A ligand-inducible anaplastic lymphoma kinase ...

Oncogene (2004) 23, 1098–1108

& 2004 Nature Publishing Group All rights reserved 0950-9232/04 $25.00 www.nature.com/onc

A ligand-inducible anaplastic lymphoma kinase chimera is endocytosis impaired Michela Serresi1, Gina Piccinini1, Elisa Pierpaoli1 and Francesca Fazioli*,1 1

Laboratory of Cellular and Molecular Biology, Institute of Clinical Medicine, University of Ancona, Via Tronto 10/A, 60020 Ancona, Italy

Ligand-induced membrane trafficking of the anaplastic lymphoma kinase (ALK) was studied using a chimeric receptor in which the extracellular and transmembrane domain of ALK was substituted for the corresponding regions of epidermal growth factor receptor (EGFR). Wild-type EGFR, EGFR/ALK and an EGFR/ALK kinase negative mutant were independently expressed in mouse NR6 fibroblasts. The capacity of EGFR/ALK to mediate [125I]-EGF internalization, receptor degradation and downregulation, which has never been previously described, was assayed. The rate of [125I]-EGF-induced internalization mediated by the cytoplasmic domain of ALK was reduced several fold compared with the wildtype EGFR. The low rate of EGF internalization promoted by EGFR/ALK correlated with an impaired degradation and downregulation of the receptor and indicate that ALK is not subject to traditional mechanisms used to regulate receptor tyrosine kinase function. Accordingly, ALK-activated intracellular domain does not associate in vivo with c-cbl and does not undergo ligand-mediated ubiquitination. The current study provides new insight into the function and regulation of ALK suggesting that the relative long membrane residence of activated ALK might confers a more potent and prolonged signaling activity. Indeed NR6-EGFR/ALK cells exhibited a B3-fold increase in a maximal mitogenic response than NR6-EGFR. Oncogene (2004) 23, 1098–1108. doi:10.1038/sj.onc.1207227 Published online 22 December 2003 Keywords: internalization; EGFR-ALK downregulation; receptor degradation

chimera;

Introduction Growth factors bind to and activate the kinase domain of cell surface receptor tyrosine kinases (RTKs), which propagate the ligand-encoded signal across the plasma membrane and translate it into a cellular response. The complete understanding of the cell physiology mediated by the activation of cell surface receptors relies not only *Correspondence: F Fazioli; E-mail: [email protected] Received 29 April 2003; revised 16 September 2003; accepted 22 September 2003

in the identification of downstream effectors but also in the regulation of receptors membrane trafficking, which controls the location of specific protein–protein interactions and the duration of the signal (Kholodenko, 2002). Indeed, both initiation and termination of signal transmission depend on regulated movement of multiple components through vesicle compartments. Accumulating evidences have shown that ligand-induced endocytosis plays a dual role in receptor signaling: apart from its function as a biophysical mechanisms for attenuating the signal, endocytosis of RTKs is also implicated in the selective distribution of the activated receptors in the proper intracellular location to allow the interaction with specific downstream effector molecules (Di Fiore and De Camilli, 2001; Wiley and Burke, 2001). The mechanism of ligand-induced receptor internalization and downregulation has been extensively described for the epidermal growth factor receptor (EGFR) and subsequently confirmed in a wide variety of RTKs (Sorkin and Waters, 1993). However, it has been shown that this phenomenon is not a general property of RTKs and indeed substantial differences in ligand-induced receptor trafficking exist also between member of the same receptor family (Baulida et al., 1996; Pinkas-Karmarski et al., 1996). It is generally accepted that differences and/or alterations in receptors intracellular trafficking contribute to the potency of mitogenic and transforming activity and therefore are an important component of altered signal transduction in cancer (Di Fiore and Gill, 1999; Ceresa and Schmid, 2000; Waterman and Yarden, 2001). This idea is based on the correlation between low rates of EGFR internalization and cell transformation (Huang et al., 1997) and is supported by the observation that v-cbl transform cells at least in part by shunting EGFR back to the cell surface (Levkowitz et al., 1998). The ALK proto-oncogene encodes a receptor-type tyrosine kinase, which belongs to the insulin receptor superfamily (Iwahara et al., 1997; Morris et al., 1997). The ligand for ALK has been recently identified as pleiotrophin (Stoica et al., 2001). The oncogenic activity of ALK was initially associated with the chromosomal translocation t(2;5)(p23;q35) found in a high proportion of anaplastic large cells lymphomas (ALCLs). This translocation results in the expression of an oncoprotein in which the N-terminal region of the nuclear protein nucleophosmin (NPM) is fused to the intracellular domain of ALK (Morris et al., 1994). Eight variant

Internalization of EGFR/ALK chimera M Serresi et al

1099

ALK fusion proteins have been recently characterized (Hernandez et al., 1999; Lamant et al., 1999; Ma et al., 2000; Touriol et al., 2000; Trinei et al., 2000; Tort et al., 2001). In each case, the intracellular domain of ALK, which encompasses the tyrosine kinase activity, is fused to the N-terminus of an activating gene that determines an unscheduled expression and constitutive activation of ALK. Some ALK oncoproteins have been detected only in ALCLs, some others have been found only in nonhematopoietic neoplasms such as inflammatory myofibroblastic tumors, and some are present in both types of malignancy (Morris, 2001). In addition, a growing body of evidences indicates that ALK is expressed in a significant portion of human cancer cell lines, suggesting a potentially wider role for ALK products in the genesis of neoplastic processes. Indeed, recent studies indicate a direct casual link between overexpression of wild-type ALK and cellular transformation (Weber et al., 2000; Stoica et al., 2001; Piccinini et al., 2002). In agreement with that, Miyake et al. (2002) reported gene amplification of ALK in different neuroblastoma cells. A number of cytoplasmic signaling proteins have been shown to be physically associated with activated ALK (Fujimoto et al., 1996; Bishof et al., 1997; Stoica et al., 2001) although only a few have been shown to be strictly

specific (Bai et al., 1998, 2000; Souttou et al., 2001; Slupianek et al., 2001; Piccinini et al., 2002). Despite studies concerning ALK-associated downstream effectors are in expansion, the endocytic trafficking of this RTK is at the moment completely undefined. Using a chimeric receptor containing the EGFR extracellular domain and ALK cytoplasmic domain (EGFR/ALK), we have functionally characterized the ability of ALK to undergo ligand-induced internalization, downregulation and degradation. Results Expression of EGFR and EGFR/ALK chimera in mouse fibroblasts Expression vectors for wild-type EGFR and the EGFR/ ALK chimera have been previously described (Piccinini et al., 2002). In addition, we engineered an EGFR/ALK kinase negative mutant (EGFR/ALK-KN) in which the nucleotide-binding lysine residue at position 1150 (numbered as specified in Morris et al., 1997) has been exchanged for an arginine residue (see Materials and methods). Mouse NR6 fibroblasts cells were independently transfected with each of the described constructs (schematically represented in Figure 1). Since NR6 cells

Figure 1 Expression of EGFR, EGFR/ALK and EGFRALK-KN in NR6 transfectants. Top: Schematic representation of the molecules employed in this study. EGFR sequences are indicated by empty boxes; ALK sequences are indicated by filled boxes. TM, transmembrane domain: dashed box. EGFR TM domain; grey box. ALK TM domain. Lys to Arg substitution at codon 1150 of the ALK sequence is indicated by a white asterisk. Bottom: Analysis of the expression levels of EGFR (a), EGFR/ALK (b) and EGFR/ ALK-KN (c) in NR6 transfectants. Total cellular proteins (100 mg) obtained from different transfectants were analysed by SDS–PAGE and immunoblotted with an anti-EGFR polyclonal antibody to evidence the expression levels of EGFR or with the anti-ALK monoclonal antibody to reveal the expression levels of EGFR/ALK and EGFR/ALK-KN. As an internal control, each blot was reprobed with an anti-vinculin antibody without prior stripping

Oncogene


1100 Table 1 Analysis of the [125I]EGF-binding properties of EGFR and EGFR/ALK chimera in NIH-3T3 and NR6 fibroblastsa High affinity Cell line NIH–EGFR NIH–EGFR/ALK NR6–EGFR NR6–EGFR/ALK

Low affinity

No. of receptor/cell

Kd (nM)

No. of receptor/cell

Kd (nM)

2.3 105 1.8 105 1.1 104 1.28 104

0.09 0.1 0.1 0.08

2.7 106 2.5 106 1.5 105 1.7 105

6.2 6.4 6.3 6.4

a 125 [ ]EGF binding was assessed by Scatchard analysis over a range of concentration from 0.016 to 16 nM in triplicate wells. Specificity of binding was controlled in a parallel competition experiment by using a 100-fold molar excess of unlabeled EGF. Data were analysed with the LIGAND software

lack endogenous EGFR, expression of EGFR, EGFR/ ALK and EGFR/ALK-KN into these cells allow study of EGF-triggered processes that are exclusively dependent on the biological properties of the transfected molecule. Western blot analysis of the stable transfectants revealed the expression of each receptor in NR6 cells (Figure 1a–c). Various transfectants were selected that express either moderate (1–2 105 receptor/cell) or high (2–3 106 receptor/cell) levels of surface receptors. We have previously described the biochemical characteristics and biological activities of the chimeric EGFR/ ALK receptors expressed in NIH 3T3 cells (Piccinini et al., 2002). In particular, we demonstrated that the [125I]EGF-binding affinities and the partition of highversus low-affinity binding sites were not altered in the chimeric receptors compared with the wild-type EGFR (Table 1). Binding of [125I]EGF to the EGFR/ALK-KN revealed comparable affinities values (data not shown). Similar results were observed when these molecules were expressed in NR6 cells (Table 1). Analysis of the tyrosine kinase activity revealed that both EGFR and EGFR/ALK receptors responded to the addition of EGF with increased levels of receptor autophosphorylation and cellular protein phosphorylation (Figure 2) as previously reported in NIH-3T3 cells (Piccinini et al., 2002). The loss of tyrosine kinase activity by the EGFR/ALK-KN mutant was indicated by its failure to react with antibodies to phosphotyrosine in immunoblot analysis (Figure 2).

Internalization of chimeric ALK receptors To compare the rate of endocytosis of bound ligand, cells were briefly exposed to [125I]EGF at 371C for the indicated time periods. The redistribution of [125I]EGF from the acid-extractable (surface-bound) to the acidresistant (internalized) pool was determined as described in Materials and methods. EGF endocytosis is a saturable process and has a maximal rate when low number of receptors are occupied by EGF (Wiley, 1988; Lund et al., 1990; Waters et al., 1992); therefore, internalization experiments were performed with [125I]-EGF concentration ranging from 0.5 to 2 ng/ml so that fewer than 20.000 surface receptors were occupied during the assay. Oncogene

Figure 2 EGF-dependent activation of chimeric receptors. NR6 cells transfected with EGFR (EGFR), the EGFR/ALK chimera (EGFR/ALK) or the kinase-negative EGFR/ALK chimera mutant (EGFR/ALK-KN) were starved overnight in serum-free medium and then treated for 10 min at 371C without () or with EGF ( þ ) (100 ng/ml) before lysis. (a) Total cellular proteins (1 mg) of each lysate were immunoprecipitated with Ab-1 antibody and analysed by immunoblot with an anti-PTyr antibody. (b) An aliquot (100 mg) of the same lysates was directly fractionated by electrophoresis and subjected to immunoblot analysis with the anti-PTyr antibody to detect tyrosine phosphorylated cellular proteins. Molecular mass markers are indicated in kilodaltons

We have initially analysed the internalization of [125I]EGF by EGFR, EGFR/ALK and EGFR/ALKKN expressed in NR6 cells at a relatively low level (about 1–2 105 receptor/cell). As shown in Figure 3a,


1101

Figure 4 [125I]EGF internalization by EGFR/ALK in NIH-3T3 cells. NIH-3T3 cells expressing EGF receptors (black circles) or the EGFR/ALK chimera (open circles) were incubated for 1–10 min at 371C with 1 ng/ml of [125I]EGF. The amount of surface-bound and internalized radioactivity was determined as described in Materials and methods. Data represent the mean value7s.d. of four independent experiments

Figure 3 [125I]EGF internalization by EGFR/ALK and EGFR/ ALK-KN in NR6 cells. Cells expressing EGF receptors (black circles), EGFR/ALK (open circles) or the EGFR/ALK-KN (black triangles) were incubated for 1–10 min at 371C with 1 ng/ml of [125I]EGF. At the indicated times, cells were washed with cold binding medium to remove unbound [125I]EGF and then acidwashed to remove surface-bound [125I]EGF as described in Materials and methods. Radioactivity present in the acid washes was expressed as surface associated [125I]EGF. The remaining cellassociated radioactivity was quantitated following solubilization of the cells. To compare the internalization capacity of the different cell lines, data at each time point are expressed as the ratio of internalized to surface radioactivity. The values are the means7s.d. (n ¼ 3)

[125I]EGF was rapidly internalized by wild-type EGFRs, whereas internalization mediated by the ALK cytoplasmic domain was extremely low. In particular, in NR6EGFR cells the amount of internalized ligand exceeded the amount of the surface-bound [125I]EGF after 4–5 min of incubation at 371C. By contrast, the rate of [125I]EGF internalization mediated by the chimeric EGFR/ALK receptor was several fold lower (Figure 3a). [125I]EGF internalization mediated by the kinase negative EGFR/ ALK mutant was similar compared with that of the EGFR/ALK chimera (Figure 3b). Similar rates of [125I]EGF internalization were measured in NIH3T3 transfectants expressing either comparable (data not shown) or higher numbers of surface receptors (about

2 106 receptor/cell) (Figure 4). Therefore, under our experimental conditions, the rate of EGF internalization was independent on the level of receptor expression. The [125I]EGF internalization rate (Ke) (calculated from experiments in which the mean surface EGF receptor and EGFR/ALK occupancy was less than 20 000) was B0.38 for EGFR, B0.035 for EGFR/ALK and 0.022 for EGFR/ALK-KN. Metabolic degradation, downregulation and recycling of EGFR/ALK To gain insight into the mechanism of the slow [125I]EGF internalization promoted by the EGFR/ALK chimera, we have analysed the rate of receptors degradation using an [35S]-labeled amino acids pulse-chase experiment. Therefore, NR6-EGFR, NR6EGFR/ALK and NR6-EGFR/ALK-KN were metabolically labeled with [35S]TransLabel and then incubated at 371C for different time periods in absence or in presence of 200 ng/ml EGF. Quantitation of radioactivity recovered in receptor immunoprecipitates showed that the half-life of wild-type EGFR was decreased approximately B4-fold in the presence of EGF (Figure 5). In the absence of EGF, the half-lives of EGFR/ALK and EGFR/ALK-KN receptors (data not shown) were similar to that of the EGFR; however, the addition of EGF had no effect on the rate of degradation of EGFR/ALK (Figure 5a and b). The net result of ligand-induced receptor internalization and degradation is a decrease in the number of functional receptors on the cell surface, a phenomenon known as receptor downregulation. To substantiate the [125I]-EGF internalization and degradation rate observed in NR6-EGFR/ALK cells, we measured the ability of Oncogene


1102

Figure 5 EGF-induced degradation of metabolically labeled NR6-EGFR and NR6-EGFR/ALK cells. Cells were labeled overnight with [35S]TransLabel, washed and then incubated without or with 200 ng/ml EGF for various periods of time at 371C. At each time point, cells were lysed and the receptors were immunoprecipitated with the anti-EGFR Ab-1 antibody. The level of radioactivity in each receptor band was quantitated using a Phosphoimager. (a) Autoradiograms of a representative experiment. (b) Relative amount of each receptor recovered from EGF-treated cells expressed as % of that in the corresponding untreated sample. Each point represents the mean7s.d. of three independent experiments

EGF to promote downregulation of the EGFR/ALK chimera. NR6 transfectants were incubated at 371C with a saturating concentration of unlabeled EGF (100 ng/ml) for various periods of time and subsequently washed to remove unbound ligand- and surface-associated EGF. The number of residual binding sites at the cell surface was then measured by incubating the cells at 41C with a saturating concentration of [125I]-EGF. As shown in Figure 6, EGFR was rapidly downregulated by approximately 75%, while downregulation of the EGFR/ALK chimera was approximately 24%. Therefore, the reduced rates of internalization and degradation of EGFR/ALK correlates with an impaired downregulation of the chimera. Downregulation is also influenced by the relative rates of receptor recycling. It is unlikely that the observed differences were the consequence of a more rapid EGF recycling in NR6-EGFR/ALK cells. In fact, it is known that when measurements are made over short periods of time, differences between apparent rates of internalizaOncogene

tion (ratio of internalized and surface ligand) reflect differences between specific internalization rates (Wiley et al., 1991; Sorkin et al., 1992). However, to exclude definitively this possibility, cells were preloaded by incubation with [125I]-EGF (10 ng/ml) at 371C for 10 min. Residual cell surface [125I]-EGF was then removed by a mild acid wash at 41C. Cells were then shifted to 371C and the reappearance of [125I]-EGF was followed. These experiments indicated that similar rates of release of [125I]-EGF into the culture medium in NR6EGFR and NR6-EHGFR/ALK cells (data not shown). Therefore, we conclude that EGFR/ALK was unable to internalize [125I]-EGF efficiently in comparison to EGFR. Analysis of EGFR/ALK and native ALK internalization by immunofluorescence staining Ligand-mediated internalization of the EGFR/ALK chimera was also examined through fluorescence microscopy using the EGFR-specific antibody Ab-1, which


1103

Figure 6 EGF-induced downregulation of the EGFR-ALK chimera. NR6 cells expressing EGFR (black circles) or EGFR/ ALK (open circles) were incubated without or with EGF (100 ng/ ml) for the indicated times at 371C. Thereafter, cells were washed with cold binding medium and surface-bound EGF was removed by cold acid washes followed by two additional rinses with binding medium. The number of EGF-binding sites on the cell surface was then determined by incubating the cells with 100 ng/ml [125I]EGF at 41C for 2 h. Data are expressed as the percentage of [125I]EGFbinding capacity relative to cells not exposed to EGF and represent the mean value7s.d. from three independent experiments

Figure 7 Analysis of EGFR, EGFR/ALK and native ALK internalization by immunofluorescence staining. NR6-EGFR (a– c), NR6-EGFR/ALK (d–f) and NR6-ALK (g–i) were incubated for 15 min at 371C in serum-free medium (a, d and g) or incubated with 100 ng/ml of EGF (b, e and h) or 2 nM pleiotrophin (c, f and i). The cells were fixed and stained with the anti-EGFR antibody Ab-1 (a– c) or the ALK-c antibody (d–i), followed by FITC-conjugated secondary antibody. Magnification, 400

was sufficient to induce rapid ALK tyrosine phosphorylation. reacts with the external domain of EGFR. To this end, NR6-EGFR and NR6-EGFR/ALK were exposed to saturating concentration of EGF (100 ng/ml) for 15 min at 371C and processed for immunostaining. These experiments indicated that both in absence and in presence of EGF, the EGFR/ALK chimera was localized at the plasma membrane and no endosome association was detected: similar results were obtained when the anti-ALK-c monoclonal antibody was used for immunostaining (Figure 7d and e). On the contrary, under the same experimental conditions, EGF induced the appearance of brightly stained vescicles in the cell periphery and perinuclear region in NR6-EGFR cells, indicating the presence of a significant portion of EGFR in endosome (Figure 7a and b). Although the ALK cytoplasmic domain is tyrosine phosphorylated in response to EGF binding and does transduce biological responses such as mitogenesis and transforming activity, it may be argued that receptor trafficking functions are not accurately reproduced in the chimeric molecule. Therefore, in the same experiments we have analysed NR6 cells transfected with the native ALK molecule (NR6-ALK), which is a highaffinity receptor for the recently identified pleiotrophin (Stoica et al., 2001). As shown in Figure 7i, after 15 min incubation with 2 ng/ml pleiotrophin, ALK was localized only in the plasma membrane and showed the same distribution pattern observed in absence of stimulation (Figure 7g). Similar results were obtained when NR6ALK cells were incubated with higher concentrations of pleiotrophin (up to 8 ng/ml) (data not shown), although in NR6-ALK cells stimulation with 2 ng/ml pleiotrophin

Receptors association to c-cbl and subsequent receptors polyubiquitination The results described above indicate that ALK does not undergo ligand-mediated rapid internalization: this could be related to an unefficient interaction of ALK with the endocytic apparatus. Different studies have shown that c-cbl is required for ligand-dependent ubiquitination of EGFR, a modification that target the protein to the lysosome degradative pathway. We have, therefore, determined whether impaired ALK internalization is paralleled by deficient association with c-cbl. Lysates obtained from untreated and EGF-stimulated NR6-EGFR and NR6-EGFR/ALK cells were immunoprecipitated with anti-c-cbl antibodies. Immunoblotting with anti-EGFR (Figure 8a, left panel) demonstrated in vivo association of c-cbl with activated EGFR, as previously reported (Galisteo et al., 1995). Under the same experimental conditions, no c-cbl association of EGFR/ALK was revealed, despite similar amount of c-cbl were present in the immunocomplexes (Figure 8, right panel). Accordingly, no ubiquitin immunoreactivity was evident in EGF-stimulated EGFR/ALK immunocomplexes (Figure 8b). Mitogenic potency of EGFR/ALK The system analysed above offers the opportunity to ask whether a receptor that does not undergo ligandinduced internalization and downregulation may be more mitogenic than a receptor, which is rapidly downregulated, particularly at low growth factor Oncogene


1104

Figure 9 EGF-stimulated DNA synthesis in NR6-EGFR and NR6-EGFR/ALK chimera. Cells expressing comparable levels of receptors (see Table 1) were serum-starved for 72 h and then stimulated for 22 h with either 1% fetal bovine serum or the indicated concentration of EGF, in the presence of 4 mCi of [methyl-3 H]thymidine/well. Data from triplicate wells are expressed as the ratio [(EGF cpmbackground cpm/1% serum cpmbackground cpm)] 100. Background values were obtained in the absence of any mitogen. Results are representative of one of three similar experiments, performed with triplicate wells for each of the indicated EGF concentrations as well as for controls. Black circles, NR6-EGFR; open circles, NR6-EGFR/ALK Figure 8 Receptors association with c-cbl and receptor polyubiquitination. NR6 cells expressing comparable levels of wild-type EGFR (EGFR) or EGFR/ALK chimera (EGFR/ALK) were starved overnight in serum-free medium and then treated for 10 min at 371C without or with EGF (100 ng/ml) before lysis. (a) Total cellular proteins (4 mg) of each lysate were immunoprecipitated with the anti-c-cbl antibody and analysed by immunoblotting either with an anti-EGFR polyclonal antibody (left panel) or with the anti-ALK monoclonal antibody (right panel). As an internal control, each blot was reprobed with the anti-c-cbl antibody without prior stripping. (b) Lysates (1 mg of total cellular proteins) obtained as described in Materials and methods, were immunoprecipitated with the anti-EGFR Ab-1 antibody and analysed by immunoblot with the anti-ubiquitin monoclonal antibody. Molecular mass markers are indicated in kilodaltons

concentrations. Therefore, we have used NR6 cells expressing either EGFR or EGFR/ALK receptors to assay the mitogenicity of increasing concentrations of EGF. For these experiments, we used two markerselected mass populations (NR6-EGFR and NR6EGFR/ALK) displaying a comparable number of EGF-binding sites. Figure 9 shows a typical EGF-dose experiment observed in these cell lines. NR6-EGFR/ ALK cells were more sensitive to EGF, as evidenced by an B3-fold increase in maximal mitogenic response as compared to NR6-EGFR, despite similar levels of EGFbinding sites and affinities. In addition, NR6-EGFR/ ALK cells also were more sensitive to EGF, as demonstrated by higher levels of mitogenic stimulation, than were NR6-EGFR cells, particularly at low EGF doses. Similar results have been previously obtained in NIH-3T3 cells (Piccinini et al., 2002). Based on these data, we conclude that the lack of a rapid ligandOncogene

induced endocytosis of EGFR/ALK might be involved in the enhanced biological activity of ALK.

Discussion Regulated trafficking of the EGFR has been extensively investigated over several decades and probably represents the best-characterized receptor trafficking system. It is well established that EGF binding to its receptor results in downregulation of EGFR: this downregulation is attributed to the rapid internalization of the activated receptors via clathrin-coated pits followed by the efficient sorting of the internalized receptors to the lysosome degradation pathway (Sorkin and Waters, 1993). This mechanism has been demonstrated for other growth factor receptors and for a long time the concept of rapid ligand-mediated endocytosis and downregulation of growth factor receptors has been improperly considered a general rule. Indeed, new and interesting concepts in our understanding of receptors trafficking have been provided in the last years. First, it is now evident that RTKs might present completely different endocytic properties and therefore follow distinct fate after ligand stimulation (Sorkin et al., 1993; Baulida et al., 1996; Pandi et al., 1997; Waterman et al., 1998). Furthermore, it is becoming clear that endocytosis and signal transduction are two processes tightly connected, regulating one other in both a positive and negative manner (Ceresa and Schmid, 2000; Di Fiore and De Camilli, 2001; Wiley and Burke, 2001). Consequently, analysis of receptors trafficking is an important


1105

component to define the biological and biochemical properties of RTKs since differences and/or alterations in their intracellular trafficking contribute to the potency of mitogenic and transforming activity. Our previous study has demonstrated that the EGFR/ ALK chimera respond to EGF and is significantly more active in inducing transformation and DNA synthesis than the wild-type EGFR when either are expressed at similar levels in NIH 3T3 cells. Comparative analysis of the biochemical pathways implicated in the transduction of mitogenic signals showed that EGFR/ALK chimera was much more efficient than EGFR to induce PI3kinase activity, suggesting that the stronger mitogenic potency of ALK is partially related to its more efficient coupling with the PI3-K pathway (Piccinini et al., 2002). The mechanisms controlling membrane trafficking of ALK have never been investigated. To gain insight into the activated ALK endocytic properties, we have comparatively analysed the rate of ligand-induced internalization of EGFR/ALK chimeric receptors and wild-type EGFRs. EGFR/ALK internalized EGF at a rate that was several fold lower than the rate of ligand internalization by wild-type EGFR. Our data also reveal a minimal extent of downregulation of EGFR/ALK. Receptors immunoprecipitation analysis on [35S]-labeled cells demonstrates a relative long membrane residence of the EGFR/ALK chimera, which did not substantially differ in presence of EGF. Therefore, the reduced rate of internalization of EGFR/ALK also resulted in an impaired degradation and downregulation of the receptor. It is unlike that the observed differences are the consequence of a rapid EGFR/ALK recycling, since recovery of intact [125I]EGF in the chase-medium of [125I]EGF-loaded cells were similar in NR6-EGFR and NR6-EGFR/ALK cells. Studies addressing the importance of the tyrosine kinase activity in endocytosis have been performed on different RTKs. It is generally accepted that receptor tyrosine kinase activity is required for cellular trafficking: for that reason, we have also included in our study a kinase-defective EGFR/ALK mutant. The extent of [125I]EGF internalization was slightly lower with the mutationally inactivated kinase. However, considering the slow internalization rate observed for the EGFR/ ALK, employment of such a mutant does not add further information to our present knowledge. Despite a large amount of phenomenological data, the molecular details of the mechanisms controlling ligandinduced endocytosis of RTKs are still confused and not fully understood. It is generally accepted that receptor recruitment into plasma membrane clathrin-coated pits requires ‘endocytic recognition sequences’ that allow specific interaction with coated pit components. Several mutagenesis studies have been used to map these critical amino-acid sequences in the RTKs cytoplasmic tail: identified endocytic signals are represented by the sequences NPXY and YXX+ (where + represents an amino acid with a bulky hydrophobic group and X stands for any amino acid) or by a di-leucine motif (Di Fiore and Gill, 1999). Under this assumption, lack of a rapid, ligand-induced endocytosis of EGFR/ALK

should indicate an inefficient interaction of the ALK cytoplasmic domain with the cellular endocytic apparatus. However, different sequences in the cytoplasmic domain of ALK fulfill the criteria for the described putative internalization codes (NPXY: ALK residues 1093–1096, NPNY and residues 1504–1507, NPTY; YXX+: ALK residues 1401–1404, YGPL plus several di-leucine motifs), suggesting that the presence of these putative endocytic codes are not sufficient to promote ligand-mediated receptor internalization. Indeed, it has been shown that mutations of the FYRAL sequence within the cytoplasmic domain of EGFR, which abrogate its interaction with the adaptor protein AP-2, do not significantly affect EGF-induced receptor internalization (Nesterov et al., 1995; Sorkin et al., 1996). In addition, although the erbB-2 intracellular domain contains putative internalization sequences at approximately the same relative position of the EGFR, an EGFR/erbB-2 chimeric receptor displays very low rate of ligand-induced internalization, downregulation and degradation and does not associate with AP-2 (Sorkin et al., 1993; Baulida et al., 1996). Possibly additional unidentified sequences are involved. Alternatively, a much more complex scenario should be considered, involving not only the presence of correct information for capture in clathrin-coated pits, but also proper signals that allow the receptor to reach them. In line with this point of view, it has been shown that in quiescent fibroblasts a large fraction of EGFRs are concentrated in caveolae but rapidly move out of this membrane domain in response to EGF (Mineo et al., 1999). In addition, it has been demonstrated that all the EGFR mutants that are endocytosis impaired also are unable to leave caveolae in response to ligand stimulation. Based on their observations, the authors of this work propose an endocytosis three-step model which involves (1) exit from caveolae; (2) migration in the bulk plasma membrane; (3) capture by coated pits (Mineo et al., 1999). Under this possibility, an inefficient movement of ALK out of caveolae may be related to its impaired internalization, despite the presence of functionally relevant endocytic codes in its cytoplasmic domain. The current study provide new insight into the function and regulation of ALK. Future research into the function of endocytic accessory factors and signaling proteins will likely reveal additional connections between ALK endocytosis and signaling, and improve our information on the oncogenic property of this receptor.

Materials and methods Eukaryotic expression vectors and site-directed mutagenesis The engineering of the expression vectors LTR-EGFR 5 M and LTR-EGFR/ALK have been previously described (Lonardo et al., 1990; Piccinini et al., 2002). To generate the EGFR/ALK kinase-negative mutant, a 2.2 kbp BamHI– BamHI fragment from the LTR-EGFR/ALK plasmid was cloned into M13mp18 phage RF DNA. Oligonucleotideprimed site-directed mutagenesis was performed on phage Oncogene


1106 ssDNA template according to the method of Kunkel (1985), by mutagenizing the AAG (Lys) codon at position 1150 of the ALK open reading frame (Morris et al., 1997) to AGG (Arg). After modification of the desired sequence, the BamHI– BamHI fragment was cloned back into the vector. The presence of the mutation in the expression vector was confirmed by DNA sequencing in both strands. The LTRALK expression vector was obtained by cloning the entire human ALK coding sequence (ATCC) into the LTR-2 expression vector (Di Fiore et al., 1987).

and surface-bound EGF was removed by incubating cells with ice-cold 0.2 M sodium acetate buffer (pH 4.5) containing 0.5 M NaCl for 2 min and 20 s followed by two rinses with binding medium. This acid wash procedure did not affect the binding properties of receptors. The number of binding sites on the cell surface was then determined by incubating the cells 100 ng/ml [125I]EGF at 41C for 2 h. The amount of surface-bound [125I]EGF corresponding to the number of surface receptors was then determined by the acid wash (pH 2.8) as described for the internalization experiments.

Cell culture and transfection assays

Receptor degradation

NR6 and NIH-3T3 mouse fibroblasts were maintained in Dulbecco’s modified Eagle’s medium (D-MEM) containing 10% fetal bovine serum (FBS) (Invitrogen) and transfected with circular plasmid DNA according to the calcium phosphate precipitation technique. Mass cell populations containing the LTR-based vectors were selected for their ability to form colonies in a medium containing mycophenolic acid, taking advantage of the presence of the selectable marker Ecogpt (Mulligan and Berg, 1981).

To study EGF receptor turnover, cells were metabolically labeled with [35S]TransLabel (ICN Biochemical) overnight at 371C in methionine- and cysteine-free medium containing 1% calf serum and 60 mCi [35S]TransLabel/ml. Cells were then washed three times with D-MEM to remove unincorporated radiolabeled amino acids. The monolayers were then incubated at 371C with or without 200 ng/ml EGF in binding medium for the indicated period before lysis and immunoprecipitation with the anti-EGFR Ab-1 antibody as described below. The amount of radioactivity in the protein bands corresponding to each receptor was quantitated using a Phosphoimager apparatus (Molecular Dynamics).

EGF-binding assays Scatchard analysis were performed as described (Piccinini et al., 2002) The number of receptors per cell and their dissociation constants (Kd) for EGF were determined from Scatchard plots. Analysis of the binding was performed with the LIGAND software. Internalization and receptor recycling assays To monitor [125I]EGF internalization, cells cultured in 24-well dishes were incubated with 0.5–2 ng/ml [125I]EGF in binding medium at 371C for various periods of time (up to 10 min) and then rapidly rinsed three times with binding medium to remove unbound ligand. Cells were then incubated at 41C with an acid-stripping solution containing 0.2 M acetic acid (pH 2.8) and 0.5 M sodium chloride. The acid wash was combined with another short rinse with the same acidic solution to determine the amount of surface-bound [125I]EGF. Finally, the cells were solubilized with 1 N sodium hydroxide to quantitate internalized radioactivity. The ratio of internalized and surface radioactivity was plotted against time and the specific rate constant for internalization (Ke) was calculated as linear regression coefficient. Nonspecific binding was measured for each time point in the presence of 100-fold molar excess of unlabeled EGF (Upstate Biotechnology Inc.). To monitor receptor recycling, cells were rinsed with ice-cold binding buffer and incubated with [125I]EGF by incubation at 41C for 1 h. Cells were then rinsed twice with cold binding medium and allowed to internalize the ligand for 10 min at 371C. Next, cells were rinsed with cold binding medium and [125I]EGF remaining on the cell surface was removed by using a mild acidic solution (0.2 M sodium acetate buffer, pH 4.5 containing 0.5 M NaCl) for 2 and 0.5 min consecutively and then washed twice with binding medium. These [125I]EGF ‘loaded cells’ were further incubated in binding medium with 300 ng/ml unlabeled EGF at 371C for different time periods and the amount of released [125I]EGF in the medium as well as surface and intracellular labeled ligand was determined. Receptor downregulation To monitor receptor downregulation, cells were incubated for various times (0–120 min) with or without EGF (100 ng/ml) in binding medium at 371C, rinsed with cold binding medium, Oncogene

Immunofluorescence Cells were grown on glass coverslips to subconfluence. Following serum starvation for 24 h, cells were either mocktreated or treated with EGF (Upstate Biotechnology Inc.) or pleiotrophin (Sigma) at the indicated concentrations for 15 min at 371C. Cells were fixed for 10 min in 4% paraformaldehyde in PBS and thereafter permeabilized by incubation for 15 min in phosphate-buffered saline (PBS) supplemented with 0.1 and 0.5% BSA. After blocking with 1% BSA in PBS, cells were incubated at room temperature in the presence of specific antibodies (anti-EGFR Ab-1 or anti-ALK-c) followed by donkey anti-mouse IgG conjugated with FITC (1 : 50 in PBS). Nuclear counterstaining was performed by incubating coverslips for 15 min at 251C with 4’6’-diamidino-2-phenylindole (Sigma). Samples were analysed and photographed using an Olympus AX70 fluorescence microscope. Protein analysis Cells treatment and lysis was performed as previously described (Piccinini et al., 2002). For immunoprecipitation procedures, cellular lysates were incubated with the indicated antibody for at least 2 h at 41C under gentle rotation and immunocomplexes were recovered on Protein-G bound to Sepharose beads (Zymed). After several washes with buffer containing 0.1% Triton X-100, 20 mM HEPES, 10% glycerol and 150 mM NaCl immunocomplexes were resuspended in Laemmli buffer and boiled. Lysates or immunocomplexes were separated by SDS–PAGE, transferred onto nitrocellulose filters and immunoblotted according to previously published procedures (Piccinini et al., 2002). Bound proteins were visualized with [125I]labeled anti-mouse-IgG (0.2 mCi/ml) (Amersham Pharmacia Biotech) or by enhanced chemiluminescence (Amersham Pharmacia Biotech). The monoclonal antibody recognizing the intracellular domain of ALK (ALK-c) was kindly provided by Professor B Falini (Institute of Haematology, University of Perugia, Italy) and Professor PG Pelicci (European Institute of Oncology, Milan, Italy) (Falini et al., 1998). Ab-1, the monoclonal antibody directed against the extracellular domain of EGFR, was purchased from Calbiochem. The rabbit anti-


1107 peptide polyclonal sera recognizing the EGFR carboxylterminal tail (residues 985–996) was previously described (Piccinini et al., 2002). Other antibodies used were the antiphosphotyrosine (anti-pTyr) monoclonal antibody (Upstate Biotechnology), the affinity purified polyclonal C-15 anti c-cbl antibodies (Santa Cruz Biotechnology) and the anti-vinculin monoclonal antibody (clone hVIN-1) (Sigma). Receptor polyubiquitination was analysed according to Kassenbrock et al. (2002) using the monoconal anti-ubiquitin antibody MMS-258 (BABCO). Briefly, cells were starved of growth factors for 18 hours and then stimulated for 10 minutes at 371C with 100 ng/ml EGF. N-ethylmaleimide (5 mM) was added to the lysing buffer. Immunocomplexes, obtained by incubating total cellular proteins with Ab-1 were resolved on a 7.5% SDS–PAGE and electrotransfered to an Immobilon-P membrane (Millipore). Mitogenic assay The [3H]-incorporation assay was performed as described previously (Piccinini et al., 2002). Data are expressed as a mitogenic index calculated as the fraction of stimulation obtained in the presence of EGF with respect to the

stimulation obtained in the presence of nonsaturating concentrations of an optimal mitogen (1% FBS). The mitogenic index was calculated as follows: [(EGF cpmbackground cpm)/(1% serum cpmbackground cpm)] 100.

Abbreviations ALCL, anaplastic large cell lymphoma; ALK, anaplastic lymphoma kinase; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; PTyr, phosphotyrosine; RTK, receptor tyrosine kinase. Acknowledgements We are grateful to Professor A Rappelli for his continues support and encouragement during the course of this work. We also thank Professor PP Pelicci and Professor B Falini for providing us with the anti-alk monoclonal antibody. The authors appreciate the excellent assistance of Dr R Bacchiocchi, D Carbonari and M Cesaroni. This research was fully supported by grant from the Associazione Italiana Ricerca sul Cancro. MS is supported by a Fellowship from Fondazione Italiana Ricerca Cancro.

References Bai RY, Dieter P, Pescel C, Morris SW and Duyster J. (1998). Mol. Cell. Biol., 18, 6951–6961. Bai RY, Ouyang T, Miething C, Morris SW, Peschel C and Duyster J. (2000). Blood, 96, 4319–4327. Baulida J, Kraus MH, Alimandi M, Di Fiore PP and Carpenter G. (1996). J. Biol. Chem., 271, 5251–5257. Bishof D, Puldorf K, Mason DY and Morris SW. (1997). Mol. Cell. Biol., 17, 2312–2325. Ceresa BP and Schmid SL. (2000). Curr. Opin. Cell Biol., 12, 204–210. Di Fiore PP and De Camilli P. (2001). Cell, 106, 1–4. Di Fiore PP and Gill GN. (1999). Curr. Opin. Cell Biol., 11, 483–488. Di Fiore PP, Pierce JH, Kraus MH, Segatto O, King CR and Aaronson SA. (1987). Science, 237, 178–182. Falini B, Bigerna B, Fizzotti M, Pulford K, Pileri SA, Delsol G, Carbone A, Paulli M, Magrini U, Menestrina F, Giardini R, Pilotti S, Mezzelani A, Ugolini B, Billi M, Pucciarini A, Pacini R, Pelicci PG and Flenghi L. (1998). Am. J. Pathol., 153, 875–886. Fujimoto J, Shiota M, Iwahara T, Seki N, Satoh H, Mori S and Yamamoto T. (1996). Proc. Natl. Acad. Sci. USA, 93, 4118–4186. Galisteo ML, Dikic L, Batzer AG, Langdon WY and Schlessinger J. (1995). J. Biol. Chem., 270, 20242–20245. Hernandez L, Pinyol M, Hernandez S, Bea` S, Puldorf K, Rosenwald A, Lamant L, Falini B, Ott G, Mason DY, Delsol G and Campo E. (1999). Blood, 94, 3265–3268. Huang HS, Nagane M, Klingbeil CK, Lin H, Nishikawa R, Ji, XD, Huang CM, Gill GN, Wiley HS and Cavenee WK. (1997). J. Biol. Chem., 272, 2927–2935. Iwahara T, Fujimoto J, Wen D, Cupples R, Bucay N, Arakawa T, Mori S, Ratzkin B and Yamamoto T. (1997). Oncogene., 14, 439–449. Kassenbrock CK, Hunter S, Garl P, Johnson GL and Anderson SM. (2002). J. Biol. Chem., 277, 24967–24975. Kholodenko BN. (2002). Trends Cell Biol., 12, 173–177. Kunkel TA. (1985). Proc. Natl. Acad. Sci. USA, 82, 488–492. Lamant L, Dastugue N, Puldorf K, Delsol G and Mariame B. (1999). Blood, 93, 3088–3095.

Levkowitz G, Waterman H, Zamir E, Kam Z, Oved S, Langdon WY, Bequinot L, Geiger B and Yarden Y. (1998). Genes Dev., 12, 3663–3674. Lonardo F, Di Marco E, King CR, Pierce J, Segatto O, Aaronson SA and Di Fiore PP. (1990). New Biol., 2, 992– 1003. Lund KA, Opresko LK, Starbuck C, Walsh BJ and Wiley HS. (1990). J. Biol. Chem., 265, 15713–15723. Ma Z, Cools J, Marynen P, Cui X, Sieberrt R, Gesk S, Schlegelberger B, Peeters B, De Wolfe-Peeters C, Wlodarska I and Morris SW. (2000). Blood, 95, 2144–2149. Mineo C, Gill GN and Anderson RG and W. (1999). J. Biol. Chem., 274, 30636–30643. Miyake I, Hakomori Y, Shinohara A, Gamou T, Saito M, Iamatsu A and Sakai R. (2002). Oncogene, 21, 5823–5834. Morris SW. (2001). Br. J. Haematol., 113, 275–295. Morris SW, Kirsten MN, Valentine MB, Dittmer KG, Shapiro DN, Saltman DL and Look AT. (1994). Science, 263, 1281– 1284. Morris SW, Naeve C, Mathew P, James PL, Kirstein MN, Cui X and Witte DP. (1997). Oncogene, 14, 2175–2188. Mulligan RC and Berg P. (1981). Proc. Natl. Acad. Sci. USA, 78, 2072–2076. Nesterov A, Wiley HS and Gill G and N. (1995). Proc. Natl. Acad. Sci. USA, 92, 8719–8723. Pandi SD, O’Hare T, Donis-Keller H and Pike LJ. (1997). J. Biol. Chem., 272, 2199–2206. Piccinini G, Bacchiocchi R, Serresi M, Vivani C, Rossetti S, Gennaretti C, Carbonari D and Fazioli F. (2002). J. Biol. Chem., 277, 9526–9531. Pinkas-Karmarski R, Soussan L, Waterman H, Levkowitz G, Alroy I, Klapper L, Lavi S, Seger R, Ratzkin BJ, Sela M and Yarden Y et al. (1996). EMBO J., 15, 2452–2467. Slupianek A, Nieborowska-Skorska M, Hoser G, Morrione A, Majewski M, Xue L, Morris SW, Wasik MA and Skorski T. (2001). Cancer Res., 61, 2194–2199. Sorkin A, Mazzotti M, Sorkina T, Scotto L and Bequinot L. (1996). J. Biol. Chem., 271, 13377–13384. Sorkin A and Waters CM. (1993). BioEssays, 15, 375–382. Sorkin A, Di Fiore PP and Carpenter G. (1993). Oncogene, 8, 3021–3028. Oncogene


1108 Sorkin A, Helin K, Waters CM, Carpenter G and Bequinot L. (1992). J. Biol. Chem., 267, 8672–8678. Souttou B, Brunet-De Carvalho N, Raulais D and Vigny M. (2001). J. Biol. Chem., 276, 9526–9531. Stoica GE, Kuo A, Aigner A, Sunitha I, Souttou B, Malerczyk C, Caughey DJ, Wen D, Karavanov A, Riegel AT and Wellstein A. (2001). J. Biol. Chem., 276, 16772–16779. Tort F, Pynol M, Puldorf K, Roncador G, Hernandez L, Nayach I, Kluin-Nelemans HC, Kluin F, Turiol C, Gelsol G, Mason D and Campo E. (2001). Lab. Invest., 81, 419– 426. Touriol C, Greenland C, Lamant L, Puldorf K, Bernard F, Roussel T, Mason DY and Delsol G. (2000). Blood, 95, 3204–3207.

Oncogene

Trinei M, Lanfrancone L, Campo E, Puldorf K, Mason DY, Pelicci PP and Falini B. (2000). Cancer Res., 60, 793–798. Waterman H, Sabanai I, Geiger B and Yarden Y. (1998). J. Biol. Chem., 273, 13819–13827. Waterman H and Yarden Y. (2001). FEBS Lett., 490, 142–152. Waters CM, Overholser Ka, Sorkin A and Carpenter G. (1992). J. Cell. Physiol., 152, 253–263. Weber D, Klomp HJ, Czubayko F, Wellstein A and Juhl H. (2000). Cancer Res., 60, 5284–5288. Wiley HS. (1988). J. Cell Biol., 107, 801–810. Wiley HS, Herbst JJ, Walsh BJ, Lauffenburger DA, Rosenfeld MG and Gill GN. (1991). J. Biol. Chem., 266, 11083–11094. Wiley HS and Burke PM. (2001). Traffic, 2, 12–18.