Background: HIV-1 infected individuals with certain HLA-B*35 allelic variants, B*3502/3503 (B*35-Px), have more rapid disease progression than those with the B*3501 (B*35-PY) variant. Our previous work showed that the overall magnitude of HIV-specific CD8⁺ T-cell responses restricted by all class I alleles does not differ between B*35-Px and B*35-PY positive individuals. However, we hypothesized that, as a single class I allele, B*35-PY is more effective than B*35-Px at eliciting CD8⁺ T-cell responses, and that this difference in CD8⁺ T-cell responses may not be detected by measuring total CD8⁺ T-cell responses.

Methods: To test our hypothesis, we constructed stable CHO cell lines expressing either B*3501, B*3502, or B*3503 molecules on their surface, and then used them as APC to present 26 peptide pools to T-cells in patients (pts) with matching B*35 genotypes. We used 15–20mer overlapping peptides covering the entire HIV-Gag, Nef, Tat, and Rev, as well as 6 known B35 restricted CTL epitopes from HIV proteins. The numbers of HIV-specific T-cells in PBMC were enumerated using an IFNγ ELISpot assay. Fourteen (14) chronically HIV-infected, HAART naïve, and B*35 positive individuals were included in the study, 10 of whom are B*35-PY and 4 are B*35-Px. The results were analyzed using T-test.

Results: We found that only 1 of 4 B*35-Px pts had a detectable CD8⁺ T-cell response, in contrast to 7 of 10 B*35-PY pts. On average, twice as many IFNγ spots were produced by peptide-stimulated PBMC from B*35-PY pts than by PBMC from B*35-Px pts (983 vs 1,956 per 10⁶ PBMC). In addition, 5 of 26 peptide pools stimulated significantly more IFNγ spots in the PBMC of B*35-PY pts than those from B*35-Px pts (p < 0.05).

Conclusions: HIV-specific CD8⁺ T-cell responses restricted by the B*35-PY allele appear to be broader and stronger than those restricted by the B*35-Px allele. Such differences may help explain why individuals with B*35-PY genotypes have a slower rate of HIV disease progression than those with B*35-Px genotypes.