Near Full-Length Identification of a Novel HIV-1 CRF01_AE/B/C Recombinant in Northern Myanmar

Cocirculation of multiple HIV-1 genotypes in the same place can create great opportunities for intersubtype recombination, which contributes significantly to the genetic diversity of HIV-1. Myanmar, one of the countries hit worst by HIV-1 in Southeast Asia, has multiple HIV-1 subtypes, including CRF01_AE, B, C, and numerous unique recombinant forms (URFs). ^{1 –4} Since 2003, the geographic hot spots for HIV-1 recombination have increasingly been found, such as Northern (e.g., Maizayang) and Central Myanmar (e.g., Mandalay) (15.3–86.1%), as well as Dehong of China, bordering Myanmar (30.9–87.9%). ^{3 –7} However, the majority of HIV-1 URFs in these regions were identified through partial genomic sequences, not by full-length or near full-length genomic sequences. The latter can provide much more genetic information for understanding HIV-1 genetic diversity.

Long-distance truck drivers (LDTDs) are under high risk of HIV-1 infection due to behaviors such as having commercial sex and/or injecting drugs frequently. ^8,9 They also have a high possibility of transmitting the virus to their spouses or commercial sexual workers (CSWs). Our previous data showed that multiple HIV-1 subtypes were circulating among Burmese LDTDs, including CRF01_AE, B, C, and some new recombinants involved with those three subtypes, based on analyses of five gene fragments. ¹⁰ To further understand the genetic characterization of these new HIV-1 recombinants in this population, we amplified and obtained the near full-length genome (NFLG) sequence of an HIV-1 recombinant previously identified from one Burmese LDTD (08mLDTD011). The NFLG analysis showed that 08mLDTD011 was a novel CRF01_AE/B/C recombinant that contained at least 14 breakpoints.

Under the support of the Australia HIV/AIDS Asia Regional Program (HAARP) and the Yunnan HAARP project, a plasma sample was collected in 2008 from a male Burmese LDTD (08mLDTD011) who was 40 years old, residing in Lasio, north of Myanmar, and self-reported to be infected with HIV-1 via heterosexual behavior. This study was approved by the Ethics Committee of the Kunming Institute of Zoology, Chinese Academy of Sciences. Written informed consent was obtained from this subject before blood collection.

Viral RNA was extracted from 200 μl plasma using the High Pure Viral RNA kit (Roche, Mannheim, Germany) and then reverse transcribed with the PrimeScript II cDNA Synthesis kit (TaKaRa Biotechnology, Dalian, China). Nested polymerase chain reactions (nested PCR) were performed to amplify NFLG through three overlapping fragments implementing High Fidelity DNA Polymerase (TransGen Biotech, Beijing, China) and primers reported previously. ¹¹ The same conditions of amplification were used in two rounds and were described as follows: 94°C for 2 min and then 10 cycles of 94°C for 10 s, 60°C for 30 s, and 68°C for 4 min, then 20 cycles of 94°C for 10 s, 55°C for 30 s, and 68°C for 4 min, followed by 68°C for 10 min. The amplicons were purified and sent to sequence directly. All the sequenced data were assembled by Seqman, an element of DNAstar software (Madison, WI), and merged into one sequence under at least 98% homology. The resulting sequence was queried against all sequences obtained in our laboratory to check potential laboratory cross-contamination. In addition, we performed HIV BLAST to search the related HIV-1 viruses with 08mLDTD011 using partial (gag, pol, and env genes) and NFLG sequences of it, and selected the top 10 hit NFLG sequences in each search.

Taking all hit sequences together and deleting repeat sequences, we obtained 19 NFLG sequences for phylogenetic analyses. The accession numbers of them are as follows: AF468970, AY082968, EF165539, KJ184177, AB097868, EF165541, AF259954, KC870030, KC870027, KC183778, KC870028, KF850149, AY008717, AY008715, AY008716, AB746345, KF250380, AB023804, and KC898978 (www.hiv.lanl.gov/content/sequence/BASIC_BLAST/basic_blast.html).

Then, a neighbor-joining (NJ) tree was constructed with MEGA 5.0 using the Kimura two-parameter model. The alignment contained HIV-1 reference sequences and additional complete genome sequences from Southeast Asia in addition to those obtained from the results of HIV BLAST. The recombinant pattern of 08mLDTD011 was determined by the Recombinant Identification Program (RIP) using all default settings except the window size of 300 bp (www.hiv.lanl.gov/content/sequence/RIP/RIP.html). Bootscanning and similarity plot analyses were also performed via SimPlot v.1.5.3 (http://sray.med.som.jhmi.edu/SCRoftware/simplot/). To further define the precise boundaries of the 08mLDTD011 strain, subregion NJ trees were constructed separately for each segment based on the estimation of informative site analysis.

The NFLG of 08mLDTD011 consisted of 8,341 base pairs (848–9,188 nt, position in HXB2), beginning from the latter part of the gag gene to the pol, env, tat, rev, vif, vpr, vpu, and nef genes and a 5′ part of the 3′ long terminal repeat (LTR). Sequence blasting and similarity analysis showed that there was no sample cross-contamination during the experiment. NFLG phylogenetic analysis showed that 08mLDTD011 clustered with CN.2010.JL.RF05.KJ184177 and MM1999.mIDU107 AB097868, two 01_AE/B/C recombinants from Jilin province in Northeast China and Myanmar, respectively, suggesting that 08mLDTD011 might be the same recombinant or a different 01_AE/B/C recombinant (Fig. 1). To determine its genotype, the sequence was submitted to RIP. The result showed a recombination form probably consisting of subtypes B, C, F2, and CRF01_AE (Fig. 2A). Because the F2 subsubtype had not been detected in this region previously, it was less likely that F2 was involved in this recombinant. To gain a more comprehensive recombinant pattern, we conducted bootscanning and similarity plot analyses with SimPlot v.1.5.3 using subtypes/subsubtypes A, B, C, D, F2, and CRF01_AE as reference sequences.

OPEN IN VIEWER

FIG. 1.

Phylogenetic tree and bootscanning plot for the near full-length sequence of the 08mLDTD011 strain isolated from Myanmar. (A) A neighbor-joining tree for 08mLDTD011 was constructed with MEGA 5.0 using the Kimura two-parameter model. The alignment contained HIV-1 reference sequences and additional complete genome sequences from the HIV database. SIV_CPZ was used as an out-group. The stability of the nodes was assessed by bootstrap analyses with 1,000 replications. Only bootstrap values >70% were shown at the corresponding nodes. (B) Bootscanning plots of 08mLDTD011 and other CRF01_AE/B/C recombinants (KJ184177, AB097868, and AB097866). Subtypes A, B, C, D, and CRF01_AE were used as reference. The parameters were used as follows: window size of 300 bp, step size of 20 bp with 100 replicates, 50% consensus.

OPEN IN VIEWER

FIG. 2.

Recombinant analyses of 08mLDTD011. (A) The Recombinant Identification Program (RIP) was performed on line in the Los Alamos HIV sequence database with defaulting settings except for a window size of 300 bp. (B) Similarity plots analysis was done by using SimPlot v.3.5.1 with subtypes/subsubtypes A, B, C, D, F2, and CRF01_AE as reference sequences. The details for the parameters are shown in Fig. 1.

Furthermore, informative site analyses were performed using sequences of subtype A, B, C, D, and CRF01_AE as parental strains. The results revealed that 08mLDTD011 was composed of subtypes B, C, and CRF01_AE, with at least 14 breakpoints corresponding to HXB2 nucleotide sequence positions 1169, 1345, 2917, 3157, 3747, 4024, 4695, 5002, 5839, 6123, 6783, 7738, 8560, and 8825, respectively (Figs. 1B, 2B, and 3). In addition, separate NJ trees were constructed for each segment extracted from 08mLDTD011 under the estimation of informative site analyses to explore their subtype origins, and it was obvious that each segment clustered well with the corresponding subtype references, respectively.

OPEN IN VIEWER

FIG. 3.

Schematic representation and subregion neighbor-joining trees of 15 mosaic segments defined by informative site analyses for 08mLDTD011. (A) The map was generated using the Map-Draw Tool available at the Los Alamos HIV sequence database. (B) The phylogenetic trees of these segments were constructed using MEGA 5.0 with all standard references downloaded from the HIV database. The stability of the nodes was assessed by bootstrap analysis with 1,000 replications; only >70% were shown in the trees.

Taken together, the map of this complex recombinant can be drawn as follows: CRF01_AE (848–1169), B (1170–1346), C (1347–2917), B (2918–3157), C (3158–3747), B (3748–4024), C (4025–4695), B (4696–5002), C (5003–5839), B (5840–6123), CRF01_AE (6124–6783), C (6784–7738), CRF01_AE (7739–8560), B (8561–8825), and CRF01_AE (8826–9188) (Fig. 3).

To determine whether this recombinant is a novel recombinant, we performed HIV BLAST using partial and NFLG sequences of 08mLDTD011 as the query. The top 10 NFLG sequences hit in each search were selected and a total of 19 NFLG sequences were obtained after deleting repeat sequences. We reconstructed the NJ tree using 08mLDTD011 and these sequences, and found that 08mLDTD011 clustered closely with two CRF01_AE/B/C recombinants (access numbers KJ184177 and AB0977868) (Fig. 1).

We further compared 08mLDTD011 with them and other NFLG sequences of CRF01_AE/B/C, B/C, and CRF01_AE/C previously reported in Southeast Asia. None of the above sequences shared the same breakpoints. As shown in Fig. 1B, 08mLDTD011 not only had a recombinant pattern different from KJ184177, but also was much more complicated than AB0977868. Another CRF01_AE/B/C recombinant, which was named mCSW503 and was reported in Myanmar, contained 13 breakpoints and was previously thought to be the most complex chimera. ³ However, 08mLDTD011, consisting of four CRF01_AE, six B segments, and five C segments, had one more breakpoint than mCSW503 (14 vs 13). Together with the BLAST results, we thought that 08mLDTD011 was a novel HIV-1 recombinant strain and one of the most complex chimeric structures that had been reported in the Myanmar–China border region.

Diverse HIV-1 subtypes and a high proportion of URFs have been detected in the Myanmar–China border region. ^{4 –7,10 –12} In this region, CRF01_AE-related and CRF01_AE-related intersubtype recombinants were the most common HIV-1 strains circulating among sexually acquired cases and IDUs, respectively. ^4,12 Our recent study indicated that both of them were also the two predominant HIV-1 subtypes circulating among Burmese LDTDs. ¹⁰ Here, we further showed a novel complex CRF01_AE/B/C recombinant from Burmese LDTDs. These results demonstrated strongly that HIV-1 recombination was occurring with high frequency among these LDTDs.

From an interview, we knew that 08mLDTD011 acquired HIV-1 through heterosexual contact. However, there was no explanation as to how this complex recombinant originated in his body, whether by frequent recombination among CRF01_AE, B, C, and even related primary recombinant strains circulating in this individual, or by direct acquisition of the recombinant or its primary recombinant strain from his high-risk partners. In our previous study, we had obtained four fragments (p17, pol, vif-env, and C2V3) of HIV-1 from this individual. After aligning them with the near-full-length sequence obtained in this study, we found that each fragment was nearly identical to the same region of the NFLG, indicating that the recombinant was less likely to be the false-positive result of coinfection with different subtypes.

Further evidence was from our previous study ¹⁰ in which we detected all suspected samples with potential dual infection and did not find dual infection in this individual of 08mLDTD011. In addition, although the origination of this complex recombinant remains unclear, we believed that the exposure to or contact with other high-risk populations (e.g., female sexual workers and IDUs) in the China–Myanmar border region might have been responsible for the generation of this complex recombinant. ¹⁰ Our previous study showed that Burmese LDTDs could fuel HIV-1 bidirectional transmission between Myanmar and China since they cross this border frequently and engage in high-risk behaviors. ¹⁰ The identification of this complex recombinant further indicated that much more attention needs to be paid to this group.

In 2003, Takebe et al. ³ reported six NFLG sequences from Myanmar. A decade later, we reported another NFLG sequence of HIV-1 obtained from a Burmese LDTD, which was a new and highly complex CRF01_AE/B/C recombinant, carrying at least 15 chimeric segments. This study not only added one NFLG sequence from Myanmar to the HIV sequence database, which can be used as reference for further studies, but also suggested that HIV-1 recombination events were ongoing in Myanmar. Therefore, continuous monitoring of the dynamic change in HIV-1 diversity in this region is urgently needed in determining how to control and prevent HIV-1 cross-border transmission. In addition, more investigations should be conducted based on near full-length genomic analyses, which will enhance our understanding of HIV-1 diversity and recombination and provide necessary sequence references for related basic research (e.g., vaccine design).

Sequence Data

The nearly full-length genomic sequence of 08mLDTD011 is available in GenBank under accession number KP455640.