1 Background/Rationale: An alternative pre-exposure prophylaxis modality of growing interest is based on the
2 premise that prevention can be safely achieved by passive immunization with broadly neutralizing anti-HIV
3 antibodies (bNAbs) against the HIV envelope. A pan-neutralizing antibody could provide a feasible means to
4 prevent HIV infection worldwide.
5 Objectives: Our hypothesis is that by rational engineering of near pan-neutralizing bNAbs against the CD4
6 binding site (CD4bs) bNAbs N49P7 and N49P9.3, using rare resistance variants of HIV as guides, a pan-
7 neutralizing with clinically relevant pharmacokinetic profiles can be made. Such a pan-neutralizing antibody
8 would even cover variants that are cross-resistant to other anti-CD4bs bNAbs (3BNC117 and N6). We have
9 preliminary data demonstrating that N49P7 and N49P9.3 are near pan-neutralizing and exhibit novel structural
10 features (e.g., N49P7 has access to the highly conserved inner domain) that can be used to guide improved
11 bNAb design. These unique features provide the closest possible starting point for building a highly potent pan-
12 neutralizing antibody. In parallel, our collaborators have developed techniques to engineer bNAbs, increasing
13 the potency of PGT121 approximately 50 times, while retaining favorable qualities (solubility and non-
14 auto/polyreactivity). The specific aims of this proposal are 1) Engineer N49P7 and N49P9.3 to improve
15 neutralization of HIV-1 viruses resistant to CD4bs bNAbs; 2) Downselection of bNAb variants from Aim 1
16 based on in vivo pharmacokinetics and biophysical properties; 3) Test engineered N49 bNAbs in two
17 humanized mouse models using resistant viruses to determine their efficacy in preventing HIV-1 infection.
18 Methods: Our plan is to make a collection of N49P7 and N49P9.3 variants with greater potency and breadth
19 using yeast surface display, focusing on selecting variants with enhanced recognition of recombinantly
20 produced gp120s and SOSIPs derived from neutralization resistant viruses. Combinatorial libraries with input
21 from x-ray crystallography, antibody lineage analysis, neutralization escape residues, and liability analysis will
22 also inform design. Emphasis will be placed on reduction of auto/polyreactive properties by aggressively
23 negatively selecting against those binding membrane proteins, immunogenic prediction algorithms, and
24 germline reversion. After 3 rounds of engineering, we will produce 48 mAbs as human IgG1 containing the “LS”
25 half-life extension and evaluate the bNAbs for in vitro neutralization. Thereafter, 8 bNAbs will be tested for
26 auto/polyreactivity, expression titer, biochemical behavior, and in vivo half-life. The top two bNAbs will be
27 tested in humanized mouse models for ability to protection against cell-free and cell-associated HIV challenge.
28 Impact: The development of a single, fully pan-neutralizing antibody with desirable PK and solubility properties
29 absent discernable polyreactivity would be a major advance in efforts to develop approaches for immune-
30 based treatment and prevention of HIV.