Share this article on:

HLA-B alleles associate consistently with HIV heterosexual transmission, viral load, and progression to AIDS, but not susceptibility to infection

Gao, Xiaojianga,d; O'Brien, Thomas Rb; Welzel, Tania Mb; Marti, Darlenea,d; Qi, Yinga,d; Goedert, James Jb; Phair, Johnc; Pfeiffer, Ruthb; Carrington, Marya,d

doi: 10.1097/QAD.0b013e32833c3219
Basic Science

Objective: HLA class I polymorphism is known to affect the rate of progression to AIDS after infection with HIV-1. Here we test the consistency of HLA-B allelic effects on progression to AIDS, heterosexual HIV transmission, and ‘set point’ viral levels.

Methods: We used adjusted Cox proportional hazard models in previously published relative hazard values for the effect of HLA-B alleles on progression to AIDS (n = 1089). The transmission study included 303 HIV-1-infected men with hemophilia and their 323 female sex partners (Multicenter Hemophilia Cohort Study cohort). Among 259 HIV-1 seroconverters (Multicenter AIDS Cohort Study cohort), HIV RNA levels at ‘set point’ were determined in stored plasma samples by a reverse-transcription polymerase chain reaction assay. HLA-B genotyping was performed by sequence-specific oligonucleotide hybridization and DNA sequencing.

Results: Several HLA-B alleles showed consistent associations for AIDS risk, infectivity, and ‘set point’ HIV RNA. HLA-B*35 was associated with more rapid progression to AIDS (relative hazard 1.39; P = 0.008), greater infectivity (odds ratio 3.14; P = 0.002), and higher HIV RNA (P = 0.01), whereas the presence of either B*27 or B*57 associated with slower progression to AIDS (B*27: relative hazard 0.49, P < 0.001; B*57: relative hazard 0.40, P < 0.0001), less infectivity (odds ratio 0.22 and 0.31, respectively, though not significant), and lower viral levels (P < 0.0001). Importantly, HLA-B polymorphism in female partners was not associated with susceptibility to HIV-1 infection.

Conclusion: HLA-B polymorphisms that affect the risk of AIDS may also alter HIV-1 infectivity, probably through the common mechanism of viral control, but they do not appear to protect against infection in our cohort.

aCancer and Inflammation Program, Laboratory of Experimental Immunology, SAIC Frederick Inc., NCI-Frederick, Frederick, USA

bDivision of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Rockville, Maryland, USA

cNorthwestern University Medical School, Chicago, Illinois, USA

dRagon Institute of MGH, MIT and Harvard, Boston, Massachusetts, USA.

Received 8 May, 2009

Revised 14 May, 2010

Accepted 14 May, 2010

Correspondence to Mary Carrington, Cancer and Inflammation Program, Laboratory of Experimental Immunology, P.O. Box B, Bldg 560, Room 21-89, Frederick, MD 21702, USA. Tel: +1 301 846 1390; fax: +1 301 846 6771; e-mail:

Back to Top | Article Outline


Among a dozen or so genes affecting the natural history of HIV infection, HLA class I shows the strongest association with progression to AIDS [1–3]. In response to HIV infection, human leukocyte antigen (HLA) class I molecules play important roles in both T-cell-mediated adaptive immunity and natural killer (NK) cell-mediated innate immunity. Individual HLA class I may influence AIDS pathogenesis by binding and presenting immunodominant HIV epitopes to T cells and/or interacting with killer immunoglobulin-like receptors (KIRs) to regulate NK cell activity [4,5]. In particular, the HLA-B*35Px alleles (a group of B*35 and B*53 subtypes encoding molecules that share similar peptide-binding characteristics, mainly consisted of B*3502, B*3503, and B*5301 in our cohorts [6]) associate with rapid progression to AIDS. B*27 and B*57 have been consistently associated with reduced rate of progression to AIDS in multiple epidemiological studies [7]. However, effects of HLA on HIV transmission are relatively poorly defined, though the influence of specific HLA alleles or shared HLA alleles [8,9] on risk of mother-to-infant HIV transmission [10,11] has been reported. HLA influence on heterosexual HIV transmission has also been reported in a Zambian cohort [12].

We recently showed that alleles encoding the HLA-Bw4 and Bw6 epitopes of the HLA-B molecules are associated with risk for HIV transmission in heterosexual couples [13]. The presence of HLA-Bw4 in HIV-1-infected men was associated with a decreased risk of male-to-female HIV-1 transmission, perhaps because HLA-Bw4 allotypes serve as ligands for KIR3DL1 expressed on NK cells. In the present report, we further investigated individual HLA-B alleles for their contributions to the risk of HIV-1 transmission from HIV-infected men to their female partners (infectivity). We also examined the effect of these alleles on set point HIV RNA levels in HIV seroconverters. These data allowed us to compare the pattern of HLA associations with AIDS progression, HIV-1 transmission, and HIV-1 viral loads in an attempt to examine potential connections among these associations.

Back to Top | Article Outline


Study participants

The participants for the analysis of HIV transmission were participants in the Multicenter Hemophilia Cohort Study (MHCS), a prospective study of patients with hemophilia and their female partners enrolled at hemophilia treatment sites in the United States and Europe [14]. Between 1986 and 1998, MHCS investigators enrolled women who had had vaginal intercourse with an MHCS participant into a study of heterosexual HIV-1 transmission. At enrollment, the female partner completed a standardized self-administered questionnaire that included questions about HIV-1 exposure and sexual practices. Blood samples were collected for HIV-1 testing. Women were invited for follow-up visits at 6-month intervals for HIV-1-testing and to complete additional questionnaires. The present study includes 303 HIV-infected men with hemophilia and their 323 female partners, of whom 43 (13.3%) had become infected with HIV-1 [13]. Twenty men had two partners (two of the 40 partners became infected).

To examine the effect of HLA alleles on HIV-1 RNA ‘set point’, 259 white men from the Multicenter AIDS Cohort Studies (MACS) were typed for HLA class I. These participants were ‘seroconverters’ whose estimated date of seroconversion was based on a negative result for antibody to HIV followed by a positive result within an average of 18 months. The effects of HLA on transmission and viral load set point were compared to previously published survival analyses in which the influence of HLA on progression to AIDS-1987 was determined in seroconverters [6,7].

Back to Top | Article Outline

Viral assays and HIV-1 RNA set point

The measurement of HIV-1 RNA levels for MACS participants has been described previously [15]. HIV-1 RNA was quantified in stored plasma samples by the Amplicor reverse transcription-PCR assay (Roche Diagnostic Corporation, Indianapolis, Indiana, USA), which has a lower detection limit of 400 copies/ml. Measurements between 12 and 24.5 months after seroconversion were used to calculate the set point value, excluding any measures after combination antiretroviral therapy was begun. The HIV-1 RNA measure from the first positive visit was also excluded if it fell within this 12–24.5-month period. The set point was then calculated as the mean of all HIV-1 RNA measures. Only patients with two or more measures were included in the present analysis.

Back to Top | Article Outline

HLA typing

High-resolution (four-digit) HLA typing was carried out using sequence-specific oligonucleotide (SSO) hybridization as described previously [6,7]. Genomic DNA was amplified using locus-specific primers flanking exons 2 and 3 of the class I HLA genes. The PCR products were fixed on nucleic acid transfer membranes and hybridized with SSO probes matching polymorphic nucleotide motifs of all known HLA alleles. Alleles and genotypes were assigned by SSO hybridization patterns. Ambiguous typing results were resolved by sequencing analysis of exons 2 and 3. High-resolution (four-digit) class I typing was informative for 303 hemophiliac men and 276 female partners. Data analyses were performed for four-digit and two-digit alleles separately. Except for the B*35Px/PY allele groups, data analyses shown in this report were based on two-digit alleles. B*35Px alleles are defined as B*3502, 3503, 3504, and 5301, and B*35PY alleles are defined as B*3501 and 3508.

Back to Top | Article Outline

Statistical analysis

In the analysis of HIV transmission, the effect of HLA alleles on male HIV-1 infectivity and female partner susceptibility was evaluated in two-by-two tables using Fisher's exact test (Proc FREQ, SAS 9.1; The SAS Institute, Cary, North Carolina, USA) and with logistic regression by calculating adjusted odds ratios (aORs) and 95% confidence intervals (CI), adjusted for race in two categories (white/nonwhite) and for age of the male partner at estimated date of seroconversion. The effect of the HIV-infected male's HLA types on transmission was, therefore, tested using a total of 323 pairs, in which the 283 men who had one female partner were each counted once in the analysis and the 20 men having two partners were each counted twice (data shown in tables). We also performed the analyses with only one randomly selected partner for patients with two partners and the results remained the same (data not shown in tables). To evaluate the influence of racial background on clinical outcomes, all analyses were performed separately with whites only and with all racial groups combined. Nonwhite groups were too small to be analyzed separately.

For HLA associations with HIV RNA levels in MACS seroconverters, t-tests were used to examine the difference of mean viral load between those carrying a particular allele and those who did not carry that allele. We also analyzed the viral load data using linear regression models that adjusted for age in 5-year categories.

Back to Top | Article Outline


The demographic information for the MHCS participants is summarized in Table 1. The median age at the estimated date of HIV-1 seroconversion was 26 years for men and 24 years for women. Whites accounted for 85.5% of men and 82.3% of women.

Table 1

Table 1

The effect of HLA on HIV-1 transmission could be examined in 323 couples (323 female partners of 303 HIV-infected men with hemophilia). HIV-1 transmission occurred in 43 of these 323 pairs (13.3%). The results for all participants are presented in Tables 2 and 3, but similar results were found when the analysis was restricted to whites only (not shown). Table 2 shows the effect of two-digit HLA-B alleles present in the HIV-positive hemophiliac men on HIV-1 transmission to their female partners after adjustment for race and age. Only B*35, which was the second most frequent allele, showed a significantly different distribution between the men who transmitted HIV compared to those who did not (aOR 3.14; P = 0.002). B*27 and B*57, the two protective alleles, were each detected only once in the transmitting group, resulting in two of the lowest ORs (0.22 and 0.31, respectively), though neither was statistically significant.

Table 2

Table 2

Table 3

Table 3

The alleles that had previously been associated with AIDS progression (HLA B*35, B*53, B*27, and B*57) were examined for consistency of their effects on AIDS progression, HIV-1 infectivity, and HIV-1 RNA levels (Table 3). Individuals with B*35 progressed to AIDS more rapidly (relative hazard 1.39; P = 0.008), had higher risk of infectivity among men with hemophilia (aOR 3.14; P = 0.002), and had a higher mean HIV RNA set point compared to those without this allele (4.61 HIV RNA log10 copies/ml vs. 4.32 HIV RNA log10 copies/ml, respectively; P = 0.01). The B*35Px subgroup was associated with more rapid progression to AIDS (relative hazard 1.92; P < 0.0001), showed the highest risk for HIV-1 transmission (aOR 3.5; P = 0.007), and had a significantly higher mean HIV RNA set point, as compared to all others (4.63 vs. 4.36 HIV RNA log10 copies/ml, respectively; P = 0.04). The B*35PY subgroup was not significantly associated with rapid progression to AIDS (relative hazard 1.12; P = 0.45), increased infectivity (aOR 2.04; P = 0.09), or higher HIV RNA (P = 0.07). HLA-B*53 (relative hazard = 1.85; P = 0.004) was not significantly associated with greater infectivity or higher mean HIV RNA, though both analyses tended toward susceptibility. B*53 is a common allele in people of African descent, but has a low frequency in Europeans. As our transmission cohort is primarily composed of European Americans, it is not surprising that any effect of B*53 did not reach significance.

Individually, both B*27 and B*57 had very low aORs (0.22 and 0.31, respectively) for HIV-1 infectivity among hemophiliac men, even though neither allele reached statistical significance due probably to small numbers.

The potential effect of HLA-B on susceptibility to HIV-1 infection in female partners was evaluated in 276 women (39 HIV-1-infected and 237 HIV-1-uninfected) who were successfully genotyped for HLA-B. No alleles showed a significant association with HIV-1 susceptibility (data not shown), including the HLA-B genotypes that are either deleterious or beneficial for progression to AIDS, infectivity, or HIV RNA set point (Table 4).

Table 4

Table 4

Back to Top | Article Outline


Diversity in the MHC has been identified as the most significant AIDS-regulating genetic factor [3]. Well documented evidence has indicated that HLA class I polymorphism specifically plays an important role in varied outcomes across individuals following HIV-1 infection [2]. However, analyses based on general populations comparing HIV-negative to HIV-positive individuals have shown little consistent evidence of HLA association with HIV infection. In general, the identification of appropriate HIV-negative and HIV-positive groups for comparison requires careful consideration, as the available groups have great potential for selection biases that may erroneously result in significant differences between the two groups. For example, the loss of rapid progressors among HIV-positive individuals, which is common across cohorts, may artificially result in a high frequency of variants that associate with longer term nonprogression. Analyses of serodiscordant sexual partners examine both viral transmitters and recipients, providing a better-controlled research design for examining the potential influence of HLA class I on HIV-1 transmission and susceptibility to infection.

We recently reported that HLA-Bw4 was associated with a lower risk for HIV-1 transmission from hemophiliac patients to their female partners [13]. The two low-risk alleles for HIV-1 transmission, B*27 and B*57, that we detected in the present study both belong to the Bw4 antigen group. The Bw4 epitope on the alpha-1 domain of the class I HLA molecule forms part of the F-pocket that accommodates the C-terminal anchor residue of the bound peptide. This complex is recognized by the NK cell receptor KIR3DL1, which may explain the general protection of the group of alleles containing the Bw4 epitope. In addition, B*27 and B*57 are known to have characteristics that can delay HIV-1 adaptation to the host MHC [16–18], explaining in part why these two alleles exhibit the strongest protective effect on AIDS progression, as well as HIV transmission, relative to all other Bw4-bearing HLA-B alleles.

HIV-1 transmitters (men with hemophilia) and recipients (their female partners) showed different patterns of HLA associations. In the men, B*35 (especially B*35Px, though B*35PY also tended in the same direction) was detected as a high-risk allele for HIV-1 transmission, whereas B*27 and B*57 were associated with low risk. In female partners, none of the previously detected AIDS-regulating HLA alleles showed an association with susceptibility to HIV-1 infection, consistent with a model in which HLA class I diversity affects outcome after HIV infection, but not infection itself. Although our sample size was small, there were no consistent tendencies for effects of these alleles on infection with those observed for viral load control, AIDS progression, and transmission. Our study was largely composed of European Americans infected with clade B virus, such that our results pertain to this ethnic group. Nevertheless, our data concur with a recent report showing that HLA alleles/haplotypes associated with transmission in an African cohort do not associate with acquisition in their heterosexual partners [12]. The different patterns of genetic associations illustrate that HIV-1 transmission and susceptibility to infection involve different host/viral mechanisms and, therefore, are subject to different genetic influences.

Our null data on susceptibility of heterosexual acquisition of HIV-1 infection among women are also consistent with a study on HLA and mother-to-infant HIV-1 transmission in which the HLA-B genotype of the recipient (the infant) did not alter the risk of transmission, whereas the HLA-B genotype of the HIV transmitter (the mother) was associated with varied risks for HIV-1 transmission [9]. If there is indeed an effect of HLA polymorphism on susceptibility to HIV-1 infection, it may very well differ from that for viral transmission. For example, MHC polymorphism might affect susceptibility by influencing innate immunity against HIV-1 infection through specific interactions with KIR. A larger sample with more HIV-1-infected women is needed to test this hypothesis.

B*35 showed consistent, significant associations with AIDS progression, HIV transmission, and HIV viral load levels, suggesting that the molecular variation of B*35 impacts these related processes in a coherent manner. Although these associations appeared to be driven primarily by B*35Px for both progression to AIDS and HIV infectivity, lack of viral load control was observed for both B*35Px and B*35PY. Thus, both subgroups of B*35 may confer some level of susceptibility, especially in terms of viral load control, though B*35Px appears to have a stronger effect on AIDS progression and infectivity.

The consistent effects of B*35Px, B*27, and B*57 on AIDS progression and transmission suggest that HLA polymorphism may influence these events through a common mechanism. A likely explanation is the control of HIV viremia. In support of this model, the high-risk B*35Px was associated with higher set point HIV RNA levels, an indicator of more rapid AIDS progression [19,20], whereas the two protective alleles B*27 and B*57 were associated with lower set point HIV RNA levels. Furthermore, an association between HIV RNA level in blood and heterosexual transmission has been reported previously [21], as has faster progression to AIDS with higher likelihood of heterosexual transmission [22,23]. Our analysis suggests that HLA class I polymorphisms contribute to these relationships.

Back to Top | Article Outline


We would like to thank Dr Pat Martin for helpful comments regarding this article. This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government. This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research.

The MACS is funded by the National Institute of Allergy and Infectious Diseases, with additional supplemental funding from the National Cancer Institute; and the National Heart, Lung, and Blood Institute: UO1-AI-35042, 5-M01-RR-00052 (GCRC), UO1-AI-35043, UO1-AI-37984, UO1-AI-35039, UO1-AI-35040, UO1-AI-37613, and UO1-AI-35041.

Back to Top | Article Outline


1. Carrington M, O'Brien SJ. The influence of HLA genotype on AIDS. Ann Rev Med 2003; 54:535–551.
2. Martin MP, Carrington M. Immunogenetics of viral infections. Curr Opin Immunol 2005; 17:510–516.
3. Fellay J, Shianna KV, Ge D, Colombo S, Ledergerber B, Weale M, et al. A whole-genome association study of major determinants for host control of HIV-1. Science 2007; 317:944–947.
4. Lanier LL. NK cell recognition. Ann Rev Immunol 2005; 23:225–274.
5. Yawata M, Yawata N, Draghi M, Little AM, Partheniou F, Parham P. Roles for HLA and KIR polymorphisms in natural killer cell repertoire selection and modulation of effector function [erratum appears in J Exp Med 2006; 203:1131]. J Exp Med 2006; 203:633–645.
6. Gao X, Nelson GW, Karacki P, Martin MP, Phair J, Kaslow R, et al. Effect of a single amino acid change in MHC class I molecules on the rate of progression to AIDS [see comment]. N Engl J Med 2001; 344:1668–1675.
7. Gao X, Bashirova A, Iversen AK, Phair J, Goedert JJ, Buchbinder S, et al. AIDS restriction HLA allotypes target distinct intervals of HIV-1 pathogenesis. Nat Med 2005; 11:1290–1292.
8. MacDonald KS, Embree JE, Nagelkerke NJ, Castillo J, Ramhadin S, Njenga S, et al. The HLA A2/6802 supertype is associated with reduced risk of perinatal human immunodeficiency virus type 1 transmission. J Infect Dis 2001; 183:503–506.
9. Winchester R, Pitt J, Charurat M, Magder LS, Goring HH, Landay A, et al. 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 2004; 36:659–670.
10. MacDonald KS, Embree J, Njenga S, Nagelkerke NJ, Ngatia I, Mohammed Z, et al. Mother-child class I HLA concordance increases perinatal human immunodeficiency virus type 1 transmission. J Infect Dis 1998; 177:551–556.
11. Mackelprang RD, John-Stewart G, Carrington M, Richardson B, Rowland-Jones S, Gao X, et al. Maternal HLA homozygosity and mother-child HLA concordance increase the risk of vertical transmission of HIV-1. J Infect Dis 2008; 197:1156–1161.
12. Tang J, Shao W, Yoo YJ, Brill I, Mulenga J, Allen S, et al. Human leukocyte antigen class I genotypes in relation to heterosexual HIV type 1 transmission within discordant couples. J Immunol 2008; 181:2626–2635.
13. Welzel TM, Gao X, Pfeiffer RM, Martin MP, O'Brien SJ, Goedert JJ, et al. HLA-B Bw4 alleles and HIV-1 transmission in heterosexual couples. AIDS 2007; 21:225–229.
14. Goedert JJ, Kessler CM, Aledort LM, Biggar RJ, Andes WA, White GC 2nd, et al. A prospective study of human immunodeficiency virus type 1 infection and the development of AIDS in subjects with hemophilia [see comment]. N Engl J Med 1989; 321:1141–1148.
15. Kirstein LM, Mellors JW, Rinaldo CR Jr, Margolick JB, Giorgi JV, Phair JP, et al. Effects of anticoagulant, processing delay, and assay method (branched DNA versus reverse transcriptase PCR) on measurement of human immunodeficiency virus type 1 RNA levels in plasma. J Clin Microbiol 1999; 37:2428–2433.
16. Kelleher AD, Long C, Holmes EC, Allen RL, Wilson J, Conlon C, et al. Clustered mutations in HIV-1 gag are consistently required for escape from HLA-B27-restricted cytotoxic T lymphocyte responses. J Exp Med 2001; 193:375–386.
17. Gillespie GM, Kaul R, Dong T, Yang HB, Rostron T, Bwayo JJ, et al. Cross-reactive cytotoxic T lymphocytes against a HIV-1 p24 epitope in slow progressors with B*57. AIDS 2002; 16:961–972.
18. Altfeld M, Addo MM, Rosenberg ES, Hecht FM, Lee PK, Vogel M, et al. Influence of HLA-B57 on clinical presentation and viral control during acute HIV-1 infection. AIDS 2003; 17:2581–2591.
19. Mellors JW, Rinaldo CR Jr, Gupta P, White RM, Todd JA, Kingsley LA. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 1996; 272:1167–1170.
20. Rodriguez B, Sethi AK, Cheruvu VK, Mackay W, Bosch RJ, Kitahata M, et al. Predictive value of plasma HIV RNA level on rate of CD4 T-cell decline in untreated HIV infection. JAMA 2006; 296:1498–1506.
21. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group [see comment]. N Engl J Med 2000; 342:921–929.
22. O'Brien TR, Busch MP, Donegan E, Ward JW, Wong L, Samson SM, et al. Heterosexual transmission of human immunodeficiency virus type 1 from transfusion recipients to their sex partners. J Acquir Immune Defic Syndr 1994; 7:705–710.
23. Lee TH, Sakahara N, Fiebig E, Busch MP, O'Brien TR, Herman SA. Correlation of HIV-1 RNA levels in plasma and heterosexual transmission of HIV-1 from infected transfusion recipients. J Acquir Immune Defic Syndr Hum Retrovirol 1996; 12:427–428.

AIDS; AIDS progression; HIV; HIV transmission; human leukocyte antigen; MHC diversity; viral load

© 2010 Lippincott Williams & Wilkins, Inc.