p130Cas mediates the transforming properties of the anaplastic lymphoma kinase

Nucleophosmin–anaplastic lymphoma kinase (NPM-ALK) is an oncogenic protein generated by the t(2;5)(p23;q35) translocation¹  that characterizes most anaplastic large-cell lymphomas (ALCLs). Each of the 2 fusion-derived fractions that form NPM-ALK plays a role in the constitutive tyrosine kinase activity of the chimeric protein. NPM is a nucleolar phosphoprotein involved in the ribosome genesis,²  in the centrosome duplication,³  and in the regulation of the basal level of p53.⁴  NPM promoter drives the constitutive expression of NPM-ALK in lymphoid cells; the oligomerization domain of NPM allows the dimerization⁵  and hence the autophosphorylation of NPM-ALK. ALK is a 210-kDa tyrosine kinase (TK) receptor belonging to the insulin growth factor receptor superfamily, which is widely expressed in the nervous system during embryogenesis but only focally in the adult brain.^1,6,7  Although the physiologic role of ALK receptor is still unknown, it seems to be involved in neuronal cell differentiation as suggested by the ability to induce neurite outgrowth in vitro⁸  and by its fundamental role in synapse formation shown in Caenorhabditis elegans.⁹ 

The constitutive NPM-ALK activity transforms rat-1 fibroblast¹⁰  and is responsible for the malignant transformation of lymphoid cells.¹¹  Furthermore, NPM-ALK has the ability to induce B- and T-cell lymphomas in vivo.^12,13  The signaling pathways downstream from NPM-ALK follow a transduction pattern shared with most of deregulated tyrosine kinases, leading to increased proliferation, resistance to apoptosis, and morphologic changes. The enhancement of cell growth in NPM-ALK–positive cells results from the activation of the rat sarcoma viral oncogene homolog (Ras)/extracellular signal-related kinase (ERK) pathway through the recruitment of the adaptors SHC (Src homology 2 domain containing protein) and insulin receptor substrate 1 (IRS-1).¹⁴  In addition, NPM-ALK controls cellular mitogenity by binding to phospholipase C γ (PLC-γ).¹⁵ 

In parallel, the inhibition of apoptosis in the NPM-ALK signal transduction derives from the activation of signal transducer and activator of transcription 3 (Stat3),¹⁶  leading to an increased expression of the antiapoptotic factor Bcl-x(L) and from the stimulation of the phosphatidylinositol 3–kinase (PI3K)/Akt pathway,¹⁷  which confers resistance to apoptosis combined with a deregulated proliferation rate due to the transcription factor forkhead box O3a (FOXO3a).¹⁸ 

In contrast to these fairly well-characterized pathways, minimal data are available regarding the effects of ALK tyrosine kinase activity on cell shape. Upon transduction in Hodgkin lymphoma cell lines, NPM-ALK induces in these cells morphologic changes that make them similar to ALCL cells.¹⁹  Moreover, in PC12 cells, ectopic expression of ALK induces a neurite outgrowth that has been shown to depend on the mitogen-activated protein (MAP) kinase pathway.⁸  In C elegans, the degradation of ALK is required for synapse stabilization,⁹  whereas in Drosophila, ALK is activated by its ligand Jelly Belly to specify visceral mesoderm migration and differentiation into muscle cells.^20,21  Moreover, ALK ectopic expression in Drosophila hemocytes promotes the formation of lamellocytes, specialized cells that participate in the encapsulation and killing of parasites, a phenotype that is coupled with the activation of both the Janus kinase (JAK)/STAT and the Jun kinase pathways through Rho guanosine triphosphatase (GTPase) Rac1.²² 

In this paper we investigated molecules involved in cell morphology that could also be relevant for the NPM-ALK–mediated transformation process. We started from a mass spectrometry–based screening of proteins interacting with NPM-ALK and involved in the cytoskeleton morphology. Through this analysis we identified p130Cas (p130 Crk-associated substrate) and then demonstrated that NPM-ALK is able to bind and phosphorylate p130Cas. The p130Cas activation is mediated by the adaptor protein growth factor receptor–bound protein 2 (Grb2) and is partially independent from Src. Importantly, we show that p130Cas activation is required for NPM-ALK–mediated cell shape modifications and actin filament depolymerization as well as for cell transformation, but not for cell migration, induced by NPM-ALK.

Taken together, these results suggest not only that p130Cas acts as one of the effectors of the cytoskeleton modifications induced by NPM-ALK but also that morphologic and growth cues are deeply embedded in NPM-ALK–mediated oncogenic processes.

Human lymphoid cells TS, DHL, and Karpas (NPM-ALK–positive) and CEM, K562, and Namalwa (NPM-ALK–negative) were obtained from New York University and maintained in RPMI 1640 (BioWhittaker, Verviers, Belgium) containing 10% fetal calf serum, 2 mM glutamine (Eurobio Biotechnology, Les Ulis, France), 100 U/mL penicillin, and 100 μg/mL streptomycin (Eurobio Biotechnology).

Human embryonal kidney cells 293T, 293GP, and 293 T-Rex Tet-On (Invitrogen, Carlsbad, CA) and murine fibroblasts MEF, NIH3T3, Syf^-/-, Cas^-/- (also known as Bcar1^-/-), and MEF Tet-Off (Clontech, Palo Alto, CA) were maintained in Dulbecco modified Eagle medium (DMEM) supplemented with 10% fetal calf serum, 2 mM glutamine (Eurobio Biotechnology), 100 U/mL penicillin, and 100 μg/mL streptomycin (Eurobio Biotechnology). In MEF and 293 Tet systems, the working concentration of tetracycline or doxycycline in the medium was 1 μg/mL.

Inducible ALK-shRNA interference TS cells were obtained by cotransduction with pLVTH vector containing the H1 promoter ALK-shRNA cassette and pLV-tTRKRAB vector,²³  as described by Piva et al.⁵³  These cells undergo NPM-ALK silencing when 1 μg/mL of doxycycline is added to the medium for at least 72 hours.

Total cellular proteins were extracted with 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA (ethylenediaminetetraacetic acid), 0.1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 mM NaF, 1 mM Na₃VO₄, and protease inhibitors (Roche, Mannheim, Germany). Cell lysates were centrifuged at 10 000g (13 000 rpm) for 15 minutes and the supernatants were collected and assayed for protein concentration using the Bio-Rad protein assay method. Thirty micrograms of proteins was run on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) under reducing conditions.

For immunoprecipitation experiments, 300 μg to 800 μg protein was immunoprecipitated with the appropriate antibody for 1 hour at 4°C followed by 30 μL protein A–Sepharose beads (Amersham, Freiburg, Germany). After SDS-PAGE, proteins were transferred to nitrocellulose, incubated with the specific antibody, and then detected with peroxidase-conjugated secondary antibodies and enhanced chemiluminescent reagent (Amersham).

PP2 Src-family kinase inhibitor was purchased from Calbiochem (Darmstadt, Germany). The following antibodies were used: polyclonal anti-ALK (1:2000; Zymed, Basel, Switzerland), monoclonal anti-ALK (1:4000; Zymed), monoclonal anti-p130Cas (1:1000; Becton Dickinson, Mountain View, CA), monoclonal anti–phospho-Tyr (PY20; 1:1000; Transduction Laboratories, Lexington, KY), polyclonal anti-Grb2 (1:2000; Santa Cruz Biotechnology, Santa Cruz, CA), monoclonal anti-Crk (1:1000; Transduction Laboratories), polyclonal anti-Src (1:1000; Santa Cruz Biotechnology), monoclonal antipaxillin (1:1000; Transduction Laboratories), monoclonal antiactin (1:1000; Santa Cruz Biotechnology), and monoclonal anti–focal adhesion kinase (anti-FAK; 1:1000; hybridome immunoglobulin G1 [IgG1] clone 4-4A, kindly gifted by Prof Silengo, University of Turin, Italy). Secondary antimouse or antirabbit peroxidase-conjugated antibodies were purchased from Amersham.

For anti-ALK immunoprecipitation, 293 T-Rex Tet-on cells (Invitrogen) transfected with a wild-type (wt) NPM-ALK and a kinase-dead mutant control NPM-ALK^K210R were grown to confluence and induced with 1 μg/mL of tetracycline for 24 hours. Cells were lysed in 20 mM Tris (pH 7.4), 150 mM NaCl, 5 mM EDTA, and 0.1% Triton X-100 in the presence of protease and phosphatase inhibitors. Ten milligrams of cleared cell lysate was mixed with 100 μg monoclonal anti-ALK coupled to agarose beads (Amersham) and rotated at 4°C for 2 hours. Precipitated immunocomplexes were washed 3 times with lysis buffer, eluted, and boiled in Laemli buffer for 5 minutes. Samples were resolved on SDS-PAGE and silver stained. Bands of interest were excised from the gel and subjected to in-gel reduction, alkylation, and digestion with trypsin (Promega, Madison, WI) and analyzed by matrix-assisted laser desorption ionization/time of flight (MALDI-TOF) mass spectrometry. Spectra were obtained using either a Reflex III MALDI-TOF mass spectrometer or an Ultraflex TOF/TOF mass spectrometer (Bruker Daltonics, Bremen, Germany), operating in the positive ion delayed extraction reflector mode. Ions were generated by irradiation of analyte/matrix deposits by a nitrogen laser at 337 nm and analyzed with an accelerating voltage of 20 kV. Each MALDI-TOF spectrum was generated by accumulating data corresponding to more than 400 laser shots. Mass calibration in the range mass over electronic charge (m/z) 800 to 4000 was performed by using a bovine beta-lactoglobulin tryptic peptide mixture. Postacquisition internal calibration was applied by using theoretic masses of trypsin autodigestion peptides. Peptide mass spectra were analyzed using Mascot Peptide Mass Fingerprinting software.²⁴ 

Some samples were analyzed by liquid chromatography–tandem mass spectrometry (LC-MS/MS) using a nanoflow–high-performance liquid chromatography (HPLC) system (Ultimate; LC Packings, Amsterdam, The Netherlands) interfaced to electrospray quadrupole time-of-flight (Q-TOF) tandem mass spectrometers (QTOF Ultima or QTOF Micro; Waters/Micromass, Manchester, United Kingdom). Protein identification via peptide MS/MS spectra was achieved by using Mascot software²⁴  for searching the National Center for Biotechnology Information (NCBI) nonredundant protein database.²⁵ 

For antiphosphotyrosine immunoprecipitation, 293 T-Rex Tet-On cells were grown in nonadherent conditions on Poly(2-hydroxyethylmethacrylate) (Poly-HEME; Sigma, St Louis, MO) coated plates, starved for 12 hours, and induced with 1 μg/mL of tetracycline for 24 hours, pelleted, and stored at -80°C. Cells were lysed in 20 mM Tris (pH 7.4), 150 mM NaCl, 5 mM EDTA, and 1% Nonidet P-40 in the presence of protease and phosphatase inhibitors. Ten milligrams cleared cell lysate was mixed with 50 μg agarose-conjugated monoclonal 4G10 (Upstate Biotechnology, Lake Placid, NY) and 10 μg monoclonal PY20 (Transduction Laboratories) and incubated rotating at 4°C overnight. Samples were processed as described in “Cell lysis, immunoprecipitation, and immunoblotting.”

Cells were grown for 12 hours on glass coverslips pretreated with fibronectin (10 μg/mL phosphate-buffered saline [PBS]) at 37°C for 1 hour to facilitate cell adhesion. Samples were fixed in PBS containing 4% paraformaldehyde at room temperature for 10 minutes and permeabilized with PBS containing 0.3% Triton X-100 for 5 minutes. Coverslips were incubated with PBS containing 3% bovine serum albumin (BSA) for 1 hour at room temperature and then stained with primary antibody for 1 hour followed by fluorescein isothiocyanate (FITC)– or tetramethyl rhodamine isothiocyanate (TRITC)–conjugated secondary antibody (1:200; Sigma) for 1 hour at room temperature.

Primary polyclonal anti-ALK (Zymed) and monoclonal antipaxillin (Transduction Laboratories) antibodies were diluted 1:400 before use. Phycoerythrin (PE)–conjugated phalloidin (1:200 PBS; Sigma) was used to stain actin filaments. Nuclei were stained 10 minutes at room temperature with HOECHST (300 ng/mL; Sigma). Coverslips were mounted in antifading solution and viewed using a Leica TCS SP2 laser-scanning confocal microscope driven by Leica Confocal Software; the images were acquired at room temperature by means of a 63×/1.32 PL APO objective (Leica, Heidelberg, Germany). Brightfield images were acquired on a Leica DM IRE2 microscope using a DC300F camera and were analyzed with IM 50 software.

Wild-type NPM-ALK or the kinase-dead mutant NPM-ALK^K210R16 was cloned in the plasmid vector pcDNA5TO (Invitrogen) at HindIII/XhoI sites and stably transfected into 293 T-Rex Tet-On cells using Effectene reagents as described by manufacturer (Qiagen, Valencia, CA). Single cell–derived clones were selected for NPM-ALK or NPM-ALK^K210R expression levels. Wild-type NPM-ALK, digested to obtain a HindIII/XhoI-blunted fragment, was cloned in the plasmid vector pBIEGFP (Clontech) at MluI-blunted sites and then stably transfected into MEF Tet-Off. Murine p130Cas was cloned in the plasmid vector pEGFP-C1 (Becton Dickinson) at EcoRI/BamHI sites or digested to obtain an AseI/BamHI-blunted fragment and cloned in the plasmid vector Pallino²⁶  at EcoRI/XhoI-blunted ends. Wild-type Grb2 or the dominant-negatives Grb2-P49L (mutated in the N-terminal SH3 domain) and Grb2-R86K (mutated in the SH2 domain) were cloned in the plasmid vector pRK5 (a kind gift from Dr A. Pellicer, New York University). Pallino vectors containing NPM-ALK or NPM-ALK^K210R or ATIC-ALK (aminoimidazole-4-carboxamide ribonucleotide formyltransferase/inosine 5′-monophosphate [IMP] cyclohydrolase–ALK) were previously described.¹⁶  Full-length human ALK receptor was purchased from ATCC (Manassas, VA).

NPM-ALK and p130Cas retroviruses were obtained by cotransfection of Pallino expression vector containing wild-type NPM-ALK or p130Cas with pMD₂VSV-G plasmid into the 293GP packaging cell line (Invitrogen). Retroviruses released in culture medium were collected 24 hours after transfection. Five hundred microliters filtered (pore size 0.45 mm) supernatants from GP cells was supplemented with 8 μg/mL polybrene (Sigma) and added to 5 × 10⁴ to 10 × 10⁴ target cells. After 12 hours of incubation, 1 mL complete medium was added and the cells were cultured for an additional 2 days. The percentages of transduced cells were analyzed for GFP expression by fluorescence-activated cell sorter (FACS). Cells were analyzed for GFP content on a FACSCalibur flow cytometer (Becton Dickinson). Calibration of the instrument was performed using Calibration beads (CALIBrite; Becton Dickinson). The CELLQuest software (Becton Dickinson) was used for the data acquisition and analysis. For cell sorting, cells were suspended at the concentration of 10 × 10⁶/mL in basic sorting buffer (5 mM EDTA, 25 mM HEPES [N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid; pH 7.0], 1% heat-inactivated fetal bovine serum [FBS]) and then sorted for GFP expression on a MoFlo High-Performance cell sorter (DAKO Cytomation, Glostrup, Denmark).

For clonogenic assay, 5 × 10⁵ cells/well were seeded in 0.35% agar on the top of a base layer containing 0.7% agar into 6-well plates. After 3 to 5 weeks of cultivation, colonies were counted under a phase-contrast microscope. All of the experiments were done using triplicate plates for experimental point.

For DNA content determination, cells were fixed for 1 hour in 70% ethanol at 4°C. After washing, cells were treated with RNase (0.25 mg/mL) and stained with propidium iodide (50 μg/mL). The S-phase fraction was calculated using the Modfit program from Becton Dickinson.

Cell migration was evaluated using 24-well, 5-μm pore size Transwell system plates (Costar, Cambridge, MA). Purified cells were washed once in RPMI 1640 (BioWhittaker) containing 0.1% fetal calf serum, 2 mM glutamine (Eurobio Biotechnology), 100 U/mL penicillin, and 100 μg/mL streptomycin (Eurobio Biotechnology) and then adjusted to 1 × 10⁶ cells/mL in the same medium. Cells (1 × 10⁵ in 100 μL) were placed on the top of the Transwell. Stromal-derived factor-1 α (SDF-1α; R&D Systems, Abingdon, United Kingdom), prepared at the indicated concentrations in the same RPMI medium (600 μL total volume), was added to the bottom of the Transwell system. After 2 hours incubation at 37°C in a 5% CO₂ atmosphere, the inserts of the transwell chambers were removed and the number of cells that had migrated into the lower well was counted by staining with a DIF-Quick Kit (IMEB, San Marcos, CA).

Cells were seeded at a density of 1 × 10⁵ cells onto Matrigel (Becton Dickinson)–coated inserts (100 μg/insert) in 24-well, 8-μm pore size Transwell system plates (Costar). After a 48-hour incubation, the inserts were removed, the cells on the upper side of the inserts were removed with a cotton swab, and the number of cells that migrated to the lower side of the filters was counted under a phase-contrast microscope by staining with DIF-Quick Kit (IMEB).

We selected different cell lines to test the effects of NPM-ALK on cell morphology, adhesion, and migration. We used 2 tetracycline-inducible cell lines: MEF Tet-Off fibroblasts were grown in DMEM medium containing tetracycline (1 μg/mL) and were induced to express NPM-ALK whenever tetracycline was depleted from the medium. The 293 T-Rex cells were induced to express NPM-ALK when 1 μg/mL of tetracycline was added to the medium; NPM-ALK was detectable by Western blot (WB) as early as 6 hours after tetracycline induction. NIH3T3 fibroblasts were retrovirally infected to stably express NPM-ALK. Upon NPM-ALK expression, MEF Tet-Off and NIH3T3 fibroblasts assumed the typical transformed phenotype with spindle-shaped cell morphology (Figure 1A-B), whereas 293 T-Rex cells detached from the culture plate (Figure 1C-D).

In fibroblasts, the transformed morphology corresponded to a disruption of the actin cytoskeleton in immunofluorescence (Figure 1E-F). Together with actin, we stained for the scaffolding protein paxillin as a marker of the cellular structural organization because it is found at the interface between the plasma membrane and the actin cytoskeleton.^27,28  Noninfected NIH3T3 cells were characterized by a strong paxillin signal at the focal adhesions, whereas NIH3T3 NPM-ALK cells showed few residual paxillin clusters at the edge of the elongated cellular processes (Figure 1E-F green arrows). Moreover, the actin cytoskeleton turned from a polarized organization to a scattered pattern. The actin filaments were clearly visible in noninfected NIH3T3 cells, whereas only a few were detectable in NIH3T3 NPM-ALK (Figure 1E-F red arrows).

For cellular migration we used T-lymphoblastoid CD4⁺ CEM cells, which are described to migrate in response to SDF-1α, the ligand of the chemokine receptor CXCR4, predominantly expressed on naive CD4⁺ and CD8⁺ T cells.²⁹  CEM cells were infected with Pallino retroviruses containing NPM-ALK, or the kinase-dead mutant (NPM-ALK^K210R) as control, and then sorted for the GFP expression. In a chemotaxis assay, upon stimulation with the chemokine SDF-1α, the migration rate of CEM NPM-ALK was significantly higher than CEM wt or CEM NPM-ALK^K210R (Figure 1G). Similarly, in inducible ALK-shRNA interference TS cells both the spontaneous and the SDF-1α–induced migration rates were significantly reduced when NPM-ALK expression was ablated (Figure 1H). Taken together, these data indicate a role of NPM-ALK in the migration process of transformed lymphoid cells.

Since few data are available on the signaling pathways that NPM-ALK activates to influence cytoskeleton organization and cell migration, we decided to apply mass spectrometry to identify new interactors or substrates that are tyrosine phosphorylated after NPM-ALK expression. To this end, we performed immunoprecipitations with a monoclonal anti-ALK antibody or with a mixture of antiphosphotyrosine antibodies, respectively. The 293T-Rex cells were induced with tetracycline for 24 hours to express NPM-ALK or NPM-ALK^K210R as control. For antiphosphotyrosine immunoprecipitations, cells were grown on Poly-HEME in order to reduce the basal phosphorylation due to adhesion-mediated signaling. After comparing the patterns of NPM-ALK and NPM-ALK^K210R samples, the bands present only in the NPM-ALK sample were cut from the gel (Figure 2). Proteins were identified by peptide mass mapping by MALDI tandem mass spectrometry and by peptide sequencing by LC-MS/MS. From the list of the identified proteins we focused on proteins known to play a role in cytoskeleton reorganization or cell migration. Table 1 lists the identified proteins related to the cytoskeleton. Most of the proteins found were structural proteins, in keeping with recently published data.³⁰  One of these proteins was the p130Cas that is an adaptor protein involved in many cellular processes related to the cytoskeleton organization, migration, and transformation.³¹ 

We confirmed that NPM-ALK was able to associate with p130Cas in 293 T-Rex cells by immunoprecipitation (Figure 3A). Next, we extended our analysis to other molecules known to be often in complex with p130Cas: paxillin and Src were shown to coprecipitate with NPM-ALK, whereas FAK and Crk were almost undetectable (Figure 3A). Importantly, the binding between NPM-ALK and p130Cas was also demonstrated in NPM-ALK–positive cells derived from human lymphomas (Figure 3B). In lymphoid cell lines, NPM-ALK associates not only with p130Cas (Figure 3B arrow) but also with Cas lymphocyte-type (Cas-L), a p105 kDa member of the Cas family exclusively expressed in lymphocytes^32,33  (Figure 3B arrowhead). In ALK-positive lymphoid cells, both p130Cas and Cas-L are present in a phosphorylated form (Figure 3C).

Moreover, in the presence of NPM-ALK, p130Cas was phosphorylated (Figure 3D). This effect was specifically due to the tyrosine kinase activity of ALK because the mutated kinase-dead NPM-ALK^K210R failed to bind and to phosphorylate p130Cas (Figure 3E). The phosphorylation of p130Cas was detected also expressing the ATIC-ALK fusion protein (Figure 3F-G) as well as the full-length ALK receptor (data not shown), thus suggesting a general mechanism of the ALK kinase domain.

NPM-ALK has recently been reported to associate with Src in Karpas and DHL lymphoid cells.³⁴  Considering that p130Cas is a major substrate of Src,³⁵  we checked the potential role of Src in NPM-ALK–mediated phosphorylation of p130Cas. We confirmed the binding between NPM-ALK and Src both in 293 T-Rex cells (Figure 3A) and in lymphoid cell lines (Figure 4A). We took advantage of Src, Yes1, and Fyn triple knock-out fibroblasts to elucidate a potential role of Src kinases in NPM-ALK–mediated p130Cas phosphorylation. NPM-ALK could induce p130Cas phosphorylation even in the absence of Src, thus demonstrating the presence of a Src-independent pathway (Figure 4B). In addition, no differences in NPM-ALK–dependent p130Cas phosphorylation were found following treatment with the Src family inhibitor PP2 both in MEF NIH3T3 cells and in lymphoid cell lines (Figures 4C-D), thus confirming the results obtained in Src/Yes1/Fyn triple knock-out fibroblasts.

Since p130Cas has no SH2 domains for a direct binding with phosphoproteins,³¹  we looked for an adaptor protein that could mediate the association between NPM-ALK and p130Cas. Grb2 has previously been shown to bind both NPM-ALK³⁶  and p130Cas³⁷  and we confirmed the association between NPM-ALK and Grb2 in ALK-positive lymphoid cell lines (Figure 5A). Moreover, the R86K-mutated Grb2 has been reported to disrupt the association between p130Cas and Grb2.³⁷  By using 2 dominant-negative forms of Grb2, R86K and P49L, respectively mutated in the SH2 and in the N-terminus SH3 domain, we were able to show that Grb2 was involved in the binding of NPM-ALK to p130Cas. In fact the R86K Grb2 construct was able to disrupt the interaction between NPM-ALK and p130Cas, whereas the P49L mutation had no effect, indicating that the SH2 domain of Grb2 was fundamental for the association between NPM-ALK and p130Cas (Figure 5B). Moreover, as a result of the releasing of NPM-ALK from the complex, p130Cas was no longer phosphorylated (Figure 5C).

Regarding the morphologic features, we asked if p130Cas could influence NPM-ALK–induced cytoskeleton modifications. We showed that the expression of NPM-ALK in NIH3T3 MEF cells led to spindle-shaped morphology and the outgrowth of elongated processes (Figure 6A-B), but both of the effects were not evident in Cas^-/- MEFs (Figure 6C-D). In immunofluorescence, NIH3T3 MEFs expressing NPM-ALK displayed a clear loss of the actin structure (Figure 6F) in contrast to Cas^-/- fibroblasts in which the actin filaments were still detectable (Figure 6H), thus suggesting an important role of p130Cas in NPM-ALK–mediated actin filament remodeling.

Next, we investigated the functional role of p130Cas in some NPM-ALK–mediated biologic effects, such as cell transformation and migration. First, we tested the colony-forming ability of NPM-ALK–expressing cells by soft-agar assay. We infected NIH3T3 MEFs and Cas^-/- MEFs with retroviruses expressing NPM-ALK and NPM-ALK^K210R as control (Figure 6I). NPM-ALK was efficient in transforming NIH3T3 MEFs as expected but failed to transform Cas^-/- fibroblasts, thus underlining the importance of p130Cas in mediating the transforming properties of NPM-ALK (Figure 6K).

To exclude the possibility of an intrinsic proliferative defect of Cas^-/- fibroblasts, we performed a cell-cycle analysis on nonconfluent MEFs in logarithmic growth phase and found no proliferative differences between Cas^-/- cells and NIH3T3 MEFs (Figure 6J). In addition, we rescued Cas^-/- fibroblasts by retroviral infection with Pallino p130Cas. In soft-agar assay the ectopic hyperexpression of p130Cas was sufficient to transform Cas^-/- fibroblasts (data not shown) in accordance to data previously described.³⁸ 

Finally, we asked whether the overexpression of p130Cas could increase the cellular migration rate. For this purpose, we infected both Cas^-/- and Cas^+/+ fibroblasts with Pallino NPM-ALK retrovirus (Figure 7A), sorted them for green fluorescent protein (GFP) content, and then performed a Matrigel migration assay. The expression of NPM-ALK was sufficient to increase cell migration as expected, although p130Cas expression did not provide an additional effect on cell migration (Figure 7B). Nevertheless, we also investigated the role of p130Cas in lymphoid cells migration; we overexpressed p130Cas together with either NPM-ALK or NPM-ALK^K210R in CEM cells via retroviral infection, and GFP-positive cells were enriched by cell sorting. Despite the strong expression levels of p130Cas and NPM-ALK obtained in CEM cells (Figure 7C) in a chemotaxis assay with SDF-1α, the synergistic effect of SDF-1α and NPM-ALK on cell migration was not further increased by the overexpression of p130Cas, thus suggesting that p130Cas could not be a fundamental player in this process (Figure 7D).

Many canonic steps of malignant transformation in which the fusion protein NPM-ALK is directly involved have been fairly well described and include increased mitogenity, protection from apoptosis, and growth in the absence of adhesion or growth factors.^10,39,40  The molecular mechanisms underlying these processes are similar to those employed by other tyrosine kinases and rely on the activation of signaling molecules such as PI3K, PLC-γ, Stat3, and MAPK.^15-17,39,41 

In line with these oncogenic properties, NPM-ALK is also capable of changing cell morphology in a similar fashion to other well-known oncogenes. In fibroblasts, NPM-ALK induces an elongated and birifrangent morphology and in lymphoid cells is able to increase the size of the cells as well as to drive anaplastic features.¹⁹ 

In the present study we started from a proteomic approach to identify molecules involved in these effects induced by NPM-ALK. By mass spectrometry on 293 T-Rex cells expressing a tetracycline-inducible form of NPM-ALK, we identified some cytoskeleton-associated proteins that precipitate or are phosphorylated as an effect of the tyrosine kinase activity of NPM-ALK. Interestingly, similar proteins, such as beta actin and tubulin, have also been found to be phosphorylated in anaplastic lymphoma cell lines (Karpas and DHL) with a recently described method of immunoaffinity profiling of tyrosine phosphorylation,³⁰  thus rendering our approach based on inducible 293 T-Rex cells a reliable method. Together with structural proteins, in the MALDI analysis of precipitated bands we found p130Cas, a 130-kilodalton (kDa) phosphotyrosine (pTyr)–scaffolding protein that normally associates with 2 oncoproteins, pp60v-src (v-Src) and p47gag-crk (v-Crk).^35,42 

The fundamental role of p130Cas in actin cytoskeleton remodeling and cell migration has been suggested by several lines of evidence including its phosphorylation following integrin engagement and its presence in the focal adhesions that form the molecular bridges between the extracellular matrix and the actin cytoskeleton.⁴³  In agreement with these functions, Cas^-/- fibroblasts show normal focal adhesion formation but have defects in cell migration.^38,44  In these events, the phosphorylation of p130Cas is mediated by FAK, Src, and protein tyrosine kinase 2β (Pyk2) in an integrin-dependent fashion.^45,46 

Here we show that p130Cas precipitates with the oncogenic tyrosine kinase NPM-ALK both in 293 T-Rex cells and in human-derived anaplastic lymphoma cells Karpas, DHL, and TS. In addition, p130Cas is phosphorylated when NPM-ALK is expressed and this phosphorylation is strictly dependent on the intrinsic tyrosine kinase activity of ALK since a kinase-dead mutant is no longer able to bind and phosphorylate p130Cas. Similarly to NPM-ALK, the ATIC-ALK fusion protein and the full-length ALK receptor phosphorylate p130Cas, thus excluding an essential role of the NPM portion of the molecule and indicating a general mechanism most likely shared by all ALK fusion proteins. Therefore, these findings add ALK to other tyrosine kinase receptors known to phosphorylate p130Cas such as EGF, fibroblast growth factor (FGF), insulin-like growth factor 1 (IGF-1), nerve growth factor (NGF), and platelet-derived growth factor (PDGF)³¹  and, more interestingly, are comparable to another fusion protein involved in hematopoietic malignancies, break point cluster region–Abelson (Bcr-Abl), which was demonstrated to bind and phosphorylate p130Cas in Bcr-Abl–positive cell lines and in samples obtained from chronic myeloid leukemia (CML) and acute lymphoblastic leukemia (ALL) patients.^47,48 

The phosphorylation of p130Cas at a molecular level has been shown in vitro to be mediated by 4 different kinases (ie, Src, Fak, Pyk2, and Abl). Since it has recently been described that Src kinase can bind to and be activated by NPM-ALK,³⁴  we considered the possibility that p130Cas phosphorylation by NPM-ALK required Src as intermediate kinase. However, in triple knock-out Src, Yes1, and Fyn cells NPM-ALK was still able to phosphorylate p130Cas at least to some extent and p130Cas phosphorylation was still detectable both in MEF NIH3T3 fibroblasts and in lymphoid cell lines even after the treatment with the Src-family inhibitor PP2, thus indicating the existence of a Src-independent pathway in p130Cas activation. This pathway could pass through the adaptor protein Grb2, a molecule previously known to bind both NPM-ALK³⁶  and p130Cas.³⁷  Indeed, we show in this paper that in 293 T-Rex cells, an SH2 Grb2 mutant, acting as a dominant-negative, impaired both the NPM-ALK binding to and the phosphorylation of p130Cas, thus revealing Grb2 as a key regulator of this interaction.

The role of p130Cas in ALK-mediated cytoskeleton organization was evident by studying the polymerization of the actin filaments. Deregulated ALK kinase activity forces cells to acquire a transformed phenotype with reduced adherence to the plate and a spindle, more birifrangent shape. In addition, ALK-transformed MEFs show long processes, in a way similar to the described neurite outgrowth induced by ALK in PC12 cells,⁸  indicative of a direct role of ALK in cytoskeleton organization. Indeed, ALK-transformed MEFs showed a prominent actin depolymerization similar to that induced by other oncogenic fusion proteins such as Bcr-Abl.⁴⁹  Interestingly, these effects were clearly impaired in Cas^-/- MEFs, indicating an essential role of p130Cas in ALK-mediated actin depolymerization. It still remains to be clarified if p130Cas could be involved also in the cellular shaping of NPM-ALK–expressing lymphoid cells.

Besides the functions on organizing the cell cytoskeleton, p130Cas is involved in broader oncogenic effects that include cell proliferation and apoptosis, indicating p130Cas as a major player in the complex network that is connecting cell adhesion and migration to cell growth and survival.³¹  Indeed, p130Cas overexpression is sufficient to induce an anchorage-independent growth in cells,^38,50  possibly through the activation of the c-Jun N-terminal kinase (JNK)/c-jun pathway,⁵¹  as well as protection from apoptosis.³¹  In this study we tested the role of p130Cas in NPM-ALK–mediated transformation, showing that Cas^-/- fibroblasts can no longer be transformed by NPM-ALK. These effects on anchorage-independent growth could derive from a combination of cytoskeleton modifications and proliferative signals both provided by NPM-ALK in wt MEFs but impaired in the absence of p130Cas. Overall the data are suggesting that NPM-ALK could control a Src-dependent and a Src-independent transformation pathway that both use p130Cas as pivotal adaptor. The Src-independent pathway may require Grb2 as key adaptor molecule, as previously discussed. In this scenario, it is possible to further speculate that NPM-ALK activates both Src and p130Cas, where phosphorylated p130Cas may further enhance Src activity in a sort of activating loop similar to what has already been described in other cancers.⁵² 

In this paper, we also describe a novel function of NPM-ALK (ie, its capability of increasing the migration rate of T lymphocytes). CEM lymphoblastoid cells have previously been proven to migrate in a gradient of SDF-1α, which is the ligand for CXCR4, a chemokine receptor expressed in T cells.²⁹  Upon retroviral expression of NPM-ALK, CEM cells almost doubled their migration rate, indicating that NPM-ALK tyrosine kinase activity influences the activation state of molecules involved in this process. However, in contrast to transformation, we could not find a clear effect of p130Cas overexpression in NPM-ALK–induced migration. Regarding lymphoid cells, this could be explained by considering that in T lymphocytes a major role in beta1 integrin– and T-cell receptor (TCR)–mediated migration has been shown for Cas-L, a p105-kDa protein that is a member of the Cas family selectively expressed in lymphocytes.^32,33  Indeed, we showed that NPM-ALK is capable of phosphorylating Cas-L in lymphocytes. It is therefore possible that NPM-ALK in lymphocytes may promote migration through the activation of different molecules than p130Cas. Further studies in lymphocytes are needed to solve this issue.

In conclusion, in this study we show p130Cas as a novel downstream interactor of NPM-ALK capable of modulating its transforming properties. This role could involve not only the p130Cas effects on cytoskeleton remodeling but also its involvement in important proliferative and survival pathways. Further studies are required to unravel this differential requirement of ALK for p130Cas or its family members in different cell subtypes.