling conditions were 94°C for 5 min, 94°C for 30 s, 63°C for 30 s, 72°C 2.5 min, followed with an extension at 72°C for 10 min. The resulting PCR product of 1191 bp in length contained a full length protease (PR) gene of 99 amino acid codons and the first 298-codon segment of the reverse transcriptase (RT) gene. PCR products were visualized by 1% agarose gel electrophoresis. Successfully amplified samples were sequenced by Biomed Co. (Beijing, China) on an ABI 3730xl automated DNA analyzer (Applied Biosystems, Foster City, CA) with primers: PROS3 (5^′-GCCAACAGCCCCACCA-3^′), RTAS (5^′-CTCAGATTGGTTGCAC-3^′), RTB (5^′-CCTAGTATAAACAATGAGACAC-3^′), PROC1S (5^′-GCTGGGTGTGGTATTCC-3^′) and RT20S3 (5^′-GTTCTAGCTCTGCTTC-3^′). Each step was carried out with appropriate negative controls to detect PCR-related contamination during the experiments.

The sequence contig assembly was performed using analysis software, Sequencher 4.9 (Gene Codes Corporation, Ann Arbor, MI). The ClustalW Multiple alignment and manual editing were performed using BioEdit Sequence Alignment Editor (Ibis Biosciences, Carlsbad, CA). Edited sequences were evaluated for DR mutations using both the Surveillance Drug Resistance Mutations (SDRM) list recommended by the World Health Organization [18], and the Stanford University Algorithm (http://​hivdb.​stanford.​edu). The HIV-1 subtypes of pol sequences were identified by the REGA HIV subtyping tool (http://​www.​bioafrica.​net/​subtypetool/​html/​) [19]. To demonstrate possible intersubtype mosaicism, candidate sequences were analyzed using the Recombination Identification Program (RIP; version 3.5.1) available at the HIV sequence database site (http://​hiv-web.​lanl.​gov). Bootscanning analysis was performed with default parameters of window: 200bp, step: 20bp, gapsrip: on, reps: 100, Kinura (2-parameter), T/t: 2.0. Similarity plot analysis (SimPlot version 3.5.1; S. Ray, Johns Hopkins University, Baltimore, MD; http://​sray.​med.​som.​jhmi.​edu/​RaySoft/​SimPlot/​) was performed using reference A1, B, C, CRF07_BC, CRF08_BC and CRF01_AE strains. All sequences obtained in this study were submitted to GenBank with accession numbers from JQ658474 to JQ658772.

A total of 320 individuals diagnosed with HIV-1 infection between 2009 and 2010 were recruited from two counties in Dehong prefecture. Two hundred ninety nine sequences (93.4%) of HIV-1 pol region were successfully amplified and sequenced. The demographics of these individuals are shown in Table 1. The ratio of males to females was 1:0.78 and the mean age was 32.5 years (range: 2-74 years). Heterosexual transmission constituted the most frequent transmission route (75.9%, n=227), followed by intravenous drug injection (17.4%, n=52), mother-to-infant transmission (1.3%, n=4) and homosexual transmission (0.7%, n=2). The transmission route of the remaining 4.7% (n=14) individuals was unknown. The ethnicities of the studied individuals were Han (40.8%, n=122), Jingpo (32.8%, n=98), Dai (21.4%, n=64), Achang (2.3%, n=7), Deang (1.3%, n=4), Bai (1.0%, n=3), and Dongxiang (0.3%, n=1). There were also 10 Burmese individuals. In addition, 28.4% (n=85) of the individuals had CD4 cell count lower than 350 cells/μl and were ART-eligible.

HIV-1 pol gene sequences (1191 bp) were subjected to subtype determination [20]. As revealed in Table 1, 6 basic subtypes were identified. HIV-1 C subtype was found to be the most common subtype (43.1%, n=129), followed by URFs (18.4%, n=55), CRF01_AE (17.7%, n=53), B (10.7%, n=32), CRF08_BC (8.4%, n=25) and CRF07_BC (1.7%, n=5). Bootscanning analysis revealed the presence of 55 (18.4%) URFs including 48 BC recombinants (16.1%, Figure 1C and D), 6 CRF_01AE/C (2.0%, Figure 1E-I) and 1 CRF_01AE/B/C (0.3%, Figure 1J). The prevalence of URF (18.4%) was similar to that of CRF01_AE (17.7%) and revealed frequent recombination among the main strains (subtype C, CRF01_AE and subtype B). Most URFs (76.4%, 42/55) were found in individuals infected through heterosexual contact (Table 2). Seven of the HIV sequences derived from the 10 Burmese individuals were subtype C and the other 3 were URF_BC.

The presence of subtypes/CRFs varied with different transmission routes. As shown in Table 2, all 6 genotypes were found in the heterosexually transmitted population. Five genotypes, with the exception of CRF07_BC, were detected in IDUs. One subtype C, 1 URF and 2 CRF01_AE strains were detected in the 4 young children infected through mother-to-child transmission; 1 CRF01_AE strain and 1 CRF07_BC strain in 2 persons who were men having sex with men. The proportions of HIV-1 genotypes in heterosexually transmitted population and IDUs showed a statistical difference (χ²=13.801, p=0.013) suggesting that different subtype distribution patterns were associated with different infection routes. The proportion of subtype C among IDUs (61.5%) was significantly higher than that among the heterosexually transmitted population (39.2%, χ²=8.591, p=0.005). Likewise, the proportion of CRF01_AE in the heterosexually transmitted population (18.9%) was significantly higher than that among IDUs (5.8%, χ²=5.332, p=0.022). The proportions of URFs, subtype B and CRF08_BC showed no statistical difference in these two populations.

To detect TDR-associated mutations, the pol sequences were screened against those within the Stanford HIV Drug Resistance Database (http://​hivdb.​stanford.​edu). DR-related mutations were identified in 13 of the 299 viral sequences (4.3%). With the exception of one individual who had 2 NNRTI mutations, the other 12 individual harbored virus with 1 DR mutation each. The demographic characteristics of these 13 ART-naive individuals are shown in Table 3. The mean age was 32.5 years. The mean CD4 cell count was 467 cells/μl and 4 individuals were ART eligible. No statistical difference was observed in distribution of DR strains by gender, age, infection route, ethnicity, CD4 cell count or HIV-1 subtype. Eight were infected through heterosexual transmission and the other 5 through intravenous drug injection. One individual (YN10S0137) was Burmese.

The proportion of sequences with TDR mutation associated with NRTIs, NNRTIs, and PIs was 0.3% (n=1), 2.7% (n=8), and 1.3% (n=4), respectively. Sequences containing mutations associated with 2 or 3 classes of antiretroviral drugs were not found. The PI-related DR included: I85V (0.67%, n=2), M46I (0.33%, n=1) and L90M (0.33%, n= 1). Although I85V is a nonpolymorphic PI-selected mutation, it does not result in resistance to PI. Only 1 NRTI-related DR mutation, M184I (0.33%, n=1), was detected. Among NNRTI-related DR mutations, K103N and Y181C were found in 3 samples (1.0%). K103N causes high-level resistance to nevirapine (NVP) and efavirenz (EFV). Y181C causes high-level resistance to NVP and intermediate resistance to EFV, etravirine (ETR) and rilpivirine (RPV). Importantly, Y181C forms the foundation for high-level ETR and RPV resistance as an addition of some single mutations and many double mutations causes high-level resistance to these drugs. Three individuals were found to have one other NNRTI-related mutation (K103NS, K101E and G190A) each (0.33%).

Based on the interpretation of the Stanford HIV Drug Resistance Database, the 13 drug resistant viruses showed different levels of resistance to 11 antiretroviral drugs (Figure 2). Although lopinavir (LPV) and indinavir (IDV) were the main PIs used in Dehong, low-level resistance to atazanavir (ATV), fosamprenavir (FPV), IDV, nelfinavir (NFV) and saquinavir (SQV) was detected. Among RTIs, intermediate and high-level resistance to 4 NNRTIs (EFV, ETR, NVP and RPV) and high-level resistance to 2 NRTIs (lamivudine (3TC) and emtricitabine (FTC)) were detected, of which 2 NNRTIs (ETR and RPV) and 1 NRTI (FTC) were not used in local patients. The reason for their presence might be due to cross-resistance, for example 3TC induced M184I/V but also rendered resistance to FTC. Further, cross-resistance was a common phenomenon among NNRTIs. The existence of cross-resistance could increase the complexity and the length of ART.