Rosen A, Casciola-Rosen L (Johns Hopkins University School of Medicine, Baltimore, MD, USA). Autoantigens in systemic autoimmunity: critical partner in pathogenesis (Review). J Intern Med 2009; 265: 625-631.Understanding the mechanisms of human autoimmune rheumatic diseases presents a major challenge, due to marked complexity involving multiple domains, including genetics, environment and kinetics. In spite of this, the immune response in each of these diseases is largely specific, with distinct autoantibodies associated with different disease phenotypes. Defining the basis of such specificity will provide important insights into disease mechanism. Accumulating data suggest an interesting paradigm for antigen selection in autoimmunity, in which target tissue and immune effector pathways form a mutually reinforcing partnership. In this model, distinct autoantibody patterns in autoimmunity may be viewed as the integrated, amplified output of several interacting systems, including: (i) the specific target tissue, (ii) the immune effector pathways that modify antigen structure and cause tissue damage and dysfunction, and (iii) the homeostatic pathways activated in response to damage (e.g. regeneration/differentiation/cytokine effects). As unique antigen expression and structure may occur exclusively under these amplifying circumstances, it is useful to view the molecules targeted as 'neo-antigens', that is, antigens expressed under specific conditions, rather than ubiquitously. This model adds an important new dynamic element to selection of antigen targets in autoimmunity, and suggests that the amplifying loop will only be identified by studying the diseased target tissue in vivo.