Induction of broadly neutralizing antibodies (bnAbs) is a critical unmet goal of HIV vaccine development. BnAb 10E8 is of particular interest as a vaccine lead due to its (1) very high breadth, (2) low number of mutations required to develop broad neutralization, (3) low auto-reactivity, (4) expected high frequency of precursors in the human na'ive B-cell pool and (5) ability to provide surprisingly potent in vivo sterilizing protection in passive transfer studies. Key challenges for inducing 10E8-like bnAbs are: (a) the low germline affinity by HIV peptides and proteins, (b) the severe restriction on bnAb angle of approach imposed by the recessed, membrane-proximal epitope environment and (c) the absence of the 10E8 epitope from most soluble, native-like trimers (e.g. SOSIP or NFL or UFO). A promising strategy to initiate 10E8-like bnAb induction is germline targeting, in which suitable 10E8-class precursors are specifically activated using engineered immunogens, thus selecting BCRs with the potential to develop broad neutralization in the absence of autoreactivity. This approach will also help circumvent steric problems associated with the recessed location of the epitope, by priming precursors with known genetic and structural potential to mature into bnAbs compatible with MPER steric restraints. In this proposal, we will engineer epitope-scaffold immunogens that activate 10E8-like precursors, using computational design and directed evolution. We will initially test immunogens ex vivo using human na'ive B cell sorting. We will further generate knock-in mice that over-express 10E8-like precursors, and we will use those mice to test B-cell priming and boosting in vivo, first adoptively transferring knockin B cells to wild- type mice in order to mimic frequencies of 10E8 bnAb precursor and competitor B cells in humans. As known bnAbs are highly mutated, vaccine induction of bnAbs following a germline prime will likely require sequential immunization with other immunogens designed to shepherd affinity maturation of the B-cell receptor. We will develop different classes of boosting immunogens, including epitope-scaffolds with more native epitopes, membrane-protein scaffolds and membrane-bound Env variants stabilized in a conformation to which 10E8 binds strongly. We will conduct sequential prime/boost immunization experiments in knock-in mice and use ELISA, cytometry, single B-cell sorting and sequencing and neutralization assays to track and optimize affinity maturation. In summary, these studies seek to develop novel HIV vaccine candidates and also to shift HIV vaccine research towards a reductionist approach based on strategic identification of bNAb precursors, state-of-the-art protein engineering to develop germline-targeting and boosting immunogens, development of human Ig knock-in mouse models to enable testing of human-repertoire-specific vaccines, and in-depth analysis of vaccine-induced affinity maturation pathways in vivo to guide iterative vaccine optimization.
An optimal HIV vaccine will likely need to elicit multiple different types of broadly neutralizing antibodies (bnAbs) targeting multiple epitopes on the HIV glycoprotein trimer. A promising strategy to elicit bnAbs involves designing sequences of vaccine immunogens to first activate bnAb precursor B cells and then guide their affinity maturation to elicit broad neutralizing activity. In this proposal we will design such vaccine regimens to elicit one particularly promising class of HIV bnAb and we will test and optimize our vaccines in knock-in mice that we engineer to produce human germline- precursor B cells.