In this article the authors propose a dynamic model of autoimmunity with T-cell recruitment and selection leading to changes in the specificity of the anti-self response during the course of disease. They argue that these change~ are due to alterations in self-antigen presentation that lead to the display of previously cryptic self-determinants. Mechanisms that could underlie this differential self-presentation are proposed.

Our understanding of T-cell mediated autoimmunity is derived to a large extent from the study of animal models of disease, for example, experimental allergic encephalomyelitis (EAE). This inflammatory, demyelinating condition can be induced by immunization with central nervous system (CNS) proteins such as myelin basic protein (MBP) or proteolipid protein (PLP), and resembles human multiple sclerosis (MS). Data obtained in the EAE system have led to the notion that autoimmune T-cell responses are restricted to a single self-determinant and utilize a very limited range of T-cell receptor (TCR) genes 2-4. Thus, in disease susceptible H-2 u mice, the early T-cell response focuses on the amino-terminal region of the MBP molecule corresponding to the N-acetylated peptide Acl-9, and the range of T-cell receptors utilized in the recognition of this determinant is confined almost entirely to the V~ gene segments 8.2 and 13. These restrictions in the diversity of the response permit specific immune intervention in EAE both at the T-cell level, by interference witl'pathogenic T cells using anti-Vl3 antibodies or by TCR peptide vaccination, and at the MHC level by blocking with anti-class 1I antibodies or by peptide competition (reviewed in Ref. 1). Do similar restrictions occur in human autoimmunity? Is specific immunotherapy a feasible therapeutic approach?