Sorafenib Activates CD95 and Promotes Autophagy and Cell Death Via Src Family Kinases in Gastrointestinal Tumor Cells

Published OnlineFirst August 3, 2010; DOI: 10.1158/1535-7163.MCT-10-0274

Therapeutic Discovery Molecular Cancer Therapeutics Sorafenib Activates CD95 and Promotes Autophagy and Cell Death via Src Family Kinases in Gastrointestinal Tumor Cells

Margaret A. Park1, Roland Reinehr6, Dieter Häussinger6, Christina Voelkel-Johnson5, Besim Ogretmen5, Adly Yacoub1,4, Steven Grant1,2,3, and Paul Dent1,3,4

Abstract Sorafenib and vorinostat interact in a synergistic fashion to kill carcinoma cells by activating CD95; the present studies have determined how sorafenib and vorinostat individually contribute to CD95 activation. Sorafenib (3–6 μmol/L) promoted a dose-dependent increase in Src Y416, ERBB1 Y845 and CD95 Y232/ Y291 phosphorylation, and Src Y527 dephosphorylation. Low levels of sorafenib-induced (3 μmol/L) CD95 tyrosine phosphorylation did not promote surface localization whereas sorafenib (6 μmol/L), or sorafenib (3 μmol/L) and vorinostat (500 nmol/L) treatment promoted higher levels of CD95 phosphorylation which correlated with DISC formation, receptor surface localization, and autophagy. CD95 (Y232F, Y291F) was not tyrosine phosphorylated and was unable to localize plasma membrane or induce autophagy. Knockdown/knockout of Src family kinases abolished sorafenib-induced CD95 tyrosine phosphorylation, DISC formation, and the induction of cell death and autophagy. Knockdown of platelet-ived growth factor receptor-β enhanced Src Y416 and CD95 tyrosine phosphorylation, which correlated with elevated CD95 plasma membrane levels and autophagy, and with a reduced ability of sorafenib to promote CD95 membrane localization. Vorinostat increased reactive oxygen species levels, and in a delayed NFκB-dependent fashion, those of FAS ligand and CD95. Neutralization of FAS-L did not alter the initial rapid drug-induced activation of CD95; however, neutralization of FAS-L reduced sorafenib + vorinostat toxicity by ∼50%. Thus, sorafenib contributes to CD95 activation by promoting receptor tyrosine phosphorylation, whereas vorinostat contributes to CD95 activation via the initial facilitation of reactive oxygen species generation and subse- quently of FAS-L expression. Mol Cancer Ther; 9(8); 2220–31. ©2010 AACR.

Introduction developed as an inhibitor of RAF-1 but which was sub- sequently shown to inhibit multiple other kinases, in- In the United States, hepatoma patients have a 5-year cluding class III tyrosine kinase receptors (6). The survival rate of less than 10% (1). We have recently devel- antitumor effects of sorafenib in renal cell carcinoma oped a novel drug therapy combining the multikinase in- and in hepatoma have been ascribed to the antiangio- hibitor sorafenib with the histone deacetylase inhibitor genic actions of this agent through inhibition of the (HDACI) vorinostat, and this combination is entering growth factor receptors (7–9). However, several groups, phase I trial in hepatomas (2–5). including ours, have shown in vitro that sorafenib kills Sorafenib (Bay 43-9006, Nexavar; a RAF family kinase human leukemia cells at concentrations below the maxi- C μ inhibitor) is a multikinase inhibitor that was originally mum achievable dose ( max)of15to20 mol/L, through a mechanism involving the downregulation of MCL-1 (10, 11). Sorafenib-mediated MCL-1 downregulation oc- Authors' Affiliations: Departments of 1Biochemistry and 2Medicine, curred through a translational rather than a transcrip- 3Institute for Molecular Medicine and 4Neurosurgery, Virginia Commonwealth University, Richmond, Virginia; 5Medical University of tional or posttranslational process that was mediated South Carolina, Charleston, South Carolina; and 6Clinic for by endoplasmic reticulum stress signaling (12, 13). This Gastroenterology, Hepatology, and Infectiology, Heinrich-Heine- suggests that the previously observed antitumor effects University Düsseldorf, Düsseldorf, Germany of sorafenib are mediated by a combination of inhibition Note: Supplementary material for this article is available at Molecular of RAF family kinases and the ERK1/2 pathway, receptor Cancer Therapeutics Online (http://mct.aacrjournals.org/). tyrosine kinases that signal angiogenesis, and the induc- P. Dent is the holder of a Universal, Inc. Professorship in Signal Transduction Research. tion of endoplasmic reticulum stress signaling. HDACI represent a class of agents that act by block- Corresponding Author: Paul Dent, Department of Biochemistry and Mo- lecular Biology, 401 College Street, Massey Cancer Center, Room 280a, ing histone deacetylation, thereby modifying chromatin Box 980035, Virginia Commonwealth University, Richmond, VA 23298- structure and gene transcription. HDACs, along with 0035. Phone: 804-628-0861; Fax: 804-827-1309. E-mail: [email protected] histone acetyl-transferases, reciprocally regulate the acet- doi: 10.1158/1535-7163.MCT-10-0274 ylation status of the positively charged NH2-terminal ©2010 American Association for Cancer Research. histone tails of nucleosomes. HDACIs promote histone

2220 Mol Cancer Ther; 9(8) August 2010

Src and Sorafenib

acetylation and neutralization of positively charged of vehicle (DMSO) in medium was 0.02% (v/v). Cells lysine residues on histone tails, allowing chromatin to were not cultured in reduced serum medium during assume a more open conformation, which favors tran- any study in this article. scription (14). However, HDACIs also induce the acety- In vitro cell treatments, microscopy, SDS-PAGE, and lation of other non-histone targets, actions that might Western blot analysis. For in vitro analyses of short-term have pleiotropic biological consequences, including inhi- cell death effects, cells were treated with vehicle or bition of HSP90 function, induction of oxidative injury, vorinostat/sorafenib for the indicated times in the figure and upregulation of death receptor expression (15–17). legends. For apoptosis assays, where indicated, cells With respect to combinatorial drug studies with a multi- were isolated at the indicated times, and either subjected kinase inhibitor such as sorafenib, HDACIs are of inter- to trypan blue cell viability assay by counting in a light est in that they also downregulate multiple oncogenic microscope or fixed to slides, and stained using a kinases by interfering with HSP90 function, leading to commercially available Diff-Quick (Giemsa) assay kit. Al- the proteasomal degradation of these proteins. Vori- ternatively, using a Becton Dickinson FACScan flow nostat (suberoylanilide hydroxamic acid; Zolinza) is a cytometer, the Annexin V/propidium iodide assay hydroxamic acid HDACI that has shown preliminary was carried out according to the instructions of the man- preclinical evidence of activity in hepatoma and other ufacturer to determine cell viability (BD PharMingen). C ∼ μ – malignancies with a max of 9 mol/L (18 28). Vorinostat/sorafenib lethality, as determined by Annexin We recently published that sorafenib and vorinostat in- V/propidium iodide, was first evident ∼24 hours after teract to kill a wide range of tumor cell types via activation drug exposure (data not shown). of the CD95 extrinsic apoptotic pathway, concomitant For SDS-PAGE and immunoblotting, cells were plated with drug-induced reduced expression of c-FLIP-s via at 5 × 105 cells/cm2 and treated with drugs at the indicated PERK/eIF2α activation (29, 30). The present studies have concentrations and after the indicated time of treatment, extended in greater molecular detail our analyses to un- lysed in whole-cell lysis buffer [0.5 mol/L Tris-HCl (pH derstanding how sorafenib and vorinostat individually in- 6.8), 2% SDS, 10% glycerol, 1% β-mercaptoethanol, and teract to promote CD95 activation and tumor cell death. 0.02% bromophenol blue], and the samples were boiled for 30 minutes. The boiled samples were loaded onto Materials and Methods 10% to 14% SDS-PAGE and electrophoresis was run overnight. Proteins were electrophoretically transferred onto Materials 0.22 μm of nitrocellulose, and immunoblotted with vari- Sorafenib tosylate (Bayer) and vorinostat (Merck) were ous primary antibodies against different proteins. provided by the Cancer Treatment and Evaluation Pro- Transfection of cells with siRNA or with plasmids gram, National Cancer Institute/NIH (Bethesda, MD). For plasmids. Cells were plated as described above and Trypsin-EDTA, DMEM, RPMI, and penicillin-streptomycin 24 hours after plating, transfected. For mouse embryonic were purchased from Life Technologies. HEPG2, HEP3B, fibroblasts (2–5 μg), or other cell types (0.5 μg), plasmids and HuH7 (hepatoma) cells were purchased from expressing a specific mRNA (or siRNA) or appropriate American Type Culture Collection. Commercially avail- vector control plasmid DNA was diluted in 50 μLof able validated short hairpin RNA molecules to knock serum-free and antibiotic-free medium (one portion for down RNA/protein levels were from Qiagen: CD95 each sample). Concurrently, 2 μLofLipofectAMINE (SI02654463; SI03118255), ATG5 (SI02655310), and Beclin 2000 (Invitrogen), was diluted into 50 μL of serum-free 1 (SI00055573, SI00055587). For confirmatory purposes, and antibiotic-free medium (one portion for each we also used the short hairpin RNA construct targeting sample). Diluted DNA was added to the diluted ATG5 (pLVTHM/Atg5), which was a generous gift from LipofectAMINE 2000 for each sample and incubated at Dr. S. Yousefi, Department of Pharmacology, University room temperature for 30 minutes. This mixture was of Bern, Bern, Switzerland. The reagents used and the added to each well/dish of cells containing 20 μLof performance of experimental procedures were described serum-free and antibiotic-free medium for a total volume in refs. (2–5, 21, 29–37). of 300 μL, and the cells were incubated for 4 hours at 37°C. An equal volume of 2× medium was then added to each Methods well. Cells were incubated for 48 hours, then treated with Culture and in vitro exposure of cells to drugs. All es- vorinostat/sorafenib. Transfection with siRNA. tablished cell lines were cultured at 37°C, 5% (v/v) CO2 Cells were plated into 60 mm in vitro using RPMI supplemented with 5% (v/v) FCS dishes from a fresh culture growing in log phase as de- and 10% (v/v) nonessential amino acids. For short-term scribed above, and transfected 24 hours after plating. cell-killing assays and immunoblotting studies, cells were Prior to transfection, the medium was aspirated and plated at a density of 3 × 103 per cm2 (∼2×105 cells per 1mLofserum-freemediumwasaddedtoeachplate. well of a 12-well plate), and 48 hours after plating, trea- For transfection, 10 nmol/L of the annealed siRNA, the ted with various drugs, as indicated. In vitro vorinostat positive sense control double-stranded siRNA targeting and sorafenib treatments were from 100 mmol/L stock glyceraldehyde-3-phosphate dehydrogenase or the nega- solutions of each drug and the maximal concentration tive control (a “scrambled” sequence with no significant

www.aacrjournals.org Mol Cancer Ther; 9(8) August 2010 2221

Park et al.

homology to any known gene sequences from mouse, of 1.00 indicates additivity, and a value of >1.00 indicates rat, or human cell lines) were used. siRNA (10 nmol/L; antagonism of action between the agents. scrambled or experimental) was diluted in serum-free medium. Four microliters of HiPerFect (Qiagen) was Results added to this mixture and the solution was mixed by pi- petting up and down several times. This solution was in- Treatment of gastrointestinal tumor cells (HEPG2, cubated at room temperature for 10 minutes, then added HEP3B, and UOK121LN) with low concentrations of sor- dropwise to each dish. The medium in each dish was afenib (3 and 6 μmol/L) increased the tyrosine phos- swirled gently to mix, then incubated at 37°C for 2 hours. phorylation of Src Y416, ERBB1 Y845, and the basal One milliliter of 10% (v/v) serum-containing medium tyrosine phosphorylation of platelet-derived growth fac- was added to each plate, and cells were incubated tor receptor-β (PDGFRβ) and caused Src Y527 dephos- 3 at 37°C for 48 hours before replating (50 × 10 cells each) phorylation (Fig. 1A). Although sorafenib (3 μmol/L) onto 12-well plates. Cells were allowed to attach modestly increased CD95 tyrosine phosphorylation, this overnight, then treated with vorinostat/sorafenib level of phosphorylation neither promoted CD95 surface (0–48 hours). Trypan blue exclusion/TUNEL/flow cyto- localization nor promoted caspase 8 association with metry assays and SDS-PAGE/immunoblotting analyses CD95 (DISC formation; Fig. 1B, C, and D; Supplementa- were done at the indicated time points. ry Fig. S1). Treatment of cells with a higher concentration Microscopy for LC3-GFP expression. Cells were trans- of sorafenib (6 μmol/L) promoted a further increase in fected with a plasmid to express an LC3-GFP fusion pro- CD95 tyrosine phosphorylation above that induced by tein, and were then cultured for 24 hours. Cells were then sorafenib (3 μmol/L), and which correlated with CD95 treated with drugs, as indicated/LC3-GFP–transfected surface localization and promoted caspase 8 association cells were visualized at the indicated time points on a with CD95 (Fig. 1B, C, and D). Concomitant treatment of Zeiss Axiovert 200 microscope using an FITC filter. cells with vorinostat enhanced sorafenib-induced (3 μmol/L) Recombinant adenoviral vectors; infection in vitro. tyrosine phosphorylation of CD95 that correlated with We generated and purchased previously noted recombi- CD95 surface localization and promoted caspase 8 associ- nant adenoviruses to express a wide variety of proteins ation with CD95 (Fig. 1B, C, and D). or to knock down p21 expression (Vector Biolabs). Cells HuH7 hepatoma cells lack endogenous CD95 expres- were infected with these adenoviruses at an approximate sion (29, 30, 37). Treatment of HuH7 cells with sorafe- multiplicity of infection of 50. Cells were incubated for nib (3 μmol/L) and vorinostat resulted in a very 24 hours to ensure adequate expression of transduced modest increase in tumor cell killing, which was signif- gene products prior to drug exposures. icantly enhanced when HuH7 cells were transiently Assessment of reactive oxygen species generation. transfected with a plasmid to express wild-type CD95 Hepatoma cells were plated in 96-well plates. Cells were (Fig. 2A). Transfection of cells with a plasmid to ex- preincubated with dihydro-DCF (5 mmol/L for 30 press a CD95 protein lacking the sites of tyrosine phos- minutes) which is nonfluorescent in its dihydro form phorylation (Y232F, Y291F) did not facilitate sorafenib but, upon reaction with reactive oxygen species (ROS), and vorinostat lethality. In general agreement with becomes highly fluorescent. Dihydro-DCF is sensitive to these data, sorafenib and vorinostat promoted DISC oxidation by hydroxyl radicals and peroxy-nitrite formation in cells transfected with wild-type CD95 directly and hydrogen peroxide in the presence of oxi- but not with CD95 (Y232F, Y291F; Fig. 2B). In Fig. 1, dases. Fluorescence measurements were obtained 0 to we noted that sorafenib promoted the activation of 30 minutes after the addition of the drug with a Vector Src and enhanced the basal levels of PDGFRβ tyrosine 3 plate reader. Data are presented corrected for basal phosphorylation, and it has been shown in other cell fluorescence of vehicle-treated cells at each time point systems that one of the protein kinases which phos- and expressed as a fold increase in ROS levels. Each phorylate CD95 is the Src family member c-Yes (37). time point represents the mean of at least six data points Knockdown of PDGFRβ promoted a compensatory ac- per experiment and of a total of three independent tivation of Src within 24 hours as judged by the in- experiments. creased Y416 phosphorylation and the decreased Y527 Data analysis. Comparison of the effects of various phosphorylation (Fig. 2C). Knockdown of PDGFRβ en- treatments was done using ANOVA and Student's t test. hanced CD95 tyrosine phosphorylation and procaspase Differences with a P < 0.05 were considered statistically 8 association with CD95, the phosphorylation and significant. Experiments shown are the means of multiple association of which, respectively, in the absence of individual points (±SEM). Median dose-effect isobolo- PDGFRβ expression were not further enhanced by sor- gram analyses to determine the synergism of drug inter- afenib (Fig. 2D, top). Enhanced levels of CD95 tyrosine action were done according to the methods of T-C. Chou phosphorylation caused by knockdown of PDGFRβ and P. Talalay using the CalcuSyn program for Windows were blocked by the expression of a dominant negative (Biosoft). Cells were treated with agents at a fixed con- (kinase inactive) Src protein (Fig. 2D, bottom). centration dose. A combination index value of <1.00 in- To confirm our findings in hepatoma cells using other dicates synergy of interaction between the drugs, a value genetic approaches, we used the SV40 large T-antigen

2222 Mol Cancer Ther; 9(8) August 2010 Molecular Cancer Therapeutics

Src and Sorafenib

Figure 1. Sorafenib causes a dose-dependent activation of Src, ERBB1, and CD95. A, HEPG2 cells were treated with vehicle (DMSO), sorafenib (3 and 6 μmol/L), as indicated. Thirty minutes after treatment, cells were isolated and lysed. Portions of lysates were immunoprecipitated (anti–phospho-tyrosine) and immunoprecipitates probed for PDGFRβ. After SDS-PAGE, immunoblotting was done to determine the phosphorylation of the indicated proteins. Representative blots (n = 3). B, cells were plated in eight-well chamber slides and were treated with vehicle (DMSO), sorafenib (3 and 6 μmol/L), or vorinostat (500 nmol/L), as indicated. Cells were fixed 6 h after exposure and surface levels of CD95 determined by immunohistochemistry. The density of CD95 staining was determined in 40 cells (n = 2, ±SEM). C, HEPG2 cells were treated with vehicle (DMSO), sorafenib (3 and 6 μmol/L), or vorinostat (500 nmol/L), as indicated. Cells were isolated 6 h after exposure and CD95 immunoprecipitated. The amount of caspase 8 association with CD95 was determined. Representative blots (n = 3). D, HEPG2 cells were treated with vehicle (DMSO), sorafenib (3 and 6 μmol/L), or vorinostat (500 nmol/L), as indicated. Thirty minutes after treatment, cells were isolated and lysed. Portions of lysates were immunoprecipitated (anti-CD95). After SDS-PAGE, immunoblotting was done to determine the phosphorylation of CD95. Representative blots (n = 3). transformed mouse embryonic fibroblasts lacking expres- the levels of a protective form of autophagy (29, 30, 36). sion of c-Src, c-Fyn, and c-Yes. In transformed MEFs: sor- In agreement with data in Fig. 2D showing that knock- afenib (3 μmol/L) and vorinostat treatment, or sorafenib down of PDGFRβ increased CD95 tyrosine phosphoryla- (6 μmol/L) treatment, promoted surface localization of tion; knockdown of PDGFRβ increased basal levels of CD95 in a Src family kinase-dependent fashion (Fig. 3A). plasma membrane–associated CD95 and suppressed the Transfection of wild-type MEFs with a plasmid to express ability of sorafenib (6 μmol/L) to promote CD95 plasma wild-type CD95, followed by sorafenib (6 μmol/L) expo- membrane levels (Fig. 4A). Knockdown of PDGFRβ or sure resulted in enhanced CD95 tyrosine phosphorylation treatment with sorafenib (6 μmol/L) increased autoph- and enhanced DISC formation (Fig. 3B). In MEFs lacking agy in a CD95- and ATG5-dependent fashion (Fig. 4B). expression of c-Src, c-Fyn, and c-Yes, sorafenib (6 μmol/L) The ability of sorafenib to kill hepatoma cells was re- exposure did not cause CD95 tyrosine phosphorylation or duced by the knockdown of PDGFRβ and sorafenib tox- DISC formation. The toxicity of sorafenib was reduced in icity (6 μmol/L) was suppressed by the knockdown of fibroblasts lacking expression of c-Src, c-Fyn, and c-Yes CD95 (Fig. 4C). Expression of wild-type CD95, but not (Fig. 3C). In hepatoma cells, expression of dominant neg- mutant CD95 (Y232F, Y291F), facilitated sorafenib- ative Src suppressed sorafenib-induced (6 μmol/L) induced (6 μmol/L) autophagy in HuH7 cells (Supple- CD95 activation; dominant negative Src increased basal mentary Fig. S2). In agreement with the above findings, levels of cell death but suppressed drug-induced toxicity the expression of dominant negative Src suppressed (Fig. 3D). sorafenib-induced (6 μmol/L) autophagy as well as tumor − − Prior studies by our group had shown that sorafenib cell killing (Figs. 3D and 4D). In ATG5 / MEFs, which can- and vorinostat-induced activation of CD95 increased not undergo autophagy, sorafenib- and vorinostat-induced

www.aacrjournals.org Mol Cancer Ther; 9(8) August 2010 2223

Park et al.

toxicities were enhanced (Supplementary Fig. S3). Vorino- and bile acid–induced autophagy (36). Knockdown of stat treatment of malignant blood cancer cells strongly and p21 suppressed the ability of sorafenib + vorinostat treat- rapidly increases p21Cip-1/WAF1/mda-6 (p21) expression; ment to increase CD95 surface localization and DISC for- however, in hepatoma cells, it more modestly elevated mation (Supplementary Fig. S4, immunohistochemistry). p21 levels and did so in a delayed fashion (Supplementary We next determined how vorinostat promoted sorafenib- Fig. S4, top blot; refs. 20, 21). Increased p21 levels were re- induced CD95 activation. Sorafenib (3 μmol/L) and duced by cotreatment with sorafenib. Knockdown of p21 vorinostat (500 nmol/L) rapidly promoted the generation reduced sorafenib + vorinostat toxicity and reduced of ROS in HEPG2 cells that was quenched by the expres- drug-induced autophagy (Supplementary Fig. S4, bottom sion of thioredoxin (Supplementary Fig. S5). Similar data graphs). These findings are similar to our prior studies in were obtained in HEP3B and UOK121LN cells (data not primary hepatocytes treated with bile acids, in which over- shown). Quenching drug-induced ROS suppressed sora- expression of p21 promoted increased CD95 expression fenib and vorinostat toxicity and blocked CD95 activa- and activation, and loss of p21 reduced CD95 activation tion (Fig. 5A); sorafenib and vorinostat increase cell

Figure 2. Sorafenib-induced Src-dependent CD95 tyrosine phosphorylation is due to inhibition of PDGFRβ. A, HuH7 cells were transfected with empty vector plasmid (CMV), a plasmid to express CD95-YFP, or a plasmid to express mutant inactive CD95-YFP (Y232F, Y291F). Twenty-four hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L). Cells were isolated 48 h after exposure and viability determined by trypan blue exclusion (n = 3, ±SEM). B, HuH7 cells were transfected with a plasmid to express CD95-YFP or a plasmid to express mutant inactive CD95-YFP (Y232F, Y291F) and 24 h later were treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L). Cells were isolated 6 h after exposure and CD95 immunoprecipitated. The amount of caspase 8 association with CD95 was determined. Representative blots (n = 3). C, HEPG2 cells were transfected with scrambled siRNA (siSCR) or an siRNA to knock down PDGFRβ. Thirty-six hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (6 μmol/L). Cells were isolated 30 min later and the phosphorylation of Src determined by immunoblotting. Representative blots (n = 3). D, top blots: HEPG2 cells were transfected with scrambled siRNA (siSCR) or an siRNA to knock down PDGFRβ. Thirty-six hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (6 μmol/L). Cells were isolated 6 h later and CD95 immunoprecipitated. The association of caspase 8 with CD95 and the tyrosine phosphorylation of CD95 were determined. Representative blots (n = 3). Bottom blots: cells were transfected with scrambled siRNA (siSCR) or an siRNA to knock down PDGFRβ and in parallel were transfected with an empty vector plasmid (CMV) or a plasmid to express dominant negative c-Src. Thirty-six hours after transfection, cells were isolated and CD95 immunoprecipitated. The tyrosine phosphorylation of CD95 was determined. Representative blots (n = 3).

2224 Mol Cancer Ther; 9(8) August 2010 Molecular Cancer Therapeutics

Src and Sorafenib

Figure 3. Src family kinase signaling is essential for sorafenib-induced CD95 activation and drug toxicity. A, wild-type or Src/Fyn/Yes null SV40 large T transformed MEFs plated in eight-well chamber slides and were treated with vehicle (DMSO), sorafenib (3 and 6 μmol/L), or vorinostat (500 nmol/L), as indicated. Cells were fixed 6 h after exposure and surface levels of CD95 determined by immunohistochemistry. The density of CD95 staining was determined in 40 cells (n = 2, ±SEM). B, wild-type or Src/Fyn/Yes null SV40 large T transformed MEFs were transfected with a plasmid to express CD95-YFP or a plasmid to express mutant inactive CD95-YFP (Y232F, Y291F) and 24 h later were treated with vehicle (DMSO) or sorafenib (6 μmol/L). Cells were isolated 6 h later and CD95 immunoprecipitated via the YFP tag. The association of caspase 8 with CD95 and the tyrosine phosphorylation of CD95 were determined. Representative blots (n = 3). C, wild-type or Src/Fyn/Yes null SV40 large T transformed MEFs were treated with vehicle (DMSO) or increasing concentrations of sorafenib (0.1–6.0 μmol/L). Cells were isolated 48 h after drug exposure and viability determined by trypan blue exclusion (n = 3, ±SEM). D, HEPG2 cells were transfected with empty vector plasmid (CMV) or a plasmid to express dominant negative Src. Twenty-four hours after transfection, cells were treated with sorafenib (3 and 6 μmol/L), as indicated. Top, cells were isolated 6 h after drug exposure, CD95 immunoprecipitated and the levels of CD95 tyrosine phosphorylation and the association of procaspase 8 with CD95 determined. Representative blots (n = 3). Bottom, cells were isolated 48 h after treatment and viability determined by trypan blue exclusion (n = 3, ±SEM).

www.aacrjournals.org Mol Cancer Ther; 9(8) August 2010 2225

Park et al.

Figure 4. PDGFRβ regulates CD95-dependent sorafenib-induced autophagy. A, HEPG2 cells were plated in eight-well chamber slides and transfected with scrambled siRNA (siSCR) or an siRNA to knock down PDGFRβ. Thirty-six hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (6 μmol/L). Cells were fixed 6 h after exposure and surface levels of CD95 determined by immunohistochemistry. The density of CD95 staining was determined in 40 cells (n = 2, ±SEM). B, HEPG2 cells were plated in eight-well chamber slides and transfected with a plasmid to express LC3-GFP and in parallel with scrambled siRNA (siSCR) or siRNA molecules to knock down PDGFRβ or CD95. Thirty-six hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (6 μmol/L). The number of intense punctate GFP-LC3 staining vesicles was determined in 40 cells (n = 2, ±SEM). C, HEPG2 cells were transfected with scrambled siRNA (siSCR) or siRNA molecules to knock down PDGFRβ or CD95. Thirty-six hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (6 μmol/L). Forty-eight hours later, cells were isolated and viability determined by trypan blue exclusion (n =3, ±SEM). D, HEPG2 cells were plated in eight-well chamber slides and transfected with a plasmid to express LC3-GFP and in parallel with a plasmid to express dominant negative Src. Thirty-six hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (6 μmol/L). The number of intense punctate GFP-LC3 staining vesicles was determined in 40 cells (n = 2, ±SEM).

2226 Mol Cancer Ther; 9(8) August 2010 Molecular Cancer Therapeutics

Src and Sorafenib

surface CD95 levels by 1.97 ± 0.08-fold, which was re- In prior studies, we noted that ∼6to∼24 hours follow- duced by transfection of thioredoxin to a 1.12 ± 0.06-fold ing vorinostat exposure, hepatoma cells increased the increase (P < 0.05). In agreement with our CD95 activation expression of FAS ligand and CD95 in a cell type–dependent data, and the role CD95 plays in stimulating autophagy, fashion (29). Incubation of hepatoma cells with an anti– quenching of ROS also blocked drug combination–induced FAS-L neutralizing antibody did not alter the ability of autophagy (Supplementary Fig. S6). sorafenib (3 μmol/L) and vorinostat treatment to activate Treatment of hepatoma cells with sorafenib and vori- CD95 6 hours after drug exposure (Fig. 6A, immunohisto- nostat initially suppressed total cellular protein tyrosine chemistry, top). However, neutralization of FAS-L sup- phosphatase activity by ∼25%, which was blocked by the pressed the ability of sorafenib (3 μmol/L) and vorinostat expression of thioredoxin (Fig. 5B). However, within treatment to kill tumor cells by ∼50% (Fig. 6A, bottom 3 hours, vorinostat and sorafenib treatment profoundly graph). Vorinostat activates NFκB, an effect that is modest- enhanced tyrosine phosphatase activity, again, in an ly enhanced by sorafenib (3 μmol/L), and an effect that ROS-dependent fashion. The ability of vorinostat and was blocked by the expression of dominant negative IκB sorafenib to induce CD95 tyrosine phosphorylation in super-repressor (Fig. 6B, left graph; data not shown). Ex- multiple gastrointestinal tumor types was blocked by pression of the dominant negative IκB super-repressor sup- the expression of thioredoxin (Fig. 5C). pressed drug-induced tumor cell killing (Fig. 6B, right

Figure 5. Vorinostat-induced ROS plays a key role in promoting PTPase inactivation and CD95 tyrosine phosphorylation. A, HEPG2 cells were transfected with empty vector plasmid (CMV) or a plasmid to express thioredoxin (TRX). Twenty-four hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L). Cells were isolated 48 h after exposure and viability determined by trypan blue exclusion (n = 2, ±SEM). B, HEPG2 cells were plated in 96-well plates and 24 h after plating were transfected with empty vector plasmid (CMV) or a plasmid to express thioredoxin. Twenty-four hours after transfection, cells were treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L). PTPase activity was measured using a commercial kit as described in Methods (n = 2, ±SEM). C, left, HuH7 cells transfected with plasmids to express CD95-YFP or CD95-YFP FF. Twenty-four hours after plating in eight-well chamber, slides were treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L) in combination. Cells were fixed after 6 h and cell surface CD95 levels determined. Right, HEPG2 cells were transfected with empty vector (CMV) or to express wild-type thioredoxin. Twenty-four hours after transfection, cells were treated with vehicle (DMSO) or with sorafenib (3.0 μmol/L) and vorinostat (500 nmol/L). Cells were isolated after 6 h and CD95 immunoprecipitated to determine DISC formation and CD95 tyrosine phosphorylation (n = 3, ±SEM).

www.aacrjournals.org Mol Cancer Ther; 9(8) August 2010 2227

Park et al.

Figure 6. Vorinostat promotes sorafenib toxicity by increasing expression of FAS-L. A, top immunohistochemistry, HEPG2 cells plated in eight-well chamber slides were pretreated with control IgG or an IgG to neutralize FAS-L (1 μg/mL) and 30 min later treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L) in combination. Cells were fixed 6 h after exposure and surface levels of CD95 determined by immunohistochemistry. The density of CD95 staining was determined in 40 cells (n = 2, ±SEM). Bottom graph, HEPG2 cells were pretreated with control IgG or an IgG to neutralize FAS-L (1 μg/mL) and 30 min later treated with vehicle (DMSO) or sorafenib (3 μmol/L) and vorinostat (500 nmol/L) in combination. Forty-eight hours later, cells were isolated and viability determined by trypan blue exclusion (n = 3, ±SEM). B, left graph, HEPG2 cells 24 h after plating in 96-well plates were transfected with NFκB-luciferase and β-galactosidase constitutive reporter constructs. Thirty-six hours after transfection, cells were treated with vehicle (DMSO), sorafenib (3.0 μmol/L), vorinostat (500 nmol/L), or both sorafenib and vorinostat. Cells were assayed for NFκB-luciferase and β-galactosidase activity 24 h after treatment (±SEM, a representative from two separate studies). Control studies showed that overexpression of dominant negative IκB blocked vorinostat-induced activation of NFκB-luciferase activity (data not shown). Right graph, HEPG2 cells 24 h after plating were infected with control empty vector virus (CMV) or a recombinant virus to express dominant negative IκB S32A S36A (dn IκB). Twenty-four hours after infection, cells were treated with vehicle (DMSO), sorafenib (3.0 μmol/L), vorinostat (500 nmol/L), or both sorafenib and vorinostat. Forty-eight hours after drug exposure, cells were isolated, spun onto glass slides, and stained using established methods for double-stranded DNA breaks indicative of apoptosis (TUNEL) as described in Materials and Methods (n = 2, ±SEM). C, HEPG2 and HEP3B cells 24 h after plating were infected with control empty vector virus (CMV) or a recombinant virus to express dominant negative IκB S32A S36A (dn IκB). Twenty-four hours after infection, cells were treated with vehicle (DMSO) or vorinostat (500 nmol/L). Cells were isolated 12 and 24 h after treatment. SDS-PAGE and immunoblotting were done to determine changes in the expression of CD95 and FAS-L, as indicated. A representative is shown of three separate studies.

2228 Mol Cancer Ther; 9(8) August 2010 Molecular Cancer Therapeutics

Src and Sorafenib

graph). In other systems, it has been shown that HDACIs Knockdown of PDGFRβ increased basal plasma mem- could, via NFκB, increase the expression of death receptors brane levels of CD95 that was Src-dependent, and when and their cognate ligands (38). We have previously pub- PDGFRβ was knocked down, the relative ability of sora- lished that vorinostat induced expression of FAS-L and fenib to cause further activation of CD95 was reduced. CD95 in hepatoma cells and we now note that this effect This finding was mirrored by changes in autophagy pre- was blocked by the expression of dominant negative IκB sented in Fig. 4B, wherein knockdown of PDGFRβ in- (Fig. 6C; ref. 29). Collectively, our data argue that vorino- creased autophagy in a CD95-dependent fashion and in stat contributes to sorafenib-induced activation of CD95 cells lacking PDGFRβ, the ability of sorafenib to cause shortly following drug exposure by elevating ROS levels additional autophagy was reduced. Although we ob- that inhibit PTPase function and promote CD95 activation, served enhanced levels of CD95 activation after PDGFRβ whereas at later time points, vorinostat promotes CD95 knockdown in Fig. 4A, this did not translate into in- signaling and tumor cell death by increasing FAS-L ex- creased basal levels of cell killing in Fig. 4C; possibly this pression in an NFκB-dependent fashion. was because we were observing elevated levels of protective autophagy after knockdown as was shown in Fig. 4B Discussion (29, 30). Thus, logically, based on data in Fig. 4A and B, the relative reduction in sorafenib-induced cell killing in Sorafenib is a small molecule drug that inhibits RAF PDGFRβ knockdown cells in Fig. 4C was because with- family serine/threonine kinases and class III receptor ty- out expression of the key target, i.e., PDGFRβ, sorafenib rosine kinases such as the PDGF and FLT receptors. Pre- cannot instantaneously initiate the series of events that vious studies have shown that sorafenib and the HDACI would lead to PDGFRβ inhibition, compensatory Src ac- vorinostat interact in vitro and in vivo in a greater than tivation, CD95 tyrosine phosphorylation, etc. additive fashion to kill transformed cells that was mech- Based on our data, if sorafenib is promoting the activa- anistically dependent on activation of CD95 (29, 30). The tion of Src family kinases, we reasoned that it is probable present studies attempted to determine, in much greater that other targets of Src kinases are also phosphorylated detail, the molecular mechanisms by which sorafenib and in response to sorafenib treatment. For example, in a pre- vorinostat, as individual agents, interacted to activate liminary study, we noted that FAK and IGF1R tyrosine CD95 and promote drug toxicity. phosphorylation are enhanced in response to sorafenib Sorafenib (3–6 μmol/L) caused a dose-dependent in- exposure. Increased FAK signaling has the potential to crease in CD95 tyrosine phosphorylation and, based on promote metastatic spread of tumor cells via Src (39). multiple criteria, in parallel activated Src family nonre- Signaling by Src kinases is known to facilitate ERBB1 ceptor tyrosine kinases. Expression of dominant negative Y845 phosphorylation and we observed this in our sys- SrcorknockoutofSrc/Fyn/Yesblockedsorafenib- tem following sorafenib treatment. Phosphorylation of induced CD95 tyrosine phosphorylation. CD95 (Y232F, Y845 has been linked to ERBB1-induced tumor cell Y291F) was not activated by either higher sorafenib growth, independent of ERK1/2 and phosphoinositide- (6 μmol/L) concentrations or by sorafenib (3 μmol/L) 3-kinase signaling (40). In at least one study, low concen- and vorinostat treatment. Knockdown of class III receptor trations of sorafenib have been shown to promote tyrosine kinase PDGFRβ expression enhanced Src activity; MEK1/2 activation via IGF1 receptor signaling, and in- in a Src-dependent fashion, knockdown of PDGFRβ en- hibition of MEK1/2 signaling enhanced sorafenib toxici- hanced CD95 tyrosine phosphorylation and also sup- ty (41). We have recently published similar data in pressed the ability of sorafenib to induce CD95 tyrosine malignant blood cancer cells (42). Thus, the practical out- phosphorylation. Vorinostat rapidly enhanced the ability come of sorafenib promoting Src kinase activation is that of lower sorafenib concentrations to increase ROS levels while this drug acts to suppress tumor growth through and to transiently suppress total cellular PTPase activity Src-dependent activation of CD95, the induction of endo- within 1 hour, which was due to ROS generation. However, plasmic reticulum stress and inhibition of proangiogenic ROS generation also significantly enhanced PTPase activ- growth factor receptors; it also has the capacity to pro- ity 3 hours after drug treatment and correlated with Src mote tumor cell survival and migration through elevated Y527 dephosphorylation and with vorinostat-stimulated signaling by Src, FAK, ERBB1, IGF1R, and possibly, the CD95 tyrosine phosphorylation. Thus, one potential model ERK1/2 pathway. Additional studies will be required to for molecular drug action is that sorafenib-mediated define the precise nodal enzymes within the sorafenib- inhibition of PDGFRβ causes a compensatory activation induced signaling network that can promote and suppress of Src family tyrosine kinases that in turn phosphorylate sorafenib lethality. and activate CD95. In addition, based on the degree to We recently noted, using higher concentrations of which PDGFRβ is inhibited by sorafenib, CD95 becomes vorinostat than those used in this article, that the drug tyrosine-phosphorylated in a Src-dependent fashion. induced DNA damage, via ROS generation, and this Vorinostat, by interacting with sorafenib to elevate ROS played a key role in the activation of NFκB by the drug levels, rapidly suppresses cellular PTPase activity which (43). Vorinostat facilitated CD95 activation in hepatoma promotes greater levels of CD95 tyrosine phosphorylation cells by two mechanisms: increased CD95 tyrosine and CD95 activation. phosphorylation and increased expression of FAS-L.

www.aacrjournals.org Mol Cancer Ther; 9(8) August 2010 2229

Park et al.

Vorinostat induced ROS generation, and when in com- combination–induced autophagy that was associated bination with sorafenib, inhibited PTPase activity and with reduced CD95 activation. We also noted that loss then profoundly activated PTPase activity; effects that of Src function blocked sorafenib-induced (6 μmol/L) were both ROS-dependent. These effects also correlated protective autoph‐agy; autophagy that was CD95- with increased CD95 phosphorylation, increased Src dependent. In contrast, in malignant blood cancer cells, Y416 phosphorylation, and dephosphorylation of Src we have previously shown that expression of p21 sup- Y527. PTPases have a ∼10-fold higher specific activity pressed the lethality of vorinostat and sorafenib treat- than the protein kinases, the actions of which are re- ment (20). Thus, vorinostat as a single agent has the versed, meaning that a small reduction in PTPase activ- potential to kill tumor cells by increasing FAS-L and ity could significantly modify a site of phosphorylation p21 levels but could also act to promote tumor cell sur- which is being actively increased due to the actions of a vival through NFκB activation. Thus, sorafenib, in part, kinase (44). The precise PTPase that regulates CD95 ty- subverts vorinostat-induced cell killing processes by rosine phosphorylation or the Src Y527 phosphorylation blocking increased p21 expression. Of note was our find- in hepatoma cells has not been identified, although it ing that sorafenib suppressed vorinostat-induced p21 ex- has been shown that both CD45 and SHP1 could prox- pression but not the enhancement of FAS-L expression; imally modulate CD95 signaling in immune cell types, because p21 protein levels, unlike those of FAS-L, are reg- and that both SHP1 and Src family kinases could asso- ulated by both protein and mRNA stabilization as well as ciate with CD95 (45, 46). This data, combined with our for both proteins at the level of transcription, our data present findings, argues that specific inhibitors of SHP-1 suggests that the HDACI-induced expression of p21 most could be of value in promoting sorafenib toxicity. probably occurs at a posttranscriptional level (48, 49). Using a neutralizing antibody, inhibition of FAS-L func- We have previously shown that sorafenib and vorino- tion suppressed sorafenib + vorinostat toxicity by ∼50%. stat combination therapy are effective at killing hepato- This correlated with vorinostat, in an NFκB-dependent ma cells in vivo, and sorafenib and HDACI combination fashion, promoting increased expression of FAS-L 12 to therapy is entering phase I evaluation in patients with 24 hours after drug exposure. However, a multitude of hepatoma,aswellasinpatientswithrenalcarcinoma studies have shown that NFκB activation could also stim- and non–small cell lung cancer (29, 30). The present stud- ulate the expression of antiapoptotic proteins and promote ies provide additional mechanistic information as to how cell growth. Previously, we have shown that the rapidly these agents interact and, furthermore, predict that inhi- (∼6 hours) induced activation of CD95 was due, in part, bitors of survival signaling receptors, e.g., ERBB1 and to ceramide generation, and our present findings showed IGF1R, which are activated in response to sorafenib expo- that CD95 activation at this time point was insensitive to sure, will enhance the antitumor efficacy of this drug the neutralizing antibody. Multiple HDACIs have shown combination. an increase in the protein levels of FAS-L and/or CD95 (e.g., refs. 38, 47), and this has been suggested as one im- Disclosure of Potential Conflicts of Interest portant mechanism for the antitumor effects of HDACIs. One well-described additional action of HDACIs in tu- No potential conflicts of interest were disclosed. mor cells is to increase expression of the cyclin dependent kinase inhibitor p21Cip-1/WAF1/mda-6 (p21; refs. 20, 29, 43). Grant Support We noted that vorinostat, in contrast with our findings in PHS grants R01-DK52825, P01-CA104177, and R01-CA108520 (P. Dent); malignant blood cancer cells, modestly increased p21 PHS grants R01-CA63753, R01-CA77141, and R01-CA93738 (S. Grant); levels in hepatoma cells and that this effect was sup- and a Leukemia Society of America grant 6405-97. These studies were also funded in part by The Jimmy V Foundation. pressed in cells treated with vorinostat and sorafenib. The costs of publication of this article were defrayed in part by the Increased expression of p21 not only acts to suppress payment of page charges. This article must therefore be hereby marked tumor growth but, in some cell types, suppresses toxic advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. JNK pathway signaling and also facilitates CD95 activation (36, 48). Knockdown of p21 expression suppressed Received 03/19/2010; revised 05/24/2010; accepted 06/11/2010; vorinostat and sorafenib toxicity and suppressed drug published OnlineFirst 08/03/2010.

References 1. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. ing pathways in liver diseases. Springer Press; 2005, p. 223–8, CA Cancer J Clin 2005;55:74–108. Chapter 19. 2. Bilimoria KY, Bentrem DJ, Lillemoe KD, Talamonti MS, Ko CY. Pan- 4. Dent P, Yacoub A, Fisher PB, Hagan MP, Grant S. MAPK pathways creatic cancer quality indicator development expert panel, American in radiation responses. Oncogene 2003;22:5885–96. College of Surgeons. Assessment of pancreatic cancer care in the 5. Valerie K, Yacoub A, Hagan MP, et al. Radiation-induced cell signal- United States based on formally developed quality indicators. J Natl ing: inside-out and outside-in. Mol Cancer Ther 2007;6:789–801. Cancer Inst 2009;101:848–59. 6. Grant S, Dent P. Kinase inhibitors and cytotoxic drug resistance. Clin 3. Dent P. MAP kinase pathways in the control of hepatocyte growth, Cancer Res 2004;10:2205–7. metabolism and survival. In: Dufour JF, Clavien P-A, editors. Signal- 7. Allan LA, Morrice N, Brady S, Magee G, Pathak S, Clarke PR. Inhibition

2230 Mol Cancer Ther; 9(8) August 2010 Molecular Cancer Therapeutics

Src and Sorafenib

of caspase-9 through phosphorylation at Thr 125 by ERK MAPK. Nat gistically kill tumor cells via FLIP suppression and CD95 activation. Cell Biol 2003;5:647–54. Clin Cancer Res 2008;14:5385–99. 8. Mori M, Uchida M, Watanabe T, et al. Activation of extracellular signal- 30. Park MA, Zhang G, Martin AP, et al. Vorinostat and sorafenib in- regulated kinases ERK1 and ERK2 induces Bcl-xL up-regulation via crease ER stress, autophagy and apoptosis via ceramide-dependent inhibition of caspase activities in erythropoietin signaling. J Cell CD95 and PERK activation. Cancer Biol Ther 2008;7:1648–62. Physiol 2003;195:290–7. 31. Park MA, Zhang G, Mitchell C, et al. Mitogen-activated protein ki- 9. Ley R, Balmanno K, Hadfield K, Weston C, Cook SJ. Activation of the nase kinase 1/2 inhibitors and 17-allylamino-17-demethoxygeldana- ERK1/2 signaling pathway promotes phosphorylation and mycin synergize to kill human gastrointestinal tumor cells in vitro via proteasome-dependent degradation of the BH3-only protein, Bim. suppression of c-FLIP-s levels and activation of CD95. Mol Cancer J Biol Chem 2003;278:18811–6. Ther 2008;7:2633–48. 10. Wang YF, Jiang CC, Kiejda KA, Gillespie S, Zhang XD, Hersey P. 32. Yacoub A, Park MA, Gupta P, et al. Caspase-, cathepsin- and PERK- Apoptosis induction in human melanoma cells by inhibition of MEK dependent regulation of MDA-7/IL-24-induced cell killing in primary is caspase-independent and mediated by the Bcl-2 family members human glioma cells. Mol Cancer Ther 2008;7:297–313. PUMA, Bim, and Mcl-1. Clin Cancer Res 2007;13:4934–42. 33. Mitchell C, Park MA, Zhang G, et al. 17-Allylamino-17-demethoxy- 11. Qiao L, Han SI, Fang Y, et al. Bile acid regulation of C/EBPβ CREB, geldanamycin enhances the lethality of deoxycholic acid in primary c-Jun function, via the extracellular signal-regulated kinase and rodent hepatocytes and established cell lines. Mol Cancer Ther c-Jun NH2-terminal kinase pathways, modulates the apoptotic 2007;6:618–32. response of hepatocytes. Mol Cell Biol 2003;23:3052–66. 34. Gupta S, Natarajan R, Payne SG, et al. Deoxycholic acid activates 12. Li N, Batt D, Warmuth M. B-Raf kinase inhibitors for cancer treat- the c-Jun N-terminal kinase pathway via FAS receptor activation in ment. Curr Opin Investig Drugs 2007;8:452–6. primary hepatocytes. Role of acidic sphingomyelinase-mediated 13. Davies BR, Logie A, McKay JS, et al. AZD6244 (ARRY-142886), a ceramide generation in FAS receptor activation. J Biol Chem potent inhibitor of mitogen-activated protein kinase/extracellular signal- 2001;279:5821–8. regulated kinase kinase 1/2 kinases: mechanism of action in vivo, 35. Qiao L, Studer E, Leach K, et al. Deoxycholic acid (DCA) causes pharmacokinetic/pharmacodynamic relationship, and potential for ligand-independent activation of epidermal growth factor receptor combination in preclinical models. Mol Cancer Ther 2007;6:2209–19. (EGFR) and FAS receptor in primary hepatocytes: inhibition of 14. Flaherty KT. Sorafenib: delivering a targeted drug to the right targets. EGFR/mitogen-activated protein kinase-signaling module enhances Expert Rev Anticancer Ther 2007;7:617–26. DCA-induced apoptosis. Mol Biol Cell 2001;12:2629–45. 15. Rini BI. Sorafenib. Expert Opin Pharmacother 2006;7:453–61. 36. Zhang G, Park MA, Mitchell C, et al. Multiple cyclin kinase inhibitors 16. Strumberg D. Preclinical and clinical development of the oral multi- promote bile acid-induced apoptosis and autophagy in primary he- kinase inhibitor sorafenib in cancer treatment. Drugs Today (Barc) patocytes via p53-CD95-dependent signaling. J Biol Chem 2008; 2005;41:773–84. 283:24343–58. 17. Gollob JA. Sorafenib: scientific rationales for single-agent and com- 37. Reinehr R, Häussinger D. Hyperosmotic activation of the CD95 sys- bination therapy in clear-cell renal cell carcinoma. Clin Genitourin tem. Methods Enzymol 2007;428:145–60. Cancer 2005;4:167–74. 38. Emanuele S, Lauricella M, Carlisi D, et al. SAHA induces apoptosis in 18. Rahmani M, Davis EM, Bauer C, Dent P, Grant S. Apoptosis induced hepatoma cells and synergistically interacts with the proteasome in- by the kinase inhibitor BAY 43-9006 in human leukemia cells involves hibitor Bortezomib. Apoptosis 2007;12:1327–38. down-regulation of Mcl-1 through inhibition of translation. J Biol 39. Huang J, Zheng DL, Qin FS, et al. Genetic and epigenetic silencing of Chem 2005;280:35217–27. SCARA5 may contribute to human hepatocellular carcinoma by ac- 19. Rahmani M, Nguyen TK, Dent P, Grant S. The multikinase inhibitor tivating FAK signaling. J Clin Invest 2010;120:223–41. sorafenib induces apoptosis in highly imatinib mesylate-resistant 40. Chung BM, Dimri M, George M, et al. The role of cooperativity with Src bcr/abl+ human leukemia cells in association with signal transducer in oncogenic transformation mediated by non-small cell lung cancer- and activator of transcription 5 inhibition and myeloid cell leukemia-1 associated EGF receptor mutants. Oncogene 2009;28:1821–32. down-regulation. Mol Pharmacol 2007;72:788–95. 41. Huynh H, Ngo VC, Koong HN, et al. AZD6244 enhances the anti-tumor 20. Dasmahapatra G, Yerram N, Dai Y, Dent P, Grant S. Synergistic in- activity of sorafenib in ectopic and orthotopic models of human hepa- teractions between vorinostat and sorafenib in chronic myelogenous tocellular carcinoma (HCC). J Hepatol 2010;52:79–87. leukemia cells involve Mcl-1 and p21CIP1 down-regulation. Clin 42. Nguyen TK, Jordan N, Friedberg J, Fisher R, Dent P, Grant S. Cancer Res 2007;13:4280–90. Inhibition of MEK/ERK1/2 sensitizes lymphoma cells to sorafenib- 21. Rahmani M, Davis EM, Crabtree TR, et al. The kinase inhibitor sora- induced apoptosis. Leuk Res 2010;34:379–86. fenib induces cell death through a process involving induction of en- 43. Rosato RR, Almenara JA, Maggio SC, et al. Role of histone deace- doplasmic reticulum stress. Mol Cell Biol 2007;27:5499–513. tylase inhibitor-induced reactive oxygen species and DNA damage in 22. Gregory PD, Wagner K, Horz W. Histone acetylation and chromatin LAQ-824/fludarabine antileukemic interactions. Mol Cancer Ther remodeling. Exp Cell Res 2001;265:195–202. 2008;7:3285–97. 23. Marks PA, Miller T, Richon VM. Histone deacetylases. Curr Opin 44. Tonks NK. Redox redux: revisiting PTPs and the control of cell sig- Pharmacol 2003;3:344–51. naling. Cell 2005;121:667–70. 24. Bali P, Pranpat M, Swaby R, et al. Activity of suberoylanilide hydro- 45. Gupta VA, Hermiston ML, Cassafer G, Daikh DI, Weiss A. B cells drive xamic acid against human breast cancer cells with amplification of lymphocyte activation and expansion in mice with the CD45 wedge her-2. Clin Cancer Res 2005;11:6382–9. mutation and Fas deficiency. J Exp Med 2008;205:2755–61. 25. Kwon SH, Ahn SH, Kim YK, et al. Apicidin, a histone deacetylase 46. Daigle I, Yousefi S, Colonna M, Green DR, Simon HU. Death recep- inhibitor, induces apoptosis and Fas/Fas ligand expression in human tors bind SHP-1 and block cytokine-induced anti-apoptotic signaling acute promyelocytic leukemia cells. J Biol Chem 2002;277:2073–80. in neutrophils. Nat Med 2002;8:61–7. 26. Pang RW, Poon RT. From molecular biology to targeted therapies for 47. Gillenwater AM, Zhong M, Lotan R. Histone deacetylase inhibitor hepatocellular carcinoma: the future is now. Oncology 2007;72 Suppl suberoylanilide hydroxamic acid induces apoptosis through both mi- 1:30–44. tochondrial and Fas (Cd95) signaling in head and neck squamous 27. Venturelli S, Armeanu S, Pathil A, et al. Epigenetic combination ther- carcinoma cells. Mol Cancer Ther 2007;6:2967–75. apy as a tumor-selective treatment approach for hepatocellular car- 48. Kim GY, Mercer SE, Ewton DZ, Yan Z, Jin K, Friedman E. The stress- cinoma. Cancer 2007;109:2132–41. activated protein kinases p38α and JNK1 stabilize p21(Cip1) by 28. Wise LD, Turner KJ, Kerr JS. Assessment of developmental toxicity phosphorylation. J Biol Chem 2002;277:29792–802. of vorinostat, a histone deacetylase inhibitor, in Sprague-Dawley rats 49. Park JS, Qiao L, Gilfor D, et al. A role for both Ets and C/EBP tran- and Dutch Belted rabbits. Birth Defects Res B Dev Reprod Toxicol scription factors and mRNA stabilization in the MAPK-dependent 2007;80:57–68. increase in p21 (Cip-1/WAF1/mda6) protein levels in primary hepato- 29. Zhang G, Park MA, Mitchell C, et al. Vorinostat and sorafenib syner- cytes. Mol Biol Cell 2000;11:2915–32.

www.aacrjournals.org Mol Cancer Ther; 9(8) August 2010 2231

Sorafenib Activates CD95 and Promotes Autophagy and Cell Death via Src Family Kinases in Gastrointestinal Tumor Cells

Margaret A. Park, Roland Reinehr, Dieter Häussinger, et al.

Mol Cancer Ther 2010;9:2220-2231. Published OnlineFirst August 3, 2010.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-10-0274

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2010/08/04/1535-7163.MCT-10-0274.DC1

Cited articles This article cites 48 articles, 23 of which you can access for free at: http://mct.aacrjournals.org/content/9/8/2220.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/9/8/2220.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://mct.aacrjournals.org/content/9/8/2220. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.