Abstract Body

Background: The RV144 clinical trial was shown to have an estimated vaccine efficacy of 31%. IgG antibodies (Abs) to V2 significantly inversely correlated with infection risk. Recently it was shown that Abs to V3 also inversely correlated with infection risk when vaccinees had low levels of Env-specific plasma IgA, and ADCC and neutralizing Abs. In addition, sieve analysis of breakthrough viruses identified significant differences in vaccine- and placebo-recipients in the proportion of certain residues present at positions 307 and 317 in V3. Studies were therefore performed to determine the fine specificity of V3 Abs in RV144 plasma and their possible role in reducing HIV infection. Methodology: Plasma specimens drawn two weeks after the last immunization from 40 vaccinees and 20 placebo recipients were tested in ELISA at a 1:100 dilution for reactivity vs. biotinylated cyclic V3 peptides (cV3s). Results: Of 8 cyclic V3 peptides studied, reactivity was strongest vs. clade B cV3s. The next strongest responses were with clade A, AG and C cV3s. The weakest responses were seen vs. clade AE cV3s (92TH023 & A244). The relative strengths of responses were consistent with analyses of the distance between the V3 sequences used in the assays. Plasma from placebo recipients were non-reactive. Seven variants of BaL and A244 cV3s with substitutions for I307 and F317 were also tested. Vaccinees’ plasma were most reactive with cV3 peptides containing I307 (which was present in the vaccine), and much less reactive with substitutions (V, T or M) that were more frequent in vaccinees’ breakthrough viruses than in those from placebo recipients. In contrast, plasma were equally reactive with wildtype A244 cV3 containing F317 and with A244 cV3 containing L317 despite the fact that vaccinee breakthrough viruses more frequently contained F317 than those from placebo recipients. Conclusions: RV144-induced V3 Abs were cross-reactive with cyclic V3 peptides from clades A, AG, B, C, and AE. Data are consistent with V3 Abs being induced primarily by the clade B gp120 boost and suggest that Abs that target I307 provided immune pressure. In contrast, since vaccinees’ plasma did not distinguish between A244 cV3s containing F317 or L317, and since the proportion of F317 was increased in vaccinees’ breatkthough viruses, it is possible that the presence of F317 confers an advantage to the virus, possibly by obscuring the epitope in the N-terminal V3 beta-strand that includes I307.