We applied G protein-derived beta gamma-subunits to permeabilized mast cells to test their ability to regulate exocytotic secretion. Mast cells permeabilized with streptolysin-O leak soluble (cytosol) proteins over a period of 5 min and become refractory to stimulation by Ca2+ and GTPgammaS over approximately 20-30 min. beta gamma-Subunits applied to the permeabilized cells retard this loss of sensitivity to stimulation (run-down) and it can be inferred that they interact with the regulatory mechanism for secretion. While alpha-subunits are without effect, beta gamma-subunits at concentrations >10(-8 )M enhance the secretion due to Ca2+ and GTPgammaS. Unlike the small GTPases Rac and Cdc42, beta gamma-subunits cannot induce secretion in the absence of an activating guanine nucleotide, and thus further GTP-binding proteins (likely to be Rho-related GTPases) must be involved. The enhancement due to beta gamma-subunits is mediated largely through interaction with pleckstrin homology (PH) domains. It remains manifest in the face of maximum activation by PMA and inhibition of PKC with the pseudosubstrate inhibitory peptide. Soluble peptides mimicking PH domains inhibit the secretion due to GTPgammaS and block the enhancement due to beta gamma-subunits. Our data suggest that beta gamma-subunits are components of the pathway of activation of secretion due to receptor-mimetic ligands such as mastoparan and compound 48/80.

The septins are a family of GTPases involved in cytokinesis in budding yeast, Drosophila, and vertebrates (see for review). Septins are associated with a system of 10 nm filaments at the S. cerevisiae bud neck, and heteromultimeric septin complexes have been isolated from cell extracts in a filamentous state. A number of septins have been shown to bind and hydrolyze guanine nucleotide. However, the role of GTP binding and hydrolysis in filament formation has not been elucidated. Furthermore, several lines of evidence suggest that not all the subunits of the septin complex are required for all aspects of septin function. To address these questions, we have reconstituted filament assembly in vitro by using a recombinant Xenopus septin, Xl Sept2. Filament assembly is GTP dependent; moreover, the coiled-coil domain common to most septins is not essential for filament formation. Septin polymerization is preceded by a lag phase, suggesting a cooperative assembly mechanism. The slowly hydrolyzable GTP analog, GTP-gamma-S, also induces polymerization, indicating that polymerization does not require GTP hydrolysis. If the properties of Xl Sept2 filaments reflect those of native septin complexes, these results imply that the growth or stability of septin filaments, or both, is regulated by the state of bound nucleotide.

Antisera directed against specific subunits of guanine nucleotide-binding regulatory proteins (G proteins) were used to immunoprecipitate these polypeptides from metabolically labeled cells. This technique detects, in extracts of a human astrocytoma cell line, the alpha subunits of Gs (stimulatory) (alpha 45 and alpha 52), a 41-kDa subunit of Gi (inhibitory) (alpha 41), a 40-kDa protein (alpha 40), and the 36-kDa beta subunit. No protein that comigrated with the alpha subunit of Go (unknown function) (alpha 39) was detected. In cells grown in the presence of [3H]myristic acid, alpha 41 and alpha 40 contained 3H label, while the beta subunit did not. Chemical analysis of lipids attached covalently to purified alpha 41 and alpha 39 from bovine brain also revealed myristic acid. Similar analysis of brain G protein beta and gamma subunits and of Gt (transducin) subunits (alpha, beta, and gamma) failed to reveal fatty acids. The fatty acid associated with alpha 41, alpha 40, and alpha 39 was stable to treatment with base, suggesting that the lipid is linked to the polypeptide via an amide bond. These GTP binding proteins are thus identified as members of a select group of proteins that contains myristic acid covalently attached to the peptide backbone. Myristate may play an important role in stabilizing interactions of G proteins with phospholipid or with membrane-bound proteins.

1. The action of GTP-binding proteins on ATP-sensitive potassium (KATP) channels was investigated. KATP channels were expressed in a mammalian cell line (COS-1 cells) by cotransfecting vectors carrying the sulphonylurea receptor (SUR1) and BIR (Kir6.2), a member of the inward rectifier K+ channel family. G proteins were also tested on KATP channels composed of an isoform of SUR1, SUR2A, in combination with Kir6.2. 2. The alpha and beta gamma subunits of the GTP binding protein G1 were tested separately in inside-out patches under continuous recording. G alpha-11 increases the activity of SUR1-Kir6.2 and SUR2A-Kir6.2 channels by 200 and by 30%, respectively. 3. G alpha-12 does not increase the activity of SUR1-Kir6.2 channels, but increase the activity of SUR2A-Kir6.2 channels by 30%. 4. Control experiments showed that GTP gamma S, a specific activator of G proteins, and heat-inactivated G alpha-11 do not increase the single channel activity. 5. No effects of the other subunits (beta gamma) from either G11 or G12 on the single channel activity were observed. 6. The protein kinase C inhibitors H7 and an inhibitory peptide (FARKGALRQKNV) had no effect on the modulatory action of G alpha-11 on SUR1-Kir6.2 channels. 7. We conclude that both types of reconstituted KATP channels are modulated by G proteins.