A Novel Mechanism for Autoantigenicity: Condensate Conformational Change

Autoantigens, targets of B cell antibodies, account for ~10% of the human proteome. It is not well understood why or how particular self-proteins become autoantigens. Prominent theories posit that failures of the immune system at various levels or exposure of aberrant proteins to the immune system are responsible for autoantigenicity. In contrast, we propose a completely unrelated mechanism of autoantigenicity that is based on normal cell functioning. We suggest that, when a protein associates with one of the ~100 types of biomolecular condensates, protein-protein interaction therein causes changes in protein structure or conformation that signal an untolerized state to the immune system. This proposal is derived from the fact that many macromolecules are more concentrated within condensates than in the extra-condensate environment. Thus, the propensity for macromolecular interaction may be greater in condensates compared with the non-condensate environment. In support of this proposal, we present examples of predicted conformational differences, with and without accompanying secondary structure differences, between proteins in known condensate heteromeric complexes and in their free monomer forms. In each case, the differences correlated with predicted B-cell binding, supporting our model. Further, from a compilation of ~1900 autoantigens, we estimate that autoantigens are twice as prevalent in condensates compared with the entire human proteome (Chi squared p < 0.0001), further suggesting that a property of condensates, e.g., protein conformational change, may contribute to autoantigenicity.