A novel mutation within the extracellular domain of TrkA ... - Nature

7 downloads 0 Views 162KB Size Report
Salamanca, Spain; 2Departamento de Ciencias Morfologicas, Universidad de Zaragoza, Spain; 3Hematology/Oncology ..... Porter AC and Vaillancourt RR.
Oncogene (2001) 20, 1229 ± 1234 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

A novel mutation within the extracellular domain of TrkA causes constitutive receptor activation Juan C Arevalo1, Blanca Conde2, Barbara I Hempstead3, Moses V Chao4, Dionisio MartõÂ n-Zanca1 and Pilar PeÂrez*,1 1

Instituto de Microbiologia Bioquimica, Departamento de Microbiologia y Genetica, CSIC Universidad de Salamanca, 37007 Salamanca, Spain; 2Departamento de Ciencias Morfologicas, Universidad de Zaragoza, Spain; 3Hematology/Oncology Division, Cornell University Medical College, New York, NY 10021, USA; 4Molecular Neurobiology Program, Skirball Institute, Departments of Cell Biology and Physiology and Neuroscience, New York University Medical Center, New York, NY 10016, USA

The TrkA NGF receptor extracellular region contains three leucine repeats ¯anked by cysteine clusters and two immunoglobulin-like domains that are required for speci®c ligand binding. Deletion of the immunoglobulinlike domains abolishes NGF binding and causes ligand independent activation of the receptor. Here we report a speci®c mutation that increases the binding anity of the TrkA receptor for NGF. A change of proline 203 to alanine (P203A) in the linker region between the leucine repeats and the ®rst Ig-like domain increased NGF binding by decreasing the ligand rate of dissociation. This mutated receptor was appropriately expressed on the cell surface and promoted ligand-independent neurite outgrowth in PC12nnr5 cells. The mutant receptor was capable of spontaneous dimerization and was constitutively phosphorylated in the absence of ligand. Moreover, expression of TrkA-P203A receptor in ®broblasts induced DNA synthesis and transformation and generated tumours in nude mice. These data suggest that domains outside of the immunoglobulin-like structure contribute to ligand binding and constitutive activation of Trk receptors. Oncogene (2001) 20, 1229 ± 1234. Keywords: receptor; tyrosine kinases; dimerization; cell transformation

Receptor tyrosine kinases (RTK) participate in many di€erent biological processes such as cell survival, proliferation or di€erentiation. All RTKs contain an ectodomain responsible for ligand binding, a single transmembrane domain and an intracellular tyrosine kinase domain. Upon ligand binding, RTKs undergo conformational changes that induce and stabilize dimerization of the receptor and subsequent autophosphorylation. Gain-of-function mutations resulting in constitutive RTK activation have been linked to numerous human disorders (for reviews see Porter and Vaillancourt, 1998; Robertson et al., 2000). These

*Correspondence: P PeÂrez Received 20 October 2000; revised 28 December 2000; accepted 3 January 2001

mutations cause spontaneous dimerization of the receptors and result in the activation of their intrinsic catalytic activity. The TrkA receptor is the founding member of a subfamily of RTKs, which also includes TrkB and TrkC that encode receptors for the nerve growth factor (NGF) family of neurotrophins. NGF-triggered TrkA signalling has been involved in growth, survival and/or di€erentiation of neurons in the central nervous system and in neural crest derived cells (review in Barbacid, 1995; Lewin and Barde, 1996; Chao, 2000; Kaplan and Miller, 2000). The extracellular region of TrkA is characterized by a number of structural motifs (Schneider and Shweiger, 1991): three amino-terminal leucine repeats (LRM) ¯anked by two cysteine clusters and two immunoglobulin-like I-set type domains which participate in ligand binding (Perez et al., 1995; Urfer et al., 1995; MacDonald and Meakin, 1996; Holden et al., 1997). A co-crystal of NGF bound to the second IgG of TrkA con®rmed that the majority of binding contacts were between this IgG domain and NGF (Wiesmann et al., 1999), but the contribution of other extracellular regions of TrkA to the dimerization and activation of the receptor is incompletely understood. Indeed, there have been several reports implicating the involvement of other regions of the Trk receptors such as the leucine repeats in ligand binding (Windisch et al., 1995a,b; Ninkina et al., 1997). TrkA gain of function mutations have been described in some human tumours. Activation occurs through genomic rearrangements that generate a hybrid kinase containing a novel N-terminus that favours constitutive dimerization (Greco et al., 1997, 1998). Di€erent TrkA oncogenic forms were also identi®ed by in vitro transformation of NIH3T3 in culture (Coulier et al., 1990) and by analysis of several deletion and chimeric TrkA mutants (Arevalo et al., 2000). Also, the extent of Trk receptor dimerization and activation can be in¯uenced by speci®c point mutations in the second Ig-like domain (Arevalo et al., 2000). These studies were consistent with a model in which the two Ig-like domains of the TrkA receptor are involved in binding to NGF and regulate the dimerization of the receptor. TrkA constitutive activation can also be achieved through intracellular

Constitutive TrkA activation JC Arevalo et al

1230

mutations in the TrkA activation loop suggesting that the autophosphorylation of the activation loop tyrosines allows the interaction with nearby basic residues that helps stabilize a functional active conformation (Cunninghan and Green, 1998). To further understand the mechanism of TrkA dimerization and the role played by other regions in the extracellular domain, we generated several single amino acid mutations in the ®rst Ig-like domain and in the linker region between this domain and the leucine repeats (LRM) of TrkA. Speci®cally, one of the cysteine residues that can form a disul®de bridge in the Ig-like structure was changed to serine (C266S), and a proline residue located at the linker region between the leucine repeats and the ®rst Ig-like domain was mutated to alanine (P203A) (Figure 1a). As a

Figure 1 (a) Schematic representation of the mutations within the TrkA receptor. (b) Scatchard plot analysis of 125I-NGF binding to TrkA (&) and TrkA-P203A (*) was performed as described (Perez et al., 1995). HEK293 cells stably transfected with those receptors were incubated with increasing concentrations of 125I-NGF in the presence or absence of an excess of unlabelled NGF for 2 h at 48C. (c) Dissociation of 125I-NGF from cells expressing TrkA (&) and TrkA-P203A (*) receptors. Cells were incubated with 125I-NGF (5610711 M) for 2 h at 48C, then samples were 100-fold diluted in PBS containing 1 mg/ml BSA and 1 mg/ml glucose. Values represent percentage of speci®c binding before dilution. The mean of at least three independent experiments with triplicate samples was calculated and error bars re¯ect the standard deviation Oncogene

control, we used the previously described L92V, L95V double mutation that disrupted the ®rst leucine repeat of the TrkA extracellular domain (Arevalo et al., 2000). Equilibrium 125I-NGF binding assays were performed with HEK293 cells expressing TrkA, TrkA-L92VL95V, TrkA-P203A and TrkAC266S. As described previously, the double mutation L92VL95V did not a€ect NGF binding (Arevalo et al., 2000; data not shown). By contrast, the C266S point mutant in the ®rst Ig-like domain eliminated NGF binding (data not shown), verifying that this domain is somehow involved in ligand-binding (Perez et al., 1995; Holden et al., 1997). The topology of the ®rst Ig-like domain closely resembles that of V-CAM (J Murray-Rust and N McDonald, personal communication). The bond between Cys217 and Cys266 corresponds to the standard immunoglobulin disulphide bridge of the I-set type of immunoglobulins (Smith and Xue, 1997). The C266S mutation probably indirectly interfered with NGFbinding by altering the Ig-like structure of the domain. Interestingly, the P203A mutation increased the NGF anity of the receptor (Figure 1b) (Kd 0.561079 M). This mutation resided outside the two Ig-like domains, a€ecting a proline residue located before the ®rst Ig-like domain. Additional experiments were performed to determine the rate of dissociation of NGF from the TrkA-P203A mutant receptor, compared to that of the wild type receptor. The results indicated that the increase in the equilibrium Kd constant for the mutant TrkA-P203A receptor could in part be due to a slower rate of dissociation (Figure 1c). Addition of NGF to PC12 rat pheochromocytoma cells activates TrkA tyrosine kinase activity, resulting in the cessation of cell proliferation and initiation of neuronal di€erentiation. To study the ability of TrkA point mutants to induce di€erentiation in the presence or absence of NGF, we transfected the cDNA encoding the mutant receptors into the PC12nnr5 cell line. This mutant cell line was derived from PC12 cells and lost the expression of endogenous TrkA receptors (Green et al., 1986). The cDNAs, cloned in the pCDNA3 plasmid, were co-transfected with an expression plasmid carrying the lacZ gene (pCMV-lacZ). Three days after transfection the percentage of transfected cells (b-galactosidase-positive) bearing neurites twice the length of the cell bodies was scored. Introduction of wild type TrkA receptors induced neurite formation in 7% of the transfected cells in the absence of NGF, and in 60 ± 70% of the cells when NGF was added (Figure 2). An oncogenic version of the receptor, Trk5, that has a 51 residue deletion in the second Ig-like domain (Coulier et al., 1990), induced neurite formation in 38 ± 40% of the transfected cells independently of the NGF presence. The double mutation L92V L95V, as the wild type TrkA, did not cause any signi®cant neurite outgrowth in the absence of NGF and only stimulated neurite outgrowth in 30% of the transfected cells when NGF was added. Interestingly, the TrkA-P203A receptor promoted neurite outgrowth in the absence of NGF. Quantitative measurements indicated that the response from the

Constitutive TrkA activation JC Arevalo et al

Figure 2 Neurite formation in PC12nnr5 cells transfected with TrkA mutant receptors. Neurite outgrowth was quanti®ed 3 days after transfection in the presence (®lled bars) or absence (empty bars) of NGF (100 ng/ml) by assessing the percentage of bgalactosidase-positive cells bearing neurites twice the length of their cell bodies. Values were calculated from at least ®ve independent experiments

TrkA-P203A receptor was comparable to the Trk5 oncogene (39%). This result suggested that the P203A mutation causes spontaneous activation of the receptor. On the other hand, 58% of the cells developed neurites in the presence of saturating concentrations of NGF, a percentage similar to that obtained with wild type TrkA in the presence of NGF. Therefore, mutation of the P203 residue in the TrkA receptor did not interfere with a ligand-dependent response. The TrkA-C266S receptor did not cause any cell di€erentiation with or without NGF. This result is noteworthy since deletion or substitution of the ®rst Iglike domain caused activation of the receptor, and a similarly placed mutation in the second Ig-like domain impeded NGF binding but caused constitutive activation of the receptor (Arevalo et al., 2000). A mutation similar to TrkA-C266S in the stem cell growth factor receptor, cKit, a€ected the disul®de-bonded cysteine residue in the second Ig-like domain (C136R) and disabled the c-Kit receptor to respond to ligand without resulting in receptor activation (Ezoe et al., 1995). It is possible that some mutations alter the structure of the ligand binding domains in such a way that block the binding of the ligand but still favour the monomeric form of the receptor. In any case, the results obtained with TrkAC266S would disfavour the likelihood of intermolecular disul®de bridging described for other receptors, in which some mutations creating unpaired cysteines lead to ligand-independent activation of those receptors (Neilson and Friesen, 1996; Siegel and Muller, 1996). To analyse if the P203A mutation results in ligand independent dimerization of the receptors, we tagged TrkA, TrkA-L92VL95V and TrkA-P203A with either HA or Myc epitopes (Arevalo et al., 2000) and cotransfected them into HEK293 cells. Two HA epitopes were introduced between amino acid 43 and 70, after the signal peptide of the molecule and the sequence encoding the Myc epitope was introduced at the MluI site, previously engineered into rat TrkA cDNA, after

the second Ig-like domain (Perez et al., 1995). The transfected cells were treated with NGF (100 ng/ml) for 10 min before lysis or left untreated. Receptor expression was detected by Western-blotting using either anti-HA, anti-Myc or anti-Trk antibodies. The level of expression was similar for all the HA and Myctagged receptors (Figure 3a). The cell extracts were immunoprecipitated with anti-HA antibodies and immunoprecipitates were detected by Western-blotting using anti-Myc antibodies. As shown in Figure 3b, Myc-TrkA, Myc-TrkA-L92VL95V and Myc-TrkAP203 were present in the HA-immunoprecipitates of the corresponding NGF-treated cells. In contrast, only Myc-TrkA-P203A was detected in the HA-TrkAP203A immunoprecipitates of cells in the absence of NGF. Therefore signi®cant constitutive receptor dimerization occurred in the TrkA-P203A expressing cells, but not in cells carrying TrkA or TrkAL92VL95V tagged receptors. Immunoblot analysis of the HA-immunoprecipitates with anti-Trk (Figure 3b lower panel) demonstrated that the level of receptor present in all the immunoprecipitates was similar.

1231

Figure 3 (a, b) Dimerization of TrkA mutant receptors. HEK293 cells were transiently transfected with the following pairs of expression plasmids: HA-TrkA / Myc-TrkA; HA-TrkAL92VL95V / Myc-TrkA-L92VL95V and HA-TrkA-P203A / MycTrkA-P203A. Cells were grown for 2 days and treated with or without NGF (100 ng/ml) for 10 min before lysis. Cell lysis and immunoprecipitation were performed as described previously (Arevalo, 2000). (a) Western-blot of cell extracts using 12CA5 anti-HA antibody; 9E10 anti-Myc antibody and anti-203 pan-Trk antiserum. (b) Immunoprecipitations performed using the 12CA5 anti-HA antibody and Western-blot done with either 9E10 antiMyc antibody or 203 anti-Trk antiserum. Immunoreactive protein bands were detected by chemiluminescence. (c) Autophosphorylation of TrkA mutant receptors. HEK293 cells transfected with TrkA, TrkA-L92VL95V and TrkA-P203A were grown during 2 days with low serum concentrations (0.5%) and treated with or without NGF (100 ng/ml) for 10 min before lysis. Cell extracts were analysed by Western-blot with 4G10 anti-phosphotyrosine antibody (left panel) or anti-203 pan-Trk antiserum (right panel) Oncogene

Constitutive TrkA activation JC Arevalo et al

1232

Oncogene

Dimerization of receptor tyrosine kinases leads to autophosphorylation of speci®c tyrosine residues within the kinase domain. We therefore analysed if there was ligand-independent phosphorylation of the TrkA-P203A receptors. HEK293 cells were transiently transfected with expression plasmids containing the mutant receptors cDNAs and were kept in low serum (0.5%), with or without NGF. The expression level and the tyrosine phosphorylation state were analysed by Western-blot using anti-203 pan-Trk antiserum (Figure 3c right panel) or 4G10 anti-phosphotyrosine antibody (Figure 3c left panel) respectively. TrkA, TrkA-L92VL95V and TrkA-P203A were phosphorylated in the presence of NGF. In contrast, only the TrkA-P203A receptors were phosphorylated without addition of NGF. These results demonstrated constitutive dimerization and activation of the TrkAP203A mutant receptor. The relative stabilization of the dimer state of the receptor may explain the observed dissociation rate of the bound NGF. Di€erent mutations in the extracellular domain of other RTKs have been described that promote dimerization or conformational changes, without a€ecting the ligand binding domains (Van Daalen et al., 1992; Eng and Mulligan, 1997). Interestingly, substitutions of amino acids P252 in the human FGFR1 and P253 in the FGFR2, are associated with the proliferative Pfei€er and Appert syndromes (Muenke et al., 1994; Wilkie et al., 1995; reviewed in Naski and Ornitz, 1998). Those proline residues are located in the linker region between domains Ig-2 and Ig-3, and the receptors bind slightly more radiolabeled FGF than wild type receptors (Neilson and Friesen, 1996). Recent analysis of the relative disposition of Ig2 and Ig3 domains in the FGFFGFR structures predicts that these mutations result in aberrant interactions between FGFRs and FGFs leading to enhanced signaling when availability of ligand is limiting (Plotnikov et al., 1999, 2000; Stauber et al., 2000). To evaluate potential oncogenic e€ects of the TrkAP203A receptor, transfected Rat-1 cells were subjected to assays for 3H-thymidine incorporation after serum starvation as well as for growth on soft agar. We obtained cell lines derived from stably transfected Rat1 cells with levels of mutant receptor expression similar to those found in PC12 cells, as assessed by immunoprecipitation and Western-blot using panTrkA antiserum (Figure 4a). The amount of 3Hthymidine incorporation was normalized relative to the incorporation obtained in each cell line grown in the presence of 10% serum. The results were expressed as the percentage of incorporation measured in cells expressing the Trk5 oncogene. As expected, cells expressing either TrkA or TrkA-L92VL95V receptors did not show any measurable thymidine incorporation in the absence of serum (Figure 4b). In contrast, cells expressing TrkA-P203A incorporated 3H-thymidine, indicating that those cells were undergoing proliferation, although the level of incorporation was lower than that of Trk5 expressing cells.

Figure 4 (a) TrkA expression levels of Rat-1 derived cell lines stably expressing mutant TrkA receptors. Cell extracts were analysed by immunoprecipitation and Western-blot using anti-203 pan-Trk antiserum. PC12 cells were used as control. (b) [3H]thymidine incorporation by the Rat-1 derived cell lines. 26104 cells/well were incubated for 2 days in DMEM medium with 10% foetal calf serum. Cells were starved for 22 h in medium with 0.2% serum, then, 10% calf serum was added to some cultures for 12 h. [3H]-Thymidine was added (0.5 mCi/ml) and cells were further incubated for 4 h, washed with PBS containing 1 mM CaCl2 and 0.5 mM MgCl2, precipitated with 10% TCA and resuspended in 0.1 ml of 0.2 N NaOH. The solution was neutralized by addition of 0.1 ml of 0.2 N HCl. Data were normalized relative to the [3H]-thymidine incorporated by the cells in the presence of serum and expressed as percentage of the incorporation measured in Trk5 expressing cells (100%). Values are the mean from at least ®ve experiments performed with two independent clones of each mutant, and error bars represent standard deviation. (c) Soft agar colony formation of Rat-1 cell lines expressing mutant TrkA receptors. 103 cells were resuspended in 2 ml DMEM medium with 10% foetal calf serum and 0.5% agar, and overlaid onto plates containing 4 ml of solidi®ed DMEM medium supplemented with 10% foetal calf serum and 1% agar. Foci were counted after 21 days. Data normalized relative to the foci formed by the Trk5 expressing clones (100%) are shown in parenthesis. Values are the means +standard deviations calculated from at least ®ve experiments performed with two independent clones of each mutant. (d) Tumour volume and tumour progression in nude mice injected with 104 Rat-1 cells stably expressing Trk5 or TrkA-P203A. 104 cells/mouse were injected subcutaneously into Swiss nu/nu mice (two sites/mouse). Tumour growth was assessed twice a week by caliper measurements of the tumour in three dimensions. When xenografts reached a mean diameter of 6 mm, the cell lines were scored as tumorigenic. Three mice (two injections/each) were used for each cell line. Tumour progression was calculated according to the equation: (Tumour volume on day n7Tumour volume on day 0)/Tumour volume on day 0. Being day 0 the ®rst day in which tumour growth was observed

To analyse the ability to grow in soft agar, the Rat-1 transfected cell lines were cultured in DMEM solidi®ed medium with 10% calf serum and 0.5% agar during 21

Constitutive TrkA activation JC Arevalo et al

days, and growth was analysed by counting the number of colonies formed. The growth in soft agar was normalized relative to cells expressing the Trk5 oncogene (100%). Cells expressing either TrkA or the TrkA-L92VL95V mutant receptor did not grow in soft agar (Figure 4c). However, TrkA-P203A expressing cells exhibited growth in soft agar and formed colonies. These ®ndings con®rm the transforming capability of TrkA-P203A mutant receptor. We also assessed the ability of the Rat-1 cell lines expressing TrkA-P203A mutant receptors to form tumours when injected in nude mice. The transfected Rat-1 cell line was injected subcutaneously (104 cells/ mouse) and the mice were examined twice per week for tumour formation. In three inoculated mice, the P203A expressing Rat-1 cells progressively formed tumours that were as aggressive as those caused by Trk5 expressing cells (Figure 5a). The period of tumour latency was 9 days for TrkA-P203A cells and 11 days for Trk5 expressing cells, but the tumour progression was faster in the mice injected with Trk5 expressing Rat-1 cells (Figure 5b). Rat-1 cells expressing wild type TrkA did not cause tumours (data not shown). These data correlated with those of the proliferation and soft agar growth assays and indicated that a single point mutation outside of the Ig-like structures caused activation of the TrkA tyrosine kinase activity. It is possible that P203A mutation favours a greater ¯exibility in the region between the LRM and the ®rst

Ig-like domain, resulting in increased receptor dimerization through the LRM. The role of the leucine repeats in ligand-binding is unclear (Urfer et al., 1995; McDonald and Meakin, 1996; Holden et al., 1997) but they might act to favour or stabilize dimerization once NGF is bound. In other RTKs, dimerization domains distinct from the ligand-binding domains has been described (Blechman et al., 1995; Omura et al., 1997). For the TrkA receptor, NGF-induced neurite outgrowth of PC12nnr5 cells expressing the TrkAL92VL95V mutant receptor lacking the ®rst leucine repeat was less ecient than in cells transfected with wild type TrkA. Similarly, there was less NGF-induced dimerization of the TrkA-L92VL95V than of the wild type TrkA receptor, suggesting a modulatory role for the ®rst leucine repeat in TrkA dimerization. Further structural analysis of the entire TrkA extracellular domain will be necessary to understand the precise molecular mechanisms by which mutations such as P203A mutation lead to receptor activation.

1233

Acknowledgements This work was supported by grants from the Fundacion Ramon Areces and the European Union Program BIO4CT96-0285. JC Arevalo was a recipient of fellowships from those grants. Grant support for MV Chao and BL Hempstead were from the NIH.

References Arevalo JC, Conde B, Hempstead BL, Chao MV, MartinZanca D and Perez P. (2000). Mol. Cell. Biol., 20, 5908 ± 5916. Barbacid M. (1995). Curr. Opin. Cell Biol., 7, 148 ± 155. Blechman JM, Lev S, Barg J, Eisenstein M, Vaks B, Vogel Z, Givol D and Yarden Y. (1995). Cell, 80, 103 ± 113. Chao MV. (2000). J. Neurosci. Res., 59, 353 ± 355. Coulier F, Kumar R, Ernst M, Klein R, Martin-Zanca D and Barbacid M. (1990). Mol. Cell. Biol., 10, 4202 ± 4210. Cunnigham ME and Greene LA. (1998). EMBO J., 17, 7282 ± 7293. Eng C and Mulligan LM. (1997). Hum. Mutat., 9, 97 ± 109. Ezoe K, Holmes SA, Ho L, Bennett CP, Bolognia JL, Brueton L, Burn J, Falabella R, Gatto EM, Issii N, et al. (1995). Am. J. Hum. Gen., 56, 58 ± 66. Greco A, Miranda C, Pagliardini S, Fusetti L, Bongarzone I and Pierotti MA. (1997). Genes Chromosomes Cancer, 19, 112 ± 123. Greco A, Fusetti L, Miranda C, Villa R, Zanotti S, Pagliardini S and Pierotti MA. (1998). Oncogene, 16, 809 ± 816. Green SH, Rydel RE, Connolly JL and Greene LA. (1986). J. Cell Biol., 102, 830 ± 843. Holden PH, Asopa V, Robertson AGS, Clarke AR, Tyler S, Bennett GS, Brain SD, Wilcock GK, Allen SJ, Smith SKF and Dawbarn D. (1997). Nature Biotechnol., 15, 668 ± 672. Kaplan DR and Miller FD. (2000). Curr. Opin. Neurobiol., 10, 381 ± 391. Lewin GR and Barde YA. (1996). Ann. Rev. Neurosci., 19, 289 ± 317.

MacDonald JIS and Meakin SO. (1996). Mol. Cell. Neurosci., 7, 371 ± 390. Muenke M, Schell U, Hehr A, Robin NH, Losken HW, Schinzel A, Pulleyn LJ, Rutland P, Reardon W, Malcolm S and Winter RM. (1994). Nature Genet., 8, 269 ± 274. Naski MC and Ornitz DM. (1998). Front. Biosci., 3, 781 ± 794. Neilson KM and Friesel R. (1996). J. Biol. Chem., 271, 25049 ± 25057. Ninkina N, Grashchuck M, Buchman VL and Davies AM. (1997). J. Biol. Chem., 272, 13019 ± 13025. Omura T, Heldin CH and Ostman. (1997). J. Biol. Chem., 272, 12676 ± 12682. Perez P, Coll PM, Hempstead BL, Martin-Zanca D and Chao MV. (1995). Mol. Cell. Neurosci., 6, 97 ± 105. Plotnikov AN, Hubbard SR, Schlessinger J and Mohammadi M. (2000). Cell, 101, 413 ± 424. Plotnikov AN, Schlessinger J, Hubbard SR and Mohammadi M. (1999). Cell, 98, 641 ± 650. Porter AC and Vaillancourt RR. (1998). Oncogene, 16, 1343 ± 1352. Robertson SC, Tynan JA and Donoghue DJ. (2000). Trends Genet., 16, 265 ± 271. Schneider R and Schweiger M. (1991). Oncogene, 6, 1807 ± 1811. Siegel PM and Muller WJ. (1996). Proc. Natl. Acad. Sci. USA, 93, 8878 ± 8883. Smith DK and Xue H. (1997). J. Mol. Biol., 274, 530 ± 545. Stauber DJ, DiGabriele AD and Hendrickson WA. (2000). Proc. Natl. Acad. Sci. USA, 97, 49 ± 54.

Oncogene

Constitutive TrkA activation JC Arevalo et al

1234

Oncogene

Urfer R, Tsoulofas P, O'Conell L, Shelton DL, Parada LF and Presta LG. (1995). EMBO J., 14, 2795 ± 2805. Van Daalen Wetters T, Hawkins SA, Roussel MF and Sherr CJ. (1992). EMBO J., 11, 551 ± 557. Wiesmann C, Ultsch MH, Bass SH and de Vos AM. (1999). Nature, 401, 184 ± 188. Wilkie AOM, Slaney SF, Oldrigde M, Poole MD, Ashworth GJ, Hockley AD, Hayward RD, David DJ, Pulleyn L and Rutland P. (1995). Nature Genet., 9, 165 ± 172.

Windish JM, Marksteiner R, Lang ME, Auer B and Schneider R. (1995a). Biochemistry, 34, 11256 ± 11263. Windish JM, Marksteiner R and Schneider R. (1995b). J. Cell Biol., 270, 28133 ± 28138.