- Email: [email protected]
Structural basis of the auto-inhibition mechanism of nonreceptor tyrosine kinase PTK6
Biochemical and Biophysical Research Communications 384 (2009) 236–242
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Structural basis of the auto-inhibition mechanism of nonreceptor tyrosine kinase PTK6 Sunggeon Ko a, Kyo-Eun Ahn a, Young-Min Lee a, Hee-Chul Ahn b, Weontae Lee a,* a b
Department of Biochemistry, College of Life Science & Biotechnology, Yonsei University, Seoul 120-749, Republic of Korea Advanced Analysis Center, Korea Institute of Science and Technology, Seoul 136-791, Republic of Korea
a r t i c l e
i n f o
Article history: Received 16 April 2009 Available online 3 May 2009
Keywords: PTK6 Auto-regulation SH3 Intra-molecular interaction NMR spectroscopy
a b s t r a c t Protein tyrosine kinase 6 (PTK6) is composed of SH3, SH2, and Kinase domains, with a linker region (Linker) between the SH2 and Kinase domains. Here, we report the structural basis of the SH3–Linker interaction that results in auto-inhibition of PTK6. The solution structures of the SH3 domain and SH3/Linker complex were determined by NMR spectroscopy. The structure of the SH3 domain forms a conventional b-barrel with two b-sheets comprised of five b-strands. However, the molecular topology and charge distribution of PTK6-SH3 slightly differs from that of the other SH3 domains. The structure of the N-terminal Linker within the complex showed that the proline-rich region (P175–P187) of the Linker forms a compact hairpin structure through hydrophobic interactions. The structure of the SH3/Linker complex revealed intra-molecular interaction between the amino acid pairs R22/E190, W44/W184, N65/P177, and Y66/P179. Mutations in PTK6 at R22, W44, N65, and Y66 residues in the SH3 domain increased catalytic activity compared with wild-type protein, implying that specific interactions between hydrophobic residues in the proline-rich linker region and hydrophobic residues in the SH3 domain are mainly responsible for down-regulating the catalytic activity of PTK6. Ó 2009 Elsevier Inc. All rights reserved.
Introduction Protein tyrosine kinases (PTKs) play an important role as regulators of various cellular functions including proliferation, differentiation, and apoptosis, and are often involved in tumorigenesis through hyper-activation processes [1–3]. Human PTK6 (also known as Brk) was identified during a survey of PTK mRNAs of human melanocytes [4], and shown to be over-expressed in breast carcinomas and colon tumors [5,6]. PTK6 is composed of Src homology 3 (SH3), Src homology 2 (SH2), and tyrosine kinase catalytic (kinase) domains. Phosphorylation of a highly conserved tyrosine residue in the SH2 domain stabilizes an inactive conformation of the protein [7]. The solution structure of the SH2 domain shows that it consists of a consensus a/b-fold with a pTyr peptide binding surface common to other SH2 domains. However, two of the a-helices (aA and aB) are located on opposite faces of the central b-sheet, and the topological arrangement of the central four-stranded b-sheet differs from that of other Src family members [8]. Activation of Src family kinases is modulated by intra-molecular interactions between the SH3 and SH2 domains that lock the
* Corresponding author. Address: Department of Biochemistry, College of Life Science & Biotechnology, Yonsei University, 134 Shinchon-Dong, Seodaemoon-Gu, Seoul 120-749, Republic of Korea. Fax: +82 2 363 2706. E-mail address: [email protected] (W. Lee). 0006-291X/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2009.04.103
kinase into a rigid conformation [9]. Modulation of the kinase activity of PTK6 by intra-molecular interaction and autophosphorylation has been reported, suggesting that interaction between the SH2 and SH3 domains and phosphorylation event play important roles in the auto-regulation of PTK6 [10,11]. Recently, we reported that mutation of key conserved residues in the N-terminal linker region between the SH2 and kinase domains of PTK6 increased its catalytic activity, indicating that the Linker plays an important role in down-regulating kinase activity [12]. Based on molecular modeling and surface plasmon resonance experiments, we proposed that the W44 residue in the SH3 domain of PTK6 participates in intra-molecular interaction of the SH3 domain with the proline-rich region of the Linker [13]. Furthermore, disruption of the intra-molecular interaction between the SH3 domain and Linker by mutation of the W44, P175, P177, and P179 residues increases the kinase activity of PTK6 [12,13]. However, the precise nature of interactions between the SH3 domain and the Linker remains unclear due to lack of information on their three-dimensional structures. Since interactions between the SH3 domain and Linker are mainly responsible for keeping PTK6 in an inactive conformation, it is important to determine the three-dimensional structures of both the SH3 domain and of the Linker in the presence of the SH3 domain. Here, we present solution structures of the SH3 domain and SH3/Linker complex and demonstrate precise molecular interactions between the SH3 domain and Linker. These results provide structure–function information on the mechanism
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
of auto-inhibitory regulation of the nonreceptor tyrosine kinase PTK6. Materials and methods Cloning, protein expression, and purification of PTK6. PTK6 (residues 1–451), PTK6-SH3 domain (residues 1–72), and PTK6-Linker (residues 171–191) were cloned into pGEX 4T-1 (Amersham). Sense primers containing the BamHI site and TEV protease recognition sequence (ENLYFQG), and antisense primers containing the EcoRI site were used for PCR experiments. All plasmids encoding PTK6 proteins were transformed into Escherichia coli BL21(DE3), and over-expression of target proteins was induced by 0.1 mM isopropyl b-D-thiogalactopyranoside (IPTG) at 25 °C. Fusion proteins were purified by GST-affinity chromatography using GSTrap FF columns (GE Healthcare), and the GST-tag was removed by TEV protease. Size exclusion chromatography was performed using
237
Superdex 75 columns. The sample buffer was exchanged for 10 mM ammonium bicarbonate buffer and lyophilized to yield the final sample. For heteronuclear NMR experiments, the SH3 domain was overexpressed in M9 media containing 15NH4Cl or 15NH4Cl/13C-D-glucose (Cambridge Isotope Laboratories) and purified by GST-affinity chromatography and size exclusion chromatography as described above. Peptide synthesis and sample preparation. Peptides corresponding to the Linker (RKHEPEPLPHWDDWEREPEEF; residues 171–191 of the human PTK6 sequence) were synthesized by an improved version of the solid-phase method using Fmoc chemistry (Anygen Inc., Kwangju, Korea). The peptide was purified by high-performance liquid chromatography (HPLC) on a Shimazu Prep LC-8A system with a Shim-pack ODS(C18) prep column. The sample was eluted with a linear gradient (A = 0.1%TFA in water and B = 0.1% TFA in CH3CN) with detection by absorbance at 230 nm. Samples were
Fig. 1. 1H–15N HSQC spectrum and solution structures of the SH3 domain of PTK6. (A) Backbone assignment of the SH3 domain (residues 1–72) is displayed in the 1H–15N HSQC spectrum. (B) The 20 NMR structures were superimposed using the MOLMOL program. All structures were fitted for Ca, C0 , and N atoms of residues 12–70 with respect to the restrained energy minimized (REM) average structure. (C) Ribbon diagram showing molecular topology of the SH3 domain. The five antiparallel b-strands are labeled.
238
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
prepared for NMR by dissolving the peptides in 90% H2O/10% D2O or 99.9% D2O in 50 mM potassium phosphate buffer (pH 7.0) and adjustment of the final peptide concentration to 2–4 mM. NMR experiments and structure calculation of the SH3 domain and Linker peptide. 15N- or 13C/15N-labeled PTK6-SH3 domain was used for heteronuclear NMR experiments. Stable isotope-labeled PTK6SH3 domain was exchanged to NMR buffer (90% H2O/10% D2O, 25 mM HEPES, 100 mM NaCl, pH 7.0) during size exclusion chromatography and concentrated to 0.2 mM using Amicon Ultra filter units (Amicon). 1H–15N heteronuclear single quantum coherence (HSQC), 3D HNCACB, CBCA(CO)NH, HNCA, HBHA(CO)NH, and HNCO experiments were used for backbone assignment [14]. 15N edited NOESY (s = 150 ms), 13C-edited NOESY (s = 150 ms), and HCCH-TOCSY were used for NOE and side chain assignment [15– 17]. For NMR analysis of the Linker, lyophilized peptide was dissolved in NMR buffer to a concentration of 1 mM. Two-dimensional total correlation spectroscopy (TOCSY) with a MLEV-17 mixing pulse of 69.7 ms and two-dimensional nuclear Overhauser effect spectroscopy (NOESY) with mixing times of 100–600 ms were performed for Linker peptide. All NMR experiments were collected in the phase-sensitive mode using the time proportional phase incrementation (TPPI) method with 2048 data points in the t2 and 256 in the t1 domains. To determine the structure of Linker bound to the SH3 domain, we performed transferred nuclear Overhauser effect spectroscopy (TR-NOESY), 2D NOESY (s = 600 ms) and 2D TOCSY [18]. All spectra were obtained using Bruker DRX 500 MHz equipped with CryoprobeTM system and Varian 900 MHz spectrometer. XWIN-NMR, NMRPipe/NMRDraw, and SPARKY software [19] were used for data processing and chemical shift assignment. CYANA 2.1 with torsion angle dynamics was used for structure calculations [20]. Distance constraints and angle constraints from TALOS were also used for final structure refinement. CYANA calculations were executed using a 16-node linux cluster computer, and 20 lowest energy structures were evaluated by the PROCHECK program [21]. Final structures were displayed and analyzed by Pymol (DeLano Scientific LLC) and MOLMOL programs [22]. NMR titration and structure calculation of the SH3/Linker complex. 1 H–15N HSQC titrations were performed to identify the specific residues involved in the interaction between the SH3 domain and linker. We prepared NMR samples containing various molar ratios of SH3:Linker (1:1, 1:3, and 1:5) and a final concentration of 15N-labeled SH3 domain of 0.2 mM. Chemical shift changes were calculated using the equation: Ddtotal = [(DdHN)2 + (DdN/5)2]1/2. The structure of the SH3/Linker complex was calculated by the HADDOCK 2.0 program integrated with CNS 1.2 in a 16-node linux cluster computer [23]. We determined active residues and passive residues based on HSQC titration experiments, GST pull down assay and solvent accessibility calculations in NACCESS program [24]. Active residues of the SH3 domain (R22, H36, W44, H64, N65, and Y66) were defined by chemical shift changes (Dd) above 0.1 in HSQC titration experiments and solvent accessibility above 40%. Active residues of the Linker (P175, P177, and P197) were defined by GST pull down assay [13] and solvent accessibility above 40%. All passive residues were determined by the neighbors to active residue and solvent accessibility above 40%. Clustering analysis of HADDOCK calculation results was executed using a cut off value of RMSD 7.5, 5.0, and 3.0 Å. Final structures were analyzed and displayed using the Pymol program. In vitro kinase assay of PTK6 and mutant proteins. An in vitro kinase assay was used to examine kinase activities of wild-type and mutant PTK6 proteins. PTK6 mutants (R22A, W44A, N65A, and Y66A) were generated by PCR mutagenesis, and proteins were purified as described previously. For the kinase assay, 10 lg PTK6 substrate (poly(Glu,Tyr), Sigma) was incubated in the absence or presence of 20 lg GST-PTK6 protein in 20 ll kinase reaction buffer
(20 mM Tris–HCl, 10 mM MgCl2, 1 mM MnCl2, 50 lM Na3VO4, pH 7.4) containing 100 lM [c-32P] ATP for 30 min at 30 °C. SDS–PAGE and autoradiography with scintillation counting were performed as described previously [12]. Results and discussion Resonance assignments and structure of the PTK6-SH3 domain The backbone resonance assignments of the SH3 domain were completed using data from 1H–15N 2D HSQC, 3D HNCACB, CBCA(CO)NH, HNCO, and HCACO spectra (Fig. 1A). The side-chain and NOE assignments were completed by combined use of 3D HCCHTOCSY, 15N-edited TOCSY-HSQC, and 13C/15N-edited NOESY experiments. The secondary structures were determined from the chemical shift indices, NOEs, and TALOS analysis. A total of 50 distance geometry (DG) structures served as starting structures for dynamical simulated-annealing calculations. The 20 lowest energy structures were selected for final structural analysis. The average structure was calculated from the geometrical average of the 20 final structures and was subjected to restraint energy minimization (REM) to correct covalent bonds and angle distortions. The structural statistics associated with the 20 final structures are summarized in Table 1. The calculated structures are well-converged with a root-mean-square deviation (RMSD) of 0.46 Å for secondary structural regions and 0.60 Å for all backbone heavy atoms. A best fit backbone superposition of all final 20 structures is displayed in Fig. 1B and the REM structure clearly demonstrates the relative orientations of its secondary structures. The backbone torsion angles, U and W for the final 20 structures ensured that all of the U, W values of the final structures were distributed in energetically favorable regions. The ribbon diagram of the REM average structure clearly shows that the topology of the structure consisted of a five-stranded antiparallel b-sheet (Fig. 1C). Four loops including RT-loop were well characterized as well. The SH3 domain forms a compact b-barrel structure with five antiparallel b-strands
Table 1 Structural statistics of the PTK6-SH3 domain and Linker.
NOE distance restraints (no) All Short range (|i j| 6 1) Medium range (1 < |i j| 6 5) Long range (|i j| > 5) Dihedral angle restraints (no) All
U W Mean CYANA target function (Å2)
PTK6-SH3 domain
PTK6-Linkera
783 503 55 225
372 248 103 21
95 47 48
0 0 0
0.44 ± 0.15
0.38 ± 0.0062
Mean RMS deviations from the average coordinate (Å) Backbone atoms (N, Ca, C0 , O) 0.60 ± 0.14b (b-strand regions only)d (0.46 ± 0.15) Heavy atoms 1.31 ± 0.18b
0.13 ± 0.06c 0.88 ± 0.18c
e
Ramachandran plot (%) Most favored regions Additional allowed regions Generously allowed regions Disallowed regions a
79.2 20.6 0.1 0.1
33.4 65.9 0.3 0.3
The solution structure of PTK6-Linker is in the presence of the SH3 domain. RMSD deviation was calculated using residues 12–70. c RMSD deviation was calculated using residues 175–188. d b-Strand regions were defined as residues 12–14 (b1), 35–40 (b2), 45–50 (b3), 56–62 (b4), and 68–70 (b5). e Ramachandran plot was calculated by PROCHECK program. b
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
239
Fig. 2. 2D NOESY spectrum and solution structure of Linker in a complex with the SH3 domain. Transferred 2D NOESY spectrum was collected with a mixing time of 600 ms. NOEs between backbone NH and side chain (A), aliphatic (B), and backbone amide (C) proton resonances are shown, respectively. (D) REM NMR structure of Linker bound to the SH3 domain is displayed as a stick representation. Three proline residues, P175, P177, and P179 were drawn by spheres. Carbon, nitrogen, and oxygen atoms were displayed by white, blue, and red color, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
240
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
(K12-V14, F35-K40, W45-L50, A56-V62, A68-R70), which is similar to that of other Src family members. However, the molecular topology of PTK6-SH3 slightly differs from that of the other SH3 domains. The first antiparallel b-sheet is formed by short strands, b1 and b5 whereas the second sheet is formed by long antiparallel strands, b2, b3, and b4 (Fig. 1C). Solution structure of the Linker peptide The NMR structures of the Linker peptide were calculated using the experimental restraints obtained from 2D NOESY spectra (Table 1). To determine the solution structure of linker in the presence of the SH3 domain, a transferred 2D NOESY experiment was performed. Although a number of NH/NH NOEs were observed, the Linker peptide does not show any regular secondary structure and, with the exception of the P175–P179 region, it is highly flexible in solution state. However, in the presence of the SH3 domain, the Linker peptide forms a well-structured turn conformation near the proline-rich region (Fig. 2A). The solution structure of Linker bound to the SH3 domain shows that the proline-rich region (P175–P187) forms a compact hairpin loop structure through a number of hydrophobic interactions, which allows molecular interaction with the SH3 domain (Fig. 2B).
Analysis of the SH3–Linker interaction The residue-specific interactions between the SH3 domain and the proline-rich linker region were analyzed by 1H–15N 2D HSQC titration experiments. The 2D HSQC spectrum of the SH3 domain in the presence of Linker is displayed in Fig. 3A. Most of the resonance shifts are observed upon Linker binding except for small deviations due to changes in the local chemical environment. Residues R22, W44, N65, and Y66 in the SH3 domain show chemical shift changes greater than 0.15 ppm (Fig. 3B). These residues are located on the surface of the molecule, where they form a hydrophobic patch for Linker binding (Fig. 3C). Our previous studies suggest that residue W44 of the SH3 domain is involved in molecular interaction with the Linker and that W44 is an important residue for auto-inhibition activity by site-directed mutagenesis and kinase assay [13]. We also showed that the P175, P177, and P179 residues in the Linker are essential for SH3 binding and catalytic activity of PTK6 [12,13]. Here, we demonstrate the structural model of the SH3/Linker complex based on NMR spectroscopy. The structure clearly shows that molecular interaction occurs between residues in the solvent-exposed loop regions of the SH3 domain and residues of the proline-rich linker region, specifically between W17/P179, E25/R186, W44/D182, W44/
Fig. 3. Molecular interaction between the SH3 domain and Linker peptide. (A) 1H–15N HSQC spectrum of the SH3 domain upon addition of the Linker peptide is shown for different molar ratios. Black, green, blue, and red colors represent molar ratios of SH3:Linker (1:0, 1:1, 1:3, and 1:5), respectively. (B) Histogram of the chemical shift deviations of the SH3 domain after titration with 1:5 molar ratios. Ranges of deviation were denoted by different colors. Chemical shift changes were calculated using the equation: Ddtotal = [(DdHN)2 + (DdN/5)2]1/2. (C) The surface model of SH3 domain labeled with chemical shift changed residues was presented. Red, blue and green colors indicated the chemical shift changes based on NMR titration experiment in (B) R22, W44, H64, N65 and Y66, were located on the surface of SH3 domain and A7, K12 and H36, were located on the opposite site. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
241
Fig. 4. A model of SH3/Linker complex together with kinase activity assay of the wild-type and mutant proteins. (A) SH3/Linker complex was calculated by HADDOCK 2.0 integrated with CNS 1.2. The SH3 domain is shown by an electrostatic surface model, and the residues that interact with Linker are labeled. Electrostatic charge distribution of the SH3 domain was calculated using the APBS program and presented by the Pymol program. Electrostatic surfaces are shown in red for negative potential, blue for positive, and white for neutral, respectively. The structure of the Linker peptide is displayed in stick model and P175, P177, and P179 residues are drawn in yellow color. (B) The detailed molecular interactions between SH3 domain and Linker are demonstrated. Both hydrophobic and electrostatic interactions are indicated as dashed lines. (C) Phosphorylation of an oligopeptide substrate (poly(Glu, Tyr)) by wild-type and four mutant PTK6. Each activity represents the average value with respect to that of wild type. The standard deviations were derived from four independent measurements. (D) Primary sequence alignment of PTK6 with other nonreceptor tyrosine kinase proteins. Multiple sequence alignment for human PTK6 (protein tyrosine kinase 6 from Homo sapiens), FRK (tyrosine protein kinase FRK from H. sapiens), SRMS (tyrosine protein kinase Srms from Mus musculus), and SRC42 (tyrosine protein kinase Src42 from Drosophila melanogaster) was performed using the CLUSTALW program. The sequence identities of each domain with respect to that of PTK6 are displayed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
W184, H64/E174, N65/P177, and Y66/P179 (Fig. 4A and B). Especially, Y66 interacts with both W17 of the SH3 domain and P179 of Linker, having an extensive network of the hydrophobic interaction (Fig. 4B). Functional implication of the SH3–Linker interaction We performed kinase activity assays using mutant PTK6 proteins to study the structure–function relationship of the SH3/Linker interaction. Wild-type PTK6 protein was used as a positive control, fixed at 100% activity. All of the mutant PTK6 proteins demonstrated higher kinase activity than wild-type PTK6, implying that the mutations increased the catalytic activity of PTK6; the Y66A, R22A, W44A, and N65A mutants demonstrated 143 ± 17%, 133 ± 5%, 110 ± 4%, and 126 ± 22% kinase activity, respectively
(Fig. 4C). These results are consistent with data from chemical shift perturbation experiments (Fig. 3B). It is interesting to note that the mutation at Y66 increases its catalytic activity even though it does not affect SH3 domain binding. Since both Y66 and W17 residues of the SH3 domain anchor the Linker through hydrophobic interactions, the size of the side-chain of Y66 would not interfere the hydrophobic interactions between the SH3 domain and Linker, however, the minor structural change might affect the auto-inhibited conformation of PTK6. Sequence comparison with nonreceptor tyrosine kinase proteins shows that the Linker sequence of PTK6 does not share a high level of homology with that of other proteins, although the SH2 and SH3 domains do demonstrate high sequence homology (Fig. 4D). Therefore, we propose that intra-molecular interaction between the SH3 domain and the Linker would be specific to PTK6.
242
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
In this study, we demonstrate that hydrophobic interaction between the SH3 domain and proline residues in the Linker is important for catalytic activity of PTK6. Specifically, we show that the R22, W44, N65, and Y66 residues of the SH3 domain are involved in auto-regulation of PTK6 activity. We propose that hydrophobic patch of the SH3 domain play an important role in auto-regulation of the catalytic function of PTK6 through interaction with the Nterminal Linker region between the SH2 and Kinase domains. Since disruption of this interaction enhances the oncogenic property of PTK6, these data will be directly applicable to the development of PTK6 inhibitors for the treatment of breast cancer. Acknowledgments This work was supported by a Grant (No. 070292) from the National R&D program for Cancer Control from the Ministry of Health & Welfare. This work was also supported in part by the Brain Korea 21 Project. References [1] K.T. Barker, L.E. Jackson, M.R. Crompton, BRK tyrosine kinase expression in a high proportion of human breast carcinomas, Oncogene 15 (1997) 799–805. [2] S.T. Lee, K.M. Strunk, R.A. Spritz, A survey of protein tyrosine kinase mRNAs expressed in normal human melanocytes, Oncogene 8 (1993) 3403–3410. [3] A. Haegebarth, R. Nunez, A.L. Tyner, The intracellular tyrosine kinase Brk sensitizes non-transformed cells to inducers of apoptosis, Cell Cycle 4 (2005) 1239–1246. [4] H. Lee, M. Kim, K.H. Lee, K.N. Kang, S.T. Lee, Exon–intron structure of the human PTK6 gene demonstrates that PTK6 constitutes a distinct family of nonreceptor tyrosine kinase, Mol. Cells 8 (1998) 401–407. [5] P.J. Mitchell, K.T. Barker, J.E. Martindale, T. Kamalati, P.N. Lowe, M.J. Page, B.A. Gusterson, M.R. Crompton, Cloning and characterisation of cDNAs encoding a novel non-receptor tyrosine kinase, brk, expressed in human breast tumours, Oncogene 9 (1994) 2383–2390. [6] X. Llor, M.S. Serfas, W. Bie, V. Vasioukhin, M. Polonskaia, J. Derry, C.M. Abbott, A.L. Tyner, BRK/Sik expression in the gastrointestinal tract and in colon tumors, Clin. Cancer Res. 5 (1999) 1767–1777. [7] J.J. Derry, S. Richard, H. Valderrama Carvajal, X. Ye, V. Vasioukhin, A.W. Cochrane, T. Chen, A.L. Tyner, Sik (BRK) phosphorylates Sam68 in the nucleus and negatively regulates its RNA binding ability, Mol. Cell. Biol. 20 (2000) 6114–6126.
[8] E. Hong, J. Shin, H.I. Kim, S.T. Lee, W. Lee, Solution structure and backbone dynamics of the non-receptor protein-tyrosine kinase-6 Src homology 2 domain, J. Biol. Chem. 279 (2004) 29700–29708. [9] T.J. Boggon, M.J. Eck, Structure and regulation of Src family kinases, Oncogene 23 (2004) 7918–7927. [10] H. Qiu, W.T. Miller, Role of the Brk SH3 domain in substrate recognition, Oncogene 23 (2004) 2216–2223. [11] H. Qiu, W.T. Miller, Regulation of the nonreceptor tyrosine kinase Brk by autophosphorylation and by autoinhibition, J. Biol. Chem. 277 (2002) 34634– 34641. [12] H. Kim, S.T. Lee, An intramolecular interaction between SH2-kinase linker and kinase domain is essential for the catalytic activity of protein-tyrosine kinase6, J. Biol. Chem. 280 (2005) 28973–28980. [13] H. Kim, J. Jung, E.S. Lee, Y.C. Kim, W. Lee, S.T. Lee, Molecular dissection of the interaction between the SH3 domain and the SH2-kinase linker region in PTK6, Biochem. Biophys. Res. Commun. 362 (2007) 829–834. [14] S. Grzesiek, H. Dobeli, R. Gentz, G. Garotta, A.M. Labhardt, A. Bax, 1H, 13C, and 15 N NMR backbone assignments and secondary structure of human interferongamma, Biochemistry 31 (1992) 8180–8190. [15] A.L. Davis, J. Keeler, E.D. Laue, D. Moskau, Experiments for recording pureabsorption heteronuclear correlation spectra using pulsed field gradients, J. Magn. Reson. 98 (1992) 207–216. [16] A.G. Palmer III, J. Cavanagh, P.E. Wright, M. Rance, Sensitivity improvement in proton-detected two-dimensional heteronuclear correlation NMR spectroscopy, J. Magn. Reson. 93 (1991) 151–170. [17] M. Piotto, V. Saudek, V. Sklenar, Gradient-tailored excitation for singlequantum NMR spectroscopy of aqueous solutions, J. Biomol. NMR 2 (1992) 661–665. [18] M. Mayer, B. Meyer, Mapping the active site of angiotensin-converting enzyme by transferred NOE spectroscopy, J. Med. Chem. 43 (2000) 2093–2099. [19] F. Delaglio, S. Grzesiek, G.W. Vuister, G. Zhu, J. Pfeifer, A. Bax, NMRPipe: a multidimensional spectral processing system based on UNIX pipes, J. Biomol. NMR 6 (1995) 277–293. [20] P. Guntert, Automated NMR structure calculation with CYANA, Methods Mol. Biol. 278 (2004) 353–378. [21] R.A. Laskowski, J.A. Rullmannn, M.W. MacArthur, R. Kaptein, J.M. Thornton, AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR, J. Biomol. NMR 8 (1996) 477–486. [22] R. Koradi, M. Billeter, K. Wuthrich, MOLMOL: a program for display and analysis of macromolecular structures, J. Mol. Graph. 14 (1996) 51–55. 29–32. [23] C. Dominguez, R. Boelens, A.M. Bonvin, HADDOCK: a protein–protein docking approach based on biochemical or biophysical information, J. Am. Chem. Soc. 125 (2003) 1731–1737. [24] G. Guerrero-Ruiz, A. Ocadiz-Ramirez, R. Garduno-Juarez, ESFERA: A program for exact calculation of the volume and surface area of fused hard-sphere molecules with unequal atomic radii, Comput. Chem. 15 (1991) 351–352.
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Structural basis of the auto-inhibition mechanism of nonreceptor tyrosine kinase PTK6 Sunggeon Ko a, Kyo-Eun Ahn a, Young-Min Lee a, Hee-Chul Ahn b, Weontae Lee a,* a b
Department of Biochemistry, College of Life Science & Biotechnology, Yonsei University, Seoul 120-749, Republic of Korea Advanced Analysis Center, Korea Institute of Science and Technology, Seoul 136-791, Republic of Korea
a r t i c l e
i n f o
Article history: Received 16 April 2009 Available online 3 May 2009
Keywords: PTK6 Auto-regulation SH3 Intra-molecular interaction NMR spectroscopy
a b s t r a c t Protein tyrosine kinase 6 (PTK6) is composed of SH3, SH2, and Kinase domains, with a linker region (Linker) between the SH2 and Kinase domains. Here, we report the structural basis of the SH3–Linker interaction that results in auto-inhibition of PTK6. The solution structures of the SH3 domain and SH3/Linker complex were determined by NMR spectroscopy. The structure of the SH3 domain forms a conventional b-barrel with two b-sheets comprised of five b-strands. However, the molecular topology and charge distribution of PTK6-SH3 slightly differs from that of the other SH3 domains. The structure of the N-terminal Linker within the complex showed that the proline-rich region (P175–P187) of the Linker forms a compact hairpin structure through hydrophobic interactions. The structure of the SH3/Linker complex revealed intra-molecular interaction between the amino acid pairs R22/E190, W44/W184, N65/P177, and Y66/P179. Mutations in PTK6 at R22, W44, N65, and Y66 residues in the SH3 domain increased catalytic activity compared with wild-type protein, implying that specific interactions between hydrophobic residues in the proline-rich linker region and hydrophobic residues in the SH3 domain are mainly responsible for down-regulating the catalytic activity of PTK6. Ó 2009 Elsevier Inc. All rights reserved.
Introduction Protein tyrosine kinases (PTKs) play an important role as regulators of various cellular functions including proliferation, differentiation, and apoptosis, and are often involved in tumorigenesis through hyper-activation processes [1–3]. Human PTK6 (also known as Brk) was identified during a survey of PTK mRNAs of human melanocytes [4], and shown to be over-expressed in breast carcinomas and colon tumors [5,6]. PTK6 is composed of Src homology 3 (SH3), Src homology 2 (SH2), and tyrosine kinase catalytic (kinase) domains. Phosphorylation of a highly conserved tyrosine residue in the SH2 domain stabilizes an inactive conformation of the protein [7]. The solution structure of the SH2 domain shows that it consists of a consensus a/b-fold with a pTyr peptide binding surface common to other SH2 domains. However, two of the a-helices (aA and aB) are located on opposite faces of the central b-sheet, and the topological arrangement of the central four-stranded b-sheet differs from that of other Src family members [8]. Activation of Src family kinases is modulated by intra-molecular interactions between the SH3 and SH2 domains that lock the
* Corresponding author. Address: Department of Biochemistry, College of Life Science & Biotechnology, Yonsei University, 134 Shinchon-Dong, Seodaemoon-Gu, Seoul 120-749, Republic of Korea. Fax: +82 2 363 2706. E-mail address: [email protected] (W. Lee). 0006-291X/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2009.04.103
kinase into a rigid conformation [9]. Modulation of the kinase activity of PTK6 by intra-molecular interaction and autophosphorylation has been reported, suggesting that interaction between the SH2 and SH3 domains and phosphorylation event play important roles in the auto-regulation of PTK6 [10,11]. Recently, we reported that mutation of key conserved residues in the N-terminal linker region between the SH2 and kinase domains of PTK6 increased its catalytic activity, indicating that the Linker plays an important role in down-regulating kinase activity [12]. Based on molecular modeling and surface plasmon resonance experiments, we proposed that the W44 residue in the SH3 domain of PTK6 participates in intra-molecular interaction of the SH3 domain with the proline-rich region of the Linker [13]. Furthermore, disruption of the intra-molecular interaction between the SH3 domain and Linker by mutation of the W44, P175, P177, and P179 residues increases the kinase activity of PTK6 [12,13]. However, the precise nature of interactions between the SH3 domain and the Linker remains unclear due to lack of information on their three-dimensional structures. Since interactions between the SH3 domain and Linker are mainly responsible for keeping PTK6 in an inactive conformation, it is important to determine the three-dimensional structures of both the SH3 domain and of the Linker in the presence of the SH3 domain. Here, we present solution structures of the SH3 domain and SH3/Linker complex and demonstrate precise molecular interactions between the SH3 domain and Linker. These results provide structure–function information on the mechanism
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
of auto-inhibitory regulation of the nonreceptor tyrosine kinase PTK6. Materials and methods Cloning, protein expression, and purification of PTK6. PTK6 (residues 1–451), PTK6-SH3 domain (residues 1–72), and PTK6-Linker (residues 171–191) were cloned into pGEX 4T-1 (Amersham). Sense primers containing the BamHI site and TEV protease recognition sequence (ENLYFQG), and antisense primers containing the EcoRI site were used for PCR experiments. All plasmids encoding PTK6 proteins were transformed into Escherichia coli BL21(DE3), and over-expression of target proteins was induced by 0.1 mM isopropyl b-D-thiogalactopyranoside (IPTG) at 25 °C. Fusion proteins were purified by GST-affinity chromatography using GSTrap FF columns (GE Healthcare), and the GST-tag was removed by TEV protease. Size exclusion chromatography was performed using
237
Superdex 75 columns. The sample buffer was exchanged for 10 mM ammonium bicarbonate buffer and lyophilized to yield the final sample. For heteronuclear NMR experiments, the SH3 domain was overexpressed in M9 media containing 15NH4Cl or 15NH4Cl/13C-D-glucose (Cambridge Isotope Laboratories) and purified by GST-affinity chromatography and size exclusion chromatography as described above. Peptide synthesis and sample preparation. Peptides corresponding to the Linker (RKHEPEPLPHWDDWEREPEEF; residues 171–191 of the human PTK6 sequence) were synthesized by an improved version of the solid-phase method using Fmoc chemistry (Anygen Inc., Kwangju, Korea). The peptide was purified by high-performance liquid chromatography (HPLC) on a Shimazu Prep LC-8A system with a Shim-pack ODS(C18) prep column. The sample was eluted with a linear gradient (A = 0.1%TFA in water and B = 0.1% TFA in CH3CN) with detection by absorbance at 230 nm. Samples were
Fig. 1. 1H–15N HSQC spectrum and solution structures of the SH3 domain of PTK6. (A) Backbone assignment of the SH3 domain (residues 1–72) is displayed in the 1H–15N HSQC spectrum. (B) The 20 NMR structures were superimposed using the MOLMOL program. All structures were fitted for Ca, C0 , and N atoms of residues 12–70 with respect to the restrained energy minimized (REM) average structure. (C) Ribbon diagram showing molecular topology of the SH3 domain. The five antiparallel b-strands are labeled.
238
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
prepared for NMR by dissolving the peptides in 90% H2O/10% D2O or 99.9% D2O in 50 mM potassium phosphate buffer (pH 7.0) and adjustment of the final peptide concentration to 2–4 mM. NMR experiments and structure calculation of the SH3 domain and Linker peptide. 15N- or 13C/15N-labeled PTK6-SH3 domain was used for heteronuclear NMR experiments. Stable isotope-labeled PTK6SH3 domain was exchanged to NMR buffer (90% H2O/10% D2O, 25 mM HEPES, 100 mM NaCl, pH 7.0) during size exclusion chromatography and concentrated to 0.2 mM using Amicon Ultra filter units (Amicon). 1H–15N heteronuclear single quantum coherence (HSQC), 3D HNCACB, CBCA(CO)NH, HNCA, HBHA(CO)NH, and HNCO experiments were used for backbone assignment [14]. 15N edited NOESY (s = 150 ms), 13C-edited NOESY (s = 150 ms), and HCCH-TOCSY were used for NOE and side chain assignment [15– 17]. For NMR analysis of the Linker, lyophilized peptide was dissolved in NMR buffer to a concentration of 1 mM. Two-dimensional total correlation spectroscopy (TOCSY) with a MLEV-17 mixing pulse of 69.7 ms and two-dimensional nuclear Overhauser effect spectroscopy (NOESY) with mixing times of 100–600 ms were performed for Linker peptide. All NMR experiments were collected in the phase-sensitive mode using the time proportional phase incrementation (TPPI) method with 2048 data points in the t2 and 256 in the t1 domains. To determine the structure of Linker bound to the SH3 domain, we performed transferred nuclear Overhauser effect spectroscopy (TR-NOESY), 2D NOESY (s = 600 ms) and 2D TOCSY [18]. All spectra were obtained using Bruker DRX 500 MHz equipped with CryoprobeTM system and Varian 900 MHz spectrometer. XWIN-NMR, NMRPipe/NMRDraw, and SPARKY software [19] were used for data processing and chemical shift assignment. CYANA 2.1 with torsion angle dynamics was used for structure calculations [20]. Distance constraints and angle constraints from TALOS were also used for final structure refinement. CYANA calculations were executed using a 16-node linux cluster computer, and 20 lowest energy structures were evaluated by the PROCHECK program [21]. Final structures were displayed and analyzed by Pymol (DeLano Scientific LLC) and MOLMOL programs [22]. NMR titration and structure calculation of the SH3/Linker complex. 1 H–15N HSQC titrations were performed to identify the specific residues involved in the interaction between the SH3 domain and linker. We prepared NMR samples containing various molar ratios of SH3:Linker (1:1, 1:3, and 1:5) and a final concentration of 15N-labeled SH3 domain of 0.2 mM. Chemical shift changes were calculated using the equation: Ddtotal = [(DdHN)2 + (DdN/5)2]1/2. The structure of the SH3/Linker complex was calculated by the HADDOCK 2.0 program integrated with CNS 1.2 in a 16-node linux cluster computer [23]. We determined active residues and passive residues based on HSQC titration experiments, GST pull down assay and solvent accessibility calculations in NACCESS program [24]. Active residues of the SH3 domain (R22, H36, W44, H64, N65, and Y66) were defined by chemical shift changes (Dd) above 0.1 in HSQC titration experiments and solvent accessibility above 40%. Active residues of the Linker (P175, P177, and P197) were defined by GST pull down assay [13] and solvent accessibility above 40%. All passive residues were determined by the neighbors to active residue and solvent accessibility above 40%. Clustering analysis of HADDOCK calculation results was executed using a cut off value of RMSD 7.5, 5.0, and 3.0 Å. Final structures were analyzed and displayed using the Pymol program. In vitro kinase assay of PTK6 and mutant proteins. An in vitro kinase assay was used to examine kinase activities of wild-type and mutant PTK6 proteins. PTK6 mutants (R22A, W44A, N65A, and Y66A) were generated by PCR mutagenesis, and proteins were purified as described previously. For the kinase assay, 10 lg PTK6 substrate (poly(Glu,Tyr), Sigma) was incubated in the absence or presence of 20 lg GST-PTK6 protein in 20 ll kinase reaction buffer
(20 mM Tris–HCl, 10 mM MgCl2, 1 mM MnCl2, 50 lM Na3VO4, pH 7.4) containing 100 lM [c-32P] ATP for 30 min at 30 °C. SDS–PAGE and autoradiography with scintillation counting were performed as described previously [12]. Results and discussion Resonance assignments and structure of the PTK6-SH3 domain The backbone resonance assignments of the SH3 domain were completed using data from 1H–15N 2D HSQC, 3D HNCACB, CBCA(CO)NH, HNCO, and HCACO spectra (Fig. 1A). The side-chain and NOE assignments were completed by combined use of 3D HCCHTOCSY, 15N-edited TOCSY-HSQC, and 13C/15N-edited NOESY experiments. The secondary structures were determined from the chemical shift indices, NOEs, and TALOS analysis. A total of 50 distance geometry (DG) structures served as starting structures for dynamical simulated-annealing calculations. The 20 lowest energy structures were selected for final structural analysis. The average structure was calculated from the geometrical average of the 20 final structures and was subjected to restraint energy minimization (REM) to correct covalent bonds and angle distortions. The structural statistics associated with the 20 final structures are summarized in Table 1. The calculated structures are well-converged with a root-mean-square deviation (RMSD) of 0.46 Å for secondary structural regions and 0.60 Å for all backbone heavy atoms. A best fit backbone superposition of all final 20 structures is displayed in Fig. 1B and the REM structure clearly demonstrates the relative orientations of its secondary structures. The backbone torsion angles, U and W for the final 20 structures ensured that all of the U, W values of the final structures were distributed in energetically favorable regions. The ribbon diagram of the REM average structure clearly shows that the topology of the structure consisted of a five-stranded antiparallel b-sheet (Fig. 1C). Four loops including RT-loop were well characterized as well. The SH3 domain forms a compact b-barrel structure with five antiparallel b-strands
Table 1 Structural statistics of the PTK6-SH3 domain and Linker.
NOE distance restraints (no) All Short range (|i j| 6 1) Medium range (1 < |i j| 6 5) Long range (|i j| > 5) Dihedral angle restraints (no) All
U W Mean CYANA target function (Å2)
PTK6-SH3 domain
PTK6-Linkera
783 503 55 225
372 248 103 21
95 47 48
0 0 0
0.44 ± 0.15
0.38 ± 0.0062
Mean RMS deviations from the average coordinate (Å) Backbone atoms (N, Ca, C0 , O) 0.60 ± 0.14b (b-strand regions only)d (0.46 ± 0.15) Heavy atoms 1.31 ± 0.18b
0.13 ± 0.06c 0.88 ± 0.18c
e
Ramachandran plot (%) Most favored regions Additional allowed regions Generously allowed regions Disallowed regions a
79.2 20.6 0.1 0.1
33.4 65.9 0.3 0.3
The solution structure of PTK6-Linker is in the presence of the SH3 domain. RMSD deviation was calculated using residues 12–70. c RMSD deviation was calculated using residues 175–188. d b-Strand regions were defined as residues 12–14 (b1), 35–40 (b2), 45–50 (b3), 56–62 (b4), and 68–70 (b5). e Ramachandran plot was calculated by PROCHECK program. b
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
239
Fig. 2. 2D NOESY spectrum and solution structure of Linker in a complex with the SH3 domain. Transferred 2D NOESY spectrum was collected with a mixing time of 600 ms. NOEs between backbone NH and side chain (A), aliphatic (B), and backbone amide (C) proton resonances are shown, respectively. (D) REM NMR structure of Linker bound to the SH3 domain is displayed as a stick representation. Three proline residues, P175, P177, and P179 were drawn by spheres. Carbon, nitrogen, and oxygen atoms were displayed by white, blue, and red color, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
240
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
(K12-V14, F35-K40, W45-L50, A56-V62, A68-R70), which is similar to that of other Src family members. However, the molecular topology of PTK6-SH3 slightly differs from that of the other SH3 domains. The first antiparallel b-sheet is formed by short strands, b1 and b5 whereas the second sheet is formed by long antiparallel strands, b2, b3, and b4 (Fig. 1C). Solution structure of the Linker peptide The NMR structures of the Linker peptide were calculated using the experimental restraints obtained from 2D NOESY spectra (Table 1). To determine the solution structure of linker in the presence of the SH3 domain, a transferred 2D NOESY experiment was performed. Although a number of NH/NH NOEs were observed, the Linker peptide does not show any regular secondary structure and, with the exception of the P175–P179 region, it is highly flexible in solution state. However, in the presence of the SH3 domain, the Linker peptide forms a well-structured turn conformation near the proline-rich region (Fig. 2A). The solution structure of Linker bound to the SH3 domain shows that the proline-rich region (P175–P187) forms a compact hairpin loop structure through a number of hydrophobic interactions, which allows molecular interaction with the SH3 domain (Fig. 2B).
Analysis of the SH3–Linker interaction The residue-specific interactions between the SH3 domain and the proline-rich linker region were analyzed by 1H–15N 2D HSQC titration experiments. The 2D HSQC spectrum of the SH3 domain in the presence of Linker is displayed in Fig. 3A. Most of the resonance shifts are observed upon Linker binding except for small deviations due to changes in the local chemical environment. Residues R22, W44, N65, and Y66 in the SH3 domain show chemical shift changes greater than 0.15 ppm (Fig. 3B). These residues are located on the surface of the molecule, where they form a hydrophobic patch for Linker binding (Fig. 3C). Our previous studies suggest that residue W44 of the SH3 domain is involved in molecular interaction with the Linker and that W44 is an important residue for auto-inhibition activity by site-directed mutagenesis and kinase assay [13]. We also showed that the P175, P177, and P179 residues in the Linker are essential for SH3 binding and catalytic activity of PTK6 [12,13]. Here, we demonstrate the structural model of the SH3/Linker complex based on NMR spectroscopy. The structure clearly shows that molecular interaction occurs between residues in the solvent-exposed loop regions of the SH3 domain and residues of the proline-rich linker region, specifically between W17/P179, E25/R186, W44/D182, W44/
Fig. 3. Molecular interaction between the SH3 domain and Linker peptide. (A) 1H–15N HSQC spectrum of the SH3 domain upon addition of the Linker peptide is shown for different molar ratios. Black, green, blue, and red colors represent molar ratios of SH3:Linker (1:0, 1:1, 1:3, and 1:5), respectively. (B) Histogram of the chemical shift deviations of the SH3 domain after titration with 1:5 molar ratios. Ranges of deviation were denoted by different colors. Chemical shift changes were calculated using the equation: Ddtotal = [(DdHN)2 + (DdN/5)2]1/2. (C) The surface model of SH3 domain labeled with chemical shift changed residues was presented. Red, blue and green colors indicated the chemical shift changes based on NMR titration experiment in (B) R22, W44, H64, N65 and Y66, were located on the surface of SH3 domain and A7, K12 and H36, were located on the opposite site. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
241
Fig. 4. A model of SH3/Linker complex together with kinase activity assay of the wild-type and mutant proteins. (A) SH3/Linker complex was calculated by HADDOCK 2.0 integrated with CNS 1.2. The SH3 domain is shown by an electrostatic surface model, and the residues that interact with Linker are labeled. Electrostatic charge distribution of the SH3 domain was calculated using the APBS program and presented by the Pymol program. Electrostatic surfaces are shown in red for negative potential, blue for positive, and white for neutral, respectively. The structure of the Linker peptide is displayed in stick model and P175, P177, and P179 residues are drawn in yellow color. (B) The detailed molecular interactions between SH3 domain and Linker are demonstrated. Both hydrophobic and electrostatic interactions are indicated as dashed lines. (C) Phosphorylation of an oligopeptide substrate (poly(Glu, Tyr)) by wild-type and four mutant PTK6. Each activity represents the average value with respect to that of wild type. The standard deviations were derived from four independent measurements. (D) Primary sequence alignment of PTK6 with other nonreceptor tyrosine kinase proteins. Multiple sequence alignment for human PTK6 (protein tyrosine kinase 6 from Homo sapiens), FRK (tyrosine protein kinase FRK from H. sapiens), SRMS (tyrosine protein kinase Srms from Mus musculus), and SRC42 (tyrosine protein kinase Src42 from Drosophila melanogaster) was performed using the CLUSTALW program. The sequence identities of each domain with respect to that of PTK6 are displayed. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
W184, H64/E174, N65/P177, and Y66/P179 (Fig. 4A and B). Especially, Y66 interacts with both W17 of the SH3 domain and P179 of Linker, having an extensive network of the hydrophobic interaction (Fig. 4B). Functional implication of the SH3–Linker interaction We performed kinase activity assays using mutant PTK6 proteins to study the structure–function relationship of the SH3/Linker interaction. Wild-type PTK6 protein was used as a positive control, fixed at 100% activity. All of the mutant PTK6 proteins demonstrated higher kinase activity than wild-type PTK6, implying that the mutations increased the catalytic activity of PTK6; the Y66A, R22A, W44A, and N65A mutants demonstrated 143 ± 17%, 133 ± 5%, 110 ± 4%, and 126 ± 22% kinase activity, respectively
(Fig. 4C). These results are consistent with data from chemical shift perturbation experiments (Fig. 3B). It is interesting to note that the mutation at Y66 increases its catalytic activity even though it does not affect SH3 domain binding. Since both Y66 and W17 residues of the SH3 domain anchor the Linker through hydrophobic interactions, the size of the side-chain of Y66 would not interfere the hydrophobic interactions between the SH3 domain and Linker, however, the minor structural change might affect the auto-inhibited conformation of PTK6. Sequence comparison with nonreceptor tyrosine kinase proteins shows that the Linker sequence of PTK6 does not share a high level of homology with that of other proteins, although the SH2 and SH3 domains do demonstrate high sequence homology (Fig. 4D). Therefore, we propose that intra-molecular interaction between the SH3 domain and the Linker would be specific to PTK6.
242
S. Ko et al. / Biochemical and Biophysical Research Communications 384 (2009) 236–242
In this study, we demonstrate that hydrophobic interaction between the SH3 domain and proline residues in the Linker is important for catalytic activity of PTK6. Specifically, we show that the R22, W44, N65, and Y66 residues of the SH3 domain are involved in auto-regulation of PTK6 activity. We propose that hydrophobic patch of the SH3 domain play an important role in auto-regulation of the catalytic function of PTK6 through interaction with the Nterminal Linker region between the SH2 and Kinase domains. Since disruption of this interaction enhances the oncogenic property of PTK6, these data will be directly applicable to the development of PTK6 inhibitors for the treatment of breast cancer. Acknowledgments This work was supported by a Grant (No. 070292) from the National R&D program for Cancer Control from the Ministry of Health & Welfare. This work was also supported in part by the Brain Korea 21 Project. References [1] K.T. Barker, L.E. Jackson, M.R. Crompton, BRK tyrosine kinase expression in a high proportion of human breast carcinomas, Oncogene 15 (1997) 799–805. [2] S.T. Lee, K.M. Strunk, R.A. Spritz, A survey of protein tyrosine kinase mRNAs expressed in normal human melanocytes, Oncogene 8 (1993) 3403–3410. [3] A. Haegebarth, R. Nunez, A.L. Tyner, The intracellular tyrosine kinase Brk sensitizes non-transformed cells to inducers of apoptosis, Cell Cycle 4 (2005) 1239–1246. [4] H. Lee, M. Kim, K.H. Lee, K.N. Kang, S.T. Lee, Exon–intron structure of the human PTK6 gene demonstrates that PTK6 constitutes a distinct family of nonreceptor tyrosine kinase, Mol. Cells 8 (1998) 401–407. [5] P.J. Mitchell, K.T. Barker, J.E. Martindale, T. Kamalati, P.N. Lowe, M.J. Page, B.A. Gusterson, M.R. Crompton, Cloning and characterisation of cDNAs encoding a novel non-receptor tyrosine kinase, brk, expressed in human breast tumours, Oncogene 9 (1994) 2383–2390. [6] X. Llor, M.S. Serfas, W. Bie, V. Vasioukhin, M. Polonskaia, J. Derry, C.M. Abbott, A.L. Tyner, BRK/Sik expression in the gastrointestinal tract and in colon tumors, Clin. Cancer Res. 5 (1999) 1767–1777. [7] J.J. Derry, S. Richard, H. Valderrama Carvajal, X. Ye, V. Vasioukhin, A.W. Cochrane, T. Chen, A.L. Tyner, Sik (BRK) phosphorylates Sam68 in the nucleus and negatively regulates its RNA binding ability, Mol. Cell. Biol. 20 (2000) 6114–6126.
[8] E. Hong, J. Shin, H.I. Kim, S.T. Lee, W. Lee, Solution structure and backbone dynamics of the non-receptor protein-tyrosine kinase-6 Src homology 2 domain, J. Biol. Chem. 279 (2004) 29700–29708. [9] T.J. Boggon, M.J. Eck, Structure and regulation of Src family kinases, Oncogene 23 (2004) 7918–7927. [10] H. Qiu, W.T. Miller, Role of the Brk SH3 domain in substrate recognition, Oncogene 23 (2004) 2216–2223. [11] H. Qiu, W.T. Miller, Regulation of the nonreceptor tyrosine kinase Brk by autophosphorylation and by autoinhibition, J. Biol. Chem. 277 (2002) 34634– 34641. [12] H. Kim, S.T. Lee, An intramolecular interaction between SH2-kinase linker and kinase domain is essential for the catalytic activity of protein-tyrosine kinase6, J. Biol. Chem. 280 (2005) 28973–28980. [13] H. Kim, J. Jung, E.S. Lee, Y.C. Kim, W. Lee, S.T. Lee, Molecular dissection of the interaction between the SH3 domain and the SH2-kinase linker region in PTK6, Biochem. Biophys. Res. Commun. 362 (2007) 829–834. [14] S. Grzesiek, H. Dobeli, R. Gentz, G. Garotta, A.M. Labhardt, A. Bax, 1H, 13C, and 15 N NMR backbone assignments and secondary structure of human interferongamma, Biochemistry 31 (1992) 8180–8190. [15] A.L. Davis, J. Keeler, E.D. Laue, D. Moskau, Experiments for recording pureabsorption heteronuclear correlation spectra using pulsed field gradients, J. Magn. Reson. 98 (1992) 207–216. [16] A.G. Palmer III, J. Cavanagh, P.E. Wright, M. Rance, Sensitivity improvement in proton-detected two-dimensional heteronuclear correlation NMR spectroscopy, J. Magn. Reson. 93 (1991) 151–170. [17] M. Piotto, V. Saudek, V. Sklenar, Gradient-tailored excitation for singlequantum NMR spectroscopy of aqueous solutions, J. Biomol. NMR 2 (1992) 661–665. [18] M. Mayer, B. Meyer, Mapping the active site of angiotensin-converting enzyme by transferred NOE spectroscopy, J. Med. Chem. 43 (2000) 2093–2099. [19] F. Delaglio, S. Grzesiek, G.W. Vuister, G. Zhu, J. Pfeifer, A. Bax, NMRPipe: a multidimensional spectral processing system based on UNIX pipes, J. Biomol. NMR 6 (1995) 277–293. [20] P. Guntert, Automated NMR structure calculation with CYANA, Methods Mol. Biol. 278 (2004) 353–378. [21] R.A. Laskowski, J.A. Rullmannn, M.W. MacArthur, R. Kaptein, J.M. Thornton, AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR, J. Biomol. NMR 8 (1996) 477–486. [22] R. Koradi, M. Billeter, K. Wuthrich, MOLMOL: a program for display and analysis of macromolecular structures, J. Mol. Graph. 14 (1996) 51–55. 29–32. [23] C. Dominguez, R. Boelens, A.M. Bonvin, HADDOCK: a protein–protein docking approach based on biochemical or biophysical information, J. Am. Chem. Soc. 125 (2003) 1731–1737. [24] G. Guerrero-Ruiz, A. Ocadiz-Ramirez, R. Garduno-Juarez, ESFERA: A program for exact calculation of the volume and surface area of fused hard-sphere molecules with unequal atomic radii, Comput. Chem. 15 (1991) 351–352.