Crystal Structure of the Rac Activator, Asef, Reveals Its Autoinhibitory Mechanism
2006; Elsevier BV; Volume: 282; Issue: 7 Linguagem: Inglês
10.1074/jbc.c600234200
ISSN1083-351X
AutoresKazutaka Murayama, Mikako Shirouzu, Yoshihiro Kawasaki, M. Kato-Murayama, Kyoko Hanawa‐Suetsugu, Ayako Sakamoto, Yasuhiro Katsura, Atsushi Suenaga, Marcos Hikari Toyama, Takaho Terada, Makoto Taiji, Tetsu Akiyama, Shigeyuki Yokoyama,
Tópico(s)Cell Adhesion Molecules Research
ResumoThe Rac-specific guanine nucleotide exchange factor (GEF) Asef is activated by binding to the tumor suppressor adenomatous polyposis coli mutant, which is found in sporadic and familial colorectal tumors. This activated Asef is involved in the migration of colorectal tumor cells. The GEFs for Rho family GTPases contain the Dbl homology (DH) domain and the pleckstrin homology (PH) domain. When Asef is in the resting state, the GEF activity of the DH-PH module is intramolecularly inhibited by an unidentified mechanism. Asef has a Src homology 3 (SH3) domain in addition to the DH-PH module. In the present study, the three-dimensional structure of Asef was solved in its autoinhibited state. The crystal structure revealed that the SH3 domain binds intramolecularly to the DH domain, thus blocking the Rac-binding site. Furthermore, the RT-loop and the C-terminal region of the SH3 domain interact with the DH domain in a manner completely different from those for the canonical binding to a polyproline-peptide motif. These results demonstrate that the blocking of the Rac-binding site by the SH3 domain is essential for Asef autoinhibition. This may be a common mechanism in other proteins that possess an SH3 domain adjacent to a DH-PH module. The Rac-specific guanine nucleotide exchange factor (GEF) Asef is activated by binding to the tumor suppressor adenomatous polyposis coli mutant, which is found in sporadic and familial colorectal tumors. This activated Asef is involved in the migration of colorectal tumor cells. The GEFs for Rho family GTPases contain the Dbl homology (DH) domain and the pleckstrin homology (PH) domain. When Asef is in the resting state, the GEF activity of the DH-PH module is intramolecularly inhibited by an unidentified mechanism. Asef has a Src homology 3 (SH3) domain in addition to the DH-PH module. In the present study, the three-dimensional structure of Asef was solved in its autoinhibited state. The crystal structure revealed that the SH3 domain binds intramolecularly to the DH domain, thus blocking the Rac-binding site. Furthermore, the RT-loop and the C-terminal region of the SH3 domain interact with the DH domain in a manner completely different from those for the canonical binding to a polyproline-peptide motif. These results demonstrate that the blocking of the Rac-binding site by the SH3 domain is essential for Asef autoinhibition. This may be a common mechanism in other proteins that possess an SH3 domain adjacent to a DH-PH module. The Rho family GTPases, including Rho, Rac, and Cdc42, participate in actin cytoskeletal network reorganization, thus resulting in cell migration and cell-cell adhesion (1.Raftopoulou M. Hall A. Dev. Biol. 2004; 265: 23-32Crossref PubMed Scopus (1137) Google Scholar). The Rho family GTPases are activated by guanine nucleotide exchange factors (GEFs), 3The abbreviations used are GEF, guanine nucleotide exchange factor; APC, adenomatous polyposis coli; PH, pleckstrin homology; DH, Dbl homology; SH3, Src homology 3; ABR, APC-binding region; MD, molecular dynamics; HA, hemagglutinin.3The abbreviations used are GEF, guanine nucleotide exchange factor; APC, adenomatous polyposis coli; PH, pleckstrin homology; DH, Dbl homology; SH3, Src homology 3; ABR, APC-binding region; MD, molecular dynamics; HA, hemagglutinin. which possess the Dbl homology (DH) domain responsible for the GEF activity. The DH domain is followed by the PH domain, and a number of DH-PH fragment structures have been reported (2.Rossman K.L. Der C.J. Sondek J. Nat. Rev. Mol. Cell. Biol. 2005; 6: 167-180Crossref PubMed Scopus (1308) Google Scholar). The Asef protein contains the DH and PH domains (Fig. 1a) and exhibits a Rac-specific GEF activity when it is bound to the armadillo repeat of the tumor suppressor adenomatous polyposis coli (APC) with a mutation (Asef stands for APC-stimulated guanine nucleotide exchange factor) (3.Kawasaki Y. Senda T. Ishidate T. Koyama R. Morishita T. Iwayama Y. Higuchi O. Akiyama T. Science. 2000; 289: 1194-1197Crossref PubMed Scopus (296) Google Scholar). The Asef protein bound with the mutated APC is involved in the migration of colorectal tumor cells (4.Kawasaki Y. Sato R. Akiyama T. Nat. Cell Biol. 2003; 5: 211-215Crossref PubMed Scopus (179) Google Scholar). The N-terminal region of Asef includes the APC-binding region (ABR) (Fig. 1a), which is presumably a flexible peptide segment. In contrast, Asef by itself showed very weak GEF activity (3.Kawasaki Y. Senda T. Ishidate T. Koyama R. Morishita T. Iwayama Y. Higuchi O. Akiyama T. Science. 2000; 289: 1194-1197Crossref PubMed Scopus (296) Google Scholar). On the other hand, a mutant form of Asef, lacking the N-terminal region, displayed GEF activity as strong as that of the APC-stimulated, full-length Asef. Therefore, the N-terminal region "autoinhibits" the DH domain of Asef in isolation, but the molecular mechanism of the autoinhibition has been elusive. A Src-homology 3 (SH3) domain resides between the ABR and the DH domain (Fig. 1a). Protein Expression and Purification−Selenomethionine-labeled Asef (residues 66–540) with an N-terminal histidine tag was expressed in the cell-free expression system (5.Kigawa T. Yamaguchi-Nunokawa E. Kodama K. Matsuda T. Yabuki T. Matsuda N. Ishitani R. Nureki O. Yokoyama S. J. Struct. Funct. Genomics. 2002; 2: 29-35Crossref PubMed Scopus (102) Google Scholar). The protein was purified by chromatography on a HisTrap column (GE Healthcare) and was subjected to tobacco etch virus protease digestion. The Asef protein was subsequently purified by MonoQ and Superdex75 gel filtration chromatography steps (GE Healthcare). The protein was concentrated in 20 mm Tris-HCl buffer (pH 7.5) containing 150 mm NaCl and 2 mm dithiothreitol to a final concentration of 9.3 mg ml–1. Decomposition of the sample was detected after a few months, and therefore, to monitor the hydrolysis of the sample during crystallization, matrix-assisted laser desorption/ionization-time of flight mass spectrometry was conducted for the crystallized samples. The spectrum indicated that the sample was intact (residues 66–540). Crystallization and Structure Determination−Crystals of Asef were grown in 20% polyethylene glycol 3350, 0.2 m MgCl2, and 0.1 m HEPES-HCl buffer (pH 7.5) at 20 °C by the hanging drop vapor diffusion method. The crystals belong to the space group C2, with unit cell dimensions a = 100.93 Å, b = 79.82 Å, c = 68.00 Å, and β = 123.6°. Three sets of x-ray diffraction data at different wavelengths (peak, 0.9789 Å; edge, 0.9794 Å; and high remote, 0.964 Å) were collected at beamline BL44B2 of SPring-8 (Harima, Japan) with an ADSC Quantum-315 CCD detector under cryogenic conditions with Paratone-N. The diffraction data were processed and scaled with the HKL2000 program package (6.Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38446) Google Scholar). The three-dimensional structure of Asef was determined by the multiple anomalous dispersion method at 2.36 Å resolution using SOLVE (7.Terwilliger T.C. Berendzen J. Acta Crystallogr. Sect. D Biol. Crystallogr. 1999; 55: 849-861Crossref PubMed Scopus (3219) Google Scholar). Subsequently, density modification was conducted by RESOLVE (8.Terwilliger T.C. Acta Crystallogr. Sect. D Biol. Crystallogr. 2000; 56: 965-972Crossref PubMed Scopus (1633) Google Scholar). The model building was performed with O (9.Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (13006) Google Scholar) and refined with CNS (10.Brunger A.T. Adams P.D. Clore G.M. DeLano W.L. Gros P. Grosse-Kunstleve R.W. Jiang J.S. Kuszewski J. Nilges M. Pannu N.S. Read R.J. Rice L.M. Simonson T. Warren G.L. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 905-921Crossref PubMed Scopus (16948) Google Scholar). The calculated electron densities around 66–113, 185–205, 485–497, 506–511, and 539–540 were not clear, and thus, they were omitted from the following refinement processes. The final model was assessed by PROCHECK in the CCP4 suite (11.Collaborative Computational Project No. 4Acta Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19733) Google Scholar). The Ramachandran plot revealed that 90.4% of the residues are in the most favored regions, with 9.6% in the additionally allowed regions. The current model yielded Rwork and Rfree values of 0.232 and 0.299, respectively. The data collection and refinement statistics are summarized in Supplemental Table 1. The ribbon and molecular surface models in the figures were depicted by PyMol. 4W. L. DeLano, personal communication. Molecular Dynamics Calculations−The x-ray structure of Asef was used as the initial structure for the SH3(+) simulation (residues 114–540) and for the SH3(–) simulations (residues 185–540). All initial structures were fully solvated by the spherical water of the TIP3P water molecules (13.Jorgensen W.L. Chandrasekhar J. Madura J.D. Impey R.W. Klein M.L. J. Chem. Phys. 1983; 79: 926-935Crossref Scopus (29474) Google Scholar), and each solvated system underwent energy minimization before the production molecular dynamics (MD) run. During the simulations, the water molecules were restrained by a soft harmonic potential (with 1.5 kcal mol–1 Å–2) to keep them associated with the protein. All MD simulations were carried out with the Amber 8.0 5D. A. Case, T. A. Darden, T. E. I. Cheatham, C. L. Simmerling, J. Wang, R. E. Duke, R. Luo, K. M. Merz, B. Wang, D. A. Pearlman, M. Crowley, S. Brozell, V. Tsui, H. Gohlke, J. Mongan, V. Hornak, G. Cui, P. Beroza, C. Schafmeister, J. W. Caldwell, W. S. Ross, and P. A. Kollman, personal communication. program on a personal computer equipped with an MDGRAPE-3 board (2 Tflops) (15.Taiji M. Proc. Hot Chips. 2004; 16Google Scholar, 16.Taiji M. Narumi T. Ohno Y. Futatsugi N. Suenaga A. Takada N. Konagaya A. Proc. Supercomputing. 2003; 2003Google Scholar). The parm99 force field (17.Wang J. Cieplak P. Kollman P.A. J. Comput. Chem. 2000; 21: 1049-1074Crossref Scopus (3472) Google Scholar) was adopted, and the time step was set to 2.0 fs. All non-bonded interactions were calculated accurately using the MDGRAPE-3 board. All bond lengths were constrained to equilibrium lengths by the SHAKE method (18.Ryckeart J.-P. Ciccotti G. Berendsen H.J.C. J. Comput. Chem. 1997; 23: 327-431Google Scholar). The temperature of each system was gradually heated at a rate of 6 K/ps and was kept constant by coupling to a temperature bath at 300 K, with a coupling constant of 1.0 ps (19.Berendsen H.J.C. Postma J.P.M. van Gunsteren W.F. DiNola A. Haak J.R. J. Chem. Phys. 1984; 81: 3684-3690Crossref Scopus (23150) Google Scholar). Cell Culture and Transfection and Antibodies−COS-7 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Plasmids were transfected into COS-7 cells using Lipofectamine 2000 (Invitrogen). The rat monoclonal antibody against the HA tag (3F10) was obtained from Roche Diagnostics, and the mouse monoclonal antibody against the Myc tag (9E10) was from Santa Cruz Biotechnology. Immunoblotting−Cell lysates in Laemmli's SDS sample buffer were separated by SDS-PAGE, electrophoretically transferred to a polyvinylidene difluoride membrane filter (Immobilon P; Millipore), and analyzed by immunoblotting using an alkaline phosphatase-conjugated goat antibody against mouse IgG or rat IgG (Promega) as a second antibody. Rac Activity Assay−To determine the cellular activation state of Rac, transfected cells were washed once with ice-cold phosphate-buffered saline and immediately lysed in 50 mm Tris-HCl buffer (pH 7.5) containing 10 mm MgCl2, 0.3 mm NaCl, and 2% IGEPAL. After centrifugation, the supernatants were mixed with 20 μg of bacterially produced GST·PAK·CRIB fusion protein bound to glutathione-Sepharose beads for 1 h. The beads and the proteins bound to the fusion protein were washed three times with 25 mm Tris-HCl buffer (pH 7.5) containing 30 mm MgCl2 and 40 mm NaCl, and then the bound proteins were separated by SDS-PAGE before the immunoblotting analysis. In the present study, we determined the crystal structure of Asef, consisting of the ABR and the SH3, DH, and PH domains (Fig. 1b). The present Asef structure reveals how the DH and PH domains interact with the SH3 domain (Fig. 1b). The SH3 domain of Asef adopts the typical fold of SH3 domains, with five antiparallel β-strands packed to form two perpendicular β-sheets and one turn of a 310 helix with the conserved Pro-173. The region connecting the SH3 and DH domains is disordered. The DH domain is composed of elongated helical bundles with six major helical segments (α1–α6) (Fig. 1b). The DH and PH domains are connected in the conventional way, which is widely conserved in many structures of GEFs for Rho family GTPases (20.Snyder J.T. Worthylake D.K. Rossman K.L. Betts L. Pruitt W.M. Siderovski D.P. Der C.J. Sondek J. Nat. Struct. Biol. 2002; 9: 468-475Crossref PubMed Scopus (191) Google Scholar). In the Asef structure, the last α-helix (α6) of the DH domain, consisting of 37 amino acid residues, is longer than those of other DH domains. Furthermore, this helix is characteristically bent in an arched shape with a curvature radius of 55 Å. Consequently, the relative location of the DH and PH domains in the Asef structure is different from those in other GEF structures. The intramolecular interactions of the SH3 domain with the DH and PH domains in Asef are shown in Fig. 2a. The SH3 domain associates mainly with the DH domain, with a contacting surface area of 350 Å2. The RT-loop (the loop between the first and second β-strands) and the C-terminal part of the SH3 domain are involved in the interactions. The main chain of the RT-loop parallels the α6 helix, with many hydrophilic interactions (Fig. 2b). In the Asef structure, the interdomain interaction surface of the SH3 domain is perpendicular to its polyproline peptide (PXXP)-binding groove and thus blocks one end of the groove. Trp-132, Trp-160, and Phe-176 of the Asef SH3 domain correspond to the three aromatic residues involved in the canonical PXXP binding. Trp-132 interacts with the α6 helix, whereas the other two are exposed to the solvent region. In the crystal structure of the p53·53BP2 complex (21.Gorina S. Pavletich N.P. Science. 1996; 274: 1001-1005Crossref PubMed Scopus (395) Google Scholar), the PXXP-binding groove and some other residues in the RT-loop of the SH3 domain of 53BP2 interact with a non-PXXP peptide of the p53 L3 loop. The intermolecular interaction surface of the 53BP2 SH3 domain (Supplemental Fig. 1a) is different from the intradomain interaction surface of the Asef SH3 domain (Fig. 2a). Furthermore, recent NMR studies revealed the binding surfaces of the p67phox, Pex13P, and Kalirin SH3 domains for non-PXXP motifs (22.Kami K. Takeya R. Sumimoto H. Kohda D. EMBO J. 2002; 21: 4268-4276Crossref PubMed Scopus (146) Google Scholar, 23.Schiller M.R. Chakrabarti K. King G.F. Schiller N.I. Eipper B.A. Maciejewski M.W. J. Biol. Chem. 2006; 281: 18774-18786Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar) (Supplemental Fig. 1a). These non-PXXP-binding surfaces are located on a different side than the PXXP-binding groove and partially overlap each other (Supplemental Fig. 1, a and b). Therefore, the interdomain interaction surface of the Asef SH3 domain is unique. The DH and PH domains constitute a unit responsible for the guanine nucleotide exchange activity in Dbl family proteins. Asef is a Rac-specific GEF (3.Kawasaki Y. Senda T. Ishidate T. Koyama R. Morishita T. Iwayama Y. Higuchi O. Akiyama T. Science. 2000; 289: 1194-1197Crossref PubMed Scopus (296) Google Scholar). The common Rho family GTPase-binding site has been identified on the DH domain in the structures of DH-PH units complexed with RhoA, Cdc42, and Rac (20.Snyder J.T. Worthylake D.K. Rossman K.L. Betts L. Pruitt W.M. Siderovski D.P. Der C.J. Sondek J. Nat. Struct. Biol. 2002; 9: 468-475Crossref PubMed Scopus (191) Google Scholar, 24.Derewenda U. Oleksy A. Stevenson A.S. Korczynska J. Dauter Z. Somlyo A.P. Otlewski J. Somlyo A.V. Derewenda Z.S. Structure (Camb.). 2004; 12: 1955-1965Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar, 25.Rossman K.L. Worthylake D.K. Snyder J.T. Siderovski D.P. Campbell S.L. Sondek J. EMBO J. 2002; 21: 1315-1326Crossref PubMed Scopus (187) Google Scholar, 26.Kristelly R. Gao G. Tesmer J.J. J. Biol. Chem. 2004; 279: 47352-47362Abstract Full Text Full Text PDF PubMed Scopus (121) Google Scholar, 27.Worthylake D.K. Rossman K.L. Sondek J. Nature. 2000; 408: 682-688Crossref PubMed Scopus (305) Google Scholar, 28.Xiang S. Kim E.Y. Connelly J.J. Nassar N. Kirsch J. Winking J. Schwarz G. Schindelin H. J. Mol. Biol. 2006; 359: 35-46Crossref PubMed Scopus (59) Google Scholar). In the present Asef structure, the Rac-binding site of the DH domain is occupied by the SH3 domain (Supplemental Fig. 2). Therefore, the present Asef structure represents the autoinhibited form, with respect to the GEF activity for Rac. We created a model structure by superimposing the DH domain of Asef on that in the Rac·Tiam1 complex (27.Worthylake D.K. Rossman K.L. Sondek J. Nature. 2000; 408: 682-688Crossref PubMed Scopus (305) Google Scholar) (Fig. 3). The model revealed that the α6 helix of the Asef DH domain causes steric hindrances with the switch 2 region and the helix α3b C-terminal region of Rac. Thus, we speculated that the α6 helix may assume a straight form, as in the other DH-PH structures, if the association between the DH and SH3 domains is disrupted. An MD calculation was carried out for the DH-PH part of Asef in the absence of the SH3 domain, and the α6 helix was stretched after 2 ns of calculations (Fig. 3). Accordingly, the Asef SH3 domain competes with Rac for the binding site on the Asef DH domain, and Rac becomes able to interact with the binding site after the release of the SH3 domain. In the absence of the mutated APC, the SH3 domain should still occupy the Rac-binding site, and therefore, the guanine nucleotide exchange activity of Asef is autoinhibited (Supplemental Fig. 3). To test this autoinhibition mechanism involving the interaction between the SH3 and DH domains, several Asef mutants were prepared with respect to Trp-132, Asp-133, and Arg-178 of the SH3 domain and Glu-365 of the DH domain. Trp-132 makes a cation-π interaction with Lys-398 (DH), and Asp-133 forms hydrogen bonds with Asn-295 (DH) (Fig. 2b, upper right). Glu-365 forms hydrogen bonds with Arg-178, Val-181, and Asn-182 of the SH3 domain (Fig. 2b, lower right). The amounts of the GTP-bound Rac1 were analyzed for COS-7 cells expressing the mutants, in comparison with the wild-type Asef protein and its truncated form (residues 66–540), the structure of which was determined in this study (Fig. 2, b and c). The amount of the GTP-bound Rac1 in COS-7 cells was increased significantly by the expression of the full-length Asef protein with the R178A, E365A, W132L/D133A, or D133A/E365A mutation but only negligibly by the expression of the wild-type, full-length Asef protein or the truncated form (Fig. 2c). These results confirmed that the SH3-DH interactions found in the present Asef structure are important for the autoinhibition of the GEF activity of Asef in cells. As the effects of the R178A and E365A mutations are slightly greater than that of the D133A/W132L mutation (Fig. 2c), the hydrogen-bonding network involving Arg-178 and Glu-365 (Fig. 2b, lower right) is more important than the other interactions (upper right) for the autoinhibition. Therefore, the SH3 domain plays the crucial role in autoinhibiting the Rac1 GEF activity of the DH domain in Asef. In proteins with the DH and PH domains following the SH3 domain(s), such as mouse/human collybistin I (14.Harvey K. Duguid I.C. Alldred M.J. Beatty S.E. Ward H. Keep N.H. Lingenfelter S.E. Pearce B.R. Lundgren J. Owen M.J. Smart T.G. Luscher B. Rees M.I. Harvey R.J. J. Neurosci. 2004; 24: 5816-5826Crossref PubMed Scopus (213) Google Scholar, 28.Xiang S. Kim E.Y. Connelly J.J. Nassar N. Kirsch J. Winking J. Schwarz G. Schindelin H. J. Mol. Biol. 2006; 359: 35-46Crossref PubMed Scopus (59) Google Scholar) and intersectin-1 (12.Zamanian J.L. Kelly R.B. Mol. Biol. Cell. 2003; 14: 1624-1637Crossref PubMed Scopus (48) Google Scholar), the GEF activity is negatively regulated by the SH3 domain. Intriguingly, a mutant intersectin-1, with the proximal SH3 domain mutated so as to abolish the PXXP binding ability, revealed that the PXXP-binding groove is not involved in inhibiting the GEF activity. These findings are consistent with our Asef structure, thus suggesting that the structural mechanism of the intramolecular GEF inhibition by the SH3 domain is conserved in these proteins. We thank Y. Nabeshima for GST·PAK·CRIB, H. Sugimura for Myc-tagged Tiam1-C1199, and Hiroko Uda-Tochio for purification of the protein. We also thank Dr. Nobuo Kamiya, Taiji Matsu, and Dr. Hisashi Naito for help in data collection at RIKEN beamline BL44B2 of SPring-8. Download .pdf (.11 MB) Help with pdf files
Referência(s)