Optimization of genetic distance threshold for inferring the CRF01_AE molecular network based on next-generation sequencing.

Introduction: HIV molecular network based on genetic distance (GD) has been extensively utilized. However, the GD threshold for the non-B subtype differs from that of subtype B. This study aimed to optimize the GD threshold for inferring the CRF01_AE molecular network. Methods: Next-generation sequencing data of partial CRF01_AE pol sequences were obtained for 59 samples from 12 transmission pairs enrolled from a high-risk cohort during 2009 and 2014. The paired GD was calculated using the Tamura-Nei 93 model to infer a GD threshold range for HIV molecular networks. Results: 2,019 CRF01_AE pol sequences and information on recent HIV infection (RHI) from newly diagnosed individuals in Shenyang from 2016 to 2019 were collected to construct molecular networks to assess the ability of the inferred GD thresholds to predict recent transmission events. When HIV transmission occurs within a span of 1-4 years, the mean paired GD between the sequences of the donor and recipient within the same transmission pair were as follow: 0.008, 0.011, 0.013, and 0.023 substitutions/site. Using these four GD thresholds, it was found that 98.9%, 96.0%, 88.2%, and 40.4% of all randomly paired GD values from 12 transmission pairs were correctly identified as originating from the same transmission pairs. In the real world, as the GD threshold increased from 0.001 to 0.02 substitutions/site, the proportion of RHI within the molecular network gradually increased from 16.6% to 92.3%. Meanwhile, the proportion of links with RHI gradually decreased from 87.0% to 48.2%. The two curves intersected at a GD of 0.008 substitutions/site. Discussion: A suitable range of GD thresholds, 0.008-0.013 substitutions/site, was identified to infer the CRF01_AE molecular transmission network and identify HIV transmission events that occurred within the past three years. This finding provides valuable data for selecting an appropriate GD thresholds in constructing molecular networks for non-B subtypes.

Low-level viremia episodes appear to affect the provirus composition of the circulating cellular HIV reservoir during antiretroviral therapy.

Low-level viremia (LLV) ranging from 50 to 1,000 copies/ml is common in most HIV-1-infected patients receiving antiretroviral therapy (ART). However, the source of LLV and the impact of LLV on the HIV-1 reservoir during ART remain uncertain. We hypothesized that LLV may arise from the HIV reservoir and its occurrence affect the composition of the reservoir after LLV episodes. Accordingly, we investigated the genetic linkage of sequences obtained from plasma at LLV and pre-ART time points and from peripheral blood mononuclear cells (PBMCs) at pre-ART, pre-LLV, LLV, and post-LLV time points. We found that LLV sequences were populated with a predominant viral quasispecies that accounted for 67.29%∼100% of all sequences. Two episodes of LLV in subject 1, spaced 6 months apart, appeared to have originated from the stochastic reactivation of latently HIV-1-infected cells. Moreover, 3.77% of pre-ART plasma sequences were identical to 67.29% of LLV-3 plasma sequences in subject 1, suggesting that LLV may have arisen from a subset of cells that were infected before ART was initiated. No direct evidence of sequence linkage was found between LLV viruses and circulating cellular reservoirs in all subjects. The reservoir size, diversity, and divergence of the PBMC DNA did not differ significantly between the pre- and post-LLV sampling points (P > 0.05), but the composition of viral reservoir quasispecies shifted markedly before and after LLV episodes. Indeed, subjects with LLV had a higher total PBMC DNA level, greater viral diversity, a lower proportion of variants with identical sequences detected at two or more time points, and a shorter variant duration during ART compared with subjects without LLV. Overall, our findings suggested that LLV viruses may stem from an unidentified source other than circulating cellular reservoirs. LLV episodes may introduce great complexity into the HIV reservoir, which brings challenges to the development of treatment strategies.