Optimizing vaccine-elicited cross-neutralizing antibody responses to the HIV Env trimer apex - Summary For an effective HIV vaccine, broadly neutralizing antibodies (bNAbs) directed to the surface envelope glycoprotein (Env) need to be generated by candidate immunogens. A major site of vulnerability is the trimer apex, or ‘top of the trimer’. Human bNAbs (PG145, PG9, CH01) elicited by natural HIV infection, access the apex vertically, display long negatively charged HCDR3s. These bNAbs interact with the topmost N-160 glycan and, as well, the positively charged Env V2 C-strand by their HCDR3s. To target this area by vaccination, Q23 NFL trimers were developed. These trimers are recognized by germline-reverted apex bNAbs (CH01 and SHIV- elicited RHA1). Leveraging avidity, we arrayed the Q23 trimers on covalent liposomes and, followed by ‘tandem ZM233 NFL boosting’, elicited autologous neutralizations. We followed this ‘tandem priming’ with heterologous boosting using WITO and 1428 NFL trimer-liposomes, eliciting serum cross-neutralization in all NHPs. Following cloning of cross-neutralizing mAbs, high-resolution cryoEM structures of mAb:trimer complexes were generated, revealing they access the trimer ‘top’ like the apex-directed bNAb, PG9. Post-analysis reveals that these immunogens lack the ‘top’ N130 glycan, allowing cross-mAbs access through this space to contact the semi- conserved, basically charged V2 C-strand. This affords us a unique opportunity to utilize these naturally N-glycan deleted trimers to optimize these initial exciting results. From these data we now understand all the blocks at the apex. Here, we will systematically address each one to overcome – C-strand semi-variability; N130 glycan lacking/presence; in addition, we will manipulate the N160 glycan by natural or engineered partial deletion to attract B cell precursors with vertical “top way in” access to the Env trimer. To begin, we will test whether elicitation of apex Abs by the Q23/ZM233 trimer-liposome ‘dual priming’ can be accomplished in transgenic mouse models: the apex-directed CH01 unmutated common ancestor (UCA) mice and the Alloy human IgG mice. In mice, priming is generally more achievable than boosting (requiring memory B cells to re-enter GCs). Following priming in the CH01.UCA mice, we will sort/clone mAbs from memory B cells of serum neutralization+ mice. We will rapidly boost other mice with 1428 NFL trimer-liposomes to maintain GCs and drive SHM. We will both clone mAbs and, in parallel, boost littermates with NFL trimer-liposomes derived from other apex-sensitive viruses to elicit apex-directed bNAbs. We will prime with trimers possessing less basic C-strands to increase neutralization breadth. We will test the best prime/regimen in NHPs (yr2). In yrs 3/4, “TTT trimers”, encoded by a single reading frame design, will be generated. This new platform allows the structure-guided production of trimers with N-160 glycan deletions in 1 or 2 protomers of the timer, enhancing “top hole” immunogenicity to elicit apex bNAbs like RHA1 or PG9. We will clone cross-neutralizing mAbs from the huIgG mice for neutralization and structural analysis (Ward lab). NHP immunization with N160-glycan-deleted trimers, followed by N160(+) boosting, will allow us to detect serum neutralization and clone mAbs for structure/function analysis.
