|Home | About | Journals | Submit | Contact Us | Français|
Human immunodeficiency virus type 1 (HIV-1)-specific immune responses during primary HIV-1 infection appear to play a critical role in determining the ultimate speed of disease progression, but little is known about the specificity of the initial HIV-1-specific CD8+ T-cell responses in individuals expressing protective HLA class I alleles. Here we compared HIV-1-specific T-cell responses between subjects expressing the protective allele HLA-B27 or -B57 and subjects expressing nonprotective HLA alleles using a cohort of over 290 subjects identified during primary HIV-1 infection. CD8+ T cells of individuals expressing HLA-B27 or -B57 targeted a defined region within HIV-1 p24 Gag (amino acids 240 to 272) early in infection, and responses against this region contributed over 35% to the total HIV-1-specific T-cell responses in these individuals. In contrast, this region was rarely recognized in individuals expressing HLA-B35, an HLA allele associated with rapid disease progression, or in subjects expressing neither HLA-B57/B27 nor HLA-B35 (P < 0.0001). The identification of this highly conserved region in p24 Gag targeted in primary infection specifically in individuals expressing HLA class I alleles associated with slower HIV-1 disease progression provides a rationale for vaccine design aimed at inducing responses to this region restricted by other, more common HLA class I alleles.
Acute human immunodeficiency virus type 1 (HIV-1) infection is marked by high titers of viral load accompanied by fever, lymphadenopathy, cutaneous rash (18), and a significant loss of CD4+ T cells, in particular in the gut-associated lymphoid tissue (26, 27). The resolution of these symptoms and the subsequent decline in viremia are associated with the emergence of HIV-1-specific CD8+ T cells (7, 22). The importance of CD8+ T-cell responses in controlling HIV-1 viremia is further supported by the successive selection of viral escape variants within targeted cytotoxic T-lymphocyte (CTL) epitopes (16). Furthermore, the human leukocyte antigen (HLA) class I haplotype that restricts virus-specific CD8+ T-cell epitopes has a significant predictive value for the rate of HIV-1 disease progression as well as the level of HIV-1 replication (9). In particular, subjects expressing HLA-B27 or -B57 experience a significantly slower progression to AIDS and constitute up to 50% of the subjects in established long-term nonprogressor cohorts (12), while subjects expressing HLA-B35 face a more rapid progression towards AIDS (14, 19). These differences in HIV-1 disease outcome depending on the HLA class I genotype indicate that HIV-1-specific CD8+ T-cell responses restricted by different HLA class I molecules are not equally effective but may differ in their specificity and their ability to control viral replication (16).
The initial virus-specific CD8+ T-cell response in acute HIV-1 infection is of low magnitude and is narrowly directed against a limited number of epitopes, but it subsequently broadens during prolonged antigen stimulation in the chronic phase of infection (5, 10, 28). However, despite the generally more broadly directed and vigorous immune response in chronic infection, control over viral replication wanes during later stages of infection. These findings further suggest that the quality and specificity of the initial virus-specific CD8+ T-cell response, rather than the magnitude of the response, may be associated with the initial control of viral replication (29) and that responses primed early in infection are crucial for determining the long-term control over viral replication.
Several cross-sectional studies in chronic HIV-1-infected individuals have shown that the CD8+ T-cell responses in subjects with low viral load and slow disease progression are preferentially directed against HIV-1 Gag (1, 17, 21, 25). However, these studies in chronic infection did not allow determination of whether Gag-specific responses emerged early in infection, contributing to the initial control of viral replication and the determination of a viral set point, or simply developed in chronic infection in the setting of immune control of HIV-1. We therefore investigated the specificity of the virus-specific T-cell response in primary infection in individuals expressing protective or nonprotective HLA class I alleles and demonstrate that a specific region within p24 Gag is highly targeted by HIV-1-specific CD8+ T cells in subjects expressing protective HLA class I molecules but not in subjects expressing nonprotective HLA class I molecules.
Fresh peripheral blood mononuclear cells (PBMCs) from 43 patients diagnosed with primary HIV-1 infection were screened comprehensively for HIV-1-specific CD8+ T-cell responses against the entire expressed HIV-1 proteome using a gamma interferon (IFN-γ) enzyme-linked immunospot (ELISPOT) assay. In addition, cryopreserved samples from a second large cohort of 293 subjects collected 8 weeks (±10 days) after the initial presentation with primary HIV-1 infection were screened for HIV-1-specific CD8+ T-cell responses using HLA-matched optimal peptides in an IFN-γ ELISPOT assay (4). Samples were collected from different sites of the AIEDRP network: Massachusetts General Hospital, Boston (n = 154); Laboratoire d'Immunologie, Centre de Recherche du CHUM, Montreal, Quebec, Canada (n = 85); National Centre in HIV Epidemiology and Clinical Research, Sydney, Australia (n = 50); and Antiviral Research Center, University of California, San Diego (n = 4). More than 95% of subjects in the cohorts were infected with HIV-1 clade B. The study was approved by the respective institutional review boards and was conducted according to the human experimentation guidelines of the Massachusetts General Hospital.
High- and intermediate-resolution HLA class I typing was performed by sequence-specific PCRs according to standard procedures. DNA was extracted from PBMCs using the Purgene DNA isolation kit for blood (Gentra Systems, Minneapolis, MN). The HLA class I allele distribution in the study population generally reflected the distribution in a typical Caucasoid population.
Freshly isolated PBMCs of each of the 43 subjects with primary HIV-1 infection were screened using 410 overlapping peptides (generally 18-mers), varying from 15 to 20 amino acids (aa) in length, which overlapped by 10 aa and spanned the entire clade B consensus sequence 2001, as described previously (13). In the additional 293 study subjects, previously described optimal peptides corresponding to HLA-matched epitopes were tested (4). Peptides were synthesized commercially (Research Genetics) or at the Massachusetts General Hospital Peptide Synthesis Core Facility.
PBMCs were plated in 96-well polyvinylidene difluoride-backed plates (MAIP S45; Milipore) that had been coated previously with 100 μl of an anti-IFNγ monoclonal antibody, 1-D1k (2 μg/ml; Mabtech), overnight at 4°C. Peptides were added directly to the wells at a final concentration of 14 μg/ml. Cells were added to the wells at 50,000 to 100,000 cells/well. The plates were incubated at 37°C in 5% CO2 overnight (14 to 16 h) and then processed as described previously (1). IFN-γ-producing cells were counted by direct visualization and are expressed as spot-forming cells (SFCs) per 106 cells. The number of specific IFN-γ-secreting T cells was calculated by subtracting the negative control value from the established SFC count. The negative controls were always <30 SFCs/106 input cells (median, 5; range, 0 to 25 SFCs/106 input cells). The positive control consisted of incubation of 100,000 PBMCs with phytohemagglutinin.
Statistical analysis and graphical presentation were done using SigmaPlot 5.0 (SPSS Inc., Chicago, IL) and GraphPad Prism. Results are given as the mean ± standard deviation (SD) or the median and range. Statistical analysis of significance (P values) was based on a one-way analysis of variance (ANOVA) test.
Control over viral replication in primary HIV-1 infection has been linked to the early emergence of HIV-1-specific CD8+ T-cell responses leading to a subsequent decline of HIV-1 viremia to a set point (7, 22). To characterize the areas of the HIV-1 proteome that are initially targeted by these few early and effective virus-specific responses, we identified 43 subjects during primary HIV-1 infection. Twenty-seven individuals were identified prior to complete HIV-1 seroconversion (negative or indeterminate anti-p24 enzyme-linked immunosorbent assay and less than three positive bands in the HIV-1 Western blot) with a median viral load of >750,000 copies/ml (range, 6,220 to 6 × 107 copies/ml) and a median CD4+ count of 453 cells/μl (range, 191 to 929 cells/μl). Sixteen individuals were identified within 6 months of acute infection with a median viral load of 28,150 copies/ml (range, 216 to 780,000 copies/ml) and CD4+ count of 480 cells/μl (range, 292 to 1,069 cells/μl). We comprehensively screened HIV-1-specific T-cell responses in these 43 individuals using a panel of 410 overlapping peptides spanning the entire proteome of the HIV-1 clade B 2001 consensus sequence in an IFN-γ ELISPOT assay (13).
As described previously (5, 10, 28), total HIV-1-specific T-cell responses in these individuals with primary infection were of lower magnitude (median of 132 SFCs/106 PBMCs; range, 56 to 2,360) and more narrowly directed against a very limited number of epitopes (median of 9 targeted epitopes; range, 2 to 38) than HIV-1-specific T-cell responses usually detected in chronic infection (1, 2, 13). The targeted epitopes were located evenly within the proteins encoded by the three large structural genes of HIV-1 gag, env, and pol, as well as by the smaller accessory nef gene (average numbers of recognized epitopes within the HIV-1 proteins: Gag, 2.9 ± 2.9; Pol, 2.8 ± 3.7; Env, 2.4 ± 2.2; and Nef, 1.7 ± 1.3). Responses directed against each of these proteins contributed between 14 and 25% to the total virus-specific T-cell response (Fig. (Fig.1A).1A). Taking into account the amino acid length of each of these four HIV-1 proteins that represented the major targets of virus-specific T-cell responses in primary HIV-1 infection (by dividing the magnitude of the average T-cell response to each protein by the number of amino acids of the respective protein), Nef was the most immunogenic component of HIV-1 (data not shown), as previously described (24). Taken together, these data demonstrate that early HIV-1-specific T-cell responses do not generally favor a certain region of the HIV-1 proteome, but that HIV-1 Nef is predominantly targeted if stratified by the amino acid length of each protein.
To further investigate whether the recognition of specific regions within HIV-1 by T cells during primary infection differed between individuals expressing protective or nonprotective HLA class I molecules, we separated study subjects into individuals expressing the two HLA class I alleles most strongly associated with protection from HIV-1 disease progression (HLA-B27 and HLA-B57); individuals expressing HLA-B35, an HLA class I allele that has been associated with more rapid disease progression; and individuals expressing neither of these alleles. Of the 43 subjects identified during primary HIV-1 infection, 8 subjects carried the protective HLA class I alleles HLA-B27 (n = 4) and HLA-B57 (n = 4), while 6 individuals expressed HLA-B35 (4 individuals with the subtype HLA-B3501 and 2 individuals with the subtype HLA-B3502). These frequencies generally reflect the frequencies of the respective HLA types in a Caucasoid population (HLA-B27, 4.9%; HLA-B35, 19%; and HLA-B57, 8.7%) (8).
In the eight individuals expressing HLA-B27 or -B57, the majority of HIV-1-specific T-cell responses were directed towards Gag early in primary HIV-1 infection (Fig. (Fig.1B),1B), and notably HIV-1 Gag-specific responses contributed significantly more to the total HIV-1-specific IFN-γ response in these individuals (44%; SD, ±13.1%) than in subjects expressing nonprotective alleles (21.5%; SD, ±19.8%; P < 0.01 [Fig. [Fig.1B]).1B]). Furthermore, Gag-specific T-cell responses contributed only negligibly to the total virus-specific response in individuals who expressed HLA-B35 (average of 5.7%; SD, ±6.7% [Fig. [Fig.1B])1B]) and were significantly lower than those in subjects expressing HLA-B57/B27 (P < 0.001) or expressing neither HLA-B57/B27 nor HLA-B35 (P < 0.01). Instead, HIV-1-specific T-cell responses in HLA-B35+ individuals were preferentially directed against HIV-1 Env (Fig. (Fig.1B).1B). Differences in the recognition of Nef, Pol, or the accessory proteins were not significantly different among the three groups.
When we analyzed the contribution of the individual Gag proteins (p17, p24, and p15 [p2-p7-p1-p6]) to the total HIV-1-specific T-cell response individually, responses to p24 Gag contributed significantly more (38%; SD, ±17.8%; P < 0.001) to the total response in HLA-B27/57+ individuals than in the individuals expressing nonprotective alleles (13.2% [SD, ±15.8] in subjects expressing neither HLA-B27/B57 nor HLA-B35 and 7.7% [SD, ±15.0] in subjects expressing HLA-B35 [Fig. [Fig.1C]).1C]). Although there was no statistically significant difference in the magnitude (P = 0.18) and breadth (P = 0.07) of the HIV-1-specific T-cell responses between the three groups, a trend towards lower breadth and magnitude was observed in subjects expressing HLA-B27/B57 (median number of recognized epitopes, 4.5; median magnitude, 1,251 SFCs/106 cells) compared to the other two groups (subjects expressing neither HLA-B27/B57, median number of recognized epitopes, 10; median magnitude 2,944 SFC/106 cells; subjects expressing HLA-B35, median number of recognized epitopes, 6; median magnitude, 2,239 SFCs/106 cells). These data, derived from individuals early in primary HIV-1 infection, show that T cells from subjects expressing HLA-B27 and HLA-B57, two alleles strongly associated with slower HIV-1 disease progression (9), preferentially target epitopes within p24 Gag.
To determine whether the identified Gag-specific responses were indeed CD8+ T-cell responses restricted by HLA-B27/B57, we further investigated the fine specificity of the Gag-specific T-cell responses on the single-epitope level in these individuals. Only a very limited number of overlapping peptides (OLPs) within p24 Gag were consistently targeted in the eight individuals expressing HLA-B27 and HLA-B57, including OLP-36 (PVGEIYKRWIILGLNKIV [Gag aa 257 to 274]), which was targeted by all four HLA-B27+ individuals, and OLP-33 (SDIAGTTSTLQEQIGWM [Gag aa 234 to 250]), which was targeted by all four HLA-B57+ individuals (Fig. (Fig.2).2). These two OLPs contained two previously described HLA-B27- and HLA-B57-restricted CD8+ T-cell epitopes, B27 (KRWIILGLNK; KK10) and B57 (TSTLQEQIGW; TW10). Figure Figure22 demonstrates the individual OLPs within Gag targeted in the eight study subjects, as well as the responses to the peptides corresponding to the optimal described epitopes contained within these OLPs when tested individually. Strikingly, the responses against these few immunodominant epitopes contributed 84% (range, 30 to 100%) to the total Gag-specific CD8+ T-cell responses and 36% (range, 14 to 64%) to the total HIV-1-specific CD8+ T-cell responses in these eight B57/B27+ subjects. Both epitopes KK10 and TW10 are located between aa 240 and 272 within the same region of p24 Gag, confined to the functionally important alpha helices 6 and 7. None of the 6 subjects expressing HLA-B35 and only 3 out of the 29 remaining subjects that expressed neither HLA-B57/B27 nor HLA-B35 recognized this specific region within p24 Gag. In these three individuals, responses to this area only contributed a median of 2.8% (range, 1 to 14%) to the total HIV-1-specific CD8+ T-cell response. None of these three subjects expressed HLA alleles with a previously reported association with slow disease progression (Ac-144, HLA-A2, -A2, -B14, -B44, -Cw5, and -Cw8; Ac-151, HLA-A34, -A24, -B1517, -B49, -Cw07, and -Cw07; and Ac-158, HLA-A1, -A68, -B7, -B8, -Cw7, and -Cw7). These data suggest that individuals expressing protective HLA class I alleles specifically target very few immunodominant epitopes within a very defined region of p24 Gag during primary HIV-1 infection.
To further evaluate this narrow targeting of a specific region in HIV-1 p24 Gag by CD8+ T-cell responses restricted by HLA-B27 and -B57 early in HIV-1 infection, we extended the analysis to a second, larger, multicenter cohort of 293 individuals identified during primary HIV-1 infection. HIV-1-specific CD8+ T-cell responses in these individuals were characterized using peptides corresponding to optimal CD8+ T-cell epitopes described for the respective HLA class I allotype of the study subject (4). The assessment of HIV-1-specific CD8+ T-cell responses was performed on frozen PBMC samples collected 8 weeks (± 10 days) following the initial presentation with primary infection (4). In this large cohort of 293 subjects, we identified 21 subjects expressing HLA-B57 and 13 subjects expressing HLA-B27. While 94% of the subjects expressing HLA-B57/B27 recognized described p24 Gag CD8+ T-cell epitopes, only 64% of non-B27/B57/B35 subjects recognized p24 Gag epitopes and only 9.8% of HLA-B35-expressing subjects recognized them. Among the 51 individuals expressing HLA-B35, only 5 recognized the epitope PPIPVGDIY (B-35-PY9) within p24, with no statistically significant difference between HLA-B35Px+ and HLA-B35Py+ individuals in this small data set (1 HLA-B35Px and 3 HLA-35Py; one subtype was not available).
In the HLA-B57+ individuals, the epitope TW10 in p24 Gag again represented the immunodominant HIV-1-specific CD8+ T-cell response and was targeted in 86% of individuals, followed by IVLPEKDSW in the reverse transcriptase (IW9) in 61.9% and KAFSPEVIPMF in p24 (KF11) in 52.4%, while other HLA-B57-restricted epitopes were less frequently targeted (Fig. (Fig.3A).3A). In HLA-B27+ subjects, the epitope KK10 (p24 Gag) was recognized in 92% of subjects, while other HLA-B27-restricted epitopes were rarely recognized (Fig. (Fig.3A).3A). These data reconfirmed in a second large cohort of 293 study subjects that the immunodominant epitopes targeted in individuals expressing the protective HLA class I alleles B27 and B57 are located consistently within a specific area of p24 Gag (aa 240 to 272, containing HLA-B57-TW10 and HLA-B27-KK10).
We next investigated in this cohort of 293 study subjects how much each targeted HIV-1-specific CD8+ T-cell epitope contributed to the total virus-specific CD8+ T-cell response of the HLA-B57+ and HLA-B27+ individuals during primary HIV-1 infection. The tested HIV-1-specific CD8+ T-cell epitopes described within HIV-1 Gag contributed 44% in subjects with HLA-B57 and 41% in subjects with HLA-B27 to the total HIV-1-specific immune response. The HLA-B57-restricted response to the epitope TW10 alone contributed already 31% (SD, ±33.3%) to the total virus-specific response in B57+ subjects, and the HLA-B27-restricted response to the epitope KK10 contributed 40% (SD, ±20.5%) to the total response in B27+ subjects. Other HLA-B57-restricted responses or responses restricted by other HLA class I alleles in the HLA-B57+ individuals (data not shown) had a significantly lower contribution to the total HIV-1-specific CD8+ T-cell response (Fig. (Fig.3B).3B). Similarly, in HLA-B27+ subjects, other HLA-B27-restricted responses outside p24 Gag contributed little to the total virus-specific CD8+ T-cell response (Fig. (Fig.3B).3B). In conclusion, the overall majority of the HIV-1-specific CD8+ T-cell response in subjects expressing HLA-B27 or HLA-B57 was geared towards p24 Gag and was mainly composed of two epitopes, KK10 and TW10, located within the same region of Gag (aa 240 to 272).
Why might CD8+ T-cell responses directed against this region of p24 Gag be more beneficial than responses directed against other regions of the virus? The region between aa 240 and aa 272 of p24 Gag is highly conserved (more than 98.3% [range, 48 to 100%] conservation per amino acid position according to the HIV-1 clade B sequences published in the Los Alamos Database), and previous studies have indicated that sequence variations within this region may incur a significant loss of viral fitness (23, 25). It has been shown that under the selection pressure of the CD8+ T-cell response in HLA-B57+ individuals, the epitope TW10 mutates after the acute phase of the infection in 74% of the subjects by a Thr→Asn amino acid substitution at residue 242 (23, 25). Furthermore, the CTL escape mutations within the TW10 epitope revert quickly back to wild type when transmitted to an HLA-B57-negative individual (23). This reversion of the escape mutation back to wild type is most likely due to a fitness cost associated with a loss in the capability to stabilize helices 6 and 7 within p24 in the presence of the mutation (25). In contrast, escape mutations from the HLA-B27-restricted epitope KK10 occur only in later stages of the infection and after the prior accumulation of compensatory mutations (20). Furthermore, CTL escape mutations within the KK10 epitope have been associated with HIV-1 disease progression (6, 11, 15). Overall, these data indicate that individuals expressing HLA-B57 and -B27 are capable of targeting this vulnerable “Achilles’ heel” of HIV-1 p24 Gag early in acute infection, allowing for the rapid control of viral replication and the preservation of an effective antiviral immune response (3). Recent studies also suggest that Gag epitopes are the earliest presented targets on infected cells, even prior to the reverse transcription and integration of the virus, as relative large quantities of Gag protein are contained in the capsid of the incoming virus, while the intracellular processing of Env epitopes follows the de novo synthesis of Env protein (30).
In conclusion, we have demonstrated that the majority of HIV-1-specific CD8+ T-cell responses in individuals expressing the protective HLA class I alleles B27 and B57 are directed against a highly conserved region within p24 Gag that is rarely targeted in other individuals during primary HIV-1 infection. Interestingly however, several potential epitopes restricted by common HLA class I alleles such as HLA-A2, HLA-A3, HLA-B7, and HLA-B8, as well as conserved CD4+ T-cell epitopes (31), are contained within this area of p24 Gag studied here (http://www.hiv.lanl.gov/content/immunology/maps/ctl/p24.html) but not the immunodominant HLA-A2-restricted p17 Gag epitope SLYNTVATL (A2-SL9). While targeting of these p24 Gag HIV-1 epitopes by CD8+ T cells is rarely observed early in natural HIV-1 infection (less than 10% of the non HLA-B27/B57 individuals targeted this region), these epitopes can be targeted in chronic HIV-1 infection (5, 13). Future studies will need to evaluate the possibility of overcoming immunodominance patterns emerging in early natural infection by prior vaccinations, with the aim of targeting vulnerable areas of HIV-1 with CD8+ T-cell responses restricted by common HLA class I alleles (3).
We thank all study subjects for their participation. We thank in particular Robert Finlayson (Taylor Square Clinic, Darlinghurst, Sydney, Australia), Robert MacFarland (407 Doctors, Darlinghurst, Sydney, Australia), Cassy Workman, (AIDS Treatment Initiative, Darlinghurst, Sydney, Australia), and Mark Bloch (Holdsworth House General Practice, Darlinghurst, Sydney, Australia) for enrolling study participants from the respective Australian sites.
This study was supported by the National Institutes of Health (RO1 AI50429; Acute Infection Early Disease Research Network U01 AI052403). In addition, A.K. is supported by a Program grant and Practitioner fellowship from the Australian National Health and Medical Research Council (NHMRC). The NHMRC is supported by the Commonwealth Department of Health and Ageing, Australia. The work of S.L. was supported by National Institutes of Health grant AI43638. J.-P.R. is a physician-scientist supported by Fonds de Recherche en Sauté du Quebec (FRSQ). We thank all the members of the Montreal HIV Infection Study receiving financial support from FRSQ.
Published ahead of print on 9 May 2007.