Acquisition of human immunodeficiency virus type 1 (HIV-1) infection is mediated by a combination of characteristics of the infectious and the susceptible member of a transmission pair, including human behavioral and genetic factors, as well as viral fitness and tropism. Here we report on the impact of established and potential new HLA class I determinants of heterosexual HIV-1 acquisition in the HIV-1-exposed seronegative (HESN) partners of serodiscordant Zambian couples.
We assessed the relationships of behavioral and clinically documented risk factors, index partner viral load, and host genetic markers to HIV-1 transmission among 568 cohabiting couples followed for at least nine months. We genotyped subjects for three classical HLA class I genes known to influence immune control of HIV-1 infection. From 1995 to December 2006, 240 HESNs seroconverted and 328 remained seronegative. In Cox proportional hazards models, HLA-A*68:02 and the B*42-C*17 haplotype in HESN partners were significantly and independently associated with faster HIV-1 acquisition (relative hazards = 1.57 and 1.55; p = 0.007 and 0.013, respectively) after controlling for other previously established contributing factors in the index partner (viral load and specific class I alleles), in the HESN partner (age, gender), or in the couple (behavioral and clinical risk score). Few if any previously implicated class I markers were associated here with the rate of acquiring infection.
A few HLA class I markers showed modest effects on acquisition of HIV-1 subtype C infection in HESN partners of discordant Zambian couples. However, the striking disparity between those few markers and the more numerous, different markers found to determine HIV-1 disease course makes it highly unlikely that, whatever the influence of class I variation on the rate of infection, the mechanism mediating that phenomenon is identical to that involved in disease control.
Citation: Song W, He D, Brill I, Malhotra R, Mulenga J, Allen S, et al. (2011) Disparate Associations of HLA Class I Markers with HIV-1 Acquisition and Control of Viremia in an African Population. PLoS ONE 6(8): e23469. doi:10.1371/journal.pone.0023469
Editor: Johan K. Sandberg, Karolinska Institutet, Sweden
Received: March 9, 2011; Accepted: July 18, 2011; Published: August 17, 2011
Copyright: © 2011 Song et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the National Institute of Allergy and Infectious Diseases (NIAID) through grants AI040951 (to S.A.), AI071906 (to R.A.K.), and AI064060 (to E.H.). J.T. is the recipient of an Independent Scientist Award (AI076123) from NIAID. The funding agency had no role in study design, data collection, data analysis, decision to publish, or preparation of this manuscript.
Competing interests: The authors have declared that no competing interests exist.
In sub-Saharan Africa, heterosexual exposure accounts for much of the spread of human immunodeficiency virus type 1 (HIV-1) infection, especially among HIV-1 discordant couples . Epidemiological evidence suggests that acquisition of HIV-1 infection is often mediated by risk behaviors (exposure), viral characteristics (subtypes and co-receptor tropism), coinfection with other pathogens , and host genetic variations that mediate innate and adaptive immune responses , . Longitudinal studies in Lusaka, Zambia have attested to the role of several factors in the index partner as determinants of HIV-1 transmission within HIV-1 discordant couples, namely HIV-1 viral load (VL) , , genital ulcer or inflammation (GUI) , , and certain polymorphisms in HLA class I genes (HLA A*36, B*57 and C*18) . For the susceptible HIV-1 exposed seronegative (HESN) heterosexual partners, major involvement of human genetic variants other than those in CCR5 receptor/ligand system have been less firmly established , . Few investigations have included relatively large numbers of paired index and susceptible partners, followed them for long enough, and included sufficient detail to allow persuasive tests of immunogenetic hypotheses.
Human leukocyte antigen (HLA) class I genes in the major histocompatibility complex (MHC) are important determinants of effective immune surveillance. Their allelic variants have been associated with various outcomes in the natural course of HIV-1 infection, including viremia and disease progression (time to manifest immunodeficiency after infection) , , , , , , . Favorable HLA alleles like HLA-B*57 and B*27 have strong and durable impact on both early and late outcomes including set-point VL , , , , , , , , and they appear to reduce or delay viral transmission by suppressing viremia in an infected potential transmitter (e.g., a sexual partner) ,. In contrast, evidence that HLA variants influence acquisition in HESNs is less convincing; associations reported for various class I alleles (A*23, A*68:02, A*74 and B*18) have been less consistent in studies of mother-infant pairs , , commercial sex workers , ,  and other high-risk groups . Occasional detection of HIV-1-specific cytotoxic T-lymphocytes (CTLs) in genital mucosa of HESNs  has been taken to imply a role for HLA class I alleles in preventing viruses from disseminating and inducing systemic antibody responses, but multiple studies have not shown enough consistency to establish unequivocally the involvement of HLA class I polymorphisms in variable susceptibility to HIV infection . Here in a large cohort of serodiscordant Zambian couples we further document influences of HLA class I alleles on the rate of HIV-1 acquisition that are disparate from those that control viremia.
Overall characteristics of HIV-1 discordant Zambian couples in this study
We analyzed 568 HIV-1 serodiscordant couples with complete HLA class I genotyping as well as adequate follow-up between 1995 and 2006 (Figure 1). These couples included 240 who seroconverted (SCs) with viruses (predominantly subtype C) closely linked to those found in their index partners and 328 susceptibles who were persistently HESN (pHESN) during quarterly follow-up visits. The pHESNs differed from SCs in sex ratio (p = 0.0002), age (p = 0.014 for men and p = 0.004 for women), time of follow-up from enrollment (p<0.0001), and behavioral and clinical risk scores (p<0.0001) (Table 1). Mean VL was higher in male than in female index partners (p<0.0001). Other characteristics, including enrollment date, cohabitation time and age difference between the man and the woman within a partnership, were highly comparable between SCs and pHESNs (data not shown). Therefore, VL of index partner, age and gender of HESN, and risk score were included as covariates in primary analyses of HIV-1 acquisition.
In all, 568 couples (240 transmission pairs and 328 non-transmission pairs) were eligible (see Table 1). Non-genetic factors included age of HESN, sex, direction of transmission (M to F or F to M), genital ulcer/inflammation (GUI), and viral load (VL).
Selective testing of HLA markers with previously reported and newly detected associations
We first examined all HLA class I alleles or haplotypes implicated in HIV-1 infection or disease control in earlier studies of associations in Africans by both Cox proportional hazards and logistic regression analysis (Table 2, Group I). Two markers, A*68:02 and one or both of the virtually inseparable alleles in the B*42-C*17 haplotype, were associated with increased likelihood or rate of acquisition (p<0.05). Only the A*68:02 association withstood analysis for false discovery rate (FDR, q = 0.025 for Cox proportional hazards model and q = 0.050 for logistic regression model). The unfavorable effects of these markers here contrast starkly with their apparent protective roles previously reported: for A*68:02 against infection in another African population  and for the B*42-C*17 haplotype with lower viremia in both our and other cohorts , .
Three additional markers involving HLA-B that have occasionally been implicated in studies of HIV-1 disease control in other populations but not in Zambians showed associations with acquisition of infection here with nominal p≤0.05: B*44 and either B*51 or the B*51-C*16 haplotype with lower risk (ORs = 0.52 and 0.37 or 0.33, respectively) and B*14 with higher risk (OR = 1.79) (Table 2, Group II). However, the statistical significance of those three markers exceeded the FDR threshold (all q≥0.2). No other variant previously associated with control of viremia  or transmission from the index partner  in Zambians and no other variant associated with acquisition in other cohorts (e.g., A*23 and B*18) ,  appeared to influence the acquisition of HIV-1 infection in Zambian couples.
No other HLA class I allele that has been reproducibly found to influence HIV-1 infection or disease progression showed a corresponding association with HIV-1 acquisition in the full cohort of discordant Zambian couples (Table 2, Group III). Only in the subset of couples with female HESNs was there a suggestion that the African allele A*74:01, recently recognized in association with both lower VL and decreased likelihood of HIV-1 infection , , , , , was associated with slower rate of acquisition (adjusted univariate RH = 0.49, p = 0.04) (data not shown). The significance of this association was above the FDR threshold (q = 0.57). Although each comparison was adjusted for age and gender, this adjustment did not alter any of the associations.
Ten HLA class I supertypes were also analyzed in both Cox proportional hazards and logistic regression models. None of the 10 showed statistically significant effects on time to infection or on overall occurrence of infection by the end of study period (Table S1). A specifically designated portion of the A2 supertype, A02/A6802, was previously reported to retard acquisition ; however, here it was associated with accelerated acquisition (RH = 1.43, p = 0.013 and q = 0.165), and this effect was due entirely to the contribution of A*68:02, the dominant allele in this supertype (85 of 137 with the supertype).
HLA-A*68:02 and B*42-C*17 as independent correlates of HIV-1 acquisition
All alleles and haplotypes in Table 2 were further analyzed in multivariable Cox proportional hazards models with stepwise elimination. These models dismissed several probable or suspected HLA factors, including (i) B*57, B*81, and C*18 in index partners, (ii) B*14, B*44, B*81, B*51, and C*18 in HESNs (adjusted p≥0.05 for all). In the end, only A* 68:02 and B*42-C*17 in HESNs and A*36 in index partners retained statistical significance (adjusted relative hazards, RH = 1.76, 1.46 and 2.06; p = 0.0004, 0.025 and <0.0001, respectively) in a reduced model for comparing time to HIV-1 acquisition (Table 3, model I). The associations of all three of these HLA factors were independent of the behavioral, clinical and biological factors documented for the study population. The impact of these genetic markers persisted in a model including VL in index partners (Table 3, model II) as well as a model including both VL in index partners and risk score in couples (RH = 1.57, 1.55 and 1.78; p = 0.007, 0.013 and 0.001) (Table 3, model III). Of the 11 initial HESNs who had both A*68:02 and B*42-C*17, five seroconverted at a rate similar to that seen with B*42-C*17 alone. The infection-free time among initial HESNs bearing A*68:02 (median = 33 months, 95% confidence interval, CI = 21–45 months) and those bearing B*42-C*17 (median = 39 months, 95% CI = 27–51 months) contrasted with that among other initial HESNs bearing neither variant (median = 69 months, 95% CI = 51–111 months). During the 11-year study interval, 59.0% of A*68:02-positive and 51.2% of B*42-C*17-positive initial HESNs became infected. HESNs with either of those two markers showed significantly more rapid rates of seroconversion than the remaining HESNs (p<0.001 by log-rank and Wilcoxon tests) (Figure 2).
Of the 11 individuals with both A*68:02 and B*42-C*17, five acquired HIV-1 at rates similar to those in carriers of B*42-C*17 alone; these 11 were treated as part of the B*42-C*17 group. Differences in acquisition-free time among the three HLA subgroups were tested for significance by the log-rank and Wilcoxon methods. Estimates of relative hazards (RH) of HIV-1 acquisition are based on Cox proportional hazards models. The numbers of subjects remaining at-risk at successive follow-up intervals are tabulated below the Kaplan-Meier plots. The seroconversion rate does not reflect the overall rate in the entire ZEHRP cohort, because couples in this study were selected for clinical and behavioral HIV-1 risk characteristics.
Analysis of GUI as a potential confounder
Among HESN partners, 23% had GUI during the 6 months prior to transmission or the end of follow-up. However, GUI was not associated with HLA class I variants of interest (Table S2).
In a comprehensive analysis of a cohort of HIV-1-serodiscordant couples including the largest known set of transmission pairs, none of the HLA-B alleles most widely reported to influence viremia, CD4 count or disease progression in subtype C infection were associated with occurrence or rate of acquisition of HIV-1 infection in the HESN partners. Specifically, although B*13:02, B*18, B*45:01, B*57:01-03, B*58:01, B*58:02, and B*81:01 have all been implicated as determinants of VL in sub-Saharan African populations , , , they had no corresponding effect on acquisition of HIV-1 subtype C infection in Zambians.
One HLA-B allele (B*42:01) that has shown some consistency in its association with better control of viremia , , had a clearly unfavorable effect on acquisition (Table 1). The B*42-C*17 haplotype is carried in 18.8% of Zambians. Strong linkage disequilibrium (LD) between B*42 and C*17 in our study population (r = 0.97)  precluded separate analysis of either B*42 or C*17 alone. However, these two alleles occur frequently in the A*30-C*17-B*42 haplotype, and the favorable effect of A*30 seen earlier  makes it an unlikely contributor to the acceleration of infection seen here with the B–C haplotype.
Additional HLA-B alleles (B*14, B*44 and B*51) were differentially distributed between SCs and pHESNs; however, their impact diminished with more comprehensive analyses. Relatively slow disease progression has been documented for B*51 in individuals infected largely with subtype B , , but that pattern has not extended to other populations, and association with lower susceptibility to infection has been inconsistent . At the population level, mutations in HIV-1 induced by B*51 may account for viral evolution to fixation in infected individuals, but it is difficult to predict how such changes in the virus might affect acquisition in HESNs . Findings with B*14 and B*44 have been even less consistent across ethnic and viral subtype boundaries , . The contributions of these alleles to susceptibility are unclear and appear to be minor. Taken as a whole, such strikingly disparate effects of HLA-B alleles on the two outcome measures (control of VL and susceptibility to infection) in the same cohort imply that if HLA-B alleles play any substantial role in modulating resistance or susceptibility to infection, the convincingly documented mechanism by which those alleles mediate CTL response to specific viral epitopes during the course of subtype C infection is unlikely to represent the dominant mechanism regulating heterosexual acquisition.
As for HLA-A alleles, only the relatively recently investigated African HLA-A*74:01 has shown a comparably favorable effect on both VL in HIV-infected individuals , ,  and acquisition among Tanzanians  as well as in HESN Zambian women in our study. Its linkage disequilibrium (LD) with B*57 could not explain the association with slower acquisition in Zambian women because the latter allele showed no such effect in the absence of A*74:01. This HLA-A allele does not show predominant LD with any particular HLA-C allele, but its pattern of LD in the class I region raises the possibility that it tags a variant involved in one of the natural killer cell pathways. The findings for A*74:01 across populations of African ancestry merit further investigation.
The strongest association seen among Zambian HESNs with any class I allele was between HLA-A*68:02 and accelerated HIV-1 acquisition. The relationship was opposite to the protective association previously reported with this allele as part of the A02 supertype in Kenyan commercial sex workers heavily exposed to HIV-1 subtype A , . A*68:02 also appeared to be associated with relative resistance to HIV-1 subtype B infection among men of European ancestry . Differential effects in the setting of different viral subtypes could explain these discrepant associations. A haplotype effect would be another possible reason for the population-specific findings. However, in Zambians A*68:02 is found most frequently in the A*68:02-C*03-B*15 haplotype–but in only 5% of HESNs and not in tight LD. Regardless of the population-specific findings, if there is a biologic basis for the association of A*68:02 with susceptibility to subtype C infection in Zambians, a similarly unfavorable relationship has not been reported for viremia or disease control with any HIV-1 subtype. Conversely, A*36 in HESN Zambians did not contribute to the acquisition of infection, even though this allele was associated with both higher VL in their index partners and accelerated viral transmission from them to their HESN partners , . Thus, for both of these HLA-A alleles our data suggest that the mechanisms underlying immune control of viremia and acquisition of HIV-1 infection are indeed distinct , .
Our earlier analysis of killer immunoglobulin-like receptor (KIR) genes has revealed some evidence that genetic associations with HIV-1 transmission may operate indirectly through prominent cofactors (e.g., coinfections). Specifically, KIR2D4*001 as an unfavorable marker in the Zambian cohort had dual associations with HIV-1 transmission by index partners and genital ulcer in index partners . Genital ulcers and inflammation (GUI) also represent a prominent cofactor for HIV-1 acquisition by HESNs, but the HLA class I variants highlighted here had no appreciable impact on GUI; that is, GUI could not have significantly confounded the observed HLA effects on HIV-1 acquisition.
The frequency of HLA-B supertypes has been inversely related to the degree of HIV-1 disease control in infected individuals of European and of African ancestry , , . However, our analyses excluded any substantial contribution by either HLA-A or HLA-B supertypes to HIV-1 acquisition in Zambians. As with the individual alleles, our data imply that the differences in peptide-binding patterns captured by supertype clustering do not influence acquisition of HIV-1 infection in the way they do HIV-1 disease control.
These findings could have important implications for the design of CTL-based vaccines. The evidence for the occurrence of CTL responses in HESNs has been inconsistent , , . However, even if it could be conclusively shown that HLA allele-specific CTL responses differentially promote or retard acquisition of infection, the profile of the genetic polymorphisms that enhance or impede that process appears to be so different from the profile of those controlling post-seroconversion immunologic events that the strategies for designing CTL-based prophylactic and therapeutic vaccines would almost surely have to diverge along those lines .
In summary, our data indicate that variation according to individual HLA-A, -B and –C alleles in HESNs and, by implication, whatever differential CTL responses may be mediated by those different alleles, fail to explain much if any variation in resistance to infection among Zambian partners exposed to HIV-1 subtype C. Of course, these disparate results for individual alleles do not preclude possible contributions of class I molecules by any of several alternative mechanisms. These alternatives include differential interaction of class I proteins with non-classical class I molecules ,  receptors in NK cell pathways , , chaperones , T-cell receptors , and immunoglobulin-like transcript 4 molecules . Hopefully, the context of cohabiting serodiscordant couples will afford further opportunity to elucidate the distinctive roles of HLA class I polymorphism in acquisition of HIV-1 infection and in disease control.
Materials and Methods
This study followed the human experimentation guidelines of the United States Department of Health and Human Services, and all enrolled patients provided written informed consent. The work presented here was further approved by Institutional Review Boards at University of Alabama at Birmingham, on a yearly basis.
From 1995 through 2006 in the Zambia-Emory HIV Research Project (ZEHRP) we prospectively evaluated cohabiting HIV-1 serodiscordant couples consisting of a seropositive index partner and an HESN partner. The procedures for participant recruitment, counseling, quarterly follow-up, clinical examination (for genital ulcer/inflammation) and laboratory testing (for HIV-1 serology and viral load) have been described elsewhere , , , . All couples whose HESN partner acquired virologically linked HIV-1 from the known index partner during follow-up were included in this study. Non-transmitting couples were selected on the basis of behavioral or clinical measures of sexual exposure. To concentrate on the level of increased risk experienced by each couple during the most recent six months of follow-up, we cumulated known predictors of risk for heterosexual transmission present at the two most recent quarterly visits [circumcision, genital ulcer/inflammation (GUI) by history or examination, pregnancy, and sperm in vaginal fluid] in a composite risk score. Although average risk scores were lower in the non-transmitting couples available for study than in transmitting couples, we disproportionally sampled non-transmitting couples with greater exposure to risk predictors. For transmitting couples, viruses detected in seroconverters had to be similar to those present in their cohabiting index partners by phylogenetic analysis of subgenomic HIV-1 sequences corresponding to gag, env and the long terminal repeat regions , . Couples with unlinked or ambiguous viruses or inadequate follow-up time (<9 months) or missing information were excluded. Through December 2006, 568 couples became eligible for analyses (Figure 1).
HLA class I typing
Methods for HLA class I genotyping have been described elsewhere , . In brief, genomic DNA extracted from buffy coats or whole blood was used for PCR with sequence-specific primers (SSP) (Dynal/Invitrogen, Brown Deer, WI), sequence-specific oligonucleotide (SSO) probe hybridization (Innogenetics, Alpharetta, GA), and sequencing-based typing (SBT) (Abbott Molecular, Inc., Des Plaines, IL). Most HLA class I alleles were defined to 4-digit specificities. HLA haplotypes and supertypes were assigned using the expectation-maximization (EM) algorithm in SAS Genetics (SAS Institute, Cary, NC) , .
HLA class I supertypes were assigned according to previous classification and a recent update , ,  designating four HLA-A supertypes (A01, A02, A03 and A24) and six HLA-B supertypes (B07, B08, B27, B44, B58 and B62). For analytic purposes here A*3001 and A*2902 were left unclassified along with other unclassified supertypes according to the previous framework . For comparison with previous studies we also clustered several related HLA-A alleles into a supertype designated A02/6802 .
Earlier analyses of the Zambian cohort have addressed various aspects of HIV-1 transmission and VL , , , , , . Here, we focused on deciphering the role of HLA class I variants [2- and 4-digit allele levels (Figure 1b), haplotypes, and supertypes] in HIV-1 acquisition among Zambian HESNs. Statistical software packages in SAS 9.2 with SAS/Genetics (SAS Institute Inc., Cary, NC) were used for all analyses. We first assessed non-genetic factors, including age, gender, risk score, direction of HIV-1 transmission (male-to-female and female-to-male) and index partner viral load (VL) (Table 1). Distribution of class I alleles were evaluated for a) Hardy-Weinberg equilibrium (HWE), b) association with time to HIV-1 acquisition (Cox proportional hazards models and Kaplan-Meier plots), c) association with HIV-1 infection status at the end of follow-up (logistic regression models). Kaplan-Meier plots illustrate differences in transmission associated with specific genetic markers; because non-transmitting couples with relatively higher frequencies of risk predictors were selectively included, these transmission rates do not reflect rates in the entire prospectively observed discordant couple population in Zambia. The overall annual HIV-1 seroincidence (7–9/100 PY) represents a one-half to two-thirds reduction in transmission following the introduction of joint testing and counseling .
Multivariable analysis highlighted genetic and non-genetic factors that showed independent associations with HIV-1 acquisition, after stepwise elimination (p>0.05) of probable or suspected HLA factors. In light of the multiple comparisons performed, we also calculated a false discovery rate q value for all HLA variants formally tested in univariate models (Table 2) . Alleles with q<0.2 received further consideration, as suggested by earlier work .
Lack of association between HLA class I supertypes and HIV-1 acquisition among 568 Zambians cohabiting with HIV-1 seropositive partners.
Lack of association between HLA class I variants and genital ulcer/inflammation in 568 Zambians who were HIV-1 seronegative at enrollment.
We thank staff and study participants in the Zambia-Emory HIV-1 Research Project for their valuable contributions to various aspects of this study. We are also grateful to Paul Farmer and Amanda Tichacek for data management, to Yun Joo Yoo for templates of structured SAS programming, and to Kui Zhang for advice on FDR calculations.
Conceived and designed the experiments: SA EH JM JT RAK. Performed the experiments: WS DH. Analyzed the data: WS IB RM JT RAK. Contributed reagents/materials/analysis tools: DH RM JM SA EH RAK. Wrote the paper: WS IB JT RAK.
- 1. Dunkle KL, Stephenson R, Karita E, Chomba E, Kayitenkore K, et al. (2008) New heterosexually transmitted HIV infections in married or cohabiting couples in urban Zambia and Rwanda: an analysis of survey and clinical data. Lancet 371: 2183–2191.
- 2. Corey L, Wald A, Celum CL, Quinn TC (2004) The effects of herpes simplex virus-2 on HIV-1 acquisition and transmission: a review of two overlapping epidemics. J Acquir Immune Defic Syndr 35: 435–445.
- 3. Kaslow RA, Dorak T, Tang JJ (2005) Influence of host genetic variation on susceptibility to HIV type 1 infection. J Infect Dis 191: Suppl 1S68–77.
- 4. Borrow P, Shattock RJ, Vyakarnam A (2010) Innate immunity against HIV: a priority target for HIV prevention research. Retrovirology 7: 84.
- 5. Fideli US, Allen SA, Musonda R, Trask S, Hahn BH, et al. (2001) Virologic and immunologic determinants of heterosexual transmission of human immunodeficiency virus type 1 in Africa. AIDS Res Hum Retroviruses 17: 901–910.
- 6. Tang J, Shao W, Yoo YJ, Brill I, Mulenga J, et al. (2008) Human leukocyte antigen class I genotypes in relation to heterosexual HIV type 1 transmission within discordant couples. J Immunol 181: 2626–2635.
- 7. Tang J, Penman-Aguilar A, Lobashevsky E, Allen S, Kaslow RA (2004) HLA-DRB1 and -DQB1 alleles and haplotypes in Zambian couples and their associations with heterosexual transmission of HIV type 1. J Infect Dis 189: 1696–1704.
- 8. Fellay J, Shianna KV, Telenti A, Goldstein DB (2010) Host genetics and HIV-1: the final phase? PLoS Pathog 6: e1001033.
- 9. Kaslow RA, Carrington M, Apple R, Park L, Munoz A, et al. (1996) Influence of combinations of human major histocompatibility complex genes on the course of HIV-1 infection. Nat Med 2: 405–411.
- 10. Carrington M, Nelson GW, Martin MP, Kissner T, Vlahov D, et al. (1999) HLA and HIV-1: heterozygote advantage and B*35-Cw*04 disadvantage. Science 283: 1748–1752.
- 11. Flores-Villanueva PO, Yunis EJ, Delgado JC, Vittinghoff E, Buchbinder S, et al. (2001) Control of HIV-1 viremia and protection from AIDS are associated with HLA-Bw4 homozygosity. Proc Natl Acad Sci U S A 98: 5140–5145.
- 12. Gao X, Nelson GW, Karacki P, Martin MP, Phair J, et al. (2001) Effect of a single amino acid change in MHC class I molecules on the rate of progression to AIDS. N Engl J Med 344: 1668–1675.
- 13. Trachtenberg E, Korber B, Sollars C, Kepler TB, Hraber PT, et al. (2003) Advantage of rare HLA supertype in HIV disease progression. Nat Med 9: 928–935.
- 14. Tang J, Kaslow RA (2003) The impact of host genetics on HIV infection and disease progression in the era of highly active antiretroviral therapy. AIDS 17: Suppl 4S51–60.
- 15. Fellay J, Shianna KV, Ge D, Colombo S, Ledergerber B, et al. (2007) A whole-genome association study of major determinants for host control of HIV-1. Science 317: 944–947.
- 16. Migueles SA, Sabbaghian MS, Shupert WL, Bettinotti MP, Marincola FM, et al. (2000) HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors. Proc Natl Acad Sci U S A 97: 2709–2714.
- 17. O'Brien SJ, Gao X, Carrington M (2001) HLA and AIDS: a cautionary tale. Trends Mol Med 7: 379–381.
- 18. Carrington M, O'Brien SJ (2003) The influence of HLA genotype on AIDS. Annu Rev Med 54: 535–551.
- 19. Lazaryan A, Lobashevsky E, Mulenga J, Karita E, Allen S, et al. (2006) Human leukocyte antigen B58 supertype and human immunodeficiency virus type 1 infection in native Africans. J Virol 80: 6056–6060.
- 20. Tang J, Malhotra R, Song W, Brill I, Hu L, et al. (2010) Human leukocyte antigens and HIV type 1 viral load in early and chronic infection: predominance of evolving relationships. PloS one 5: e9629.
- 21. Gao X, O'Brien TR, Welzel TM, Marti D, Qi Y, et al. (2010) HLA-B alleles associate consistently with HIV heterosexual transmission, viral load, and progression to AIDS, but not susceptibility to infection. AIDS 24: 1835–1840.
- 22. Lazaryan A, Song W, Lobashevsky E, Tang J, Shrestha S, et al. (2010) Human leukocyte antigen class I supertypes and HIV-1 control in African Americans. J Virol 84: 2610–2617.
- 23. MacDonald KS, Embree J, Njenga S, Nagelkerke NJ, Ngatia I, et al. (1998) Mother-child class I HLA concordance increases perinatal human immunodeficiency virus type 1 transmission. J Infect Dis 177: 551–556.
- 24. Mackelprang RD, Carrington M, John-Stewart G, Lohman-Payne B, Richardson BA, et al. (2010) Maternal human leukocyte antigen A*2301 is associated with increased mother-to-child HIV-1 transmission. J Infect Dis 202: 1273–1277.
- 25. Beyrer C, Artenstein AW, Rugpao S, Stephens H, VanCott TC, et al. (1999) Epidemiologic and biologic characterization of a cohort of human immunodeficiency virus type 1 highly exposed, persistently seronegative female sex workers in northern Thailand. Chiang Mai HEPS Working Group. J Infect Dis 179: 59–67.
- 26. MacDonald KS, Fowke KR, Kimani J, Dunand VA, Nagelkerke NJ, et al. (2000) Influence of HLA supertypes on susceptibility and resistance to human immunodeficiency virus type 1 infection. J Infect Dis 181: 1581–1589.
- 27. McLaren PJ, Ball TB, Wachihi C, Jaoko W, Kelvin DJ, et al. (2010) HIV-exposed seronegative commercial sex workers show a quiescent phenotype in the CD4+ T cell compartment and reduced expression of HIV-dependent host factors. J Infect Dis 202: Suppl 3S339–344.
- 28. Koehler RN, Walsh AM, Saathoff E, Tovanabutra S, Arroyo MA, et al. (2010) Class I HLA-A*7401 Is Associated with Protection from HIV-1 Acquisition and Disease Progression in Mbeya, Tanzania. J Infect Dis 202: 1562–1566.
- 29. Rowland-Jones SL, Dong T, Fowke KR, Kimani J, Krausa P, et al. (1998) Cytotoxic T cell responses to multiple conserved HIV epitopes in HIV-resistant prostitutes in Nairobi. J Clin Invest 102: 1758–1765.
- 30. Addo MM, Altfeld M, Brainard DM, Rathod A, Piechocka-Trocha A, et al. (2011) Lack of Detectable HIV-1–Specific CD81 T Cell Responses in Zambian HIV-1–Exposed Seronegative Partners of HIV-1–Positive Individuals. J Infect Dis 203: 258–262.
- 31. Tang J, Tang S, Lobashevsky E, Myracle AD, Fideli U, et al. (2002) Favorable and unfavorable HLA class I alleles and haplotypes in Zambians predominantly infected with clade C human immunodeficiency virus type 1. J Virol 76: 8276–8284.
- 32. Kiepiela P, Leslie AJ, Honeyborne I, Ramduth D, Thobakgale C, et al. (2004) Dominant influence of HLA-B in mediating the potential co-evolution of HIV and HLA. Nature 432: 769–775.
- 33. Leslie A, Matthews PC, Listgarten J, Carlson JM, Kadie C, et al. (2010) Additive contribution of HLA class I alleles in the immune control of HIV-1 infection. J Virol 84: 9879–9888.
- 34. Matthews PCAE, Listgarten J, Leslie A, Mkhwanazi N, Carlson JM, et al. (2011) HLA-A*7401-Mediated Control of HIV Viremia Is Independent of Its Linkage Disequilibrium with HLA-B*5703. J Immunol 186: 5675–5686.
- 35. Brumme Z, Wang B, Nair K, Brumme C, de Pierres C, et al. (2009) Impact of select immunologic and virologic biomarkers on CD4 cell count decrease in patients with chronic HIV-1 subtype C infection: results from Sinikithemba Cohort, Durban, South Africa. Clin Infect Dis 49: 956–964.
- 36. Teixeira SL, Bastos FI, Hacker MA, Morgado MG (2009) Distribution of CCR5 genotypes and HLA Class I B alleles in HIV-1 infected and uninfected injecting drug users from Rio de Janeiro, Brazil. Infect Genet Evol 9: 638–642.
- 37. Kawashima Y, Pfafferott K, Frater J, Matthews P, Payne R, et al. (2009) Adaptation of HIV-1 to human leukocyte antigen class I. Nature 458: 641–645.
- 38. Li S, Jiao H, Yu X, Strong AJ, Shao Y, et al. (2007) Human leukocyte antigen class I and class II allele frequencies and HIV-1 infection associations in a Chinese cohort. J acquir Immune Defic SYndr 44: 121–131.
- 39. Xu MY, Hong KX, Deng XL, Li J, Peng H, et al. (2004) Association of HLA-B alleles with human immunodeficiency virus type 1 infection in the Yi ethnic group in Sichuan province. Biomed Environ Sci 17: 203–208.
- 40. Liu C, Carrington M, Kaslow RA, Gao X, Rinaldo CR, et al. (2003) Association of polymorphisms in human leukocyte antigen class I and transporter associated with antigen processing genes with resistance to human immunodeficiency virus type 1 infection. J Infect Dis 187: 1404–1410.
- 41. Merino A, Malhotra R, Morton M, Mulenga J, Allen S, et al. (2011) Impact of a Functional KIR2DS4 Allele on Heterosexual HIV-1 Transmission among Discordant Zambian Couples. J Infect Dis 203: 487–495.
- 42. Doytchinova IA, Guan P, Flower DR (2004) Identifiying human MHC supertypes using bioinformatic methods. J Immunol 172: 4314–4323.
- 43. Lund O, Nielsen M, Kesmir C, Petersen AG, Lundegaard C, et al. (2004) Definition of supertypes for HLA molecules using clustering of specificity matrices. Immunogenetics 55: 797–810.
- 44. Kaul R, Rutherford J, Rowland-Jones SL, Kimani J, Onyango JI, et al. (2004) HIV-1 Env-specific cytotoxic T-lymphocyte responses in exposed, uninfected Kenyan sex workers: a prospective analysis. AIDS 18: 2087–2089.
- 45. Hladik F, Desbien A, Lang J, Wang L, Ding Y, et al. (2003) Most highly exposed seronegative men lack HIV-1-specific, IFN-gamma-secreting T cells. J Immunol 171: 2671–2683.
- 46. Walker B (2002) Harnessing the immune system in the fight against AIDS. Keystone Conference on HIV Pathogenesis Keystone, Colorado, USA.
- 47. Yunis EJ, Romero V, Diaz-Giffero F, Zuniga J, Koka P (2007) Natural Killer Cell Receptor NKG2A/HLA-E Interaction Dependent Differential Thymopoiesis of Hematopoietic Progenitor Cells Influences the Outcome of HIV Infection. J Stem Cells 2: 237–248.
- 48. Hoare HL, Sullivan LC, Pietra G, Clements CS, Lee EJ, et al. (2006) Structural basis for a major histocompatibility complex class Ib-restricted T cell response. Nat Immunol 7: 256–264.
- 49. Winchester R, Pitt J, Charurat M, Magder LS, Goring HH, et al. (2004) Mother-to-child transmission of HIV-1: strong association with certain maternal HLA-B alleles independent of viral load implicates innate immune mechanisms. J Acquir Immune Defic Syndr 36: 659–670.
- 50. Khakoo SI, Carrington M (2006) KIR and disease: a model system or system of models? Immunol Rev 214: 186–201.
- 51. Turnquist HR, Schenk EL, McIlhaney MM, Hickman HD, Hildebrand WH, et al. (2002) Disparate binding of chaperone proteins by HLA-A subtypes. Immunogenetics 53: 830–834.
- 52. Tomiyama H, Yamada N, Komatsu H, Hirayama K, Takiguchi M (2000) A single CTL clone can recognize a naturally processed HIV-1 epitope presented by two different HLA class I molecules. Eur J Immunol 30: 2521–2530.
- 53. Huang J, Goedert JJ, Sundberg EJ, Cung TD, Burke PS, et al. (2009) HLA-B*35-Px-mediated acceleration of HIV-1 infection by increased inhibitory immunoregulatory impulses. J Exp Med 206: 2959–2966.
- 54. McKenna SL, Muyinda GK, Roth D, Mwali M, Ng'andu N, et al. (1997) Rapid HIV testing and counseling for voluntary testing centers in Africa. AIDS 11: Suppl 1S103–110.
- 55. Allen S, Meinzen-Derr J, Kautzman M, Zulu I, Trask S, et al. (2003) Sexual behavior of HIV discordant couples after HIV counseling and testing. AIDS 17: 733–740.
- 56. Kempf MC, Allen S, Zulu I, Kancheya N, Stephenson R, et al. (2008) Enrollment and retention of HIV discordant couples in Lusaka, Zambia. J Acquir Immune Defic Syndr 47: 116–125.
- 57. Trask SA, Derdeyn CA, Fideli U, Chen Y, Meleth S, et al. (2002) Molecular epidemiology of human immunodeficiency virus type 1 transmission in a heterosexual cohort of discordant couples in Zambia. J Virol 76: 397–405.
- 58. Sidney J, Peters B, Frahm N, Brander C, Sette A (2008) HLA class I supertypes: a revised and updated classification. BMC Immunol 9: 1.
- 59. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of Royal Statistical Society, Series B 57: 289–300.
- 60. Bhattacharya T, Daniels M, Heckerman D, Foley B, Frahm N, et al. (2007) Founder effects in the assessment of HIV polymorphisms and HLA allele associations. Science 315: 1583–1586.