One or more of three Gi proteins, Gi1-3, mediates hormonal inhibition of adenylyl cyclase. Whether this inhibition is mediated by the alpha or by the beta gamma subunits of Gi proteins is unclear. Mutations inhibiting the intrinsic GTPase activity of another G protein, the stimulatory regulator of adenylyl cyclase (Gs), constitutively activate it by replacing either of two conserved amino acids in its alpha subunit (alpha s). These mutations create the gsp oncogene which is found in human pituitary and thyroid tumours. In a second group of human endocrine tumours, somatic mutations in the alpha subunit of Gi2 replace a residue cognate to one of those affected by gsp mutations. This implies that the mutations convert the alpha i2 gene into a dominantly acting oncogene, called gip2, and that the mutant alpha i2 subunits are constitutively active. We have therefore assessed cyclic AMP accumulation in cultured cells which stably or transiently express exogenous wild-type alpha i2 complementary DNA or either of two mutant alpha i2 cDNAs. The results show that putatively oncogenic mutations in alpha i2 constitutively activate the protein's ability to inhibit cAMP accumulation.

Hormones inhibit synthesis of adenosine 3',5'-monophosphate (cAMP) in most cells via receptors coupled to pertussis toxin (PTX)-sensitive guanine nucleotide-binding (G) proteins. Mutationally activated alpha subunits of Gi2 (alpha i2) constitutively inhibit cAMP accumulation when transfected into cells. Cells have now been transfected with mutant alpha subunits of four other G proteins--Gz, a PTX-insensitive G protein of unknown function, and Gi1, Gi3, and G(o), which are PTX-sensitive. Mutant alpha z, alpha i1, and alpha i3 inhibited cAMP accumulation but alpha o did not. Moreover, expression of wild-type alpha z produced cells in which PTX did not block hormonal inhibition of cAMP accumulation. Thus, Gz can trigger an effector pathway in response to hormone receptors that ordinarily interact with PTX-sensitive Gi proteins.

The~i-opioidreceptor has recently been shown to stimulate phosphoinositide-specific phospholipase C via the pertussis toxin-sensitive G15 protein. Given the promiscuous nature of G16 and the high degree of resemblance of signaling properties of the three opioid receptors, both 6-and K-opioid receptors are likely to activate G16. Interactions of 6-and K-opioid receptors with G15 were examined by coexpressing the oploid receptors and Ga15 in COS-7 cells. The 6-selective agonist [D-Pen 2,DPen5] enkephalin potently stimulated the formation of mositol phosphates in cells coexpressing the 6-opioid receptorand Ga 16. The 6-opioid receptor-mediated stimulation of phospholipase C was absolutely dependent on the coexpression of Ga16 and exhibited appropriate ligand selectivity and dose dependency. Similar transfection studies revealed only weak stimulation by the pi-opioid receptor, whereas the K-opioid receptor produced moderate phospholipase C activity. Ga16 thus appeared to interact differentially with the three opioid receptors. Radioligand binding assays indicate that the 1i-opioid receptor was expressed at a lower level than those of the 6-and K-Opioid receptors. To examine if differential coupling to Ga16 is prevalent, a panel of G5-or G1-coupled receptors was coexpressed with Ga16 in COS-7 cells and assayed for agonist-induced stimulation of phospholipase C. Activation of a2-and /12-adrenergic, dopamine D1 and D2, adenosine A1, somatostatin-l and -2, C5a, formyl peptide, and luteinizing hormone receptors all resulted in stimulation of phospholipase C, with maximal stimulations ranging from 1 .5-to almost 17-fold. These findings suggest that the promiscuous Ga16 can in fact discriminate among different receptors and that such preferential interaction might in part be due to the abundance of receptors.

In a recently discovered class of oncogenes, GTPase-inhibiting mutations constitutively activate a subunits of signal-transducing guanine nucleotide-binding proteins (G proteins). Somatic mutations in a subclass of endocrine tumors are found in the arginine-179 codon ofthe a subunit ofGu2 (aj2), creating the putative gip2 oncogene. We have tested the ability ofgip2 to mediate neoplastic transformation of Rat-l and NIH 3T3 fibroblasts in tissue culture. Expression of a mutant aj2 cDNA encoding cysteine in place of arginine-179 (au2-R179C)caused Rat-1 cells to grow to a higher density in monolayer culture, to lose anchorage dependence, and to form tumors when injected subcutaneously into nude mice. In contrast, expression of aj2-R179C failed to alter growth or tumorigenicity of NIH 3T3 cells. We conclude that gip2 is an oncogene, by the criterion that it induces neoplastic transformation of Rat-1 cells. Failure of gip2 to transform NIH 3T3 cells is in keeping with clinical indications that gip2 is a tissue-selective oncogene.Hormone receptors activate heterotrimeric guanine nucleotide-binding proteins (G proteins) by promoting replacement of GDP by GTP in the guanine nucleotide-binding sites of the G proteins' a subunits. Each GTP-bound a subunit then regulates the activity of its specific downstream effector(s) until an intrinsic GTPase activity converts bound GTP to GDP, turning off the a subunit and returning it to its GDP-bound, inactive state. In the a subunit of Gs (as), point mutations (1-4) or cholera toxin-catalyzed covalent modification (5) inhibit this GTPase turnoff mechanism, thereby inducing constitutive activation of adenylyl cyclase, the effector of Gs.Recently discovered mutations in genes for G protein a subunits have revealed an additional class of human oncogenes. As with many other oncogenes, these a subunit mutations presumably constitutively activate mitogenic signaling pathways that are normally regulated by their corresponding protooncogene products. The gsp oncogene results from somatic point mutations that inhibit the GTPase activity of as, causing constitutive activation of adenylyl cyclase. By mimicking mitogenic effects of trophic hormones that normally stimulate cAMP synthesis via Gs-coupled receptors, the gsp oncogene contributes to the growth oftumors derived from pituitary somatotrophs (2, 6-9) and thyroid cells (7).A second putative G protein oncogene, gip2, results from somatic point mutations in the gene for the a subunit of Gi2 (ai2) (7). Three inferences suggest that these a12 mutations create an oncogene: (i) gip2 mutations are found in a substantial proportion (z30%) of a restricted subclass of human tumors-i.e., those derived from the adrenal cortex or from endocrine cells of the ovary (7). (ii) gip2 mutations were found by a method-allele-specific hybridization to tumor DNA-that can only detect clonally expanded alterations in DNA; clonal expansion is a hallmark of neoplasia. (iii) gip2 mutations substitute cysteine or histidine for the arginine residue at positi...