Potent and broad HIV-1 neutralization in fusion peptide-primed SHIV-infected macaques

Abstract

An antibody-based HIV-1 vaccine will require the induction of potent cross-reactive HIV-1-neutralizing responses. To demonstrate feasibility toward this goal, we combined vaccination targeting the fusion-peptide site of vulnerability with infection by simian-human immunodeficiency virus (SHIV). In four macaques with vaccine-induced neutralizing responses, SHIV infection boosted plasma neutralization to 45%-77% breadth (geometric mean 50% inhibitory dilution [ID₅₀] ∼100) on a 208-strain panel. Molecular dissection of these responses by antibody isolation and cryo-electron microscopy (cryo-EM) structure determination revealed 15 of 16 antibody lineages with cross-clade neutralization to be directed toward the fusion-peptide site of vulnerability. In each macaque, isolated antibodies from memory B cells recapitulated the plasma-neutralizing response, with fusion-peptide-binding antibodies reaching breadths of 40%-60% (50% inhibitory concentration [IC₅₀] < 50 μg/mL) and total lineage-concentrations estimates of 50-200 μg/mL. Longitudinal mapping indicated that these responses arose prior to SHIV infection. Collectively, these results provide in vivo molecular examples for one to a few B cell lineages affording potent, broadly neutralizing plasma responses.

Keywords: HIV-1; NHP; SHIV infection; antibody-lineage tracing; broadly neutralizing antibodies; cryo-EM; fusion peptide; molecular dissection; rhesus macaque; vaccine.

Figure 1.. SHIV-infection boosting yields broadly neutralizing plasma responses in vaccinated macaques in just eight weeks

(A) Immunization scheme. Two groups of five rhesus macaques were primed with FP8v1 coupled to a carrier protein (rTTHc or KLH) and boosted with Env trimers (BG505 DS-SOSIP FP8v1 and ConC FP8v2), followed by SHIV.CH1012 (Group 1) or SHIV.Ce1176 (Group 2) inoculation at week 124. Red dots represent FP8v1, and grey balls and cylinders represent rTTHc and KLH carriers, respectively, for FP8v1-rTTHc or FP8v1-KLH immunogens. (B) Fusion peptide-competition neutralization. The percentage reduction indicates plasma neutralizing activity at week 70 against BG505.N611Q, a fusion-peptide sensitive mutant, to be mediated by fusion peptide-competed response in 4 of 10 macaques. ID₅₀ (for plasma) indicates 50% inhibitory dilution; IC₅₀ (μg/ml; for antibody) indicates 50% inhibitory concentration. (C) Longitudinal plasma ELISA endpoint titers and ID₈₀-neutralization breadth. After SHIV inoculation, plasma ELISA endpoint titers against FP8v1 peptide (top) and BG505 trimer (middle) reached similar levels as before SHIV inoculation. Plasma neutralizing ID₈₀ titers (bottom) on the 19-strain panel dramatically increased in just 8 weeks after SHIV inoculation. (D) Dendrograms of ID₈₀ titers at wk132 (Post-SHIV wk8) on 19-strain panel. See also Figures S1, S2 and Tables S1, S2.

Figure 2.. SHIV infection boosts plasma neutralizing responses in the top macaques to ~100 ID₅₀ with 45-77% breadth on 208-strain panel, with neutralization signatures indicating fusion peptide-directed responses

(A) Neutralization ID₅₀ titers of representative plasma from each macaque on a cross-clade 208-strain panel. Geometric mean ID₅₀ titers calculated for ID₅₀ >20 values. (B) Dendrograms showing ID₅₀ titers for DJ85 wk132, GPZ6 wk148, TRNM wk132 and HERH wk156 plasma samples on 208-strain panel. (C) Neutralization fingerprinting analysis on 129 strains reveals fusion peptide-directed neutralizing responses to predominate in plasma. See also Table S2.

Figure 3.. Mutational selection at the fusion peptide site of vulnerability observed in a majority of fusion peptide-primed macaques after SHIV infection

(A) Locations of Env mutations displayed on a five-residue sliding window, with residue ranges shown for FP8v1 and Env trimer immunogens, and targeted VRC34.01 epitope highlighted in green for N88-proximal sites and in red for fusion peptide. Only Env sequences at week 16 after SHIV infection are shown. (B) SHIV sequences with mutations in selected regions, for 10 fusion peptide-primed macaques and 5 unvaccinated control macaques. The sequences of unvaccinated control group for “week 16” analysis are from week 18/20 and for “week 32” are from weeks 32/40 post-SHIV inoculation (see Figure S2 and STAR Methods). Welch’s t-test used to compare vaccinated and unvaccinated groups. See also Figures S1, S2 and Table S5.

Figure 4.. Single B cell-sorting and sequencing reveals a broadly neutralizing fusion peptide-directed antibody in each of the top macaques

(A) Identified antibodies and lineages. Antibodies were isolated by single B cell sorting at plasma neutralization peaks after SHIV-infection boosting. Positive B cells that bound at least two of three probes (FP8, BG505 trimer, or ConC trimer) were “positive” sorted for either FP8 and Env trimer probes, or for both BG505 and ConC trimer probes. Pie charts summarize identified antibodies (N represents the number of isolated antibody), with individually lineages colored by binding and neutralization characteristics: red if positive for FP8 binding; blue if negative for FP8 binding; and gray if non-neutralizing. (B) Binding and neutralization of representative antibodies from neutralizing lineages. ELISA endpoint titers of representative antibodies against FP8v1 and FP8v2 peptides as well as BG505 DS-SOSIP FP8v1 and ConC FP8v2 trimers revealed neutralizing antibodies to derive from either FP8-binding or FP8-non-binding antibodies. IC₅₀ neutralizing titers are provided for representative antibodies on the 19-strain panel. SHM levels of heavy and light chains and whether the lineage present at wk70 are also indicated. (C) Neutralization IC₅₀ (left) and IC₈₀ (right) titers on 208-strain panel shown for representative antibodies from each macaque as scatter plots (each dot represents the value of an antibody against a single pseudovirus strain), along with published fusion peptide-directed antibodies, DFPH-a.01 and VRC34.01. Geometric mean IC₅₀ and IC₈₀ calculated for IC₅₀ <50 μg/ml and IC₈₀ <50 μg/ml values, respectively. See also Figures S3-S8 and Tables S1-S5.

Figure 5.. Isolated monoclonal antibodies can recapitulate plasma neutralization in each of the top macaques

(A) Correlation between observed (experimental) and predicted ID₅₀ values as defined by linear regression of the contributions of each monoclonal antibody needed to recapitulate plasma neutralization. As some of the predicted monoclonal neutralization was absent (IC₅₀ >50 μg/ml), a linear scale was used. To account for outliers, DJ85 and HERH scales were expanded, and correlations recalculated, to test whether correlations were dependent on single large values. (B) Bar plot of antibody concentrations predicted to reproduce experimental neutralization shown in (A). Red bars depict concentration for fusion peptide-binding antibodies, and blue bars depict trimer-reactive antibodies that did not recognize the fusion peptide-N terminus; gray stripes on bars indicate strain- or clade-specific neutralizing antibodies. Antibodies shown on horizontal axes are representative monoclonals of different lineages from each macaque (e.g., on the far left plot, antibody “a” is DJ85-a.01). (C) Experimental replication of plasma neutralization as simulated by monoclonal antibody combinations shown in (B). ID₅₀ values are shown for monoclonal combination compared to plasma. (D) Correlations (r) between monoclonal combinations and experimental plasma neutralization (ID₅₀) on the 19-strain panel. ID₅₀ values of <20 are below the limit of assay detection (these have been graphed on the axes and used in correlations with ID₅₀ = 10). See also Figures S3-S7 and Tables S2, S7.

Figure 6.. Longitudinal phylogenetic analysis reveals each of the top macaques to have only 1-2 broadly neutralizing fusion peptide-binding lineages, each induced prior to SHIV infection

Paired heavy and light chain-maximum likelihood phylogenetic trees for monoclonal antibodies identified from each of the top macaque neutralizers. ELISA titers of each antibody to FP8v1, FP8v2, and Env trimers BG505 and ConC are depicted with colored “dots”. Lineages defined genetically are shown as red or blue bars, according to FP8 binding in Figure 4B, with CDR H3 lineage designation. Notably, genetic analysis indicated CDR H3-defined lineages DJ85 a-f and TRNM a-e to result from a single recombination that occurred prior to SHIV infection. See also Figures S3-S7 and Tables S1, S3, S4.

Figure 7.. Cryo-EM structures of isolated broad neutralizing antibodies provide residue-level detail for HIV-1 neutralization in fusion peptide-primed SHIV-boosted macaques

(A-D) The left panels show a composite of Env trimer (gray, with fusion peptide in red) recognized by antibodies, with heavy chains colored by lineage (CDR H3 designation), and light chains in black, with details of lineage, neutralization breadth on 19-strain panel, and resolution summarized (ND: not determined); right panels provide selected structural details. (A) In DJ85, an expanded lineage of multidonor antibodies directed against the fusion peptide-site of vulnerability provides neutralization breadth. (B) In GPZ6, two fusion peptide-directed antibodies provide breadth, with GPZ6-b.01 accommodating FP8v1 flexibility. FP8v1 flexibility of GPZ6-b.01 was analyzed by 100 ns MD simulations and RMSF calculation. Degree of red indicates residue flexibility. (C) In TRNM, the CD4-binding site-directed antibody TRNM-f.01 induces an open-occluded conformation, whereas all fusion-peptide-directed antibodies recognize the prefusion-closed conformation of Env trimer. (D) In HERH, antibody HERH-c.01 focuses on the recognition of glycan N88, similar to GPZ6-b.01. See also Data S1 and Figures S4-S8 and Tables S1-S3.