Crystal Structure of the RUN Domain of the RAP2-interacting Protein x
2006; Elsevier BV; Volume: 281; Issue: 42 Linguagem: Inglês
10.1016/s0021-9258(19)84099-1
ISSN1083-351X
AutoresMutsuko Kukimoto‐Niino, Tetsuo Takagi, Ryogo Akasaka, Kazutaka Murayama, Tomomi Uchikubo‐Kamo, Takaho Terada, Makoto Inoue, Satoru Watanabe, Akiko Tanaka, Yoshihide Hayashizaki, T. Kigawa, Mikako Shirouzu, Shigeyuki Yokoyama,
Tópico(s)Ubiquitin and proteasome pathways
ResumoRap2-interacting protein x (RPIPx) is a homolog of RPIP8, a specific effector of Rap2 GTPase. The N-terminal region of RPIP8, which contains the RUN domain, interacts with Rap2. Using cell-free synthesis and NMR, we determined that the region encompassing residues 83-255 of mouse RPIPx, which is 40-residues larger than the predicted RUN domain (residues 113-245), is the minimum fragment that forms a correctly folded protein. This fragment, the RPIPx RUN domain, interacted specifically with Rap2B in vitro in a nucleotide-dependent manner. The crystal structure of the RPIPx RUN domain was determined at 2.0Å of resolution by the multiwavelength anomalous dispersion (MAD) method. The RPIPx RUN domain comprises eight anti-parallel α-helices, which form an extensive hydrophobic core, followed by an extended segment. The residues in the core region are highly conserved, suggesting the conservation of the RUN domain-fold among the RUN domain-containing proteins. The residues forming a positively charged surface are conserved between RPIP8 and its homologs, suggesting that this surface is important for Rap2 binding. In the crystal the putative Rap2 binding site of the RPIPx RUN domain interacts with the extended segment in a segment-swapping manner. Rap2-interacting protein x (RPIPx) is a homolog of RPIP8, a specific effector of Rap2 GTPase. The N-terminal region of RPIP8, which contains the RUN domain, interacts with Rap2. Using cell-free synthesis and NMR, we determined that the region encompassing residues 83-255 of mouse RPIPx, which is 40-residues larger than the predicted RUN domain (residues 113-245), is the minimum fragment that forms a correctly folded protein. This fragment, the RPIPx RUN domain, interacted specifically with Rap2B in vitro in a nucleotide-dependent manner. The crystal structure of the RPIPx RUN domain was determined at 2.0Å of resolution by the multiwavelength anomalous dispersion (MAD) method. The RPIPx RUN domain comprises eight anti-parallel α-helices, which form an extensive hydrophobic core, followed by an extended segment. The residues in the core region are highly conserved, suggesting the conservation of the RUN domain-fold among the RUN domain-containing proteins. The residues forming a positively charged surface are conserved between RPIP8 and its homologs, suggesting that this surface is important for Rap2 binding. In the crystal the putative Rap2 binding site of the RPIPx RUN domain interacts with the extended segment in a segment-swapping manner. The small GTPases of the Ras family function as molecular switches, regulating diverse cellular processes such as proliferation and differentiation (1Bourne H.R. Sanders D.A. McCormick F. Nature. 1990; 348: 125-132Crossref PubMed Scopus (1836) Google Scholar, 2Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2684) Google Scholar). They cycle between the GTP-bound, activated forms and the GDP-bound, inactivated forms, with the aid of regulatory proteins, such as guanine nucleotide exchange factors, GTPase-activating proteins, and guanine nucleotide dissociation inhibitors. Between the GTP-bound and GDP-bound forms, conformational changes occur in two conserved regions, the switch-I and switch-II regions. The switch-I region encompasses the "effector region," through which the GTP-bound forms interact with their effectors that activate downstream targets. The Ras family consists of Ras (H-, K-, and N-Ras), M-Ras, R-Ras, TC21, Rap1 (Rap1A and B), Rap2 (Rap2A and B), Ral, Rheb, Rin, and Rit (3Campbell S.L. Khosravi-Far R. Rossman K.L. Clark G.J. Der C.J. Oncogene. 1998; 17: 1395-1413Crossref PubMed Scopus (919) Google Scholar). Rap1 and Rap2 are very close relatives (∼60% sequence identity), and they exhibit ∼50% sequence identity with Ras (4Bokoch G.M. Biochem. J. 1993; 289: 17-24Crossref PubMed Scopus (83) Google Scholar). Rap1 and Ras share an identical effector region, whereas in Rap2, the single replacement of Ser-39 by Phe is observed. Several experiments have shown that Rap1 functions as an antagonist in Ras-signaling pathways, probably by competing for the Ras effectors (5Kitayama H. Sugimoto Y. Matsuzaki T. Ikawa Y. Noda M. Cell. 1989; 56: 77-84Abstract Full Text PDF PubMed Scopus (762) Google Scholar, 6Yatani A. Quilliam L.A. Brown A.M. Bokoch G.M. J. Biol. Chem. 1991; 266: 22872-22877Abstract Full Text PDF PubMed Google Scholar, 7Sakoda T. Kaibuchi K. Kishi K. Kishida S. Doi K. Hoshino M. Hattori S. Takai Y. Oncogene. 1992; 7: 1705-1711PubMed Google Scholar, 8Cook S.J. Rubinfeld B. Albert I. McCormick F. EMBO J. 1993; 12: 3475-3485Crossref PubMed Scopus (334) Google Scholar). In contrast, Rap2 did not antagonize Ras-induced transformation (9Jimenez B. Pizon V. Lerosey I. Beranger F. Tavitian A. de Gunzburg J. Int. J. Cancer. 1991; 49: 471-479Crossref PubMed Scopus (26) Google Scholar). Rap2 interacts with Ras effectors, such as Raf, phosphatidylinositol 3-kinase (PI3K), and Ral guanine nucleotide dissociation stimulator (GDS), but it only inhibits the activation of the downstream targets of Raf and PI3K and not RalGDS (10Ohba Y. Mochizuki N. Matsuo K. Yamashita S. Nakaya M. Hashimoto Y. Hamaguchi M. Kurata T. Nagashima K. Matsuda M. Mol. Cell. Biol. 2000; 20: 6074-6083Crossref PubMed Scopus (93) Google Scholar, 11Christian S.L. Lee R.L. McLeod S.J. Burgess A.E. Li A.H. DangLawson M. Lin K.B. Gold M.R. J. Biol. Chem. 2003; 278: 41756-41767Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar, 12Nancy V. Wolthuis R.M. de Tand M.F. Janoueix-Lerosey I. Bos J.L. de Gunzburg J. J. Biol. Chem. 1999; 274: 8737-8745Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar). Furthermore, Rap2 is regulated differently by guanine nucleotide exchange factors and GTPase-activating proteins as compared with Rap1 (10Ohba Y. Mochizuki N. Matsuo K. Yamashita S. Nakaya M. Hashimoto Y. Hamaguchi M. Kurata T. Nagashima K. Matsuda M. Mol. Cell. Biol. 2000; 20: 6074-6083Crossref PubMed Scopus (93) Google Scholar), suggesting that Rap1 and Rap2 have distinct signaling functions. Rap2 interacts with its specific effector, Rap2-interacting protein 8 (RPIP8) 2The abbreviations used are: RPIP, Rap2-interacting protein 8; GFP, green fluorescent protein; GST, glutathione S-transferase; GTPγS, guanosine 5′-O-(thiotriphosphate); HSQC, heteronuclear single-quantum coherence; CARD, caspase recruitment domain; r.m.s.d., root mean square deviation; GDPNP, phosphoaminophosphonic acid-guanylate ester. 2The abbreviations used are: RPIP, Rap2-interacting protein 8; GFP, green fluorescent protein; GST, glutathione S-transferase; GTPγS, guanosine 5′-O-(thiotriphosphate); HSQC, heteronuclear single-quantum coherence; CARD, caspase recruitment domain; r.m.s.d., root mean square deviation; GDPNP, phosphoaminophosphonic acid-guanylate ester. (13Janoueix-Lerosey I. Pasheva E. de Tand M.F. Tavitian A. de Gunzburg J. Eur. J. Biochem. 1998; 252: 290-298Crossref PubMed Scopus (43) Google Scholar). The homologs of RPIP8 include Rap2-interacting protein 9 (RPIP9) (14Wang S. Zhang Z. Ying K. Chen J.Z. Meng X.F. Yang Q.S. Xie Y. Mao Y.M. Biochem. Genet. 2003; 41: 13-25Crossref PubMed Scopus (11) Google Scholar) and Rap2 interacting protein x (RPIPx). RPIP9 expression is activated during malignant breast epithelial transformation and increases with the progression toward an invasive phenotype (15Raguz S. De Bella M.T. Slade M.J. Higgins C.F. Coombes C. Yagu¨e E. Int. J. Cancer. 2005; 117: 934-941Crossref PubMed Scopus (15) Google Scholar). Tumor necrosis factor receptor-associated factor 2 (Traf2)- and Nck-interacting kinase (TNIK) is a recently identified Rap2-specific effector that is distinct from RPIP8 (16Machida N. Umikawa M. Takai K. Sakima N. Myagmar B.E. Taira K. Uezato H. Ogawa Y. Kariya K. J. Biol. Chem. 2004; 279: 15711-15714Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar, 17Taira K. Umikawa M. Takai K. Myagmar B.E. Shinzato M. Machida N. Uezato H. Nonaka S. Kariya K. J. Biol. Chem. 2004; 279: 49488-49496Abstract Full Text Full Text PDF PubMed Scopus (123) Google Scholar). TNIK interacts with Rap2 through its C-terminal regulatory domain, termed the CNH domain, and regulates the actin cytoskeleton. RPIP8 and its homologs contain a RUN domain (named after RPIP8/UNC-14/NESCA) (Fig. 1A), which is composed of six conserved blocks (18Callebaut I. de Gunzburg J. Goud B. Mornon J.P. Trends Biochem. Sci. 2001; 26: 79-83Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar). The RUN domain of RPIP8 is probably responsible for the Rap2 binding because the N-terminal region of RPIP8, which contains the RUN domain, interacts with Rap2 (13Janoueix-Lerosey I. Pasheva E. de Tand M.F. Tavitian A. de Gunzburg J. Eur. J. Biochem. 1998; 252: 290-298Crossref PubMed Scopus (43) Google Scholar). Although the RUN domain exists in several proteins that could function as specific effectors of the Rap and Rab families of GTPases, it is not clear whether the binding of such GTPases is the common role of the RUN domains. In Rabip4, a Rab4 effector, the RUN domain is not linked to the binding of Rab4 or Rap2 but is involved in the subcellular localization of the protein (19Cormont M. Mari M. Galmiche A. Hofman P. Le Marchand-Brustal Y. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 1637-1642Crossref PubMed Scopus (86) Google Scholar, 20Mari M. Macia E. Le Marchand-Brustal Y. Cormont M. J. Biol. Chem. 2001; 276: 42501-42508Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). On the other hand, Iporin, a novel Rab1-interacting protein, has been shown to interact with Rab1 through its RUN domain (21Bayer M. Fischer J. Kremerskothen J. Ossendorf E. Matanis T. Konczal M. Weide T. Barnekow A. BMC Cell Biol. 2005; 6: 15Crossref PubMed Scopus (26) Google Scholar). The RUN domain of mouse RPIPx shares 31% sequence identity with that of mouse RPIP8. Here, we identified a 173-residue fragment of mouse RPIPx (residues 83-255), including the predicted RUN domain, as a structural domain that specifically binds to Rap2B. The crystal structure of the RPIPx RUN domain revealed a novel bundle structure, which is distinct from any other known Ras/Rap-binding domain. Screening of Folded Proteins—The mouse RPIPx cDNA clone is from the FANTOM RIKEN full-length cDNA clone collection (FANTOM clone ID 2810428M05). A screening of soluble proteins using green fluorescent protein (GFP) as a C-terminal fusion, was carried out in the batch mode of an Escherichia coli cell-free protein synthesis system (22Yokoyama S. Hirota H. Kigawa T. Yabuki T. Shirouzu M. Terada T. Ito Y. Matsuo Y. Kuroda Y. Nishimura Y. Kyogoku Y. Miki K. Masui R. Kuramitsu S. Nat. Struct. Biol. 2000; 7: 943-945Crossref PubMed Scopus (325) Google Scholar). The fluorescence of reaction mixtures expressing each GFP fusion protein was measured. Next, the solubility of the proteins was confirmed by SDS-PAGE. Briefly, proteins without GFP fusions were synthesized in the small-scale dialysis mode of the cell-free reaction (23Kigawa T. Yabuki T. Matsuda N. Matsuda T. Nakajima R. Tanaka A. Yokoyama S. J. Struct. Funct. Genomics. 2004; 5: 63-68Crossref PubMed Scopus (271) Google Scholar). After an aliquot was removed, the reaction mixtures were centrifuged to remove the insoluble proteins, and then the total and soluble fractions were subjected to SDS-PAGE. Finally, the protein folding was examined by NMR. Uniformly 15N-labeled fragments of His-tagged RPIPx (residues 83-245, 83-255, 83-265, and 83-275) were synthesized in the middle-scale dialysis mode of the cell-free reaction. The proteins were partially purified by chromatography on nickel-Sepharose resin (GE Healthcare). The folding of the proteins was analyzed by measuring their 1H and 1H,15N heteronuclear single-quantum coherence (HSQC) spectra. In Vitro Binding Assay—The C-terminal-truncated form of mouse Rap2B (residues 1-167) was synthesized by the E. coli cell-free system as a fusion protein with glutathione S-transferase (GST) and was immobilized on glutathione-Sepharose 4B resin (GE Healthcare). The Histagged RPIPx RUN domain (residues 83-255) was synthesized by the cell-free system and was dialyzed against 25 mm Tris-HCl buffer (pH7.5) containing 5 mm dithiothreitol. It was then incubated with the glutathione-Sepharose 4B resin carrying immobilized GST-Rap2B or GST-Rap1A, preloaded with GDP, GTPγS, or GDPNP. After an incubation at 4 °C for 3 h, the resin was washed 4 times with 25 mm Tris-HCl buffer (pH 7.5) containing 5 mm dithiothreitol, 5 mm MgCl2, 1 mm EDTA, and 0.1% Tween 20. The bound proteins were eluted with glutathione and subjected to SDS-PAGE followed by Western blotting with a polyclonal anti-HAT antibody (Clontech). Protein Expression and Purification—The fragment encoding residues 83-255 of mouse RPIPx was cloned into the TA vector, pCR2.1TOPO (Invitrogen), as a fusion with an N-terminal His tag and a tobacco etch virus protease cleavage site. The selenomethionine (SeMet)-substituted protein was produced in the large scale dialysis mode of the cell-free reaction (24Kigawa T. Yamaguchi-Nunokawa E. Kodama K. Matsuda T. Yabuki T. Matsuda N. Ishitani R. Nureki O. Yokoyama S. J. Struct. Funct. Genomics. 2001; 2: 29-35Crossref Scopus (102) Google Scholar). The reaction solution was centrifuged at 16,000 × g at 4°C for 20 min. The resulting supernatant was loaded on a HisTrap (GE Healthcare) column (5 ml) previously equilibrated with 20 mm Tris-HCl buffer (pH 8.0) containing 1 m NaCl and 15 mm imidazole and was eluted with 20 mm Tris-HCl buffer (pH 8.0) containing 500 mm NaCl and 500 mm imidazole. The sample buffer was exchanged to 20 mm Tris-HCl buffer (pH 8.0) containing 1 m NaCl and 15 mm imidazole with a HiPrep 26/10 desalting column (GE Healthcare). The His tag was cleaved by 200 μl of tobacco etch virus protease (4 mg/ml) at 30°C for 1 h and was removed by a second passage through the HisTrap column. The sample buffer was exchanged to 20 mm Tris-HCl buffer (pH 8.5) containing 5 mm 2-mercaptoethanol with a HiPrep 26/10 desalting column. Next, the protein sample was loaded on a HiTrap Q (GE Healthcare) column (5 ml) previously equilibrated with 20 mm Tris-HCl buffer (pH 8.5) containing 5 mm 2-mercaptoethanol and was eluted with a linear gradient of 0-1 m NaCl in 20 mm Tris-HCl buffer (pH 8.5) with 5 mm 2-mercaptoethanol. Finally, the protein sample was loaded on a HiLoad 16/60 Superdex 75 (GE Healthcare) column previously equilibrated with 20 mm Tris-HCl buffer (pH 8.0) containing 150 mm NaCl and 5 mm 2-mercaptoethanol and was eluted with this buffer. The yield of the purified protein was 0.69 mg/1-ml cell-free extract. Crystallization and Data Collection—The crystallization conditions for the mouse RPIPx RUN domain, corresponding to residues 83-255, were screened using Crystal Screen 1 and 2 kits (Hampton Research) by the 96-well sitting drop vapor diffusion method. The preliminary crystals were obtained under condition number 32 of Crystal Screen 2 (0.1 m HEPES buffer at pH 7.5 containing 0.1 m sodium chloride and 1.6 m ammonium sulfate). The crystal used for structure determination was grown at 20 °C in the hanging drop vapor diffusion method in a drop composed of 1 μl of 30.3 mg/ml protein solution (20 mm Tris-HCl buffer at pH 8.0 containing 150 mm NaCl and 5 mm 2-mercaptoethanol) and 1 μl of reservoir solution (0.1 m Tris buffer at pH 8.5 containing 0.05 m NaCl and 1.4 m ammonium sulfate) equilibrated against 400 μl of the reservoir solution. Data collection was carried out at 100 K in a cryoprotectant solution containing 90% Paratone-N and 10% glycerol. The MAD data were collected at three different wavelengths at BL26B1, SPring-8 (Harima), and were recorded on a Jupiter210 CCD detector (Rigaku). The wavelengths were chosen to optimize the anomalous signals by measuring the x-ray fluorescence spectrum. All diffraction data were processed with the HKL2000 program (25Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref PubMed Scopus (38527) Google Scholar). Structure Determination and Refinement—The program SOLVE (26Terwilliger T.C. Berendzen J. Acta Crystallogr. Sect. D. 1999; 55: 849-861Crossref PubMed Scopus (3219) Google Scholar) was used to locate the selenium sites and to calculate the phases, and RESOLVE was used for the density modification (27Terwilliger T.C. Acta Crystallogr. Sect. D. 2001; 57: 1763-1775Crossref PubMed Scopus (78) Google Scholar). Automatic tracing using RESOLVE BUILD (28Terwilliger T.C. Acta Crystallogr. Sect. D. 2002; 59: 1174-1182Crossref Scopus (74) Google Scholar) was used to partially build the model, and the rest of the model was built and refined with the programs O (29Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-118Crossref PubMed Scopus (13009) Google Scholar) and CNS (30Bru¨nger 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. Acta Crystallogr. Sect. D. 1998; 54: 905-921Crossref PubMed Scopus (16957) Google Scholar). Refinement statistics are presented in Table 1. The quality of the model was inspected by the program PROCHECK (31Laskowski R.A. MacArthur M.W. Moss D.S. Thornton J.M. J. Appl. Crystallogr. 1993; 26: 283-291Crossref Google Scholar). Structural similarities were calculated with DALI (32Holm L. Sander C. J. Mol. Biol. 1993; 233: 123-138Crossref PubMed Scopus (3562) Google Scholar), VAST (33Gibrat J.F. Madej T. Bryant S.H. Curr. Opin. Biol. 1996; 6: 377-385Crossref Scopus (887) Google Scholar), and MATRAS (34Kawabata T. Nucleic Acids Res. 2003; 31: 3367-3369Crossref PubMed Scopus (216) Google Scholar). Graphic figures were created using the programs Molscript (35Kraulis P.J. J. Appl. Crystallogr. 1991; 24: 946-950Crossref Google Scholar) and Raster3D (36Merritt E.A. Bacon D.J. Methods Enzymol. 1997; 277: 505-524Crossref PubMed Scopus (3875) Google Scholar). The molecular surface was created with the program PyMOL (37DeLano W.L. PyMOL. DeLano Scientific, San Carlos, CA2002Google Scholar). The atomic coordinates have been deposited in the Protein Data Bank with the accession code 2DWK.TABLE 1X-ray data collection, phasing and refinement statisticsPeakEdgeRemoteData collectionWavelength (Å)0.97920.97950.9640Resolution (Å)50-2.0 (2.07-2.00)50-2.0 (2.07-2.00)50-2.0 (2.07-2.00)Unique reflections16,51316,49916,505Redundancy10.210.110.1Completeness (%)99.6 (100.0)99.6 (100.0)99.6 (100.0)I/σ(I)14.5 (5.2)17.4 (5.5)16.7 (5.3)Rsym (%)aRsym = ∑|Iavg - Ii|/∑Ii, where Ii is the observed intensity, and Iavg is the average intensity.10.9 (49.7)9.0 (45.2)9.4 (45.1)MAD analysisResolution (Å)15-2.0No. of sites8FOMMADbFigure of merit after SOLVE phasing.0.60FOMRESOLVEcFigure of merit after RESOLVE.0.75RefinementResolution (Å)33.6-2.0No. of reflections16,483No. of protein atoms1260No. of water molecules68Rwork (%)20.5Rfree (%)dRfree is calculated for 10% of randomly selected reflections excluded from refinement.23.1r.m.s.d. bond length (Å)0.010r.m.s.d. bond angles (°)1.4a Rsym = ∑|Iavg - Ii|/∑Ii, where Ii is the observed intensity, and Iavg is the average intensity.b Figure of merit after SOLVE phasing.c Figure of merit after RESOLVE.d Rfree is calculated for 10% of randomly selected reflections excluded from refinement. Open table in a new tab Analytical Ultracentrifugation—All analytical ultracentrifugation experiments were carried out with a Beckman Optima XL-I analytical ultracentrifuge. The sample buffer was 20 mm Tris-HCl buffer (pH 8.0) containing 150 mm NaCl and 5 mm 2-mercaptoethanol, and all experiments were performed at 20 °C. The solvent density and the protein partial specific volume (ν-) were estimated with SEDNTERP (38Laue T.M. Shah B. Ridgeway T.M. Pelletier S.L. Analytical Ultracentrifugation in Biochemistry and Polymer Science. Royal Society of Chemistry, London1992: 90-125Google Scholar). Sedimentation velocity data were obtained at 40,000 rpm using an Epon two channel centerpiece, with a loading concentration of 1.0 mg/ml. The data were analyzed with the program SEDFIT (39Schuck P. Biophys. J. 1998; 75: 1503-1512Abstract Full Text Full Text PDF PubMed Scopus (306) Google Scholar). Sedimentation equilibrium experiments were carried out with six channel centerpieces, with loading concentrations of 0.25, 0.5, and 1.0 mg/ml. Data were obtained at 15,000, 17,000, and 19,000 rpm. A total equilibration time of 24 h was used for each speed, with scans taken at 20 and 22 h to ensure that equilibrium had been attained. The absorbance wavelength was 280 nm. The equilibrium data were fitted using the manufacturer's software. Identification of the RPIPx RUN Domain—The N-terminal region (residues 1-298) of mouse RPIP8, which contains the RUN domain (residues 53-189), is involved in a specific interaction with Rap2 (Fig. 1A) (13Janoueix-Lerosey I. Pasheva E. de Tand M.F. Tavitian A. de Gunzburg J. Eur. J. Biochem. 1998; 252: 290-298Crossref PubMed Scopus (43) Google Scholar). In mouse RPIPx, residues 113-245 were predicted as the RUN domain. To conduct a structural analysis of the RUN domain of RPIPx, we carried out a screening of folded proteins using the E. coli cell-free system and NMR. Because the formation of the GFP chromophore is related to the correct folding of the protein that is expressed as a fusion with GFP (40Waldo G.S. Standish B., M. Berendzen J. Terwilliger T.C. Nat. Biotechnol. 1999; 17: 691-695Crossref PubMed Scopus (723) Google Scholar, 41Pedelacq J.D. Piltch E. Liong E.C. Berendzen J. Kim C.Y. Rho B.S. Park M.S. Terwilliger T.C. Waldo G.S. Nat. Biotechnol. 2002; 20: 927-932Crossref PubMed Scopus (143) Google Scholar), we used this system for the primary screening of soluble proteins generated by the cell-free system. First, we expressed a series of RPIPx RUN domains with variable lengths of flanking sequences (Fig. 1B) as C-terminal fusions with GFP. We measured the fluorescence of the cell-free extract expressing each protein (Fig. 1C). Four proteins starting with residue 83 (residues 83-245, 83-255, 83-265, and 83-275) showed high levels of fluorescence, suggesting that these proteins are soluble. Next, we expressed the RPIPx RUN domains without GFP at the C terminus and investigated their solubility by SDS-PAGE (Fig. 1D). This revealed that the four proteins starting with residue 83 were highly expressed in the soluble fractions, which is consistent with the results in Fig. 1C. These results indicate that residues 83-112, including 30-residues upstream of the predicted RUN domain (Fig. 1B), are necessary for the solubility of the RPIPx RUN domain. Finally, to investigate the protein folding, we partially purified the soluble proteins that we identified (residues 83-245, 83-255, 83-265, and 83-275) and measured their 1H and 1H,15N HSQC spectra (Fig. 2). In the 1H spectra of the RPIPx RUN domains that end with residues 255, 265, and 275 (Fig. 2B-D), signals higher than 0.5 ppm, which seem to be up-field shifted methyl signals, were visible. In addition, the signals were well dispersed in the 1H,15N HSQC spectra of these proteins. These results strongly suggest that the RPIPx RUN domains corresponding to residues 83-255, 83-265, and 83-275 are correctly folded. On the other hand, in the 1H spectrum of the RPIPx RUN domain corresponding to residues 83-245 (Fig. 2A), signals higher than 0.5 ppm were not visible. In addition, signal broadening occurred in the 1H,15N HSQC spectrum, suggesting that this protein is not correctly folded. These results indicate that residues 246-255 are required for the correct folding of the RPIPx RUN domain (Fig. 1B). Therefore, we identified residues 83-255 of RPIPx as the minimum fragment that forms a correctly folded protein. The RPIPx RUN Domain Binds to Rap2—We next investigated whether the RUN domain of RPIPx binds to Rap2 in vitro. The structural domain we identified, the RPIPx RUN domain (residues 83-255), was used for the binding analysis. Because the two Rap2 proteins (Rap2A and B) share a sequence identity of 95% and have an identical effector region, only Rap2B was used for the binding analysis. A His-tagged RPIPx RUN domain was examined for its interaction with an immobilized GST fusion protein of Rap2B bound with GDP, GTPγS, or GDPNP (Fig. 3). The His-tagged RPIPx RUN domain interacted with the GTPγS-bound and GDPNP-bound forms of GST-Rap2B but not with the GDP-bound form of GST-Rap2B. This indicates that the RPIPx RUN domain interacts with the activated form of Rap2. This interaction is Rap2-specific because the His-tagged RPIPx RUN domain did not interact with the GDP-bound, GTPγS-bound, or GDPNP-bound form of GST-Rap1A. Crystal Structure of the RPIPx RUN Domain—Because the RPIPx RUN domain is relatively large (173 residues) for a structure determination by NMR, we crystallized the RPIPx RUN domain and determined its structure. The crystal belongs to the primitive hexagonal space group P6422, with unit cell constants of a = b = 86.52 Å, c = 106.58 Å and contains one protein molecule per asymmetric unit. The structure of the RPIPx RUN domain was refined to 2.0 Å by the MAD method. The crystallographic data are summarized in Table 1. The final model includes 162 residues of RPIPx (residues 83-136 and 144-251) (Fig. 4) and 68 water molecules in the asymmetric unit. Two short regions, residues 137-143 and residues 252-255, which are located in a loop region and a C-terminal region of the RUN-domain fragment, respectively, are disordered. The overall structure of the RPIPx RUN domain is shown in a ribbon representation in Fig. 5A. The structure adopts a single globular fold (residues 83-241), consisting of eight α-helices (α1-α8) and four additional regions of 310-helix (η1-η4). Among these helices, α1-α6 and α8 are aligned anti-parallel to one another, making a seven-helix bundle with multiple hydrophobic interactions inside the bundle. α7 is a one-turn α-helix that is also arranged in an anti-parallel manner with the remaining helices. After the C terminus of the globular fold, an extended segment runs from residues 242 to 251. This segment, containing another one-turn α-helix, α9 (residues 244-247), is fixed by hydrophilic contacts with its crystallographic neighbors, forming a tetramer in the crystal (Fig. 5B). The RUN Domain in Solution—To identify the predominant species in solution, the molar mass of the RPIPx RUN domain was determined by analytical ultracentrifugation. Both sedimentation velocity and equilibrium experiments were carried out using UV absorption. The sedimentation velocity data revealed that the RPIPx RUN domain sedimented as a single species, with a molar mass estimated as 19.2 kDa. Sedimentation equilibrium yielded a molar mass value of 20,478 Da. The expected molar mass of the monomer is 19,762 Da. The fit of the data to a single ideal species model is shown in Fig. 6. The analytical centrifugation indicated that the RPIPx RUN domain exists as a monomer in solution. Therefore, the crystal structure of the tetramer, in which the extended segment (residues 242-251) interacts with its crystallographic neighbors (Fig. 5B), is possibly due to crystal packing. Considering that the C-terminal region of the RUN domain fragment (residues 246-255), encompassing the extended segment, is required for the correct folding of the protein (Fig. 2), some intramolecular interactions between the extended segment and the rest of the molecule seem to occur in solution. Residue 241 in one monomer is close (4.26 Å) to that in the other monomer (Fig. 5B), and thus, the intermolecular interactions between the two monomers might represent a possible mode of segment swapping. We propose a model of the monomeric structure of the RPIPx RUN domain in solution in which the extended segment winds around the bundle (Fig. 5C). The RPIPx RUN domain could assume both the crystal and modeled structures, but the equilibrium is probably toward the modeled structure in solution. The RUN-domain Fold—The sequence analysis predicted that the RUN domain is composed of six conserved blocks, which constitute the core of a globular structure (18Callebaut I. de Gunzburg J. Goud B. Mornon J.P. Trends Biochem. Sci. 2001; 26: 79-83Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar). Our structure of the RPIPx RUN domain revealed that the conserved blocks contain seven α helices (α2-α8), which make hydrophobic interactions involving highly conserved residues (Fig. 4). Besides the conserved blocks, α1 also constitutes the core of the RPIPx RUN domain fold (Fig. 5A). The interactions between α1 and the rest of the structure are mostly hydrophobic as are those between the other helices (α2-α8). The interactions between α1 and the other helices involve the side chains of hydrophobic residues, Leu-90, Met-91, Met-93, Ala-94, Ile-98, Leu-101, Ile-102, Ala-105, Leu-106, and Leu-108 in α1, Leu-119, and Val-125 in α2, Ile-226 and Leu-230 in α8, and Leu-233, Val-235, and Ile-236 in η4 (Fig. 4). A comparison of the amino acid sequences of other RUN domains revealed that most of these sites are also occupied by hydrophobic residues in the RUN domains of RPIP8 (13Janoueix-Lerosey I. Pasheva E. de Tand M.F. Tavitian A. de Gunzburg J. Eur. J. Biochem. 1998; 252: 290-298Crossref PubMed Scopus (43) Google Scholar), RPIP9 (14Wang S. Zhang Z. Ying K. Chen J.Z. Meng X.F. Yang Q.S. Xie Y. Mao Y.M. Biochem. Genet. 2003; 41: 13-25Crossref PubMed Scopus (11) Google Scholar), Rabip4 (22Yokoyama S. Hirota H. Kigawa T. Yabuki T. Shirouzu M. Terada T. Ito Y. Matsuo Y. Kuroda Y. Nishimura Y. Kyogoku Y. Miki K. Masui R. Kuramitsu S. Nat. Struct. Biol. 2000; 7: 943-945Crossref PubMed Scopus (325) Google Scholar), UNC-14 (42Ogura K. Shirakawa M. Barnes T.M. Hekimi S. Ohshima Y. Genes Dev. 1997; 11: 1801-1811Crossref PubMed Scopus (78) Google Scholar), and NESCA (4
Referência(s)