p140Sra-1 (Specifically Rac1-associated Protein) Is a Novel Specific Target for Rac1 Small GTPase*

Rac1 small GTPase plays pivotal roles in various cell functions such as cell morphology, cell polarity, and cell proliferation. We have previously identified IQGAP1 from bovine brain cytosol as a target for Rac1 by an affinity purification method. By using the same method, we purified a specifically Rac1-associated protein with a molecular mass of about 140 kDa (p140) from bovine brain cytosol. This protein interacted with guanosine 5′-(3-O-thio)triphosphate (GTPγS)·glutathioneS-transferase (GST)-Rac1 but not with the GDP·GST-Rac1, GTPγS·GST-Cdc42, or GTPγS·GST-RhoA. The amino acid sequences of this protein revealed that p140 is identified as a product of KIAA0068 gene. We denoted this protein as Sra-1 (SpecificallyRac1-associated protein). Recombinant Sra-1 interacted with GTPγS·GST-Rac1 and weakly with GDP·Rac1 but not with GST-Cdc42 or GST-RhoA. The N-terminal domain of Sra-1 (1–407 amino acids) was responsible for the interaction with Rac1. Myc-tagged Sra-1 and the deletion mutant capable of interacting with Rac1, but not the mutants unable to bind Rac1, were colocalized with dominant active Rac1Val-12 and cortical actin filament at the Rac1Val-12-induced membrane ruffling area in KB cells. Sra-1 was cosedimented with filamentous actin (F-actin), indicating that Sra-1 directly interacts with F-actin. These results suggest that Sra-1 is a novel and specific target for Rac1.

Rac1, a member of the Rho small GTPases, has been shown to regulate actin filament reorganization and cell substratum adhesion such as focal contact (for a review, see Ref. 1). Rac1 has been shown to be involved in platelet-derived growth factor-induced membrane ruffling (2) and insulin-induced membrane ruffling in KB cells (3). In addition to the regulation of actin cytoskeleton, Rac1 has been shown to regulate multiple cellular processes; Rac1 has been shown to stimulate the arachidonic acid release in Swiss 3T3 cells or in Rat-1 cells (4) and to regulate the activities of c-Jun N-terminal kinase and p38 (5)(6)(7), members of mitogen-activated protein kinases. Ectopic expression of dominant active Rac1 in mice results in the reduction of Purkinje cell axon terminals (8).
Rac1 has two interconvertible forms as follows: GDP-bound inactive form and GTP-bound active form (for reviews, see Refs. 1 and 9), and the GTP-bound form interacts with its target molecules and exerts its biological functions. The target molecules for Rac1 have been identified to be serine/threonine kinase PAK (10,11), WASP 1 (12,13), POR1 (14), and p67 phox (15,16) in neutrophil. Recently, we and other groups (17)(18)(19)(20) have identified IQGAP as a target for Cdc42 and Rac1. Among these targets, POR1 has been shown to specifically interact with Rac1 but not with Cdc42. POR1 has been shown to be involved in the Rac1-induced membrane ruffling formation (14). However, it still remains to be clarified how Rac1-specific function is achieved. Therefore, it is important to identify its novel targets.
In the present study, we purified putative targets for Rac1 with molecular masses of 140 kDa and 120 kDa and identified them as a KIAA0068 gene product (Sra-1) (21) and HEM-2 (22), respectively. We found here that Sra-1 directly interacts with activated Rac1 and F-actin, suggesting that Sra-1 is a novel target for Rac1.

EXPERIMENTAL PROCEDURES
Materials and Chemicals-The cDNA of human Sra-1 (KIAA0068) was identified as described (21). Actin was prepared from rabbit skeletal muscle as described (23). KB cells were kindly provided from Dr. Y. Miyata (Kyoto University, Japan). Anti-Myc polyclonal antibody and anti-HA monoclonal antibody (12CA5) were purchased from Santa Cruz Co. (Santa Cruz, CA) and Boehringer Mannheim, respectively. Reticulocyte in vitro translation system kit was purchased from Promega Co. (Madison, WI). Other materials and chemicals were obtained from commercial sources.
Plasmid Construction-Recombinant wild-type and mutant small GTPases were expressed as GST fusion proteins and purified as described (20). For expression in COS7 cells, pEF-BOS-HA-small GT-Pases were prepared as described (20). To obtain pBluescript KS(Ϫ) full-length cDNA of Sra-1, cDNA of Sra-1 subcloned into EcoRV and NotI sites of pBluescript SK(Ϫ) was digested with SacII, blunted, and inserted into ClaI-digested and blunt-ended pBluescript KS(Ϫ). To produce pGEX4T-1 Sra-1 (1-407 aa), the cDNA fragment corresponding * This study was supported by grants-in-aid for Scientific Research and for Cancer Research from the Ministry of Education, Science and Culture, Japan (1996) and by grants from the Mitsubishi Foundation and Kirin Brewery Co. Ltd. The cDNA Research Program was supported in part by the Kazusa DNA Research Institute. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) D38549.
§ Present address: Division of Microbial Molecular Genetics, Nara Institute of Science and Technology, Ikoma 630-01, Japan.
GST-Rac1 Affinity Column Chromatography-The affinity purification was performed essentially as described (20). Briefly, 30 ml of bovine brain cytosol (25 mg/ml protein) was passed through glutathione beads (1 ml) to remove endogenous GST. Then, the pass fraction was loaded on glutathione beads (1 ml) coated with GST-small GTPases preloaded with guanine nucleotides as described (20). After washing the columns, bound proteins were coeluted with GST-small GTPases by the addition of 5 ml of reduced glutathione. After p140 was resolved by SDS-PAGE, the protein was transferred to polyvinylidene difluoride membrane and subjected to amino acid sequencing as described (24). The eluates (40 l) were subjected to SDS-PAGE followed by immunoblotting using anti-Sra-1 antibody.
Interaction of Sra-1 with Small GTPases-The insect cells overexpressing Myc-Sra-1 were homogenized with buffer containing 20 mM Tris/HCl at pH 7.5, 1 mM dithiothreitol, 1 mM EDTA, and 10 M (p-amidinophenyl)methanesulfonyl fluoride on ice and centrifuged at 100,000 ϫ g for 1 h at 4°C and purified from the supernatant by MonoQ column chromatography. Myc-Sra-1 (100 g) was loaded on the affinity beads coated with GST-small GTPases (50 g each) preloaded with guanine nucleotides as described (20). After washing the columns, elution was performed by the addition of 500 l of reduced glutathione. Forty l of the eluates were resolved by SDS-PAGE followed by silver staining. In vitro translation was performed using reticulocyte lysate and T7 polymerase, and the fragments were labeled with [ 35 S]Met. The indicated fragments of in vitro translated Sra-1 were mixed with affinity beads coated with the GST-small GTPases. After washing the beads, the bound proteins were coeluted with the small GTPases by the addition of glutathione. The eluates were resolved by SDS-PAGE, and radioactivities were detected using the bioimaging analyzer BAS2000 (Fujix, Tokyo, Japan).
Cosedimentation Assay-Cosedimentation assay was performed as described (25). Myc-Sra-1 (260 nM) and F-actin (3 M) were mixed and incubated for 2 h at 4°C. The mixture (50 l) was loaded onto 20% (w/v) sucrose layer (100 l) and then centrifuged at 200,000 ϫ g for 1 h at 4°C. The supernatant and the pellet were then subjected to SDS-PAGE followed by immunoblotting with anti-Sra-1 antibody, and the amounts of Sra-1 were quantified.

RESULTS AND DISCUSSION
To identify Rac1-interacting molecules, the bovine brain cytosol was loaded onto a GST-Rac1 affinity column as described (20). The proteins bound to the affinity column were coeluted with GST-Rac1 by the addition of glutathione. Proteins with molecular masses of about 180 kDa (p180), 170 kDa (p170), 140 kDa (p140), 120 kDa (p120), 110 kDa (p110), and 90 kDa (p90) were detected in the glutathione eluate from the GTP␥S⅐GST-Rac1 affinity column but not from the GST or GDP⅐GST-Rac1 affinity column (20) (Fig. 1A). We have previously identified p170, p110, and p90 as IQGAP1, p110, and p85 subunits of phosphatidylinositol-3 kinase, respectively (20). To determine the molecular identities of the p140 and p120, they were subjected to amino acid sequencing as described (24). Seven peptide sequences derived from p140 were determined (Fig. 1B). The amino acid sequences of p140 revealed that p140 was identical to those of a putative protein encoded by a human cDNA, KIAA0068 (DDBJ/EMBL/GenBank accession number D38549), which was identified by the cDNA project (21). The amino acid sequences of the p120 are 1) AAEDLFVNIRGY, 2) ELATVLSDQPG, 3) ASLSLADHREL, and 4) RLSSVDSVLK, all of which are identical to the deduced amino acid sequence of rat HEM-2 (22). 2 We have also determined the amino acid sequence of p180 and found that p180 is a novel protein which is highly homologous to myosin heavy chain. 3 The molecular weights of a KIAA0068 gene product and HEM-2 are calculated to be 145,180 and 128,250, respectively, which are almost the same as the apparent molecular masses of p140 and p120 estimated by SDS-PAGE. Moreover, anti-Sra-1 antibody recognized p140 in the eluate from GST-GTP␥S⅐Rac1 affinity column ( Fig. 2A). Therefore, we concluded that p140 and p120 were the bovine counterparts of a human KIAA0068 gene product and HEM-2, respectively. We also confirmed that p140 was detected specifically in the GST-Rac1-affinity eluate but not in the GST-Cdc42 nor GST-RhoA affinity eluates from bovine brain cytosol (20). Since p140 specifically interacted with Rac1 as described below, we denoted this protein as Sra-1 (Specifically Rac1-interacting protein) and hereafter refer to it as Sra-1.
To examine whether recombinant Sra-1 interacts with GTP␥S⅐Rac1, the affinity beads coated with GST-small GT-Pases were mixed with Myc-Sra-1 purified from overexpressing insect cells. After washing the affinity beads, GST-small GT-Pases were eluted by the addition of reduced glutathione. Myc-Sra-1 coeluted strongly with GTP␥S⅐GST-Rac1 and weakly with GDP⅐GST-Rac1 (Fig. 2B), indicating that Sra-1 directly interacts with GTP␥S⅐GST-Rac1. Recombinant Sra-1 was not detected in the eluate of GST-RhoA, GST-Cdc42, or GST affinity beads.
The indicated mutants of Sra-1 were translated in vitro, and their interactions with GST-Rac1 were examined. The in vitro translated full-length Sra-1 interacted with GTP␥S⅐GST-Rac1 and slightly with GDP⅐GST-Rac1 (Fig. 3). Sra-1 N-1 (1-959 aa) interacted with GTP␥S⅐GST-Rac1, whereas Sra-1 C-1 (659 -1253 aa) did not. The shorter fragments of Sra-1, N-2 (1-632 aa), and N-3 (1-407 aa) were also interacted with GTP␥S⅐GST-Rac1. This result indicates that the N-terminal domain of Sra-1 (1-407 aa) is responsible for the interaction with Rac1. The consensus sequence of the Rac1-binding domain of the target proteins such as PAK and WASP has been determined, and it was designated as CRIB (13,27). The shortest fragment Sra-1 N-3 did not contain the CRIB domain. Therefore, Sra-1 is categorized into the target group that does not have the CRIB domain.
Rac1 has been shown to regulate membrane ruffling (2,3). To explore whether Sra-1 is also involved in the Rac1-mediated regulation of actin cytoskeleton, we microinjected cDNAs of Myc-Sra-1 with HA-Rac1 Val-12 , structurally equivalent to Ras-Val-12 (9), into serum-starved KB cells. Membrane ruffling was observed in the injected cells as described (3) (Fig. 4A). Sra-1 specifically accumulated at Rac1-induced membrane ruffling area and colocalized with Rac1 Val-12 (Fig. 4, B and C). Myc-Sra-1 was also colocalized with cortical actin filaments (Fig. 4, E and F). We also attempted to examine whether Sra-1 interacts with Rac1 in COS7 cells overexpressing Sra-1 and Rac1 Val-12, but we failed to do so (data not shown). When Sra-1 was coexpressed with Rac1, most Sra-1 was recovered in the membrane fraction and hardly extracted by detergents. However, colocalization of Sra-1 with Rac1 strongly suggests that these proteins interact in vivo.
To explore further the interaction of Sra-1 with Rac1, we microinjected various deletion mutants of Sra-1 with Rac1 Val-12 into serum-starved KB cells and analyzed their distributions. When the deletion mutants were microinjected with Rac1 Val-12 into the cells, the membrane ruffling was elicited (Fig. 5, A, D,  G, and J), and Rac1 Val-12 was accumulated at the membrane ruffling area (Fig. 5, C, F, I, and L). The deletion mutants capable of interacting with Rac1 Val-12 were accumulated at the membrane ruffling area (Fig. 5, B, E, and H), whereas the mutant unable to interact with Rac1 Val-12 was not (Fig. 5K), suggesting that the interaction of Sra-1 with Rac1 is necessary for the accumulation of Sra-1 at the membrane ruffling area. This result is consistent with the in vitro interaction result and suggests that Rac1 may recruit Sra-1 to the membrane ruffling area.
Activated Rac1 and Cdc42 induce distinct phenotypes, membrane ruffling, and filopodia formation, respectively (2,28,29). It is likely that the distinct phenotype elicited by Cdc42 and Rac1 reflects their specific target activities. POR1 enhances the Ha-Ras Val-12 -induced membrane ruffling, and truncated POR1 inhibited the Rac1-induced membrane ruffling (14). Although this suggests that POR1 is required for the membrane ruffling formation, it is possible that other targets are also required for the membrane ruffling formation elicited by Rac1. Our results suggest that Sra-1, in addition to POR1, is involved in the membrane ruffling formation.
We explored the relationship between Sra-1 and actin filament. When Myc-Sra-1 was incubated with F-actin and centrifuged, Myc-Sra-1 was cosedimentated with F-actin (Fig. 6). Myc-Sra-1 was not cosedimentated in the absence of F-actin. This result indicates that Sra-1 directly interacts with F-actin. It is difficult to examine whether Rac1 affects the interaction of Sra-1 with F-actin, because Sra-1 was sedimentated even in the absence of F-actin when it was mixed with GST-Rac1 for an unknown reason. Taken together, it is likely that Sra-1 is involved in the regulation of membrane ruffling formation induced by Rac1 through the interaction with F-actin.
Recently, it is reported that introduction of activated PAK, a common target for Rac1 and Cdc42, resulted in the formation of filopodia in Swiss 3T3 cells (30) and induces the morphological change of HeLa cells (31). We have recently found that IQ-GAP1, another common target for Rac1 and Cdc42, directly interacts with F-actin (32). It is likely that these targets may be required for alteration of actin filament necessary for both filopodia and membrane ruffling formation. Our results suggest that Sra-1 is the specific machinery necessary for the membrane ruffling formation. Taken together, it is plausible that Sra-1 can function in concert with other targets such as PAK, IQGAP1, and POR1, resulting in the formation of membrane ruffling.