Do B cells drive the diversification of immune responses?

Abstract

Although the phenomenon of immunodominance of epitopes within protein antigens is well established, immune responses in which antigen persists are characterized by a great diversity of Tand B-cell responses. This is observed especially in chronic infection and autoimmunity. In this brief paper, we would like to propose a mechanism for diversification of immune responses based on the unique ability of the B cell to present the diversity of peptides from an antigen or antigenic complex required for a diverse CD4 T-cell response based on our recent studies on the role of B cells in antigen presentation to CD4 T cells in vivo (Refs 1,2 and M.J. Mamula, S. Fatehnejad and J. Craft, submitted). The activation of CD4 T cells by an exogenous protein antigen requires preseotation by a cell that can internalize the antigen, process it, and present it on the surface bound to major histocompatibility complex (MHC) class II molecules. The same APC must provide suitable costimulatory signals, the best characterized of which is the CD28 ligand B7/BB1 ~'a. Much recent data supports a role for dendritic cells (DC) in the initiation of CD4 T-cell responses to protein antigens in vWO s-7. However, earlier studies suggested that B cells were important for the induction of recall proliferative responses to protein antigens, in that mice treated from ~-;'h --';'~ anti-la ~'-: . . . . :,._a.. o-co. . . . . . . . . . . .s~a~t~ a ~ t ~ u o u y tO suppress n __,, development had defective T-cell proliferative responses that could be restored with antigen-specific B cells 8-1°. It is important to note, however, that several studies have shown that such B-cell deficient mice have normal levels of MHC class II restricted CD4 helper T cells n. Thus, these studies suggest that B cells are required for some but not all aspects of CD4 T-cell priming. Dendritic cells have been shown to be the most potent stimulators of CD4 T-cell responses to non-self MHC molecules 5, and as they constitutively express B7/BB112 and high levels of MHC class II molecules s, they are also excellent candidates for initiating CD4 T-cell responses to extrinsic antigens. However, DC are quite inefficient at antigen uptake and lack a means for discriminating between extrinsic protein antigens and self proteins whose peptides compete with them for presentation. As it is estimated that 100-200 MHC class II peptide complexes are required to signal a CD4 T cell, and as a DC has around 10 s MHC class II molecules, a maximum of 1000 peptides could be presented by such a cell. Moreover, direct evidence shows that some peptides are very much overexpressed on DC ~3, suggesting even less complexity. Thus, it seems likely that extrinsic antigens will be presented relatively inefficiently by DC, consistent with the very high protein doses required for effective presentation of peptides by the MHC class II molecules of DC 6. Thus, only those peptides that bind extremely well to the MHC class II alleles of the DC are likely to achieve levels on the cell surface sufficient to prime a helper CD4 T cell. This probably accounts for the phenomenon of immunodominant peptide epitopes in primary CD4 T-cell responses. Once helper CD4 T cells are primed to immunodominant peptides on DC, they will in turn activate specific antigen-binding B cells that internalize antigen 10 000-fold more efficiently than non-specific cell¢ 41~. When these antigen-binding B cells are activated by the primed antigen-specific helper T cells, they become competent to prime naive CD4 T cells. Because of the large amounts of protein internalized by the antigen-specific B cell, a much greater range of peptides from the original protein, or from proteins associated with it, can be presented to these naive CD4 T cells. Thus, the antigen that binds to the B cell can diversify the CD4 T-cell response. Recent studies from Eynon and Parker I~ and from Fuchs and Matzinger ~7 demonstrate that when resting B cells present antigen to CD4 T cells, tolerance resu!ts, an observation thought to reflect the absence of co-stimuiatory activity on resting B cells. By contrast, B cells activated with a variety of microbial substances can stimulate naive CD4 T-cell proliferative responses in vitro ~s, although Fuchs and Matzinger ~ find that even LPS-acti-