Initiation of HIV neutralizing B cell lineages with sequential envelope immunizations

Abstract

A strategy for HIV-1 vaccine development is to define envelope (Env) evolution of broadly neutralizing antibodies (bnAbs) in infection and to recreate those events by vaccination. Here, we report host tolerance mechanisms that limit the development of CD4-binding site (CD4bs), HCDR3-binder bnAbs via sequential HIV-1 Env vaccination. Vaccine-induced macaque CD4bs antibodies neutralize 7% of HIV-1 strains, recognize open Env trimers, and accumulate relatively modest somatic mutations. In naive CD4bs, unmutated common ancestor knock-in mice Env⁺B cell clones develop anergy and partial deletion at the transitional to mature B cell stage, but become Env^- upon receptor editing. In comparison with repetitive Env immunizations, sequential Env administration rescue anergic Env⁺ (non-edited) precursor B cells. Thus, stepwise immunization initiates CD4bs-bnAb responses, but immune tolerance mechanisms restrict their development, suggesting that sequential immunogen-based vaccine regimens will likely need to incorporate strategies to expand bnAb precursor pools.

Fig. 1

Immunogenicity of CH505 Envelopes in macaques. Rhesus macaques were immunized with CH505 transmitted-founder (TF) or natural variants of TF (weeks (w) 53, 78, and 100) envelope (Env) alone, or sequential combinations of TF and TF⁻ variants in AS01E or GLA-SE adjuvants. Immunization time points are indicated by an arrow. a, b Neutralization profiles of plasma post sixth immunization from macaques immunized with CH505 Envs in AS01E (a) or GLA-SE (b). Neutralization was assessed in TZM-bl cells once, except for titers <100 ID50 that are representative of two independent assays. c Post-sixth immunization plasma autologous neutralization titers. Each symbol represents the neutralization titer per animal, and the neutralization titers per group of animals are shown via box and whisker plots; box indicates median, and lower and upper quartile ranges, and whiskers indicate minimum and maximum titers. Neutralization positivity cutoff was 20 ID50 (dotted line). P = 0.07, exact Wilcoxon test; CH505.w4.3 neutralization titers for macaques immunized with CH505 TF alone or sequential Envs in AS01E. P = 0.3, Kruskal–Wallis test; CH505.w4.3 neutralization titers for macaques immunized with all CH505 Env regimens in GLA-SE. P = 0.8, exact Wilcoxon test; plasma neutralization titers for macaques immunized with CH505 TF Env in AS01E or GLA-SE. P = 0.4, exact Wilcoxon test; plasma neutralization titers from macaques immunized with sequential CH505 Envs in AS01E or GLA-SE. d, e Vaccination peaks and troughs for plasma reactivity with CH505 TF Env gp120; macaques were immunized with CH505 TF Env in AS01E or GLA-SE (d), and CH505 sequential Envs in AS01E or GLA-SE (e). Binding was performed in a single ELISA, each symbol represents plasma binding per animal, and the best fit line represents mean plasma binding within each group of animals. P < 0.05 is significant (exact Wilcoxon test); post sixth immunization, CH505 TF gp120 plasma-binding titers in macaques immunized with CH505 TF in AS01E (9.3 ± 1.0) or GLA-SE (8.9 ± 1.6) (P = 0.8) (d) and sequential CH505 Envs in AS01E (10.2 ± 0.9) or GLA-SE (10.0 ± 0.8) (P = 0.7) (e). f, g Plasma, post sixth immunization, competitive inhibition of soluble CD4 and CD4-binding site broadly neutralizing antibody (bnAb) CH106 in ELISA. Each symbol represents blocking levels per animal and the best fit line represents the mean blocking activity within each group of animals. Shown are representative plasma blocking activities of sCD4 and CH106 from macaques immunized with CH505 TF alone or sequential Env combinations (sequential, additive)

Fig. 2

Repertoire analysis of CH505 Envelope vaccine-induced antibodies. Memory B cells bearing candidate CD4-binding site receptors, or recombinant CD4-binding site antibodies, are termed CH505 differential binders. a Frequency of CH505 differential-binding memory B cells from 12 macaques immunized with CH505 TF envelope (Env) gp120 alone and 16 macaques immunized with sequential combinations of CH505 Env gp120s (sequential + additive). One million peripheral blood cells (PBMCs) from immunized macaques were phenotyped by FACS analysis for memory B cells that had CH505 Env differential binding. Shown is frequency (%) of memory B cells in PBMC that demonstrated CH505 Env differential binding. Number of animals with CH505 differential-binding memory B cells above background at 2 weeks post each immunization is listed above each graph. b Total Env-reactive antibodies isolated from blood memory collected 2 weeks post fourth and sixth immunizations in 16 macaques immunized with CH505 TF alone or sequential combinations of Env gp120, regardless of adjuvant; CH505 differential (CH505-D) and non-differential binders (CH505-ND). c, d Immunogenetics of antibodies isolated from macaques who received similar CH505 Envs, but different adjuvant. Antibody heavy variable (IGHV) gene mutation frequencies (c) and CDR3 lengths (d) were inferred by Cloanalyst software program and the results shown via box and whiskers plots; box indicates median, and lower and upper quartile ranges, and whiskers indicate minimum and maximum numbers evaluated. e IGHV mutation frequencies, and (f) IGHV CDR3 lengths, of total Env-reactive antibodies isolated from macaques regardless of CH505 Env vaccines or adjuvants. c–f Statistical analyses: (c, d) P > 0.05 (exact Wilcoxon test with false discovery rate (FDR) correction)—(i) IGHV mutation frequencies or CDR3 lengths of CH505 differentials or non-differentials isolated from animals receiving similar Env vaccines (TF or sequential Envs) in different adjuvants, and (ii) immunogenetics of CH505-D and CH505-ND antibodies isolated from macaques immunized with sequential or additive CH505 Envs in GLA-SE; (e, f) P = 0.008 (exact Wilcoxon test with FDR correction)—IGHV mutation frequency and CDR3 lengths for CH505-D and CH505-ND antibodies, regardless of Env vaccines and adjuvants, and (ii) for IGHV CDR3 lengths of CH505-D and non HIV-1-reactive antibodies

Fig. 3

Characteristics of macaque DH522 lineage antibodies. a Immunogenetics of clonally related DH522 lineage antibodies inferred by the Cloanalyst software program. DH522.1 and DH522.2 antibodies were isolated after 4 (21 weeks) and 6 (59 weeks) immunizations, respectively. b Neutralization profile of DH522.1 and DH522.2 monoclonal antibodies (mAbs) for viruses previously tested by plasma from CH505 Env-vaccinated macaques. Neutralization was performed in TZM-bl cells and titers reported as µg/mL IC50. c DH522.1 and DH522.2 competition for binding epitopes targeted by CD4-binding site bnAb CH106 and soluble CD4. Palivizumab was used as a negative control. d Binding profile of DH522 lineage antibodies to a panel of >9600 autoantigens via luminex as previously described. Shaded gray region; autoantigen binding to each mAb, and a polyreactivity profile for each mAb. e Footprints of DH522.2 and two related antibodies, on the surface of deglycosylated gp120 Env monomeric cores. Surface of deglycosylated gp120 in gray and antibody footprint in blue; the surface of deglycosylated HIV-1 YU2 gp120 core in the DH522.2 complex structure is underlain by a trace of the deglycosylated gp120 core in the compound-probed structure. (DH522.2) Structure of DH522.2—hydrophobic side chains of the Val-Leu-Phe motif at the tip of DH522.2 HCDR3 interact with a pocket on the surface of gp120 lined by loop B and adjacent to the CD4-binding site; (F105) the Val-Phe-Tyr hydrophobic tip of the F105 HCDR3 interacts with the same hydrophobic pocket on gp120 (PDB: 3HI1); (CH103) the CH103 footprint rests heavily on the CD4-binding loop and V5 (PDB: 4JAN). Crystal structure of the CH103 complex had only the gp120 outer domain; the inner domain (lighter gray) was modeled from the DH522.2-bound structure. f Superposition onto the BG505 SOSIP.664 gp140 trimer model of the deglycosylated gp120 cores in structures of complexes with DH522.2, F105, and CH103; view is along the gp140 threefold axis. g Footprints of CD4-mimic, CD4-binding site bnAbs (VRC01, CH235), and CD4 attachment inhibitor N-(4-bromophenyl)-N′-(2,2,6,6-tetramethylpiperidin-4-yl)ethanediamide (NBD-557). h–j Negative stain electron microscopy analysis of DH522.2 in complex with the fully glycosylated CH505 transmitted-founder (TF) SOSIP.664 trimer. h Electron microscopy (EM) reconstruction of a DH522.2-SOSIP.664 complex; b12-bound gp120 was docked into the EM map (PDB: 2NY7). i 2D class averages of the DH522.2-SOSIP.664 complex. Fabs indicated by the red arrows. j The Fourier shell correlation curve is shown along with the resolution and is determined using FSC = 0.5

Fig. 4

Immune tolerance in CH103 germ line knock-in mice. a–c Bone Marrow B cell development in naive CH103 unmutated common ancestor (UCA) double knock-in (dKI) mice, compared to wild type (B6), showing anergic phenotype and reduced frequencies of mature B cells associated with the second tolerance checkpoint. a Flow histograms indicating percentages in progenitor/precursor (Pro/pre), Immature (Imm), Transitional (Trans), and Mature (Mat) subsets. Data gated on live, total (CD19⁺ B220⁺) B cells. b Graphical summary of data shown in (a) for several mice. c–e Peripheral B cell development in naive CH103 UCA dKI mice, showing reduced total B cell numbers and frequencies of follicular mature B cells. c Representative flow histograms indicating percentages of total, live gated splenic B cells in transitional (Trans), marginal zone (MZ), and mature follicular (Mat) subsets. Graphical representation of total splenic B cell numbers (d) or mature to transitional ratios (e), as a measure of developmental arrest (transitional B cell accumulation) in periphery. Circles represent individual mice, and means are denoted by black bars. *P < 0.05; **P < 0.005, two-tailed Student’s t-test. n.s. not significant

Fig. 5

Analysis of receptor editing in CH103 germ line knock-in mice. a–c Env reactivity in mature B cell compartments of CH103 unmutated common ancestor (UCA) double knock-in (dKI) mice. a Representative pseudocolored dot plots showing typical patterns of reduced TF Env⁺ mature (follicular) B cells in naive CH103 UCA dKI mice. Shown as controls for background TF Env staining are WT (B6) mice. b Graphical summaries of percentages of differential binding in transitional and mature compartments in naive CH103 UCA dKI mice. Each dot represents an individual animal. ***P < 0.001, two-tailed Student’s t test. c Regression analysis of developmental blockade severity and residual mature B cell fraction retaining Env specificity. d Loss of CH103 KI light chain (LC) rescues splenic mature B cell development and reverses anergic phenotype in CH103 UCA V_HDJ_H ⁺⁺ (“heavy-chain (HC)-only”) KI mice. **P < 0.005, two-tailed Student’s t test. e Progressively decreased differential binding of splenic transitional and mature B cells (means + SEM) in CH103 UCA dKI (n = 6), heterozygous (het) dKI (n = 5), and V_HDJ_H ^+/+ KI (n = 4) mice. CH01 het dKI mice (n = 2), a model that undergoes no negative selection, and WT B6 mice (n = 4) are shown as positive and negative controls for Env binding, respectively. f, g Extensive and highly-restricted receptor editing of the CH103 _λ3-20 LC by “master editor” Vκ5-39, as revealed in CH103 UCA het dKI mice. f Receptor editing in CH103 UCA het dKI mice. Shown are pie charts of HC/LC usage and breakdown of Vκ family and Jκ usage, among LCs paired to the KI V_H4-59-bearing HC. Individual mature B cells from naive CH103 het dKI mice were obtained by single-cell sorting on the total (unselected for TF Env binding) mature follicular B cell repertoire, and V_HDJ_H/V_λJ_λ Ig gene pairs from single cells were recovered by RT-PCR for sequencing and immunogenetic analysis. g Polyreactivity/autoreactivity profiles of the CH103 UCA and Vκ5-39 editor M5808 monoclonal antibodies (mAbs) on human protoarrays. The >500-fold binding compared to control mAb, that is, the official autoreactivity “cutoff”, is indicated by dashed lines. Also shown are candidate autoantigens bound by the CH103 UCA mAb (circled in blue, in the red shaded box) and to which binding by M5808 is eliminated (green-shaded box)

Fig. 6

Signaling responsiveness of transitional and mature B cell subsets from CH103 germ line knock-in mice. a Representative flow histograms showing selectively reduced IgM mean fluorescence intensity (MFI), in differential-binding transmitted-founder (TF) Envelope (Env⁺) mature B cell fractions in CH103 unmutated common ancestor (UCA) double knock-in (dKI) mice, relative to Env⁻ (light-chain (LC)-edited) fractions. Data show gating for Env-specific bone marrow (BM) and splenic Env⁺ and Env⁻trans or mat B cell fractions, and is gated on live, total (B220⁺) B cells or live, total splenocytes, respectively. b Graphical summary (means + SEM) of data shown in (a) performed on eight mice per strain. c Ex vivo proximal signaling responses to BCR cross-linking in naive B cells from CH103 UCA dKI and control B6 mice, based on Ca⁺⁺ levels (Fluo-4 MFI) before/after anti-IgM stimulation. Shown are representative data (from a single CH013 UCA dKI mouse) of Ca⁺⁺ responses, to varying concentrations of anti-BCR, in mature B cells, either recirculating in BM (upper row) or in spleen follicles (lower row). Also shown for comparison is proximal signaling of mature B cells from control WT B6 mice. d Graphic representation of Ca⁺⁺-signaling responses (means ± SD) in both transitional and mature (splenic or BM B cells). Data are cumulative from four mice and two independently performed experiments. e CH103 UCA dKI transitional and mature B cell subset responses to priming immunogen TF Env. Data shown are from experiments performed in BM B cell fractions, using saturating amounts (100 nM) of tetramerized TF Env gp120 or anti-BCR (as a positive control for 100% of B cells signaling). *P < 0.05; **P < 0.005, ***P < 0.001; two-tailed Student’s t-test. n.s. not significant

Fig. 7

Early (TF + week 53) phase of Env “4-valent” sequential immunization rescues unedited (anergic) mature B cells in CH103 germ line knock-in mice. a Representative FACS dot plot histograms of IgG-switched splenic memory B cells in control or experimental groups denoted. Shown is transmitted-founder (TF) Env reactivity of splenic IgG⁺ memory B cells, either gated for wild-type (WT) TF Env (non-differential; upper row) or differential, that is, lineage-specific (∆371 TF− vs. WT TF+; lower row) gating strategies. Data are shown after second immunizations in CH103 unmutated common ancestor (UCA) heterozygous (het) double knock-in (dKI) (V_HDJ_H ^+/− × V_λJ_λ ^+/−) mice. Gating was performed on live, class-switched memory (B220⁺ CD19⁺ CD93⁻ CD38⁺ IgG⁺ IgM⁻) B cells. b Graphical representation of differential-binding IgG⁺ memory B cells per mouse spleen. Circles represent individual mice, and means are denoted by black bars. **P < 0.005, two-tailed Student’s t test. c Area under the curve (AUC) (mean + SEM) ELISA binding of serum IgG⁺ from TF Env-immunized mice, to WT or mutant ∆371 TF Env. Sera was collected 10 days after each immunization. Note the low-titer, differential serum IgGs early, but higher-titer, non-differential IgG Ab responses later in the sequential regimen. Also note that the last two boosts were carried out using the week 100 Env. d, e Distribution of CH103 prec. human V_H4-59 and V_λ3-1 segments vs. other, endogenous murine V_H and Vκ families utilized, in differential-sorted splenic IgG⁺ CD38⁺ memory B cells from immunized mice. Shown as comparisons are single IgG⁺ cells sorted from control-immunized mice. Middle circles; number of single-cell pairs sequenced. Relative sizes of LC pie charts indicate the % of LCs pairing to the KI HCs, as a proportion of total single-cell clone numbers amplified/sequenced. f Amino acid mutation status of memory IgG⁺ clones, pooled from all immunized CH103 UCA het dKI mice (n = 4), shown as all pairs retaining a residual heavy-chain (HC) and/or KI allele, or broken down into three groups: either as dKI clones, or as HC-KI-only or light-chain (LC)-KI-only-bearing clones