Further observations upon HLA-B27, Yersinia enterocolitica, and ankylosing spondylitis

Sulphasalazine in ankylosing spondylitis. A radiological, clinical and laboratory assessment

In a 12-month double-blind placebo-controlled trial, the effect of sulphasalazine was studied in 40 patients with ankylosing spondylitis. The treatment group showed significant improvement in pain, stiffness, sleep disturbance (p less than 0.05), finger/floor distance, erythrocyte sedimentation rate, C-reactive protein, orosomucoid and IgA levels (p less than 0.01). There was improvement in sleep disturbance (p less than 0.05), finger/floor distance and erythrocyte sedimentation rate (p less than 0.01) in the placebo group. Sulphasalazine did not retard radiological progression as measured either by plain X-ray or computerised tomographic scans. Multiple analysis of variance did not show a significant difference in disease activity indicators between the 2 groups.

Antigenic mimicry and autoimmune diseases

The finding of cross-reactive autoantibodies or sequence homology does not necessarily mean that this molecular mimicry is biologically meaningful or associated with disease pathogenesis. For example, relatives of persons with putative autoimmune insulin-dependent diabetes [123], and elderly humans [124] have a high incidence of autoantibodies which are generally not associated with autoimmune disease. In addition, natural antibodies to cell constituents [125] may be present in normal sera. These antibodies need to be directed against biologically important domains of host cell proteins in order to mediate autoimmune disease [27]. In spite of extensive homology between two sequences, a cross-reactive immune response may not be generated. The dissimilar amino acids should not be radical substitutions or affect the binding properties of the molecule. For instance, antibodies to synthetic peptides with only one substitution in a 19 amino acid sequence may not bind the whole protein [126]. Despite an identical six amino acid sequence shared by HLA-B27 and an EBV protein, no cross-reactive antibodies to EBV peptides were found in HLA-B27 positive patients with AS or RS. Unless the homology and subsequent crossreactive immune response can recognize a host protein intimately involved in disease pathogenesis, autoimmune disease is unlikely to occur.

Lack of serologically defined arthritogenic Shigella flexneri cell envelope antigens in post-dysenteric arthritis

Post-dysenteric or reactive arthritis (ReA) is closely associated with HLA-B27. This histocompatibility antigen is heterogeneous and consists of 2 serologically defined variants: B27M1+M2+ and B27M1+M2-. This paper gives a qualitative evaluation of the antibodies present in the sera of 62 patients with dysentery due to Shigella flexneri 2a, a known arthritogenic bacterium. The patients were classified in 4 groups: B27M1+M2+ReA+ (n = 5), B27M1+M2+ReA- (n = 7); B27M1+M2-ReA- (n = 1); B27-ReA- (n = 49). The isolated infectant possessed cell envelope antigens with B27M2-like epitopes (Mr 20,000). Analysis of the spectrum of antibodies directed against the separated cell envelope antigens of S. flexneri in the sera of these patients revealed 7 main patterns of reactivity. The detectable immunogens encompassed protein stainable antigens (Mr 98, 78, 68, 54, 50, 44, 41, 35, 14 and 13 kDa), lipopolysaccharides and peptidoglycan. None of the sera possessed detectable antibodies to the B27M2-like antigen. Consequently, this antigen is unlikely to be associated with ReA, and this applies equally to other antigens or patterns of antigens. The arthritogenicity of S. flexneri may therefore not be determined by the presence or absence of detectable antibody titers to certain cell envelope antigens. We hypothesize that other properties of these antigens could be of significance.