AXL regulates mesothelioma proliferation and invasiveness - Nature

Oncogene (2011) 30, 1643–1652

& 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc

ORIGINAL ARTICLE

AXL regulates mesothelioma proliferation and invasiveness W-B Ou1, JM Corson1, DL Flynn2, W-P Lu2, SC Wise2, R Bueno3, DJ Sugarbaker3 and JA Fletcher1 1 Department of Pathology, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA; 2Department of Research Discovery, Deciphera Pharmaceuticals, LLC, Lawrence, KS, USA and 3Division of Thoracic Surgery, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA

Mesothelioma is an asbestos-associated and notoriously chemotherapy-resistant neoplasm. Activation of the receptor tyrosine kinases (RTKs), epidermal growth factor receptor and MET, has been described in subsets of mesothelioma, suggesting that TKs might represent therapeutic targets in this highly lethal disease. We employed proteomic screening by phosphotyrosine immunoaffinity purification and tandem mass spectrometry to characterize RTK activation in mesothelioma cell lines. These assays demonstrated expression and activation of the AXL protein, which is an RTK with known oncogenic properties in non-mesothelial cancer types. AXL was expressed and activated strongly in 8 of 9 mesothelioma cell lines and 6 of 12 mesothelioma biopsies, including each of 12 mesotheliomas with spindle-cell histology. Somatic AXL mutations were not found, but all mesotheliomas expressed an alternatively spliced AXL transcript with in-frame deletion of exon 10, and six of seven mesothelioma cell lines expressed the AXL ligand, growth arrest-specific 6 (GAS6). GAS6 expression appeared to be functionally relevant, as indicated by modulation of AXL tyrosine phosphorylation by knockdown of endogeneous GAS6, and by administration of exogenous GAS6. AXL silencing by lentivirus-mediated short hairpin RNA suppressed mesothelioma migration and cellular proliferation due to G1 arrest. The AXL inhibitor DP-3975 inhibited cell migration and proliferation in mesotheliomas with strong AXL activation. DP3975 response in these tumors was characterized by inhibition of PI3-K/AKT/mTOR and RAF/MAPK signaling. AXL inhibition suppressed mesothelioma anchorage-independent growth, with reduction in colony numbers and size. These studies suggest that AXL inhibitors warrant clinical evaluation in mesothelioma. Oncogene (2011) 30, 1643–1652; doi:10.1038/onc.2010.555; published online 6 December 2010 Keywords: AXL; receptor tyrosine kinases; proliferation; invasiveness; mesothelioma

Correspondence: Dr W-B Ou or Dr JA Fletcher, Department of Pathology, Brigham and Women’s Hospital and Harvard Medical School, 75 Francis Street, Boston, MA 02115, USA. E-mail: [email protected] or jfl[email protected] Received 24 February 2010; revised 13 September 2010; accepted 28 October 2010; published online 6 December 2010

Introduction Malignant pleural mesothelioma is a locally aggressive and highly lethal neoplasm, linked epidemiologically to asbestos exposure (Craighead and Mossman, 1982) and SV40 virus infection (Carbone et al., 1994). Pathologically, mesothelioma histological subtypes are epithelioid, spindle cell and mixed (epithelioid and spindled) (Janne et al., 2002), of which the spindled type has the worst prognosis. Conventional chemotherapies and radiation therapy have limited efficacy against mesothelioma, and substantial improvements in survival will require development of novel and more effective pharmacological interventions. A better understanding of mesothelioma biology—including key growth factor signaling pathways—will likely be useful in identifying biologically rational targets for novel therapies. Various lines of evidence support roles of receptor tyrosine kinase (RTK) in mesothelioma pathogenesis. Epidermal growth factor receptor expression and activation have been demonstrated in some mesotheliomas, resulting in downstream activation of MAPK proliferation-associated signaling pathways (Pache et al., 1998; Yang et al., 2007). These studies suggest that epidermal growth factor receptor activation might play an early role in asbestos-induced mesothelial mitogenicity and carcinogenesis. The RTK protein, MET, and its ligand (hepatocyte growth factor) are co-overexpressed in mesothelioma compared with nonneoplastic mesothelial cells (Klominek et al., 1998; Tolnay et al., 1998), and MET activation results in increased mesothelioma cell proliferation (Klominek et al., 1998). PHA-665752, a specific small-molecular inhibitor of MET, caused mesothelioma cell cycle arrest and inhibited phosphorylation of MET, p70S6K, AKT and MAPK (Mukohara et al., 2005). Furthermore, MET small interfering RNA also inhibited mesothelioma cell line proliferation (Jagadeeswaran et al., 2006). VEGF (vascular endothelial growth factor) and VEGFtype C have roles in angiogenesis and lymphangiogenesis in mesothelioma (Ohta et al., 1999; Catalano et al., 2002). VEGF regulates mesothelioma cell proliferation via activation of its receptors, VEGFR1 and VEGFR2 (Strizzi et al., 2001). Insulin-like growth factor 1 and IGF1R expression levels are comparable in nonneoplastic and malignant mesothelial cells (Lee et al., 1993), and antisense IGF1R inhibition resulted in

AXL regulates mesothelioma proliferation and invasiveness W-B Ou et al

1644

decreased proliferation and tumorigenicity in a mesothelioma cell line (Pass et al., 1996). Recently, the EPHB4 RTK was found to be highly expressed in mesothelioma but not in normal mesothelial cells. EPHB4 knockdown inhibited mesothelioma cell proliferation, migration and invasion by initiating caspase-8mediated apoptosis and downregulating antiapoptotic Bcl-xl (Xia et al., 2005). More recently, c-Src expression and activation were reported in mesothelioma, and dasatinib-mediated c-Src inhibition resulted in mesothelioma apoptosis and cell cycle arrest (Tsao et al., 2007). These varied studies suggest that RTK activation contributes to oncogenic progression from non-neoplastic mesothelial progenitor cell to mesothelioma. AXL (also known as UFO/ARK/Tyro) is a RTK with oncogenic potential and transforming activity (Wimmel et al., 2001), which is stimulated by its ligand GAS6 (growth arrest-specific 6), a vitamin K-dependent growth potentiating factor (Varnum et al., 1995). AXL was first identified as a transforming gene in chronic myeloid leukemia (Janssen et al., 1991; O’Bryan et al., 1991), but is also overexpressed in metastatic colon and prostate cancer, thyroid carcinoma, breast cancer and melanoma (Quong et al., 1994; Craven et al., 1995; Ito et al., 1999; Jacob et al., 1999; Meric et al., 2002). Various studies have demonstrated roles of GAS6/AXL in regulating cell growth and survival in normal and cancer cells (Stitt et al., 1995; Wimmel et al., 2001). AXL is characterized by a unique extracellular composition consisting of immunoglobulin-like and fibronectin-type III-like domains, which are also found in adhesion molecules of the cadherin and immunoglobulin superfamily, therefore suggesting that AXL might regulate cell adhesion (Burchert et al., 1998). Other evidence is consistent with roles of AXL in tumor cell invasion, metastasis and survival (Jacob et al., 1999; Holland et al., 2005; Mahadevan et al., 2007). In the present study, we used phosphotyrosine immunoaffinity and tandem mass spectrometry to identify activated RTKs in mesotheliomas. These assays demonstrated AXL activation, and we found that AXL inhibition by RNA interference and a small-molecular inhibitor exerted antiproliferative effects in mesothelioma cell lines because of suppression of growth-mediating signaling pathways accompanied by G1-phase arrest.

Results AXL receptor tyrosine kinase is activated and strongly expressed in mesothelioma Phosphotyrosine eluates and panRTK immunoprecipitations were prepared from MESO257 and MESO924 cell lines, and phosphotyrosine immunoblot stains demonstrated B140 kDa proteins in both preparations (Figure 1). Mass spectrometry evaluation of the B140 kDa protein in a coomassie-stained gel demonstrated the AXL receptor tyrosine kinase. This finding was confirmed by stripping the immunoblots and restaining for AXL (Figure 1). Oncogene

Figure 1 Expression of tyrosine-phosphorylated tyrosine kinase proteins in MESO257 and MESO924: demonstrated by panRTK immunoprecipitations and phosphotyrosine immunopurifications, which were immunoblotted and stained for phosphotyrosine and AXL. The predominant tyrosine-phosphorylated proteins in the panRTK IP lanes (upper panels) were AXL (B150 kDa), epidermal growth factor receptor (B175 kDa) and an unidentified protein at B250 kDa.

Using MESO924 as an internal standard, AXL protein expression was evaluated by immunoblotting in a variety of human cancers: mesothelioma, myeloma (B cell), Jurkat (T cell), gastrointestinal stromal tumor, leiomyosarcoma, solitary fibrous tumor, adenoid cystic carcinoma, non-small-cell lung cancer, colon carcinoma, renal cancer, prostate cancer, undifferentiated carcinoma, mast cell leukemia and melanoma (Figure 2a). AXL expression was stronger in mesothelioma than in the other cancer types. Further immunoblotting studies demonstrated robust AXL expression in eight of nine mesothelioma cell lines, whereas expression was nearly undetectable in normal mesothelial cells (Figure 2b). AXL expression was then evaluated in frozen mesothelioma biopsy materials (uncultured) of various histological subtypes, with a spindle-cell sarcoma as non-mesothelial comparator (histologically similar to spindle-cell mesothelioma) (Figure 2c). AXL expression, particularly the mature, glycosylated form (upper band, B140 kDa) was strongest in the mesotheliomas containing spindle-cell components (Figure 2c). AXL activation in mesothelioma cell lines and tumor tissues was further evaluated by AXL immunoprecipitation, followed by phosphotyrosine immunoblotting (Supplementary Figure 1). AXL was strongly tyrosine phosphorylated in eight of nine mesothelioma cell lines (MESO59, MESO257, MESO296, MESO428, MESO542, MESO589, MESO647 and MESO924) and five primary frozen tumor tissues (95–99, 95–697, 97–150, 97–470 and 98–542). By contrast, AXL phosphorylation was undetectable in normal mesothelial cells (97–510) (Supplementary Figure 1). Sequencing analyses did not demonstrate AXL mutations in any of five mesothelioma cell lines (MESO924, MESO257, MESO428, MESO647 and JMN1B), although


1645

Figure 2 (a) Immunoblotting demonstrates strong AXL expression in mesothelioma compared with a broad range of non-mesothelial tumors, including gastrointestinal stromal tumor, adenoid cystic carcinoma, non-small-cell lung cancer, colon carcinoma, renal cancer, prostate cancer, carcinoma, mast cell leukemia, melanoma, leiomyosarcoma, solitary fibrous tumor and BPH. Actin staining is a loading control. (b) Immunoblotting comparison of AXL expression in mesothelioma cell lines and normal mesothelial cells. (c) Immunoblotting evaluations of AXL expression in primary mesotheliomas of different histological subtypes. The left panel has primarily epithelial-type mesothelioma, but includes one spindle-cell mesothelioma and one spindle-cell sarcoma (histologically similar to spindle-cell mesothelioma) for comparison. The right panel contains mesotheliomas with spindle-cell components.

all expressed an alternatively spliced AXL transcript with inframe deletion of exon 10, which encodes the extracellular juxtamembrane region. AXL regulation of mesothelioma proliferation and cell cycle checkpoints AXL shRNA (short hairpin RNA)-mediated knockdown resulted in B95% inhibition of AXL protein expression in mesothelioma cell lines, MESO924, MESO296 and MESO428 (Figure 3a). AXL knockdown in these mesotheliomas resulted in B50% inhibition of cell viability (Promega CellTiter-Glo assay; Madison, WI, USA) at 3 and 7 days after AXL silencing, compared with the empty vector control (Figure 3b). Comparable antiproliferative effects were seen after infection of the mesothelioma cells with AXL shRNA for 5 and 9 days, and all studies were confirmed using at least two independent shRNA transductions (data not shown). AXL knockdown in MESO296 was associated with a reduction in the G2 peak and a slight increase in the G1/0 peak (63.54% for empty vector versus 65.89% for AXL shRNA) (Figure 3c). AXL knockdown in MESO924 and MESO428 resulted in dramatic G1 block. The G1/0 peaks were 52.72% (MESO924) and 59.44% (MESO428) in the empty vector-treated cells compared with 64.38% (MESO924) and 68.17%

(MESO428) in AXL shRNA-treated cells (Figure 3c). AXL knockdown also induced apoptosis: MESO924 and MESO296 nuclear fragmentations were demonstrated in 0.44% and 0.7 of cells, respectively, treated with empty vector control, but in 3.27% and 4.07 cells, respectively, treated with AXL shRNA (Figure 3c). To determine whether AXL kinase activity regulates mesothelioma viability, the AXL kinase was inactivated using a small-molecule inhibitor, DP-3975 (Deciphera Pharmaceuticals, LLC (Lawrence, KS, USA)). These studies were performed in four mesothelioma cell lines expressing strongly tyrosine-phosphorylated AXL (MESO296, MESO924, MESO257 and MESO428) and in mesothelioma JMN1B as a comparator that expresses non-phosphorylated AXL. AXL immunoprecipitation and phosphotyrosine immunoblotting were performed after dimethyl sulfoxide versus DP3975 treatment (Figure 4a). DP-3975 inhibited AXL substantially in MESO296, MESO924, MESO257 and MESO428. AXL response to DP-3975, evaluated further in MESO296, was dose dependent (IC50B100 nM) and associated with inhibition of AXL:PI3-K interaction (Figure 4a). DP-3975 treatment inhibited downstream signaling intermediates, AKT and S6, in a dose-dependent manner in MESO296 and MESO257 but not in JMN1B (Figure 4b). Likewise, cell viability Oncogene


1646

Figure 3 Immunoblotting, cell viability and cell cycle evaluations for mesothelioma cell lines (MESO924, MESO296 and MESO428) at 96 h after infection by lentiviral AXL shRNA constructs. (a) Immunoblotting demonstrates efficiency of AXL knockdown. Actin immunostains show equivalence of lane loading. Control lanes for each cell line include uninfected cells (untreated lane) and cells infected with empty lentiviral vector. (b) Cell viability was evaluated by a Cell-titer Glo ATP-based luminescence assay in MESO924 (black bars), MESO296 (gray bars),and MESO428 (white bars) at days 3 and 7 after infection with lentiviral AXL shRNA. Data were normalized to the empty lentivirus infections and represent the mean values (±s.d.) of quadruplicate cultures. (c) Cell cycle analyses were performed 4 days after infection by lentiviral AXL shRNA constructs. MESO924 and MESO428 cells show substantial G1-block after AXL silencing compared with empty vector control. MESO924 and MESO296 show substantial nuclear fragmentation after AXL silencing.

was inhibited by DP-3975 in the AXL-phosphorylated mesotheliomas, but not in JMN1B (Figure 4c). Cell viability IC50s at day 3 were 1.1, 1.1 and 0.9 mM, respectively, in MESO296, MESO924 and MESO257, compared with 42.0 mM in JMN1B. AXL regulation of mesothelioma migration and invasiveness Mesothelioma clinical spread is characterized by extension and invasive growth along the pleura, pericardium Oncogene

or peritoneum. Therefore, assays were performed to evaluate roles of AXL in mesothelioma migration and invasion. Wound-healing assays in AXL-phosphorylated MESO296 cells demonstrated that AXL shRNAmediated knockdown resulted in marked inhibition of wound closure at 24–48 h, whereas complete wound closure was seen in control cells infected with empty lentivirus vector (Figure 5a). Additional wound-healing assays were performed at 48–72 h in mesothelioma cells treated with AXL-inhibitor DP-3975 versus dimethyl


1647

by AXL shRNA knockdown or DP-3975, and after PI3K inhibition by BEZ235 (Supplementary Figure 2). MESO296 and MESO428 mesothelioma colonies were three- to eightfold smaller when grown in the presence of DP-3975, BEZ235 or with AXL shRNA knockdown (Supplementary Figure 2A and B). Similarly, colony formation was substantially reduced by these interventions (Supplementary Figure 2C). AXL phosphorylation of ligand GAS6 dependence Ligand-mediated AXL modulation was evaluated by suppressing endogenous GAS6 in mesothelioma cell lines and by treating the cell lines with exogenous GAS6 (Figures 6a, b and c). Immunoblotting studies demonstrated coexpression of GAS6 and phospho-AXL in six of seven mesothelioma cell lines, whereas GAS6 expression was nearly undetectable in normal mesothelial cells (Figure 6a). Notably, the highest GAS6 expression was seen in mesotheliomas with the highest phospho-AXL expression (MESO924, MESO589, MESO296 and MESO428). GAS6 shRNA knockdown inhibited mesothelioma AXL tyrosine phosphorylation substantially, whereas total AXL expression increased, consistent with accumulation of inactive AXL (MESO428 cell line, Figure 6b). Mesothelioma AXL phosphorylation could also be regulated by exogenous GAS6 (Figure 6c), resulting in increased phosphorylation of downstream signaling intermediates, AKT, MAPK and S6 (Figure 6d). Conversely, AXL inhibition by DP-3975 resulted in dose-dependent inactivation of AKT, MAPK and S6 (Figure 6d).

Discussion

Figure 4 (a) Inhibition of AXL phosphorylation by smallmolecular inhibitor, DP-3975 (1 mM), and dose response for DP3975-mediated AXL inhibition in MESO296. All assays were performed after 4 h of DP-3975 treatment. (b) DP-3975 inhibits AXL downstream signaling intermediates, AKT and S6, in MESO296 and MESO257, but not in JMN1B. b-Actin stain is a loading control. (c) Cell viability assay (Cell-titer Glo) after 72 h of DP-3975 treatment. Data were normalized to the dimethyl sulfoxide control and represent the mean values (±s.d.) of quadruplicate cultures.

sulfoxide control. Cell migration was inhibited by DP-3975 in AXL-phosphorylated MESO296 and MESO257, but not in JMN1B, whereas the wounds were healed in all cell lines after 72 h of dimethyl sulfoxide exposure (Figure 5c). Similarly, matrigel assays demonstrated 45% inhibition of invasiveness after AXL knockdown in AXL-phosphorylated mesotheliomas, compared with infection with empty vector (Figure 5b). AXL and PI3-K regulation of mesothelioma anchorage-independent growth was determined by evaluating clonogenic growth in soft agarose after AXL inhibition

Malignant pleural mesothelioma is a highly lethal malignancy that is often associated with asbestos exposure. Current therapies, involving intensive surgery, radiation and chemotherapy, have improved survival for some patients, although nearly all succumb, eventually, to the disease. Tyrosine kinase proteins are key regulatory elements of proliferation and survival in many cancers (Gschwind et al., 2004), and tyrosine kinase activation is essential in the development of mesothelioma from the starting point of a nonneoplastic mesothelial progenitor cell. Herein, we used phosphotyrosine immunoaffinity and mass spectrometry to identify activated tyrosine kinase proteins in mesothelioma (Figure 1): these studies demonstrated AXL activation, which we show to be a frequent event in mesothelioma cell lines and mesothelioma surgical specimens, but not in normal mesothelial cells. Notably, AXL expression, overall, was even stronger in mesothelioma than in various other cancer types that have been reported to feature high levels of AXL expression (Figure 2a)—including prostate cancer, melanoma and colon carcinoma (Quong et al., 1994; Craven et al., 1995; Jacob et al., 1999). We show that AXL expression and activation are particularly strong in the spindle-cell subtypes of mesothelioma, which are known to be Oncogene


1648

Figure 5 (a) Lentiviral shRNA-mediated AXL knockdown inhibited migration of MESO296, as assessed by in vitro wounding assays. (b) AXL knockdown inhibited migration of mesothelioma cells (MESO924, MESO296 and MESO428) as assessed by matrigel invasion and transwell migration assays. (c) DP-3975 treatment inhibited migration of MESO257 and MESO296, but not JMN1B, as assessed by in vitro wounding assays.

associated with short survival (Figure 2c). Various recent studies implicate AXL as an oncogenic factor in advanced solid tumors: AXL is overexpressed in highly invasive lung adenocarcinoma cell lines, metastatic DU145 prostate cancer models and metastatic colon tumors (Craven et al., 1995; Jacob et al., 1999; Shieh et al., 2005). These and other studies suggest that AXL overexpression promotes invasiveness (Lay et al., 2007; Tai et al., 2008; Li et al., 2009). We show that in vitro mesothelioma migration and invasion are inhibited by AXL antagonists, either by AXL shRNA knockdown or the DP-3975 small-molecular inhibitor of AXL kinase activity (Figures 5a, b and c). These findings are in keeping with a recent publication showing that increased AXL expression correlates with poor prognosis in mesothelioma (Pass et al., 2004). In preliminary genomic evaluations, we found no evidence for AXL gain-of-function mutations in five mesothelioma cell lines, although all cases expressed an alternatively spliced AXL transcript with in-frame deletion of exon 10. Each of these five mesothelioma Oncogene

cell lines expressed both types 1 and 2 AXL transcripts, in which the type 2 transcript lacks the nine-codon exon 10, encoding part of the extracellular juxtamembrane region, of the type 1 transcript. It is not known whether the type 1 and 2 transcripts differ in biological activity, and AXL genomic rearrangements or amplifications were not observed by cytogenetic and single-nucleotide polymorphism array analyses in 18 mesotheliomas (Fletcher, data not shown). However, AXL activation was not an inevitable consequence of AXL expression, because certain mesotheliomas (for example JMN1B) and non-neoplastic mesothelial cells featured strong AXL expression in absence of evident activation. Notably, expression of AXL ligand (GAS6) correlated strongly with mesothelioma AXL phosphorylation (Figure 6): furthermore, mesothelioma AXL activity was modulated by shRNA-mediated inhibition of endogenous GAS6 expression and by administration of exogenous GAS6. These findings are strong evidence for autocrine/paracrine mechanisms as one of the contributors to AXL activation in mesothelioma.


1649

that AKT and S6 are in part AXL dependent in MESO296, MESO257 and MESO924 (mesothelioma cell lines expressing activated AXL) but not in JMN1B (mesothelioma cell line expressing similar levels of AXL, but non-activated) (Figures 4b and 6d), whereas MAPK was AXL dependent in MESO296 and MESO924 (Figure 6d). Two AXL autophosporylation residues (Y779 and Y821) are putative PI3-K-binding sites (O’Bryan et al., 1991). Interactions between AXL and PI3-K were blocked in a dose-dependent manner by the AXL inhibitor, DP-3975 (Figure 4a), suggesting that AXL regulation of PI3-K/AKT/mTOR signaling requires AXL:PI3-K interaction. AKT is a crucial intermediate in RTK/PI3-K signaling, with contributions to cell proliferation, survival and migration/ invasiveness, depending on cell context (Sordella et al., 2004; Stommel et al., 2007; Harir et al., 2008; Faber et al., 2009). Soft agarose assays confirmed that the PI3-K/AKT/mTOR signaling pathway contributes to AXL-mediated mesothelioma tumorigenic properties (Supplementary Figure 2A and 2C). Similarly, AXL inhibition by shRNA-mediated transcript knockdown or biochemical inhibition was associated with decreased mesothelioma viability and decreased cell migration in a wound-healing assay, but this was seen only in mesothelioma cells expressing activated AXL (Figures 3b, 4c, 5a and c). Furthermore, shRNA-mediated AXL silencing resulted in G1 phase arrest and decreased invasiveness in mesotheliomas expressing activated AXL (Figures 3c and 5b). AXL inhibition also suppressed anchorageindependent colony formation, as manifested by reduction in colony numbers and colony size, supporting a role for AXL activation in the mesothelioma growth typically seen in pleural and peritoneal fluids (Supplementary Figure 2A, B, and C). In conclusion, our studies suggest that AXL activation contributes to aberrant viability, migration and invasiveness in mesothelioma. Mesothelioma is a highly lethal disease in which most afflicted individuals succumb to consequences of local tumor spread/invasion. Therefore, AXL warrants evaluation as a therapeutic target to inhibit clinical progression of mesothelioma. Figure 6 Ligand GAS6 relevance to mesothelioma AXL activity. (a) GAS6 expression is associated with increased phospho-AXL expression in mesothelioma cell lines and normal mesothelial cells. (b) GAS6 stable shRNA knockdown was performed by puromycin selection for 10 days, resulting in substantial inhibition of AXL phosphorylation. (c) GAS6-dependent AXL phosphorylation in serum-starved mesothelioma cells: treated with GAS6 for 10 min. (d) Inhibition of AXL and downstream signaling intermediates (AKT, MAPK and S6) by DP-3975 in serum-starved MESO924 and MESO296 cells stimulated with 400 ng/ml GAS6 for 30 min. Evaluations were performed after 4 h of DP-3975 treatment. b-Actin stain is a loading control.

AXL was phosphorylated in mesothelioma cell lines and primary tumor tissues, but not in normal mesothelial cells (Figures 1 and 6, and Supplementary Figure 1), suggesting that AXL contributes to regulation of crucial signaling pathways in mesothelioma. Indeed, we show

Materials and methods Antibodies and reagents We developed an anti-panRTK polyclonal rabbit antibody that recognizes highly conserved epitopes in the tyrosine kinase family (Heinrich et al., 2003). Polyclonal antibodies to AXL and GAS6 were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Polyclonal antibodies to AKT and all phosphospecific antibodies were from Cell Signaling Technology (Beverly, MA, USA). Monoclonal mouse antibodies were to phosphotyrosine (PY99, Santa Cruz), S6 (Cell Signaling Technology) and actin (Sigma-Aldrich Corp., St Louis, MO, USA). Lentiviral GAS6 shRNA constructs were from Open Biosystems (Huntsville, AL, USA). Recombinant human GAS6 was from R&D Systems (Minneapolis, MN, USA) and was reconstituted in sterile phosphate buffered saline. Mouse anti-phosphotyrosine-sepharose 4B, Protein A and Protein G sepharose beads, NOVEX Coomassie colloidal blue Oncogene


1650 staining kit, NuPAGE TM 4–12% Bis-Tris Gel, Lipofectamine and Plus reagent were from Invitrogen life Technologies (Carlsbad, CA, USA). Phenyl phosphate, crystal violet, ultrapure agarose and polybrene were from Sigma. The AXL inhibitor, DP-3975, was developed by Deciphera Pharmaceuticals, LLC. Novel mesothelioma cell lines Eight mesothelioma cell lines were established from surgical materials from previously untreated patients. These studies were approved by the Brigham and Women’s Hospital Institutional Review Board, under a discarded tissues protocol. The MESO59, MESO257, MESO542 and MESO924 cell lines were established from epithelial-type mesotheliomas; MESO296, MESO589 and MESO647 from mixed-histology mesotheliomas; and MESO188 and MESO428 from spindlecell mesotheliomas. An additional mesothelioma cell line, JMN1B, was established from an epithelial-type mesothelioma (Demetri et al., 1989). Derivation of each cell line from the corresponding surgical specimen was corroborated by demonstrating the persistence of unique clonal cytogenetic aberrations—as seen in the primary tumors—in each of the cell lines (data not shown). Normal mesothelial cells (97–510) were established in our laboratory from a non-neoplastic pleural effusion. Frozen tumor specimens All frozen tumor specimens were discarded tissues, obtained from the Brigham and Women’s Hospital. In all, 95–776, 96– 338, 96–677, 97–301, 97–470 and 98–542 were epithelioid-type mesotheliomas; 95–599, 95–697, 96–299, 96–300, 96–975 and 97–150 were mixed-histology mesotheliomas. Phosphotyrosine immunoaffinity purification and tandem mass spectrometry RTKs were immunoprecipitated from mesothelioma cell lysates using panRTK polyclonal antibodies and were then immunoblotted and stained for phosphotyrosine (PY99) to identify the molecular mass of putative-activated RTKs. The activated RTKs were then characterized in MESO257 and MESO924 cell lines by mass spectrometry. This was accomplished by first purifying the activated proteins on a phosphotyrosine immunoaffinity column in which 0.5 ml of sepharose-conjugated anti-phosphotyrosine was loaded in a 10 ml polypropylene column and washed with a 5 volume of 1 phosphate buffered saline containing 0.02% sodium azide. The column was pre-equilibrated with a 5 volume of lysis buffer, followed by incubation with 10 mg of protein lysates at 4 1C with end-over-end mixing overnight. The flow-through fraction was collected and reloaded on the column, 10 times. The column was then washed with 1 ml lysis buffer, 5 times, and the tyrosine-phosphorylated proteins were eluted with 1 ml 100 mM phenyl phosphate in lysis buffer, 10 times. The eluates were dialyzed against 1l of 1 phosphate buffered saline buffer overnight at 4 1C. The sample was then concentrated to B30 ml of final volume using centricon filters (Millipore Corporation, Bedford, MA, USA). Finally, 20% of the eluates were immunoblotted and stained for phosphotyrosine, whereas the remaining 80% eluates were gel electrophoresed and stained using a NOVEX Coomassie colloidal blue stain kit protocol. A strongly tyrosine-phosphorylated protein band at 140 kDa was excised from the Coomassie gel, and protein sequences were determined by mass spectrometry. The blots were then stripped and restained with specific antibodies to validate candidate tyrosine kinase oncoproteins. Oncogene

Western blotting analysis Whole-cell lysates were prepared using lysis buffer (1% NP-40, 50 mM Tris–HCl (pH 8.0), 100 mM sodium fluoride, 30 mM sodium pyrophosphate, 2 mM sodium molybdate, 5 mM EDTA and 2 mM sodium orthovanadate) containing protease inhibitors (10 mg/ml aprotinin, 10 mg/ml leupeptin and 1 mM phenylmethylsulfonyl fluoride). Frozen tumor samples were diced into small pieces in cold lysis buffer on dry ice and homogenized using a Tissue Tearor (Model 398, Biospec Products Inc., USA) for two seconds, 3–5 times, on ice, and the cell lysate was then rocked overnight at 4 1C. Lysates were cleared by centrifugation at 14 000 r.p.m. for 20 min at 4 1C, and lysate protein concentrations were determined using a BioRad protein assay (Bio-Rad Laboratories Hercules, CA, USA). Electrophoresis and western blotting were performed as described previously (Rubin et al., 2001). The hybridization signals were detected by chemiluminescence (ECL, Amersham Pharmacia Biotechnology) and captured using a FUJI LAS1000-plus chemiluminescence imaging system. Immunoprecipitation Sepharose-protein G beads to goat polyclonal antibody and the sepharose-protein A to rabbit polyclonal antibody were used. In all, 1 mg of protein lysates (500 ml) was preadsorbed for 30 min using 20 ml of protein G or protein A beads at 4 1C. In all, 10 ml of primary antibodies against rabbit panRTK (10 mg/ml) or goat AXL (0.2 mg/ml) were rocked with the lysates for 2 h at 4 1C. Then 20 ml of sepharose-protein G or -protein A beads were added and rocked overnight at 4 1C, followed by centrifugation at 10 000 r.p.m. for 2 min at 4 1C, after which the sepharose beads were washed three times with 750 ml of IP buffer (25 min/each time) and once with 750 ml 10 mM Tris–Cl buffer (pH 7.6). Loading buffer (20 ml) was added to the beads and boiled for 5 min at 95 1C. RNA preparation and reverse transcriptase–PCR for mutation analysis RNA was prepared using Invitrogen Trizol reagent according to the manufacturer’s protocol. In all, 2 mg of total RNA were used for reverse transcription reactions using a Sigma Enhanced Avian HS RT–PCR Kit with a two-step format. PCR-amplified AXL complementary DNA fragments were gel purified before sequencing. Lentiviral AXL shRNA constructs and virus preparation The pLKO.1puro (7 kb) lentivirus construct contains an U6 promoter and HIV-1 RNA-packaging signal with a puromycin- and ampicillin-resistant element cloned 30 of the human phosphoglycerate kinase promoter. A cpptCTE was inserted 50 of the human phosphoglycerate kinase promoter. Human AXL shRNA constructs were generated by ligating the following oligomers into the unique AgeI and EcoRI sites of pLKO.1puro: AXL forward 50 -CCGGCGAAATCCTCTAT GTCAACATCTCGAGATGTTGACATAGAGGATTTCG TTTTTG-30 and reverse 50 -AATTCAAAAACGAAATCCT CTATGTCAACATCTCGAGATGTTGACATAGAGGAT TTCG-30 . Lentivirus preparations were produced by cotransfecting empty vector pLKO.1puro with AXL shRNA or GAS6 shRNAs, and helper virus-packaging plasmids pCMVDR8.91 and pMD.G (at a 10:10:1 ratio) into 293T cells. Transfections were carried out using Lipofectamine and Plus reagent. Lentiviruses were harvested at 24, 36, 48 and 60 h after transfection. Virus was frozen at 80 1C in appropriately sized aliquots for infection.


1651 Cell culture and virus infection Mesothelioma cells were cultured in RPMI 1640 medium with 15% fetal bovine serum and seeded in six-well plates. Lentiviral shRNA infections were carried out in the presence of 8 mg/ml polybrene. Cells were lysed for western blot analysis or harvested for cell cycle analysis at 96 h after infection. Following transduction, MESO428 cells were selected for stable expression of the GAS6 shRNAs using 2 mg/ml puromycin. Cell viability analysis MESO924, MESO257, MESO296, MESO428 and JMN1B cells were plated at 3000 cells/well in a 96-well flat-bottomed plate (Falcon, Lincoln NJ, USA) and cultured in RPMI1640 for 24 h before transduction with lentiviral empty vector or AXL shRNA, or treatment with AXL inhibitor. Proliferation studies were carried out after 3, 5, 7 and 9 days using the CellTiter-Glo luminescent assay from Promega, and quantitated using a Veritas Microplate Luminometer from Turner Biosystems (Sunnyvale, CA, USA). The data were normalized to the empty vector group or dimethyl sulfoxide. All assays were performed in quadruplicate wells and were averaged from two independent experiments for each cell line. Cell cycle analysis MESO924, MESO296 and MESO428 cells in six-well plates were trypsinized and washed once with Hank’s balanced salt solution at room temperature after infection with lentivirus for 96 h. For nuclear staining after lentiviral transduction, a DAPI-containing solution (nuclear isolation and staining solution; NPE systems, Pembroke Pines, FL, USA) was added to the cells, and the cell suspension was immediately analyzed in a flow cytometer (NPE Quanta, NPE Systems). Data analyses were performed using Modfit LT software 3.1 (Verity Software House, Topsham, ME, USA). In vitro wound healing assays Cell wounding studies were carried out using methods similar to those previously described (Shaw et al., 2001). A slash was created in near-confluent cell cultures using the tip of a P-1000 pipetman, at 96 h after shRNA transduction, or after adding 1 mM DP-3975. The plates were photographed at 0, 24, 48 and 72 h using Spot software (version 3.5.9 for MacOS) and a Nikon Eclipse TE2000-5 microscope. Cell migration and invasion assays The migratory ability and invasiveness of mesothelioma cells were evaluated by matrigel assay as previously described (Yang et al., 2007). Matrigel (Collaborative Research Inc., Boston, MA, USA) was diluted with RPMI1640 (1:2) and then

coated onto 12-well inserts (Boyden chamber) with 12 mm pore size, with incubation at 37 1C overnight. Cells were treated with lentiviral vector or AXL shRNA for 4 days, followed by suspension in 0.5 ml of 0.5% serum-containing RPMI and seeded on the upper chamber of each well with 1.5 ml of 15% serum-containing medium added to the lower chamber, the higher serum content in the lower chamber providing a chemotactic gradient. After 24 h, non-invading cells that remained on the upper surface of the filter were removed using a cotton swab and cells that remained adherent to the underside of membrane were fixed in 4% formaldehyde and stained with Hoechst 33342 dye (Sigma-Aldrich Corp.). Migrated cells were counted using a fluorescence microscope. Five contiguous fields of each sample were examined using a 20 objective to obtain a representative number of cells that had invaded across the matrigel. Soft agarose assays Following transduction, MESO296 and MESO428 cells were selected for stable expression of AXL shRNA using 2 mg/ml puromycin. Mesothelioma cells (5 103) were resuspended in 1.5 ml of 0.3% agarose in complete media for AXL shRNA knockdown, or with the addition of DP-3975 (1 mM) or BEZ235 (0.5 mM) for biochemical inhibitor evaluations. This agarose suspension was then layered over 3 ml of 0.6% agarose in complete media in six-well plates. Every 4 days, 350 ml of fresh media, alone or containing DP-3975 or BEZ235, were added to each of the wells. The cells were incubated for 3 weeks, followed by staining with 1 ml of 0.005% crystal violet for 1 h. Colonies were counted manually with a Nikon Eclipse TS100 microscope. All experiments were performed in triplicate. Colonies were photographed using Spot software (version 3.5.9 for MacOS) and a Nikon Eclipse TE2000-5 microscope.

Conflict of interest Dr Flynn is an employee of Deciphera Pharmaceuticals, LLC and has an equity position in Deciphera Pharmaceuticals, LLC. Dr Wise and Dr Lu are employees of Deciphera Pharmaceuticals, LLC. The authors declare no conflict of interest.

Acknowledgements This research was supported by funding from the International Mesothelioma Program at Brigham and Women’s Hospital.

References Burchert A, Attar EC, McCloskey P, Fridell YW, Liu ET. (1998). Determinants for transformation induced by the Axl receptor tyrosine kinase. Oncogene 16: 3177–3187. Carbone M, Pass HI, Rizzo P, Marinetti M, Di Muzio M, Mew DJ et al. (1994). Simian virus 40-like DNA sequences in human pleural mesothelioma. Oncogene 9: 1781–1790. Catalano A, Romano M, Martinotti S, Procopio A. (2002). Enhanced expression of vascular endothelial growth factor (VEGF) plays a critical role in the tumor progression potential induced by simian virus 40 large T antigen. Oncogene 21: 2896–2900.

Craighead JE, Mossman BT. (1982). The pathogenesis of asbestosassociated diseases. N Engl J Med 306: 1446–1455. Craven RJ, Xu LH, Weiner TM, Fridell YW, Dent GA, Srivastava S et al. (1995). Receptor tyrosine kinases expressed in metastatic colon cancer. Int J Cancer 60: 791–797. Demetri GD, Zenzie BW, Rheinwald JG, Griffin JD. (1989). Expression of colony-stimulating factor genes by normal human mesothelial cells and human malignant mesothelioma cells lines in vitro. Blood 74: 940–946. Faber AC, Li D, Song Y, Liang MC, Yeap BY, Bronson RT et al. (2009). Differential induction of apoptosis in HER2 and EGFR Oncogene


1652 addicted cancers following PI3K inhibition. Proc Natl Acad Sci USA 106: 19503–19508. Gschwind A, Fischer OM, Ullrich A. (2004). The discovery of receptor tyrosine kinases: targets for cancer therapy. Nat Rev Cancer 4: 361–370. Harir N, Boudot C, Friedbichler K, Sonneck K, Kondo R, MartinLanneree S et al. (2008). Oncogenic Kit controls neoplastic mast cell growth through a Stat5/PI3-kinase signaling cascade. Blood 112: 2463–2473. Heinrich MC, Corless CL, Duensing A, McGreevey L, Chen CJ, Joseph N et al. (2003). PDGFRA activating mutations in gastrointestinal stromal tumors. Science 299: 708–710. Holland SJ, Powell MJ, Franci C, Chan EW, Friera AM, Atchison RE et al. (2005). Multiple roles for the receptor tyrosine kinase axl in tumor formation. Cancer Res 65: 9294–9303. Ito T, Ito M, Naito S, Ohtsuru A, Nagayama Y, Kanematsu T et al. (1999). Expression of the Axl receptor tyrosine kinase in human thyroid carcinoma. Thyroid 9: 563–567. Jacob AN, Kalapurakal J, Davidson WR, Kandpal G, Dunson N, Prashar Y et al. (1999). A receptor tyrosine kinase, UFO/Axl, and other genes isolated by a modified differential display PCR are overexpressed in metastatic prostatic carcinoma cell line DU145. Cancer Detect Prev 23: 325–332. Jagadeeswaran R, Ma PC, Seiwert TY, Jagadeeswaran S, Zumba O, Nallasura V et al. (2006). Functional analysis of c-Met/hepatocyte growth factor pathway in malignant pleural mesothelioma. Cancer Res 66: 352–361. Janne PA, Taffaro ML, Salgia R, Johnson BE. (2002). Inhibition of epidermal growth factor receptor signaling in malignant pleural mesothelioma. Cancer Res 62: 5242–5247. Janssen JW, Schulz AS, Steenvoorden AC, Schmidberger M, Strehl S, Ambros PF et al. (1991). A novel putative tyrosine kinase receptor with oncogenic potential. Oncogene 6: 2113–2120. Klominek J, Baskin B, Liu Z, Hauzenberger D.. (1998). Hepatocyte growth factor/scatter factor stimulates chemotaxis and growth of malignant mesothelioma cells through c-met receptor. Int J Cancer 76: 240–249. Lay JD, Hong CC, Huang JS, Yang YY, Pao CY, Liu CH et al. (2007). Sulfasalazine suppresses drug resistance and invasiveness of lung adenocarcinoma cells expressing AXL. Cancer Res 67: 3878–3887. Lee TC, Zhang Y, Aston C, Hintz R, Jagirdar J, Perle MA et al. (1993). Normal human mesothelial cells and mesothelioma cell lines express insulin-like growth factor I and associated molecules. Cancer Res 53: 2858–2864. Li Y, Ye X, Tan C, Hongo JA, Zha J, Liu J et al. (2009). Axl as a potential therapeutic target in cancer: role of Axl in tumor growth, metastasis and angiogenesis. Oncogene 28: 3442–3455. Mahadevan D, Cooke L, Riley C, Swart R, Simons B, Della CK et al. (2007). A novel tyrosine kinase switch is a mechanism of imatinib resistance in gastrointestinal stromal tumors. Oncogene 26: 3909–3919. Meric F, Lee WP, Sahin A, Zhang H, Kung HJ, Hung MC. (2002). Expression profile of tyrosine kinases in breast cancer. Clin Cancer Res 8: 361–367. Mukohara T, Civiello G, Davis IJ, Taffaro ML, Christensen J, Fisher DE et al. (2005). Inhibition of the met receptor in mesothelioma. Clin Cancer Res 11: 8122–8130. O’Bryan JP, Frye RA, Cogswell PC, Neubauer A, Kitch B, Prokop C et al. (1991). axl, a transforming gene isolated from primary human myeloid leukemia cells, encodes a novel receptor tyrosine kinase. Mol Cell Biol 11: 5016–5031. Ohta Y, Shridhar V, Bright RK, Kalemkerian GP, Du W, Carbone M et al. (1999). VEGF and VEGF type C play an important role in angiogenesis and lymphangiogenesis in human malignant mesothelioma tumours. Br J Cancer 81: 54–61.

Pache JC, Janssen YM, Walsh ES, Quinlan TR, Zanella CL, Low RB et al. (1998). Increased epidermal growth factor-receptor protein in a human mesothelial cell line in response to long asbestos fibers. Am J Pathol 152: 333–340. Pass HI, Liu Z, Wali A, Bueno R, Land S, Lott D et al. (2004). Gene expression profiles predict survival and progression of pleural mesothelioma. Clin Cancer Res 10: 849–859. Pass HI, Mew DJ, Carbone M, Matthews WA, Donington JS, Baserga R et al. (1996). Inhibition of hamster mesothelioma tumorigenesis by an antisense expression plasmid to the insulin-like growth factor1 receptor. Cancer Res 56: 4044–4048. Quong RY, Bickford ST, Ing YL, Terman B, Herlyn M, Lassam NJ. (1994). Protein kinases in normal and transformed melanocytes. Melanoma Res 4: 313–319. Rubin BP, Singer S, Tsao C, Duensing A, Lux ML, Ruiz R et al. (2001). KIT activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res 61: 8118–8121. Shaw RJ, Paez JG, Curto M, Yaktine A, Pruitt WM, Saotome I et al. (2001). The Nf2 tumor suppressor, merlin, functions in Rac-dependent signaling. Dev Cell 1: 63–72. Shieh YS, Lai CY, Kao YR, Shiah SG, Chu YW, Lee HS et al. (2005). Expression of axl in lung adenocarcinoma and correlation with tumor progression. Neoplasia 7: 1058–1064. Sordella R, Bell DW, Haber DA, Settleman J. (2004). Gefitinibsensitizing EGFR mutations in lung cancer activate anti-apoptotic pathways. Science 305: 1163–1167. Stitt TN, Conn G, Gore M, Lai C, Bruno J, Radziejewski C et al. (1995). The anticoagulation factor protein S and its relative, Gas6, are ligands for the Tyro 3/Axl family of receptor tyrosine kinases. Cell 80: 661–670. Stommel JM, Kimmelman AC, Ying H, Nabioullin R, Ponugoti AH, Wiedemeyer R et al. (2007). Coactivation of receptor tyrosine kinases affects the response of tumor cells to targeted therapies. Science 318: 287–290. Strizzi L, Catalano A, Vianale G, Orecchia S, Casalini A, Tassi G et al. (2001). Vascular endothelial growth factor is an autocrine growth factor in human malignant mesothelioma. J Pathol 193: 468–475. Tai KY, Shieh YS, Lee CS, Shiah SG, Wu CW. (2008). Axl promotes cell invasion by inducing MMP-9 activity through activation of NF-kappaB and Brg-1. Oncogene 27: 4044–4055. Tolnay E, Kuhnen C, Wiethege T, Konig JE, Voss B, Muller KM. (1998). Hepatocyte growth factor/scatter factor and its receptor cMet are overexpressed and associated with an increased microvessel density in malignant pleural mesothelioma. J Cancer Res Clin Oncol 124: 291–296. Tsao AS, He D, Saigal B, Liu S, Lee JJ, Bakkannagari S et al. (2007). Inhibition of c-Src expression and activation in malignant pleural mesothelioma tissues leads to apoptosis, cell cycle arrest, and decreased migration and invasion. Mol Cancer Ther 6: 1962–1972. Varnum BC, Young C, Elliott G, Garcia A, Bartley TD, Fridell YW et al. (1995). Axl receptor tyrosine kinase stimulated by the vitamin K-dependent protein encoded by growth-arrest-specific gene 6. Nature 373: 623–626. Wimmel A, Glitz D, Kraus A, Roeder J, Schuermann M. (2001). Axl receptor tyrosine kinase expression in human lung cancer cell lines correlates with cellular adhesion. Eur J Cancer 37: 2264–2274. Xia G, Kumar SR, Masood R, Koss M, Templeman C, Quinn D et al. (2005). Up-regulation of EphB4 in mesothelioma and its biological significance. Clin Cancer Res 11: 4305–4315. Yang MH, Chang SY, Chiou SH, Liu CJ, Chi CW, Chen PM et al. (2007). Overexpression of NBS1 induces epithelial-mesenchymal transition and co-expression of NBS1 and Snail predicts metastasis of head and neck cancer. Oncogene 26: 1459–1467.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene