貝淵 弘三

We have previously shown that IQGAP1, a recently identified target for Cdc42 and Rad small GTPases, showed a distribution similar to that of cortical actin cytoskeleton at the membrane ruffling area induced by insulin and Rac1(val12) (Kuroda, S., Fukata, M,, Kobayashi, K., Nakafuku, M., Nomura, N., Iwamatsu, A., and Kaibuchi, K. (1996) J. Biol. Chem. 271, 28363-23367). Here we identified an IQGAP1-interacting molecule with molecular mass of 43 kDa (p43) from bovine brain cytosol, using glutathione S-transferase (GST)-IQGAP1 affinity column chromatography. The amino acid sequencing of the protein revealed that p43 was identical to beta- and gamma-actin. IQGAP1 was cosedimentated with filamentous actin (F-actin). The amino-terminal domain (amino acids 1-216) of IQGAP1 was responsible for the interaction with F-actin. Falling ball viscometry assay revealed that IQGAP1 cross-linked the F-actin. This IQGAP1 activity was further enhanced by guanosine 5'-(3-O-thio)triphosphate (GTP gamma S) GST-Cdc42 but not by GDP-GST-Cdc42. The gel filtration analysis of IQGAP1 revealed that IQGAP1 appeared as oligomers and that GTP gamma S-GST-Cdc42 but not GDP GST-Cdc42 enhanced the oligomerization of IQGAP1. These results strongly suggest that IQGAP1, acting downstream of Cdc42, can cross-link the actin filament through its oligomerization.

Small GTPase Rho plays pivotal roles in the Ca2+ sensitization of smooth muscle. However, the GTP-bound active form of Rho failed to exert Ca2+-sensitizing effects in extensively Triton X-100-permeabilized smooth muscle preparations, due to the loss of the important diffusible cofactor (Gong, M. C., Iizuka, K., Nixon, G., Browne, J. P., Hall, A., Eccleston, J. F., Sugai, M., Kobayashi, S., Somlyo, A. V., and Somlyo, A. P. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 1340-1345), Here we demonstrate the contractile effects of Rho-associated kinase (Rho-kinase), recently identified as a putative target of Rho, on the Triton X-100-permeabilized smooth muscle of rabbit portal vein, Introduction of the constitutively active form of Rho-kinase into the cytosol of Triton X-100-permeabilized smooth muscle provoked a contraction and a proportional increase in levels of monophosphorylation of myosin light chain in both the presence and the absence of cytosolic Ca2+. These effects of constitutively active Rho-kinase were wortmannin (a potent myosin light chain kinase inhibitor)-insensitive. Immunoblot analysis revealed that the amount of native Rho-kinase was markedly lower in Triton X-100-permeabilized tissue than in intact tissue. Our results demonstrate that Rho-kinase directly modulates smooth muscle contraction through myosin light chain phosphorylation, independently of the Ca2+-calmodulin-dependent myosin light chain kinase pathway.

We examined the intracellular localization of the myosin binding subunit (MBS) of smooth muscle myosin phosphatase. In MDCK cells in a confluent monolayer of polarized epithelial sheet, MBS was concentrated to the cell-cell adhesion sites. Double-immunofluorescence analysis with anti-MBS and anti-β-catenin antibodies showed that MBS was mainly localized at the adherens junction. Furthermore, MBS was translocated reversibly between the cytosol and the cell-cell adhesion sites during the formation and disappearance of cell-cell contacts. These data suggest that MBS may play an important role in the regulation of the cell-cell adhesion.

We have identified, in Xenopus oocyte cytosol, a protein kinase named REKS (Ras-dependent extracellular signal-regulated kinase (ERK)/mitogen-activated protein kinase kinase (MER) stimulator), which phosphorylates and activates recombinant ERK2 through recombinant MEK in a recombinant GTP gamma S (guanosine 5'-(3-O-thio)triphosphate)-Ras-dependent manner. We show here that this REKS activity is synergistically enhanced by a combination of mammalian recombinant GTP gamma S-Ki-Ras and 14-3-3 protein purified from rat brain. 14-3-3 protein is known to activate tyrosine and tryptophan hydroxylases, to modulate the protein kinase C activity, to stimulate secretion, and to show phospholipase A(2) activity per se. 14-3-3 protein did not affect the MEK activity. 14-3-3 protein neither interacted with Ki-Ras nor affected the neurofibromin activity to stimulate the GTPase activity of Ki-Ras under the conditions where the recombinant N-terminal fragment of c-Raf-1 inhibited it. These results suggest that 14-3-3 protein has an additional function in the regulation of the Ras-MEK-ERK cascade pathway through the activation of REKS.

The small GTP-binding protein Rab3A (identical to smg p25A) is expressed in neural/exocrine/endocrine cells, is distributed between the cytosol and secretory vesicle membranes, and may cycle between these locations to regulate exocytosis. It is proposed that the GTP/GDP state of Rab3A controls this distribution. In PC12 cells, cytosolic Rab3A is predominantly GDP-bound, whereas membrane-associated Rab3A is approximately 50% GTP-bound. Two cytosolic factors, GDP dissociation inhibitor (GDI) and guanine nucleotide releasing factor (GRF), act only on GDP.Rab3A, and preferentially with post-translationally modified Rab3A. Rab3A GTPase-activating protein (GAP) does not preferentially act on processed Rab3A, and interacts selectively with GTP.Rab3A. GDI antagonizes GRF but not GAP activity toward Rab3A. These data are consistent with the concept of an ordered Rab3A cycle controlled by factors that regulate the guanine-nucleotide binding state of Rab3A.

The point-mutated active form of ras p21 is known to activate mitogen-activated protein (MAP) kinase/extracellular signal-regulated kinase (ERK) in intact mammalian cells and Xenopus oocytes, although the direct target molecule of ras p21 remains to be identified. To elucidate the role of the post-translational processing of ras p21 for the MAP kinase activation, we established the cell-free system in which ras p21 activated MAP kinase. The guanosine 5'-(3-O-thio)triphosphate (GTPgammaS) bound form of post-translationally processed Ki-ras 4B p21 activated MAP kinase in the cytosol fraction of Xenopus oocytes, but the GTPgammaS bound form of post-translationally unprocessed Ki-ras 4B p21 or the GDP bound form of processed or unprocessed Ki-ras 4B p21 was far less effective. The GTPgammaS bound form of processed Ki-ras 4B p21 activated recombinant ERK2 in the presence of the cytosol fraction of Xenopus oocytes, but the unprocessed protein was far less effective. These results provide a complete biochemical assay for ras p21 to activate MAP kinase in a cell-free system and indicate that all the elements downstream of ras p21 necessary for the MAP kinase activation are cytosolic and that the post-translational processing of ras p21 is important for the MAP kinase activation.

The superoxide-generating NADPH oxidase system in phagocytes consists of at least membrane-associated cytochrome b558 and three cytosolic components named SOCI/NCF-3/sigma-1/C1, SOCII/NCF-1/p47-phox, and SOCIII/NCF-2/p67-phox. p47-phox and p67-phox were isolated, and their primary structures were determined, but SOCI has not been well characterized. In the present study, we first purified SOCI to homogeneity from the cytosol fraction of the differentiated HL-60 cells. The purified SOCI was a small GTP-binding protein (G protein) with a M(r) of about 22,000. The guanosine 5'-(3-O-thio)triphosphate-bound form, but not the GDP-bound form, of this small G protein showed the SOCI activity. The partial amino acid sequence of SOCI thus far determined was identical to the amino acid sequence deduced from the cDNA encoding rac2 p21. None of the purified small G proteins, including Ki-ras p21, smg p21B/rap1B p21, rhoA p21, and rac1 p21, showed the SOCI activity. These results indicate that SOCI is a small G protein very similar, if not identical, to rac2 p21. The GDP/GTP exchange reaction of SOCI was stimulated and inhibited by stimulatory and inhibitory GDP/GTP exchange proteins for small G proteins, named smg GDS and rho GDI, respectively. The NADPH oxidase activity was also stimulated and inhibited by smg GDS and rho GDI, respectively. These results indicate that the superoxide-generating NADPH oxidase system is regulated by both smg GDS and rho GDI through rac2 p21 or the rac2-related small G protein in phagocytes.

We have previously purified smg GDP dissociation stimulator (GDS) from bovine brain and isolated its cDNA from a bovine brain cDNA library. smg GDS stimulates the GDP/GTP exchange reaction of a group of small GTP-binding proteins (G proteins), including at least c-Ki-ras p21, smg p21A, smg p21B, rhoA p21 and rhoB p21, by stimulating the dissociation of GDP from and the subsequent binding of GTP to each small G protein. In this study, we have isolated and sequenced the cDNA of smg GDS from a human brain cDNA library using the cloned bovine smg GDS cDNA. The cDNA has an open reading frame encoding a protein of 558 amino acids with a calculated M(r) value of 61 122. Human smg GDS shares 93% nucleotide and 96% amino acid sequence homologies with bovine smg GDS. The isolated cDNA is expressed in Escherichia coli, and the encoded protein shows the physical and functional properties similar to those of bovine smg GDS.

A stimulatory GDP/GTP exchange protein for smg p21 (smg p21 GDS) stimulated the dissociation of GDP from smg p21B. This reaction was inhibited by acidic membrane phospholipids such as phosphatidylinositol, phosphatidylinositol-4-monophosphate, phosphatidylinositol-4,5-bisphosphate, phosphatidic acid, and phosphatidylserine but not by phosphatidylcholine or phosphatidylethanolamine. These acidic phospholipids inhibited the smg p21 GDS action in a manner competitive with both smg p21 GDS and smg p21B. smg p21 GDS has other actions to inhibit the binding of smg p21B to membranes and to induce the dissociation of prebound smg p21B from the membranes. The acidic phospholipids also inhibited these two actions of smg p21 GDS. smg p21B has a polybasic region and an isoprenoid moiety in its C-terminal region which are necessary for its membrane-binding activity and its sensitivity to the smg p21 GDS actions. Therefore, it is possible that acidic membrane phospholipids interact with this polybasic region and thereby inhibit the smg p21 GDS actions.

We have recently found, by use of the rhoA p21 purified from bovine aortic smooth muscle, that it is similarly post-translationally processed as described for ras p21s: it is first geranylgeranylated at the cysteine residue in the C-terminal region followed by removal of the three C-terminal amino acids and the subsequent carboxyl methylation of the revealed C-terminal cysteine residue. In the present study, we investigated the function(s) of these post-translational modifications of the C-terminal region of rhoA p21 by use of the rhoA p21s purified from bovine aortic smooth muscle and rhoA p21-overexpressing Escherichia coli since the bacterial protein was not modified with a geranylgeranyl moiety. Bovine rhoA p21 bound to plasma membranes and phosphatidylserine-linked Affigel, but bacterial rhoA p21 did not bind to them. The inhibitory GDP/GTP exchange protein for rhoA p21, named GDP dissociation inhibitor (GDI), made a complex with the GDP-bound form of bovine rhoA p21 and thereby inhibited the dissociation of GDP from and the subsequent binding of GTP to it. However, rho GDI neither made a complex with the GDP-bound form of bacterial rhoA p21 nor affected these reactions of the bacterial protein. The stimulatory GDP/GTP exchange protein for rhoA p21, named GDP dissociation stimulator (GDS), stimulated the dissociation of GDP from bovine rhoA p21, but was inactive for the bacterial protein. In contrast, the GTPase activating protein for rhoA p21 is active not only for bovine rhoA p21 but also for the bacterial protein. These results suggest that the post-translational modifications of the C-terminal region of bovine rhoA p21, most presumably the geranylgeranylation, which are absent in bacterial rhoA p21, play important roles in its interaction with membranes and the stimulatory and inhibitory GDP/GTP exchange proteins but not with the GAP.