Ankylosing spondylitis, HLA-B27, and klebsiella: cross reactivity and antibody studies.

SIR, The association of klebsiella with HLA-B27 involves two separate theoretical models which have been the subject of some debate, recently discussed by Singh et al,' and deserve further comment. The first model deals with cross reactivity or molecular mimicry (cross tolerance hypothesis) between HLA-B27 and antigens found in some Gram negative bacteria (klebsiella, shigella, and yersinia), which we first proposed some 10 years ago.2 In this model there is no difference between HLA-B27 positive lymphocytes obtained from either patients with ankylosing spondylitis (AS) or healthy controls. The second model deals with bacterial cross reactivity which is present only in HLA-B27 positive lymphocytes obtained from patients with AS but not in those lymphocytes obtained from healthy controls.3 We would like to bring to your notice that three other independent centres have published reports of cross reactivity between Gram negative bacteria (klebsiella, shigella, and yersinia) and HLA-B27. Van Bohemen and coworkers from Amsterdam, using anti-B27 monoclonal antibodies provided by Grumet from Stanford in California, showed that anti-B27(M1) reacted with a protein component in two klebsiella strains (K21 and K43) and Yersinia enterocolitica type 9.4 Kono and coworkers from Los Angeles produced an anti-Yersinia enterocolitica monoclonal antibody which reacted with 12112 B27 lymphoblastoid cell lines, but with only 4/31 B27 negative ones.5 Three of the four reactive B27 negative cell lines carried B7, an HLA antigen that cross reacts with HLA-B27. Trull and Panayi from Guy's Hospital (London) produced a mouse anti-klebsiella serum which reacted in microlymphocytotoxicity assays with 5/7 B27 positive Iymphocytes and with 0/13 B27 negative lymphocytes. It was interesting to note that 2/5 of the reactive lymphocytes came from healthy B27 positive individuals.