TY - JOUR
AU - Wu, Joe C.
AB - Abstract The activation of receptor tyrosine kinases (RTKs) is tightly regulated through a variety of mechanisms. Kinetic studies show that activation of c-Kit RTK occurs through an inter-molecular autophosphorylation. Phosphopeptide mapping of c-Kit reveals that 14–22 phosphates are added to each mol of wild-type (WT) c-Kit during the activation. Phosphorylation sites are found on the JM, kinase insert (KID), c-terminal domains and the activation loop (A-loop), but only the sites on the JM domain contribute to the kinase activation. The A-loop tyrosine (Y823) is not phosphorylated until very late in the activation (>90% completion), indicating that the A-loop phosphorylation is not required for c-Kit activation. A sunitinib-resistant mutant D816H that accelerates auto-activation by 184-fold shows no phosphorylation on the A-loop tyrosine after full activation. A loss-of-phosphorylation mutation Y823F remains fully competent in auto-activation. Similar to WT and D816H, the unactivated Y823F mutant binds sunitinib and imatinib with high affinity (KD = 5.9 nM). But unlike the WT and D816H where the activated enzymes lose the ability to bind the two drugs, activated Y823F binds the two inhibitors effectively. These observations suggest that the A-loop of activated Y823F remains flexible and can readily adopt unactivated conformations to accommodate DFG-out binders. A-loop, autophosphorylation, tyrosine kinase, c-Kit, sunitinib, imatinib Receptor tyrosine kinases (RTK) play a key role in a host of cellular processes including differentiation and cell growth. The kinase c-Kit is a class III RTK consisting of five IgG-like extracellular domains, a transmembrane domain followed by a cytoplasmic kinase domain. This class also includes the Platelet Derived Growth Factor (PDGF), fms-related tyrosine kinase 3 (FLT3) and the Colony Stimulating Factor (CSF-1) receptors (1–3). Activation of c-Kit occurs upon binding Stem Cell Factor (SCF) to the extracellular domain which leads to autophosphorylation of the cytoplasmic kinase domain and activation of downstream-signalling molecules (4–6). Despite the wealth of information on the effectors and downstream-signalling events that occur upon stimulation of cells with SCF, the molecular mechanism of activation of c-Kit remains not fully elucidated. There are multiple phosphor-acceptor sites in the kinase domain, including Y823 on the activation loop, known to be phosphorylated on c-Kit either through autophosphorylation or by other kinases in trans (1, 7, 8). Activation loop tyrosine autophosphorylation is generally thought to regulate kinase activation of RTKs (9, 10). However, A-loop tyrosine phosphorylation may play very different roles in regulating kinase activation, depending on the individual kinase. For example, EGFR activation loop phosphorylation has no effect on the catalytic activity of the kinase but rather provides a docking site for the SH2 domain of STAT5 (11, 12). This is in contrast to the Tie2 RTK in which the A-loop tyrosine is phosphorylated early in a sequential manner in autophosphorylation reactions followed by Y1110. This residue is required for downstream interactions with SH2 domains (13). The resulting active Tie2 has a 100-fold increase in kcat and a 460-fold increase in kcat/Km compared to the unphosphorylated enzyme. In the case of c-Kit, the specific role of the A-loop tyrosine Y823 in the activation process is unknown. Structural studies of c-Kit indicate that upon activation, a shift in the conformation of the Asp810-Phe811-Gly812 motif (DFG) at the N-terminus of the A-loop occurs. In the active conformation, the A-loop extends over the C terminus of the catalytic pocket and the Phe811 moves away from the ATP-binding region (‘DFG-in’), thereby creating catalytically active conformation of the kinase. Although it is known that the shift between active and inactive conformation must be controlled by phosphorylation/dephosphorylation, the exact mechanism is not clear. Little is known about the functional role of the A-loop tyrosine plays, nor about the multiple phosphorylation sites in the JM (Juxtamembrane) and KID (Kinase Insert) domains. In addition, the sequence of 14–22 phosphorylation events of c-Kit leading to full activation of kinase activity is poorly understood. In H526 cells, Y823 is phosphorylated upon SCF stimulation (14). This may imply a possible functional role of A-loop phosphorylation in c-Kit activation. On the other hand, a crystal structure of active c-Kit shows Y823 unphosphorylated (15), suggesting that the phosphorylation of Y823 may not be required. In another crystal structure where c-Kit assumes an enzymatically unactivated conformation, Y823 is positioned in the kinase active site and hydrogen bonded to the catalytic base (D792) thus blocking access to the catalytic centre and maintaining the enzyme in the inhibited state (16). Hence phosphorylation of c-Kit Y823 could function to disengage the A-loop from the inhibitory state rather than to activate the enzyme. In this report, several mutants are generated to investigate the roles phosphorylated and unphosphorylated Y823 play in shifting the conformational equilibrium between the activated and unactivated state. Material and Methods Expression and purification of c-Kit wild-type (WT) and mutants c-Kit constructs were expressed and purified as previously described (15). Briefly, the construct encoding an N-terminal 6× HIS tag + r3c protease site + c-Kit residues 544–976 was cloned into baculovirus vector pVL1393. High titre stock was made and expressed in Sf9 cells at multiplicity of infection of 1 for 48 h. Cells were harvested and frozen –80°C. The pellet was lysed in 25 mM Tris pH 7.4, 250 mM NaCl, 0.25 mM TCEP, 20 mM imidazole and bound to Ni–NTA column. The protein was step eluted using same buffer with 250 mM imidazole. Peak fractions were dialysed overnight at 4°C. For the WT protein, the HIS tag was cleaved with r3C protease. Protein was reloaded on Ni–NTA to remove cleaved HIS tag. Flow through was collected from column and concentrated. Samples were run on Superdex-200 column equilibrated with 25 mM Tris pH 7.4, 250 mM NaCl, 0.5 mM TCEP and 1 mM EDTA. Peak fractions were pooled, concentrated and flash frozen. c-Kit activation and phosphopeptide mapping The kinase domain construct of WT (1.0 µM), D816H (0.7 µM) or Y823F (1.0 µM) c-Kit (amino acids 544–976) was activated by the addition of ATP to 4 mM in buffer consisting of 50 mM Tris pH 7.6, 10 mM MgCl2, 3 µM Na3VO4, 1 mM EGTA. Samples were incubated at room temperature for WT and Y823F or on ice for D816H. Aliquots were then removed at various time points and the activation reaction was quenched by addition of EDTA to 12 mM. Trypsin and chymotrypsin digestion of c-Kit was carried out in 50 mM NH4HCO3 buffer as described (17) with slight modifications. Samples were denatured by incubation at 95°C for 5 min. Followed by reduction with DTT (4.5 mM) at 56°C for 20 min. Denatured samples were then alkylated with iodoacetamide (10 mM) for 15 min. Trypsin, chymotrypsin and lys-C digestion were then performed by adding 1 : 10 enzyme weight/weight ratio protease and incubated overnight at 37°C for trypsin and lys-C and room temperature for chymotrypsin. LC–MS–MS analysis was performed on a Thermo-fisher LTQ mass spectrometer coupled with an Eksigent 2D nano-LC with a nano-electrospray source from New ObjectiveTM (MA, USA). A Picofrit ProteoPep2 column (10 cm) was used. The MS–MS data were collected in positive mode using data-dependent zoom scan. Phosphorylation site search was performed using SpectrumMill and Mascot software. Direct binding of sunitinib to c-Kit mutants Direct sunitinib binding to unactivated mutants Y823A, Y823D and Y823F was detected by measuring intrinsic Trp fluorescence quenching that occurs upon drug binding to c-Kit as described in (18). Determination of auto-activation rate constant kact Activation of c-Kit at saturating ATP concentration was monitored using ATP-regenerating system composed of 4 mM ATP, 4 mM PEP, 0.5 mM NADH, 2.5 mg/ml poly Glu-Tyr, 25 U/ml PK and 31 U/ml LDH, in Buffer A (50 mM HEPES, pH 7.5, 10 mM MgCl2, 1 mM EGTA, 3 µM Na3VO4, 0.01% Brij-35). For a bimolecular auto-activation reaction coupled with a kinase activity assay performed at saturated substrate concentration, the integrated rate equation relating the change in ADP concentration to the auto-activation rate constant (kact) is:   (1) Data from auto-activation reaction progress curves were fit the above equation to determine the kact and kcat as described in (18). IC50 determinations For IC50 determinations of sunitinib and imatinib against pre-activated c-Kit, 10 μl of 400 nM pre-activated c-Kit was incubated with 10 μl of 0–90 μM sunitinib or imatinib at 20°C for 30 min. This was followed by addition of 80 μl of the ATP-regenerating system as described above with the exception that 0.25 mg/ml poly Glu-Tyr substrate was used. Kinase activity was then monitored by following the linear decrease of A340. In order to measure IC50 for the inhibition of c-Kit auto-activation, 10 μl of 400 nM unactivated c-Kit kinase was pre-incubated with 10 μl of 0–90 μM sunitinib or imatinib at 20°C for 30 min, followed by addition of 10 μl of 12 mM ATP to initiate the c-Kit auto-phosphorylation. The autophosphorylation/activation was allowed to continue for 1 h at 20°C. Reaction wells were sealed to avoid evaporation. Progress of the auto-activation reaction was followed by determining the resultant active c-Kit activity by the addition of 70 μl of ATP-regenerating coupled assay mixture. Eleven-point dose response curves with inhibitor concentration ranging from 0.3 nM to 30 μM were used to determine IC50. Steady-state determination of Km, Vmax Kinetics experiments to determine Vmax and Km for ATP were preformed using pre-activated c-Kit prepared by incubating 10-µM unactivated enzyme with 4 mM ATP in Buffer A for 3 h at room temperature. Free ADP/ATP was removed prior to performing experiments by running samples through two sequential Sephadex G-50 columns (Sigma) bed volume 3 ml. The activated enzyme was then stored at –80oC for future uses. c-Kit kinase activity was determined using an ATP-regenerating system as described above. The steady-state rate of poly Glu-Tyr phosphorylation was calculated from the observed linear decrease of A340 due to the coupled oxidation of NADH, using 6,220 as the molar absorbance of NADH. In the absence of phosphor-acceptor poly Glu-Tyr, NADH depletion was negligible. To assure that our derived Vmax parameters reflected concentrations of active enzyme, active enzyme concentration was determined by active site titration as described (19, 20). Measurement of IC50 as a function of activation time WT c-Kit (0.5 µM) in buffer consisting of 50 mM Tris pH 7.6, 10 mM MgCl2, 3 µM Na3VO4, 1 mM EGTA was auto-activated by the addition of 4 mM ATP. Activation reaction was stopped by the addition of EDTA to 12 mM. The IC50 for sunitinib was then measured in Buffer A as described above using 10 nM c-Kit. Results Auto-activation and phosphopeptide mapping In order to study the sequence of phosphorylation events that lead up to c-Kit activation, phosphorylation sites and kinase activity were monitored closely during an activation reaction. As shown in Fig. 1, under the indicated experimental conditions, maximal kinase activity of WT c-Kit was obtained within 3 h. The phosphorylation of c-Kit in this auto-activation reaction yielded multiple phosphorylated species with 0–1 phosphate per c-Kit molecule before the reaction, and 14–22 phosphates on each c-Kit after full activation (Fig. 1A and B). Throughout the course of activation, aliquots of the reaction mixture were collected and analysed in phosphopeptide mapping experiments as well as in a kinase activity assay. The results are summarized in Table I. Mass spectra of intact c-Kit prior to trypsin and chymotrypsin digestion show 14–22 phosphates incorporated after full activation, however phosphopeptide mapping experiments revealed 9, 8 and 11 phosphates in the fully activated kinase for WT, D816H and Y823F mutants, respectively. This is most likely due to the presence of phosphopeptide species not detectable as a result of poor ionization efficiencies and/or low abundance of phosphopeptide resulting from the protease digestion. Fig. 1 View largeDownload slide LC–MS spectra of intact WT c-Kit. (A) Mass spectra of unactivated c-Kit WT prior to incubation ATP. (B) Mass spectra for fully phosphorylated activated form following incubation with 4 mM ATP for 3 h at 10-µM c-Kit concentration. (C) Specific activity of c-Kit vs time for samples analyzed by Mass spectrometry. Auto-activation reaction was carried out at 1-µM c-Kit with 4 mM ATP for a time course of 3.5 h. Fig. 1 View largeDownload slide LC–MS spectra of intact WT c-Kit. (A) Mass spectra of unactivated c-Kit WT prior to incubation ATP. (B) Mass spectra for fully phosphorylated activated form following incubation with 4 mM ATP for 3 h at 10-µM c-Kit concentration. (C) Specific activity of c-Kit vs time for samples analyzed by Mass spectrometry. Auto-activation reaction was carried out at 1-µM c-Kit with 4 mM ATP for a time course of 3.5 h. Table I. Phosphopeptide mapping of WT, D816H and Y823F c-Kit. Time (min)  % activity  specific activity (µmol/min/mg)  Phosphorylation sites        Wild Type  0  0  0.006 ± 0.004    60  20  0.058 ± 0.010  Y547, Y553, Y703, Y721  90  64  0.172 ± 0.008  Y547, Y553, Y568,Y703, Y721, (S729/Y730), S959  120  90  0.239 ± 0.007  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  150  99  0.261 ± 0.024  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730),Y823, S959  180  100  0.265 ± 0.013  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730),Y823, S959  210  100  0.264 ± 0.007  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730),Y823, S959        D816H  0  0  0.007 ± 0.012    1  20  0.041 ± 0.014  Y568,Y703, Y721, S959  2  31  0.060 ± 0.023  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  4  50  0.090 ± 0.015  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  6  90  0.156 ± 0.021  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  8  93  0.161 ± 0.036  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  10  100  0.172 ± 0.024  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959        Y823F  0  0  0.008 ± 0.005  S943  4  6  0.020 ± 0.003  Y703, Y721, S943  10  7  0.021 ± 0.008  Y547, Y568,Y703, Y721, S729, S943  20  22  0.051 ± 0.031  Y547, Y553, Y568,Y703, Y721, S729,Y747, S943  40  60  0.125 ± 0.006  Y547, Y553, Y568, Y570,Y703, Y721, S729, Y747, S943  60  96  0.194 ± 0.041  Y547, Y553, Y568, Y570,Y703, Y721, S729, Y747, S943  120  100  0.203 ± 0.002  Y547, Y553, Y568, Y570, Y703, Y721, S729, Y747, Y900, S937, S943  Time (min)  % activity  specific activity (µmol/min/mg)  Phosphorylation sites        Wild Type  0  0  0.006 ± 0.004    60  20  0.058 ± 0.010  Y547, Y553, Y703, Y721  90  64  0.172 ± 0.008  Y547, Y553, Y568,Y703, Y721, (S729/Y730), S959  120  90  0.239 ± 0.007  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  150  99  0.261 ± 0.024  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730),Y823, S959  180  100  0.265 ± 0.013  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730),Y823, S959  210  100  0.264 ± 0.007  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730),Y823, S959        D816H  0  0  0.007 ± 0.012    1  20  0.041 ± 0.014  Y568,Y703, Y721, S959  2  31  0.060 ± 0.023  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  4  50  0.090 ± 0.015  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  6  90  0.156 ± 0.021  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  8  93  0.161 ± 0.036  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959  10  100  0.172 ± 0.024  Y547, Y553, Y568, Y570,Y703, Y721, (S729/Y730), S959        Y823F  0  0  0.008 ± 0.005  S943  4  6  0.020 ± 0.003  Y703, Y721, S943  10  7  0.021 ± 0.008  Y547, Y568,Y703, Y721, S729, S943  20  22  0.051 ± 0.031  Y547, Y553, Y568,Y703, Y721, S729,Y747, S943  40  60  0.125 ± 0.006  Y547, Y553, Y568, Y570,Y703, Y721, S729, Y747, S943  60  96  0.194 ± 0.041  Y547, Y553, Y568, Y570,Y703, Y721, S729, Y747, S943  120  100  0.203 ± 0.002  Y547, Y553, Y568, Y570, Y703, Y721, S729, Y747, Y900, S937, S943  JM domain are depicted in italic; KID domain are depicted by underline; activation loop are depicted by bold; C-terminal domain are in the normal font; unable to distinguish between either of these two phosphorylation sites are included in the parenthesis. View Large In the course of WT c-Kit auto-activation, phosphorylation first takes place in the JM domain on tyrosine residues Y547, Y553, Y568 and Y570. Phosphorylation at these positions accounts for 85% of what’s needed for full activation. The activation loop tyrosine Y823 is subsequently phosphorylated, but does not further contribute to activation. In addition to these two phosphorylation events, multiple residues in the KID are phosphorylated as well, including S729 and/or Y730, Y703 and Y721. Assignment of pS729 and pY730 was not possible for WT and D816H c-Kit since there was insufficient MS–MS fragmentation data for this peptide in either trypsin or chymotrypsin digestions. It is not clear whether phosphorylation in KID contributes to kinase activation as it takes place concurrently with the phosphorylation in JM domain. pY703 and pY721 in KID have been postulated to be responsible for relaying signals to downstream effector proteins such as Grb2 (21) and Csk homologous kinase (22) in the process of signal transduction. The same experiment was performed for a sunitinib resistant mutant of c-Kit, D816H. Recent studies indicated that the A-loop mutant D816H is activated 184-fold faster than WT c-Kit and the increased rate of auto-activation leads to resistance to sunitinib in patients with Gastrointestinal Stromal Tumors (18). As shown in Table I, phosphorylation of D816H follows exactly the same path as WT c-Kit albeit on a much shorter timescale. Under similar conditions (0.7 µM c-Kit with 4 mM ATP) full activation was observed after only 6 min at 4°C, compared to 3 h for the WT enzyme at 25°C. The second order rate constant for auto-activation (kact) was reported to be 0.25 and 46 mM–1s–1 for WT and D816H, respectively (18). After full activity was reached, the same phosphorylation sites are found in D816H and WT, with the exception of Y823 in D816H. This residue was phosphorylated only after extended incubation with 4 mM ATP. Thus, phosphorylation of Y823 is not required for the activation of D816H, and is likely not needed for WT c-Kit as well. Activation loop phosphorylation is a common mechanism of controlling kinase activation. Y823 is the only potential tyrosine phosphorylation site on the A-loop of c-Kit. If a phosphorylated Y823 is not required for activation, what role does the A-loop play in c-Kit, and what is the significance of Y823? To address these questions, Y823F, Y823A and Y823D were expressed and purified for further studies. While no detectable kinase activity was observed for Y823A and Y823D after extended activation with ATP, Y823F mutant can be promptly activated to a similar specific activity as WT and D816H. The binding studies in Fig. 2 indicate Y823D and Y823F bind sunitinib with similar affinity as WT (18). This result suggests that the conformation of these variants remain unaffected, particularly in the area responsible for binding sunitinib. However the mutation Y823A most likely leads to misfolded enzyme as it is unable to bind sunitinib. Phosphopeptide-mapping experiments indicate that phosphorylation of Y823F follows the same path as WT and D816H, with virtually all identical residues phosphorylated (Table I). These results suggest that Y823F substitution does not alter the sequence of phosphorylation on other residues that lead to activation of c-Kit. However, given that Y823D is enzymatically impaired and cannot undergo auto-activation, and that Y823F is activated at a much accelerated rate compared to WT c-Kit, it seems that an appropriate side chain at this position such as Tyr or Phe is necessary for c-Kit activity and that substitution of Y823 with a negatively charged residue does not mimic phosphorylated Y823 in the context studied here. Fig. 2 View largeDownload slide Binding of sunitinib to unactivated Y823F, Y823D and Y823A. The change in intrinsic Trp fluorescence that occurs upon the interaction with sunitinib was used to monitor the binding of sunitinib to unactivated mutant forms of c-Kit. Samples of unactivated enzyme (100 nM) in kinase activity assay buffer were titrated with sunitinib. The decrease in fluorescence that occurs upon sunitinib binding was monitored using PTI fluorimeter with excitation and emission wavelengths of 293 and 340 nm respectively. Binding curves were fit to the Morrison equation as described in (18) yielding KD values of 5.9 and 4.8 nM for Y823F (Panel A) and Y823D (Panel B) mutants. These values are within 3- to 4-fold those reported for the WT kinase (18). No detectable change in fluorescence is observed with Y823A (Panel C) variant by titration of up to 200 nM of the drug. Fig. 2 View largeDownload slide Binding of sunitinib to unactivated Y823F, Y823D and Y823A. The change in intrinsic Trp fluorescence that occurs upon the interaction with sunitinib was used to monitor the binding of sunitinib to unactivated mutant forms of c-Kit. Samples of unactivated enzyme (100 nM) in kinase activity assay buffer were titrated with sunitinib. The decrease in fluorescence that occurs upon sunitinib binding was monitored using PTI fluorimeter with excitation and emission wavelengths of 293 and 340 nm respectively. Binding curves were fit to the Morrison equation as described in (18) yielding KD values of 5.9 and 4.8 nM for Y823F (Panel A) and Y823D (Panel B) mutants. These values are within 3- to 4-fold those reported for the WT kinase (18). No detectable change in fluorescence is observed with Y823A (Panel C) variant by titration of up to 200 nM of the drug. As shown in Table I, after a 1 h activation reaction for Y823F, full activation is achieved and no increase in kinase activity is seen with extended incubation with ATP. However, phosphorylation at the KID (Y747) and the C-terminal domain (Y900, S937 and S943) continues. This may suggest that the phosphorylation of the KID in Y823F mutant contributes very little to the kinase activation. Despite having the same final phosphorylated state and similar enzymatic activity as WT enzyme, mutant Y823F and the sunitinib resistant variant D816H auto-activate much faster than WT c-Kit. Quantitative determination of the rate of auto-activation and catalytic efficiency of activated c-Kit was performed in an integrated assay system that couples active c-Kit production with a c-Kit catalysed poly(Glu-Tyr) phosphorylation. Fig. 3 shows representative progression curves of Y823F and WT c-Kit in this assay wherein the activation of c-Kit leads to continuous increase in kinase activity. This results in exponential accumulation of kinase reaction product ADP. Experiments performed at several concentrations of WT and Y823F variant provide data to fit to equation 1 in the ‘Materials and Methods’ section and afford determination of the activation and catalytic rate constants, kact and kcat, respectively. Accordingly, a kact of 0.7 mM–1s–1 and a kcat of 2.8 s–1 for Y823F were obtained compared to a kact of 0.04 mM–1s–1 and a kcat of 3.6 s–1 for WT (Table II). These results indicate that activated Y823F has a similar kinase activity as WT, but the auto-activation proceeds at a 18-fold faster rate than WT. Fig. 3 View largeDownload slide Auto-activation progress curves for WT and Y823F c-Kit. Unactivated c-Kit (WT and Y823F) was incubated with ATP in the presence of 10 mM MgCl2 to initiate auto-activation reaction. The readout of activated enzyme is from the phosphorylation of poly Glu-Tyr in an ATP regeneration system as described in the ‘Materials and Methods’ section. Fitting of the curves to Equation (1) gives auto-activation rates (kact) of 0.04 and 0.7 mM–1s–1 for WT and Y823F mutants respectively. (inset) Reaction scheme for the activation of c-Kit by auto-phosphorylation. Once activated, c-Kitact catalyses the phosphorylation of poly Glu-Tyr substrate in NADH coupled reaction system. Once sample is completely activated, a linear reaction progress curve is observed with poly Glu-Tyr as a substrate. Fig. 3 View largeDownload slide Auto-activation progress curves for WT and Y823F c-Kit. Unactivated c-Kit (WT and Y823F) was incubated with ATP in the presence of 10 mM MgCl2 to initiate auto-activation reaction. The readout of activated enzyme is from the phosphorylation of poly Glu-Tyr in an ATP regeneration system as described in the ‘Materials and Methods’ section. Fitting of the curves to Equation (1) gives auto-activation rates (kact) of 0.04 and 0.7 mM–1s–1 for WT and Y823F mutants respectively. (inset) Reaction scheme for the activation of c-Kit by auto-phosphorylation. Once activated, c-Kitact catalyses the phosphorylation of poly Glu-Tyr substrate in NADH coupled reaction system. Once sample is completely activated, a linear reaction progress curve is observed with poly Glu-Tyr as a substrate. Table II. Auto-activation rates, kinase activity and IC50 values for WT, Y823F and ΔJMΔKID-c-Kit.   WT   Y823F  ΔJMΔKID    Auto-activation  kact (mM−1 s−1)   0.04 ± 0.002    0.7 ± 0.1  >150  kcat (s−1)      3.6 ± 0.2    2.8 ± 1.2  –    Steady-State Kinase Activity  Km (10−6 M)  115.0 ± 46.1  273.5 ± 160  91.5 ± 26.1  Vmax (10−8 M s−1)   2.46 ± 0.54   1.86 ± 0.12  3.50 ± 1.55    IC50 activated (µM)  sunitinib   20.5 ± 0.7  0.200 ± 0.090  >30  imatinib    2.0 ± 1.4  0.051 ± 0.022  >30    IC50 unactivated (µM)  sunitinib  0.068 ± 0.020  0.027 ± 0.018  0.033 ± 0.025  imatinib  0.046 ± 0.005  0.093 ± 0.025  0.010 ± 0.001    WT   Y823F  ΔJMΔKID    Auto-activation  kact (mM−1 s−1)   0.04 ± 0.002    0.7 ± 0.1  >150  kcat (s−1)      3.6 ± 0.2    2.8 ± 1.2  –    Steady-State Kinase Activity  Km (10−6 M)  115.0 ± 46.1  273.5 ± 160  91.5 ± 26.1  Vmax (10−8 M s−1)   2.46 ± 0.54   1.86 ± 0.12  3.50 ± 1.55    IC50 activated (µM)  sunitinib   20.5 ± 0.7  0.200 ± 0.090  >30  imatinib    2.0 ± 1.4  0.051 ± 0.022  >30    IC50 unactivated (µM)  sunitinib  0.068 ± 0.020  0.027 ± 0.018  0.033 ± 0.025  imatinib  0.046 ± 0.005  0.093 ± 0.025  0.010 ± 0.001  View Large Inhibitor sensitivity of WT and mutants The clinically observed mutation D816H confers resistance to sunitinib. Enzymological and structural studies demonstrate that the drug resistance is a result of the greatly increased rate of auto-activation of the mutant that leads to a rapid loss of molecules in the preferred unactivated conformation for binding sunitinib. Thus, sunitinib may be used to probe the equilibrium dynamic between activated and unactivated conformations of Y823F in phosphorylated or unphosphorylated states. As demonstrated in Fig. 4A, sunitinib prefers unactivated c-Kit conformation and is a more potent inhibitor of unactivated WT c-Kit than fully activated, phosphorylated enzyme. IC50 values were determined to be >20 µM and 68 nM for activated and unactivated c-Kit, respectively. The same experiment was performed using Y823F shown in Fig. 4B. Unlike WT c-Kit, fully activated Y823F is much more sensitive to sunitinib with an IC50 = 200 nM, whereas the unactivated form of the enzyme remains similar to WT c-Kit, giving an IC50 = 27 nM. Thus, the affinity for sunitinib binding to activated and unactivated Y823F only differs by 7-fold, in comparison to >300-fold for WT c-Kit. Similar results were obtained using imatinib as the probe (Table II). Imatinib is known to prefer unactivated conformation of the kinase. As such, Y823F seems to be more accommodating than the WT enzyme in shifting the equilibrium dynamic between activated and unactivated conformations, such that when sunitinib is present, fully activated Y823F can quickly adopt the unactivated conformation preferred by sunitinib binding. The increased dynamic in conformation does not affect ATP binding or kinase catalysis. As shown in Fig. 4C, ATP Km and Vmax of Y823F remain very similar to that of WT c-Kit. Fig. 4 View largeDownload slide Kinetic parameters and IC50 values for c-Kit WT and Y823F. IC50 curves for WT (A) and Y823F (B) c-Kit. For activated form, samples of c-Kit were pre-incubated with saturating ATP prior to IC50 assay as described in the ‘Materials and Methods’ section. (C) Kinetics for WT and A-loop mutant Y823F. ATP serially diluted 2-fold from 1 mM. Twelve point saturation curves fit to Michaelis–Menton equation yielding Km and Vmax values of 115.0 µM and 2.46 × 10–8 M s–1 for WT and 273.5 µM and 1.86 × 10–8 M s–1 for Y823F. Vmax values derived from fitting were determined using the active enzyme concentration measured by titration with tight binding inhibitor sunitinib as described in the Material and Methods section. Fig. 4 View largeDownload slide Kinetic parameters and IC50 values for c-Kit WT and Y823F. IC50 curves for WT (A) and Y823F (B) c-Kit. For activated form, samples of c-Kit were pre-incubated with saturating ATP prior to IC50 assay as described in the ‘Materials and Methods’ section. (C) Kinetics for WT and A-loop mutant Y823F. ATP serially diluted 2-fold from 1 mM. Twelve point saturation curves fit to Michaelis–Menton equation yielding Km and Vmax values of 115.0 µM and 2.46 × 10–8 M s–1 for WT and 273.5 µM and 1.86 × 10–8 M s–1 for Y823F. Vmax values derived from fitting were determined using the active enzyme concentration measured by titration with tight binding inhibitor sunitinib as described in the Material and Methods section. Previous biochemical and structural studies show that the JM domain plays a critical role in the activation of c-Kit (15, 16, 18, 23). Phosphopeptide mapping data presented in this study indicate that the initial phosphorylation occurs on four tyrosine residues on the JM domain and that the specific activity of the enzyme can be largely attributed to phosphorylation of these residues. Thus, the JM domain seems to function as a built-in inhibitor of the enzyme, and phosphorylation of the domain moves it away from the auto-inhibitory position. Is releasing the JM domain from the auto-inhibitory position sufficient to activate c-Kit? As shown in Table II, a c-Kit construct with the JM and KID domains deleted is indeed activated almost instantly upon exposing to ATP (kact > 150 mM–1s–1), consistent with the previously observed instantaneous activation of an oncogenic JM mutant V560D (18). The ΔJMΔKID-c-Kit was also tested against sunitinib and imatinib. Like D816H, the activated ΔJMΔKID-c-Kit is insensitive to the two DFG-out inhibitors (IC50 > 30 µM, Table II). But unlike D816H where the drug sensitivity quickly diminishes during auto-activation, unactivated ΔJMΔKID-c-Kit remains sensitive to the drugs (Table II). This sensitivity is maintained throughout the 16-h course of the auto-activation reaction giving IC50 = 540 and 5 nM for sunitinib and imatinib after 16 h, respectively (18). Therefore, removal of JM domain does not significantly affect the A-loop such that ΔJMΔKID-c-Kit maintains its ability to bind the two DFG-out inhibitors in an unactivated conformation. These results suggest that the activation process requires at least two transitional domain movements: Release of JM domain from its auto-inhibitory position followed by the A-loop transitioning to the DGF-in state. Removal of the JM domain alone does not automatically convert the kinase from the unactivated to activated state. Taken together, results from the phosphopeptide mapping and IC50 determinations suggest that for WT c-Kit, the sensitivity to sunitinib should be maintained if the A-loop is unphosphorylated. Y823 phosphorylation occurs late in the auto-activation, as such it should be possible to capture the kinase in a state that is active yet still remains sensitive to sunitinib. To examine this, we measured the IC50 as a function of auto-activation time. As shown in Fig. 5, the overlay of IC50 and the relative activity show the curves are not superimposable. Under the conditions tested, at 300 min, the kinase has reached full activity yet the IC50 is 37 nM. Only after extended activation time of 480 min does the IC50 reach the upper limit of detection of 30 µM. Thus, simply reaching full activity by autophosphorylation of tyrosine residues in the JM domain does not correlate with insenstivity to sunitinib, but rather A-loop phosphorylation is the major determinant that locks the conformation of c-Kit into a state that is not compatible with high affinity binding of sunitinib. Fig. 5 View largeDownload slide IC50 as a function of auto-activation time for WT c-Kit. WT c-Kit (0.5 μM) was activated by addition of 4 mM ATP in the presence of 10 mM MgCl2. Reaction was then quenched by the addition of 12 mM EDTA. Samples were then diluted to 10 nM in buffer A and the IC50 for sunitinib was measured. Activity of c-Kit reaches a maximum at 300 min while the course of IC50 increases at a slower rate, reflecting the slow phosphorylation of Y823 in the auto-activation process. Under these conditions, intermediate phosphorylated forms of c-Kit in which Y823 is unphosphorylated are the predominant species from 0–360 min auto-activation time. Fig. 5 View largeDownload slide IC50 as a function of auto-activation time for WT c-Kit. WT c-Kit (0.5 μM) was activated by addition of 4 mM ATP in the presence of 10 mM MgCl2. Reaction was then quenched by the addition of 12 mM EDTA. Samples were then diluted to 10 nM in buffer A and the IC50 for sunitinib was measured. Activity of c-Kit reaches a maximum at 300 min while the course of IC50 increases at a slower rate, reflecting the slow phosphorylation of Y823 in the auto-activation process. Under these conditions, intermediate phosphorylated forms of c-Kit in which Y823 is unphosphorylated are the predominant species from 0–360 min auto-activation time. Discussion It has been observed that there are multiple phosphorylation sites in the JM, KID and C-terminal domains c-Kit (7, 21, 22, 24–27). In addition to those responsible for kinase activation in the JM domain, several phosphorylated residues have been implicated in the recruitment of downstream effector proteins in both the JM and KID domains, e.g. pY568 for the adaptor protein Lnk (28) and pY703/pY721 for Grb2 and PI 3-kinase respectively (2)· Deletion of the KID results in decreased in SCF-stimulated PI-3 kinase activity in T18 and COS-1 cells (26, 29). Data presented here indicate that the activation process of c-Kit occurs sequentially with respect to the JM and activation loop tyrosine phosphorylation. In addition to these sites, the autophosphorylation reaction leads to the presence of phospho-serine sites (S729, S937, S943 and S959), these most likely are due to non-specific activities as no other reports indicate c-Kit is a dual specificity kinase in cells. In addition, the data reported here have been determined utilizing the soluble cytoplasmic kinase domain which may alleviate constraints on kinase specificity imposed by membrane proximal regions in the context of the full length receptor. The earliest detected phosphorylation sites in all three variants are located on the KID domain (Y703 and Y721) as shown in Table I. Phosphorylation of the KID domain is unlikely to be required for activation since KID truncated c-Kit retains full kinase activity (data not shown). Residues on the KID are essential for binding downstream effectors [reviewed in (1, 2)], thus initial phosphorylation of the KID would facilitate the rapid transduction of signals in response to SCF stimulation. Phosphopeptide mapping results presented here are in contrast to the autophosphorylation of fibroblast growth factor receptor 1 (FGFR1) receptor kinase domain, in which initial autophosphorylation of a A-loop tyrosine (Y653) leads to 50- to 100-fold increase in catalytic activity followed by phosphorylation of residues involved in binding signalling molecules (30, 31). Further 10-fold activation is achieved by phosphorylation of a second tyrosine in the A-loop (Y654) in the last step of auto-activation. These differences reflect diverse mechanisms of controlling activation of the cytoplasmic kinase domain. c-Kit is maintained in the unactivated state primarily through auto-inhibition by the JM domain blocking the catalytic site in addition to the A-loop tyrosine interacting with the catalytic aspartate (D792) whereas FGFR is auto-inhibited by a conserved proline (P663) and residues N-terminal which are situated to obstruct the protein substrates-binding site [reviewed in (32)]. Full activation of FGFR1 kinase domain requires phosphorylation of two adjacent A-loop tyrosine residues, Y653 and Y654 (33). Both Y823F and D816H variants of c-Kit are auto-activated much faster than WT. While the mechanism is not completely clear, it is possible that the two substitutions result in a subtle conformational change that doesn’t involve the sunitinib-binding site however it renders unactivated c-Kit more susceptible to auto-activation. After activation, although the two mutants are almost indistinguishable from WT with respect to their kinase activity, they respond very differently to inhibitors that prefer a DFG-out conformation such as sunitinib and imatinib. Active WT and D816H lose their sensitivity to sunitinib completely (IC50 > 20,000 nM), whereas Y823F remains very sensitive to the drug (IC50 = 200 nM). Such a difference may reflect the flexibility of activated enzyme in adopting an unactivated conformation. In the activated state, despite being phosphorylated on the same residues, Y823F and WT may be very different in their conformational dynamic such that Y823F can readily adopt the DFG-out conformation and bind sunitinib or imatinib, whereas WT and D816H can not. Thus, mutation of Y823 to F does not prevent the enzyme from activation, but it leads to a more accommodating c-Kit that remains sensitive to the drugs after activation. Accordingly, pY823 in WT is likely to be responsible for stabilizing or ‘locking’ the A-loop at the active position. Is Y823 of any significance in the unactivated state of c-Kit? Since mutant Y823F can be activated successfully, it seems reasonable to expect the same activation with a variety of mutations at this position. Surprisingly, Y823D renders the enzyme unactivatable although fully competent to bind sunitinib. This result suggests that an appropriate side chain at this position is necessary for unactivated c-Kit to maintain a conformation susceptible to auto-activation. In the physiological context, Y823 phosphorylation may be critical for stabilizing the activated state and preventing it from reverting back to the unactive form. As such, pY823 may be an important determinant for relaying downstream signalling events that require the kinase to be ‘locked’ in the activated state. The effects of the Y823F mutation on the c-Kit-signalling pathway are the subject of further studies. Taking into account the observations presented in this study, we propose a model for the activation of c-Kit shown in Fig. 6. In this model, there are two critical determinants for maintaining the unactivated state. First, the JM domain inhibits enzyme activity in cis by blocking access to the catalytic site and second, the activation loop is stabilized by the aromatic side chain of tyrosine or phenylalanine. These two mechanisms maintain the enzyme in an auto-inhibited or unactivated state. Upon activation, the JM domain is rapidly phosphorylated on tyrosine residues 547, 553, 568 and 570, thus releasing it from its auto-inhibitory position between the N- andC-terminal lobes. The resulting phosphorylated kinase remains sensitive to the inhibitors sunitinib and imatinib that bind to the DFG-out unactivated conformation as the A-loop is unphosphorylated. With longer activation time, phosphorylation of the A-loop tyrosine 823 occurs, resulting in decreased flexibility of the activation loop which is then locked into the active conformation. The ensemble of conformational states is shifted in going from Y823 to pY823 forms. Mutations in the activation loop may result in either a gain (Y823F) or loss (D816H) of sensitivity to inhibitors that prefer unactivated conformation such as imatinib and sunitinib. Fig. 6 View largeDownload slide Model for the activation of WT c-Kit kinase domain. Initial phosphorylation of tyrosine residues Y547, Y553, Y568, Y570 in the JM domain result in activation of the kinase. However, the activated kinase in this phosphorylation state lacking Y823 phosphorylation remains sensitive to sunitinib as the A-loop remains flexible. A much slower phosphorylation on Y823 occurs and this leads to decreased flexibility and insensitivity to DFG-out specific compounds sunitinib and imatinib. When Y823 is mutated to Phe, the last step cannot occur and the kinase remains sensitive to sunitinib. Mutation of Y823 to Asp is not an effective mimic of the phosphorylated state as this kinase unactivatable. Fig. 6 View largeDownload slide Model for the activation of WT c-Kit kinase domain. Initial phosphorylation of tyrosine residues Y547, Y553, Y568, Y570 in the JM domain result in activation of the kinase. However, the activated kinase in this phosphorylation state lacking Y823 phosphorylation remains sensitive to sunitinib as the A-loop remains flexible. A much slower phosphorylation on Y823 occurs and this leads to decreased flexibility and insensitivity to DFG-out specific compounds sunitinib and imatinib. When Y823 is mutated to Phe, the last step cannot occur and the kinase remains sensitive to sunitinib. Mutation of Y823 to Asp is not an effective mimic of the phosphorylated state as this kinase unactivatable. Conflict of interest None declared. Acknowledgements We thank Dr Elizabeth A. Lunney for helpful discussion. References 1 Roskoski RJr..  Structure and regulation of Kit protein-tyrosine kinase – the stem cell factor receptor, Biochem,  Biophys. Res. Commun. ,  2005, vol.  338 (pg.  1307- 1315) Google Scholar CrossRef Search ADS   2 Roskoski RJr..  Signaling by Kit protein-tyrosine kinase – the stem cell factor receptor, Biochem,  Biophys. Res. Commun. ,  2005, vol.  337 (pg.  1- 13) Google Scholar CrossRef Search ADS   3 Yarden Y,  Kuang WJ,  Yang-Feng T,  Coussens L,  Munemitsu S,  Dull TJ,  Chen E,  Schlessinger J,  Francke U,  Ullrich A.  Human proto-oncogene c-kit: a new cell surface receptor tyrosine kinase for an unidentified ligand,  EMBO J. ,  1987, vol.  6 (pg.  3341- 3351) Google Scholar PubMed  4 Herbst R,  Shearman MS,  Jallal B,  Schlessinger J,  Ullrich A.  Formation of signal transfer complexes between stem cell and platelet-derived growth factor receptors and SH2 domain proteins in vitro,  Biochemistry ,  1995, vol.  34 (pg.  5971- 5979) Google Scholar CrossRef Search ADS PubMed  5 Lemmon MA,  Pinchasi D,  Zhou M,  Lax I,  Schlessinger J.  Kit receptor dimerization is driven by bivalent binding of stem cell factor,  J. Biol. Chem. ,  1997, vol.  272 (pg.  6311- 6317) Google Scholar CrossRef Search ADS PubMed  6 Herbst R,  Lammers R,  Schlessinger J,  Ullrich A.  Substrate phosphorylation specificity of the human c-kit receptor tyrosine kinase,  J. Biol. Chem. ,  1991, vol.  266 (pg.  19908- 19916) Google Scholar PubMed  7 Blume-Jensen P,  Wernstedt C,  Heldin CH,  Ronnstrand L.  Identification of the major phosphorylation sites for protein kinase C in kit/stem cell factor receptor in vitro and in intact cells,  J. Biol. Chem. ,  1995, vol.  270 (pg.  14192- 14200) Google Scholar CrossRef Search ADS PubMed  8 Linnekin D.  Early signaling pathways activated by c-Kit in hematopoietic cells,  Int. J. Biochem. Cell. Biol. ,  1999, vol.  31 (pg.  1053- 1074) Google Scholar CrossRef Search ADS PubMed  9 Huse M,  Kuriyan J.  The conformational plasticity of protein kinases,  Cell ,  2002, vol.  109 (pg.  275- 282) Google Scholar CrossRef Search ADS PubMed  10 Johnson LN,  Noble ME,  Owen DJ.  Active and inactive protein kinases: structural basis for regulation,  Cell ,  1996, vol.  85 (pg.  149- 158) Google Scholar CrossRef Search ADS PubMed  11 Boerner JL,  Demory ML,  Silva C,  Parsons SJ.  Phosphorylation of Y845 on the epidermal growth factor receptor mediates binding to the mitochondrial protein cytochrome c oxidase subunit II,  Mol. Cell. Biol. ,  2004, vol.  24 (pg.  7059- 7071) Google Scholar CrossRef Search ADS PubMed  12 Yang S,  Park K,  Turkson J,  Arteaga CL.  Ligand-independent phosphorylation of Y869 (Y845) links mutant EGFR signaling to stat-mediated gene expression, Exp,  Cell Res. ,  2008, vol.  314 (pg.  413- 419) Google Scholar CrossRef Search ADS   13 Murray BW,  Padrique ES,  Pinko C,  McTigue MA.  Mechanistic effects of autophosphorylation on receptor tyrosine kinase catalysis: enzymatic characterization of Tie2 and phospho-Tie2,  Biochemistry ,  2001, vol.  40 (pg.  10243- 10253) Google Scholar CrossRef Search ADS PubMed  14 Maulik G,  Bharti A,  Khan E,  Broderick RJ,  Kijima T,  Salgia R.  Modulation of c-Kit/SCF pathway leads to alterations in topoisomerase-I activity in small cell lung cancer,  J. Environ. Pathol. Toxicol. Oncol. ,  2004, vol.  23 (pg.  237- 251) Google Scholar CrossRef Search ADS PubMed  15 Mol CD,  Lim KB,  Sridhar V,  Zou H,  Chien EY,  Sang BC,  Nowakowski J,  Kassel DB,  Cronin CN,  McRee DE.  Structure of a c-kit product complex reveals the basis for kinase transactivation,  J. Biol. Chem. ,  2003, vol.  278 (pg.  31461- 31464) Google Scholar CrossRef Search ADS PubMed  16 Mol CD,  Dougan DR,  Schneider TR,  Skene RJ,  Kraus ML,  Scheibe DN,  Snell GP,  Zou H,  Sang BC,  Wilson KP.  Structural basis for the autoinhibition and STI-571 inhibition of c-Kit tyrosine kinase,  J. Biol. Chem. ,  2004, vol.  279 (pg.  31655- 31663) Google Scholar CrossRef Search ADS PubMed  17 Matsudaira PT.  A Practical Guide to Protein and Peptide Purification for Microsequencing,  1993  2nd edn, Academic Press, USA 18 Gajiwala KS,  Wu JC,  Christensen J,  Deshmukh GD,  Diehl W,  DiNitto JP,  English JM,  Greig MJ,  He YA,  Jacques SL,  Lunney EA,  McTigue M,  Molina D,  Quenzer T,  Wells PA,  Yu X,  Zhang Y,  Zou A,  Emmett MR,  Marshall AG,  Zhang HM,  Demetri GD.  KIT kinase mutants show unique mechanisms of drug resistance to imatinib and sunitinib in gastrointestinal stromal tumor patients,  Proc. Natl Acad. Sci. USA ,  2009, vol.  106 (pg.  1542- 1547) Google Scholar CrossRef Search ADS   19 Morrison JF.  Kinetics of the reversible inhibition of enzyme-catalysed reactions by tight-binding inhibitors,  Biochim. Biophys. Acta ,  1969, vol.  185 (pg.  269- 286) Google Scholar CrossRef Search ADS PubMed  20 Williams JW,  Morrison JF.  The kinetics of reversible tight-binding inhibition,  Methods Enzymol ,  1979, vol.  63 (pg.  437- 467) Google Scholar PubMed  21 Thommes K,  Lennartsson J,  Carlberg M,  Ronnstrand L.  Identification of Tyr-703 and Tyr-936 as the primary association sites for Grb2 and Grb7 in the c-Kit/stem cell factor receptor,  Biochem. J. ,  1999, vol.  341(Pt 1) (pg.  211- 216) Google Scholar CrossRef Search ADS PubMed  22 Price DJ,  Rivnay B,  Fu Y,  Jiang S,  Avraham S,  Avraham H.  Direct association of Csk homologous kinase (CHK) with the diphosphorylated site Tyr568/570 of the activated c-KIT in megakaryocytes,  J. Biol. Chem. ,  1997, vol.  272 (pg.  5915- 5920) Google Scholar CrossRef Search ADS PubMed  23 Mol CD,  Fabbro D,  Hosfield DJ.  Structural insights into the conformational selectivity of STI-571 and related kinase inhibitors,  Curr. Opin. Drug Discov. Devel. ,  2004, vol.  7 (pg.  639- 648) Google Scholar PubMed  24 Kozlowski M,  Larose L,  Lee F,  Le DM,  Rottapel R,  Siminovitch KA.  SHP-1 binds and negatively modulates the c-Kit receptor by interaction with tyrosine 569 in the c-Kit juxtamembrane domain,  Mol. Cell. Biol. ,  1998, vol.  18 (pg.  2089- 2099) Google Scholar CrossRef Search ADS PubMed  25 Lennartsson J,  Wernstedt C,  Engstrom U,  Hellman U,  Ronnstrand L.  Identification of Tyr900 in the kinase domain of c-Kit as a Src-dependent phosphorylation site mediating interaction with c-Crk,  Exp. Cell. Res. ,  2003, vol.  288 (pg.  110- 118) Google Scholar CrossRef Search ADS PubMed  26 Serve H,  Hsu YC,  Besmer P.  Tyrosine residue 719 of the c-kit receptor is essential for binding of the P85 subunit of phosphatidylinositol (PI) 3-kinase and for c-kit-associated PI 3-kinase activity in COS-1 cells,  J. Biol. Chem. ,  1994, vol.  269 (pg.  6026- 6030) Google Scholar PubMed  27 Timokhina I,  Kissel H,  Stella G,  Besmer P.  Kit signaling through PI 3-kinase and Src kinase pathways: an essential role for Rac1 and JNK activation in mast cell proliferation,  EMBO J. ,  1998, vol.  17 (pg.  6250- 6262) Google Scholar CrossRef Search ADS PubMed  28 Gueller S,  Gery S,  Nowak V,  Liu L,  Serve H,  Koeffler HP.  Adaptor protein Lnk associates with Tyr(568) in c-Kit,  Biochem. J. ,  2008, vol.  415 (pg.  241- 245) Google Scholar CrossRef Search ADS PubMed  29 Lev S,  Givol D,  Yarden Y.  Interkinase domain of kit contains the binding site for phosphatidylinositol 3′ kinase,  Proc. Natl Acad. Sci. USA ,  1992, vol.  89 (pg.  678- 682) Google Scholar CrossRef Search ADS   30 Furdui CM,  Lew ED,  Schlessinger J,  Anderson KS.  Autophosphorylation of FGFR1 kinase is mediated by a sequential and precisely ordered reaction,  Mol. Cell ,  2006, vol.  21 (pg.  711- 717) Google Scholar CrossRef Search ADS PubMed  31 Lew ED,  Furdui CM,  Anderson KS,  Schlessinger J.  The precise sequence of FGF receptor autophosphorylation is kinetically driven and is disrupted by oncogenic mutations,  Sci. Signaling ,  2009, vol.  2 pg.  ra6  Google Scholar CrossRef Search ADS   32 Hubbard SR,  Mohammadi M,  Schlessinger J.  Autoregulatory mechanisms in protein-tyrosine kinases,  J. Biol.Chem. ,  1998, vol.  20 (pg.  11987- 11990) Google Scholar CrossRef Search ADS   33 Mohammadi M,  Schlessinger J,  Hubbard SR.  Structure of the FGF receptor tyrosine kinase domain reveals a novel autoinhibitory mechanism,  Cell ,  1996, vol.  86 (pg.  577- 587) Google Scholar CrossRef Search ADS PubMed  Abbreviations Abbreviations RTK receptor tyrosine kinase SCF stem cell factor JM Juxtamembrane KID kinase insert domain © The Authors 2010. Published by Oxford University Press on behalf of the Japanese Biochemical Society. All rights reserved
TI - Function of activation loop tyrosine phosphorylation in the mechanism of c-Kit auto-activation and its implication in sunitinib resistance
JO - The Journal of Biochemistry
DO - 10.1093/jb/mvq015
DA - 2010-03-26
UR - https://www.deepdyve.com/lp/oxford-university-press/function-of-activation-loop-tyrosine-phosphorylation-in-the-mechanism-MAHDjGDbAJ
SP - 601
EP - 609
VL - 147
IS - 4
DP - DeepDyve
ER -