An additional three peptides—one each in ENV, POL, and VPR—elicited positive responses in Mali only. The 27 epitopes chosen in 2009 were also assessed in ELISpot assays with five HIV-positive donors who were confirmed to be HLA-A2 negative. Four of the five donors (80%) had no positive IFNγ responses to any of the 27 peptides tested;

one donor responded to only one of 27 (3.7%) peptides tested, demonstrating HLA-A2 specificity of the peptides selected for our present study. For the cohorts of chronically HIV-1-infected subjects from both the Miriam Hospital and the clinic in Bamako, Mali, there was no clear association between viral load, CD4 T-cell count, or years of known HIV infection with responses to HLA-A2 Luminespib epitopes. In addition, no clear association was found between having multiple A2 alleles and the number of epitopes that elicited a detectable IFNγ ELISpot result for a given donor. It is worth mTOR inhibitor noting that, in general, the subjects from Mali had an impressive number of epitope responses compared to the Providence subjects (Table 3a–c). One patient in this group responded to 25 epitopes, and four others with low viral loads responded to a mean of eleven epitopes. It is possible that this is

due to the fact that these subjects were recruited for the study less than a year after they had been identified as HIV-positive and/or due to the correlate that none of the study participants in Mali had yet received long-term antiretroviral therapy. Notably, the one Providence subject (H_0865) who was not receiving ART, yet had a low viral load, responded to eight HLA-A2 epitopes. The ELISpot analysis reconfirmed eleven epitopes that were published for HLA-A2 prior to the time of selection for this study (Table 1). Five of the epitopes that were initially identified and predicted by our 2002 informatics analysis as entirely novel HLA-A2 epitopes have subsequently been validated as A2-restricted epitopes by others (Table 1). These epitopes are ENV-1004 (TMGAASITL) [65], GAG-1012 (RMYSPVSIL) [66], POL-1006

Ergoloid (ALQDSGSEV) [67], POL-1247 (HLKTAVQMAV) [54], and VIF-1237 (DLADQLIHLY) [54]. Thus sixteen of the 38 epitopes have been validated by both our group and by other laboratories as HLA-A2 epitopes. In addition, assays confirmed five peptides that had been published epitopes prior to selection for inclusion in our study, although they were not published in the context of HLA-A2 (Table 1). Four of these epitopes were immunogenic in ELISpot assays with PBMCs from HLA-A2 subjects, and while only two of these epitopes were tested in in vitro binding assays, both bound to HLA-A2. The fifth epitope, POL-1016 (GLKKKKSVTV) [67], did not elicit positive IFNγ ELISpot responses in any subjects yet was shown to bind to HLA-A2 with low affinity, indicating that this may still be a relevant candidate for inclusion in a global vaccine (Table 1).