CD8+ T-cell responses vary widely in their abilities to control HIV replication and prevent progression to AIDS. Early studies demonstrated a temporal correlation between HIV-specific CD8+ T-cell responses and establishment of set point viremia, before seroconversion.1,2 Several associations have been made between specific HLA alleles and altered disease progression after HIV infection.3 Furthermore, CD8+ T cells are a major driving force in HIV evolution, indicative of the strong selection pressure imposed by these responses.4,5 However, several studies that have assessed the entire HIV-specific CD8+ T-cell response with interferon-γ (IFN-γ) enzyme-linked immunosorbent spot (ELISPOT) assays and overlapping peptide libraries found no correlation between clinical outcome and these responses.6,7 The discrepancy between these studies may be because certain functional aspects of CD8+ T-cell responses are more effective than others. CD8+ T-cell responses are variable in numerous respects, including the restricting HLA class I allele, the HIV-1 protein targeted, the avidity of T-cell receptor-peptide-HLA complexes, the consequences of epitope mutation, the ability to cross-recognize more viral variants, and the timing and dominance of responses in acute infection. This study focuses on comparing 2 divergent functional phenotypes of responding cells: secretion of IFN-γ and proliferative capacity.
T-cell memory can be categorized in terms of immediacy of effector function (effector memory, Tem), proliferation capacity (central memory, Tcm), and localization in lymph nodes (Tcm) or periphery (Tem). Murine models of some viral infections suggest that Tcm but not Tem CD8+ T cells are protective upon rechallenge.8 In humans, Tcm responses have been associated with improved outcome in HIV infection. Previous work by our group has shown that HIV-specific Tcm responses associate with resistance to HIV infection.9 In HLA-B*57 and B*27+ long-term nonprogressors (LTNPs), retention of proliferation was found to be greater than in progressing individuals.10,11 Polyfunctional CD8+ T-cell responses (including interleukin-2) have been shown to be protective in HIV infection,12 and these cells are preferentially induced by protective vaccinia vaccination.13 Proliferation responses lost after primary HIV infection can be rescued by IL-2 secreted by CD4+ T cells.14 More recently, proliferation and not cytokine production or cytotoxicity correlated inversely with viral load (VL) in a South African cohort.15 Therefore, although responses with a Tcm component (proliferation, IL-2) seem to be protective in HIV infection, most data describing the breadth and specificity of HIV-specific CD8+ T-cell responses have been based solely on IFN-γ secretion, which may not be a correlate of protection. We compared how IFN-γ and proliferation assays defined the breadth and specificity of HIV-specific CD8+ T-cell responses and found a substantial disconnect between these functional measures of HIV-specific CD8+ T cells.
HIV-infected Kenyan female sex workers (n = 23) were enrolled in the study. All subjects gave informed consent to participate, and the study was approved by the Institutional Review Boards at the Universities of Nairobi and Manitoba. All subjects were chronically infected with HIV (median >9.5, mean >8.2, range >1-20 years). antiretroviral status, VL, and CD4 counts are shown in Table 1. Several met criteria for long-term survivors (n = 6, HIV+ >10 years) and LTNPs (n = 2, HIV+ >10 years, CD4 >400). Although many of these patients are now on antiretroviral therapy through US President's Emergency Plan for AIDS Relief (PEPFAR) funding, only 1 was on antiretroviral therapy at the time of the assay (for 6 months). IFN-γ ELISPOT responses were measured as previously described,16 with positive responses defined as those with at least twice background control, >50 spot-forming units (SFU) per million peripheral blood mononuclear cells (PBMCs), and with a positive to Staphylococcus aureus enterotoxin B. Proliferation responses were defined by measuring dilution of carboxyfluorescein succinimidyl ester and CD8-antigen presenting cell staining, with positive responses defined as those at least twice background, >1% of CD8+ T cells being carboxyfluorescein succinimidyl ester-low, and with a positive response to S. aureus enterotoxin B.
Initial approaches to compare proliferation and IFN-γ responses demonstrated an increased sensitivity in detecting a proliferation response using 9mer compared with a 15-mer overlapping peptide library, as has been shown for IFN-γ ELISPOTs.17,18 Although only 20% (3/15) of subjects proliferated to overlapping peptide pools (not shown), 52% (12/23) responded to individual optimal-length peptides (P = 0.088, Fisher exact). Therefore, we chose to use known epitopes to compare IFN-γ and proliferation responses. The sequences and lengths of these optimal-length peptides (n = 18) are shown in Table 1 and include several described and novel Env epitopes selected on the basis of >5% population recognition in an epitope mapping study in the same population (L. R. McKinnon, BSc, 2008). The majority of these peptides have been shown to be CD8+ T-cell epitopes rather than CD4 epitopes by intracellular cytokine staining (data not shown).
Positive responses were detected in either assay in 16 of 23 subjects (70%), for a total of 24 ELISPOT and 41 proliferation responses. Discordance between IFN-γ and proliferation was common, with half of responding subjects being positive in one assay and not the other (n = 4 proliferation+, IFN-γ−; n = 4 proliferation−, IFN-γ+). Of the remaining subjects who were positive in both assays, different peptides were recognized in most instances. Only in subjects 1072 (1 epitope), 1425 (1 epitope), and 1740 (3 variants of an epitope) were both IFN-γ and proliferation responses observed. This demonstrates that in >60 HIV-specific responses, assay concordance was only observed in 5 instances (Fig. 1A). The breadth of proliferation responses was minimally higher than for IFN-γ (Fig. 1B, P = 0.23, paired t test). Differences in breadth within patients were common, as 9 of 16 responders had a difference in breadth of 2 or more epitopes between assays. These results suggest a substantial disconnect between the specificity and breadth of HIV-specific proliferating and IFN-γ producing CD8+ T cells.
One possible explanation for these assay-specific differences could be due to assay sensitivities or arbitrary assay cutoffs. However, mean and median background for ELISPOT were 21 and 18 SFU per million cells, whereas mean and median of peptide-specific responses were 369 and 484 SFU per million PBMCs. For proliferation, background values were 2.13% and 2.25% of CD8+ T cells, whereas the mean and median of peptide-specific responses were 5.87% and 10.38% of CD8+ T cells. There are some instances where responses in only one assay were weak, leaving the possibility that an accompanying response was present in the other assay but fell below the limit of detection. However, there are many examples of both IFN-γ and proliferation responses that are strong (ie, above the overall median) in the absence of an accompanying response in the other assay. These include 7 IFN-γ responses (29%) that were >500 SFU while proliferation negative and 14 proliferation responses (34%) that were >10% of CD8+ cells without an accompanying IFN-γ response. These data suggest that high magnitude of HIV-specific CD8+ T-cell responses in either assay often do not have an accompanying response in the other assay.
We next correlated clinical data (VL, CD4 count, and year infected) with the breadth and magnitude of the IFN-γ and proliferative responses observed in these subjects. There were no correlations between Env responses and years infected or VL (all P > 0.26, data not shown). Although no clinical associations were found with magnitude, we found that broader recognition of optimal Env epitopes in IFN-γ ELISPOT was inversely correlated with CD4 count (r2 = −0.66, P = 0.005, Fig. 1C). However, no associations between proliferation and CD4 count were observed. A similar observation was made in regard to progression; although Env-specific ELISPOT responses were associated with a faster time to CD4 <350 (2.9 vs 6.8 years, P = 0.05), proliferation responses were not (P = 0.3, Fig. 1D). One caveat of our study is that we measured CD8 responses to HIV Env, which has been associated with higher VLs in HIV infection.19 One explanation for this could be the poor correspondence of proliferation and IFN-γ responses to Env epitopes. Similar studies in proteins such as p24 are needed to determine whether proliferation and IFN-γ responses correspond better than in Env and which best correlates with HIV disease progression. Our data suggest that broad IFN-γ responses to Env may be detrimental, whereas proliferation responses may not be. As our study was cross sectional in chronically infected subjects, we were not able to determine whether broad IFN-γ responses are the cause or effect of disease progression.
Previous studies have shown that proliferation responses are lost in chronic HIV infection but are retained in LTNPs or those bearing HLA alleles that associate with protection.10,11,14 Our data suggest that a substantial proportion (>50%) of HLA-diverse subjects have proliferation responses regardless of disease status. The discrepancy between these studies may be use of cryopreserved versus fresh cells (we used fresh), the clinical status of our subjects (many long-term survivors and LTNPs), and the length of peptides used for stimulation (9mer). The latter of these may be the strongest argument, given that data from us and others suggest an increased sensitivity when detecting responses using shorter peptides. This has been corroborated by a recent study using HLA class I tetramers and optimal epitopes, which also detected a significant number of proliferation responses.15 The lack of correlation between proliferation and clinical parameters in our study may have been due to limited numbers and epitopes, all of which were in HIV-1 Env. Nevertheless, we provide evidence for robust HIV-specific CD8+ T-cell proliferation responses in chronic HIV infection, and this is one of the first reports to demonstrate a substantial difference in intrapatient specificity and breadth between IFN-γ and proliferative responses. For practical reasons, this study focused on epitopes previously defined in IFN-γ ELISPOT assays. Despite this bias, more proliferation responses than IFN-γ responses were observed, suggesting that epitope mapping of CD8 responses using alternate functional readouts such as proliferation may provide a more comprehensive assessment of HIV-specific CD8 responses.
Tcm and Tem generation after infection is thought to be influenced by the strength of stimulation that reaches a responding T cell.20 It is plausible that throughout infection, epitopes differ in their avidities and tendencies to induce responses characteristic of Tem or Tcm. Therefore, it may not be surprising that during chronic infection, HIV-specific CD8+ T-cell proliferation and short-term IFN-γ responses differ in epitope specificity. Previous studies defining CD8+ responses have been biased toward IFN-γ, an effector molecule that, although cheap and easy to measure, perhaps does not define a population of cells that are playing an active role in limiting HIV replication. The present study describes differences in specificity and breadth between IFN-γ and proliferation using epitopes defined in IFN-γ assays. Future studies should address whether this difference in specificity exists in early infection or whether it occurs as infection progresses. Given differences in specificity between these immunological readouts, further efforts to map novel proliferative epitopes may be warranted, and if practical, using optimal-length peptide libraries. Finally, these findings have important implications for evaluation of vaccine candidates designed to induce protective T-cell responses.
We thank the staff at the Majengo clinic, including Jane Kamene, Jane Njoki, Elizabeth Bwibo, and Edith Amatiwa for patient recruitment and sampling; Sandy Koesters for reviewing the manuscript; Kenyan AIDS Vaccine Initiative for use of flow cytometry facilities; Leslie Slaney, Ian MacLean, and support staff at Universities of Manitoba and Nairobi; and the study subjects who volunteered to make this study possible.
1. Borrow P, Lewicki H, Hahn BH, et al. Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection. J Virol
2. Koup RA, Safrit JT, Cao Y, et al. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Virol
3. Trachtenberg E, Korber B, Sollars C, et al. Advantage of rare HLA supertype in HIV disease progression. Nat Med
4. Borrow P, Lewicki H, Wei X, et al. Antiviral pressure exerted by HIV-1-specific cytotoxic T lymphocytes (CTLs) during primary infection demonstrated by rapid selection of CTL escape virus. Nat Med
5. Goulder PJ, Phillips RE, Colbert RA, et al. Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nat Med
6. Addo MM, Yu XG, Rathod A, et al. Comprehensive epitope analysis of human immunodeficiency virus type 1 (HIV-1)-specific T-cell responses directed against the entire expressed HIV-1 genome demonstrate broadly directed responses, but no correlation to viral load. J Virol
7. Betts MR, Ambrozak DR, Douek DC, et al. Analysis of total human immunodeficiency virus (HIV)-specific CD4(+) and CD8(+) T-cell responses: relationship to viral load in untreated HIV infection. J Virol
8. Zaph C, Uzonna J, Beverley SM, et al. Central memory T cells mediate long-term immunity to Leishmania major in the absence of persistent parasites. Nat Med
9. Alimonti JB, Koesters SA, Kimani J, et al. CD4+ T cell responses in HIV-exposed seronegative women are qualitatively distinct from those in HIV-infected women. J Infect Dis
10. Migueles SA, Laborico AC, Shupert WL, et al. HIV-specific CD8+ T cell proliferation is coupled to perforin expression and is maintained in nonprogressors. Nat Immunol
11. Horton H, Frank I, Baydo R, et al. Preservation of T cell proliferation restricted by protective HLA alleles is critical for immune control of HIV-1 infection. J Immunol
12. Betts MR, Nason MC, West SM, et al. HIV nonprogressors preferentially maintain highly functional HIV-specific CD8+ T cells. Blood
13. Precopio ML, Betts MR, Parrino J, et al. Immunization with vaccinia virus induces polyfunctional and phenotypically distinctive CD8(+) T cell responses. J Exp Med
14. Lichterfeld M, Kaufmann DE, Yu XG, et al. Loss of HIV-1-specific CD8+ T cell proliferation after acute HIV-1 infection and restoration by vaccine-induced HIV-1-specific CD4+ T cells. J Exp Med
15. Day CL, Kiepiela P, Leslie AJ, et al. Proliferative capacity of epitope-specific CD8 T cell responses is inversely related to viral load in chronic HIV-1 infection. J Virol
16. McKinnon LR, Ball TB, Wachihi C, et al. Epitope cross-reactivity frequently differs between central and effector memory HIV-specific CD8+ T cells. J Immunol
17. Beattie T, Kaul R, Rostron T, et al. Screening for HIV-specific T-cell responses using overlapping 15-mer peptide pools or optimized epitopes. AIDS
18. Draenert R, Brander C, Yu XG, et al. Impact of intrapeptide epitope location on CD8 T cell recognition: implications for design of overlapping peptide panels. AIDS
19. Kiepiela P, Ngumbela K, Thobakgale C, et al. CD8+ T-cell responses to different HIV proteins have discordant associations with viral load. Nat Med
20. Sallusto F, Geginat J, Lanzavecchia A. Central memory and effector memory T cell subsets: function, generation, and maintenance. Annu Rev Immunol