Severin Coleon, PhD

Design and Evaluation of Stabilized Envelopes for Targeting HIV-1 V2-apex Broad Neutralizing Antibody Precursors

Coleon, Severin1,2, Lu, Xiaozhi1,2, Bal, Joena1,2, Smith, Lena1,2, Zhu, Li1,2, Velasquez, Marcela1,2, Brown, Alecia1,2, Lee, Esther1,2, Cain, Derek1,2, Wagh, Kshitij3, Hahn, Beatrice H.4, Haynes, Barton F.1,2,5, Saunders Kevin O.1,2,3,6

1Duke Human Vaccine Institute, Durham, NC 27710; 2Department of Medicine, Duke University School of Medicine, Durham, NC 27710; 3Los Alamos National Laboratory, Los Alamos, NM; 4Department of Medicine, University of Pennsylvania; Philadelphia, PA; 5Department of Immunology, Duke University School of Medicine, Durham, NC 27710; 6Duke Department of Surgery, Durham, NC 27710;

Broad neutralizing antibodies (bnAbs) directed against the second variable region peptide and proximal glycans (termed the V2-apex site) of the HIV-1 envelope (Env) trimer are known for their exceptional potency. Thus, one strategy for developing an HIV-1 vaccine consists of engineering Env immunogen trimers to elicit V2-apex bnAbs. The first step in eliciting V2-apex bnAbs is to engage their unmutated antibody precursors. However, V2 apex precursors have long complementarity determining regions, which are rare in the naive antibody repertoire. To overcome the rarity of V2-apex bnAb precursors, we sought to design Env immunogens that engage multiples precursors of V2-apex bnAbs. We hypothesized that Envs from CAP256 viruses that were isolated from an individual who developed V2 Apex bnAbs could be engineered to engage additional V2-apex bnAb precursors.

A stabilized prefusion-closed trimer SOSIP version of CAP256.wk34.c80 Env sequences was mutated in the V2-apex region to optimize bnAbs precursor reactivity. Soluble Env proteins were expressed using Freestyle 293F cells. Mutated envelopes were tested for their ability to bind V2-apex bnAbs and inferred precursors. Positive binding Envs were selected to make conjugate nanoparticles. Binding magnitude and kinetics for each Env were assessed by ELISA and bio-layer interferometry (BLI) respectively. In addition, gp150 and gp145 versions of the selected envelopes were designed as mRNA immunogens, referred as CAP256SU, and stabilized using several V3-stabilizing mutations such as Q328M. mRNA constructs were transfected into Freestyle 293F cells and examined for their ability to express HIV-1 Env trimers that preferentially bind to bnAbs as compared to non-nAbs by flow cytometry.

Soluble CAP256.wk34.c80 Env trimers bearing three different motifs: DKKKK, DKKRK or DKRRK in the V2 region at position 167-171 were produced. Strong binding was observed against six HIV-1 V2-apex bnAbs (PGT145, VRC26.25, CH01, PG9, PG16, PCT64) for all three envelopes. High binding to CH01 RUA was also observed for all three envelopes. Interestingly the DKRRK motif improved the binding to PG9 UCA and PG16 UCA as compared to the other two, but lost its ability to bind VRC26.UCA. DKKRK envelopes conjugated to nanoparticles showed a binding increase for eight of the twenty-two V2-apex glycan precursors tested as opposed to the trimer alone, highlighting the effects of avidity of the nanoparticles. In addition, a Q328M substitution stabilized CAP256SU gp150 expressed by mRNAs, with a notable binding increase for PG9 and PG16 and CH01 RUA. Overall, the gp150 expression from mRNAs was higher than gp145.

Here we show HIV-1 Env trimers can be designed to bind to multiple human and rhesus macaque V2-apex bnAb precursors. Avidity increases the reactivity with V2-apex bnAB precursors. These new antigens will be tested as part of a sequential HIV-1 vaccine strategy, and will be used as antigens for determining V2 apex bnAbs precursor frequencies and immunogenetics.