ALK Activation Induces Shc and FRS2 Recruitment: Signaling and Phenotypic Outcomes in PC12 Cells Differentiation

FEBS Letters 581 (2007) 727–734

ALK activation induces Shc and FRS2 recruitment: Signaling and phenotypic outcomes in PC12 cells diﬀerentiation

Joﬀrey Degoutin, Marc Vigny*, Jean Y. Gouzi1 INSERM, U706/UPMC, Institut du Fer a` Moulin, 17 rue du Fer a` Moulin, 4 Place Jussieu, F-75005 Paris, France

Received 22 November 2006; revised 27 December 2006; accepted 16 January 2007

Available online 25 January 2007

Edited by Jesus Avila

activation of the ERK MAP-kinase cascade. However, the Abstract Activation of the neuronal receptor tyrosine kinase ALK (anaplastic lymphoma kinase) promoted the neuron-like nature of the primary events initiated by ALK activation differentiation of PC12 cells through specific activation of the and leading to MAP-kinase activation/PC12 cells differentia- ERK MAP-kinase pathway. However, the nature of primary sig- tion remains to be clarified and can be a major clue to further naling events initiated is still poorly documented. Here, we estab- understand the mechanisms underlying the neuronal differenti- lished that Shc and FRS2 adaptors were recruited and ation processes. phosphorylated following antibody-based ALK activation. We Generally, following activation of RTKs, one of the primary further demonstrated that Shc was recruited to the consensus events leading to MAP-kinase activation is the direct recruit- 1507 phosphotyrosine site NPTpY and FRS2 was likely recruited ment of a variety of intracellular modular proteins organized to a novel non-orthodox phosphotyrosine site within ALK. Final- in multimolecular signaling complexes [13]. The 80 kDa mem- ly, we characterized a functional role for Shc and likely FRS2 in brane-anchored adaptor FRS2 (Fibroblast Receptor Sub- ALK-dependant MAP-kinase activation and neuronal differentiation of PC12 cells. These findings hence open attractive per- strate-2) [14] has been described as an essential component spectives concerning specific characteristics of ALK in the of the signaling complexes associated with the FGF (FGFR1) control of the mechanisms driving neuronal differentiation. [15,16], NGF (TrkA) [17], GDNF (RET) [18,19] and insulin 2007 Federation of European Biochemical Societies. Published [20] receptors. Activation of FGFR1 or TrkA induces the acti- by Elsevier B.V. All rights reserved. vation of FRS2, which specifically participates to sustained activation of MAP-kinase pathway and to neuronal differenti- Keywords: Anaplastic lymphoma kinase; Shc; FGF receptor ation of PC12 cells [15,17,21–23]. Following autophosphoryla- substrate 2; Mitogen-activated protein kinase; PC12 cells; tion of TrkA [17] or RET [18,19] FRS2 classically binds to Neurite outgrowth specific phosphorylated tyrosine residues located within conserved consensus sequences NPXpY or IENKLpY, respectively. Surprisingly, FRS2 is constitutively bound to a specific sequence in the juxtamembrane region of the FGFR1 that does not include any tyrosine residue: KSIPLRRQVTVS 1. Introduction [16]. Interestingly, FRS2 was found to directly bind TrkA at the same phosphotyrosine residue (within the NPXpY site) Anaplastic lymphoma kinase (ALK) is a 220 kDa receptor that binds another adaptor protein, Shc [17]. Shc (also named tyrosine kinase (RTK) recently identified in human, mouse ShcA) is expressed as three isoforms of about 46, 52 and and Drosophila [1–3]. ALK possesses a classical structure 66 kDa. This competitive binding to TrkA suggested that shared with other RTKs, a large extracellular domain, a single Shc and FRS2 could regulate a switch between proliferation transmembrane domain and an intracellular domain contain- and differentiation of PC12 cells. Indeed, according to the ing the tyrosine kinase catalytic site and 18 tyrosine residues well-established model [24], Shc/Grb2/Ras pathway only trig- forming potential sites of phosphorylation. The nature of its gers transient activation of MAP-kinase pathway leading to cognate ligands in vertebrate is still a matter of debate [4–9]. cell proliferation, whereas FRS2 signaling participates to sus- The role of this receptor also still remains to be clearly identi- tained activation of MAP-kinases leading to neuronal differen- fied, however spatio-temporal expression studies [1,2,10] tiation. strongly suggested that ALK could play an important role Finally, FRS2 and Shc adaptors exhibited a similar expres- in the development and function of the nervous system. In sion pattern compared to ALK in several tissues during agreement with this hypothesis, many studies demonstrated development including the nervous system [25,26]. Therefore, that ALK activation promoted neuronal differentiation of these data altogether led us to address in this report the ques- PC12 or neuroblastoma cell lines [8,9,11,12]. Analysis of the tion of the potential involvement of FRS2 and Shc in rela- signaling pathways pointed to an essential role of sustained tion to ERK MAP-kinase signaling and downstream neuronal differentiation of PC12 cells induced by ALK activation. In order to control the activation of ALK in the absence of clearly established ligand(s), we took advantage of * Corresponding author. Fax: +33 1 45 87 61 32. monoclonal antibodies (mAbs) directed against the extracel- E-mail address: [email protected] (M. Vigny). lular domain of ALK which we recently generated and char- 1Present address: Institute of Molecular Biology and Genetics, BSRC acterized [9] and which presented all the characteristics of ‘Alexander Fleming’, Vari, Greece. activating antibodies.

0014-5793/$32.00 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2007.01.039 728 J. Degoutin et al. / FEBS Letters 581 (2007) 727–734

2. Materials and methods aldhehyde/30 mM sucrose and washed 3 times with PBS. Then cells were permeabilized in 0.5% PBS-Triton X-100 for 5 min and washed 2.1. Plasmid constructions with 0.1 M PBS-glycine for 15 min. After 1 h of blocking in PBS con- The full-length human ALK cDNA as well as a construct coding for taining 1.5% BSA, cells were incubated in the same buffer with mono- a kinase-defective form in which the invariant lysine residue located in clonal rat anti-HA (3F10) antibody (0.9 lg/ml) to visualize cells the ATP-binding portion of the catalytic domain was changed to argi- expressing ALK-H derived proteins. The cells were then washed 5 nine (this point mutation has been previously shown to completely in- times with PBS before and after incubation with anti-rat IgG Alexa hibit the activation/phosphorylation of ALK [9,11,12]) was engineered 488-conjugated secondary antibody (Molecular Probes, Leiden, The as previously described [11]. Mutation of tyrosine residues (Y) to phen- Netherlands). Then cells were mounted in Mowiol 4-88 supplemented ylalanine residues (F) was generated with the QuikChange site-directed with Hoechst 33258 nucleic acid stain (0.5lg/ml, Molecular Probes) to mutagenesis kit (Stratagene Europe, Amsterdam, The Netherlands) visualize nuclei. Conventional fluorescence microscopy was performed within the four potential PTB binding sites present in the intracellular on a Leica microscope (Roche Applied Science) equipped with a domain of ALK: NPNY1096, NPTY1507, NVNY1584GY1586. Numbers MicroMax CCD camera (Princeton Instruments, Roper Scientific, indicate the position of the Y residues in the ALK full length receptor. Trenton, NJ). Images were assembled using Adobe Photoshop soft- ALK mutated forms coded by these constructs were respectively ware. The Metamorph software was used (Roper Scientific) for quan- named ALK-H Y1, ALK-H Y2 and ALK-H Y3-Y4. Constructs cod- tification; 100 transfected cells were counted, and cells bearing neurites ing for the mutated forms ALK-H Y1-Y3-Y4 and ALK-H Y1-Y2-Y3- longer than twice the diameter of the cell body were scored as differen- Y4 were obtained by subcloning using appropriate restriction enzymes. tiated. The experiments were performed in triplicate. Mutagenesis was verified by sequencing (Genset, Paris, France). Con- struct coding for the ALK receptor deleted of the 33 juxtamembrane amino acids 1057–1089, named ALK-HDjuxt.mb, has been previously 3. Results described [28]. All generated constructs allowing the expression ALK-derived proteins were linked to an HA tag at their C-terminal end. 3.1. Activation of ALK tyrosine kinase specifically induces The vector encoding FRS2 was kindly provided by J.K. Heath (CR- the recruitment and tyrosine phosphorylation of FRS2 UK Growth Factor Group, School of Biosciences, University of Bir- and Shc adaptors mingham, UK) [27]. To address the question of the potential association of the adaptor proteins FRS2 and Shc with ALK, we first performed 2.2. Cell culture and transfection HEK 293 cells obtained from ATCC (Manassas, VA) were main- co-immunoprecipitation experiments in HEK 293 cells stably tained in high glucose Dulbecco’s modified Eagle’s medium supple- expressing the wild type human ALK receptor (ALK-H) or mented with 10% foetal calf serum (Invitrogen) at 37 Cin its related kinase-defective form (dALK-H) [9] and further atmosphere containing 5% CO2. Stably transfected cells [9] were cul- transfected with the FRS2 vector. Cells were treated or not tured continuously in the same medium complemented by 2 lg/ml puromycin (Sigma). Parental HEK 293 cells were transiently transfec- during 15 min with 6 nM (1 lg/ml) of ALK-activating mouse ted using calcium phosphate precipitation method as previously de- monoclonal antibody (mAb) 46 we previously developed [9]. scribed [9]. Cells plated at a density of 3 · 104 cells/cm2 were cultured Immunoblotting using rat anti-HA tag revealed that equal for 2 days, serum-deprived for 12–16 h, and then treated or not with amounts of ALK-H and dALK-H proteins were immunopre- mAbs we previously described [9] directly added in the medium. cipitated (Fig. 1A). According to our previous report, the weak Rat pheochromocytoma (PC12) cells (kindly provided by T.Galli, Institut Jacques Monod, Paris, France) were grown in a complete med- level of basal phosphorylation of both species of ALK-H (220- ium (RPMI 1640, Invitrogen) supplemented with 10% horse serum kDa and 140-kDa) was strongly increased upon mAb 46 treat- (HS) and 5% foetal calf serum (FCS) at 37 C in an atmosphere con- ment. By contrast, no phosphorylation was observed in cells taining 7.5% CO2. Cells were cultured in flasks or plastic dishes coated expressing the kinase-defective mutant (dALK-H). Therefore, with collagen (1 lg/cm2). PC12 cells were electroporated as previously described [9] and immediately transferred to fresh complete medium these data confirmed that both basal and induced phosphory- for 1 day. Then, cells were serum starved in a low-serum medium lation were a result of ALK-H intrinsic kinase activity, thus (1% HS without FCS) for 16 h before the addition of the mAbs. validating the system of mAb-induced activation of the receptor. 2.3. Immunoprecipitation and immunoblotting analysis More interestingly, immunoblotting with anti-FRS2 and Cell extracts were prepared by lysing the cells in a RIPA buffer anti-Shc antibodies (Fig. 1A) revealed that both the transfec- (10 mM NaPi buffer, pH 7.8, 60 mM NaCl, 1% Triton X-100, 0.5% ted FRS2 (80-kDa) and endogenous Shc (52-, and 46-kDa spe- deoxycholic acid, 0.1% SDS, 10% glycerol, 25 mM b-glycerol phosphate, 50 mM sodium fluoride, 2 mM sodium pyrophosphate, 1 mM sodium cies) adaptors were associated with ALK-H. Upon mAb 46 orthovanadate and protease inhibitor mixture ‘‘complete’’, Roche) treatment, the adaptors are indeed more recruited and this is and analysed by direct immunoblotting as previously described [9] or strongly dependant to the phosphorylation state of the recep- subjected to immunoprecipitation (100 lg of total proteins) using the tor. In addition, co-immunoprecipitates of cells expressing 1.5 lg of mouse anti-HA tag antibody (12CA5). Bound proteins were visualized using the Odyssey Imaging System (LI-COR bioscience). dALK-H did not display any anti-FRS2 or anti-Shc immu- Antibodies and their sources were: mouse (12CA5) and rat (3F10) no-reactivity, thus revealing that ALK intrinsic kinase activity monoclonal anti-HA (Roche Diagnostics, Meylan, France), mouse was required for FRS2 and Shc association with ALK. Then, anti-PY (4G10) and rabbit anti-ERK (Upstate Biotechnology, Lake Pla- in order to determine whether these two adaptors could be cid, NY), rabbit anti-PY (AB1599, Chemicon International, Temecula, CA, USA), rabbit anti-FRS2 (H-91, Santa Cruz Biotechnology), rabbit activated in parallel to their association with ALK, immuno- anti-Shc (Transduction laboratories), mouse anti-P-ERK (Sigma, Saint- blotting experiment using anti-phosphotyrosine antibodies Quentin Fallavier, France). Rabbit anti-insulin receptor phosphospecific was performed on whole-cell lysates displaying the same antibody (pY/pY1162/1163, Biomol, Plymouth Meeting, PA) was used amounts of FRS2, Shc, ALK-H or dALK-H proteins for detection of tyrosine phosphorylated ALK due to the cross-reacting (Fig. 1B). In the absence of mAb 46 treatment, FRS2 and properties of this antibody as previously discussed [9]. Shc adaptors exhibited basal levels of tyrosine phosphorylation that were significantly increased in response to mAb 46 2.4. Neurite outgrowth assay stimulation. These phosphorylations were due to ALK kinase Electroporated PC12 cells were treated at day 2 with the mAbs. At day 4, cells were fixed and processed for immunofluorescence. Cells activity, as neither FRS2 nor Shc phosphorylation was ob- were fixed for 15 min at room temperature with pre-warmed 2% form- served in cell extracts expressing the kinase defective form J. Degoutin et al. / FEBS Letters 581 (2007) 727–734 729

ALK-H dALK-H ALK-H dALK-H

mAb 46 (6nM) : 0' 15' 0' 15' kDa mAb 46 (6nM) : 0' 15' 0' 15' kDa

220 220 ALK ALK 140 140

220 220 P-ALK P-ALK 140 140

FRS2 80 FRS2 80

52 P-FRS2 80 Shc 46

IP : HA Shc 52 46 52 P-Shc 46

44 P-Erk 42

Erk 44 42

Fig. 1. mAb46-induced activation of ALK triggers FRS2 and Shc association with the receptor and their tyrosine phosphorylation. HEK 293 cells stably transfected with the indicated ALK-H and dALK-H constructs were serum-starved for 16 h and then treated or not with activating mAb 46 (6 nM) during 15 minutes at 37 C. (A) Cells were lysed in RIPA buffer as described in Section 2 and 100 lg of proteins was then immunoprecipitated using 1.5 lg of mouse anti-HA tag antibody (12CA5). Western blot analysis was performed using the rat anti-HA tag antibody (3F10), the rabbit anti-insulin receptor phosphospecific antibody (see Section 2), anti-FRS2 antibody (H-91) or the anti-Shc antibody. (B) Cells were lysed in a RIPA buffer and whole-cell lysates (10 lg) were submitted to Western blot analysis using the rat anti-HA tag antibody (3F10) (upper panel), the anti-FRS2 antibody (H-91) or the anti-Shc antibody (median panels). The blot was then reprobed with the rabbit anti-insulin receptor phosphospecific antibody or phosphotyrosine (4G10 or rabbit) antibodies. Western blot analysis was also performed using the mouse anti-phosphoERK1/2 antibody and then reprobed with the rabbit anti-panERK antibody (lower panel). dALK-H. We finally assayed for ERK 1/2 phosphorylation HDjuxt.mb.). Second, sequence analysis revealed that ALK (Fig. 1B) since sustained activation of MAP-kinase pathway contains four putative phosphotyrosine-binding domain had been previously shown to be triggered by ALK activation (PTB) binding sites NXXpY. Since PTB domains of the [8,9,11,12]. Immunoblotting using anti-phospho-ERK 1/2 anti- modular adaptors Shc and FRS2 could classically bind to spe- bodies confirmed that mAb 46 treatment increased the cific phosphorylated tyrosine residues located within highly phosphorylation of both the ERK 1 and 2 proteins, in a conserved consensus sequences NPXpY of different RTKs long-lasting manner since it remained elevated for at least [17–19], we created various point mutants of ALK-H by site- 1 h (data not shown). In addition, as no phosphorylation directed mutagenesis, in which the different tyrosines was detected in cells expressing dALK-H, this confirmed that (Y1096F: Y1; Y1507:Y2; Y1584:Y3; Y1586:Y4) were replaced both basal and induced ERK phosphorylations were depen- by phenylalanine residues. Then, parental HEK 293 cells were dent on ALK kinase activity. transiently co-transfected with each of the various mutant constructs and with FRS2. 3.2. Tyrosine 1507 of ALK is required for Shc association As shown in Fig. 2B, immunoblotting using anti-HA tag re- with the receptor whereas any conventional PTB binding vealed that comparable amounts of WT ALK-H and its re- site of ALK is required for FRS2 association lated mutant proteins were immunoprecipitated. By contrast To determine which amino-acid motif of ALK could be in- to our previous results obtained with HEK cells stably express- volved in Shc or FRS2 adaptors association to the receptor, ing these proteins (see Fig. 1A and B), ALK-H and all the several mutants plasmid constructs derived from that encoding other kinase active forms exhibited a strong basal level of the wild-type receptor ALK-H were engineered (represented in phosphorylation that was not increased upon 15 min of mAb Fig. 2A). First, since the juxtamembrane region of the FGFR1 46 treatment. This basal phosphorylation has already been de- has been previously described as the binding site of FRS2 [16], scribed and discussed in previous studies [9,12] as a conse- we created a mutant allowing the expression of the ALK-H quence of protein overexpression in HEK cells transiently protein deleted of its 33 juxtamembrane amino-acids (ALK- transfected in comparison with cells stably expressing this 730 J. Degoutin et al. / FEBS Letters 581 (2007) 727–734

A ECD TM HA PTK ALK-HΔjuxt.mb.

juxtamembrane 1090 domain: VYRRKHQELQAMQMELQSPEYKLSKLRTSTIMT

1096 1507 Y1NPNY F NPTY F Y2

PTK ALK-H Y1 Y2 Y3 Y4

NVNY1584F GY1586F Y3 Y4

Δjuxt.mb. Y2 Y1 Y3 Y4 Y1 Y2 Y3 Y4

ALK-H dALK-H ALK-H ALK-H ALK-H ALK-H

mAb 46 (6nM) : 0' 15' 0' 15' 0' 15' 0' 15' 0' 15' 0' 15' kDa 220 ALK 140

220 P-ALK 140

FRS2 80

52 Shc 46

IP : HA

P-FRS2 80

44 P-Erk 42

44 Erk 42

Whole Extract

Fig. 2. Tyrosine 1507 in the NPTY PTB binding site of ALK is required for Shc association with the receptor whereas no classical PTB binding site of ALK is required for FRS2 association. (A) Schematic representation of the ALK-H mutant proteins: ALK-HDjuxt.mb. (ALK-H deleted of its 33 juxtamembrane amino-acids) and ALK-H Y1Y2Y3Y4 (point mutations: Y1096F Y1507F Y1584F Y1586F). ECD, extracellular domain; TM, transmembrane domain; PTK, protein tyrosine kinase domain; HA, hemagglutinin tag. (B) Western blot analysis performed on cells transfected with the ALK-H derived proteins after normalization of their levels of expression. HEK 293 cells were transiently transfected as described in Section 2 with the indicated ALK-H or ALK-H-derived mutant constructs and FRS2. Cells were serum-starved for 16 h and then treated or not with mAb 46 (6 nM) during 15 minutes at 37 C. Cells were lysed in a RIPA buffer and the lysates (10 lg) and 100 lg of proteins were then immunoprecipitated using the 1.5 lg of mouse anti-HA tag antibody (12CA5). Western blot analysis was performed using the rat anti-HA tag (3F10), the rabbit anti- insulin receptor phosphospecific, the anti-FRS2 (H-91) or the anti-Shc antibodies (upper panels). Whole-cell extracts were probed with the anti- phosphoERK1/2 antibodies and then reprobed using anti-phosphotyrosine (4G10) and anti-panERK antibodies (lower panels). receptor at a lower level. Note that the ALK-HDjuxta.mb. mu- induced phosphorylation were a result of ALK intrinsic kinase tant was indeed less phosphorylated than the different mu- activity. Furthermore, mAb 46 treatment actually increased tants, certainly relied either on a change of the structure of ERK1/2 basal phosphorylation in all cell extracts and not in its intracellular domain or on a impairment of the mobility those expressing the dALK-H form (Fig. 2B). of the protein in the membrane. In addition, no ALK phos- As described in Fig. 1A, Western blot analysis performed on phorylation was observed in cells expressing the kinase-defec- ALK-H and dALK-H co-immunoprecipitates (Fig. 2B) con- tive mutant (dALK-H), thus confirming that both basal and firmed that both the FRS2 and Shc adaptors were associated J. Degoutin et al. / FEBS Letters 581 (2007) 727–734 731 with ALK-H and that ALK intrinsic kinase activity was re- of antibody treatment. These data confirmed that ALK activa- quired. However, as a consequence of the high level of basal tion was required to induce PC12 cells differentiation phosphorylation of ALK, no discernible increase of FRS2 [8,9,11,12]. More interestingly, in the presence of mAb 46 the and Shc recruitment occurred upon mAb 46 treatment. More percentage of cells expressing the ALK-H Y2 and ALK-H surprisingly, FRS2 was also co-immunoprecipitated with all Y1Y2Y3Y4 mutants that extend neurites was significantly re- the kinase active forms of ALK, but not by the kinase defective duced (43%) compared to those expressing the WT ALK-H form dALK-H. In addition, Western blot analysis using anti- protein (63%). This weaker differentiation level strongly paral- phosphotyrosine antibody performed on whole-cell lysates leled the weaker level of MAP-kinases activation previously ob- expressing comparable amounts of FRS2 and ALK-H mutants served in these cells (Fig. 3A). Altogether these data revealed revealed that FRS2 was tyrosine phosphorylated in each case. that Y1057 in ALK was essential for the recruitment and acti- These data revealed that neither the juxtamembrane domain of vation of Shc adaptor, leading to MAP-kinase pathway activa- ALK-H nor the four tyrosine residues present in the NXXpY tion and subsequent neuronal differentiation of PC12 cells. motifs were required for FRS2 association with ALK. By contrast, Shc adaptor was not co-immunoprecipitated with the ALK-H Y2 and ALK-H Y1Y2Y3Y4 mutants. This data dem- 4. Discussion onstrated that mutation of Y1507 abrogated Shc association with ALK-H, revealing that this residue was specifically The purpose of the present study was to provide further ele- responsible for Shc association to the receptor. Despite of its ments to the understanding of the nature of the primary signal- low phosphorylation state, the ALK-HDjuxta.mb. mutant ing events initiated by activation of the receptor ALK. Indeed, could still efficiently associate with FRS2 and Shc adaptors, the role played by various modular adaptor proteins in ALK and triggered ERK activation upon mAb treatment. signaling has been poorly documented so far. Only a few studies previously described that ALK activation resulted in the association of insulin receptor substrate-1 (IRS-1), Shc, 3.3. MAP-kinase pathway activation and neuronal c-Cbl, PLCc and the p85 subunit of PI3-kinase, subsequently differentiation of PC12 cells partially require Shc leading to tyrosine phosphorylation of IRS-1 and Shc [8]. association with ALK Here, we first corroborated the information that ALK activa- To further investigate the functional outcomes of the associ- tion triggered Shc association and its tyrosine phosphoryla- ation of Shc and FRS2 with ALK, we tested their involvement tion. Furthermore, we characterized for the first time the in MAP-kinase activation and subsequent neuron-like differen- precise phosphotyrosine residue (Y1507) of ALK that was crit- tiation of PC12 cells, two major effects previously assigned to ical for Shc association. Indeed, Y1507F (Y2) mutation has ALK [9,11,12]. Whole-cell lysates of transiently transfected been shown to totally abrogate Shc recruitment. This data is PC12 cells expressing approximately identical amounts of in agreement with previous results obtained with the oncogenic ALK-H derived proteins were subjected to Western blot anal- form of the receptor, NPM-ALK [29]. The Y2 mutation failed ysis and first assayed for ERK1/2 activation. Fig. 3A illus- to abrogate association of FRS2. These findings implicitly trated that ERK1/2 phosphorylation required ALK kinase demonstrated that Shc and FRS2 were not competing for asso- activity, thus confirming in PC12 cells our first results obtained ciation with ALK as such a mechanism had been previously with HEK cells (see Fig. 1B). Indeed, mAb 46 treatment in- described with the NGF receptor (TrkA) [17]. This also implies duced the activation in cells expressing the ALK-H protein that different sites of ALK could be required for respective Shc and not in those expressing the kinase defective dALK-H. In and FRS2 association, thus disclosing a novel specific feature cells expressing the ALK-H Y2 and ALK-H Y1Y2Y3Y4 mu- of the receptor ALK. tants which previously were shown to display no association Our studies also revealed for the first time that FRS2 associ- with Shc, mAb 46 stimulation also induced ERK1/2 phosphor- ation with ALK was dependent on the kinase activity of the ylation, but to a significantly weaker level compared to cells receptor. Therefore, this suggested that FRS2 was recruited expressing the WT ALK-H. This implied that Shc association in a phosphotyrosine-dependent manner, as it had been previ- with ALK is required for a maximal ERK1/2 activation. ously shown with the receptors TrkA or RET [17–19] rather Therefore, Y1507 in ALK was essential for Shc association being constitutively bound to the receptor as it had been pre- with the receptor and partially required for subsequent viously shown with the FGFR1 receptor [16]. Indeed, yeast MAP-kinase phosphorylation in PC12 cells. two-hybrid protein–protein interaction assays revealed that Then, in order to investigate the role of Shc adaptor in ALK- FRS2 was constitutively bound to a juxtamembrane region dependent PC12 cells differentiation, we treated transfected of FGFR1, which do not contain any asparagine or tyrosine cells with 6 nM mAb 46 for two days and assayed for neurite residues. With the aim to determine which alternative binding outgrowth. In the absence or presence of mAb 46, cells trans- site was required for FRS2 association with ALK, we engi- fected with the dALK-H construct or the empty vector alone neered several mutants of the receptor. In agreement with showed essentially a round shape and undifferentiated our hypothesis, we revealed that FRS2 association with phenotype (data not shown). By contrast, cells expressing the ALK did not require the presence of its juxtamembrane region. ALK-H protein (Fig. 3B) exhibited a significant level of basal Since our results suggested that the mechanism of FRS2 bind- differentiation (19%), displaying short and thin neurites. How- ing to ALK could be identical to that of TrkA or RET, we ever, when treated with mAb 46, the percentage of these cells engineered other mutants of ALK and further showed that extending neurites was greatly increased (63%) and they FRS2 association did not require any classical consensus displayed a much larger and spread cell body. Moreover, they PTB-binding sequence of ALK, such as its four NXXpY mo- harboured thick neurite extensions (reaching several folds the tifs. As any other reported PTB-binding motif [30] has been de- cell body size), which could be visible as early as the 12th hour scribed in ALK amino-acid sequence, our results altogether 732 J. Degoutin et al. / FEBS Letters 581 (2007) 727–734

Δjuxt.mb. Y2 Y1 Y3 Y4 Y1 Y2 Y3 Y4

ALK-H dALK-H ALK-H ALK-H ALK-H ALK-H

mAb 46 (6 nM) : 0' 15' 0' 15' 0' 15' 0' 15' 0' 15' 0' 15' kDa 220 ALK 140

44 P-Erk 42

44 Erk 42

B ALK-H ALK-H + mAb46

ALK-H Y2 ALK-H Y2 + mAb46

80 70 60 50

cells extending 40 30 neurites (%) 20 10 Transfected Transfected 0 mAb46 (6nM) : - + -+ - + -++ -+-

ALK-H dALK-H ALK-HΔjuxt.mb. ALK-H Y2 ALK-H Y1 Y3 Y4 ALK-H Y1 Y2 Y3 Y4

Fig. 3. Shc association with activated ALK is involved in ERK 1/2 phosphorylation and PC12 cells differentiation. (A) PC12 cells were electroporated with normalized quantities of the indicated ALK-H or ALK-H-derived mutant constructs, cultured during 16 h in a 1% horse serum medium at 37 C and treated or not with the mAb 46 (6 nM) for 15 minutes. Cells were then lysed in a RIPA buffer and the lysates (10 lg) were submitted to Western blot analysis using the anti-HA tag (3F10) (upper panel), anti-phosphoERK1/2 or the anti-panERK antibodies (lower panels). (B) Quantification of the effect of induced activation of the different ALK-H or ALK-H-derived mutant proteins on neurite outgrowth. Transfected PC12 cells were treated or not with the mAb 46 (6 nM) for 2 days, then fixed, permeabilized and immunofluorescence assay was performed. Cells expressing the ALK-H derived proteins were labelled with the anti-HA-tag antibody (3F10) and cell nuclei were coloured with the nucleic acid stain Hoechst 33258. Bar = 100 lm. Then, PC12 cells were submitted to neurite outgrowth assay as indicated in Section 2. The experiment was performed in triplicate and values are expressed as the means ± S.E.M. (%). strongly suggested that ALK activation could lead to the lined the critical role played by sustained activation of the recruitment of FRS2 to a novel non-orthodox PTB-binding ERK MAP-kinase pathway in this process [31]. PC12 cells site not related to those comprised within TrkA, RET, FGFR1 can either differentiate from chromaffin-like cells to sympa- or any other protein. thetic neuron-like cells when treated with various neurotrophic Many studies focusing on the mechanisms of action trig- factors such as NGF or FGF, or proliferate upon epidermal gered by RTKs and leading to PC12 cells differentiation under- growth factor (EGF) stimulation [32,33]. According to the J. Degoutin et al. / FEBS Letters 581 (2007) 727–734 733 well-established model [31], sustained activation of the MAP- (ARC) and Association Française contre les Myopathies (AFM). kinase pathway induced by NGF or FGF was essential to pro- J.Y.G is a recipient of Research Fellowship from AFM and J.D. was ` ´ mote PC12 cell differentiation whereas transient activation of supported by the French Ministere de l’Enseignement Superieur et de la Recherche. We thank H. Burgar and J.K. Heath (CR-UK this pathway triggered proliferation of these cells. Our studies Growth Factor Group, School of Biosciences, University of Birming- revealed a functional role for Shc in ALK signaling, displaying ham, UK) for providing FRS2 vectors. We are particularly grateful that association with ALK was partially required for ERK to N. Brunet-de Carvalho for her continual support and helpful activation and PC12 cells differentiation. We showed that discussions. abrogation of ALK/Shc association in cells expressing the mutants resulted in a complete lack of Shc recruitment. Shc would be in turn unable to concomitantly recruit Grb2/Sos References proteins, resulting in the inhibition of Ras-dependent activa- [1] Iwahara, T. et al. (1997) Molecular characterization of ALK, a tion of ERK pathway we described here. However, it is note- receptor tyrosine kinase expressed specifically in the nervous worthy that inhibition of MAP-kinases activation and PC12 system. Oncogene 14, 439–449. cells differentiation triggered by ALK mutants was not com- [2] Morris, S.W., Naeve, C., Mathew, P., James, P.L., Kirstein, M.N., plete. This partial inhibition could be explained by the role Cui, X. and Witte, D.P. (1997) ALK, the chromosome 2 gene locus played by Shc regarding to the time-course of ERK activation. altered by the t(2;5) in non-Hodgkin’s lymphoma, encodes a novel neural receptor tyrosine kinase that is highly related to leukocyte Thus, one can propose that the remaining level of PC12 cells tyrosine kinase (LTK). Oncogene 14, 2175–2188. differentiation might be a consequence of signals induced by [3] Loren, C.E., Scully, A., Grabbe, C., Edeen, P.T., Thomas, J., the recruitment of alternative proteins to ALK, which could McKeown, M., Hunter, T. and Palmer, R.H. (2001) Identification overcome the absence of Shc signaling and inhibition of the and characterization of DAlk: a novel Drosophila melanogaster RTK which drives ERK activation in vivo. Genes Cells 6, 531– transient phase of MAP-kinases activation. As mentioned 544. above, previous studies showed that IRS-1, PI3-kinase and [4] Stoica, G.E. et al. (2001) Identification of anaplastic lymphoma PLC-c were associated with ALK following its mAb-induced kinase as a receptor for the growth factor pleiotrophin. J. Biol. activation [8]. However, our previous studies in the same sys- Chem. 276, 16772–16779. tem based on the use of pharmacological inhibitors of PI3-ki- [5] Stoica, G.E., Kuo, A., Powers, C., Bowden, E.T., Sale, E.B., Riegel, A.T. and Wellstein, A. (2002) Midkine binds to anaplastic nase and PLC-c demonstrated that both proteins were not lymphoma kinase (ALK) and acts as a growth factor for different involved in ALK-dependent promotion of PC12 cells differen- cell types. J. Biol. Chem. 277, 35990–35998. tiation [11,12]. Moreover, IRS-1 has been described as being [6] Dirks, W.G., Fahnrich, S., Lis, Y., Becker, E., MacLeod, R.A. involved in the transient phase of MAP-kinases activation and Drexler, H.G. (2002) Expression and functional analysis of the anaplastic lymphoma kinase (ALK) gene in tumor cell lines. and subsequently PC12 cells proliferation [34]. Therefore, Int. J. Cancer 100, 49–56. IRS-1 could act similarly to Shc, likely having only moderate [7] Miyake, I., Hakomori, Y., Shinohara, A., Gamou, T., Saito, M., effects on the sustained phase of MAP-kinases activation and Iwamatsu, A. and Sakai, R. (2002) Activation of anaplastic subsequent PC12 cells differentiation. However, FRS2 has lymphoma kinase is responsible for hyperphosphorylation of been reported as the preferential adaptor protein required ShcC in neuroblastoma cell lines. Oncogene 21, 5823–5834. [8] Motegi, A., Fujimoto, J., Kotani, M., Sakuraba, H. and for prolonged activation of this ERK pathway and PC12 cells Yamamoto, T. (2004) ALK receptor tyrosine kinase promotes differentiation [15,17,22,23]. Therefore, it seems reasonable to cell growth and neurite outgrowth. J. Cell Sci. 117, 3319–3329. assume that this novel ALK-associated adaptor can be a good [9] Moog-Lutz, C., Degoutin, J., Gouzi, J.Y., Frobert, Y., Carvalho, candidate for relaying the effects triggered by the receptor. N.B.-d., Bureau, J., Creminon, C. and Vigny, M. (2005) Activa- tion and Inhibition of anaplastic lymphoma kinase receptor However, since FRS2 binding site in ALK still remains a tyrosine kinase by monoclonal antibodies and absence of agonist conundrum, its identification and further mutation studies will activity of pleiotrophin. J. Biol. Chem. 280, 26039–26048. be certainly informative on the specific role of FRS2 in ALK [10] Vernersson, E., Khoo, N.K., Henriksson, M.L., Roos, G., signaling. Previous studies led in PC12 cells showed that a mu- Palmer, R.H. and Hallberg, B. (2006) Characterization of the tant of Y490 of TrkA (the unique Shc and FRS2 binding site) expression of the ALK receptor tyrosine kinase in mice. Gene Expr. Patterns. could neither activate the Ras signaling pathway nor promote [11] Souttou, B., Carvalho, N.B., Raulais, D. and Vigny, M. (2001) the neuronal differentiation of PC12 cells [35]. Thus, we could Activation of anaplastic lymphoma kinase receptor tyrosine expect that mutation of the different binding sites of both Shc kinase induces neuronal differentiation through the mitogen- and FRS2 in ALK will result in the complete lack of ERK 1/2 activated protein kinase pathway. J. Biol. Chem. 276, 9526–9531. [12] Gouzi, J.Y., Moog-Lutz, C., Vigny, M. and Brunet-de Carvalho, activation and PC12 cells differentiation. N. (2005) Role of the subcellular localization of ALK tyrosine In conclusion, our studies provided novel elements essential kinase domain in neuronal differentiation of PC12 cells. J. Cell for the understanding of the primary signaling events induced Sci. 118, 5811–5823. by ALK activation, through the characterization of Shc and [13] Pawson, T. and Scott, J.D. (1997) Signaling through scaffold, FRS2 adaptors in ALK-dependant signaling leading to the dif- anchoring, and adaptor proteins. Science 278, 2075–2080. [14] Rabin, S.J., Cleghon, V. and Kaplan, D.R. (1993) SNT, a ferentiation of PC12 cells. More particularly, we enlightened differentiation-specific target of neurotrophic factor-induced tyro- some specific features of the receptor ALK regarding to the sine kinase activity in neurons and PC12 cells. Mol. Cell. Biol. 13, mechanisms of adaptors recruitment. Finally, our results open 2203–2213. particularly attractive perspectives concerning specific charac- [15] Kouhara, H., Hadari, Y.R., Spivak-Kroizman, T., Schilling, J., Bar-Sagi, D., Lax, I. and Schlessinger, J. (1997) A lipid-anchored teristics of the receptor ALK in the control of general mecha- Grb2-binding protein that links FGF-receptor activation to the nisms driving neuronal differentiation. Ras/MAPK signaling pathway. Cell 89, 693–702. [16] Xu, H., Lee, K.W. and Goldfarb, M. (1998) Novel recognition Acknowledgments: This work was supported in part by institutional motif on fibroblast growth factor receptor mediates direct funding from INSERM and Universite´ Paris 6 (UPMC), as well as association and activation of SNT adapter proteins. J. Biol. by grants from the Association pour la Recherche sur le Cancer Chem. 273, 17987–17990. 734 J. Degoutin et al. / FEBS Letters 581 (2007) 727–734

[17] Meakin, S.O., MacDonald, J.I., Gryz, E.A., Kubu, C.J. and [27] Burgar, H.R., Burns, H.D., Elsden, J.L., Lalioti, M.D. and Verdi, J.M. (1999) The signaling adapter FRS-2 competes with Heath, J.K. (2002) Association of the signaling adaptor FRS2 Shc for binding to the nerve growth factor receptor TrkA. A with fibroblast growth factor receptor 1 (Fgfr1) is mediated by model for discriminating proliferation and differentiation. J. Biol. alternative splicing of the juxtamembrane domain. J. Biol. Chem. Chem. 274, 9861–9870. 277, 4018–4023. [18] Kurokawa, K., Iwashita, T., Murakami, H., Hayashi, H., Kawai, [28] Mourali, J. et al. (2006) Anaplastic lymphoma kinase is a K. and Takahashi, M. (2001) Identification of SNT/FRS2 dependence receptor whose proapoptotic functions are activated docking site on RET receptor tyrosine kinase and its role for by caspase cleavage. Mol. Cell. Biol. 26, 6209–6222. signal transduction. Oncogene 20, 1929–1938. [29] Fujimoto, J., Shiota, M., Iwahara, T., Seki, N., Satoh, H., Mori, [19] Melillo, R.M., Santoro, M., Ong, S.H., Billaud, M., Fusco, A., S. and Yamamoto, T. (1996) Characterization of the transforming Hadari, Y.R., Schlessinger, J. and Lax, I. (2001) Docking protein activity of p80, a hyperphosphorylated protein in a Ki-1 FRS2 links the protein tyrosine kinase RET and its oncogenic lymphoma cell line with chromosomal translocation t(2;5). Proc. forms with the mitogen-activated protein kinase signaling cas- Natl. Acad. Sci. USA 93, 4181–4186. cade. Mol. Cell. Biol. 21, 4177–4187. [30] Uhlik, M.T., Temple, B., Bencharit, S., Kimple, A.J., Siderovski, [20] Delahaye, L., Rocchi, S. and Van Obberghen, E. (2000) Potential D.P. and Johnson, G.L. (2005) Structural and evolutionary involvement of FRS2 in insulin signaling. Endocrinology 141, division of phosphotyrosine binding (PTB) domains. J. Mol. Biol. 621–628. 345, 1–20. [21] Zeng, G. and Meakin, S.O. (2002) Overexpression of the signaling [31] Marshall, C.J. (1995) Specificity of receptor tyrosine kinase adapter FRS2 reconstitutes the cell cycle deficit of a nerve growth signaling: transient versus sustained extracellular signal-regulated factor non-responsive TrkA receptor mutant. J. Neurochem. 81, kinase activation. Cell 80, 179–185. 820–831. [32] Greene, L.A. and Tischler, A.S. (1976) Establishment of a [22] Xu, H. and Goldfarb, M. (2001) Multiple effector domains within noradrenergic clonal line of rat adrenal pheochromocytoma cells SNT1 coordinate ERK activation and neuronal differentiation of which respond to nerve growth factor. Proc. Natl. Acad. Sci. USA PC12 cells. J. Biol. Chem. 276, 13049–13056. 73, 2424–2428. [23] Ong, S.H., Guy, G.R., Hadari, Y.R., Laks, S., Gotoh, N., [33] Huff, K., End, D. and Guroff, G. (1981) Nerve growth Schlessinger, J. and Lax, I. (2000) FRS2 proteins recruit intra- factor-induced alteration in the response of PC12 pheochromo- cellular signaling pathways by binding to diverse targets on cytoma cells to epidermal growth factor. J. Cell Biol. 88, 189– fibroblast growth factor and nerve growth factor receptors. Mol. 198. Cell. Biol. 20, 979–989. [34] Dikic, I., Schlessinger, J. and Lax, I. (1994) PC12 cells over- [24] Stork, P.J.S. (2005) Directing NGF’s actions: it’s a Rap. 7, 338– expressing the insulin receptor undergo insulin-dependent neuro- 339. nal differentiation. Curr. Biol. 4, 702–708. [25] Hadari, Y.R., Gotoh, N., Kouhara, H., Lax, I. and Schlessinger, [35] Obermeier, A., Bradshaw, R.A., Seedorf, K., Choidas, A., J. (2001) Critical role for the docking-protein FRS2 alpha in FGF Schlessinger, J. and Ullrich, A. (1994) Neuronal differentiation receptor-mediated signal transduction pathways. Proc. Natl. signals are controlled by nerve growth factor receptor/Trk binding Acad. Sci. USA 98, 8578–8583. sites for SHC and PLC gamma. EMBO J. 13, 1585–1590. [26] Lai, K.M. and Pawson, T. (2000) The ShcA phosphotyrosine docking protein sensitizes cardiovascular signaling in the mouse embryo. Genes Dev. 14, 1132–1145.