Antibodies are generated by B cells in a process that involves V(D)J recombination followed by somatic hypermutation and selection of high-affinity antibodies. But if there is an alternative, biologically plausible mechanism for speedier production of higher-affinity antibodies, then why is this mechanism not used by the immune system? This question has been addressed in a recent study by Jun Sun and colleagues.

The authors used a modelling system that captures important features of the immune response to vaccination and disease, as well as features of protein molecular evolution. Using this system, they replicated the typical immune-system dynamics of combinatorial joining of subdomains (that is, V(D)J recombination) followed by rounds of point mutation (PM). They also modelled an alternative system that included V(D)J recombination and PM, as well as gene-segment swapping (referred to as GSS+PM). They found that GSS+PM yields antibodies of higher affinity than PM alone and has faster dynamics.

The authors then compared the specificity of the antibodies generated by each strategy. They found that antibodies generated by GSS+PM recognized more antigens and with higher affinity than those generated by PM alone. That is, the antibodies generated by GSS+PM have greater crossreactivity than those generated by PM alone. Antibodies generated by GSS+PM could therefore crossreact with self-molecules, and crossreactivity has been proposed as one mechanism by which autoimmune responses might arise. In fact, the authors found that a typical antibody generated by GSS+PM would be 1,000 times more promiscuous than a typical antibody generated by PM alone (which usually recognizes only its intended target).

These results could explain why the immune system does not use GSS+PM, even though such antibodies would be generated faster and would have higher affinity than those generated by PM alone — the immune system must balance responses to non-self, invading pathogens while avoiding antiself, autoimmune responses.