A hypothesis for the HLA-B27 immune dysregulation in spondyloarthropathy: contributions from enteric organisms, B27 structure, peptides bound by B27, and convergent evolution.

Several human rheumatic diseases occur predominantly in persons who carry the histocompatibility (HLA) class I allele B27. They have also been related to Gram-negative enteric microorganisms. In addition, the recent recovery of peptides bound to B27 has allowed an understanding of the structural requirements for their binding. Using the accumulated data base of protein sequences, we have tested a series of hypotheses. First, we have asked whether the primary amino acid sequence of the hypervariable regions of HLA-B27 shares short sequences with the proteins of Gram-negative enteric bacteria. The data demonstrate that, unique among the HLA-B molecules, the hypervariable regions of HLA-B27 unexpectedly share short peptide sequences with proteins from these bacteria. Second, we have asked whether the enteric proteins tend to satisfy the structural requirements for peptide binding to B27 in those regions of the sequence shared with B27. This hypothesis also tends to be true, especially in an allelically variable part of the B27 sequence which is predicted to bind B27 if it were to be presented as a free peptide. We conclude that HLA-B27 and enteric Gram-negative bacteria have undergone a previously unappreciated form of convergent evolution which may be important in the process leading to these rheumatic diseases. Moreover, the regions of the enteric bacterial proteins which are contiguous with the short sequences shared with B27 tend to have structures which are also predicted to bind B27. These observations suggest a mechanism for autoimmunity and lead to the prediction that the B27-associated diseases are mediated by a subset of T-cell receptors, B27, and the peptides bound by B27.