Experimental data show that relatively low concentrations of 15-deoxyspergualin (DSG) inhibit the induction of cytotoxic T lymphocytes (CTL) and the generation of antibody-producing cells. Considerably higher concentrations of DSG are required to inhibit proliferative responses. In this in vitro study, the effects of DSG on CTL induction, on proliferative responses induced by different stimuli, and on the production of interleukins IL-1, IL-2 and IL-6 and IFN-gamma (gamma-interferon) were assessed and compared with the effects of CsA (cyclosporine A) and/or FK506. We confirmed the suppressive action of DSG on the generation of CTL. Quite unexpectedly, however, we found that, although DSG did not affect the proliferative response to allogeneic lymphocytes or a superantigen, it did inhibit proliferation of peripheral blood leucocytes (PBL) stimulated with Staphyloccus aureus. DSG was active even when added on day 2 of in vitro culture, suggesting that DSG does not inhibit early events. The fraction of CD3+ lymphoblasts and the CD4/CD8 ratio was lower in cells stimulated by S. aureus in the presence of DSG, showing a selective effect on CD3+CD4+ responder T lymphocytes. The proportion of IL-2 receptor (CD25) positive cells was also reduced by DSG treatment. Moreover, we found that DSG inhibited the proliferation induced by PHA (phytohaemagglutinin) but not by Con A (concanavalin A). This effect of DSG was time-dependent, since PHA induced proliferation was not affected until day 4 after stimulation, and indicated that DSG may inhibit proliferation induced via a CD2- but not via a CD3-mediated pathway. DSG did not influence the production of IL-2 or IFN-gamma or the lipopolysaccharide induced production of IL-2 or IL-6. In contrast, the production of IL-6 was inhibited when cells were stimulated by allogeneic lymphocytes, S. aureus, PHA or Con A. This suggested to us that the DSG-suppressed IL-6 production could be the basis for the other observed effects. We tried to mimic the DSG effects with antibodies and indeed found that the IL-6 specific antibodies had similar effects. Furthermore, recombinant IL-6 completely overcame the suppressive effects of DSG on S. aureus and PHA induced proliferation, whereas addition of IL-6 to DSG treated PBL only partly restored the cytotoxic activity of lymphoblasts induced by allogeneic cells. Thus, the inhibitory effect of DSG on de novo synthesis of IL-6 could explain some of its immunosuppressive effects, but additional DSG-sensitive steps are obviously involved in CTL induction and differentiation.