Leadbetter E. A et al. (2002). Chromatin-IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors. Nature, 416:603-607.

A common feature of systemic autoimmune diseases such as rheumatoid arthritis (RA) and systemic lupus erythematosis (SLE) is the production of self-reactive antibodies. The world of Toll-like receptors (TLRs) has suddenly become more interesting with the recent discovery that TLR-9 is required for the induction of autoantibodies in a mouse model of autoimmune diseases. Some of these target other antibodies and are known as rheumatoid factors (RFs). Another important example occurs in SLE where autoantibodies bind chromatin or other nucleic acid-protein complexes.

Using a mouse strain in which most B cells bear IgM class RF to IgG2a on their surface, Leadbetter et al. have explored the mechanism whereby the B cells produce excess RF. The model is rather complex in that the mice have B cells expressing surface RF to self IgG2a but are normal. If however they are bred with mice prone to autoimmune disease (the lpr mouse), the self IgG2a become immunogenic and this provokes the production of circulating RF autoantibodies. The authors examined these mice closely and found that the IgG2a recognizes nucleosomes, which are DNA-protein complexes normally found in the nucleus. They demonstrated that proliferation of the RF+B cells by IgG2a/nucleosome complexes required DNA, as treatment of the nucleosomes with DNAse abolished the effect. The question then asked was what was the co-receptor that would act in concert with the IgM RF on the B cell surface in this effect? This is because B cells require two signals for activation: one via the B cell receptor and the other via co-stimulation, which can be T cell-dependent or -independent. In the model here, no T cells are present, and the authors turned to TLRs, which in the case of TLR-4 has been shown to co-stimulate B cell activation in response to lipopolysaccharide. To test a role for TLRs, the authors bred the mice with MyD88-deficient mice. MyD88 is an adapter protein required for signalling by all TLRs so far analysed. These mice did not respond, implying a role for TLR(s) in the process. The obvious candidate TLR was TLR-9, which is required for responses to hypomethylated CpG DNA, which is present in bacterial DNA and to a lesser extent in mammalian DNA. They tested for TLR-9 involvement in two ways: first, TLR-9 signalling differs from other TLRs tested in its requirement for endosome acidification. Chloroquin interferes with this process and was shown to be inhibitory. Second, CpG analogues that act as antagonists to TLR-9 blocked the effect.

The model arrived at therefore was that RF-producing B cells are activated by co-stimulation through surface RF recognizing IgG2a bound to antigens in nucleosomes, and TLR-9 binding host DNA. This co-stimulation is essential for the response to occur. Importantly, other TLRs could have similar effects. Autoantibodies often recognise endogenous sub-cellular antigens, which might therefore also drive co-stimulation via RF on the B cell surface in combination with TLR-4, which has been shown to respond to sub-cellular proteins such as hsp60. If microbial products participate, this might explain why relapse in autoimmune diseases can occur following infection. The work therefore has potentially far-reaching consequences for our understanding of the pathogenesis of autoimmune inflammatory diseases and points the way to novel therapies that would limit TLR function during disease flares. The clinical focus might therefore shift away from T cells in RA back to B cell activation as an important pathogenic process, as was the case in the 1970s and 1980s.