Stochastic simulations show how passive immunization can influence the germinal centre reaction and optimize host humoral responses

Abstract

Passive immunization with antigen-specific antibodies was shown recently to induce lasting improvements in endogenous antibody production, raising the prospect of using passive immunization as a tool to engineer host humoral responses. The mechanism with which administered antibodies alter endogenous antibody production remains unknown. B cells that produce antigen-specific antibodies evolve and get selected in germinal centres (GCs). This selection requires that B cells acquire antigen presented in GCs. We hypothesized that passive immunization biases this selection in favour of B cells with high affinities for antigen. Administered antibodies form immune complexes with antigen which only B cells with higher affinities than the administered antibodies for antigen can rupture and acquire antigen, thus increasing the selection stringency in GCs. With this mechanistic hypothesis, we constructed a stochastic simulation model of the GC reaction. The simulations recapitulated and synthesized several independent experimental observations, presenting strong evidence in support of our hypothesis. Further, the simulations revealed a quality-quantity trade-off constraining the GC response. As the selection stringency increased, surviving B cells had higher affinities for antigen but fewer B cells survived. Increasing antigen availability in the GC relaxed this constraint. The affinity of the administered antibodies and/or antigen availability could thus be tuned to maximize the GC output. Comprehensively spanning parameter space, we predict passive immunization protocols that exploit the quality-quantity trade-off and maximize the GC output. Our study thus presents a new conceptual understanding of the GC reaction and a computational framework for the rational optimization of passive immunization strategies.\n\nSignificance statementWhen natural antibody production is inadequate, passive immunization with external antibodies can alleviate disease. Remarkably, passive immunization induced lasting improvements in natural antibody production in recent studies, suggesting that it could be deployed to engineer natural antibody responses. However, how administered antibodies alter natural antibody production remains unknown. B cells that produce antibodies targeting specific antigen evolve in germinal centres (GCs). We hypothesized that administered antibodies form complexes with antigen, preferentially allowing B cells with higher affinities to acquire antigen and be selected, thus altering antibody production. With this mechanistic hypothesis, we performed stochastic simulations of the GC reaction, which recapitulated experiments, unravelled a quality-quantity trade-off constraining the GC response, and predicted passive immunization protocols that maximized the GC output.