The pathogenesis of the autoimmune disease, myasthenia gravis (MG), involves an antibody-mediated attack against acetylcholine receptors (AChRs). Since the relevant antibody response is T cell dependent, a therapeutic strategy aimed at T lymphocytes actively participating in the immune reaction to AChR should result in relatively selective suppression of AChR antibody. During an active immune response, T cells express receptors for interleukin 2 (IL2). In this study, we have used a genetically engineered fusion protein comprised of the binding region of IL2 and the toxic portion of diphtheria toxin (DAB486-IL2), to attempt to treat an experimental animal model of MG in rodents. We examined the effects of treatment with DAB486-IL2 in vivo on primary, ongoing, and secondary antibody responses to purified Torpedo AChR. Treatment of mice with intraperitoneal injections of DAB486-IL2 beginning at the time of immunization inhibited the primary AChR antibody response by 50% during the treatment period. Ongoing and secondary antibody responses to AChR were not suppressed in vivo by treatment with DAB486-IL2. In comparison, DAB486-IL2 was far more potent in suppressing antibody responses and lymphoproliferation in cell culture. At a dose comparable to that given in vivo, cellular proliferation and antibody production were virtually eliminated in a secondary response in vitro. The suppressive effect of DAB486-IL2 was much more pronounced when it was given at the time of initial antigen stimulation, as compared with its effect when given during an already established antibody response. These findings suggest that the effect of the fusion toxin on AChR antibody production was due predominantly to inhibition of T cells rather than B cells.