Ultrapotent Broadly Neutralizing Human-llama Bispecific Antibodies against HIV-1

Abstract

Broadly neutralizing antibodies are proposed as therapeutic and prophylactic agents against HIV-1, but their potency and breadth are less than optimal. This study describes the immunization of a llama with the prefusion-stabilized HIV-1 envelope (Env) trimer, BG505 DS-SOSIP, and the identification and improvement of potent neutralizing nanobodies recognizing the CD4-binding site (CD4bs) of vulnerability. Two of the vaccine-elicited CD4bs-targeting nanobodies, G36 and R27, when engineered into a triple tandem format with llama IgG2a-hinge region and human IgG1-constant region (G36×3-IgG2a and R27×3-IgG2a), neutralized 96% of a multiclade 208-strain panel at geometric mean IC₈₀s of 0.314 and 0.033 µg mL^-1, respectively. Cryo-EM structures of these nanobodies in complex with Env trimer revealed the two nanobodies to neutralize HIV-1 by mimicking the recognition of the CD4 receptor. To enhance their neutralizing potency and breadth, nanobodies are linked to the light chain of the V2-apex-targeting broadly neutralizing antibody, CAP256V2LS. The resultant human-llama bispecific antibody CAP256L-R27×3LS exhibited ultrapotent neutralization and breadth exceeding other published HIV-1 broadly neutralizing antibodies, with pharmacokinetics determined in FcRn-Fc mice similar to the parent CAP256V2LS. Vaccine-elicited llama nanobodies, when combined with V2-apex broadly neutralizing antibodies, may therefore be able to fulfill anti-HIV-1 therapeutic and prophylactic clinical goals.

Keywords: HIV‐1; bNAb; bispecific antibodies; broadly neutralizing antibody; llama; neutralizing nanobodies; vaccination; vaccine.

Figure 1

BG505 DS‐SOSIP immunized llama develops broadly neutralizing serum responses. A) Immunization schema in a llama. The llama was immunized with BG505 DS‐SOSIP subcutaneously once every 21 days (except the last boost) as indicated above the arrow line, and serum was collected 10 days post each immunization starting from day 52 as indicated beneath the arrow line. B) Antibody binding response to HIV Env probes in the immunized llama as determined by ELISA. C) Llama sera neutralization response on a 14‐strain panel. Color shading represents potency as indicated on the right of the table. D) Dendrograms of the neutralization activity of day 188 and day 271 sera on a 60‐strain panel. Dendrograms display the diversity of tested viral strains, with branches colored according to neutralization potency (non‐neutralized branches shown in gray).

Figure 2

Identification of neutralizing nanobodies from BG505 DS‐SOSIP immunized llama. A) Nanobody phage library construction and screening. The four probes used for phage screening are: Env Trimer (BG505 DS‐SOSIP), Fusion peptide, Glycan base trimer (BG505 DS‐SOSIP.4mut_N502‐660), and RSC3. B) Summary of epitopes of 151 nanobodies from day 188 library. 30 nanobodies selected for small panel neutralization test are grouped into 4 categories: Non‐neutralizer (one line); Weak neutralizer (two lines); Moderate neutralizer (three lines); Broad neutralizer (four lines plus one red arrow). Control nanobody J3 is marked with four lines and one blue arrow. C) Phylogenetic tree of three selected nanobody lineages. 42 nanobodies from the three lineages were tested on a 10‐strain panel first, then the top 6 candidates were further tested on an additional 15‐strain panel. Neutralization breadth of the top 6 nanobodies on a 25‐strain panel is shown beneath the nanobody names. Scale bars indicate the distance of 16.67 nucleotides (nt) in each tree. D) 25‐strain neutralization of the top 6 nanobodies from Figure S3C (Supporting Information). The broadest (G36) and most potent (R27) nanobodies were selected for further analysis.

Figure 3

Immunization‐elicited nanobodies, G36 and R27, in nanobody x3‐IgG2a format, show broad and potent HIV‐1 neutralization. A) 25‐strain neutralization of nanobody x3‐IgG2a. B) 208‐strain panel neutralization of G36×3‐IgG2a and R27×3‐IgG2a. Dendrograms display the diversity of tested viral strains, with branches colored according to neutralization potency (non‐neutralized branches shown in gray). IC₅₀ shown is geometric mean. C) Comparison of neutralization breadth and potency for G36×3‐IgG2a and R27×3‐IgG2a with other human antibodies and vaccine‐elicited NHP antibodies on 208‐strain panel.

Figure 4

Cryo‐EM structures of nanobodies R27 and G36 in complex with HIV‐1 Env trimer reveal modes of recognition similar to J3. A) Cryo‐EM structure of nanobody R27 in complex with HIV‐1 BG505 DS‐SOSIP Env. Overall cryo‐EM density map and refined model are shown in two views with gp120 protomers colored green, cyan, and slate, respectively. The density and model of nanobody R27 is colored orange. The contour level of Cryo‐EM map is 9.5 σ. B) Cryo‐EM structure of VHH G36 in complex with HIV‐1 BG505 DS‐SOSIP Env. Overall cryo‐EM density map and refined model are shown in two views with gp120 protomers colored green, cyan, and slate, respectively. The density and model of G36 is colored magenta. The contour level of Cryo‐EM map is 9.6 σ. C) Epitopes of R27 and G36 on BG505 DS‐SOSIP. Epitopes of R27, G36, and J3 are shown in orange, magenta, and pink surfaces, respectively. R27 has a much smaller contact area on the neighboring protomer. D) Comparison of binding modes and angles. (Left) Structures of nanobodies R27, G36, and J3 are aligned with CD4 by the gp120 domain shown in green. R27, G36, and J3 are roughly in a similar position with N termini (labeled with “N”) in close proximity. (Right) The axes of R27, G36, and J3 are shown in orange, magenta, and pink rods. Axes of CD4 domain 1 and Fv domain of VRC01‐class antibody N6 are shown in yellow and olive rods for comparison. E) Detailed interactions between nanobodies and BG505 DS‐SOSIP. Residues that form hydrogen bonds and salt bridges are highlighted with sticks representation with bonds between atoms shown in gray dotted lines. Nanobodies and protomers of HIV Env are colored the same as in panels A and B. F) Alignment of nanobody sequences. Paratope residues are colored in orange and magenta, respectively. Residues interacting with neighboring protomer are colored in lighter shades. Residues interacting with both protomers are underlined. G) Nanobody mimicry of CD4 Phe43 interaction with gp120. G36, like J3, inserts Tyr99 into the “Phe43 pocket” on gp120, whereas R27 has an Ala at this position.

Figure 5

Ultra‐potent HIV‐1 bispecific antibodies from attaching nanobodies to the light chain of V2‐apex‐directed antibody CAP256V2LS. A) Schematic of CAP256L‐nanobody chimeras. B) 38‐strain neutralization of nanobodies. C) 80‐strain panel neutralization. Dendrograms display the diversity of tested viral strains, with branches colored according to neutralization potency (non‐neutralized branches shown in gray). D) Comparison of neutralization breadth and potency for R27 and G36 constructs with other potent antibodies on the 208‐strain panel. Data for CAP256L‐G36×3LS and CAP256L‐R27×3LS are estimated from 80‐strain data in panel C. Bispecific and trispecific antibodies are shown as stars.

Figure 6

Cryo‐EM structure of CAP256L‐R27LS Fab in complex with HIV‐1 Env reveals CAP256 and R27 to bind prefusion‐closed trimer at V2‐apex and CD4bs simultaneously. Cryo‐EM density A) and refined model B) for the CAP256‐R27‐BG505 DS SOSIP complex were shown in two 90°‐views and colored by chains. The chimera antibody bound to the HIV‐1 Env with CAP256L CDR H3 (colored olive) inserted into the V2‐apex and the CAP256LS light chain (skyblue)‐linked R27 contacting one of the CD4‐binding sites, however, the density of the flexible linker between R27 and CAP256L light chain was disordered. Three copies of R27 were observed to bind to each of the 3 CD4bs on the HIV‐1 Env trimer, indicating the other two R27s were from different chimeric antibodies.

Figure 7

CAP256L‐R27×3LS autoreactivity and half‐life in human FcRn‐Fc KI mice. A) Autoreactivity of antibodies determined by HEp‐2 cell binding assay. B) Summary of autoreactivity of antibodies determined by HEp‐2 cell binding assay and anti‐cardiolipin ELISA assay. C) In vivo half‐life of CAP256L‐nanobody variants assessed in a human FcRn‐Fc knock‐in mouse model.