Computer Simulations of the Humoral Immune System Reveal How Imprinting Can Affect Responses to Influenza HA Stalk with Implications for the Design of Universal Vaccines

Background Successful vaccination against the H1N1 Influenza A virus has required the continuous development of new vaccines that are antigenically similar to currently circulating strains. Vaccine strategies that can increase the cross-reactivity of the antibody response, especially to conserved regions, are essential to creating long-lasting immunity to H1N1 viruses. How pre-existing immunity affects vaccine-induced antibody cross-reactivity is still not well understood. Methods An immunological shape space of antigenic sites of hemagglutinin (HA) was constructed using viral sequence data. A Gillespie Algorithm-based model of the humoral immune system was used to simulate B cell responses to A/California/07/2009 (CA09) HA antigen after prior immunization with an antigenically similar or dissimilar strain. The effect of pre-existing memory B cells and antibody on the resulting antibody responses was interrogated. Results We found increased levels of highly-cross-reactive antibodies after immunization with antigenically dissimilar strains. This increase was dependent on pre-existing memory B cells. Furthermore, pre-existing antibody also interfered with the cross-reactive antibody response, but this effect occurred irrespective of the priming antigen. Conclusion These findings suggest that vaccination by divergent strains will boost highly-cross-reactive antibodies by selectively targeting memory B cells specific to conserved antigenic sites and by reducing the negative interference caused by pre-existing antibody.