HIV-1 infections with multiple founders associate with the development of neutralization breadth

Abstract

Eliciting broadly neutralizing antibodies (bnAbs) is a cornerstone of HIV-1 vaccine strategies. Comparing HIV-1 envelope (env) sequences from the first weeks of infection to the breadth of antibody responses observed several years after infection can help define viral features critical to vaccine design. We investigated the relationship between HIV-1 env genetics and the development of neutralization breadth in 70 individuals enrolled in a prospective acute HIV-1 cohort. Half of the individuals who developed bnAbs were infected with multiple HIV-1 founder variants, whereas all individuals with limited neutralization breadth had been infected with single HIV-1 founders. Accordingly, at HIV-1 diagnosis, env diversity was significantly higher in participants who later developed bnAbs compared to those with limited breadth (p = 0.012). This association between founder multiplicity and the subsequent development of neutralization breadth was also observed in 56 placebo recipients in the RV144 vaccine efficacy trial. In addition, we found no evidence that neutralization breath was heritable when analyzing env sequences from the 126 participants. These results demonstrate that the presence of slightly different HIV-1 variants in acute infection could promote the induction of bnAbs, suggesting a novel vaccine strategy, whereby an initial immunization with a cocktail of minimally distant antigens would be able to initiate bnAb development towards breadth.

Fig 1. No evidence of heritability of neutralization breadth.

Distribution and heritability of peak neutralization breadth across individuals. Non-broad neutralizers (indicated in blue) are defined as individuals who neutralized <35% of a 34-virus panel three to four years post-diagnosis and broad neutralizers (indicated in red) as individuals who neutralized >70% of the panel. Individuals sampled less than two years post-infection (intermediate neutralizers) are also shown (indicated in lavender). (A) Boxplot of peak neutralization breadth across HIV-1 subtypes and recombinants in the RV217 cohort. The number of individuals in each group is shown in parenthesis. (B) Histogram of neutralization breadth measured at peak neutralization breadth (RV217, n = 70) and three years after infection (RV217, n = 32; RV144, n = 39); dots for each participant are shown alongside each histogram. (C) Ancestral reconstruction of neutralization breadth across a phylogeny reconstructed from consensus sequences from the earliest post-diagnosis samples in 70 RV217 individuals (filled circles) and 56 RV144 placebo participants (open circles). The color spectrum shows the reconstructed ancestral estimates for neutralization breadth. (D) Density plot of the H² (heritability) estimates from three conditioned runs (chains) of a phylogenetic Ornstein-Uhlenbeck mixed model performed on the phylogeny in (C). The solid line shows the mean of the posterior estimates of H².

Fig 2. Association between infections with multiple founders and neutralization breadth.

Comparisons of individuals who neutralized <35% of a 34-virus panel three to four years post-infection (non-broad neutralizers) and >70% of the panel (broad neutralizers) and had single (grey fill) or multiple (pink fill) founder infections. Individuals with neutralization estimates sampled less than two years post-infection are categorized as intermediate neutralizers (35% < neutralization breadth < 70%). Percentage of (A) RV217 participants and (B) RV144 placebo recipients within each breadth category who had infections with single (grey) and multiple (pink) founders. (C) ROC graph of the performance of founder multiplicity (pink line) for predicting neutralization breadth in RV217 participants (solid pink line) and RV144 placebo recipients (dashed pink line). The null expectation is shown as a solid black line. AUC estimates are shown. (D) Neutralization breadth in RV217 participants with single (grey) or multiple (pink) founder infections sampled at one year, two years, and three years post-diagnosis, and at peak breadth. Participants with superinfections are shown in darker shades. (E) Neutralization breadth in RV144 placebo recipients sampled at one year and three years post-diagnosis. (D,E) The number of individuals in each group is shown in parenthesis and p-values for pairwise comparisons are shown above each pair. (A,C,D) Parenthetical numbers indicate values after removing individuals identified as superinfected.

Fig 3. Higher diversity in infections with multiple founders in early infection.

Comparisons of sequence diversity as a function of neutralization breadth and founder multiplicity in RV217 participants. Non-broad neutralizers (indicated in blue) are defined as individuals who neutralized <35% of a 34-virus panel three to four years post-diagnosis and broad neutralizers (indicated in red) as individuals who neutralized >70% of the panel. Individuals sampled less than two years post-infection (intermediate neutralizers) are also shown (indicated in lavender). Individuals with single founder infections are shown in grey and with multiple founder infections in pink. Median pairwise distance at one month, six months, and three years post-diagnosis in (A) non-broad, intermediate, and broad neutralizers and in (B) participants with single or multiple founder infections. The number of individuals in each group is shown in parenthesis and p-values for pairwise comparisons are shown above each pair. (C,D) Maximum sequence divergence over time from each participant’s consensus at the earliest sampling day for (C) non-broad neutralizers and (D) broad neutralizers. Circles indicate participants with single (grey) or multiple (pink) founder infections and lines indicate whether the individual achieved broad neutralization (red) or not (blue). A low-transparency representation of (D) is shown in (C) and vice versa. (A,B) P-values calculated after removing individuals identified as superinfected are shown in parentheses.

Fig 4. Multiple founders boosted diversification in broad neutralizers.

Comparisons of polymorphic sites in RV217 participants who neutralized <35% (indicated in blue) or >70% (indicated in red) of a 34-virus panel three to four years post-diagnosis. Histogram of polymorphic sites across (A) non-broad neutralizers or (B) broad neutralizers. Boxplots of the number of polymorphic sites in non-broad (blue) and broad (red) neutralizers at one month, six months, and three years post-diagnosis, shown separately for (C) all Env sites, surface sites, CD4bs, V1-V2, V3, and MPER Ab contact sites and for (D) VRC01, 35O22, 10–1074, VRC26.25, and 10E8 epitopes. Points are shown separately for individuals infected with a single founder (grey) versus multiple founders (pink). (D) Asterisks denote significant pairwise comparisons (p ≤ 0.05) of (C) Mann-Whitney U tests and (D) Fisher’s exact test between broad and non-broad neutralizers (black) and single- versus multiple-founder broad neutralizers (pink).