Abstract Body

Background:

One of the factors contributing to HIV’s persistence is its rapid evolution. Understanding how the env gene evolves under antibody-mediated pressure is particularly important for efforts to use antibodies as therapies for HIV. In a previous clinical trial, infusion of the broadly-neutralizing antibody (bNAb) VRC01, which targets the CD4 receptor binding site, lowered plasma HIV viremia in a subset of participants, but was associated with the emergence of neutralization-resistant env variants. Here we sought to develop an analytical approach for identifying env mutations that confer escape from antibodies using single-genome sequence data from participants treated with VRC01.

Methods:

We applied single-genome amplification and sequencing (HT-SGS) using unique molecular identifiers (UMIs) and the Pacific Biosciences long-read platform, to measure HIV env allele frequency changes over time in longitudinal samples from people with chronic, untreated HIV infection who received one dose of VRC01. Eight participants were studied, and a total of 29,433 sequences were analyzed. We used two approaches to identify changes adaptive to VRC01 escape: (1) a codon-based dN/dS approach executed by HYPHY with FUBAR (2) a population genetics approach where Tajima’s D was used to identify positive selective sweeps along the env gene, and then a score for the selected allele favored in evolution (SAFE) was calculated.

Results:

In six of eight hosts the viral populations pre and post infusion were distinct. Specifically, changes in the VRC01 epitope were observed. In two individuals, both FUBAR and SAFE approaches identified the same positions as adaptive. In another four individuals FUBAR and SAFE pointed to different residues as adaptive. Examining manually regions of selective sweeps as identified by negative values of Tajima’s D, revealed complex evolutionary scenarios, and suggested the existence of several adaptive mutations in each host. Furthermore, epistasis between positions is likely, due to existence of low frequency combinations, of the adaptive alleles, that do not grow in frequency. In two individuals, subpopulation structure correlating with sequence signatures of cell tropism is associated with different escape mutations.

Conclusions:

These results suggest diverse adaptive pathways to bNAb escape among different individuals. While some mutations are easily identified as adaptive, multiple escape mutations, epistasis and tropism associated mutation are more challenging to identify and interpret.