|
|
||||||||
1 Structural Biology Laboratory, Nara Institute of Science and Technology, 8916-5, Takayama, Ikoma, Nara 630-0192, Japan
2 CREST, Japan Science and Technology Agency, 8916-5 Takayama, Ikoma, Nara 630-0192, Japan
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Small GTPases possess a common GTPase domain that consisits of about 170 residues with five conserved amino acid motifs, G1G5, which are involved in GTP binding and hydrolysis and base recognition (Vetter & Wittinghofer 2001). The domain exists in two conformational states and behaves as a binary switch that cycles between the active GTP-binding and the inactive GDP-binding forms. The switch function is achieved by large conformation changes in two flexible regions referred to as switches I and II. Compared with other small GTPases, Rad has distinct features in the amino acid sequence, which might affect GTP binding or GTPase activity. The Rad GTPase core domain, which exhibits
35% identity with other small GTPases such as H-Ras, contains crucial substitutions in the conserved motifs, G1-G3 and G4 (Zhu et al. 1995). In fact, recombinant Gem, Rad and Rem are known to display exceedingly low levels of intrinsic GTPase activity (Cohen et al. 1994; Zhu et al. 1995; Finlin et al. 2003). Another notable characteristic of Rad is its N- and C-terminal extensions that consist of 85 and 50 residues, respectively. These extensions contain 14-3-3 and calmodulin binding sites (Moyers et al. 1997; Beguin et al. 2006). The C-terminal extension lacks the typical CAAX prenylation motifs needed for post-translational modifications (Takai et al. 2001). In an effort to better understand the structure-function relationship of these novel GTPases, we have determined the first crystal structure of a member of the RGK-family GTPases, human Rad GTPase in complex with GDP at 1.8 Å resolution.
| Results |
|---|
|
|
|---|
Our crystal contained two Rad molecules, molecules A and B, in the crystallographic asymmetric unit of a P21 crystal lattice. These two molecules form a dimer in the crystal lattice, though Rad exists as a monomer in solution, confirmed by gel filtration giving a single peak at 19 kDa (data not shown). The structure of these two molecules is essentially the same with a small root mean square (r.m.s) deviation of 0.30 Å. Rad contains five
-helices (A1A5) and one large ß-sheet that comprises two extended anti-parallel ß-strands (B2 and B3) and five extended parallel ß-strands (B3, B1, B4B6) (Fig. 1A). These major features of the Rad GTPase domain fold are basically conserved as found in H-Ras and other related small GTPases, consisting of a six-stranded ß-sheet surrounded by helices connected with loops. The bound GDP molecule was found at the canonical nucleotide-binding site (Fig. 1B). The current model contains three poorly defined loop regions, switches I and II and loop A3B5, which were omitted from the model (Fig. 1A,C). Excluding residues of poorly defined loops and altered helix A2, the C
-carbon atoms of Rad when superimposed on the GDP-bound forms of H-Ras (Tong et al. 1991) and RalA (Nicely et al. 2004) yield relatively small r.m.s. deviations of 0.79 Å and 0.86 Å for 124 and 128 C
-carbon atoms, respectively (Fig. 2A). Thus, the Rad GTPase domain displays no significant changes in the overall fold compared to H-Ras and other Ras-family GTPases such as RalA, though there are several unusual features in the detailed structures which will be described below.
|
|
The switch I region of small GTPases generally contains the G2 motif (FVXXYXPTXXDXY where X represents any amino acid residue) that is important for nucleotide binding. The Rad switch I region, however, lacks this motif (Fig. 1C). In our crystal, switch I was disordered and most of the residues are invisible in the current map (Fig. 2B). Helix A1, which is followed by switch I, is important for stabilization of the switch I conformation. Helix A1 of Rad is shorter than that of H-Ras, being hampered by two glycine residues (112 and 113) at the C-terminus of the helix. These small residues cause a small shift of the C-terminal part of helix A1 towards the ß-sheet (Fig. 2B). In H-Ras, Ile24 exists at the position corresponding to Gly112 of Rad and stabilizes helix A1 by making contacts with the ß-sheet (Fig. 2C). In addition to these changes in configuration, Rad helix A1 has a polar residue, Arg109, on the helix surface. This position is usually occupied by a non-polar residue, Ile21 in H-Ras, which stabilizes the loop conformation of switch I through hydrophobic interactions with two conserved residues, Val29 and Tyr32, of the G2 motif. Moreover, a hydrogen bond between conserved tyrosine residues Tyr32 and Tyr40 stabilizes the loop. The absence of the conserved valine and tyrosine residues, together with the relatively high content of mobile glycine residues, might impart flexibility to the Rad switch I region.
Rad switch I lacks the conserved phenylalanine (Phe28 of H-Ras) that traps the GDP molecule through non-polar interactions with the base and sugar. Instead, Arg109 on helix A1 serves to hold the nucleotide by forming a water-mediated hydrogen bonding network (Fig. 2B).
The switch II region
Switch II of Rad is also disordered in our crystal, suggesting conformational flexibility. The G3 motif comprises the N-terminal half of switch II. Rad possesses marked substitutions within this motif (DIWEQD) compared to the common sequence (DTAGQE) that is completely conserved in H-Ras and RalA. At the beginning of the loop, two hydrophobic residues, Ile145 and Trp146, of Rad pack against helices A2 and A3 to form a compact hydrophobic core (Fig. 3A). Ile145 interacts with Met161, which would result in unwinding of the C-terminus of helix A2. These interactions seem to induce a shift in orientation of helix A2 compared with that found in H-Ras. Rad displays a similar orientation of helix A2 to that of RalA, though the length of the helix is much shorter than that of RalA (Nicely et al. 2004). This difference in helix length is caused by the introduction of a helix breaker, Pro155, at the N-terminus of Rad helix A2.
|
The phosphate-binding loop
The phosphate-binding loop of Rad possesses a G1 motif (GXPXXGKSXL) that has a proline substitution, Pro100, at the second conserved glycine residue (G12 of Ras) of the common G1 motif (GXGXXGKSXL) (Fig. 1C). The location of this glycine residue is well known for the site of mutation to produce a transforming form of H-Ras. Crystal structures of mutated H-Ras GTPases show that the mutations cause no significant conformational change in the phosphate-binding loop. Similar results were obtained in Rad, which displays the G1 motif well overlaid on those of H-Ras and RalA with extremely small r.m.s. deviations (0.160.18 Å) (Fig. 4). This structural conservation is coincident with the fact that the Rad phosphate-binding loop retains main chain interactions with the ß-phosphate of GDP as in H-Ras.
|
Magnesium ion binding
In our Rad-GDP complex, the magnesium ion is located at the same position as in the H-Ras-GDP complex. The superposition of Rad and H-Ras proteins yields a small deviation of 0.37 Å for magnesium ions. The ion possesses the same octahedral coordination shell that includes four water molecules, one ß-phosphate oxygen atom and the side chain of Ser105 (Ser17 of H-Ras) (Fig. 3). Conserved Ser105 is located in the G1 motif and is stabilized by conserved Asp144 (Asp57 of H-Ras) of the G3 motif. Mutation of Ser105 with asparagine in Rad abolishes GTP binding (Zhu et al. 1995). Taking into account the lack of switch Inucleotide interactions, the magnesium ion seems to be more critical for GDP binding to Rad than in other GTPases. Moreover, switch I lacks the conserved threonine residue (Thr35 of H-Ras) that is well known to coordinate to the magnesium ion in the GTP-bound form, suggesting relatively weak binding of the magnesium ion to the GTP-bound form. This idea is consistent with the fact that Rad displays unique magnesium dependence where GDP and GTP binding were optimal at relatively high magnesium ion concentrations (110 mM) (Zhu et al. 1995).
The base recognition
The GDP molecule bound to Rad displays an anti-glycosyl conformation with a C2-endo sugar pucker, basically identical to that of GDP conformations bound to other small GTPases including H-Ras. The guanine base is trapped in a hydrophobic pocket, in a manner similar to H-Ras, and participates in several interactions with conserved residues (Asn203, Lys204, Asp206 and Ala234) of the G4 and G5 motifs (Fig. 5A). The hydrogen bonding network between the base and these two motifs is the same as in H-Ras. The superposition of Rad on H-Ras yields an r.m.s. deviation of 0.95 Å for atoms of the bound GDP molecules. A notable difference in the base recognition site might be pointed out in the conformational stability of the G5 motif, in addition to the absence of contacts with the conserved phenylalanine (Phe28 of H-Ras) in switch I as mentioned above. The Rad G5 motif (ETSAAL) replaces the conserved threonine residue (EXSAXT) with a leucine residue, Leu236. In H-Ras and other small GTPases, the conserved threonine residue (Thr148 of H-Ras) plays an important role in stabilizing the base recognition loop of the G5 motif by forming intraloop hydrogen bonds (Fig. 5B). Moreover, H-Ras possesses a lysine residue at the variant position of the G5 motif (ETSAKT), while Rad possesses an alanine residue at this site. This lysine residue of H-Ras, Lys147, forms a hydrogen bond to Asp119 and contributes to stabilization of the side-chain conformation of Asp119 that participates in guanine base recognition. The absence of these interactions in Rad might impart greater flexibility in the base recognition site formed by the G4 and G5 motifs. This might be related to the observation that GTP binding to Rad was slightly inhibited in the presence of high ATP concentrations (Zhu et al. 1995).
|
| Discussion |
|---|
|
|
|---|
GTPase-activating proteins (GAPs) generally interact with both switches I and II of target GTPases for their stimulation activity (Vetter & Wittinghofer 2001; Hakoshima et al. 2003). Since the conformational property of Rad switches I and II is distinct from the Ras-family members including H-Ras, Rad should resist the action of Ras-specific GAPs. Indeed, known GAPs such as Ras-GAP, NF1, p190, Rap-GAP and IQGAP1 failed to stimulate Rad GTPase activity (Zhu et al. 1995). nm23-H1, found recently, was reported to exert a GAP action against Rad as well as a GEF action (Zhu et al. 1999).
Our structure of the GDP-bound form of Rad revealed uncommon structural features of the human Rad GTPase domain, especially conformational flexibility in the switch I region. These findings provide valuable clues for further biochemical and biomedical analyses of this novel small GTPase that is closely engaged in cytoskeletal rearrangement, ion channel regulation, diabetes and cancer.
| Experimental procedures |
|---|
|
|
|---|
Over-expression, purification and crystallization of the GTPase core domain (residues 87258) of human Rad was carried out according to previously described methods (Yanuar et al. 2005). The crystals were grown for 23 days at 20 °C in a hanging drop of 1 : 1 volume mixture of 11 mg/mL protein solution and precipitant solution of 25% PEG 3350, 100 mM MOPS (pH 7.0), 25 mM MgCl2 and 10 mM DTT. The crystals were found to belong to space group P21, with unit-cell parameters a = 52.2 Å, b = 58.6 Å, c = 53.4Å, ß = 98.0Å. A diffraction data set was collected to a resolution of 1.8Å at SPring-8, Japan on beamline BL41XU using an ADSC Quantum 315 CCD detector system and the data was processed using the HKL2000 program suite (Otwinowski & Minor 1997).
Structural determination
Using the program MODELLER (Sali & Blundell 1993), starting models for molecular replacement were obtained by homology modeling with Ras-family GTPases, H-Ras (PDB code 1JAH), Rap1A (1GUA), Rap2 (3RAP) and RalA (1U90 and 1UAD). These GTPases exhibit between 34 and 37% sequence identity with Rad. The structural analysis was carried out using polyalanine models with the molecular replacement program Molrep (CCP41994). Prime-and-switch phasing was performed to remove model bias using the program RESOLVE (Terwilliger 2003) prior to creating electron denstity maps. Density modification with solvent flattening, histogram matching and-non-crystallographic symmetry averaging was also performed to improve electron density maps. Model building was performed using the program O (Jones et al. 1991). Difference Fourier maps were calculated for model building of the bound GDP molecules and picking magnesium ions. The structure was refined with the CNS program (Brünger et al. 1998). Non-crystallographic symmetry was treated during the refinement process until an R-value of 31.7% (an Rfree of 32.0%) was achieved. After several cycles of refinement, the structure was refined with an R-value of 21.4% (an Rfree of 24.0%) without any residues in disallowed regions as checked by the program PROCHECK (Laskowski et al. 1993). The results of the data collection and refinement of the structure are summarized in Table 1. The structure was inspected and rendered using the program PyMOL (DeLano & Bromberg 2004).
|
| Acknowledgements |
|---|
| Footnotes |
|---|
Communicated by: Masao Tasaka
* Correspondence: E-mail: hakosima{at}bs.naist.jp
| References |
|---|
|
|
|---|
Beguin, P., Nagashima, K., Gonoi, T., et al. (2001) Regulation of Ca2+ channel expression at the cell surface by the small G-protein kir/Gem. Nature 411, 701706.[CrossRef][Medline]
Brünger, A.T., Adams, P.D., Clore, G.M., et al. (1998) Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905921.[CrossRef][Medline]
CCP4 (Collaborative Computational Project, Number, 4). (1994) The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D Biol. Crystallogr. 50, 760763.[CrossRef][Medline]
Cohen, L., Mohr, R., Chen, Y.Y., et al. (1994) Transcriptional activation of a ras-like gene (kir) by oncogenic tyrosine kinases. Proc. Natl. Acad. Sci. USA
91, 1244812452.
DeLano, W.L. & Bromberg, S. (2004) Pymol. User's Guide (http://www.pymol.org/), San Carlos, California: DeLano Scientific LLC.
Finlin, B.S., Crump, S.M., Satin, J. & Andres, D.A. (2003) Regulation of voltage-gated calcium channel activity by the Rem and Rad GTPases. Proc. Natl. Acad. Sci. USA
100, 1446914474.
Hakoshima, T., Shimizu, T. & Maesaki, R. (2003) Structural basis of the Rho GTPase signaling. J. Biochem. (Tokyo)
134, 327331.
Ihara, K., Muraguchi, S., Kato, M., et al. (1998) Crystal structure of human RhoA in a dominantly active form complexed with a GTP analogue. J. Biol. Chem.
273, 96569666.
Jones, T.A., Zou, J.-Y., Cowan, S.W. & Kjeldgaard, M. (1991) Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A Found. Crystallogr. 47, 110119.[CrossRef]
Kelly, K. (2005) The RGK family: a regulatory tail of small GTP-binding proteins. Trends Cell Biol. 15, 640643.[CrossRef][Medline]
Krengel, U., Schlichting, L., Scherer, A., et al. (1990) Three-dimensional structures of H-ras p21 mutants: molecular basis for their inability to function as signal switch molecules. Cell 62, 539548.[CrossRef][Medline]
Laskowski, R.A., MacArthur, M.W., Moss, D.S. & Thornton, J.M. (1993) PROCHECK: a program to check the stereochemical quality of protein structures J. Appl. Crystallogr. 26, 283291.[CrossRef]
Moyers, J.S., Bilan, P.J., Reynet, C. & Kahn, C.R. (1996) Overexpression of Rad inhibits glucose uptake in cultured muscle and fat cells. J. Biol. Chem.
271, 2311123116.
Moyers, J.S., Bilan, P.J., Zhu, J. & Kahn, C.R. (1997) Rad and Rad-related GTPases interact with calmodulin and calmodulin-dependent protein kinase II. J. Biol. Chem.
272, 1183211839.
Nicely, N.I., Kosak, J., de Serrano, V. & Mattos, C. (2004) Crystal structures of Ral-GppNHp and Ral-GDP reveal two binding sites that are also present in Ras and Rap. Structure 12, 20252036.[Medline]
Olson, M.F. (2002) Gem GTPase: between a ROCK and a hard place. Curr. Biol. 12, R496R498.[CrossRef][Medline]
Otwinowski, Z. & Minor, W. (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307326.
Pai, E.F., Krengel, U., Petsko, G.A., Goody, R.S., Kabsch, W. & Wittinghofer, A. (1990) Refined crystal structure of the triphosphate conformation of H-ras p21 at 1.35 Å resolution: implications for the mechanism of GTP hydrolysis. EMBO J. 9, 23512359.[Medline]
Reynet, C. & Kahn, C.R. (1993) Rad: a member of the Ras family overexpressed in muscle of type II diabetic humans. Science
262, 14411444.
Sali, A. & Blundell, T.L. (1993) Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234, 779815.[CrossRef][Medline]
Takai, Y., Sasaki, T. & Matozaki, T. (2001) Small GTP-binding proteins. Physiol. Rev.
81, 153208.
Terwilliger, T.C. (2003) Automated main-chain model building by template matching and iterative fragment extension. Acta Crystallogr. D Biol. Crystallogr. 59, 3844.[CrossRef][Medline]
Tong, L.A., de Vos, A.M., Milburn, M.V. & Kim, S.H. (1991) Crystal structures at 2.2 Å resolution of the catalytic domains of normal ras protein and an oncogenic mutant complexed with GDP. J. Mol. Biol. 217, 503516.[CrossRef][Medline]
Tseng, Y.H., Vicent, D., Zhu, J., et al. (2001) Regulation of growth and tumorigenicity of breast cancer cells by the low molecular weight GTPase Rad and nm23. Cancer Res.
61, 20712079.
Vetter, I.R. & Wittinghofer, A. (2001) The guanine nucleotide-binding switch in three dimensions. Science
294, 12991304.
Ward, Y., Yap, S.F., Ravichandran, V., et al. (2002) The GTP binding proteins Gem and Rad are negative regulators of the Rho-Rho kinase pathway. J. Cell Biol.
157, 291302.
Yanuar, A., Sakurai, S., Kitano, K. & Hakoshima, T. (2005) Expression, purification, crystallization and preliminary crystallographic analysis of human Rad GTPase. Acta Crystallogr. F Struct. Biol. Cryst. Commun. 61, 978980.
Zhu, J., Reynet, C., Caldwell, J.S. & Kahn, C.R. (1995) Characterization of Rad, a new member of Ras/GTPase superfamily, and its regulation by a unique GTPase-activating protein (GAP)-like activity. J. Biol. Chem.
270, 48054812.
Zhu, J., Tseng, Y.H., Kantor, J.D., et al. (1999) Interaction of the Ras-related protein associated with diabetes rad and the putative tumor metastasis suppressor NM23 provides a novel mechanism of GTPase regulation. Proc. Natl. Acad. Sci. USA
96, 1491114918.
Received: 12 May 2006
Accepted: 18 May 2006
This article has been cited by other articles:
![]() |
P. Beguin, Y. J. A. Ng, C. Krause, R. N. Mahalakshmi, M. Y. Ng, and W. Hunziker RGK Small GTP-binding Proteins Interact with the Nucleotide Kinase Domain of Ca2+-channel beta-Subunits via an Uncommon Effector Binding Domain J. Biol. Chem., April 13, 2007; 282(15): 11509 - 11520. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | ADVANCED SEARCH | TABLE OF CONTENTS |