In silico examination of germinal centre B cell diversity and interlineage competition in the context of broad antibody responses

High affinity antibodies that are generated by B cells undergoing affinity matu- ration within germinal centres are vital for vaccine mediated protection against viral pathogens. However, viruses can evade antibodies by acquiring modifica- tions in their surface proteins. Therefore, vaccines capable of generating broad antibody responses by targeting viral mutants and the original virus are cur- rently an important yet unmet need. Antibody breadth can emerge due to antibodies targeting multiple epitopes or due to targeting a single conserved epitope on the viral envelope. The latter are also known as broadly neutraliz- ing antibodies or bnAbs. In this thesis, using an agent-based model of the GC reaction we performed two separate studies to systematically explore the role of GC conditions in modulating the evolution of such antibodies. In this context, one possibility to elicit broad antibody responses is by en- abling GCs to host a larger clonal diversity of antigen specific B cells. However, by selecting higher affinity clones, GC affinity maturation can diminish clonal diversity. This dependency between affinity and clonal diversity is poorly un- derstood, especially when essential drivers of GC B cell selection such as antigen availability, founder cell affinity and Tfh cell quality and quantity are altered. We addressed this important knowledge gap in the first study and reported GC environments that allowed robust B cell affinity maturation with minimal losses in clonal diversity. In the second study, we focused on HIV bnAbs. bnAb elicitation via vacci- nation has not been possible so far, partly because naive B cells that can evolve into bnAbs are very rare. To circumvent this, germline-targeting immunogens that are capable of strongly binding and activating bnAb-type naive B cells have been developed. However, recent clinical trials suggest that a vaccine will likely need to elicit multiple bnAbs for protective efficacy. Germline-targeting immunogens have not been tested before for priming multiple bnAb lineages, signaling an important knowledge gap. In this work, we constructed a mech- anistic model of B cell priming, which recapitulated the known experimental observations regarding the priming of bnAbs and unraveled antibody feedback as an additional factor that modulated bnAb lineage priming. We then extended our model to multiple bnAb lineages and reported conditions amenable to their priming. Both studies here and their findings have important implications for design- ing vaccines aimed at eliciting broad spectrum antibody responses.