Presentation of HLA class I-derived peptides: potential involvement in allorecognition and HLA-B27-associated arthritis.

Some 25 years ago, when purified HLA class I allotypes were first being analyzed, of major concern was that the papain used for solubilization might produce a mess of proteolytic fragments that would prove impossible to separate and sequence. Those fears proved unfounded (Parham et al. 1975), and the homogeneity of the preparations was sufficient to allow crystallization and determination of the three-dimensional structure (Bjorkman et al. 1987). Ironically the least ordered region of the electron density map provoked the most interest because it gave a first view of the diverse peptides bound by an MHC molecule. With this image a second chapter of HLA class I biochemistry began, its charge to determine the structures of bound peptides and their influence on the immune system. The extraordinary polymorphism of HLA class I heavy chains now seems quite manageable compared to the vast complexity of the peptides, and our present ignorance as to which ones are important for health and disease. The comparative weakness of most HLA class I associations with disease has made HLA-B27 an especially favored target for investigation, and more is known of the structure and peptide-presenting function of HLA-B27 than for any other HLA-B allotype (López de Castro 1994). Much of this information relates to the native HLA-B27 molecule and has been collected in the belief that disease is a direct consequence of its antigen-presenting function. If one subscribes to the relevance of the transgenic rodent models, this position has almost become untenable. For rats and mice 'non-functional' forms of HLA-B27 are the agents of disease, raising the possibility that B27-associated arthritis is induced by HLA class II presentation of a B27-derived peptide, a variant of the mechanism advanced for the classical HLA class II-associated diseases: type 1 diabetes, multiple sclerosis and rheumatoid arthritis (Gregersen et al. 1987, Roudier et al. 1989, Cucca & Todd 1996, Hall & Bowness 1996). Such speculation invites the obvious question as to whether other diseases associated with HLA class I and chronic inflammation, HLA-C and psoriasis for example (Tiilikainen et al. 1980, Yanagisawa et al. 1995), result from class II presentation of class I peptides.