The abilities of Escherichia coli-derived human interferon gamma (IFN-gamma) and E. coli-derived human interferon-alpha A (IFN-alpha A) or -alpha 2 (IFN-alpha 2) to augment natural killer (NK) cytotoxicity were compared. When low concentrations (less than 10 antiviral units/ml) of interferons were used, and equal numbers of antiviral units of E. coli-derived IFN-gamma and E. coli-derived IFN-alpha A or IFN-alpha 2 were compared for their ability to augment NK, E. coli-derived IFN-gamma was found to be more active in augmenting NK against the K562 targets, than E. coli-derived IFN-alpha A or IFN-alpha 2. Antiviral units in these experiments were determined by the standard cytopathic effect assay using vesicular stomatitis virus (VSV)-challenged human fibroblasts, trisomic for chromosome 21. However, when these interferons were compared on a weight basis (ng/ml) or on a molar basis, their ability to augment NK against the K562 targets was comparable. These differences in the relative abilities of these interferons (when their concentrations were expressed in antiviral units/ml) to augment NK, were due to an approximately 100-fold difference in their specific activities (antiviral units per mg of interferon). These were 1.8 X 10(6) units/mg for E. coli-derived IFN-gamma, 2.0 X 10(8) units/mg for E. coli-derived IFN-alpha A, and 1.8 X 10(8) units/mg for E. coli-derived IFN-alpha 2. At concentrations higher than 10 units/ml, all these interferons showed a similar ability to augment NK. Studies on the kinetics of the augmentation revealed that in vitro treatment with E. coli-derived IFN-gamma for several hours was necessary for augmentation of NK against targets from haemopoietic human tumour cell lines (K562, Daudi). In contrast, alpha interferons were able to augment NK after treatment in vitro for significantly shorter periods (30 min or less with certain donors). Augmentation of NK cytotoxicity of human peripheral blood mononuclear leucocytes by E. coli-derived IFN-gamma was not accompanied by the induction of interleukin 2 (IL-2) production, suggesting that IL-2 is not involved in the augmentation of NK by IFN-gamma. A monoclonal antibody specific for human IFN-gamma blocked augmentation of NK by E. coli-derived IFN-gamma and natural IFN-gamma, but not by E. coli-derived IFN-alpha A or staphylococcal enterotoxin A (SEA).(ABSTRACT TRUNCATED AT 400 WORDS)