The wasp venom, mastoparan (MP), activates reconstituted pertussis toxin (PTX)-sensitive G-proteins in a receptor-independent manner. We studied the effects of MP and its analogue, mastoparan 7 (MP 7), on G-protein activation in HL-60 cells and a reconstituted system and on nucleoside diphosphate kinase (NDPK)-catalysed GTP formation. MP activated high-affinity GTP hydrolysis in HL-60 membranes with an EC50 of 1-2 microM and a maximum at 10 microM. Unlike the effects of the formyl peptide receptor agonist, N-formyl-L-methionyl-L-leucyl-L-phenylalanine (fMet-Leu-Phe), on GTPase, those of MP were only partially PTX-sensitive. MP-induced rises in cytosolic Ca2+ concentration and superoxide-anion formation in intact HL-60 cells were also only incompletely PTX-sensitive. N-Ethylmaleimide inhibited MP-stimulated GTP hydrolysis to a greater extent than that stimulated by fMet-Leu-Phe. Unlike the latter, MP did not enhance incorporation of GTP azidoanilide into, and cholera toxin-catalysed ADP-ribosylation of, Gi-protein alpha-subunits in HL-60 membranes. By contrast to fMet-Leu-Phe, MP did not or only weakly stimulated binding of guanosine 5'-[gamma-thio]triphosphate to Gi-protein alpha-subunits. MP 7 was considerably more effective than MP at activating the GTPase of reconstituted Gi/G(o)-proteins, whereas in HL-60 membranes, MP and MP 7 were similarly effective. MP and MP 7 were similarly effective at activating [3H]GTP formation from [3H]GDP and GTP in HL-60 membranes and by NDPK purified from bovine liver mitochondria. Our data suggest the following: (1) MP activates Gi-proteins in HL-60 cells, but (2) the venom does not simply mimic receptor activation. (3) MP and MP 7 may activate GTP hydrolysis in HL-60 membranes indirectly through interaction with NDPK. (4) MP 7 is a more effective direct activator of PTX-sensitive G-proteins than MP, whereas with regard to NDPK, MP and MP 7 are similarly effective.