Rational design of a trispecific antibody targeting the HIV-1 Env with elevated anti-viral activity

Abstract

HIV-1 broadly neutralizing antibodies (bNAbs) are being explored as passively administered therapeutic and preventative agents. However, the extensively diversified HIV-1 envelope glycoproteins (Env) rapidly acquire mutations to evade individual bNAbs in monotherapy regimens. The use of a "single" agent to simultaneously target distinct Env epitopes is desirable to overcome viral diversity. Here, we report the use of tandem single-chain variable fragment (ScFv) domains of two bNAbs, specific for the CD4-binding site and V3 glycan patch, to form anti-HIV-1 bispecific ScFvs (Bi-ScFvs). The optimal Bi-ScFv crosslinks adjacent protomers within one HIV-1 Env spike and has greater neutralization breadth than its parental bNAbs. Furthermore, the combination of this Bi-ScFv with a third bNAb recognizing the Env membrane proximal external region (MPER) results in a trispecific bNAb, which has nearly pan-isolate neutralization breadth and high potency. Thus, multispecific antibodies combining functional moieties of bNAbs could achieve outstanding neutralization capacity with augmented avidity.

Fig. 1

Design of bispecific antibodies. a Structure of HIV-1 JR-FL SOSIP.664 Env trimer (PDB: 5FYK) showing the footprints of bNAbs VRC01 and PGT121 and their proximity in both intra-protomer (left) and inter-protomer (right) binding configurations. PGT122 serves as a surrogate for PGT121. b Distances between VRC01 and PGT121 VH/VL termini and two Bi-ScFv molecules of different topology

Fig. 2

Bispecific antibody constructs. a Schematic presentation of the Bi-ScFv and Bi-NAb antibody constructs. b Schematic diagram of the molecular configurations of the Bi-ScFv and Bi-NAb antibodies

Fig. 3

Binding characteristics of anti-Env bispecific antibodies. a Schematic diagram of the bispecific binding assay via biolayer interferometry (BLI). b BLI response curves of bispecific binding assay. OCTET biosensors were loaded with biotinylated RSC3 (ligand 1) presenting the CD4bs epitope, and then probed sequentially with the bispecific antibody and BG505 SOSIP.664_D368R trimer (ligand 2) presenting the V3 glycan epitope. As controls, parental IgGs were used in place of the bispecific antibody. c Upper, negative stain EM of Bi-ScFv_{dVRC01-5X-PGT121}, in complex with BG505 SOSIP.664 Env at a molar ratio of 0.5:1 (second from left and middle panel) and 6:1 (second from right and right panel). The PGT121 and VRC01 variable regions are highlighted in green and purple, respectively (middle and right panel); Lower, schematic presentation of Env trimer and the epitopes for PGT121 (green) and VRC01 (purple), side view (left), and top view (right)

Fig. 4

Neutralization breadth and potency of bispecific antibodies. a Neutralization breadth of the parental and bispecific antibodies was tested against an HIV-1 pseudovirus panel consisting of Envs of 20 viral strains. Heat maps of IC₅₀ titers were generated in Excel. In the heatmaps, each row represents a virus strain while columns represent antibodies. Warmer colors indicate more potent neutralization and blue indicates at 50 µg/mL, antibody virus neutralization is below detection threshold (see legend). Breadths based on IC₅₀s are also summarized. Potencies (IC₅₀ geometric mean values, GMT) were calculated against sensitive viruses. Total IC₅₀ GMT value, is also shown when IC₅₀ titers for all tested virus were included and an IC₅₀ value of 50 µg/mL was assigned to a neutralization-resistant virus (^† indicates that the IC₅₀ was adjusted by a factor of 3 to account for the molarity difference between the lower molecular weight Bi-ScFv and the IgG or Bi-NAb). b Neutralization breadth of the parental and bispecific antibodies was tested against a panel of 208 viral strains. Heat maps of IC₅₀, breadth and potency are shown as in a

Fig. 5

Bispecific antibody neutralization potency-breadth curves. a Potency-breadth curves comparing the bispecific antibodies to their parental IgGs in the neutralization assay with 208 viral strains (left panel) and summary of IC₅₀ titers (μg/mL) against VRC01-resistant and dual-resistant viruses that are sensitive to the bispecific antibodies (right panel). (^† indicates that the IC₅₀ was adjusted by a factor of 3 to account for the molarity difference between the lower molecular weight Bi-ScFv and the IgG or Bi-NAb). b Raw neutralization curves of dual-resistant viruses sensitive to the Bi-ScFv and Bi-NAb in a

Fig. 6

Neutralization breadth and potency of trispecific antibodies. a Neutralization breadth of the parental, bispecific and trispecific antibodies was tested against a panel of 208 viral strains. Heat maps of IC₅₀, breadth and potency are shown as in Fig. 4a. b Scatter plots of IC₅₀ titers in which each virus is represented by an individual dot. Total IC₅₀ GMT value is indicated by a black line for each antibody. Statistical differences in neutralization were evaluated using non-parametric t-test (Wilcoxon matched-pairs signed rank test) with *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. c Potency-breadth curves comparing the Tri-NAb to the Bi-NAb as well as the parental (left panel) and summary of IC₅₀ (μg/mL) against viruses that are resistant to Bi-NAb but sensitive to the Tri-NAb (right panel). ^† indicates that the IC₅₀ was adjusted by a factor of 3 to account for the molarity difference between the lower molecular weight Bi-ScFv and the IgG or Bi-NAb

Fig. 7

Analysis of multispecific antibody neutralization breadth and potency. a Comparison of Tri-NAb potency to parental IgGs potencies. Each dot on the graph represents a virus plotted in decreasing order of sensitivity to the reference parental IgG (black). The potency of the Tri-NAb (green), the second parental IgG (gray), and the third parental IgG (open circle) against the same virus is overlaid with data points below the parental IgG indicating increased potency. b Neutralization potency (IC₅₀ geometric mean values, GMT) of Bi-NAb (upper panel) and Tri-NAb (lower panel) compared to the parental antibodies. Viruses were grouped by parental antibody sensitivity, with S indicating sensitive and R, resistant. N, the number of viruses under each sensitivity group. IC₅₀ GMT titer fold change reflecting potency compared to the respective antibodies is indicated in the parentheses next to the multispecifc antibody potency. Fold changes in red indicating improved potency. (^† indicates that the IC₅₀ was adjusted by a factor of 3 to account for the molarity difference between the lower molecular weight Bi-ScFv and the IgG, Bi-NAb or Tri-NAb)