A comprehensive engineering strategy improves potency and manufacturability of a near pan-neutralizing antibody against HIV

Abstract

Anti-HIV envelope broadly neutralizing antibodies (bnAbs) are alternatives to conventional antiretrovirals with the potential to prevent and treat infection, reduce latent reservoirs, and/or mediate a functional cure. Clinical trials with "first-generation" bnAbs used alone or in combination show promising antiviral effects but also highlight that additional engineering of "enhanced" antibodies will be required for optimal clinical utility, while preserving or enhancing Current Good Manufacturing Practices (cGMP) manufacturing capability. Here, we report the engineering of an anti-CD4-binding site (CD4bs) bnAb, N49P9.3. Through a series of rational modifications, we produced a variant, N49P9.6-FR-LS, that demonstrates enhanced potency, superior antiviral activity in combination with other bnAbs, low polyreactivity, and longer circulating half-life. Additional engineering for manufacturing produced a final variant, eN49P9, with properties conducive to cGMP production. Overall, these efforts demonstrate the feasibility of developing enhanced anti-CD4bs bnAbs with greatly improved antiviral properties as well as potential translational value.

Keywords: ADCC; CD4 binding site; HIV; HIV cure; bnAb; combination therapies; elite neutralizer; engineering; monoclonal antibody; neutralization.

Figure 1.. Engraftment of heavy-chain framework-3 loop from VRC03 into N49P9.6 to engineer N49P9.6-FR

The amino acid differences between N49P9.3 and N49P9.6 heavy chain are shown in the red box. The heavy-chain framework-3 loops of the bnAbs are set within the orange box, and the part of VRC03 engrafted into N49P9.6 is shaded in green.

Figure 2.. Improvement in neutralization through engineering

Variants of N49P9.3 show lower IC50 values compared to parental forms in 118 multi-tier multi-clade pseudovirus panel. N49P9 and N49P9.1 (identical VDJ to N49P9 but with LC7 instead of LC2) also included for comparison. N49P9.6-FR has an IC50 and IC80 of 0.01 and 0.03 μg/mL, respectively. IC50 values between groups compared by Mann-Whitney U test: **p < 0.01; ****p < 0.0001.

Figure 3.. N49P9.6-FR outperforms other bnAbs individually and as part of bnAb combinations

(A) Heatmap of IC80 titers for N49P9.6 and N49P9.6 FR and other best-in-class CD4bs, V2 apex, and V3 glycan bnAbs. (B) IC80 breadth-potency curves for bnAbs in (A). (C) Spiderplots showing multimetric comparison of bnAbs using IC80 breadth and potency, IC99 breadth and potency, and IIP median and breadth (IIP > 5log10). Axes ticks are defined in (D), and for each metric, radially outwards is better performance. (D) Similar to (C) except evaluating 2- and 3-bnAb combinations. Each CD4bs bnAb is combined with V2, V3, or both bnAbs, and IC80, IC99, and IIP are predicted. IC80 and IC99 breadth values were calculated at 1 μg/mL and 10μg/mL, respectively. Colors in (C) and (D) correspond to the color of the CD4bs bNAb in (B).

Figure 4.. N49P9.6-FR-LS protective efficacy in CD34-NSG-SGM3 and CD34-NSG-IL15 mouse models

(A) Experimental schema. Mice were injected i.p. with the indicated doses of bnAb and 3 h later challenged with HIV 1086c i.p. Plasma viremia was measured by quantitative reverse-transcription PCR (RT-qPCR) weekly to determine infection status. (B) Dose-effect curves for the bnAb measured in either mouse strain. Significantly greater protection was seen in Hu-CD34-IL15 mice (p < 0.0001 by logistic regression for binary outcome at week 3). All control mice injected with Synagis became infected in both mouse strains (Table S5). The experiment was repeated once with similar results.

Figure 5.. Molecular details of the Env trimer interprotomer contacts of N49P9.6-FR

(A) Top view of the complex of BG505 SOSIP.664 HIV-1 Env trimer with N49P9.6-FR (the Fabs of PGT121 were omitted for clarity). The three N49P9.6-FR Fabs (the cryo-EM model was built with the variable regions of the Fabs only, dark and light green ribbon) are shown bound to the Env trimer (inner and outer gp120 domains colored light and dark gray, respectively). On the right, one N49P9.6-FR-gp120BG505 complex was extracted from the trimer to show the details of the Fab contacts to the primary gp120 protomer. One additional primary gp120 contact of N49P9.6-FR that is only seen in the Env trimer and is mediated by the region preceding the introduced frame sequence is encircled. On the bottom, a zoomed-in view depicts the contact between N49P9.6-FR with the adjacent BG505 protomer. All contact residues within a 4 Å distance criterion cutoff are shown as sticks, and H-bonds are shown as blue dotted lines. The residues of the introduced frame sequence are shown in red. (B) Network of interactions formed between N49P9.6-FR with the adjacent gp120BG505 protomer as defined by a 5 Å distance criterion cutoff. The contacts for VRC03 by the same criteria to the adjacent gp120BG505 protomer are shown for comparison (PDB ID: 6CDI). (C) A zoomed-in view depicts the contact between N49P9.6-FR with the adjacent BG505 protomer. All contact residues within a 4 Å distance criterion cutoff are shown as sticks, and H-bonds are shown as blue dotted lines. The residues of the introduced frame sequence are shown in red. (D) Epitope footprints of Fabs mapped onto the gp120 primary sequences. Contact residues are defined by a 5 Å cutoff and marked above the sequence with (+) for side chain and (−) for main chain to indicate the type of contact. Contact types are colored as following: hydrophilic (blue), hydrophobic (green), and both (black). Buried surface residues determined by PISA are shaded blue for primary gp120 contact residues and orange for adjacent gp120 contact residues.