Blind confirmation in Leiden of Geczy factor on the cells of Dutch patients with ankylosing spondylitis

Absence of autoantibodies to peptides shared by HLA-B27.5 and Klebsiella pneumoniae nitrogenase in serum samples from HLA-B27 positive patients with ankylosing spondylitis and Reiter's syndrome

Some microorganisms which are pathogenic in humans share amino acid sequences with human proteins (molecular mimicry). It has been suggested that molecular mimicry might be a reason for autoimmunity as a result of immunological cross reactivity. A homologous sequence of six amino acids has been found in both Klebsiella pneumoniae nitrogenase and the HLA-B27.5 molecule. In addition, (auto)antibodies to a synthetic peptide that contained the HLA-B27.5/klebsiella mimicking epitope have been detected in serum samples from HLA-B27 positive patients with ankylosing spondylitis and Reiter's syndrome. Confirmation of these data is important, because ankylosing spondylitis and Reiter's syndrome have so far been assumed to be 'seronegative' rheumatic diseases. It was, however, not possible to confirm the presence of autoantibodies against the mimicking peptide in serum samples from patients with ankylosing spondylitis and Reiter's syndrome. Serum samples from 81 patients with ankylosing spondylitis, 38 patients with Reiter's syndrome, and 81 healthy blood donors were tested against the 'mimicking peptide' in an enzyme linked immunosorbent assay (ELISA). Some of the serum samples from patients showed high but non-specific binding to the mimicking peptide. A highly significant correlation between binding to plastic coated with the mimicking peptide, to plastic coated with an irrelevant peptide, and even to non-coated plastic was observed. The nature of the serum component(s) in these patient serum samples (and some control serum samples) responsible for the high non-specific binding to plastic remains unclear. It was also shown that antibodies to the HLA-B27 peptide (containing the mimicking epitope) induced in rabbits do not cross react with the klebsiella peptide and vice versa.

Dysfunction of HLA-B27

Every HLA antigen, as defined by the WHO-HLA Nomenclature, is unique. The major function of these molecules is to present antigen-peptides to the T-cell receptor, thereby contributing to the immunological defence mechanism. This function is regulated for each MHC antigen by its unique structure, with the peptide-binding pockets of the three-dimensional groove of the corresponding molecules playing the critical role. However, HLA-B27 is special by virtue of its disease association(s). Various aspects which might provide an explanation for--or at least a clue to an understanding of the specific role of--B27 in its disease associations are reviewed. Since it appears that there are no published experimental data which would support either of the alternative hypothetical possibilities, the bulk of current theories must therefore be purely speculative. The only lead to a better understanding of the function of B27 in disease associations is the postinfectious reactive arthritis. If it is the B27 molecule itself which is involved, further in vivo work on B27 transgenic animals might help solve this problem with its numerous unknown factors.

Trial and error in producing ankylosing-spondylitis-selective antisera according to Andrew Geczy

Geczy found that rabbit sera raised against Klebsiella strain K43 cross-reacted with the cells from HLA-B27 positive patients with ankylosing spondylitis (AS). Other laboratories failed to reproduce these results. After a series of unsuccessful attempts, however, we managed to prepare one selective antiserum, using E. coli, isolated from a Dutch Bechterew patient, in offspring of rabbits Geczy sent us. Ever since we obtained irreproducible results only. This paper reports about the many attempts we have made to produce a discriminating antiserum for use in a combined vital stain and dye-exclusion assay.

Blood lymphocytes from ankylosing spondylitis patients fail to induce disease-specific cytotoxic T lymphocytes

The intriguing observation made by Geczy et al. (1) showing the possibility of generating specific ankylosing spondylitis--cytotoxic T lymphocytes by presenting HLA-B27+AS+ cells as antigen-specific stimulator cells prompted us (by using Geczy's approach) to identify cytotoxic T lymphocytes specific for this apparent B27+AS+ target structure. Peripheral blood mononuclear cells (PBMC) of 21 healthy B27+ individuals were stimulated in primary and in short-term cultures with PBMC of an HLA-identical sibling suffering from definite AS (n = 12). In addition, PBMC in vitro modified by "Geczy bacterial products" from two healthy B27+ individuals were used to stimulate B27+ AS- lymphocytes (either autologous or from a healthy HLA-identical sibling). Effector cells raised in primary AS- versus AS+ and AS- versus "modified B27" mixed lymphocyte culture combinations showed no proliferative nor cytotoxic activity at all. The scarcely observed cytotoxic reactivity of restimulated mixed lymphocyte culture was not restricted to AS+B27+ cells. These results demonstrate that PBMC from ankylosing spondylitis patients fail to induce disease-specific cytotoxic T lymphocytes and suggest that an ankylosing spondylitis--related "modified B27" structure does not exist, at least in the patient material tested.

In vitro mutagenesis of HLA-B27. Substitution of an unpaired cysteine residue in the alpha 1 domain causes loss of antibody-defined epitopes

The HLA class I molecules identified serologically as HLA-B27 are highly associated with ankylosing spondylitis and related human disorders. All known HLA-B27 amino acid sequences contain a cysteine residue at position 67; no other published HLA class I sequence contains a cysteine within the hypervariable region of the alpha 1 domain, which extends from amino acid residues 63-84. To investigate the role of this cysteine residue in the antigenic structure of HLA-B27, we isolated a genomic clone encoding a molecule of the HLA-B27.1 subtype and performed oligonucleotide-directed mutagenesis to convert the cysteine at position 67 to a tyrosine. When transfected into mouse L cells, both the wild-type and Cys67----Tyr67 mutant B27 genes directed the synthesis and surface expression of molecules reactive with the monomorphic anti-HLA class I antibody W6/32. However, only the L cells transfected with the wild-type B27 gene reacted with the anti-B27 antibody ME1; L cells transfected with the mutant B27 were completely unreactive with this antibody. Experiments with hybrid exons created from the HLA-B27 and HLA-A2 genes yielded results consistent with the mapping of the ME1 epitope to the B27 alpha 1 domain. A second anti-B27 antibody, GS145.2, also showed markedly reduced binding to the Cys67----Tyr67 mutant. These studies document the importance of the unique Cys67 residue in the antigenic structure of HLA-B27.