Autoimmunity provoked by infection: how good is the case for T cell epitope mimicry?

Autoimmune diseases remain one of the mysteries that perplex immunologists. What makes the immune system, which has evolved to protect an organism from foreign invaders, turn on the organism itself? A popular answer to this question involves the lymphoid network's primordial function: autoimmunity is a by-product of the immune response to microbial infection. For decades there have been tantalizing associations between infectious agents and autoimmunity: beta-hemolytic streptococci and rheumatic fever; B3 Coxsackieviruses and myocarditis; Trypanosoma cruzi and Chagas' disease; diverse viruses and multiple sclerosis; Borrelia burgdorfii and Lyme arthritis; and B4 Coxsackievirus, cytomegalovirus or rubella and type 1 diabetes, to name the most frequently cited examples. In addition, animal models have provided direct evidence that infection with a particular microbe can incite a particular autoimmune disease. Nonetheless, many of the associations appear less than convincing and, even for those that seem to be on solid footing, there is no real understanding of the underlying mechanism(s).