SPECIFICITY It is a basic assumption that, for reasons unknown, the constituents of the human body may become antigenic, and then the patient produces antibodies against them. Some autoantibodies are tissue specific (e.g. haemolytic anaemia, pernicious anaemia) whereas others are non-tissue specific (e.g. systemic lupus erythematosus). To prove 'autoimmunity' in the strict sense is not an easy task. In any event, the outcome of antigenic stimulation-whether antibody formation or tolerance-seems to depend on the same factors for autoantigens as it does for any other antigen [146]. Rabbits rendered tolerant to bovine serum albumin respond to immunization with human serum albumin, and in doing so frequently produce antibodies that react with bovine serum albumin, which means the termination of tolerance. This is of special interest as a model of autoimmune disease since natural tolerance of body constituents might be terminated by immunization with a cross-reactive antigen from the environment (266]. It cannot be overemphasized how difficult it could be, at least in several in- stances, to delineate specific from non-specific events in any immunological pro- cess. Cyclic AMP may be one of the examples for non-specific inducers, whereas thymopoietin (25a ], for instance, induces only the differentiation of cells with thymus-specific antigens. The participation of cells recruited by lymphokines also represents non-specific amplification. Lymphocytes of unrelated specificity con- tinue to accumulate in experimental autoimmune inflammation after elimination of the specifically sensitized cells that initiated the inflammation [262].

SPECIFICITY It is a basic assumption that, for reasons unknown, the constituents of the human body may become antigenic, and then the patient produces antibodies against them. Some autoantibodies are tissue specific (e.g. haemolytic anaemia, pernicious anaemia) whereas others are non-tissue specific (e.g. systemic lupus erythematosus). To prove 'autoimmunity' in the strict sense is not an easy task. In any event, the outcome of antigenic stimulation-whether antibody formation or tolerance-seems to depend on the same factors for autoantigens as it does for any other antigen [146]. Rabbits rendered tolerant to bovine serum albumin respond to immunization with human serum albumin, and in doing so frequently produce antibodies that react with bovine serum albumin, which means the termination of tolerance. This is of special interest as a model of autoimmune disease since natural tolerance of body constituents might be terminated by immunization with a cross-reactive antigen from the environment (266]. It cannot be overemphasized how difficult it could be, at least in several in- stances, to delineate specific from non-specific events in any immunological pro- cess. Cyclic AMP may be one of the examples for non-specific inducers, whereas thymopoietin (25a ], for instance, induces only the differentiation of cells with thymus-specific antigens. The participation of cells recruited by lymphokines also represents non-specific amplification. Lymphocytes of unrelated specificity con- tinue to accumulate in experimental autoimmune inflammation after elimination of the specifically sensitized cells that initiated the inflammation [262].