Modulation of the C-Met/Hepatocyte Growth Factor Pathway in Small Cell Lung Cancer1

620 Vol. 8, 620–627, February 2002 Clinical Cancer Research

Modulation of the c-Met/Hepatocyte Growth Factor Pathway in Small Cell Lung Cancer1

Gautam Maulik, Takashi Kijima, Patrick C. Ma, ing. We also demonstrate that the Hsp90 inhibitor geldana- Sudip K. Ghosh, Jeffrey Lin, Geoffrey I. Shapiro, mycin, which also affects c-Met, reduced the growth and Erik Schaefer, Elena Tibaldi, Bruce E. Johnson, viability of four of four SCLC cell lines by 25% to 85%, over 2 a 72-h time period. Geldanamycin caused apoptosis of SCLC and Ravi Salgia cells, as well as led to increased levels of Hsp70 but not Departments of Adult Oncology [G. M., T. K., P. C. M., J. L., G. I. S., Hsp90. These results demonstrate that c-Met/HGF pathway B. E. J., R. S.] and Immunobiology [E. T.], Dana-Farber Cancer is functional in SCLC, and it would be useful to target this Institute, and Department of Medicine, Brigham and Women’s Hospital, and Harvard Medical School, Boston, Massachusetts 02115; pathway toward novel therapy. Biosource International, Hopkinton, Massachusetts 01748 [E. S.]; and Department of Biotechnology, Indian Institute of Technology, INTRODUCTION Kharagpur, West Bengal 721302, India [S. K. G.] SCLC3 is an aggressive cancer that tends to metastasize relatively quickly. Even with chemotherapeutics available for ABSTRACT the past 30 years, long-term survival rates have been quite The c-Met receptor tyrosine kinase and its ligand HGF dismal (1). New and novel therapeutics are being tested in (hepatocyte growth factor) have been shown to be involved patients with SCLC to arrive at better control of disease with the in angiogenesis, cellular motility, growth, invasion, and dif- goal of longer survival. Recently, tyrosine kinase receptors have ferentiation. The role of c-Met/HGF axis in small cell lung been implicated in the etiology of multiple tumors, and they may cancer (SCLC) has not been reported previously. We have be important therapeutic targets (2). We have reported recently determined the expression of p170c-Met precursor and that the tyrosine kinase c-Kit is an important RTK in SCLC that p140c-Met ␤-chain in seven SCLC cell lines by immunoblot- is targeted by the tyrosine kinase inhibitor STI 571 (3). This has ting. We used the SCLC cell line H69, which expressed an been applied clinically, and STI 571 has been introduced into a abundant amount of c-Met to study the function and down- Phase II clinical trial for patients with previously untreated stream effects of c-Met activation. Stimulation of H69 cells SCLC and patients with relapsed sensitive disease. with HGF (40 ng/ml, 6-h stimulation) significantly altered c-Kit tyrosine kinase activation by stem cell factor is only cell motility of the SCLC cells with increased formation of one of many mechanisms proposed for the pathogenesis of filopodia and membrane ruffling, characterized as mem- SCLC. As shown previously, c-Met RTK mRNA is also ex- brane blebbing, as well as increased migration of the cellular pressed in lung cancer cells (4). c-Met is an RTK, which clusters were seen. We have further studied the signal trans- stimulates the invasive growth of carcinoma cells, is tumori- duction pathways of HGF/c-Met in the H69 cell line. The genic, and overexpressed in many solid tumors (5). The c-Met ␣ ␤ stimulation of H69 with HGF (40 ng/ml, >24 h, maximal at receptor is a disulfide linked - heterodimer with a molecular ␤ 1 h) increased the amount of reactive oxygen species formed weight of Mr 190,000 (6). The Mr 140,000 -chain spans the by 34%. HGF stimulation (40 ng/ml, 7.5-min stimulation) of membrane and possesses cytoplasmic tyrosine kinase activity H69 cells showed increased tyrosine phosphorylated bands and can be detected in its precursor form at Mr 170,000. The natural ligand for c-Met is the HGF (also known as scatter identified at Mr 68,000, 120,000–140,000, and 200,000. Some of these tyrosine-phosphorylated bands were identified as factor; Refs. 7 and 8). HGF stimulation of c-Met can lead to the focal adhesion proteins paxillin, FAK, PYK2, and the proliferation, increased survival, altered motility, enhanced in- c-Met receptor itself. Phospho-specific antibodies show that vasion into extracellular matrix, and more rapid formation of tyrosines at amino acid (a.a.) 31 of paxillin, and autophos- tubules (9). c-Met overexpression, as well as activating muta- phorylation sites at a.a. 397 of p125FAK, and a.a. 402 of tions in the various domains, can lead to carcinogenesis in PYK2 are phosphorylated in response to HGF/c-Met signal- multiple tumors. There are multiple activating mutations in the c-Met gene identified in hereditary papillary renal carcinomas (10). c-Met, on activation by autophosphorylation, can associate with and activate multiple signal transducing intermediates, such as Grb2, p85 subunit of PI3k, Stat-3, and Gab1 (9). Received 8/14/01; revised 11/9/01; accepted 11/12/01. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 3 The abbreviations used are: SCLC, small cell lung cancer; RTK, 1 Supported by NIH Grant 75348-04 and Lowe’s Center for Thoracic receptor tyrosine kinase; HGF, hepatocyte growth factor; ROS, reactive Oncology (to R. S.). oxygen species; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra- 2 To whom requests for reprints should be addressed, at Department of zolium bromide; TUNEL, terminal deoxynucleotidyl transferase-medi- Adult Oncology, Dana 1234B, Dana-Farber Cancer Institute, 44 Binney ated dUTP nick-end labeling; DCF-DA, dichloro-fluorescein-diacetate; Street, Boston, MA 02115. Phone: (617) 632-4389; Fax: (617) 632- DMTU, 1,3-dimethythiourea; PI3k, phosphatidylinositol 3Ј-kinase; 4379; E-mail: [email protected]. WCL, whole cell lysate; IP, immunoprecipitated lysates.

In lung cancer, c-Met mRNA levels can be elevated. As an both known CRKL SH2 domain binding sites, or tyrosine 181 in example, in a study of 25 SCLC cell lines, it was shown that 22 paxillin; tyrosine 397 (autophosphorylation site) and tyro- of 25 cell lines express c-Met mRNA and only 2 of 25 cell lines sine861 (Src phosphorylation Site) in p125FAK; and tyrosine coexpressed HGF mRNA (4). c-Met appears to be activated in 402 (autophosphorylation site) and 881 (Grb2 binding site) in a paracrine fashion, with HGF being produced by the stromal PYK2 were used. cells. c-Met has shown to be expressed in the non-small cell Preparation of Cell Lysates and Immunoblotting. Cell lung cancer cell line A431, and activation by HGF stimulates lines were lysed in lysis buffer as described previously (15). Cell cell growth, scattering, and invasion of these cells (4). HGF lysates were separated by 7.5% SDS-PAGE under reducing levels in tumor tissue have been directly associated with short- conditions, electrophoretically transferred to pure nitrocellulose ened survival in patients with non-small cell lung cancer (11). In transfer and immobilization membrane (Schleicher & Schuell, an effort to determine the role of c-Met in SCLC, we show by Keene, NH), and processed for immunoblotting using estab- immunoblotting that c-Met protein is expressed and phospho- lished methods with enhanced chemiluminescense technique rylated in SCLC, as well as functional through the cytoskeleton. (NEN Life Science Products, Boston, MA). Also immunopre- Patterns of phosphoregulation of individual tyrosine residues cipitations were performed according to standard procedures were determined for cytoskeletal proteins in response to HGF in and immunoblotting thereafter (16). SCLC, including p125FAK, PYK2, and paxillin. We further Analysis of ROS. A total of 106 cells were treated with show that targeting of c-Met by geldanamycin reduced phos- or without 5 ␮M DCF-DA (2Ј,7Ј-dichlorofluoroscein-diacetate; phorylation of signaling proteins, inhibited growth and viability Acros Organics, Pittsburgh, PA) for 7.5 min at 37°C and sub- in SCLC cells, and altered the levels of the molecular chaperone sequently washed twice in cold Dulbecco’s PBS before analysis Hsp70. using a Coulter Epics XL flow cytometer (Coulter Corp., Mi- ami, FL). The fluorescence of oxidized DCF was measured with an excitation wavelength of 480 nm and an emission wavelength MATERIALS AND METHODS of 525 nm (17). Cell Lines and Cell Culture. The SCLC cell lines were Cell Cycle Analysis and Apoptosis Detection by Flow obtained from the American Type Culture Collection (Rock- Cytometry. SCLC cells were treated at 37°C with either ville, MD). The SCLC cell lines were maintained in RPMI 1640 DMSO or geldanamycin at concentration of 100 nM and ana- (Cellgro), 10% (volume for volume) FCS, and 1% (volume for lyzed after propidium iodide staining using standard methods volume) penicillin-streptomycin (12). All cell lines were incu- (3). Apoptosis detection by flow cytometry was performed using bated at 37°C with 5% CO2. The cell lines were harvested the Apopotosis Detection System, Fluorescein (Promega, Mad- during log phase growth. For stimulation studies with HGF, ison, WI), based on the TUNEL assay, as per the manufacturer’s cells were deprived of growth factors by incubating the cells directions. without FCS for 16 h, and the media comprised of 0.5% BSA Time-lapse Video Microscopy. Serum-starved cells (Sigma Chemical Co., St. Louis, MO). The SCLC cell lines that were placed in uncoated plastic tissue culture plates (35 ϫ 10 expressed c-Met were tested for protein tyrosine phosphoryla- mm plates; Becton Dickinson Labware), and after6hofobser- tion in response to HGF at the dose of 40 ng/ml for 7.5 min. This vation, they were treated with 40 ng/ml HGF in a temperature- short-time incubation was shown to be the time for maximal controlled chamber at 37°C in standard media. The cells were phosphorylation in a kinetic analysis (data not shown). H69 cells examined by video microscopy using Olympus IX70 inverted were incubated with 40 ng/ml HGF for Յ6 h to observe the microscope, omega temperature controlled device, Optronics formation of filopodia, change in membrane ruffling, and mi- Engineering DEI-750 video camera, Olympus OEV142 TV, and grational movement of the clusters by time-lapse video micros- Panasonic AG6740 time-lapse S-VHS video recorder continu- copy. For ROS levels, we determined over 24 h since late ously. The digital video images were then processed with an oxidative species can form during this time. For ROS, H69 cells Apple computer containing a G4 microprocessor and analyzed were treated with or without HGF (40 ng/ml). with the NIH Image Analysis program. MTT Assays. Cell viability was measured by MTT col- Confocal Microscopy. F-actin was visualized in fixed orimetric dye reduction assay (Sigma Chemical Co.; Ref. 13). cells (18). FITC-phalloidin (Sigma Chemical Co.) was used to Each data point was repeated in independent experiments and determine F-actin. Confocal image analysis was performed us- performed in quadruplicates, and SDs were calculated. Geldana- ing a Leica NT-TCS confocal microscope fitted with argon and mycin was obtained from the National Cancer Institute and used krypton lasers. to determine the effects on SCLC cell growth and viability by the MTT assay (14). Antibodies. The antiphosphotyrosine monoclonal anti- RESULTS body 4G10 was obtained from UBI (Lake Placid, NY). Anti-c- Expression of c-Met in SCLC Cell Lines. A polyclonal Met antibody (C-12), anti-Hsp90 antibody, anti-Hsp70 anti- antibody against c-Met was used to detect the p170c-Met precur- body, anti-p125FAK antibody, and anti-PYK2 antibody were sor expression, as well as the p140c-Met ␤-chain (Fig. 1A). In the obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) H69 and H345 SCLC cell lines, the p170 precursor and p140 and were used per manufacturer’s directions. Antipaxillin anti- forms were strongly detected. There was moderate expression of body (clone 5H11) was used as described previously (15). both in H128 and H146 SCLC cell lines, and in H82, H209, and Polyclonal phosphorylation site-specific antibodies (Biosource H446 SCLC cell lines, only p140c-Met was detected by immu- International, Camarillo, CA) to tyrosine 31 or tyrosine 118, noblotting. In H82 and H128 cell lines, p140c-Met expression

Fig. 1 Expression of c-Met in SCLC cell lines and tyrosine phosphorylation of proteins in response to HGF. A, expression of c-Met in SCLC cell lines. Cell lysates for SCLC cells were processed as described in “Materials and Methods.” Lysates were applied on 7.5% SDS-PAGE gel, transferred to Immobilon-P membrane, and probed with anti-c-Met antibody. Note the p140c-Met ␤-chain, as well as the p170c-Met pro-form. As an internal control, the membrane was stripped and probed with ␤-actin. In B, HGF stimulated tyrosine phosphorylation of proteins in SCLC cell lines. Shown are four separate SCLC cell line (H209, H69, H82, and H146) lysates run on SDS-PAGE gel, transferred to membrane, and immunoblotted with antiphosphotyrosine with (ϩ) or without (Ϫ) stimulation with HGF (40 ng/ml, 7.5 min). In the bottom panel, the membrane was stripped and reprobed with anti-c-Met antibody. The p140 and p170 forms of c-Met are identified.

Fig. 2 c-Met is tyrosine phosphorylated in response to HGF in H69 appeared as a doublet, and H209 and H446 expressed a slightly cells. A, c-Met tyrosine phosphorylation. Cell lysates of H69 (with high larger form of p140c-Met. Four of the SCLC cell lines were expression of c-Met) and H209 (with low expression of c-Met) treated tested for tyrosine phosphorylation of proteins in response to with or without HGF (7.5 min, 40 ng/ml) were immunoprecipitated (IP) with anti-c-Met, and the WCLs and IPs were immunoblotted with HGF (40 ng/ml; Fig. 1B). There was increased tyrosine phos- antiphosphotyrosine. In B, the membrane in A was stripped and immu- phorylation of multiple proteins in three of four cell lines used, noblotted with anti-c-Met as a control. In C, cell lysates of H69 and with the most dramatic tyrosine phosphorylation change in H69 H209 were immunoprecipitated with antiphosphotyrosine, and the cell line. Increased tyrosine phosphorylation was visualized at WCLs and IPs were immunoblotted with anti-c-Met.

Mr 200,000, 170,000, 140,000, 120,000, 90,000, 70,000, 60,000, and 40,000 in response to HGF. c-Met Tyrosine Phosphorylation in Response to HGF in H69 SCLC Cell Line. Because there was a robust expression sufficient expression of c-Met to determine the impact of adding of c-Met in H69 SCLC cell line, as compared with very low HGF to other functions in SCLC. expression of c-Met in H209 cell line, we determined tyrosine As shown previously (3), H69 SCLC cells characteristi- phosphorylation of c-Met in response to HGF in these lines. As cally move together as a cluster and display membrane ruffling, seen in Fig. 2A, with anti-c-Met immunoprecipitation and phos- characterized by membrane blebs (Fig. 3A). When SCLC cells photyrosine immunoblot, there was increased tyrosine phospho- are separated from the cluster, they either unite with another rylation of c-Met in response to HGF (7.5 min, 40 ng/ml). The group of cells or, if isolated, they undergo apoptosis (data not amount of c-Met immunoprecipitation is the same with or shown). With HGF treatment (40 ng/ml) of the SCLC, the cells without HGF (Fig. 2B). In converse experiment, phosphoty- had dramatically increased membrane ruffling and formation of rosine immunoprecipitation and anti-c-Met immunoblot re- filopodia, as well as increased migrational movement of the vealed the same result (Fig. 2C). cellular clusters (Fig. 3B). As calculated by NIH Image Analysis Altered Cell Motility in H69 SCLC Cell Line. The program, the average size of filopodia formed with HGF stim- above data show that c-Met is functionally active in SCLC. We ulation was 4 Ϯ 1.2 (SD) ␮m. have used the H69 cell line as a model, because there was F-actin staining of SCLC cells treated with HGF also

Fig. 4 F-actin staining of H69 SCLC cells in response to HGF. Serum- starved SCLC cells were visualized by confocal microscopy without (A and B) or with (C and D) HGF stimulation. Shown are the F-actin staining patterns (A and C) and the confocal images (B and D). Note the F-actin staining of filopodia in C with stimulation by HGF (40 ng/ml, 7.5 min, arrows).

cytoskeletal functions. As shown in Fig. 6, p125FAK was phos- Fig. 3 Time-lapse video microscopy of H69 SCLC cells in response to phorylated on tyrosines 397 (autophosphorylation site) and 861 HGF. Serum-starved SCLC cells were visualized by time-lapse video (the major Src phosphorylation site), in response to HGF. Sim- microscopy and recorded for 0–12 h without (A) or with (B) HGF. In the ilarly, PYK2 was phosphorylated on tyrosines 402 (autophos- absence of HGF (A), SCLC cells move in clusters and are characterized as having membrane ruffles (“blebs”). In the presence of HGF (40 phorylation site) and 881 (Grb2 binding site) in response to ng/ml), SCLC cells had increased filopodia formation (B, arrows), HGF. Finally, HGF treatment resulted in higher migrating forms movement of cellular cluster, and membrane ruffling. Each numbered of paxillin coinciding with phosphorylation at tyrosine 31 (the picture represents an approximate of 2-h interval. first CRKL binding site) but not tyrosines 118 or 181. Geldanamycin Inhibits Cell Growth of SCLC Cells. Geldanamycin is an antibiotic with multiple effects on tumor shows actin-containing filopodia (Fig. 4) by confocal micros- cells and has been shown to disrupt the c-Met/HGF axis (14). copy, as seen similarly by time-lapse video microscopy. We have determined the effects of geldanamycin on growth and HGF Leads to an Increase in ROS. We analyzed the viability, cell cycle, cell motility, and signal transduction of endogenous levels of ROS in H69 cell line, with and without SCLC cells. treatment of HGF (40 ng/ml) over 24 h (Fig. 5) using 2Ј,7Ј- Cell viability curves in Fig. 7 show that over 72 h of DCF-DA by flow-cytometry. The relative ROS levels were treatment with geldanamycin (Յ200 nM), there was decreased increased in response to HGF by ϳ30%. DCF-DA can measure viability seen in the four SCLC cell lines tested (H69, H82, oxidizing agents, such as hydrogen peroxide, superoxide anion, H146, and H249). The calculated IC50s are: 89 nM for H69, 54 and hydroxyl radicals (17). Using DMTU, which is a potent nM for H146, and 86 nM for H249 cell lines. At the concentra- scavenger of hydroxyl radicals, the ROS levels decreased in tions used, 50% decrease in cell viability was not seen in H82 response to HGF. Without HGF treatment and DMTU, there cell line. was no further decrease in ROS levels. Using the H69 cell line, cell cycle analysis was performed HGF Stimulation of SCLC Cells Leads to Phosphoryl- (Fig. 8, A–D) with or without HGF and geldanamycin. In ation of Multiple Cytoskeletal Proteins. HGF activates cy- response to 100 nM geldanamycin, there was an increase (from toskeletal proteins leading to altered cell motility. We looked for 14 to 34%, no HGF treatment; from 15 to 32%, with HGF activation of specific pathways known to be involved in the treatment) in the sub-G1 population. The sub-G1 population regulation of the focal adhesion. Phosphorylation of cytoskeletal corresponded to apoptotic events because an independent proteins p125FAK, PYK2, and paxillin was determined by TUNEL assay with geldanamycin of H69 cells confirmed this using a panel of phosphorylation site-specific antibodies. We finding (Fig. 8E). We also detected decreased cell migration, determined the activation of specific phosphoylation sites in formation of filopodia, and formation of membrane ruffles these proteins, known to be important in the regulation of within 2–4 h of geldanamycin treatment (data not shown). HGF

Fig. 6 Phosphorylation of p125FAK, PYK2, and paxillin on specific amino acid residues on the stimulation of HGF in H69 cells. Shown are specific phosphorylations of p125FAK on tyrosines 397 and 861, PYK2 on tyrosines 402 and 881, and paxillin on tyrosine 31. Shown are WCLs of H69 cells with (ϩ) or without (Ϫ) stimulation with HGF (7.5 min, 40 ng/ml) separated on SDS-PAGE gel, applied to membrane, and immunoblotted with various antibodies: A, expression and specific phosphorylation of p125FAK, anti-p125FAK, and anti-phospho-FAK (tyrosine residues 397 or 861); B, expression and specific phosphorylation of Fig. 5 Kinetics of up-regulation of ROS levels in H69 SCLC cells with PYK2, anti-PYK2, and anti-phospho-PYK2 (tyrosine residues 402 or HGF treatment. A, kinetics of ROS production with HGF. ROS levels 881); C, expression and specific phosphorylation of paxillin, antipaxillin were measured with flow cytometry using DCF-DA and plotted as a clone 5H11, and anti-phospho-paxillin (tyrosine residues 31, 118, percentage increase as compared with baseline of no treatment with or 181). HGF. Error bars, SD from at least three separate experiments. B, fluorescence-activated cell sorter analysis of ROS levels with and without HGF. Graph measuring ROS formation by flow cytometry without (darkened image) or with (lightened image) HGF. C, ROS inhibition with DMTU scavenger. ROS formation, as measured by flow cytometry, figure with (darkened image) or without (lightened image) DMTU. Shown is also inhibition of ROS formation in response to the hydroxyl scavenger DMTU.

stimulation of cells did not enhance or abrogate the effects of geldanamycin. Geldanamycin Regulates c-Met Function and Expres- sion. We next investigated the effect of geldanamycin on downstream tyrosine phosphorylation after the addition of HGF (Fig. 9). In H69 cells treated with geldanamycin, there was a Fig. 7 Geldanamycin decreases viability for SCLC cell lines. Geldana- decrease in tyrosine phosphorylation of proteins (at Mr 220,000, mycin-treated (0–250 nM) SCLC cell lines (H69, H82, H146, and H249) 110,000–140,000, and 90,000). Additionally, although the were evaluated for viability by the MTT assay. Data are plotted in amount of protein on the gel is similar (as assessed by the reference to the DMSO control (0 nM) over a period of 72 h. Error bars, control PI3k immunoblot), there was a decreased amount of SDs of four separate experiments. c-Met in response to geldanamycin treatment. Because geldanamycin directly inhibits the function of the molecular chaperone Hsp90 (19), the effects on the levels of c-Met with Hsp90 and have documented that HGF/c-Met pathway is functional in its related chaperone Hsp70 (Fig. 10) were examined. In dose- SCLC, and this may be a useful target for therapeutic interven- Յ ␮ responsive fashion ( 1 M of geldanamycin), there were in- tion in SCLC. We have shown in this study that c-Met was creased levels of Hsp70 and not Hsp90. By cross-immunopre- expressed in SCLC cell lines; downstream tyrosine phosphoryl- cipitation studies, we did not appreciate any association between ation of proteins (with Mr 200,000, 170,000, 140,000, 120,000, Hsp70 or Hsp90 with c-Met itself (data not shown). 90,000, 70,000, 60,000, and 40,000) occurred in a short time interval, as well as in a dose-responsive fashion (data not DISCUSSION shown) in response to HGF. c-Met expression was heterogene- The studies of RTKs in solid tumors have exploded re- ous in the various different SCLC cell lines, and it is possible cently because of the potential utility of these molecules as there may be mutations of the c-Met gene or overamplification molecularly targeted therapies (2). In studies shown here, we of the gene in certain SCLC cell lines. This possibility is

Fig. 9 HGF-stimulated tyrosine phosphorylation in SCLC can be inhibited by treatment with geldanamycin. A, expression of tyrosine phophorylated proteins. Antiphosphotyrosine immunoblot of control or geldanamycin-treated cells. There were decreased tyrosine phosphoryl-

ation bands (Mr 120,000–150,000) in response to HGF (7.5 min, 40 ng/ml) in the geldanamycin-treated cells (100 nM,24h).B, c-Met expression. The same membrane was stripped and probed for anti-c- Met. c-Met expression appears to be decreased in the geldanamycin- treated lanes. C, PI3k expression. Control PI3k immunoblot was done of the same membrane, showing similar loading of proteins.

Fig. 8 Cell cycle analysis and apoptosis assay of H69 cells with geldanamycin treatment (A–D). Untreated (A and C) or HGF-treated (B and D; 24 h, 40 ng/ml) H69 SCLC cells with (C and D;24h,100nM) or without (A and B) geldanamycin. SCLC cells were fixed and stained with propidium iodide for flow cytometry. E, apoptosis detection using the TUNEL assay shown without or with geldanamycin (24 h, 100 nM).

currently being investigated. Stimulation of c-Met by HGF leads to increased formation of ROS and increased cell motility. This is of special interest because stromal cells in the lung are known to produce HGF and thereby have the potential to activate c-Met in lung cancer cells. We have shown here that cell motility of SCLC is enhanced by the addition of HGF. The mechanism whereby HGF stimulation of c-Met leads to increased motility, migration, and invasion in cancer cells has not been well worked out. c-Met Fig. 10 Geldanamycin treatment alters the levels of Hsp70 and not stimulation promotes cell movement, causes epithelial cells to Hsp90. H69 cells were treated with geldanamycin with various concen- disperse (“scatter”) and endothelial cells to migrate, and pro- trations (0, 100, and 1000 nM) for 24 h in starved medium. The lysates motes chemotaxis (20–23). Invasion is also important in c-Met were prepared according to “Materials and Methods.” Immunoblots of the same membrane were performed in a sequential fashion, using signaling because mutant mice nullizygous for Met show that anti-c-Met, anti-Hsp90, anti-Hsp70, and anti-␤-actin antibodies. muscles originating from dermomyotome cells that migrate to the limb, diaphragm, and tip of the tongue fail to develop (24). In this study, c-Met stimulation as reflected by tyrosine phosphorylation by HGF leads to increased migration, membrane that activation of the c-Met signaling pathway contributes to the ruffling, and filopodia formation in SCLC cells. It would now be metastatic mechanism of SCLC. interesting to determine which phosphorylation sites of c-Met Cell motility is also regulated by the focal adhesion. In are activated by HGF in SCLC. Therefore, it is quite possible suspension cells, such as SCLC, the focal adhesions are re-

flected as punctuate focal contacts (25). We have shown that that ROS are generated by HGF, maximally within 60 min of certain components of the focal adhesion, such as paxillin, treatment. ROS have been shown to contribute to cellular func- p125FAK, and PYK2, are phosphorylated at specific sites in tions, and we have recently demonstrated that ROS can induce response to HGF. Induction of phosphorylation of FAK and cell cycle progression, increase cell migration, inhibit protein PYK2 at their autophosphorylation sites indicates that the initial tyrosine phosphatases, and lead to increase phosphorylation of activation step has occurred. The fact that the Grb2 binding site, cellular proteins (17). It would now be interesting to study the on PYk2, becomes phosphorylated in response to HGF suggests expression and activity of ROS-regulating enzymes by HGF in a means to differentially activate the Ras/mitogen-activated SCLC. protein kinase pathway. Interestingly, HGF has been known for HGF also leads to plasmin activation, and Webb et al. (14) some time to activate calcium flux in a dose-dependent manner. have reported that geldanamycins lead to decreased plasmin One of the distinguishing features between the focal adhesion activation at femtomolar concentrations. Using the geldanamy- kinases, FAK and PYK2, is the ability to activate PYK2 with cins at nM concentrations led to down-regulation of c-Met, calcium. Finally, HGF treatment resulted in higher migrating inhibition of HGF-mediated cell motility and invasion, and forms of paxillin, coinciding with phosphorylation at tyrosine 31 reversion of Met-transformed cells. Using geldanamycin in (the first CRKL binding site) but not tyrosines 118 or 181. SCLC cells responsive to HGF, we have further shown alter- Paxillin is a dynamic cytoskeletal protein that can form com- ation in cell cycle, apoptosis, decreased cell motility, and de- plexes with adhesion proteins, including FAK, PYK2, Src, Csk, creased tyrosine phosphorylation. Within 72 h of treatment with and p130Cas (26). Moreover, differential phosphorylation of geldanamycin, there was decreased viability and apoptosis, of paxillin appears to influence the temporospatial regulation of all four SCLC cell lines tested. focal adhesions and the actin cytoskeleton (27). Geldanamycin has been described as an inhibitor of Hsp90 The process of cell motility, as reflected by formation of function (19). Hsp90 functions as a chaperone, to assist proteins membrane ruffles and filopodia, as well as migration, is well to acquire mature conformation, and has been shown to associ- regulated by small GTPases and PI3k. The actin cytoskeleton ate with tyrosine kinases, such as v-Src, LCK, and p210BCR/ needs to be dynamic for cell shape change, alterations in cell ABL. In a recent study by An et al. (34), geldanamycin altered contact, and cell motility. Ras-like GTPases of the Rho family, the association of Hsp90 with p210BCR/ABL, thereby leading which includes Rho, Rac, and Cdc42, can reorganize actin. In to the degradation of the cellular tyrosine kinase. We show in Swiss 3T3 cells, Rho is involved in formation of actin stress this study that there was indeed decreased amounts of c-Met in fibers and focal adhesions, Rac is involved in formation of response to geldanamycin treatment. However, this may not be lamellipodia and membrane ruffles, and Cdc42 is involved in related to association with Hsp90, because cross-immunopre- formation of filopodia (28, 29). In the cascade of regulation, cipitation did not reveal any direct interaction between these two Cdc42 activation is followed by Rac activation, thereafter stim- molecules. Possibilities are that c-Met either may not be syn- ulating Rho. Specific inhibitors of Rho activity, such as C3 thesized as fast as nongeldanamycin-treated cells or that c-Met ADP-ribosyltransferase and Rho-guanine nucleotide dissocia- is being degraded via ubiquitination in response to geldanamy- tion inhibitor, interfere with Swiss 3T3 cell motility, and this cin. The striking observation that Hsp70 levels are increased by can be reversed by the addition of constitutively Rho. The Rho geldanamycin is unique and has not been reported previously. family members are clearly under tight regulation in the cell, but Hsp70 has been implicated to be important in apoptosis events the mechanisms of regulation are not understood in SCLC and as related to ROS (35), and the levels of Hsp70 may be increas- would be quite interesting to study further. We would hypoth- ing with geldanamycin to the oxidative stress response of the esize that Rac is important in the membrane ruffling (blebbing) cells undergoing apoptosis and needs to be further explored. It and Cdc42 is important in filopodia formation in SCLC stimu- is to be noted, although geldanamycin may reduce the expres- lated by HGF. It is also not understood if small GTPases or PI3k sion of c-Met, geldanamycin may also act through other mech- affect scattering. However, in the SCLC lines studied, we were anisms than c-Met inhibition alone. This possibility is raised not able to appreciate cell scattering (data not shown). This may because HGF did not restore the biological functions in geldana- be because SCLC are suspension cells and may behave differ- mycin-treated cells. It would now be interesting to determine ently than epithelial cells. which other pathways, including Hsp90 and Hsp70, would be As another biological function of HGF, we have shown involved in the biochemical and biological effects of geldana- Ϫ that ROS levels are up-regulated in SCLC. ROS, such as O2 . , mycin in SCLC. Additionally, there are analogues of geldana- .Ϫ . OH ,NO , and H2O2, have recently begun to be appreciated mycin available, and the effects on SCLC along with HGF/c- for their role in regulation of signal transduction, gene expres- Met should be further explored. Finally, an inhibitor specifically sion, proliferation, and motility (30, 31). SCLC is directly targeting HGF/c-Met would be quite useful for SCLC studies. caused by cigarette smoking, and it is well known that cigarette smoke can generate considerable toxic ROS (32). A delicate balance between oxidants and the protective effects of intra and REFERENCES extracellular oxidant defense system is abrogated in SCLC. In a 1. Chute, J. P., Chen, T., Feigal, E., Simon, R., and Johnson, B.E. recent study from Arakaki et al. (33), it was shown that the Twenty years of phase III trials for patients with extensive-stage small- cell lung cancer: perceptible progress. J. Clin. Oncol., 17: 1794–1801, N-acetylcysteine prevented HGF-suppressed growth of Sarcoma 1999. 180 and Meth A cells and HGF-induced apoptosis. In contrast to 2. Druker, B. J., and Lydon, N. B. Lessons learned from the develop- the study by Arakaki et al. (33), we did not observe altered ment of an abl tyrosine kinase inhibitor for chronic myelogenous leu- apoptosis on HGF stimulation (data not shown) and have found kemia. J. Clin. Investig., 105: 3–7, 2000.

3. Wang, W. L., Healy, M. E., Sattler, M., Verma, S., Lin, J., Maulik, 18. Salgia, R., Quackenbush, E., Lin, J., Souchkova, N., Sattler, M., G., Stiles, C. D., Griffin, J. D., Johnson, B. E., and Salgia, R. Growth Ewaniuk, D. S., Klucher, K. M., Daley, G. Q., Kraeft, S. K., Sackstein, inhibition and modulation of kinase pathways of small cell lung cancer R., Alyea, E. P., von Andrian, U. H., Chen, L. B., Gutierrez-Ramos, cell lines by the novel tyrosine kinase inhibitor STI 571. Oncogene, 19: J. C., Pendergast, A. M., and Griffin, J. D. The BCR/ABL oncogene 3521–3528, 2000. alters the chemotactic response to stromal-derived factor-1␣. Blood, 94: 4. Rygaard, K., Nakamura, T., and Spang-Thomsen, M. Expression of 4233–4246, 1999. the proto-oncogenes c-met and c-kit and their ligands, hepatocyte 19. Neckers, L., Schulte, T., and Mimnaugh, E. Geldanamycin as a growth factor/scatter factor and stem cell factor, in SCLC cell lines and potential anti-cancer agent: its molecular target and biochemical activ- xenografts. Br. J. Cancer, 67: 37–46, 1993. ity. Investig. New Drugs, 17: 361–373, 1999. 5. To, C. T., and Tsao, M. S. The roles of hepatocyte growth factor/ 20. Ponzetto, C., Bardelli, A., Zhen, Z., Maina, F., dalla Zonca, P., scatter factor and met receptor in human cancers (Review). Oncol. Rep, Giordano, S., Graziani, A. Panayotou, G., and Comoglio, P. M. A 5: 1013–1024, 1998. multifunctional docking site mediates signaling and transformation by the hepatocyte growth factor/scatter factor receptor family. Cell, 77: 6. Faletto, D. L., Tsarfaty, I., Kmiecik, T. E., Gonzatti, M., Suzuki, T., 261–271, 1994. and Vande Woude, G. F. Evidence for non-covalent clusters of the c-met proto-oncogene product. Oncogene, 7: 1149–1157, 1992. 21. Royal, I., and Park, M. Hepatocyte growth factor-induced scatter of Madin-Darby canine kidney cells requires phosphatidylinositol 3- 7. Iyer, A., Kmiecik, T. E., Park, M., Daar, I., Blair, D., Dunn, K. J., kinase. J. Biol. Chem., 270: 27780–27787, 1995. Sutrave, P., Ihle, J. N., Bodescot, M., and Vande Woude, G. F. Struc- 22. Royal, I., Fournier, T. M., and Park, M. Differential requirement of ture, tissue-specific expression, and transforming activity of the mouse Grb2 and PI3-kinase in HGF/SF-induced cell motility and tubulogen- met protooncogene. Cell Growth Differ., 1: 87–95, 1990. esis. J. Cell Physiol., 173: 196–201, 1997. 8. Rong, S., Bodescot, M., Blair, D., Dunn, J., Nakamura, T., Mizuno, 23. Weidner, K. M., Di Cesare, S., Sachs, M., Brinkmann, V., Behrens, K., Park, M., Chan, A., Aaronson, S., and Vande Woude, G. F. Tumor- J., and Birchmeier, W. Interaction between Gab1 and the c-Met receptor igenicity of the met proto-oncogene and the gene for hepatocyte growth tyrosine kinase is responsible for epithelial morphogenesis. Nature factor. Mol. Cell. Biol., 12: 5152–5158, 1992. (Lond.), 384: 173–176, 1996. 9. Stella, M. C., and Comoglio, P. M. HGF: a multifunctional growth 24. Bladt, F., Riethmacher, D., Isenmann, S., Aguzzi, A., and Birch- factor controlling cell scattering. Int. J. Biochem. Cell Biol., 31: 1357– meier, C. Essential role for the c-met receptor in the migration of 1362, 1999. myogenic precursor cells into the limb bud. Nature (Lond.), 376: 768– 10. Schmidt, L., Junker, K., Weirich, G., Glenn, G., Choyke, P., Luben- 771, 1995. sky, I., Zhuang, Z., Jeffers, M., Vande Woude, G., Neumann, H., 25. Nobes, C. D., and Hall, A. Rho GTPases control polarity, protru- Walther, M., Linehan, W. M., and Zbar, B. Two North American sion, and adhesion during cell movement. J. Cell Biol., 144: 1235–1244, families with hereditary papillary renal carcinoma and identical novel 1999. mutations in the MET proto-oncogene. Cancer Res., 58: 1719–1722, 1998. 26. Sattler, M., Verma, S., Shrikhande, G., Byrne, C. H., Pride, Y. B., Winkler, T., Greenfield, E. A., Salgia, R., and Griffin, J. D. The 11. Siegfried, J. M., Weissfeld, L. A., Luketich, J. D., Weyant, R. J., BCR/ABL tyrosine kinase induces production of reactive oxygen spe- Gubish, C. T., and Landreneau, R. J. The clinical significance of cies in hematopoietic cells. J. Biol. Chem., 275: 24273–24278, 2000. hepatocyte growth factor for non-small cell lung cancer. Ann. Thorac. Surg., 66: 1915–1918, 1998. 27. Sattler, M., Pisick, E., Morrison, P. T., and Salgia, R. Role of the cytoskeletal protein paxillin in oncogenesis. Crit. Rev. Oncog., 11: 12. Salgia, R., Li, J. L., Ewaniuk, D. S., Wang, Y. B., Sattler, M., Chen, 63–76, 2000. W. C., Richards, W., Pisick, E., Shapiro, G. I., Rollins, B. J., Chen, L. B., Griffin, J. D., and Sugarbaker, D. J. Expression of the focal 28. Nobes, C. D., and Hall, A. Rho, rac and cdc42 GTPases: regulators adhesion protein paxillin in lung cancer and its relation to cell motility. of actin structures, cell adhesion and motility. Biochem. Soc. Trans., 23: Oncogene, 18: 67–77, 1999. 456–459, 1995. 13. Sattler, M., Salgia, R., Durstin, M. A., Prasad, K. V., and Griffin, 29. Nobes, C. D., Hawkins, P., Stephens, L., and Hall, A. Activation of J. D. Thrombopoietin induces activation of the phosphatidylinositol-3Ј the small GTP-binding proteins rho and rac by growth factor receptors. kinase pathway and formation of a complex containing p85PI3K and the J. Cell Sci., 108: 225–233, 1995. protooncoprotein p120CBL. J. Cell. Physiol., 171: 28–33, 1997. 30. Adler, V., Yin, Z., Tew, K. D., and Ronai, Z. Role of redox potential 14. Webb, C. P., Hose, C. D., Koochekpour, S., Jeffers, M., Oskarsson, and reactive oxygen species in stress signaling. Oncogene, 15: 6104– M., Sausville, E., Monks, A., and Vande Woude, G. F. The geldana- 6111, 1999. mycins are potent inhibitors of the hepatocyte growth factor/scatter 31. Kamata, H., and Hirata, H. Redox regulation of cellular signaling. factor-met-urokinase plasminogen activator-plasmin proteolytic net- Cell. Signal., 11: 1–14, 1999. work. Cancer Res., 60: 342–349, 2000. 32. Rahman, I., and MacNee, W. Role of oxidants/antioxidants in 15. Salgia, R., Avraham, S., Pisick, E., Li, J. L., Raja, S., Greenfield, smoking-induced lung diseases. Free Radic. Biol. Med., 21: 669–681, E. A., Sattler, M., Avraham, H., and Griffin, J. D. The related adhesion 1996. focal tyrosine kinase forms a complex with paxillin in hematopoietic 33. Arakaki, N., Kajihara, T., Arakaki, R., Ohnishi, T., Kazi, J. A., cells. J. Biol. Chem., 271: 31222–31226, 1996. Nakashima, H., and Daikuhara, Y. Involvement of oxidative stress in 16. Sattler, M., Salgia, R., Shrikhhande, G., Verma, S., Pisick, E., tumor cytotoxic activity of hepatocyte growth factor/scatter factor. Prasad, K. V., and Griffin, J. D. Steel factor induces tyrosine phospho- J. Biol. Chem., 274: 13541–13546, 1999. rylation of CRKL and binding of CRKL to a complex containing c-kit, 34. An, W. G., Schulte, T. W., and Neckers, L. M. The heat shock phosphatidylinositol 3-kinase, and p120(CBL). J. Biol. Chem., 272: protein 90 antagonist geldanamycin alters chaperone association with 10248–10253, 1997. p210bcr-abl and v-src proteins before their degradation by the protea- 17. Sattler, M., Winkler, T., Verma, S., Byrne, C. H., Shrikhande, G., some. Cell Growth Differ., 11: 355–360, 2000. Salgia, R., and Griffin, J. D. Hematopoietic growth factors signal 35. Tamura, Y., Peng, P., Liu, K., Daou, M., and Srivastava, P. Immu- through the formation of reactive oxygen species. Blood, 93: 2928– notherapy of tumors with autologous tumor-derived heat shock protein 2935, 1999. preparations. Science, 278: 117–120, 1997.

Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2002 American Association for Cancer Research. Modulation of the c-Met/Hepatocyte Growth Factor Pathway in Small Cell Lung Cancer

Gautam Maulik, Takashi Kijima, Patrick C. Ma, et al.

Clin Cancer Res 2002;8:620-627.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/8/2/620

Cited articles This article cites 34 articles, 17 of which you can access for free at: http://clincancerres.aacrjournals.org/content/8/2/620.full#ref-list-1

Citing articles This article has been cited by 33 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/8/2/620.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/8/2/620. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.