Background: Individuals who are repeatedly exposed to HIV-1 but remain uninfected represent a unique cohort to study the potential mechanisms of resistance against HIV-1 infection. 

Methods: To investigate if specific cellular immunity contributes to protection, we enumerated T cells recognizing HIV-1 Env, Gag, Pol, and Nef epitopes in 16 exposed seronegatives (ES) using dendritic cells (DC) for antigen presentation and an IFN-g ELISPOT assay for detection. 

Results: All ES are healthy male homosexuals maintaining high-risk sexual activities with their known HIV-1-infected partner (100%) or multiple other partners (50%).  We detected T cells specific to at least 3 HIV-1 epitopes in 6 ES (38%).  This finding is in contrast to our previous studies in which only 1 of the 16 ES demonstrated HIV-1-specific T cells in an IFN-g ELISPOT assay using PBMC alone.  The majority of responses were directed against Pol epitope pools (33 out of 55 total responses), followed by Env (16/55) and Nef (6/55).  The magnitude of individual responses ranged from 100 spot-forming cells (SFC) to 4740 SFC per million responder cells (mean 1920; median 1860).  HIV-1 Gag-specific T cells were not seen.  For most assays, PBMCs were purified into CD8+ and CD8- T-cell subsets and responses were present both within the CD8+ (28/51) and the CD8- cells (23/51).  We also detected responses within the CD8+ but not the CD8- subset in 2 out of 5 low-risk HIV-1-seronegative individuals (mean SFCs: 310 and 1050).  When we stratified the risk status of the 16 ES by scoring the overall frequencies of anal receptive (AR) and anal insertive (AI) intercourse, 5 of the 6 responders (83%) fell into the highest risk category.  Among the nonresponders, only 3 of 10 volunteers (30%) fell into the highest risk category, and one of these tested homozygous for the CCR5 D32 mutation. 

Conclusions: These findings suggest that repeated exposure to HIV-1 by very high-risk sexual activity in homosexual men stimulates broad and often pronounced HIV-1-specific T cell immunity.  Alternatively, we cannot exclude the possibility that these represent true false-positive responses resulting from recognition of non HIV-1 cross-reactive epitopes.  Repeated studies in low-risk HIV-1 seronegative control subjects in addition to the ES will be important to distinguish between these possibilities.