Evolution of HIV-1 envelope towards reduced neutralization sensitivity, as demonstrated by contemporary HIV-1 subtype B from the United States

Abstract

Subtype B HIV-1 has been the primary driver of the HIV-1 epidemic in the United States (U.S.) for over forty years and is also a prominent subtype in the Americas, Europe, Australia, the Middle East and North Africa. In this study, the neutralization profiles of contemporary subtype B Envs from the U.S. were assessed to characterize changes in neutralization sensitivities over time. We generated a panel of 30 contemporary pseudoviruses (PSVs) and demonstrated continued diversification of subtype B Env from the 1980s up to 2018. Neutralization sensitivities of the contemporary subtype B PSVs were characterized using 31 neutralizing antibodies (NAbs) and were compared with strains from earlier in the HIV-1 pandemic. A significant reduction in Env neutralization sensitivity was observed for 27 out of 31 NAbs for the contemporary as compared to earlier-decade subtype B PSVs. A decline in neutralization sensitivity was observed across all Env domains; the NAbs that were most potent early in the pandemic suffered the greatest decline in potency over time. A meta-analysis demonstrated this trend across multiple subtypes. As HIV-1 Env diversification continues, changes in Env antigenicity and neutralization sensitivity should continue to be evaluated to inform the development of improved vaccine and antibody products to prevent and treat HIV-1.

Fig 1. Phylogenetic relationship of contemporary U.S. subtype B Envs.

(A) Contemporary subtype B HIV-1 Envs were cloned from post residual HIV test samples from 30 U.S. military personnel and sequence diversity was evaluated. Fiebig stage of Env isolation is shown in blue (Fiebig I-IV) and red (Fiebig V-VI); high Env diversity is marked with an asterisk. (B) The phylogenetic relationship was evaluated between the contemporary U.S. subtype B Envs from this study (colored symbols) and subtype B Env sequences collected between 2011–2020 in the U.S. (gray lines) downloaded from the Los Alamos National Laboratory (LANL) HIV sequence database. (C) Divergence from the subtype B Env amino acid most recent common ancestor (MRCA) sequence and (D) pairwise diversity was determined for the consensus Envs from this study (n = 32, magenta), U.S. Envs from 2017–2018 (n = 26, pink), 2011–2016 (n = 239, light green), and prior to 2011 (n = 823, dark green) in the LANL database. Significant differences were determined between consensus Envs from this study and the LANL subset by a Mann Whitney U test. (E) Shannon entropy was estimated at each HXB2 site for the different env alignments, shown as matched colored lines. Pink highlights represent the Env variable loops V1 –V5, from left to right.

Fig 2. Neutralization sensitivities of contemporary subtype B HIV-1 Envs.

The neutralization sensitivity of the contemporary, CCR5-utilizing, subtype B pseudoviruses (PSVs) was determined using 31 neutralizing antibodies (NAbs) or reagents. NAbs are colored by specificity, including: V1V2 loop (orange), V3 loop (blue), CD4 binding site (CD4bs, green), membrane proximal external region (MPER, purple), glycan (CHO, yellow) and bridging region (BR, gray). Geometric mean (GM) (A) IC₅₀ and (B) IC₈₀ scores were calculated for each NAb against all PSVs and compared with the percent of PSVs neutralized by each NAb. The red dotted line at 50 indicates the negative cutoff value for this assay; the gray dotted line at 1 indicates the estimated NAb IC₈₀ value for treatment efficacy, as determined by the AMP studies. Neutralization potency of clinically relevant antibodies and antibody combinations was evaluated against early-stage, subtype B HIV-1 Envs. (C) NAbs were tested individually (shown as circles), or (D) in combination (shown as triangles; open if no bispecific antibody was present and gray if bispecific antibody was present). Combinations of NAbs included three NAbs that targeted different domains, as indicated by the x-axis title and the colored legend.

Fig 3. Changes in neutralization sensitivities of subtype B HIV-1 Envs over time.

Neutralization sensitivities of contemporary U.S. subtype B pseudoviruses (PSVs) (2017–18) were compared with U.S. subtype B PSVs from earlier decades. Each graph demonstrates the neutralization profile of one neutralizing antibody (NAb), as indicated in the title above each graph. Symbols are colored by the NAb specificity, with V1V2-specific NAbs in orange, V3-specific NAbs in blue, CD4bs NAbs in green, MPER NAbs in purple, glycan-specific (CHO) NAbs in yellow, and bridging region-specific NAbs in gray. Significant differences were determined by Mann Whitney U test; * = p<0.01, ** = p<0.001, *** = p<0.0001.

Fig 4. Changes in domain-specific neutralization sensitivities of subtype B HIV-1 Envs over time.

For each neutralizing antibody (NAb), the geometric mean (GM) IC₅₀ and percentage of pseudoviruses (PSVs) neutralized, was determined for all PSVs from each decade period, as shown in Fig 3. The trends for NAb potency (A-D) and breath (F-I) over time are shown by NAb specificity, including (A and F) V1V2 specific NAbs, (B and G) V3 NAbs, (C and H) CD4bs NAbs and (D and I) MPER NAbs. This neutralization analysis was repeated using 33 subtype B HIV+ plasma samples against PSVs from each decade period; the (E) GM ID₅₀ and (J) percentage of PSVs neutralized, was determined for all PSVs from each decade. Significant differences were determined by Wilcoxon matched-pairs signed rank test; * = p<0.01, ** = p<0.001, *** = p<0.0001.

Fig 5. Associations between initial neutralizing antibody potency and rate of declining neutralization potency over time.

For each neutralizing antibody (NAb), the geometric mean (GM) IC₅₀ was determined for all PSVs from an earlier and later decade period, as shown in the x-axis title. Fold change in NAb potency was determined between the earlier and later decade periods. A positive fold change in Env sensitivity to a given NAb indicates increasing sensitivity (earlier PSV GM IC₅₀ > later PSV GM IC₅₀), while a negative fold indicates decreasing sensitivity over time (earlier PSV GM IC₅₀ < later PSV GM IC₅₀). The vertical dashed line at 1 indicates no observed change. The correlation between the fold change in NAb potency over time and the potency of the NAb against PSVs from the earlier decade period are shown for (A) subtype B from the United States, (B) CRF01_AE from Thailand, (C) subtype A2 from East Africa, (D) subtype D from Uganda, and (E) subtype C from South Africa. Individual NAbs are shown as unique symbols colored by target specificity, including NAbs targeting the V1V2 (orange), V3 (blue), CD4bs (green) and MPER (purple) domains. Correlations were evaluated by Spearman correlation analysis. The summary of fold change in NAb potency over a ten-year period is shown for (F) all NAbs from all regions; data are stratified by initial NAb potency and significant differences were determined by Mann Whitney U test; ** = p<0.001, *** = p<0.0001.

Fig 6. Changes in subtype B Env neutralizing antibody contact site deviations over time and impact on Env neutralization sensitivity.

Subtype B Env sequences for pseudoviruses included in the neutralization analysis for each decade period were analyzed to determine the number of mismatched amino acids from the HXB2 reference strain that occurred in contact sites for neutralizing antibodies (NAbs) (A) PGDM1400, (B) 10–1074, (C) VRC01, and (D) 10E8. The relationship between the number of mismatched residues in the NAb contact site and Env NAb neutralization sensitivity were determined for (E) PGDM1400, (F) 10–1074, (G) VRC01, and (H) 10E8. Significant differences were determined by Mann Whitney U test; * = p<0.01, ** = p<0.001, *** = p<0.0001. Correlations were evaluated by Spearman correlation.