Previous studies have demonstrated that the Dictyostelium G alpha subunit G alpha 2 is essential for the cAMP-activation of adenylyl cyclase and guanylyl cyclase and that g alpha 2 null mutants do not aggregate. In this manuscript, we extend the analysis of the function of G alpha 2 in regulating downstream effectors by examining the in vivo developmental and physiological phenotypes of both wild-type and g alpha 2 null cells carrying a series of mutant G alpha 2 subunits expressed from the cloned G alpha 2 promoter. Our results show that wild-type cells expressing G alpha 2 subunits carrying mutations G40V and Q208L in the highly conserved GAGESG (residues 38-43) and GGQRS (residues 206-210) domains, which are expected to reduce the intrinsic GTPase activity, are blocked in multicellular development. Analysis of down-stream effector pathways essential for mediating aggregation indicates that cAMP-mediated activation of guanylyl cyclase and phosphatidylinositol-phospholipase C (PI-PLC) is almost completely inhibited and that there is a substantial reduction of cAMP-mediated activation of adenylyl cyclase. Moreover, neither mutant G alpha 2 subunit can complement g alpha 2 null mutants. Expression of G alpha 2(G43V) and G alpha 2(G207V) have little or no effect on the effector pathways and can partially complement g alpha 2 null cells. Our results suggest a model in which the dominant negative phenotypes resulting from the expression of G alpha 2(G40V) and G alpha 2(Q208L) are due to a constitutive adaptation of the effectors through a G alpha 2-mediated pathway. Analysis of PI-PLC in g alpha 2 null mutants and in cell lines expressing mutant G alpha 2 proteins also strongly suggests that G alpha 2 is the G alpha subunit that directly activates PI-PLC during aggregation. Moreover, overexpression of wild-type G alpha 2 results in the ability to precociously activate guanylyl cyclase by cAMP in vegetative cells, suggesting that G alpha 2 may be rate limiting in the developmental regulation of guanylyl cyclase activation. In agreement with previous results, the activation of adenylyl cyclase, while requiring G alpha 2 function in vivo, does not appear to be directly carried out by the G alpha 2 subunit. Our data are consistent with adenylyl cyclase being directly activated by either another G alpha subunit or by beta gamma subunits released on activation of the G protein containing G alpha 2.