Previously we have described the V2-directed CAP256-VRC26 lineage that includes broadly neutralizing antibodies (bNAbs) that neutralize globally diverse strains of HIV. We also identified highly mutated "off-track" lineage members that share high sequence identity to broad members but lack breadth. Here， we defined the mutations that limit the breadth of these antibodies and the probability of their emergence. Mutants and chimeras between two pairs of closely related antibodies were generated: CAP256.04 and CAP256.25 (30% and 63% breadth， respectively) and CAP256.20 and CAP256.27 (2% and 59% breadth). Antibodies were tested against 14 heterologous HIV-1 viruses and select mutants to assess breadth and epito... More
Previously we have described the V2-directed CAP256-VRC26 lineage that includes broadly neutralizing antibodies (bNAbs) that neutralize globally diverse strains of HIV. We also identified highly mutated "off-track" lineage members that share high sequence identity to broad members but lack breadth. Here， we defined the mutations that limit the breadth of these antibodies and the probability of their emergence. Mutants and chimeras between two pairs of closely related antibodies were generated: CAP256.04 and CAP256.25 (30% and 63% breadth， respectively) and CAP256.20 and CAP256.27 (2% and 59% breadth). Antibodies were tested against 14 heterologous HIV-1 viruses and select mutants to assess breadth and epitope specificity. A single R100rA mutation in the third heavy chain complementarity-determining region (CDRH3) introduced breadth into CAP256.04， but all three CAP256.25 heavy chain CDRs were required for potency. In contrast， in the CAP256.20/27 chimeras， replacing only the CDRH3 of CAP256.20 with that of CAP256.27 completely recapitulated breadth and potency， likely through the introduction of three charge-reducing mutations. In this individual， the mutations that limited the breadth of the off-track antibodies were predicted to occur with a higher probability than those in the naturally paired bNAbs， suggesting a low barrier to the evolution of the off-track phenotype. Mapping studies to determine the viral immunotypes (or epitope variants) that selected off-track antibodies indicated that unlike broader lineage members， CAP256.20 preferentially neutralized viruses containing 169Q. This suggests that this globally rare immunotype， which was common in donor CAP256， drove the off-track phenotype. These data show that affinity maturation to counter globally rare viral immunotypes can drive antibodies within a broad lineage along multiple pathways towards strain-specificity. Defining developmental pathways towards and away from breadth may facilitate the selection of immunogens that elicit bNAbs and minimize off-track antibodies.
