Circulation of HIV-1 Multiple Complexity Recombinant Forms Among Female Sex Workers Recently Infected with HIV-1 in Thailand

Abstract

The circulating subtype distribution of HIV-1 has not been well characterized in female sex worker (FSW) populations in Thailand. To understand the mechanisms and interrelationships of epidemics involving FSWs in Thailand, we performed a large molecular epidemiological study of FSWs aged 25 years with recently acquired HIV-1 infections. The samples were collected in 2005, 2007, 2009, and 2011 in 38 provinces, representing every region of Thailand. After gag (p24), pol (pro-RT), and env (C2/V3) were sequenced, comprehensive genome analysis was performed. Genetic subtypes were determined in 159 plasma samples. The percentage of circulating recombinant forms (CRFs) CRF01_AE (90.6%) predominated, while subtype B (1.3%), other CRFs (1.9%), and unique recombinant forms (URFs) (6.2%) were identified as minor populations. Interestingly, the unique recombinant nature of these HIV-1 strains was verified in 10 specimens, indicating the presence of new forms of HIV-1 intersubtypes G/A, C/B, AE/B/C, and AE/B with different recombination breakpoints. Subtype B has contributed to these new generations of unique CRF01/B recombinants, especially in the pol (RT) gene, in which the template switching of the RT genomes occurred during reverse transcription. These results imply that the several unique recombinant viruses circulating in Thailand were probably generated in the population or introduced from neighboring countries. Our study helps clarify the patterns of viral transmission and define transmission pathways in Thailand.

In Thailand, an estimated 440,000 people were living with HIV in 2013, and an estimated 8,000 people were being newly infected with HIV per year.¹ HIV infection remains a major public health concern in Thailand. The major risk populations are injected drug users (IDUs), men who have sex with men (MSMs), and female sex workers (FSWs). The overall prevalence of HIV among FSWs in Bangkok, Thailand, was 20.2% in 2007.² Notably, this percentage is substantially higher than the 11.8% (95% confidence interval [CI]: 11.6–12.0) overall prevalence of HIV reported in 50 low- and middle-income countries during 2007–2011.³ In addition, the prevalence of HIV in non–venue-based FSWs (i.e., FSWs who contact clients through phone networks, on the streets, or in other public spaces) was 22.7% in Bangkok. This percentage is almost 10 times higher than that reported for venue-based FSWs who participate in routine sentinel surveillance (2.5%).² Hence, improving HIV/AIDS prevention in FSWs is a major public health objective in Thailand.

Since 1989, HIV-1 subtype B has been the predominant subtype among IDUs, whereas CRF01_AE has been the main clade among individuals with heterosexual exposure risk.⁴ The HIV-1 epidemic in Thailand has changed, with CRF01_AE now accounting for the majority of HIV-1 infections in high-risk groups, that is, MSMs and IDUs.^5,6

However, there is no recent information about the distribution of HIV-1 subtypes/CRFs in the FSW population. Therefore, an understanding of the distribution of HIV-1 clarifies the dynamics of HIV-1 epidemics among the FSW populations in Thailand.

We determined the nucleotide sequences of the gag, pol, and env regions from plasma viral RNA. All samples were characterized by targeted genomic sequencing (Table 1). The nested polymerase chain reaction (PCR) products of the gag (p24), pol (pol-RT), and env (C2V3) regions were 671, 1,277, and 565 bp long, respectively. The samples showing possible recombinant or non-AE were further amplified and sequenced with the gp41-nef primer set. The nested PCR products were 800 bp long. The nucleotide sequences were aligned with the HIV-1 group M (subtypes A–J) reference strains, and selected references were found previously in Asia: HIV-1 subtype B-US (B) (accession numbers: AY835778 and K02007), subtype B-Thai (B′) (AY945710 and DQ354116), subtype C (AF067158 and AF361875), and CRF01_AE (JX446848, JX446654, AY713425, U51189, and AY008714) strains. Phylogenetic trees were constructed by the neighbor-joining method with 100 bootstrap replicates. These analyses were conducted using MEGA version 6.0 software.

Table 1.

Primers Used for Gene Amplification

Gene	Primers ^a	Position in HXB2	Primer sequence (5′-3′)	References
gag	gag 848_FO^b	848-869	GGTTAAGGCCAGGGGGAAAGAA
	gag-5_RO	2377-2401	TTCCYCCTATCATTTTTGGTTTCC	Li et al. ¹²
	gag1173_FI	1173-1196	CAGCCAAAATTACCCTATAGTGCA
	gag1844_RI	1817-1844	CCCTGRCATGCTGTCATCATTTCTTCTA
pol	PRO5_FO	2074-2095	AGACAGGYTAATTTTTTAGGGA	Saeng-aroon et al. (2011)¹⁷
	DRRT34_RO	3504-3529	GCTATTAAGTCTTTTGATGGGTCATA
	PRO1M_FI	2148-2167	AGAGCCAACAGCCCCACCAG
	DRRT4L_RI	3402-3425	TACTTCTGTTAGTGCTTTGGTTCC
env	Env6556_FO^b	6556-6581	ATGGGATCAAAGCCTAAAGCCATGTG
	Env7811_RO^b	7811-7781	AGTGCTTCCTGCTGCTCCCAAG
	Env6816_FI^b	6816-6844	CCTCAGCCATTACACAGGCCTGTCCAAAG
	Env7381_RI^b	7381-7360	TTACAGTAGAAAAATTCCCCTC
gp41-nef	Gp41/nef_FO	8520-8544	TTCAGCTACCACCGCTTGAGAGACT	Kijak et al. (2007)¹⁸
	gp41/nef_RO	61-82	CARTCAGGGAAGWAGCCTTGTG
	Nef_FI	8797-8815	ATGGGDRGCAARTGGTCAA
	Nef_RI	9023-9047	GCTCCCTTRWAAGTCATTGGTCTTA

Primers are assigned a primer name_primer direction (forward, F, or reverse, R), followed by a nested outward (O) or inward (I) primer.

Provided by Dr. Masashi Tatsumi.

All samples were collected from Thailand's National HIV Drug Resistance Surveillance Program. The plasma samples of venue-based FSWs were then sent from 38 of 77 provinces, representing every region of Thailand. The FSWs had neither a previous pregnancy nor an antiretroviral treatment history. The enrolled HIV-infected FSWs are shown in Table 2. In brief, the samples were screened and confirmed for the presence of HIV-1 antibodies. The World Health Organization (WHO) recommends collecting 45–65 samples for an HIV drug resistance surveillance program.⁷ When the number of enrolled samples using the BED IgG-Capture Enzyme Immunoassay (BED-CEIA) was insufficient, we used the criterion of 25 years of age at the time of enrollment.

Table 2.

Distribution of HIV-1 Subtypes in Female Sex Workers in Thailand, 2005–2011

					Pure forms			CRF		URF
Sampling year	Number of samples	HIV-positive cases, n (%)	HIV+, age <25 years, WHO recommended	Complete analyzing of sequences of three region genes	CRF01_AE, n (%)	Subtype B (US)	Subtype B (Thai)	CRF15_01B	CRF54_01B	01/B	G/A	C/B	01/B/C
2005	7,201	388 (5.3)	52	39	34 (87.2)		1	2		1	1
2007	9,377	423 (4.5)	51	44	39 (88.6)	1				3		1
2009	9,081	335 (3.7)	34	26	26 (100)
2011	8,654	262 (3.0)	64	50	45 (90)				1	2			2
Total	34,313	1,408 (4.1)	201	159	144 (90.6)	1	1	2	1	6	1	1	2

CRF, circulating recombinant form; WHO, World Health Organization.

To identify the recombination breakpoints, we used SimPlot 2.5 software, recombination detection and analysis (RDP2), and http://jphmm.gobics.de (jpHMM) to calculate the bootstrap plots. For these plots, bootscanning was performed on neighbor-joining trees for a window of 200 bp moving along the alignment in 10-bp increments, thus evaluating 100 replicates for each phylogeny. Since the results obtained by SimPlot and RDP2 were identical, only the plots generated by SimPlot software are presented. Finally, we mapped and visually inspected three fragment genes to compare them with the HIV-1 references.

As Table 1 shows, of the 34,313 FSWs enrolled in 2005, 2007, 2009, and 2011, 1,408 FSWs were positive for HIV-1 infection. The prevalence of HIV-1–positive FSWs was 4.1% (95% CI: 3.9–4.3). We successfully amplified and sequenced gag, pol, and env genes from 159 samples. Of these subjects, the mean age was 21.2 years (range, 16–25 years). The genetic subtype for each of the regions gag, pol, and env was determined in the 159 specimens (Table 1). The neighbor-joining analysis placed 90.6% (144/159) of the sequences in HIV-1 CRF01_AE with high bootstrap values (more than 90/100) for the three gene regions (Fig. 1). The prevalence of CRF01_AE in 2005, 2007, 2009, and 2011 was 87.2%, 88.6%, 100%, and 90%, respectively.

FIG. 1.

The distribution of HIV-1 subtypes circulating among female sex workers in Thailand. Phylogenetic trees based on (A) the gag, (B) the pol, and (C) the env nucleotide sequences. Subtypes are indicated on the right side of the tree. Bootstrap values >90 are shown at corresponding nodes. CRF01_AE in the present study is shown in the blue circle. The terms in red indicate representative references of subtype B and subtype C, which are parent subtypes for recombinant viruses as progeny. The other subtypes are the magnitude in the rectangle. Non-CRF01_AE is indicated by purple highlighting. The specimens with discrepant subtypes in three genes are highlighted by black. Color images available online at www.liebertpub.com/aid

In the gag trees, most of the sequences belonged to the monophyletic cluster of HIV-1 CRF01_AE, distinct from the nonrecombinant form of subtype A (Fig. 1A). There were eight gag sequences located in different clusters in the phylogenetic tree (Fig. 1A). The phylogenetic analysis of the pol region showed several clusters, and most of the sequences were located in pol CRF01_AE (Fig. 1B). Interestingly, many putative recombination patterns between CRF01_AE and B were detected in this pol tree. We found breakpoints of the pol gene in several specimens (Fig. 2), and most of them were recombinant between subtype B and CRF01_AE. Most of the env sequences formed a single cluster within env CRF01_AE, except for eight specimens that were placed in env subtypes B-US, B′, C, and A (Fig. 1C).

FIG. 2.

Genome structures of 15 samples. The nucleotide sequence determination in relation to the reading frames of corresponding HIV-1 genes. Each deduced subtype structure is illustrated based on the results of the phylogenetic tree analysis. Regions in color could be assigned to known subtypes. Orange, subtype A; blue, subtype B; yellow, subtype C; red, CRF01_AE; and purple, subtype G. The region in white was not PCR amplified. PCR, polymerase chain reaction. Color images available online at www.liebertpub.com/aid

The subtype classification and distribution of 15 samples were obtained with the use of the jpHMM program based on four genes (Fig. 2). The genetic subtype of the three gene regions was determined to be CRF01_AE in 90.6% (144/159). Of the 159 samples, 15 showed a non-AE result, and we further sequenced the gp41-nef regions of these 15 samples. There were one subtype B-US, one B′, two CRF15_01B, and one CRF54_01B. The other 10 samples showed discrepancies from all known CRFs and subtypes, indicating that we found URFs in this study. No. 117505 and no. 110098 showed complicated genome patterns, each composed of three subtypes: CRF01_AE, B, and C (Fig. 2). No. 051171 and no. 079004 were similar to, but distinct from, CRF02_AG and subtype C, respectively.

We used the recombinant identification programs SimPlot, RDP2, and jpHMM to determine whether the apparent discrepancy of subtypes in the gag, pol, and env regions in the specimens was due to coinfections or a recombination of multiple HIV-1 subtypes (Figs. 3 and 4). Since these programs gave similar results, only the plots created by the SimPlot program are presented (Fig. 3). We identified 10 putative recombinant forms between HIV-1 subtype B and CRF01_AE (Fig. 2). However, the locations of the recombination breakpoints in these specimens were not identical (Fig. 3A–H). Multiple recombination breakpoints between HIV-1 subtype B and subtype G were identified in the pol region of no. 051171 (Fig. 3I).

FIG. 3.

Recombination breakpoint analysis in the 1,277-bp region encompassing RT of pol region. The specimens (A–I) with discrepant subtypes in three genes were selected for this analysis. Bootscanning plots depict the relationships to reference sequences of HIV-1 subtypes A through K and RL42, which belong to the HIV-1 subtype B′ cluster within subtype B. At the top of each panel, the strategy of nucleotide sequence determination with the reading frame of the corresponding HIV-1 gene is illustrated. The bootscanning was plotted for a window of 200 bp moving in increments of 10 bp along the alignment. The regions for subtype B′, subtype C, CRF01_AE, and subtype G are identified by high bootstrap values (>90%). Each point of crossover of two curves indicates a recombination breakpoint. The black triangles mark the locations, where the sequences were broken to construct phylogenetic trees. The phylogenetic trees were computed by the neighbor-joining method with parsimony bootstrap values at the nodes containing the specimen of interest. The deduced subtype structure is illustrated above the bootscanning. Regions in white could not be assigned to any known subtype. Color images available online at www.liebertpub.com/aid

FIG. 4.

Recombination breakpoint analysis in the 674-bp (A and B) and 277-bp (C) regions from the gp41 region to the nef region for the 15 specimens showing possible recombinants. The analysis was performed the same as in Figure 3. The bootscanning was plotted for a window of 200 bp moving in increments of 10 bp along the alignment. The deduced subtype structure is illustrated above the bootscanning. Color images available online at www.liebertpub.com/aid

These breakpoints in the pol region of no. 051171 belonged to different genome patterns of CRF02_AG (Fig. 3I). No. 051171_2005 from Chachoengsao Province revealed a new unique recombinant exhibiting subtype G gag (not subtype A gag) and an independent cluster and subtype G pol. Similarly, in no. 072757_2007, multiple recombination breakpoints between HIV-1 subtype B and CRF01_AE were identified in the pol region (Fig. 3B). No. 072757_2007 carried gag subtype B, a pol-independent cluster, and subtype B (Fig. 3B). Recombination breakpoints were not identified in the pol regions of the remaining two specimens, no. 054100_2005 and no. 057107_2005. However, their genome patterns were similar to those of CRF15_01B (Fig. 2). These specimens will be further analyzed to clarify whether the genome patterns of the specimens are due to the recombinant form or coinfection.

To discriminate the genomic patterns of the specimens, we also analyzed the gp41-nef region using SimPlot. Two of the 10 subtype B/CRF01_AE recombinants, no. 074293_2007 and no. 116157_2011, were found to contain the breakpoints of the gp41-nef region (Fig. 4A, B). No. 079004_2007 from Tak Province was initially classified as HIV-1 subtype C (gag C/pol C/env C). However, a breakpoint was identified in the gp41-nef region and was found to be a new and unique recombinant form (URF) between subtype C and subtype B (Fig. 4C). No. 117505_2011 from Pang-Nga Province revealed three different subtypes in each fragment gene (gag CRF01_AE/pol B/env C) (Fig. 2).

This is the first large-scale molecular epidemiological study of the FSW population in Thailand to be conducted from samples collected between 2005 and 2011. Our data were derived from a large longitudinal cohort study, which is a part of the National HIV Drug Resistance Surveillance Program. Our results provide evidence of patterns of viral transmission from region to region and a clue toward defining the transmission pathways between groups and individuals based on sequence data.

Our study revealed that the predominant HIV-1 subtypes are CRF01_AE in 90.6%, B in 1.3%, CRF01_AE/B recombinants in 5.7%, and other URFs in 2.5%, as depicted in Figure 1. This distribution is mostly consistent with the results of the Bangkok MSM Cohort Study,⁵ which reported 82% CRF01_AE, 5% B, and 13% CRF01_AE/B recombinants. In Thailand, CRF01_AE and CRF01_AE/B recombinants are the major circulating strains among FSWs and MSMs. The molecular epidemiology of HIV-1 among IDUs in northern Thailand indicated that CRF01_AE (86.8%) and CRF01_AE/B′ (13.2%) recombinants are dominant in this risk group.⁶ An increasing number of infections of CRF01_AE/B′ recombinants were reported among all risk groups in Thailand.^6,8,9 Furthermore, the distribution has shifted from B to CRF01_AE among IDUs. Other studies reported similar shifts in China and Japan.^10

–13 The reasons for the changes in distribution are unclear, but the shifts could be related to different behaviors between these high-risk groups. Our present phylogenetic analyses of the gag, pol, and env regions showed that URFs similar to HIV-1 subtype C sequences from the western area of Thailand tended to be phylogenetically more closely related to those found in India, whereas they were weakly related to the HIV-1 subtype C distributed in African countries (Fig. 1). Subtype C was identified in Tak Province, near the border with Myanmar. This sample is so closely related to subtype C identified in India that it could be linked with Myanmar or India.

A limitation of the present study was that, because of the cost of analysis, we analyzed only parts of the full-length genome and used fewer clinical samples than would have been desirable. The four fragment genes might have failed to detect new recombinant form strains or URFs. Our results indicate that subtype B has not become extinct but rather has contributed to these new generations of CRF01_AE/B recombinants, especially in the pol (RT) gene. The recombination points of many URFs were identified in the pol (RT) region. One possibility among the many recombinations in the RT gene is the template switching of the RT enzyme between the two RNA genomes during reverse transcription.¹⁴ Since FSWs have many opportunities to become exposed to genetically different HIV-1 strains, many URFs might be easily generated between CRF01_AE/B recombinants and pure CRF01_AE, which are the two major dominants in Thailand.

Our analyses of intersubtype mosaicism detected possible recombination breakpoints in at least 12 specimens. Of these, 10 specimens showed intersubtype recombination between HIV-1 CRF01_AE and subtype B. One specimen appeared to have breakpoints between subtype A and subtype G. The mosaic pattern of the latter specimen was not consistent with CRF02_AG. A new recombination virus may emerge among high-risk populations exhibiting the highest HIV infection rates in the region. Two specimens appeared to be these recombinants: CRF15_01B and CRF54_01B. Actually, many CRF01_AE/B′ recombinants have been detected in Southeast Asian countries.^15,16 The present findings indicate that our study should be continued to detect the emergence of more CRFs and URFs.

The unique recombinant nature of these HIV-1 strains was verified in 10 specimens, indicating the presence of new forms of HIV-1 intersubtypes G/A, C/B, AE/B/C, and AE/B with different recombination breakpoints. These data suggest that multiple subtypes of B, C, and CRF01_AE are cocirculating in Thailand, leading to the evolution of new forms of recombinants among the risk populations that exhibit high HIV infection rates there. Thailand functions as a crucial intersection, connecting the countries in Southeast Asia as well as international traffic with southern China (Yunnan Province) and Myanmar. Reflecting Thailand's geography, the results of the present study revealed the unique distribution and occurrence of HIV-1 subtypes and their intersubtype recombinants.

In conclusion, the cocirculations of multiple subtypes of HIV-1 among persons engaging in high-risk behaviors in Thailand provide unique opportunities for coinfection of different HIV-1 subtypes and the generation of various forms of HIV-1 intersubtype recombinants. More structural information on the HIV-1 strains prevalent in these geographically intermingled regions is crucial to elucidate the dynamics of how HIV spreads and develop vaccine strategies in this unique area in Asia. To decrease the incidence of HIV, we need to continue the surveillance of HIV molecular epidemiology every 3–5 years. Thailand has intensified HIV prevention efforts among MSMs (including TG), MSWs, and FSWs as well as their clients. This includes expanded outreach to target populations, improvements in service linkages, and increased referrals for prevention and treatment as needed.

Ethical Statement

Questionnaire information was deidentified by the reviewer and recoded so that no information could be linked to any individual participant. The protocol of this study was reviewed and approved by the Ethics Committee of the Institute for the Development of Human Research Protections (IHRP) in Thailand. The Ethics Committee waived the need for consent.

Footnotes

Acknowledgments

We acknowledge the support of and funding from the Japan Initiative for Global Research Network on Infectious Diseases (J-GRID). We thank all the following individuals: the FSWs who participated in the study; the staff of the Bureau of Epidemiology; the 38 Provincial Public Health Offices (for collecting the samples); Dr. Achara Teeraratkul and Dr. Kunjanakorn Phokhasawad of the Thai MOPH–U.S. CDC Collaboration (for their supportive and helpful advice); Dr. Masashi Tatsumi (for the primer information); Mr. Nattapong Kamklai, Ms. Ratrawee Boonmuang, and Dr. Piraporn Utchee (for their laboratory support); and Ms. Dorothy L. Southern for providing guidance on quantitative article writing and critical reviews of the article.

Author Disclosure Statement

No competing financial interests exist.

Sequence Data

Partial sequences of the HIV CRF01_AE gag, pol, and env genes analyzed in this study have been deposited in GenBank under accession numbers LC114516–LC114978. The subtypes were further confirmed by phylogenetic analysis using the reference sequences representing subtypes A–D, F–H, J, K, CRF01_AE, CRF15_01B, CRF33_01B, and CRF34_01B (www.hiv.lanl.gov/).

References

UN Joint Programme on HIV/AIDS (UNAIDS), The Gap Report, 2014, available at: www.refworld.org/docid/53f1e1604.html, accessed 10 April 2016 .

Manopaiboon

, et al.: Unexpectedly high HIV prevalence among female sex workers in Bangkok, Thailand in a respondent-driven sampling survey. Int J STD AIDS, 2013; 24:34–38.

Baral

, et al.: Burden of HIV among female sex workers in low-income and middle-income countries: A systematic review and meta-analysis. Lancet Infect Dis, 2012; 12:538–549.

, et al.: Independent introduction of two major HIV-1 genotypes into distinct high-risk populations in Thailand. Lancet, 1993; 341:1171–1174.

Leelawiwat

, et al.: Increasing HIV-1 molecular complexity among men who have sex with men in Bangkok. AIDS Res Hum Retroviruses, 2015; 31:393–400.

Tovanabutra

, et al.: The changing molecular epidemiology of HIV type 1 among northern Thai drug users, 1999 to 2002. AIDS Res Hum Retroviruses, 2004; 20:465–475.

Bennett

, et al.: Recommendations for surveillance of transmitted HIV drug resistance in countries scaling up antiretroviral treatment. Antivir Ther, 2008; 13 Suppl 2:25–36.

Ramos

, et al.: New HIV type 1 CRF01_AE/B recombinants displaying unique distribution of breakpoints from incident infections among injecting drug users in Thailand. AIDS Res Hum Retroviruses, 2003; 19:667–674.

Arroyo

, et al.: HIV type 1 molecular epidemiology among high-risk clients attending the Thai Red Cross Anonymous Clinic in Bangkok, Thailand. AIDS Res Hum Retroviruses, 2010; 26:5–12.

10.

Wang

, et al.: Identification of subtype B, multiple circulating recombinant forms and unique recombinants of HIV type 1 in an MSM cohort in China. AIDS Res Hum Retroviruses, 2008; 24:1245–1254.

11.

Wang

, et al.: The dynamic face of HIV-1 subtypes among men who have sex with men in Beijing, China. Curr HIV Res, 2011; 9:136–139.

12.

, et al.: High genetic diversity of HIV-1 was found in men who have sex with men in Shijiazhuang, China. Infect Genet Evol, 2011; 11:1487–1492.

13.

Kondo

, et al.: Emergence in Japan of an HIV-1 variant associated with transmission among men who have sex with men (MSM) in China: First indication of the International Dissemination of the Chinese MSM lineage. J Virol, 2013; 87:5351–5361.

14.

Chen

, Powell

, Hu

: High frequency of genetic recombination is a common feature of primate lentivirus replication. J Virol, 2006; 80:9651–9658.

15.

, et al.: Identification of a novel second-generation circulating recombinant form (CRF48_01B) in Malaysia: A descendant of the previously identified CRF33_01B. J Acquir Immune Defic Syndr, 2010; 54:129–136.

16.

, et al.: Identification of new CRF51_01B in Singapore using full genome analysis of three HIV type 1 isolates. AIDS Res Hum Retroviruses, 2012; 28:527–530.

17.

Saeng-aroon

, et al.: Drug-resistant mutation patterns in CRF01_AE cases that failed d4T+3TC+nevirapine fixed-dosed, combination treatment: Follow-up study from the Lampang cohort. Antiviral Res, 2010; 87; 22–29.

18.

Kijak

, et al.: Distinguishing molecular forms of HIV-1 in Asia with a high-throughput, fluorescent genotyping assay, MHAbce v.2. Virology, 2007; 358:178–191.