HIV-1 envelope diversity and sensitivity to broadly neutralizing antibodies across stages of acute HIV-1 infection

Abstract

We studied the relationship between viral diversity and susceptibility to broadly neutralizing antibodies (bNAbs) in longitudinal plasma and peripheral blood mononuclear cells from 89 people with HIV who initiated antiretroviral therapy (ART) during acute and early HIV-1 infection (AEHI). HIV-1 diversity and predicted bNAb susceptibility were comparable across AEHI. Diversity evolution was not observed during ART, suggesting (pro)viruses at initiation or during treatment may identify individuals with susceptible virus for bNAb interventional trials.

Antiretroviral therapy (ART) has dramatically improved outcomes for people with HIV (PWH) but in most cases requires lifelong daily adherence. Antibodies that target HIV-1 envelope (Env) glycoproteins to neutralize a wide variety of HIV strains are being investigated as potential long-acting agents for HIV treatment, prevention, and cure strategies [1–4]; two such broadly neutralizing antibodies (bNAbs), 10-1074 and 3BNC117, have demonstrated clinical antiviral effects [5–11]. However, resistance driven by env genetic diversity may limit bNAbs’ effectiveness. Although ART initiation during acute and early HIV-1 infection (AEHI) restricts viral diversity [12–15], a better understanding of the genetic composition of env during primary infection and conservation on long-term ART may help predict clinical outcomes of interventions involving bNAbs. Here, we characterize env diversity and viral susceptibility to bNAbs before and after 60 weeks of suppressive ART in PWH who started treatment during AEHI.

We performed longitudinal analysis on samples from individuals enrolled in the AIDS Clinical Trials Group A5354 study [Early ART to Limit Infection and Establishment of Reservoir (EARLIER)], who initiated ART during Fiebig stages I–V of HIV-1 infection [16]. The study was approved by ethics committees/institutional review boards at all participating institutions; participants provided written consent. At ART initiation, staging was Fiebig I–II (n = 12), Fiebig III–IV (n = 38), and Fiebig V (n = 39) (Supplementary Table 1). Plasma and peripheral blood mononuclear cells (PBMCs) were obtained at baseline (pre-ART) and 60 weeks after ART onset. The env gene from plasma HIV-1 RNA and PBMC HIV-1 DNA was sequenced using MiSeq next-generation sequencing (Illumina, Inc., San Diego, California, USA), generating average read lengths of ∼200 base pairs, and analyzed as previously described [17]. Genotypic env data from individuals with chronic infection (≥6 months after estimated infection date) were included (data on file, Gilead): 16 post-ART PBMC samples and 12 pre-ART plasma samples. Subtyping was based on nucleotide identity of the consensus env gene against curated subtype reference sequences from GenBank (www.ncbi.nlm.nih.gov/genbank/). BLASTn v2.6.0+ software was used, as previously described [18]. Env complexity was assessed at the intra-patient level by average pairwise distance analysis using a sliding window approach [18]. Sequences were analyzed for presence of multiposition HIV-1 Env amino-acid signatures associated with in-vitro phenotypic susceptibility to bNAbs 10-1074 or 3BNC117 [17]. Briefly, genotypic assessment of viral susceptibility to 10-1074 was based on the sequential presence of the N332 glycan, D325, H330, T63, T320, and L179 in Env, with positive-predictive values for sensitivity to 10-1047 (defined as IC₅₀ <1 μg/ml) of 87, 90, 92, 98, 99, and 100%, respectively. Susceptibility to 3BNC117 was based on the sequential presence of I201, F353, I108, A281, E102, and Y318 in Env, producing positive-predictive values of 78, 84, 86, 91, 92, and 93%, respectively. The base pair variability threshold was set to less than 1% to ensure signature presence in all viral quasi-species.

Eighty-nine PWH who initiated ART during AEHI were included. Infection types were subtypes B (n = 76 participants), C (n = 8), F1 (n = 2), and A, D, and G (n = 1 for each). HIV env diversity was significantly lower in AEHI participants versus individuals with chronic infection (average pairwise distance: pre-ART plasma, 0.2492 versus 0.6153, P < 0.001; pre-ART PBMC, 0.1899 versus 0.6257, P < 0.01; post-ART PBMC, 0.1443 versus 0.6257, P < 0.001; two-sided Mann–Whitney U test); a trend toward lower diversity was observed in Fiebig I–II versus III–V (P > 0.05). Diversity was not significantly changed following more than 1 year of ART in paired proviral sequences (P > 0.05, two-sided Wilcoxon matched-pairs signed rank test), signifying lack of viral evolution after ART initiation consistent with prior reports [19,20]. Consistent with expectations that early treatment favors bNAb recognition, a higher prevalence of each Env sensitivity signature for 10-1074 and 3BNC117 was observed in virus from participants initiating ART during AEHI versus chronic infection, although these differences did not reach significance (pre-ART plasma, P > 0.05; post-ART PBMC, P > 0.05; Fisher exact test). Notably, prevalence of Env sensitivity signatures for 10-1074 and 3BNC117 were largely concordant between intraparticipant plasma virus and PBMC provirus pairs across time points. Concordance scores for signatures across paired samples were calculated by dividing the number of concordant observations by the total number of observations per signature (Fig. 1). Sensitivity signatures for 10-1074 were largely concordant between plasma viruses and PBMC proviruses before ART onset, with median concordance score of 80% across 87 participants with paired samples (Fig. 1a). Median signature concordance for 10-1074 was 72% across 78 participants with paired pre-ART plasma virus and post-ART PBMC provirus samples (Fig. 1c), and 71% for 79 participants with paired pre-ART and post-ART PBMC provirus samples (Fig. 1e). Median concordance for 3BNC117 sensitivity signatures was 89% between paired pre-ART plasma virus and pre-ART PBMC provirus pairs (Fig. 1b), 83% between paired pre-ART plasma virus and post-ART PBMC provirus pairs (Fig. 1d), and 93% between paired PBMC proviruses before and after ART (Fig. 1f).

Fig. 1.

Concordance of broadly neutralizing antibody sensitivity signatures of paired sequenced samples.

Data are shown for 87 participants with paired pre-antiretroviral therapy (ART) plasma virus and peripheral blood mononuclear cell (PBMC) provirus based on signatures for (a) 10-1074 and (b) 3BNC117 sensitivity; for 78 participants with paired pre-ART plasma virus and post-ART PBMC provirus based on presence of HIV-1 envelope signatures for (c) 10-1074 and (d) 3BNC117 sensitivity; and for 79 participants with paired pre-ART and post-ART PBMC samples based on presence of HIV-1 envelope signatures for (e) 10-1074 sensitivity and (f) 3BNC117 sensitivity. Numbers above graphs represent percentage concordance (100 × number of participants with concordant signatures/total number with signatures).

Our longitudinal investigation of env diversity and predicted bNAb susceptibility in a large cohort of PWH treated during AEHI indicates that both remained constant in HIV-1 RNA and DNA before therapy and in HIV-1 DNA following more than 1 year of viral suppression.

Study limitations included the use of bulk sequencing, which precludes analysis of longitudinal changes in phylogenetic lineages. Additionally, sequencing of only the env gene leaves uncertainty as to whether the sampled proviruses are replication-competent or reflect the full viral reservoir diversity. Despite these limitations and to the best of our knowledge, this is the largest study to date to longitudinally examine env sequences in PWH treated during AEHI. The analyses were comprehensive in assessing longitudinal changes in env diversity and bNAb susceptibility in viruses from both plasma and PBMC compartments, and the findings are consistent with previous studies [12,21]. Based on the bNAb epitope complexity present in Env, genotypic prediction of bNAb susceptibility is challenging and not fully established.

In conclusion, our findings indicate that individuals with AEHI express Env signatures associated with predicted susceptibility to 10-1047 and 3BNC117 that were concordant in viruses and proviruses; these persisted during treatment, consistent with the low env diversity observed in these samples. These data suggest that in individuals with AEHI, Env signatures from plasma or PBMCs similarly identify participants with susceptible virus for enrollment in bNAb clinical studies.

Acknowledgements

J.A., T.A.C., J.J.E., and E.S.D. enrolled study participants. B.M., C.C., and J.W.M. conceptualized the study. B.M., C.C., J.W.M., R.G., and R.H. contributed to the design of the study. A.P., B.M., J.C., J.L., and L.A.V. contributed to data collection and analyses. B.M., L.S., L.A.V., and C.C. interpreted the data and wrote the manuscript with input from all authors. All authors have read, reviewed, and approved the final version of this manuscript.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, the United States Army, the Department of Defense, or the Henry M. Jackson Foundation for the Advancement of Military Medicine. The investigators have adhered to the policies for protection of human subjects as prescribed in AR 70-25.

Conflicts of interest

L.A.V., L.S., B.M., A.P., J.L., R.H., R.G., and C.C. are employees and stockholders of Gilead. T.A.C. receives research support from the United States Army and the National Institutes of Health. J.J.E. receives research support from Gilead and ViiV, and is a consultant for Gilead, Merck, and ViiV. E.S.D. receives research support from Gilead and ViiV, and is a consultant for Gilead, ViiV, Theratechnologies, and Merck. J.W.M. receives grant support from Gilead, is a consultant for Gilead and AlloVir, and owns share options in Galapagos and Infectious Disease Connect, and shares of MingMed unrelated to the current work. For the remaining authors, nothing was declared. The AIDS Clinical Trials Group A5354 study was supported by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health (grants UM1 AI068634, UM1 AI068636, UM1 AI068618, UM1 AI106701, and P30 AI027757). Editorial support was provided by Becky O’Connor, PhD, of Parexel, funded by Gilead Sciences, Inc. This study was funded by Gilead Sciences, Inc.