South Africa has one of the largest HIV epidemics worldwide, with a prevalence of approximately 17%.1 The first South African cases of AIDS were identified in men who have sex with men (MSM) in 1982,2 and the HIV prevalence among primarily white MSM peaked in the mid 1980s.3 However, by the early 1990s, this MSM epidemic was rapidly giving way to a far larger heterosexual epidemic, predominantly among black Africans.4 An early molecular epidemiological study of circulating HIV strains in Cape Town in the late 1980s and early 1990s suggested that the 2 epidemics were independent of one another: the MSM epidemic was comprised almost exclusively of HIV-1 subtype B virus and was associated with travel to high-income countries.5 The heterosexual epidemic, by contrast, comprised mainly HIV-1 subtype C virus and most likely spread to South Africa through local or regional travel.5
Independent HIV epidemics in different population groups have also been identified in Thailand, and it has been postulated that the “epidemiological segregation” of subtypes may be because of demographic and social network factors that extend beyond sexual preference.6 Indeed, the importance of societal structures has also been noted in the spread of other sexually transmitted diseases in South Africa.7 In South Africa, the political segregation of apartheid, including legally enforced racial segregation of residential areas and prohibition of inter-racial sexual relationships may have limited social mixing and amplified this phenomenon. Furthermore, the different cultural backgrounds of South African MSM communities may influence how these groups identify and engage in MSM activity.8
The current emphasis of HIV prevention and treatment policies and research in South Africa are overwhelmingly focused on heterosexual and vertical HIV transmission. As a result, research among MSM and HIV prevention policies for this group have been largely neglected. Yet, HIV prevalence remains high in MSM groups in South Africa.9 While the impact of MSM transmission of HIV on the South African epidemic may be minor, these transmission patterns may contribute disproportionally to incident HIV infections among males generally.10
Little is known about the HIV-1 subtypes currently circulating among MSM in South Africa, or how the changing sociopolitical landscape over the past 2 decades may have impacted this epidemic. Yet, HIV subtypes may have implications for HIV testing, transmission, natural history of the disease, treatment and prevention, particularly HIV vaccine designs.11 We aimed to describe the circulating HIV-1 subtypes among MSM in Cape Town and explore possible differences and links between both the generalized epidemic and historically geographic and racially defined communities.
Self-identified MSM were recruited from urban and periurban sites in Cape Town in 2009 and 2010, over an 8-month period. Historically, Cape Town was divided into geographical racially defined areas.12 For the purposes of this study, urban areas were defined as Cape Town's Central Business District and surrounding communities that were historically designated as white communities. Periurban areas were defined as township or low-income communities that were historically designated as black African and/or colored communities. The racial categories of black African, colored, and white were used as per the terms used in the latest South African census.13 In South Africa, “colored” refers to individuals of mixed race or of Khoisan or Malay descent.14 Participants were recruited from organizations and venues frequented by MSM, through other studies and Internet advertising and through subsequent referrals from enrolled participants (quasi-snowballing technique15). Men were considered eligible for the study if they were aged 16 years or older, resided in the greater Cape Town area, had engaged in sexual activity with another man in the last 12 months and were willing to undergo an HIV test.
After enrollment, participants completed interviewer-administered questionnaires in their home language that captured demographic data and information on sexual orientation and disclosure. Information was gathered on participants' sexual history, in particular regarding sexual history with female partners, and the participant's 3 most recent sexual partners in the past 6 months. HIV testing was performed using fingerprick rapid antibody screening tests and followed the national HIV Counseling and Testing guidelines.16 Men who tested HIV positive had blood samples drawn for CD4 T-cell count testing and HIV subtyping. The study was approved by the Human Research Ethics Committee of the University of Cape Town. Participants provided written informed consent.
Viral RNA Extraction, cDNA Synthesis, and Amplification
HIV-1 subtyping was performed by DNA amplification followed by sequencing of the HIV-1 env gene. RNA was extracted from 200 to 400 μL plasma using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany). Reverse transcription of RNA to single-stranded cDNA was performed with SuperScript III Reverse Transcriptase using primer Oligo(dT)20 (Invitrogen Life Technologies, Carlsbad, CA) based on the manufacturer-recommended, previously described methods.17
To ensure that single genomes were amplified, cDNA was serially diluted to determine the exact dilution at which no more than 30% positive amplification reactions were present per dilution batch. Amplification of gp160 was then performed according to a method described elsewhere.18
For Gag p17p24 amplification and sequencing, a One-Step RT-PCR was performed (SuperScript III One-Step RT-PCR System with Platinum Taq DNA Polymerase High Fidelity, Invitrogen Life Technologies), which was adapted from a method described elsewhere.19 This consisted of 1× Reaction Mix (0.2 mM of each dNTP and 1.6 mM MgSO4), 0.2 μM of each primer DT1 5′ ATGGGTGCGAGAGCGTCAGTATT 3′ (nt 790-812 HXB2) and DT7 5′ CCCTGACATGCTGTCATCATTTCTTCT 3′ (nt 1818-1844 HXB2), and 1 μL of the SuperScript III RT/Platinum Taq Mix in a 50 μL reaction. Polymerase chain reaction (PCR) cycling parameters were 1 cycle of 55°C for 30 minutes, 1 cycle of 94°C for 2 minutes, 10 cycles of a denaturing step of 94°C for 15 seconds, an annealing step of 55°C for 30 seconds, an extension step of 68°C for 1 minute; 25 cycles of a denaturing step of 94°C for 15 seconds, an annealing step of 55°C for 30 seconds, an extension step of 68°C for 1 minute with a 5-second increase for each cycle thereafter, and followed by a final extension of 68°C for 7 minutes. The second round PCR consisted of 1× Expand High Fidelity Buffer, 1.5 mM MgCl2, 0.2 mM each dNTP, 0.3 μM of each primer DT3 5′ CATCTAGTATGGGCAAGCAGGGA 3′ (nt 886-908 HXB2) and DT6 5′ ATGCTGACAGGGCTATACATTCTTAC 3′ (nt 1609-1634 HXB2), and 2.6 units Expand High Fidelity polymerase (Roche) in a 100 μL reaction. Second round cycling parameters: 1 cycle of 94°C for 2 minutes; 10 cycles of 94°C for 15 seconds, 57°C for 30 seconds, 72°C for 1 minute; 25 cycles of 94°C for 15 seconds, 57°C for 30 seconds, 72°C for 1 minute with 5 seconds increase per cycle; 72°C for 7 minutes.
Sequencing and Sequence Analysis
PCR products were directly sequenced using BigDye Terminator chemistry as recommended by the manufacturer (Applied Biosystems, Foster City, CA). The sequences were determined using an ABI 3730xl genetic analyzer (Applied Biosystems) and edited using the Sequencher program (Gene Codes, Ann Arbor, MI).
The complete env gp160 sequences (∼2500bp) of study participants were aligned using CLUSTAL W,20 and subtypes were assigned by inferring gp160 phylogenetic relatedness of participant sequences to an HIV-1 subtype reference set from the HIV Sequence Database.21 Maximum Likelihood phylogenetic trees were generated using MEGA v5.04 software22 based on the Tamura-Nei model.23 Reliability of the tree topology was assessed by 500 bootstrap replicates. Sequences that clustered together were further investigated by reprocessing sequences in gp160 and a region in gag, p17p24 (748 bp), to exclude any possibility of contamination or sample mix-up. The integrity of cluster patterns was confirmed by this genotypic analysis of these regions in all cases except 1 (in which clustering could not be confirmed in gag).
Sequences were screened for evidence of inter-subtype recombination by the Recombination Identification Program (RIP 3.0) (http://www.hiv.lanl.gov/content/sequence/RIP/RIP.html), REGA HIV-1 Subtyping Tool-Version 2.0 (http://www.bioafrica.net/rega-genotype/html/subtypinghiv.html) and jumping profile Hidden Markov Model (jpHMM) (http://jphmm.gobics.de/).24 Recombinant viruses were analyzed for subtype assignment and breakpoint identification by SimPlot software v3.5.1 (http://sray.med.som.jhmi.edu/SCRoftware). Specifically, neighbor-joining bootscan analysis was performed with a sliding window of 200 bp incremented by 20 bp across the entire alignment that included sequences of 9 HIV-1 subtypes—A, B, C, D, F, G, H, J, and K, obtained from the Los Alamos HIV database (http://www.hiv.lanl.gov/). Phylogenetic trees were generated from portions of gp160 on either side of Simplot identified breakpoints to confirm unique HIV-1 inter-subtype recombinant viruses.
Data were analyzed using STATA 10.0 (StataCorp, College Station, TX). Bivariate analyses used Wilcoxon rank-sum, χ2 tests, and Fisher exact tests, as appropriate. Bivariate and multivariate logistic regression models were developed to examine factors associated with HIV-1 subtypes. All statistical tests were 2-sided at α = 0.05.
In this study, 194 HIV-infected MSM were enrolled, of which 93 (48%) were from urban areas.
Demographics and Sexual History
The median age of participants was 32 years [interquartile range (IQR), 26–39], and 47 (25%) participants were employed. Overall, 129 (67%) participants were black African, 47 (24%) were colored, and 17 (9%) were white men. The demographic characteristics, sexual orientation, and relationship histories are reported by race (Table 1). In total 72% of men identified as homosexual or gay, 21% identified as bisexual, and 3% as heterosexual. Black men were more likely to identify as bisexual or heterosexual compared with other races (33% compared with 6% in both colored and white men; P < 0.001). Black men were less likely to have disclosed their same-sex sexual activity to family or friends compared with colored and white MSM (P = 0.01).
Overall, 96 (49%) men reported having a female sexual partner in the past, with black African and white men more likely to have ever had a female partner compared with colored men (P < 0.001). Overall, 64% of participants were currently in a relationship: 50% with a man, 9% with both a male and a female partner, and 5% with a female partner. Black men were more likely to be in relationship with both men and women (not statistically significant). When asked about their 3 most recent male sexual partners, 31% [95% confidence interval (CI): 10% to 61%] of white participants, 66% (95% CI: 50% to 80%) of colored and 32% (95% CI: 24% to 41%) of black participants reported a partner of a different race.
HIV-1 subtypes were confirmed phylogenetically for 143 participants. There was no difference in those participants with and without subtyping data in terms of residential location (P = 0.42), race (P = 0.51), or previous or recent female sexual partner (P = 0.22 and P = 0.87, respectively). However, participants who identified as homosexual/gay were more likely to have subtyping data available (P = 0.01). Overall 116 (81%) samples were subtype C, 20 (14%) subtype B, 2 (1%) were subtypes A1, 2 (1%) F2, and 3 (3%) were unique inter-subtype recombinants AC, BC, and BF (one of each). The availability of subtyping data and the distribution of HIV subtypes among racial groups are shown in Figure 1. In total, 85% of participants with subtype B virus and 43% of those with subtype C were from urban sites.
The phylogenetic trees for the viruses in each race group are presented in Figure 2. Subtype C virus was more likely to occur in black African men [odds ratio (OR): 4.5; 95% CI: 1.7 to 12.5 compared with colored men and OR: 18.8; 95% CI: 4.5 to 78.3 compared with white men] and men who identified as bisexual/heterosexual (P = 0.01). While subtype C was not associated with a previous history of female sexual partner, it was associated with a recent female sexual partner (OR: 6.8; 95% CI: 1.4 to 33.3 “within last year” compared with “>1 year ago”). The association of demographic, sexual orientation, and sexual history with subtype C is reported in Table 2. When adjusted for location, race (P = 0.001) remained positively associated with subtype C virus, but location did not (P = 0.39). Among white men, the median age of individuals with subtype C virus was 34 years (IQR, 27–41) versus 40 years (IQR, 36–40) among men with subtype B (P = 0.77).
Some subtypes that occur at low frequencies globally were detected, including 2 subtype F2 viruses in black African men and 1 unique BF recombinant virus in a white man. The subtype F2 viruses were closely clustered suggestive of recent transmission from a common source (Fig. 3). The participant with BF virus reported recently returning from Dubai, where he had had at least 2 sexual partners. Two additional unique recombinant strains were identified, namely recombinant AC (from a black participant) and BC (from a colored participant). Both these participants reported no sexual history with a woman and reported that their last 3 sexual partners were all in South Africa. The 3 unique recombinant viruses were not recognized circulating recombinant forms.
In addition to the F2 cluster, there were a further 17 participants in 7 clusters of low diversity (range, 2–4 individuals/cluster) suggestive of recent transmission events (Fig. 3). From the available sexual network data, possible epidemiological links were identified between 15 (79%) participants occurring in clusters (ie, reported residing in or having recent sexual partners from the same communities), and no epidemiological link was found for the remaining 4 participants. While there were no differences in race within clusters, in 3 of these clusters, participants resided in different locations to 1 another (ie urban or peri-urban).
In the early 1990s, studies reported 2 independent HIV-1 epidemics in Cape Town: a predominantly subtype B epidemic among white MSM and a subtype C epidemic in the heterosexual population, mainly among black Africans.5 Although a number of studies have assessed the circulating HIV subtypes of predominantly the heterosexual epidemic since then,25,26 this study is the first dedicated study of subtypes among MSM in Cape Town in the last 2 decades. This study demonstrates that the 2 epidemics are no longer independent and reports evidence of bridging between the generalized heterosexual and concentrated MSM HIV epidemics in Cape Town.
Early MSM studies were performed among predominantly white and colored men; in contrast, this study enrolled a majority of black MSM. Despite the different study populations, subtype B virus remained more prevalent among MSM in our study compared with the generalized epidemic (14% vs. 1%–7%, respectively26,27). However, the majority of MSM in this study had subtype C virus. Subtype C was significantly associated with men from periurban locations, and more specifically with black African race. Subtype C was also strongly associated with bisexual orientation and reported sex with a female in the last year. We found that half of the self-identified MSM participants reported high levels of sex with women. In particular, black African MSM were more likely to identify as bisexual or heterosexual compared with white or colored MSM and were more likely to have a current female partner. This may be related to the finding that black African men were less likely to have disclosed their sexual preference to anyone, possibly because of high levels of stigma and discrimination within these communities.28 Together, the higher levels of bisexual activity and the association with subtype C virus among black African MSM may be evidence of bridging between the heterosexual and homosexual populations in Cape Town. This hypothesis is strengthened by reports of an increasing representation of subtype B in South Africa's heterosexual population.26 In addition, 2 of the 3 recombinants detected in this study included subtype C sequences (including 1 BC sequence), further supporting the theory of bridging between the MSM and heterosexual epidemics. This finding is in keeping with literature from Kenya where it has also been reported that MSM are predominantly infected with the same HIV subtype as the general population.29
Although predominately associated with black African MSM, an increasing prevalence of subtype C virus was noted among white MSM—no subtype C viruses were identified among this group in the 1990s,5 but subtype C accounted for 36% of HIV infections among white MSM in this study. A trend toward subtype C virus occurring in younger white MSM was also noted, although this was not statistically significant (possibly because of the small sample size in this racial group). Together with the high reported levels of inter-racial relationships, this finding may reflect a breaking down of historical social barriers in Cape Town in the recent past, with mixing across previously racially defined urban and peri-urban communities, and a bridging of epidemics across these districts.
Research from Cape Town over the past decade has reported overwhelmingly subtype C epidemics (>90%) with variable representation of subtypes B, low frequencies of A, G and F1, and occasional recombinant viruses.30,31 This study reported an unusually high diversity of HIV subtypes. These subtypes may have been introduced to Cape Town MSM networks through international tourism5 and the migration of other African MSM to South Africa, seeking the protection afforded by the country's constitution. Furthermore, although there were insufficient specimens to exclude a founder effect among subtype B viruses, the lack of structure among the subtype C component of the overall phylogenetic tree is evidence of the absence of a founder effect in this group. This is in keeping with literature that suggest multiple introductions of subtype B and C viruses into South Africa.5,30
This study has several limitations. First, because of the hard to reach nature of the MSM population, a number of recruitment strategies were used, as is commonplace when working with MSM,15 and this nonrandom recruitment strategy may limit the generalizability of our findings. Furthermore, we did not enroll a control group of ethnically diverse, heterosexually infected individuals, and comparisons with this population group were based on the published literature. The large number of linked cases identified in this study may be because of the high frequency of multiple concurrent partners among MSM.32 In this study, the evidence of possible recent transmission links may also be a result of the peer-based recruitment strategy. We were unable to confirm transmission events, as we did not collect in-depth sexual network data. Furthermore, we did not investigate the likely route of HIV acquisition. It is possible that some participants acquired HIV infection through nonsexual routes, such as blood transfusions or intravenous drug use, although intravenous drug use is reportedly low (<3%) among South African MSM.28
In conclusion, this study presents novel data on the molecular epidemiology of HIV-1 subtypes among MSM in South Africa. A substantially higher proportion of subtype C infections was found in this sample compared with previous studies, and evidence of transmission of relatively rare subtypes such as F2 and BF recombinants was also found. These data concur with behavioral evidence for the links between the heterosexual and MSM epidemics, as well as between the urban and periurban communities in Cape Town. These findings provide insights into the drivers of HIV epidemics in different population groups and may have implications for HIV prevention strategies and research, particularly the development of locally efficacious HIV vaccines. Over the next several years, vaccine products based on the subtype C constructs will be entering clinical trials. Given that MSM who report sex with women have a significant risk of subtype C infection, they should be eligible for participation in these studies.
The authors would like to acknowledge Earl Burrell for his contribution to the early design of this study and both Sheetal Manicklal and Ruwayhida Thebus for their assistance in processing study samples.
2. Ras GJ, Simson IW, Anderson R, et al.. Acquired immunodeficiency syndrome. A report of 2 South African cases. S Afr Med J. 1983;64:140–142.
3. Sher R. HIV infection in South Africa, 1982-1988–a review. S Afr Med J. 1989;76:314–318.
4. Wood R, O'Keefe EA, Maartens G. The changing pattern of transmission and clinical presentation of HIV infection in the Western Cape region of South Africa (1984-1995). South Afr J Epidemiol Infect. 1996;11:96–98.
5. van Harmelen J, Wood R, Lambrick M, et al.. An association between HIV-1 subtypes and mode of transmission in Cape Town, South Africa. AIDS. 1997;11:81–87.
6. Mastro TD, Kunanusont C, Dondero TJ, et al.. Why do HIV-1 subtypes segregate among persons with different risk behaviors in South Africa and Thailand? AIDS. 1997;11:113–116.
7. Kark S. The social pathology of syphilis in Africans. S Afr Med J. 1949;23:77–84.
8. Tucker A. Queer Visibilities: Space, Identity and Interaction in Cape Town. Hoboken, NJ: John Wiley & Sons; 2009.
9. Rispel LC, Metcalf CA, Cloete A, et al.. HIV prevalence and risk practices among men who have sex with men in two South African cities. J Acquir Immune Defic Syndr. 2011;57:69–76.
10. South African Centre for Epidemiological Modelling and Analysis (commissioned for SA National AIDS Council). The Modes of Transmission of HIV in South Africa. Pretoria, South Africa: South Africa National AIDS Council; 2009.
11. Peeters M, Sharp PM. Genetic diversity of HIV-1: the moving target. AIDS. 2000;14(suppl 3):S129–S140.
12. Group Areas Act, No.41 of. 1950. Cape Town, South Africa: Parliament of South Africa.
14. Boonzaier E, Sharp J. South African Keywords: The Uses & Abuses of Political Concepts. Cape Town, South Africa: David Philip; 1988.
15. Magnani R, Sabin K, Saidel T, et al.. Review of sampling hard-to-reach and hidden populations for HIV surveillance. AIDS. 2005;19(suppl 2):S67–S72.
17. Keele BF, Giorgi EE, Salazar-Gonzalez JF, et al.. Identification and characterization of transmitted and early founder virus envelopes in primary HIV-1 infection. Proc Natl Acad Sci U S A. 2008;105:7552–7557.
18. Salazar-Gonzalez JF, Bailes E, Pham KT, et al.. Deciphering human immunodeficiency virus type 1 transmission and early envelope diversification by single-genome amplification and sequencing. J Virol. 2008;82:3952–3970.
19. McCormack GP, Glynn JR, Crampin AC, et al.. Early evolution of the human immunodeficiency virus type 1 subtype C epidemic in rural Malawi. J Virol. 2002;76:12890–12899.
20. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994;22:4673–4680.
21. Los Alamos National Laboratory. HIV Database. 2012. Available at: http://www.hiv.lanl.gov/
. Accessed February 02, 2013.
22. Tamura K, Peterson D, Peterson N, et al.. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol. 2011;28:2731–2739.
23. Tamura K, Nei M. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evol. 1993;10:512–526.
24. Schultz AK, Zhang M, Leitner T, et al.. A jumping profile Hidden Markov Model and applications to recombination sites in HIV and HCV genomes. BMC Bioinformatics. 2006;7:265.
25. Van Harmelen JH, Van der RE, Loubser AS, et al.. A predominantly HIV type 1 subtype C-restricted epidemic in South African urban populations. AIDS Res Hum Retroviruses. 1999;15:395–398.
26. Jacobs GB, Loxton AG, Laten A, et al.. Emergence and diversity of different HIV-1 subtypes in South Africa, 2000-2001. J Med Virol. 2009;81:1852–1859.
27. Nwobegahay JM, Bessong PO, Masebe TM, et al.. Prevalence of antiretroviral drug resistance mutations and HIV-I subtypes among newly-diagnosed drug-naive persons visiting a voluntary testing and counselling centre in northeastern South Africa. J Health Popul Nutr. 2011;29:303–309.
28. Baral S, Burrell E, Scheibe A, et al.. HIV risk and associations of HIV infection among men who have sex with men in peri-urban Cape Town, South Africa. BMC Public Health. 2011;11:766.
29. Smith AD, Tapsoba P, Peshu N, et al.. Men who have sex with men and HIV/AIDS in sub-Saharan Africa. Lancet. 2009;374:416–422.
30. Romani B, Glashoff R, Engelbrecht S. Molecular and phylogenetic analysis of HIV type 1 vpr sequences of South African strains. AIDS Res Hum Retroviruses. 2009;25:357–362.
31. Jacobs GB, Laten A, van Rensburg EJ, et al.. Phylogenetic diversity and low level antiretroviral resistance mutations in HIV type 1 treatment-naive patients from Cape Town, South Africa. AIDS Res Hum Retroviruses. 2008;24:1009–1012.
32. Beyrer C, Trapence G, Motimedi F, et al.. Bisexual concurrency, bisexual partnerships, and HIV among Southern African men who have sex with men. Sex Transm Infect. 2010;86:323–327.
Keywords:© 2014 by Lippincott Williams & Wilkins
HIV-1; subtypes; men who have sex with men