The C terminus of RON tyrosine kinase plays an autoinhibitory role.

RON is a receptor tyrosine kinase in the MET family. We have expressed and purified active RON using the Sf9/baculovirus system. The constructs used in this study comprise the kinase domain alone and the kinase domain plus the C-terminal region. The construct containing the kinase domain alone has a higher specific activity than the construct containing the kinase and C-terminal domains. Purified RON undergoes autophosphorylation, and the exogenous RON C terminus serves as a substrate. Peptides containing a dityrosine motif derived from the C-terminal tail inhibit RON in vitro or when delivered into intact cells, consistent with an autoinhibitory mechanism. Phenylalanine substitutions within these peptides increase the inhibitory potency. Moreover, introduction of these Phe residues into the dityrosine motif of the RON kinase leads to a decrease in kinase activity. Taken together, our data suggest a model in which the C-terminal tail of RON regulates kinase activity via an interaction with the kinase catalytic domain.

RON is a receptor tyrosine kinase in the MET family. We have expressed and purified active RON using the Sf9/baculovirus system. The constructs used in this study comprise the kinase domain alone and the kinase domain plus the C-terminal region. The construct containing the kinase domain alone has a higher specific activity than the construct containing the kinase and C-terminal domains. Purified RON undergoes autophosphorylation, and the exogenous RON C terminus serves as a substrate. Peptides containing a dityrosine motif derived from the C-terminal tail inhibit RON in vitro or when delivered into intact cells, consistent with an autoinhibitory mechanism. Phenylalanine substitutions within these peptides increase the inhibitory potency. Moreover, introduction of these Phe residues into the dityrosine motif of the RON kinase leads to a decrease in kinase activity. Taken together, our data suggest a model in which the C-terminal tail of RON regulates kinase activity via an interaction with the kinase catalytic domain.
RON is a receptor tyrosine kinase that is involved in cell proliferation, survival, and motility (1)(2)(3). RON is one of the three members of the MET family of receptor tyrosine kinases (MET, RON, and SEA) (4,5). The MET family members share a number of unique structural properties, most notably an ␣␤ disulfide-linked heterodimeric structure. RON is composed of a 40-kDa extracellular ␣ chain and a 150-kDa transmembrane ␤ chain with intrinsic protein kinase activity (4,6,7).
RON is expressed in a variety of cells, including epithelial cells and macrophages (7, 8 -11). The ligand for the RON receptor is macrophage-stimulating protein (MSP), 1 also known as hepatocyte growth factor-like protein (6,7). In normal cells, MSP binding leads to a transient increase in RON activity, whereas tumor cells often possess elevated levels of RON protein, expression of altered forms of RON, and increased RON kinase activity (12)(13)(14).
Upon activation of RON, the receptor becomes phosphorylated within the activation loop of the kinase catalytic domain, and the enzymatic activity of RON is enhanced. RON also possesses two tyrosine residues in the C-terminal tail in a motif (Y 1353 VQLPAT 1360 YMNL) that is conserved in all MET family members (15). Ligand stimulation of MET family members leads to phosphorylation of these two tyrosines (Tyr 1353 and Tyr 1360 ). These phosphorylated tyrosine residues provide multifunctional docking sites for the p85 regulatory subunit of phosphatidylinositol 3-kinase (16), the Grb2⅐SOS complex (15,17), STAT3 (18), and the adaptor protein Gab1 (19 -21). Mutation of these two tyrosines (Y1353F/Y1360F) suppressed the transforming ability of activated forms of RON (15). A similar mutation caused a complete loss of transforming ability of the related SEA kinase (22). These results could be the result of the inability of the double mutant to engage SH2 domain-containing downstream signaling proteins. Another possibility, however, is that the intrinsic kinase activity of the Y1353F/Y1360F mutant is altered. Bardelli et al. (23) demonstrated that peptides containing C-terminal sequences inhibited MET kinase activity in vitro. A peptide derived from the C-terminal tail impaired MET-induced invasive growth in transformed epithelial cells. These studies suggested that the carboxy-terminal domain may act as an intramolecular modulator of MET receptor (23).
At present, the regulatory mechanism of MET family kinases is unclear, partly because no enzymatic studies have been carried out on a purified member of this family. We report the first purification and characterization of active RON from eukaryotic cells, using the Sf9/baculovirus expression system. We present evidence that the C-terminal region of RON plays an autoinhibitory role.
Baculovirus Expression Vectors-The RON kinase-C-terminal tail construct (RON-CT) was generated by PCR. This construct encoded amino acids Ala 1065 -Thr 1400 of RON. The PCR 5Ј-primer had the sequence CGCGGATCCGGCGCTCTTGGCTGAGGTCAAG, and the 3Ј-primer was GGAATTCGGAGTGGGCCGAGGAGGCTCTGAGAG. These primers had 31 nucleotides (5Ј-primer) and 33 nucleotides (3Ј-primer) of complementarily with the template and encoded unique restriction sites (BamHI at the 5Ј-end and EcoRI at the 3Ј-end). The PCR product was ligated into plasmid pCR-BluntII-TOPO (Invitrogen). The resulting plasmid was digested with BamHI/EcoRI, and the RON insert was purified on an agarose gel. The RON fragment was subcloned into plasmid pBACgus-9 (N-terminal T7 tag, C-terminal CBD tag, and polyhistidine tag; Novagen), and expressed in Sf9 cells using the Bacvector-3000 DNA transfection kit (Novagen). The 2YF mutant form (RON-2YF) was generated by sitedirected mutagenesis of the pBACgus-9 RON-CT construct using the QuikChange mutagenesis system (Stratagene). The mutation was confirmed by automated DNA sequencing.
To produce the isolated kinase catalytic domain of RON (RON-KIN), PCR was carried out with the 5Ј-primer 5Ј-CGCGGATCCGGCGCTCT-TGGCTGAGGTCAAG and the 3Ј-primer 5Ј-GGAATTCGGCACTATCT-GCTCCACCTCCCC. These primers had 31 and 30 nucleotides of complementarily with the template, respectively, and encoded unique restriction sites (BamHI at the 5Ј-end and EcoRI at the 3Ј end). The RON-KIN construct encodes amino acids Ala 1065 -Val 1345 . Baculovirus expression of RON-KIN was carried out using the methods described above for RON-CT. For production of RON-CT and RON-KIN proteins, 0.6 liter of Sf9 cells (1.8 ϫ 10 6 cells/ml) were infected with recombinant baculovirus at a multiplicity of infection of 7.5 and 9.0, respectively. After 4 days of infection, cells were harvested and washed twice with phosphate-buffered saline.
Purification of RON-CT-Sf9 Cells were lysed in a French pressure cell in 20 mM Tris-HCl buffer (pH 8.0) containing 2 mM Na 3 VO 4 , 5 mM 2-mercaptoethanol, 10 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride. After centrifugation at 40,000 ϫ g for 30 min, the supernatant was filtered and applied to a 1.6 ϫ 10-cm Source Q FPLC column (Amersham Biosciences), which was preequilibrated with homogenizing buffer. Flow-through fractions were applied to a 4-ml Ni-NTA column (Qiagen). The column was washed with buffer containing 20 mM imidazole, 0.5 M NaCl, 2 mM Na 3 VO 4 , 10% glycerol, 5 mM 2-mercaptoethanol, 20 mM Tris-HCl (pH. 8.0), and 1 M NaCl. RON-CT was eluted with buffer containing 100 mM imidazole, 5 mM 2-mercaptoethanol, 2 mM Na 3 VO 4, 10% glycerol, and 20 mM Tris-HCl (pH. 8.0). RON kinase activity was measured using the phosphocellulose paper binding assay (25) with peptide EAIYAAPFAKKKG as a substrate. Peaks of activity were pooled and concentrated. For some preparations, additional purification was carried out on a Mono Q FPLC column. The column was washed with buffer containing 0.05 M NaCl, and RON-CT was eluted with a linear gradient of 0.05-0.3 M NaCl in the same buffer. The RON-KIN and RON-2YF proteins were purified by similar methods.
In some experiments, the C-terminal tags were removed proteolytically from RON-CT. 100 g of RON-CT was digested with 0.25 milliunit of thrombin for 2.5 h at 30°C. The digested enzyme was applied to a Superdex 200 gel filtration column to separate away the cleaved CBD and polyhistidine tags.
RON Kinase Assay Using Synthetic Peptide Substrates-RON kinase activity was determined using the phosphocellulose paper assay. Reaction mixtures contained 20 mM Tris-HCl (pH 7.4), 10 mM MgCl 2 , 0.1 mM Na 3 VO 4 , 0.5 mM dithiothreitol, 0.25 mM ATP, 2.5 mg/ml albumin, varying concentrations of peptide substrate, and [␥-32 P]ATP (200 -400 cpm/ pmol). Reactions were terminated by the addition of 50% acetic acid, and samples were spotted on P-81 phosphocellulose paper (25). Incorporation of 32 P into peptide was determined by liquid scintillation counting. The value of K m (peptide) was determined using a range of peptide concentrations (0.05-2.0 mM) and 0.25 mM [␥-32 P]ATP. Kinetic parameters were calculated by fitting data to the Michaelis-Menten equation. For some experiments, a continuous spectrophotometric assay was employed (26). In this assay, production of ADP is coupled to the oxidation of NADH, which is measured as a reduction in absorbance at 340 nm.
Autophosphorylation of RON-Purified RON proteins were incubated with 0.25 mM [␥-32 P]ATP (400 -700 cpm/pmol) in kinase buffer containing 20 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 0.5 mM dithiothreitol, and 0.1 mM Na 3 VO 4 at 30°C. To analyze the effects of synthetic peptides on RON autophosphorylation, RON was preincubated with peptides for 10 min prior to the addition of [␥-32 P]ATP. Reactions were stopped by the addition of SDS-sample buffer and analyzed by SDS-PAGE and autoradiography.
Western Blotting-Proteins were analyzed on 8% SDS-polyacrylamide gels and transferred to Immobilon membrane (Millipore, Bedford, MA) in the presence of 0.1% SDS. The membranes were blocked using 5% milk in Tris-buffered saline plus 0.1% Tween 20, then probed with the appropriate antibodies. Blots were visualized using horseradish peroxidase-conjugated second antibody with ECL (enhanced chemiluminescence, Amersham Biosciences) or alkaline phosphatase-conjugated second antibody with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium as substrates.
GST-RON Fusion Proteins-GST-RON65 (amino acids 1335-1400 of RON) and GST-RON100 (amino acids 1300 -1400) were expressed in Escherichia coli DH5␣ cells as GST fusion proteins. The fusion proteins were purified using glutathione-agarose, as described previously (27). For phosphorylation experiments, 0.125 g of purified RON-CT was incubated with 2.0-g GST-RON65 or GST-RON100 fusion protein plus 0.25 mM [␥-32 P]ATP in kinase assay buffer for 40 min. Reactions were stopped by the addition of SDS-sample buffer, then subjected to SDS-PAGE. Phosphorylation was analyzed by autoradiography. Purified GST was used as a control.
Delivery of Peptides to NIH3T3 Cells-RON-NIH3T3 cells (28) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C with 7% CO 2 . Chariot-peptide complexes were prepared according to the manufacturer's protocol. RON-NIH3T3 cells were serum starved for 1 h and incubated further with Chariot-peptide complexes for 3 h. Cells were stimulated with 100 ng/ml MSP for 30 min before harvesting. Cell lysates were applied to a 10% SDS-polyacrylamide gel and analyzed by Western immunoblotting with anti-phosphotyrosine, anti-RON, anti-phospho-MAPK, and MAPK antibodies. (Fig. 1A). We purified RON-CT to homogeneity by chromatography on Source-Q and Ni-NTA columns (Fig. 1B). Kinase activity was monitored during the purification by the phosphocellulose paper binding assay, using the Abl consensus peptide EAIYAAPFAKKKG as a substrate. Purified RON-CT migrates with the expected molecular mass (Ϸ55.4 kDa) (Fig. 1B). RON-CT reacted with a rabbit polyclonal antibody raised against GST-RON100 (residues 1300 -1400), as well as with anti-phospho-RON antibody, suggesting that the purified protein is autophosphorylated or phosphorylated by an endogenous Sf9 cell kinase (Fig. 1B).

Purification of RON and Kinetics of Peptide Phosphorylation-We produced a RON construct containing the kinase domain plus the C-terminal tail (RON-CT) by infection of Sf9 cells with a recombinant baculovirus vector
As an initial test of the substrate specificity of RON kinase, we carried out experiments with four peptides containing recognition motifs for different subfamilies of tyrosine kinases. The kinase recognition motifs included were for Src, Abl, EGFR, and insulin receptor. RON-CT preferred the EGFR substrate from this group of peptides (Table I). We carried out initial rate kinetic measurements with saturating concentrations of ATP and varying concentrations of peptides. These experiments yielded a K m for the EGFR peptide of 520 M and a k cat of 13.72 min Ϫ1 (k cat /K m ϭ 26.2 ϫ 10 Ϫ3 min Ϫ1 M Ϫ1 ) ( Table I). The next best substrate for RON, Abl peptide, was phosphorylated with a k cat /K m ϭ 11 ϫ 10 Ϫ3 min Ϫ1 M Ϫ1 , ϳ2.4 times lower than the EGFR substrate (Table I). Phosphorylation of the Src and insulin receptor-specific peptides was barely detectable. Thus, our kinetic experiments show that RON phosphorylates the EGFR peptide with a k cat /K m value Ͼ 17-fold higher than that seen with a preferred substrate for another RTK, insulin receptor kinase. Previous experiments with immunoprecipitated RON also showed that RON preferred the EGFR peptide sequence (29). In those studies, however, a Src substrate sequence was phosphorylated roughly equally to the EGFR sequence, whereas in our studies (Table I) Autophosphorylation of RON-Many tyrosine kinases are regulated by autophosphorylation within the activation loop, a segment that lies between the N-and C-terminal lobes of the catalytic domain. MET family receptors contain a pair of tyrosine residues in the activation loop (tyrosines 1238 and 1239 in RON). For the MET receptor itself, autophosphorylation has been mapped to the residues corresponding to Tyr 1238 and Tyr 1239 , and phosphorylation at these sites activates MET (30,31). By Western blotting with an antibody that recognizes RON that is doubly phosphorylated at Tyr 1238 /Tyr 1239 , we detected phosphorylation of RON-CT after expression in Sf9 cells (Fig.  1B). We investigated whether purified RON-CT can undergo autophosphorylation. Purified RON-CT was incubated with [␥-32 P]ATP in kinase buffer, and the reaction mixtures were analyzed by SDS-PAGE and autoradiography. RON-CT was autophosphorylated in a time-dependent manner, and preincubation of RON-CT with unlabeled ATP reduced the level of autophosphorylation (Fig. 2).
Phosphorylation of the RON C-terminal dityrosine motif (Y1353/Y1360) has been suggested to be caused by autophosphorylation by RON itself (15)(16)(17), although it is also possible that another kinase phosphorylates these residues. To test whether purified RON has the capacity to phosphorylate these C-terminal tyrosines, we synthesized two peptides (peptides Y1353 and Y1360) corresponding to these sequences. Peptide Y1353 contains residues 1346 -1359 (SALLGDHYVQLPAT), and peptide Y1360 contains residues 1354 -1367 (VQLPA-TYMNLGPST). We also prepared a synthetic peptide (peptide Y1353/Y1360) containing the entire dityrosine motif (residues 1349 -1367; LGDHYVQLPATYMNLGPST). These peptides showed essentially no phosphorylation (data not shown). We next tested two GST fusion proteins containing residues 1335-1400 (GST-RON65) and 1300 -1400 (GST-RON100) from the RON C-terminal tail. These GST fusion proteins were incubated with purified RON-CT and [␥-32 P]ATP in kinase buffer. Autoradiography showed 32 P incorporation into GST-RON65 and GST-RON100 as well as into RON-CT itself (Fig. 3). GST alone was not a substrate for RON. GST-RON65 in particular contains no other tyrosines besides Tyr 1353 and Tyr 1360 ; thus, these results show that RON has the capacity to phosphorylate these sites, at least in the context of an exogenous fusion protein.
Peptides Derived from the C-terminal Tail Inhibit RON Kinase Activity-It has been demonstrated previously that a peptide corresponding to the C-terminal tail of the MET receptor inhibits MET kinase activity (23). To investigate whether the RON C-terminal tail regulates kinase activity, we tested peptides Y1353, Y1360, and Y1353/Y1360 as potential inhibitors. We also synthesized a peptide in which the two tyrosines in the C-terminal sequence (residues 1349 -1367) were replaced with phenylalanines (peptide F1353/F1360). In these experiments, we first removed the CBD and His tags from RON-CT by treatment with thrombin to avoid any interference by the tags in our assays. RON-CT activity was measured toward 0.5 mM Abl peptide substrate in the presence of various concentrations of the C-terminal peptides. As shown in Fig. 4A, peptide F1353/ F1360 dramatically inhibited RON-CT activity. Peptide Y1353 also showed inhibition, whereas peptides Y1360 and Y1353/ Y1360 had very little effect (Fig. 4A). Consistent with the results shown in Fig. 4A, phosphorylation of GST-RON65 and  GST-RON100 was also inhibited by peptide F1353/F1360 but not by peptide Y1353/Y1360 (Fig. 4B). We also analyzed the effects of the synthetic peptides on RON-CT autophosphorylation (Fig. 4C). The results of this experiment correlated well with the results presented in Fig. 4A; peptide F1353/F1360 was the most effective inhibitor of RON-CT autophosphorylation, and peptide Y1353 also gave inhibition at higher concentrations. Peptides Y1360 and Y1353/Y1360 were inactive in this assay (Fig. 4C).
We investigated whether these peptides are specific for RON kinase. We carried out inhibition experiments using the purified catalytic domain of another receptor tyrosine kinase, the insulin-like growth factor I receptor (IGF1R). We measured phosphorylation of a specific IGF1R peptide substrate by the triply phosphorylated, activated form of IGF1R. Even at concentrations as high as 1 mM, none of the RON C-terminal peptides gave any inhibition of IGF1R (data not shown).

Deletion of the C-terminal Tail Enhances RON Kinase Activity-To test further the idea that the RON C terminus is inhibitory, we produced a form of RON containing the kinase catalytic domain alone (RON-KIN). RON-KIN was expressed in
Sf9 insect cells and purified to homogeneity using a strategy similar to that for RON-CT. RON-KIN migrates with the expected molecular mass (Ϸ50.9 kDa) and reacts with anti-RON antibody (Fig. 5A). RON-KIN undergoes autophosphorylation, demonstrating that sites exist within the catalytic domain itself (data not shown). We compared the specific activity of RON-CT and RON-KIN using 0.5 mM Abl peptide as substrate. Initial experiments confirmed that the peptide assay is linear over the time period examined (Fig. 5B). RON-KIN had a specific activity ϳ3-fold higher than that of RON-CT. We carried out kinetic measurements with saturating concentrations of ATP and varying concentrations of EGFR peptide (Fig. 5C). RON-KIN had a K m for the EGFR peptide of 745 M and a k cat value of 77.82 min Ϫ1 (k cat /K m ϭ 104.46 ϫ 10 Ϫ3 min Ϫ1 M Ϫ1) . Thus, the k cat /K m for RON-KIN was about four times higher than for RON-CT (k cat /K m ϭ 26.2 ϫ 10 Ϫ3 min Ϫ1 M Ϫ1 ; Table I).
The higher activity of RON-KIN suggests that the C terminus may contain an autoinhibitory element. We tested whether RON-KIN would be susceptible to inhibition by the C-terminal peptides. As shown above for RON-CT (Fig. 4), we carried out RON-KIN kinase assays with varying concentrations of the C-terminal peptides, using the Abl peptide as substrate. Peptides Y1353, Y1353/Y1360, and F1353/F1360 all showed significant inhibition in this experiment (Fig. 6A). Peptide Y1360 did not inhibit RON-KIN. Inhibition of RON-KIN was observed at lower concentrations of peptides than was observed for RON-CT (compare Figs. 4A and 6A). These results are consistent with a model in which the peptides can more easily access the kinase domain in the absence of the C-terminal tail. The results on peptide phosphorylation were mirrored in studies of RON-KIN autophosphorylation (Fig. 6B). Peptides Y1353/ Y1360 and F1353/F1360 showed complete inhibition of RON-KIN autophosphorylation at the lowest concentration tested (25 M). Peptide Y1353 gave inhibition at a slightly higher concentration (125 M). To gain a better understanding of the mechanism of inhibition, we carried out kinetic measurements with varying concentrations of peptide F1353/F1360. Inhibition of RON-KIN by peptide F1353/F1360 was found to be competitive with respect to peptide substrate (Fig. 6C).
Tyr to Phe Substitutions in RON-CT Decrease Kinase Activity-Because peptide F1353/F1360 was a more effective inhibitor than peptide Y1353/Y1360 (Figs. 4 and 6), we produced a mutant form of RON (designated RON-2YF) containing Tyr to Phe substitutions at positions 1353 and 1360 in the tail. We expressed RON-2YF in Sf9 cells and purified the enzyme to homogeneity using procedures similar to those described above for RON-CT. The C-terminal mutations severely impaired RON kinase activity (Fig. 5D). The k cat /K m for RON-2YF (using EGFR peptide as substrate) was 0.6 ϫ 10 Ϫ3 min Ϫ1 M Ϫ1 , ϳ43 times lower than for RON-CT. Our interpretation (see "Discussion") is that the C-terminal tail of the 2YF mutant form of RON is engaged in a strengthened interaction with the kinase domain (relative to the wild-type RON-CT).
RON Inhibition in Intact Cells-To test the ability of the RON C-terminal peptides to inhibit full-length RON receptor, we introduced peptides Y1353/Y1360 and F1353/F1360 into NIH3T3 cells using Chariot, a commercial kit for protein transduction. As a control, we used the Src peptide substrate, the weakest of the RON kinase substrates (Table I). MSP treatment led to an increase in tyrosine phosphorylation of RON in these cells (Fig. 7A). Although the control peptide had no effect on RON activation at the two concentrations tested (9 and 90 M), peptide Y1353/Y1360 gave some inhibition at the higher concentration. Peptide F1353/F1360 showed inhibition at both concentrations, with a substantial reduction in phosphorylation at 90 M (Fig. 7A). The blot was reprobed with RON antibody to confirm that peptide treatment did not affect RON expression (Fig. 7A). We carried out similar experiments to examine MAPK activation in these cells. Treatment of the NIH3T3 cells with MSP led to a stimulation of MAPK, as detected by phospho-MAPK Western blotting (Fig. 7B). In these experiments, peptides Y1353/Y1360 and F1353/F1360 both gave a significant reduction in MAPK activation at 90 M. These experiments show that the RON C terminus has the capacity to inhibit kinase activity in the context of the intact, ligand-responsive receptor. DISCUSSION An emerging theme in the regulation of receptor tyrosine kinases is that domains outside the core catalytic domain play important roles in regulating kinase activity. The crystal structures of more than a dozen RTKs have been reported, typically of the kinase domains alone. These studies have highlighted the importance of the conformation of the activation loop, which regulates substrate access and the proper positioning of catalytic residues (32)(33)(34). In many cases, however, there may be aspects of the RTK regulatory mechanisms which are not present in these structures; biochemical data for several RTKs show a more complex mechanism of activation. The juxtamembrane region (JM) of some RTKs clearly plays a negative role. Tyrosine residues in the JM region of several RTKs function as direct inhibitors of catalytic function and also serve as recruitment sites for various downstream signaling proteins. These  (37), and c-Kit (38). The crystal structures of the autoinhibited forms of these three kinases have been solved, and they demonstrate the structural basis for the negative regulatory role of the JM region in RTKs. The JM regions of these RTKs prevent the activation loop from adopting an active conformation, and phosphorylation within the JM region relieves autoinhibition. In the case of c-Kit, addition of a peptide corresponding to the JM region inhibited kinase activity, providing biochemical evidence for the intramolecular regulation (39). Biochemical experiments have implicated the C-terminal portion of the cytoplasmic domain of several RTKs in kinase regulation. Substitution of the C-terminal residues of c-Fms with residues of v-Fms enhanced receptor autophosphorylation and transforming activity (40). The deletion of the C-terminal residues of ErbB2 increased receptor kinase activity as well as transforming ability (41). A peptide containing the C terminus of platelet-derived growth factor ␤ receptor inhibited its kinase activity (42). The crystal structure of the Tie2 RTK has been solved with an intact C-terminal tail. The Tie2 C-terminal tail is present in an extended conformation. The tail interacts with the C-lobe of the kinase catalytic domain and ends near the substrate binding site. The activation loop of Tie2 is in an "active-like" conformation (43), and the end of the C-terminal tail might prevent access of substrates to the active site. The side chain hydroxyls of two tyrosine residues (Tyr 1101 and Tyr 1112 ) are hydrogen bonded to surrounding residues and may stabilize this conformation of the C-terminal tail. Deletion of the Tie2 C-terminal tail increased receptor autophosphorylation and kinase activity, lending further support to a model in which the tail is involved in kinase regulation (44).
The RON receptor studied here is closely related to the MET receptor, for which a crystal structure has been solved which includes the C-terminal tail (45). The tail of MET includes two unphosphorylated tyrosines that correspond to the dityrosine motif studied here. The first tyrosine (Tyr 1349 in MET) is in an extended conformation, whereas the second (Tyr 1356 ) is part of a type I ␤ turn. As in the case of Tie2, the MET C-terminal tail approaches the substrate binding site, but it is not clear from the structure whether the tail regulates substrate access or kinase activity. Previous biochemical experiments suggested a role for the MET C-terminal tail in kinase regulation. MET was immunoprecipitated from mammalian cells, and the addition of a peptide mimicking the C terminus inhibited autophosphorylation and phosphorylation of myelin basic protein (23). Delivery of this peptide in intact cells inhibited a variety of MET signaling functions.
Here we provide biochemical evidence for C-terminal regulation of RON, a receptor tyrosine kinase in the MET family. Our experiments were carried out with purified RON, ruling out the possibility that copurifying cellular components were responsible for the effects. The C-terminal tail region of RON, especially the bidentate tyrosine motif, appears to serve an autoinhibitory function. This is evident from the higher specific activity of the kinase domain alone (RON-KIN) compared with the kinase domain with the C-terminal tail (RON-CT) (Fig. 5).
In agreement with previous results on MET (23), we showed that peptides derived from C-terminal tail of RON inhibited both substrate and autokinase activity in vitro. The inhibitory effect was more pronounced in the case of RON-KIN than for RON-CT and was competitive with respect to peptide substrate. In the case of MET, a peptide containing Phe replacements in the dityrosine motif gave similar inhibition to the natural sequence, whereas we found (Figs. 4, 6) that the Phecontaining RON peptide was a more effective inhibitor. This could be caused in part by a strengthened interaction with the kinase catalytic domain. To test this idea, we carried out preliminary experiments to compare the binding of peptides  but RON-CT showed a higher affinity for peptide F1353/F1360 than for peptide Y1353/Y1360. 2 Our results suggest a model in which the C-terminal tail (Y 1353 VQLPAT 1360 YMNL) interacts with the catalytic domain and partially inhibits RON kinase activity (Fig. 8). Upon ligand stimulation, the C terminus tail would be displaced from its autoinhibitory position. Oligomerization-induced conformational changes together with phosphorylation of the activation loop may trigger this displacement of the C-terminal tail. We also speculate that the C-terminal tail becomes phosphorylated during this process. It is unclear whether this phosphorylation is catalyzed by RON itself or by another kinase. Synthetic peptides containing the C-terminal tail sequence were not phosphorylated by RON, whereas larger C-terminal constructs (GST-RON65 and GST-RON100) were substrates. These larger proteins may contain additional determinants for substrate recognition which are absent in the peptides. In our model, the 2YF substitution in the C-terminal tail results in strengthened binding to the catalytic domain, and this interaction interferes with access of substrates to the active site. This would explain the decreased activity of the RON-2YF construct (Fig. 5D).
Peptide Y1353 was a more effective inhibitor of RON-CT and RON-KIN than peptide Y1360 under all conditions tested (Figs. 4 and 6). These results could indicate that the side chain of unphosphorylated Y1353 is particularly important in autoinhibition. It is also possible, however, that there are N-terminal amino acids present only in peptide Y1353 which strengthen binding to RON in the context of a synthetic peptide (or Cterminal residues present in peptide Y1360 which weaken binding). This might also explain why peptide Y1353 inhibited RON-CT more potently than peptide Y1353/Y1360, which includes both residues of the dityrosine motif (Fig. 4). Peptides Y1353 and Y1353/Y1360 were equally effective as inhibitors of RON-KIN (Fig. 6). The mechanism for autoinhibition remains speculative in the absence of structural information.
After phosphorylation, the bidentate motif interacts with several different signaling proteins to activate multiple downstream signaling pathways (15,46). Among the SH2 domaincontaining proteins that interact with the bidentate motif are Grb2 and the p85 subunit of phosphatidylinositol 3-kinase (15)(16)(17). Tyr to Phe mutations in the bidentate motif have been shown to eliminate MSP-stimulated growth of Ba/F3 pro-B cells and apoptosis of mouse erythroleukemia cells (46). Similar mutations blocked the cell scattering and migration activities of RON-overexpressing 293 cells in response to MSP (47). Although the mutations clearly block interaction with downstream signaling partners, our results on the RON-2YF mutant indicate that lowered kinase activity may also play a role in these studies.   (tyrosines 1238 and 1239), as well as a bidentate motif containing two tyrosines (1353 and 1360) in the Cterminal tail (dark line). The bidentate motif interacts with the catalytic domain and partially inhibits RON kinase activity (left). Upon ligand stimulation, residues 1238/1239 in the activation loop become phosphorylated, the C-terminal tail is displaced from the substrate binding region, and RON activity increases (right). Replacement of residues Tyr 1353 and Tyr 1360 with Phe strengthens the binding of the C-terminal tail, leading to inhibition of RON. Open circles, unphosphorylated tyrosine; filled circles, phosphorylated.