Structural Basis for the Autoinhibition and STI-571 Inhibition of c-Kit Tyrosine Kinase
2004; Elsevier BV; Volume: 279; Issue: 30 Linguagem: Inglês
10.1074/jbc.m403319200
ISSN1083-351X
AutoresClifford D. Mol, D.R. Dougan, T. Schneider, R.J. Skene, Michelle L. Kraus, Daniel Scheibe, G. Snell, Hua Zou, Bi‐Ching Sang, Keith P. Wilson,
Tópico(s)Coagulation, Bradykinin, Polyphosphates, and Angioedema
ResumoThe activity of the c-Kit receptor protein-tyrosine kinase is tightly regulated in normal cells, whereas deregulated c-Kit kinase activity is implicated in the pathogenesis of human cancers. The c-Kit juxtamembrane region is known to have an autoinhibitory function; however the precise mechanism by which c-Kit is maintained in an autoinhibited state is not known. We report the 1.9-Å resolution crystal structure of native c-Kit kinase in an autoinhibited conformation and compare it with active c-Kit kinase. Autoinhibited c-Kit is stabilized by the juxtamembrane domain, which inserts into the kinase-active site and disrupts formation of the activated structure. A 1.6-Å crystal structure of c-Kit in complex with STI-571 (Imatinib® or Gleevec®) demonstrates that inhibitor binding disrupts this natural mechanism for maintaining c-Kit in an autoinhibited state. Together, these results provide a structural basis for understanding c-Kit kinase autoinhibition and will facilitate the structure-guided design of specific inhibitors that target the activated and autoinhibited conformations of c-Kit kinase. The activity of the c-Kit receptor protein-tyrosine kinase is tightly regulated in normal cells, whereas deregulated c-Kit kinase activity is implicated in the pathogenesis of human cancers. The c-Kit juxtamembrane region is known to have an autoinhibitory function; however the precise mechanism by which c-Kit is maintained in an autoinhibited state is not known. We report the 1.9-Å resolution crystal structure of native c-Kit kinase in an autoinhibited conformation and compare it with active c-Kit kinase. Autoinhibited c-Kit is stabilized by the juxtamembrane domain, which inserts into the kinase-active site and disrupts formation of the activated structure. A 1.6-Å crystal structure of c-Kit in complex with STI-571 (Imatinib® or Gleevec®) demonstrates that inhibitor binding disrupts this natural mechanism for maintaining c-Kit in an autoinhibited state. Together, these results provide a structural basis for understanding c-Kit kinase autoinhibition and will facilitate the structure-guided design of specific inhibitors that target the activated and autoinhibited conformations of c-Kit kinase. The stem cell factor receptor c-Kit is a receptor protein-tyrosine kinase (RPTK) 1The abbreviations used are: RPTK, receptor protein-tyrosine kinase; KID, kinase insertion domain; MuSK, autoinhibited muscle-specific kinase; MES, 4-morpholineethanesulfonic acid; PTR, phosphotyrosine; SH, Src homolgy; RMS, root mean square. that initiates cell growth and proliferation signal transduction cascades in response to stem cell factor binding (1Linnekin D. Int. J. of Biochem. Cell Biol. 1999; 31: 1053-1074Google Scholar). c-Kit, named after its viral homolog v-Kit (2Besmer P. Murphy J.E. George P.C. Qiu F.H. Bergold P.J. Lederman L. Snyder Jr., H.W. Broudeur D. Zuckerman E.E. Hardy W.D. Nature. 1986; 320: 415-421Google Scholar), is a member of the Type III transmembrane RPTK subfamily, which includes the colony-stimulating factor-1 receptor (3Coussens L. Van Beveren C. Smith D. Chen E. Mitchell R.L. Isacke C.M. Verma I.M. Ullrich A. Nature. 1986; 320: 277-280Google Scholar), also known as the FMS receptor, the related Flt-3 receptor (4Rosnet O. Schiff C. Pebusque M.-J. Marchetto S. Tonnelle C. Toiron Y. Birg F. Birnbaum D. Blood. 1993; 82: 1110-1119Google Scholar), and the platelet-derived growth factor α- and β-receptors (5Claesson-Welsh L. Eriksson A. Westermark B. Heldin C.-H. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 4917-4921Google Scholar, 6Yarden Y. Escobedo J.A. Kuang W.-J. Yang-Feng T.L. Daniel T.O. Tremble P.M. Chen E.Y. Ando M.E. Harkins R.N. Francke U. Fried V.A. Ullrich A. Williams L.T. Nature. 1986; 323: 226-232Google Scholar), as well as c-Kit (7Yarden Y. Kuang W.-J. Yang-Feng T. Coussens L. Munemitsu S. Dull T.J. Chen E. Schlessinger J. Francke U. Ullrich A. EMBO J. 1987; 6: 3341-3351Google Scholar). The Type III RPTK family is characterized by five extracellular immunoglobulin (Ig) domains, a single transmembrane helix, an autoinhibitory juxtamembrane domain, and a cytoplasmic kinase domain that is split by a kinase insertion domain (KID) (see Fig. 1A) (6Yarden Y. Escobedo J.A. Kuang W.-J. Yang-Feng T.L. Daniel T.O. Tremble P.M. Chen E.Y. Ando M.E. Harkins R.N. Francke U. Fried V.A. Ullrich A. Williams L.T. Nature. 1986; 323: 226-232Google Scholar, 8Ullrich A. Schlessinger J. Cell. 1990; 61: 203-212Google Scholar). The binding of a stem cell factor dimer to the extracellular Ig domains of c-Kit causes two c-Kit RPTKs to dimerize and permits the kinase domains to act in trans as a substrate and enzyme for one another. The result of stem cell factor binding is the phosphorylation of specific tyrosine residues located in c-Kit juxtamembrane regions (9Heldin C.-H. Cell. 1995; 80: 213-223Google Scholar, 10Hubbard S.R. Mohammadi M. Schlessinger J. J. Biol. Chem. 1998; 273: 11987-11990Google Scholar, 11Weiss A. Schlessinger J. Cell. 1998; 94: 277-280Google Scholar, 12Blume-Jensen P. Hunter T. Nature. 2001; 411: 355-365Google Scholar). Tyrosine residue 568 is the primary site of in vivo autophosphorylation (see Fig. 1B). Phosphorylation of the tyrosine initiates a cytoplasmic serine/threonine phosphorylation cascade that promotes cell growth and proliferation (12Blume-Jensen P. Hunter T. Nature. 2001; 411: 355-365Google Scholar). Mutations that cause constitutive activation of c-Kit kinase activity in the absence of stem cell factor binding are implicated in highly malignant human cancers, including gastrointestinal stromal tumors (13Hirota S. Isozaki K. Moriyama Y. Hashimoto K. Nishida T. Ishiguro S. Kawano K. Hanada M. Kurata A. Takeda M. Tunio G.M. Matsuzawa Y. Kanakura Y. Sinomura Y. Kitamura Y. Science. 1998; 279: 577-580Google Scholar, 14Nishida T. Hirota S. Taniguchi M. Hashimoto K. Isozaki K. Nakamura H. Kanakura Y. Tanaka T. Takabayashi A. Matsuda H. Kitamura Y. Nat. Genet. 1998; 19: 323-324Google Scholar), germ cell tumors (15Tian Q. Frierson Jr., H.F. Krystal G.W. Moskaluk C.A. Am. J. Pathol. 1999; 154: 1643-1647Google Scholar), mast cell and myeloid leukemias (16Kitamura Y. Hirota S. Nishida T. Mutat. Res. 2001; 477: 165-171Google Scholar), and in mastocytosis (17Longley Jr., B.J. Metcalfe D.D. Tharp M. Wang X. Tyrrell L. Lu S.-Z. Heitjan D. Ma Y. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 1609-1614Google Scholar). Moreover, activating c-Kit mutations that occur in the kinase domain are resistant to many kinase inhibitors currently in use as chemotherapy treatments (18Frost M.J. Ferrao P.T. Hughes T.P. Ashman L.K. Mol. Cancer Ther. 2002; 12: 1115-1124Google Scholar, 19Ma Y. Zeng S. Metcalfe D.D. Akin C. Dimitrijevic S. Butterfield J.H. McMahon G. Longley B.J. Blood. 2002; 99: 1741-1744Google Scholar, 20Ueda S. Ikeda H. Mizuki M. Ishiko J. Matsumura I. Tanaka H. Shibayama H. Sugahara H. Takai E. Zhang X. Machii T. Kanakura Y. Int. J. Hematol. 2002; 76: 427-435Google Scholar, 21Zermati Y. De Sepulveda P. Feger F. Letard S. Kersual J. Casteran N. Gorochev G. Dy M. Ribadeau Dumas A. Dorgham K. Parizot C. Bieche Y. Vidaud M. Lortholary O. Arock M. Hermine O. Dubreuil P. Oncogene. 2003; 22: 660-664Google Scholar). The kinase activity of c-Kit is tightly regulated throughout its signaling cycle. Binding of the protein-tyrosine phosphatase SHP-1 to the phosphorylated c-Kit juxtamembrane region results in dephosphorylation of the tyrosine residues and termination of the intracellular signal (22Kozlowski M. Larose L. Lee F. Le D.M. Rottapel R. Siminovitch K.A. Mol. Cell. Biol. 1998; 18: 2089-2099Google Scholar). The dual role of the juxtamembrane region as an unphosphorylated autoinhibitory domain and as a phosphorylated intracellular signal has made it difficult to dissect the structural and mechanistic functions of the juxtamembrane region in the signaling cascade. Specific site-directed mutations introduced into the juxtamembrane domains of c-Kit (23Ma Y. Cunningham M.E. Wang X. Ghosh I. Regan L. Longley B.J. J. Biol. Chem. 1999; 274: 13399-13402Google Scholar) and platelet-derived growth factor (24Irusta P.M. DiMaio D. EMBO J. 1998; 17: 6912-6923Google Scholar, 25Irusta P.M. Luo Y. Bakht O. Lai C.-C. Smith S.O. DiMaio D. J. Biol. Chem. 2002; 277: 38627-38634Google Scholar) indicate that several residues are necessary to maintain the kinase in an autoinhibited state (see Fig. 1B). Based on amino acid sequence analysis the Type III RPTK juxtamembrane domains are proposed to adopt a conformation similar to WW domains, which are implicated in regulating cellular processes (26Ilsley J.L. Sudol M. Winder S.J. Cell. Signal. 2002; 14: 183-189Google Scholar, 27Macias M.J. Wiesner S. Sudol M. FEBS Lett. 2002; 513: 30-37Google Scholar, 28Sudol M. Hunter T. Cell. 2000; 103: 1001-1004Google Scholar). Similarly, the c-Kit autoinhibitory juxtamembrane region has been proposed to form a putative α-helix that exerts negative control over uninduced receptor (23Ma Y. Cunningham M.E. Wang X. Ghosh I. Regan L. Longley B.J. J. Biol. Chem. 1999; 274: 13399-13402Google Scholar). Studies with a synthetic peptide of the c-Kit juxtamembrane region suggest that it folds as an autonomous domain and directly interacts with the amino-terminal lobe of the kinase domain (29Chan P.M. Ilangumaran S. La Rose J. Chakrabartty A. Rottapel R. Mol. Cell. Biol. 2003; 23: 3067-3078Google Scholar). These models for the autoinhibition of Type III RPTKs are derived from autoregulation mechanisms inferred from the crystal structures of other protein-tyrosine kinases (reviewed in Refs. 30Hubbard S.R. Hill J.H. Annu. Rev. Biochem. 2000; 69: 373-398Google Scholar and 31Hubbard S.R. Curr. Opin. Struct. Biol. 2002; 12: 735-741Google Scholar). Autoinhibition in cis is seen in the crystal structure of the autoinhibited form of the EphB2 receptor tyrosine kinase domain (32Wybenga-Groot L.E. Baskin B. Ong S.H. Tong J. Pawson T. Sicheri F. Cell. 2001; 106: 745-757Google Scholar). This structure contains a truncated juxtamembrane domain with phenylalanine substitution mutations of signaling tyrosine residues similar to those in the juxtamembrane regions of c-Kit and the Type III RPTKs. The visible portion of the EphB2 juxtamembrane domain forms a short α-helix that packs against the amino-terminal lobe of the kinase and disrupts the interactions of the control, or C, helix with residues that form an activated kinase ATP-binding site. Similarly, the crystal structure of the autoinhibited muscle-specific kinase MuSK reveals a small ordered segment of the MuSK juxtamembrane region that forms a short α-helix, which likely impedes the ability of the C-helix to contribute to productive binding of ATP (33Till J.H. Becerra M. Watty A. Lu Y. Ma Y. Neubert T.A. Burden S.J. Hubbard S.R. Structure. 2002; 10: 1187-1196Google Scholar). An autoinhibitory structural role for a juxtamembrane region tyrosine residue was also inferred from the crystal structure of the insulin receptor in which a similar disruption of the conformation of the C-helix was observed (34Li S. Covino N.D. Stein E.G. Till J.H. Hubbard S.R. J. Biol. Chem. 2003; 278: 26007-26014Google Scholar). Targeting these distinct inactive kinase conformations for a rational drug design may allow for the design of tight binding and specific compounds. STI-571, also known as Imatinib® or Glivec®/Gleevec®, is one such drug compound that specifically binds to the inactive conformation of the Abl kinase. Abl kinase is directly implicated in the pathogenesis of chronic myelogenous leukemia (35O'Dwyer M.E. Mauro M.J Druker B.J. Cancer Investig. 2003; 3: 429-438Google Scholar). Notably, STI-571 does not inhibit many other kinases, but it does inhibit the two closely related Type III RPTKs, platelet-derived growth factor and c-Kit (36Buchdunger E. Cioffi C.L. Law N. Stover D. Ohno-Jones S. Druker B.J. Lydon N.B. J. Pharmacol. Exp. Ther. 2000; 295: 139-145Google Scholar). Recently we reported (37Mol C.D. Lim K.B. Sridhar V. Zou H. Chien E.Y. Sang B.-C. Nowakowski J. Kassel D.B. Cronin C.N. Mcree D.E. J. Biol. Chem. 2003; 278: 31461-31464Google Scholar) the 2.9-Å resolution crystal structure of an activated c-Kit kinase domain-product complex bound in trans with juxtamembrane phosphotyrosine residues and ADP and Mg2+. For comparative analysis, we recapitulate that structure here at an improved 2.65-Å resolution. We also present the 1.9-Å resolution crystal structure of unphosphorylated c-Kit kinase containing the entire autoinhibitory juxtamembrane region. The juxtamembrane domain inserts directly into the cleft between the kinase amino- and carboxyl-terminal lobes, disrupting the c-Kit control helix, and physically blocking the conserved kinase Asp-Phe-Gly (DFG) motif from attaining a productive conformation. The kinase activation loop folds back over the substrate-binding groove and interacts with the kinase-active center as a pseudosubstrate. We also present a 1.6-Å resolution co-crystal structure of a c-Kit·STI-571 complex that illustrates that portions of the inhibitor would clash with regions of the juxtamembrane domain that maintain c-Kit in the autoinhibited conformation. These results provide the detailed molecular basis for understanding the mechanism of c-Kit kinase autoinhibition and will facilitate the structure-guided design of specific and potent inhibitors that target the activated and autoinhibited conformations of c-Kit kinase. Protein Expression and Purification—The portion of the human c-Kit gene that comprises the catalytic kinase domains and the entire intact juxtamembrane domain (residues 544–935, GenBank™ accession number NM_000222 (3Coussens L. Van Beveren C. Smith D. Chen E. Mitchell R.L. Isacke C.M. Verma I.M. Ullrich A. Nature. 1986; 320: 277-280Google Scholar)) with the KID residues 694–753 deleted was cloned and expressed as described previously (37Mol C.D. Lim K.B. Sridhar V. Zou H. Chien E.Y. Sang B.-C. Nowakowski J. Kassel D.B. Cronin C.N. Mcree D.E. J. Biol. Chem. 2003; 278: 31461-31464Google Scholar). Recombinant c-Kit protein was expressed in Spodoptera frugiperda cells in 5L Wave BioReactors with ESF-921 protein-free media (Expression Systems), and the cells were harvested after 2 days. The protein yield was 3.77 mg/liter and the recombinant protein, containing an amino-terminal His6 tag followed by an rTEV protease cleavage site, was purified by Ni2+-chelate chromatography, the tag was removed with rTEV protease, and the tag and any uncleaved material were removed by a second passage over a ProBond Ni2+-chelate column. Purified c-Kit protein was judged to be >95% pure by SDS-PAGE and was concentrated to 6–12 mg/ml in buffer containing 25 mm Tris, pH 7.6, 250 mm NaCl, 5 mm dithiothreitol, and 1 mm EDTA. Purified and concentrated c-Kit protein was flash-frozen in 50-μl aliquots by direct immersion in liquid nitrogen and stored at -80 °C. Crystallization and Data Collection—Crystals of activated c-Kit kinase domain (∼0.2 × 0.2 × 0.05 mm) are grown at 20 °C as described (37Mol C.D. Lim K.B. Sridhar V. Zou H. Chien E.Y. Sang B.-C. Nowakowski J. Kassel D.B. Cronin C.N. Mcree D.E. J. Biol. Chem. 2003; 278: 31461-31464Google Scholar) by pre-incubating enzyme samples (6 mg/ml) with ATP and MgCl2 prior to performing sitting-drop crystallization experiments using 5 μl of protein solution and 5 μl of reservoir solution (18% polyethylene glycol 8000, 0.1 m MES, pH 7.1). Crystals of autoinhibited c-Kit kinase (∼0.1 × 0.1 × 0.2 mm) are obtained at 20 °C from the same protein preparation (10 mg/ml) using Nanovolume Crystallization® (38Hosfield D. Palan J. Hilgers M. Scheibe D. McRee D.E. Stevens R.C. J. Struct. Biol. 2003; 142: 207-217Google Scholar) techniques with 50 nl of protein solution and 50 nl of reservoir solution containing 13% polyethylene glycol 8000, 0.1 m HEPES, pH 7.0. For crystals of the co-complex with STI-571 (∼0.1 × 0.1 × 0.1 mm), 50 nl of enzyme-inhibitor solution and 50 nl of reservoir (2 m phosphate, 0.1 m Tris, pH 8.5) were used, and the crystals were grown at 4 °C. Single crystals of all enzyme forms were harvested in reservoir solutions supplemented with 25% ethylene glycol, and flash-frozen by direct immersion in liquid nitrogen. X-ray diffraction data were collected at the Advanced Light Source Beam Line 5.0.3 equipped with an ADSC Quantum4 CCD detector, and the diffraction intensities were integrated and scaled using the HKL2000 program suite (39Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326PubMed Google Scholar). The activated c-Kit kinase crystals belong to the orthorhombic space group P21212 with unit cell dimensions a = 95.3 Å, b = 119.5 Å, and c = 62.6Å and contain two enzyme molecules in the asymmetric unit. The autoinhibited c-Kit kinase crystals belong to the orthorhombic space group P212121 with unit cell dimensions a = 44.4 Å, b = 77.2 Å, and c = 94.6 Å and possess one enzyme molecule in the asymmetric unit. The c-Kit·STI-571 complex crystals are in the trigonal space group P3221 with unit cell dimensions a = b = 70.1 Å and c = 127.9 Å and contain one enzyme-inhibitor co-complex in the asymmetric unit. Structure Determination and Refinement—The structure of activated c-Kit kinase was determined by molecular replacement using AMoRe (40Navaza J. Acta Crystallogr. Sect. D. 2001; 57: 1367-1372Google Scholar) and refined as described previously (37Mol C.D. Lim K.B. Sridhar V. Zou H. Chien E.Y. Sang B.-C. Nowakowski J. Kassel D.B. Cronin C.N. Mcree D.E. J. Biol. Chem. 2003; 278: 31461-31464Google Scholar). This structure contains two c-Kit enzyme molecules in the asymmetric unit and clear electron density for residues 567–935 of both enzymes, with the exception of ∼20 amino-terminal amino acid residues as well as ∼10 residues comprising the truncated KID, which are disordered. The structure of autoinhibited c-Kit kinase contains only one enzyme molecule in the asymmetric unit and was also determined by molecular replacement with AMoRe, using the activated c-Kit kinase structure as a search model with the correct solution yielding a correlation coefficient of 0.30 and an initial R-value of 0.50. The initial solutions were refined with REFMAC (41Murshudov G.N. Vagin A.A. Dodson E.J. Acta Crystallogr. Sect. D. 1997; 53: 240-255Google Scholar), and the models were visually inspected and manually built and rebuilt using the XtalView/Xfit program suite (42McRee D.E. J. Struct. Biol. 1999; 125: 156-165Google Scholar). The structure of autoinhibited c-Kit kinase is fully ordered with unambiguous electron density for residues 548–935, including the truncated KID residues. For the structure of the c-Kit·STI-571 complex two molecular replacement calculations were performed utilizing the activated structure (correlation coefficient 0.28, R-value 0.47) and the autoinhibited structure (correlation coefficient 0.39, R-value 0.44). The autoinhibited structure was clearly the better model, and a refinement of this model yielded clear unambiguous electron density for the bound inhibitor. The x-ray data collection and crystallographic refinement statistics for these three structures are presented in Table I. The coordinates have been deposited in the Protein Data Bank, 1PKG (activated), 1T45 (autoinhibited), and 1T46 (STI-571 co-complex).Table IX-ray data collection and refinement statisticsAutoinhibitedActivatedSTI-571Data collectionResolution (Å)30.0-1.90 (1.97-1.90)30.0-2.65 (2.74-2.65)30.0-1.60 (1.66-1.60)Observations107,29998,443220,014Unique reflections26,26021,24348,006Completeness (%)98.3 (96.5)99.3 (95.1)98.4 (90.0)I/σI12.6 (3.1)15.4 (2.0)19.3 (2.6)RsymaRsym = ΣhΣj|〈I(h)〉 - I(h)j |/ ΣhΣj 〈I(h)〉 where 〈I(h)〉 is the mean intensity of symmetry-related reflections0.058 (0.345)0.046 (0.493)0.028 (0.367)RefinementResolution (Å)20.0-1.9010.0-2.6520.0-1.60Reflections used24,83519,68745,483Number of protein atoms2,6894,6722,381Number of ligand atoms05737Number of solvent atoms20823268RMS bonds (Å)0.0090.0120.007RMS angles (°)1.1121.3351.085Average B-value (Å2)16.954.923.6R-value, RfreebR-value = Σ||Fobs| - |Fcalc||/ Σ |Fobs|; Rfree for 5% of reflections excluded from refinement. Values in parentheses are for the respective high resolution shells0.193, 0.2220.223, 0.2760.188, 0.213a Rsym = ΣhΣj|〈I(h)〉 - I(h)j |/ ΣhΣj 〈I(h)〉 where 〈I(h)〉 is the mean intensity of symmetry-related reflectionsb R-value = Σ||Fobs| - |Fcalc||/ Σ |Fobs|; Rfree for 5% of reflections excluded from refinement. Values in parentheses are for the respective high resolution shells Open table in a new tab Conformational Flexibility Calculations—Coordinate uncertainties were estimated using the B-factor scaled version of Cruickshank's diffraction-component precision index (43Cruickshank D.W. Acta Crystallogr. Sect. D. 1999; 55: 583-601Google Scholar), which is based on B-factors including the translation, liberation, and screw-rotation contributions (44Schneider T.R. Acta Crystallogr. Sect. D. 2002; 58: 195-208Google Scholar). The coordinate uncertainties for the activated conformation ranged from 0.246 to 0.688 Å with a mean of 0.365 ± 0.076 Å. The coordinate uncertainties for the autoinhibited conformation were significantly smaller reflecting the higher resolution of the diffraction data used to refine the structure, values ranged from 0.080 to 0.313 Å with a mean of 0.138 ± 0.048 Å. Structure of Active c-Kit Kinase—The activated and autoinhibited c-Kit kinase structures were obtained from constructs containing an intact juxtamembrane region (approximately residues from Thr544 to Trp580) followed by the kinase domains. The KID is a region of variable length and unknown function that splits the coding sequence for the C-lobe of many RPTKs, and is truncated in these constructs (see"Experimental Procedures"). Crystals of the active c-Kit kinase are obtained by pre-incubating purified protein samples with ATP and Mg2+ to initiate the trans autophosphorylation reaction. Analysis of the reaction products by mass spectrometry (37Mol C.D. Lim K.B. Sridhar V. Zou H. Chien E.Y. Sang B.-C. Nowakowski J. Kassel D.B. Cronin C.N. Mcree D.E. J. Biol. Chem. 2003; 278: 31461-31464Google Scholar) shows that the primary sites of phosphorylation are Tyr568 and Tyr570, which are located near the junction of the juxtamembrane domain and the N-lobe of the kinase (Fig. 1B). The x-ray data collection and 2.65-Å resolution refinement statistics for activated c-Kit are presented in Table I. This structure contains two c-Kit enzyme molecules in the crystallographic asymmetric unit, with phosphotyrosine (PTR)-568 from the amino termini of adjacent molecules in the crystal lattice binding in trans at the kinase-active center. In the active c-Kit kinase structure, the ∼20 amino acid residues of the juxtamembrane domain preceding Tyr568 (Fig. 1B) of both enzymes in the crystallographic asymmetric unit are disordered as are the residues of the truncated KID. The structure of active c-Kit kinase domain is consistent with the common protein kinase fold (Fig. 2A). The smaller N-lobe is comprised of mostly β-strands and a single α-helix, the C-helix, that modulates activity via contacts with the nucleotide binding site and DFG motif at the base of the activation loop (A-loop) (Fig. 2, A and B). Residue Glu640 of the C-helix of active c-Kit provides a critical salt link to the buried side chain of Lys623, which bridges the α- and β-phosphates of the bound ADP. The c-Kit DFG motif is also in an active conformation, with Asp810 coordinating the Mg2+ ion bridging the α- and β-phosphates, and Phe811 positioned to allow binding of the adenine moiety of the nucleotide (Fig. 3A).Fig. 3On and off conformations of the c-Kit kinase A-loop DFG motif. A, stereo view of the nucleotide-bound c-Kit kinase-active center. The DFG motif (Asp810, Phe811, Gly812) is in a conformation that allows nucleotide binding with Asp810 ligating the bound Mg2+ ion (pink sphere). In this product complex structure, the P-loop (orange) with Phe600 shields the active center, and the γ-phosphate group of ATP has been transferred to the target tyrosine residue PTR-568. B, stereo view of the autoinhibited c-Kit kinase-active center. The penetration of the autoinhibitory juxtamembrane region (purple) inserts Trp557 into the area that Phe811 occupies in the activated structure. The DFG motif is in the autoinhibited off state with Phe811 flipped over and occluding the nucleotide binding site. In this pseudosubstrate complex, The A-loop (green) is folded back and the P-loop (orange, upper left) is shifted from its nucleotide-bound position.View Large Image Figure ViewerDownload (PPT) The active c-Kit kinase enzyme structure is a product complex, with ADP, a Mg2+ ion, and the side chain of PTR-568 bound at the kinase-active site. As PTR-568 is covalently attached to another enzyme molecule in the crystal, substrate peptide binding in trans is restricted by crystal packing interactions. Although the details of phosphotyrosine binding at the active center is identical in both molecules of the asymmetric unit, the global course of the substrate peptide binding differs. We have focused our discussion on substrate peptide binding to molecule A, as in this configuration the peptide packs against and stabilizes the active conformation of the A-loop (37Mol C.D. Lim K.B. Sridhar V. Zou H. Chien E.Y. Sang B.-C. Nowakowski J. Kassel D.B. Cronin C.N. Mcree D.E. J. Biol. Chem. 2003; 278: 31461-31464Google Scholar). The activation loop is in the active conformation, despite the fact that the target A-loop tyrosine, Tyr823, is not phosphorylated. The side chain of Tyr823 is directed toward a positively charged region formed by the guanidinium groups of Arg815 and Arg791. A PTR residue at this position would further stabilize the kinase A-loop in an active conformation. Structure of Autoinhibited c-Kit Kinase—Many protein kinases are maintained in an inactive state in the absence of activating signals, a fact first determined for the Ser/Thr protein kinases (45Kemp B.E. Pearson R.B. Biochim. Biophys. Acta. 1991; 1094: 67-76Google Scholar), and recently observed for the protein-tyrosine kinases (46Huse M. Kuriyan J. Cell. 2002; 109: 275-282Google Scholar). For many non-receptor tyrosine kinases, such as the Abl kinase, the inactive state of the kinase is maintained through SH2 and SH3 protein interaction domains that associate with and inhibit the kinase domain (47Nagar B. Hantschel O. Young M.A. Scheffzek K. Veach D. Bornmann W. Clarkson B. Superti-Furga G. Kuriyan J. Cell. 2003; 112: 859-871Google Scholar). The RPTKs, such as c-Kit, lack these inhibitory protein interaction domains, and the kinase is activated via ligand-mediated receptor dimerization. In addition to its role as a substrate for trans autophosphorylation in vivo, there is compelling evidence that the juxtamembrane domain of the Type III RPTKs also functions as an autoinhibitory domain. Unfortunately, much of this region is disordered in the active structure of c-Kit. The structure of autoinhibited c-Kit kinase, which has an ordered juxtamembrane domain, permits the dual function of the juxtamembrane region to be more clearly understood. The structure of c-Kit in an autoinhibited conformation was determined using the active c-Kit kinase domain structure as a molecular replacement search model. Even at the earliest stages of refinement the electron density maps indicated that a substantial rearrangement and shifting of the structural elements in the kinase N-lobe had occurred (Fig. 2B). The high resolution 1.90-Å data for the autoinhibited structure allowed for the entire polypeptide chain to be fit unambiguously, including the residues for the truncated KID and the aminoterminal residues that were not visible in the activated kinase structure. Contrary to the possibility that the Type III RPTK juxtamembrane regions might adopt a "WW-like" β-sheet domain (24Irusta P.M. DiMaio D. EMBO J. 1998; 17: 6912-6923Google Scholar) or an α-helical structure (23Ma Y. Cunningham M.E. Wang X. Ghosh I. Regan L. Longley B.J. J. Biol. Chem. 1999; 274: 13399-13402Google Scholar), the juxtamembrane domain forms a compact hairpin loop that inserts directly into the domain interface between the kinase N- and C-lobes (Fig. 2B). The ∼20 amino-terminal residues of the juxtamembrane domain, from Tyr547 to Gly565, comprise the inner buried half of this hairpin loop and form specific interactions that disrupt the conformation of the kinase DFG motif and A-loop (see below). The subsequent region that contains the target tyrosine residues of the autophosphorylation reaction extends along the exterior solvent-exposed half of the hairpin loop (Fig. 2B). Juxtamembrane region residues identified by site-directed mutagenesis as critical in maintaining the autoinhibited state (Fig. 2) make important interactions in the interface between the amino-terminal autoinhibitory domain and the formal kinase N- and C-lobes (Fig. 3B). The side chains of Val559 and Val560 stabilize a hydrophobic patch formed by Val643, Tyr646, Cys788, and Ile789, whereas Tyr553 penetrates deeply into the interface (Fig. 3B) and forms hydrogen bonds between its side chain–OH group and the side chains of buried and conserved residues Asp810 of the DFG motif and Glu640. In the active c-Kit kinase structure, the conformation of the polypeptide backbone is such that Phe811 of the DFG motif does not block the nucleotide-binding site (Fig. 3A). However, in the autoinhibited kinase conformation, Phe811 is blocked from this position by Trp557 and occludes the active site. (Fig. 3B). The kinase DFG motif flips between the "on" and "off" states by rotating about the ϕ main chain torsion angle of Asp810, which maintains the Asp810 carboxylate side chain in approximately the same position in the two enzyme
Referência(s)