Abstract Body

Viremic controllers can yield insights into HIV persistence, but they are underrepresented in reservoir dynamics studies. We combined single-genome sequencing (SGS), proviral quantification, phylogenetics and mathematical modeling to: i) reconstruct within-host pre-ART HIV evolutionary histories; ii) measure proviral burden, age and diversity on-ART and iii) estimate proviral half-lives in 4 viremic controllers.

Three participants broadly maintained pVL <~2000, while one eventually lost control, prior to initiating ART in chronic infection. We performed subgenomic HIV RNA SGS (nef) on a median of 12 longitudinal pre-ART plasma samples/participant. Two PBMC samples from a median of 1.1 and 1.9 years on ART were also analyzed: the first for reservoir quantification using the Intact Proviral DNA Assay (IPDA); and the second for proviral nef SGS. Proviral sequence ages were inferred using a phylogenetic approach that leverages within-host pre-ART HIV evolutionary rates; we then applied a published mathematical model of reservoir seeding and decay to infer host-specific proviral half-lives from these data.

We collected 356 unique plasma HIV RNA sequences (range 52-173/participant) and 206 intact, non-hypermutated, unique proviral sequences (12-118/participant). All within-host phylogenies exhibited molecular clock signal pre-ART (range 1.16×10-5–5.35×10-5 substitutions/base/day). Pre-ART pVL area under the curve correlated strongly with longitudinal pre-ART plasma HIV sequence diversity, total on-ART proviral burden and overall on-ART proviral diversity (all Spearman’s ?=1; p=0.08). Intact proviral %, quantified by IPDA, ranged from 10-94%, where the latter was observed in the individual who lost control prior to ART. For two participants, inferred proviral integration dates ranged from shortly following infection to cART initiation; for the other two, including the participant who lost control, proviruses largely dated well into chronic infection. For three of four participants, the best-fit proviral half-life estimates were <~1 year, suggesting relatively rapid proviral turnover pre-ART; the fourth's proviral pool was consistent with negligible decay following deposition.

Despite their viremic control, significant within-host pre-ART HIV evolution nevertheless gave rise to diverse within-host proviral pools with varying intact genome burden. HIV eradication strategies must overcome within- and between-host diversity in proviral landscape.