Vaccine responses in the context of impaired self-tolerance

Abstract

In the context of HIV vaccine design, broadly-neutralizing antibodies (bnAbs) are those which are able to neutralize a diverse population of HIV variants. The existing global diversity of HIV necessitates that any traditional vaccine intending to generate humoral immunity must elicit these bnAbs, making them an important metric in the development of HIV vaccine design strategies. BnAbs have not, to date, ever been isolated from an individual following an HIV vaccine trial. As a group, bnAbs have been shown to exhibit a number of unusual traits, including poly- and auto-reactivity, as well as a years-long delay in development, when they arise during persistent HIV infection at all. Interestingly, persistent HIV infection can erode self-tolerance, permitting the production of autoreactive antibodies. Conversely, when autoreactive bnAbs are expressed in mouse models, they are usually constrained by immune tolerance. To establish if the erosion of self-tolerance and the development of autoreactive bnAbs were related, we utilized immunogens that had been rationally designed to stimulate VRC01-like bnAb precursor B cell receptors (BCRs), mice that featured VRC01-like precursor BCRs, and Lymphocytic choriomeningitis mammarenavirus (LCMV), a paradigmatic model of persistent viral infection that impairs humoral self-tolerance. Here, we demonstrate that impaired self-tolerance does not permit the development of greater neutralization breadth. During these experiments, an unexpectedly lethal phenotype developed in the experimental animals. We identified two elements which mediated the effect: an adjuvant, and the BCR expressed in the mice. The latter was a variation of a bnAb-precursor not previously thought to be constrained by tolerance, though our data suggest that the effect was antibody-mediated. Lastly, we utilized the antigen-specificity of the same mouse model to assess the transcriptome of naïve, germinal center, memory, and antibody-secreting B cells, before pairing that dataset with a transcriptomic and proteomic analysis of the same B cell populations in humans. This permitted hypothesis-generating comparisons between molecule types, cells, and species. We were able to identify specific hits that are both meaningful and feasible to follow up on, as well as unexpected trends, including the unusual association between memory B cells and gene ontology terms associated with neuronal development.