UCSF

Naïve B cells co-express two different BCR isotypes, IgM and IgD, which have identical antigen binding domains but distinct membrane proximal regions. Despite decades of investigation, it is still unclear why B cells co-express both isotypes. Initial studies of IgD- and IgM-deficient mice concluded that IgM and IgD can largely substitute for each other. However, the isotypes differ in structure and expression pattern. IgD is highly expressed on the surface of all naïve B cells, but surface IgM expression is downregulated on autoreactive B cells. Here we demonstrate that IgD is less sensitive than IgM to self-antigen, and the isotypes differ in their ability to drive rapid antibody responses. We generated competitive chimeras in which B cells expressed either IgD or IgM alone. IgD-expressing cells adopted a developmental pattern consistent with reduced self-antigen recognition; they did not enter the B1a compartment but efficiently generated marginal zone cells. We crossed a reporter of endogenous antigen signaling, Nur77-eGFP, onto IgM- and IgD-deficient backgrounds. Differences in reporter expression suggest that IgD senses endogenous antigens more weakly than IgM in vivo. Lyn-/- mice lack a kinase essential for inhibition of BCR signaling, resulting in inappropriate B cell responses to self-antigen and aberrant short-lived plasma cell (SLPC) generation. Lyn-/- B cells expressing exclusively IgD did not differentiate into SLPCs, but were competent to generate germinal center (GC) responses. Similarly, B cells expressing exclusively IgD exhibit normal GC but impaired IgG1+ SLPC generation in response to T-dependent immunization. We propose a model in which autoreactive cells are excluded from rapid antibody responses due to predominant IgD and low IgM expression, but these cells can still participate in immune responses by entering a GC response and undergoing somatic mutation.