Selective inhibitors of MEK1/ERK and p38 mitogen- activated protein kinases potentiate apoptosis induction by sulindac sulfide in human colon carcinoma cells

The nonsteroidal anti-inflammatory drug (NSAID) sulindac prevents experimental colon cancer and can regress precancerous polyps in humans. Sulindac sulfide inhibits cyclooxygenase (COX)-mediated prostaglandin synthesis and retards the growth of cultured colon cell lines primarily by inducing apoptosis. Given the known role of mitogen-activated protein kinase (MAPK) in signal transduction and the regulation of cell survival and death, we determined the effect of sulindac sulfide on MAPK activation, COX-2 expression, and apoptosis induction in HCA-7 human colon cancer cells. Sulindac sulfide treatment was associated with activation of ERK and p38 MAPK in a dosageand time-dependent manner, and also activated upstream MEK. Similar results were seen in HCT-15 cells and also with the selective COX-2 inhibitor NS398. ERK and p38 activation were accompanied by an induction of COX-2 protein expression. Selective inhibitors of sulindac sulfide–induced ERK (PD98059) and p38 MAPK (SB203580) activation also suppressed the induction of COX-2 by this NSAID. Furthermore, both MAPK inhibitors significantly augmented sulindac sulfide– induced apoptosis, as did suppression of constitutive COX-2 using antisense oligonucleotides. In conclusion, MEK/ERK and p38 MAPK activation mediate COX-2 induction by sulindac sulfide. Selective inhibitors of these MAPKs potentiate apoptosis induction by this NSAID, suggesting a novel strategy for the prevention or treatment of colorectal cancer. [Mol Cancer Ther 2005;4(1):51–59] Introduction Epidemiologic studies have consistently shown that frequent and prolonged use of nonsteroidal anti-inflammatory drug (NSAID) can reduce the incidence of colorectal adenomas and carcinomas (reviewed in ref. 1). Furthermore, the NSAID sulindac has been shown to regress colorectal polyps in patients with familial adenomatous polyposis (2). Sulindac is a pro-drug that is metabolized in vivo to its sulfide and sulfone derivatives. Only the active sulfide derivative inhibits cyclooxygenase (COX) enzymes that regulate prostaglandin synthesis from arachidonic acid (3). The inducible COX-2 isoform is overexpressed in most human colorectal cancers and has been shown to modulate apoptosis, angiogenesis, and tumor cell invasiveness (reviewed in ref. 1). These effects may underlie its role in tumorigenesis. Sulindac sulfide has been shown to inhibit the growth of cultured colon cancer cells primarily by inducing apoptosis (4, 5). Whereas this drug is a potent inhibitor of COX enzymatic activity, its effect upon COX-2 protein expression remain unknown. COX-2 is regulated by transcriptional and post-transcriptional events (6). Evidence suggests that COX-2 induction can be mediated by mitogen-activated protein kinase (MAPK; refs. 7–11). In this regard, MAPK pathways [ERK, p38, and c-Jun NH2terminal kinase (JNK)] have been shown to be involved in COX-2 induction in a stimulusand cell type–specific manner. However, limited and conflicting data exist regarding the relationship of MAPKs with COX-2 induction in NSAID-treated tumor cells (12–14). Furthermore, the role of MAPK pathways in relation to NSAID-induced apoptosis remains unknown. The MAPK family is a well-characterized signaling pathway regulating cell survival and cell death (15). The MAPK family comprises extracellular signal-regulated kinase (ERK), p38 MAPK, and JNK. MAPK subfamilies can be activated by a variety of extracellular stimuli and recent data suggest that MAPKs may mediate apoptotic signaling induced by antineoplastic drugs (16). The ERK pathway is activated by various stimuli, including growth factors (17), lipopolysaccharide (18), and chemotherapeutic agents (19, 20). Microtubule or actin-interfering agents were found to stimulate MAPK signaling and AP-1 activity, which led to increased COX-2 expression in mammary epithelial cells (21). ERK activation is associated with cell proliferation, differentiation, and cell survival. In contrast, p38 MAPK and JNK/stress-activated protein kinase are activated in response to cellular stress (induced by UV irradiation, chemotherapeutic agents, and proinflammatory cytokines) and are involved in growth arrest and, in some contexts, apoptosis (22–24). Received 8/25/04; revised 11/4/04; accepted 11/16/04. Grant support: National Colorectal Cancer Alliance and NIH R01 grant DK56378. 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. Requests for reprints: Frank A. Sinicrope, Mayo Clinic, 200 First Street Southwest, Rochester, MN 55905. E-mail: [email protected] Copyright C 2005 American Association for Cancer Research. Molecular Cancer Therapeutics 51 Mol Cancer Ther 2005;4(1). January 2005 on July 12, 2017. © 2005 American Association for Cancer Research. mct.aacrjournals.org Downloaded from We (25) and others (26) have shown that sulindac sulfide can increase the level of membrane death receptor 5 and can activate caspase-8 to engage the extrinsic apoptotic pathway. Furthermore, sulindac sulfide can trigger the intrinsic mitochondrial apoptotic pathway with cytosolic release of cytochrome c (27). Studies have also shown that sulindac can induce apoptosis in vivo. In this regard, inhibition of intestinal tumorigenesis in animal models is accompanied by increases in the extent of enterocyte apoptosis (28, 29). Prolonged treatment with sulindac or celecoxib are also associated with modulation of apoptosis in colorectal epithelial from familial adenomatous polyposis patients (30, 31). However, the signal transduction pathways used by sulindac to trigger apoptosis, including the potential role of MAPKs, remain unknown. Whereas sulindac inhibits COX enzymatic activity, several studies have shown that NSAIDs exert their antiproliferative and proapoptotic effects by a COX-independent mechanism (1, 27, 32). In this study, we examined the role of ERK and p38 MAPKs in the regulation of sulindac sulfide–induced COX2 expression in human colon cancer cells. The influence of MAPK signaling upon apoptosis induction by this NSAID was also studied. Specifically, we determined the effects of selective inhibitors of MEK1/ERK (PD98059) and p38 (SB203580) upon alterations in cell viability and apoptosis induced by sulindac sulfide. Materials andMethods Cell Culture HCA-7 cells (obtained from S. Kirkland, United Kingdom) and HCT-15 (American Type Culture Collection, Gaithersburg, MD) human colon cancer cell lines were grown in RPMI 1640 (Life Technologies, Gaithersburg, MD), supplemented with 10% heat-inactivated fetal bovine serum, 100 IU/mL penicillin, 100 Ag/mL streptomycin and maintained at 37jC in a humidified atmosphere of 95% air and 5% CO2. HCA-7 cells are known to produce significant quantities of prostaglandin E2 (33). Cells were seeded at a density of 3 10 cells/100-mm dish and incubated in serum-free medium with 1 mg bovine serum albumin/mL for 24 hours before drug treatment. Cells were treated with sulindac sulfide (Sigma, St. Louis, MO) at the specified times and dosages (0–160 Amol/L). The drug was dissolved in 100% DMSO and then diluted for experiments with the final concentration of DMSO maintained at 0.1%. Cells were also treated with NS398 (Cayman Chemical, Ann Arbor, MI) for 3 hours. After 24 hours, fresh medium was added containing drug(s) for analysis of cell viability or apoptosis. For experiments using the p38 MAP/ERK kinase (MEK) inhibitor SB203580 (Calbiochem, La Jolla, CA) or the MEK1 inhibitor, PD098059 (Calbiochem), HCA-7 cells were maintained in serum-free media for 24 hours, then treated with either inhibitor for the indicated dosages and times alone or before the addition of sulindac sulfide. 3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazolium BromideAssay Cells were seeded in 96-well plates at a density of 2 10 and allowed to attach overnight. The cells were then treated with sulindac sulfide (120 Amol/L) for 48 hours alone or after preincubation for 20 minutes with PD98059 or SB203580 (both at 2.5–10 Amol/L). Medium was then replaced with 100 AL of 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide solution (Sigma) at a final concentration of 1 mg per mL per well. After incubation at 37jC for 4 hours, the supernatant was aspirated, and 100 AL of 10% SDS was added to each well and incubated for 8 hours at 37jC. Absorbance was measured at wavelength of 75 nm using a microplate reader. Terminal Deoxynucleotidyl Transferase ^Mediated Nick End Labeling Assay Cultured cells were treated for 48 hours with sulindac sulfide (0–160 Amol/L). DNA strand breaks, consistent with apoptosis, were detected by the terminal deoxynucleotidyl transferase–mediated nick end labeling (TUNEL) assay (In situ Cell Death Detection kit, Roche Diagnostics, Mannheim, Germany). In separate experiments, cells were incubated with sulindac sulfide (120 Amol/L) in the presence or absence of PD98059 (0 – 10 Amol/L) or SB203580 (0–10 Amol/L). Both floating and attached cells were harvested and then fixed using 4% paraformaldehyde in PBS (pH 7.4) for 1 hour at room temperature. Cells were resuspended in 100 AL of 0.1% Triton X-100 in 0.1% sodium citrate for 2 minutes on ice for permeabilization and then labeled with fluorescein dUTP and stained with propidium iodide, according to the manufacturer’s instructions. Thereafter, the cells were analyzed for apoptosis using a FACScan flow cytometer (Epics Profile; Coulter Co., Hialeah, FL). Immunoblot of MAPKPhosphorylation Cells were washed with cold PBS and suspended in a lysis buffer [20 mmol/L HEPES (pH 7.5), 100 mmol/L NaCl, 10% glycerol, 1% Triton X-100, 1 mmol/L EGTA (pH 7.5), 0.1 mmol/L Na3VO4,1 Ag/mL aprotinin, 1 Ag/ mL leupeptin, and 1 mmol/L phenylmethylsulfonyl fluoride] on ice for 20 minutes, then cell debris was removed by centrifugation at 16,000 rpm for 15 minutes. The protein concentration in cell lysates was determined using a Bradford protein assay (Bio-Rad, Richmond, CA). Samples containing 50 Ag of protein were added to SDSPAGE loading buffer with 5% h-mercaptoethanol, heated to 95jC for 5 minutes, and loaded onto 10% polyacrylamide gels. Proteins were then transferred to nitrocellulose membranes. After blocking the blot with 5% milk in PBS containing 0.1% Tween 20, the membrane was incubated for 1 hour at room temperature with an anti-COX-2 (1:1,000; Cayman Chemical) and anti-h-actin (1:3,000; Sigma) monoclonal antibodies. Membranes were incubated with highly specific antibodies against total MEK (1:1,000), ERK (1:2,000) or p38 (1:1,000). For analysis of phosphorylated MAPKs, membranes were blocked with 5% nonfat milk in TBST [10 mmol/L Tris (pH 7.4), 100 mmol/L NaCl, 0.5% Tween 20] and then probed for 1 hour with phosphospecific antibodies against p-MEK (1:1,000), phospho-ERK Thr202/Tyr204 (1:1,000), phospho-p38 MAPKs Mediate Sulindac Sulfide– Induced COX-2 Expression 52 Mol Cancer Ther 2005;4(1). January 2005 on July 12, 2017. © 2005 American Association for Cancer Research. mct.aacrjournals.org Downloaded from (1:1,000) for 1 hour (all MAPK antibodies from Cell Signaling, Beverly, MA). Blots were then incubated with secondary antibodies conjugated with peroxidase (Bio-Rad) for 1 hour at room temperature. The signal was detected by chemiluminescence using the enhanced chemiluminescence detection kit (Amersham, Arlington Heights, IL). COX-2 Antisense Oligonucleotides HCA-7 cells (3 10) in 2 mL of RPMI 1640 were plated in six-well tissue culture plates (Corning Costar Co., Cambridge, MA). Twenty-four hours later, cells were cotransfected with 20 Ag COX-2 antisense oligodeoxynucleotide [(GGAAACATCGACAGT); Sigma Genosys; TX] and 10 Ag of pcDNA1 vector containing a neomycinresistant marker (Invitrogen Co., Carlsbad, CA) using 100 AL LipofectAMINE reagent according to the vendor’s instructions. Sixteen hours posttransfection, the cells were selected in RPMI 1640 containing 500 Ag/mL geneticin (Life Technologies) and 1 mg bovine serum albumin/mL for 24 hours. Cell were incubated in the presence or absence of sulindac sulfide (80 Amol/L) for 48 hours. Cells were then harvested and analyzed for COX-2 expression by immunoblotting or apoptosis by the TUNEL assay. Results Sulindac Sulfide Activates MEK, ERK, and p38 MAPKs Using antibodies that recognize their phosphorylated forms, we detected a low level of MEK, ERK, and p38 MAPK activation in serum starved HCA-7 human colon cancer cells, consistent with their constitutive activation (Fig. 1A–C). Treatment of these cells with sulindac sulfide (0–120 Amol/L) for up to 24 hours produced early and sustained activation of ERK (p42 and p44) that was detected at 15 minutes (Fig. 1A and B) and persisted at 3 and 24 hours following incubation with drug (Fig. 2). Activation occurred at all doses of sulindac sulfide tested and at all time points. Total ERK levels (unphosphorylated and phosphorylated) were essentially unchanged at 15 minutes, 3, and 24 hours time points (Figs. 1A and 2). Activation of p38 MAPK by sulindac sulfide occurred at 15 minutes at a threshold dose of 80 Amol/L, whereas 40 Amol/L was effective at longer incubations (3 and 24 hours; Figs. 1B and 2). Total p38 levels were unchanged to slightly increased at 24 hours. Similar to sulindac sulfide, the selective COX-2 inhibitor NS398 activated phospho-ERK and p38 in HCA-7 cells (Fig. 1D). In a separate experiment, we determined whether sulindac sulfide could alter MEK phosphorylation, the upstream activator of ERK. Sulindac sulfide treatment produced an early activation of MEK that was maximal at a 15 minutes and sustained at 3 hours (Fig. 1C). Total MEK levels were unaffected by drug treatment. This result indicates that sulindac sulfide activates MEK upstream of ERK. To exclude the possibility that the effects upon MAPKs were cell line dependent, HCT-15 colon cancer cells were treated with sulindac sulfide and the selective COX-2 inhibitor NS398. In HCT-15 cells, both NSAIDs activated ERK and the increase in phosphorylation was greater with sulindac sulfide than with NS398, although similar levels of p38 MAPK activation were observed (Fig. 1D). NSAID treatment did not alter the levels of ERK1 or total p38 proteins. Sulindac Sulfide InducesCOX-2ExpressionbyActivation of ERK and p38MAPKs Treatment of serum starved HCA-7 cells with sulindac sulfide (0–120 Amol/L) was found to induce COX-2 Figure 1. Sulindac sulfide activates MEK, ERK, and p38 MAPK in human colon carcinoma cells. Cultured HCA-7 cells were treated with sulindac sulfide at the indicated concentrations and times. Cells were then harvested and lysed, and equal aliquots of extracted protein were analyzed for phospho-ERK and total ERK (A), phospho-p38 and total p38 (B), and phospho-MEK and total MEK (C) expression by immunoblotting. Incubation of cells with sulindac sulfide induces an early, dose-dependent phosphorylation of ERK and p38 MAPK without affecting total ERK or p38 levels. Phosphorylation of upstream MEK by sulindac sulfide was time dependent and maximal at 15 min. after incubation with drug. Total MEK levels were unchanged by treatment. D, sulindac sulfide (80 Amol/L) and the selective COX-2 inhibitor NS398 (200 Amol/L) activate ERK and p38 MAPK in HCA-7 cells and in the COX-2–deficient HCT15 colon cancer cell line at 3 h. Molecular Cancer Therapeutics 53 Mol Cancer Ther 2005;4(1). January 2005 on July 12, 2017. © 2005 American Association for Cancer Research. mct.aacrjournals.org Downloaded from protein expression in a dosageand time-dependent manner (Fig. 2). COX-2 induction occurred at 3 hours postincubation with drug and was sustained at 24 hours for all doses relative to untreated cells (Fig. 2). In contrast to sulindac sulfide, treatment of HCA-7 cells with 5-fluorouracil or cisplatin failed to induce COX-2 expression (data not shown). On these same blots, phosphorylated and total ERK and p38 MAPK expression were analyzed. As shown in Fig. 2, activation of these MAPKs accompanied COX-2 induction by sulindac sulfide. To confirm that COX2 induction was indeed mediated by MAPKs, we used the MEK1 inhibitor PD098059 and the p38 MAPK inhibitor SB203580. Initially, we verified that these MAPK inhibitors could block sulindac sulfide–induced ERK and p38 activation. As shown by immunoblotting, pretreatment of HCA-7 cells with PD98059 (2.5–10 Amol/L) or SB203580 (2.5–10 Amol/L) resulted in a dose-dependent inhibition of sulindac sulfide–induced ERK or p38 activation, respectively (Fig. 3A and B). Moreover, both PD098059 and SB203580 were found to inhibit sulindac sulfide–induced COX-2 expression (3 hours, Fig. 3C). PD98059 was a more potent inhibitor of sulindac sulfide–induced COX-2 than was SB203580, yet both inhibitors at 10 Amol/L restored COX-2 to its level of constitutive expression. Together, these data show that ERK and p38 MAPK activation mediate the induction of COX-2 expression by sulindac sulfide in HCA-7 cells. As shown above, treatment of HCA-7 cells with sulindac sulfide induced COX-2 protein expression via activation of MAPKs. Forced expression of COX-2 has been associated with apoptosis resistance in vitro (25, 27, 34). Therefore, we determined the effect of COX-2 suppression upon the apoptotic susceptibility of HCA-7 cells to sulindac sulfide. HCA-7 cells were transfected with COX-2 antisense oligodinucleotides (COX-2 AS ODN). COX-2 AS ODN blocked both constitutive and sulindac sulfide–induced COX-2 expression, as shown by immunoblotting (Fig. 4A). Furthermore, the dose-dependent induction of apoptosis (TUNEL assay) by sulindac sulfide was significantly augmented in HCA-7 cells transfected with the COX-2 AS ODN, compared with untransfected cells (Fig. 4B). These data suggest that constitutive COX-2 expression in HCA-7 cells confers resistance to sulindac sulfide–induced apoptosis. This finding is consistent with our previous data demonstrating that ectopic COX-2 expression in the COX-2 – deficient HCT-15 cell line attenuates mitochondrial and death receptor–mediated apoptotic signaling triggered by sulindac sulfide (25, 27). Given that sulindac sulfide is a potent inhibitor of COX Figure 2. COX-2 induction accompanies the activation of ERK and p38 MAPK by sulindac sulfide. A marked induction of COX-2 (72 kDa) protein expression is seen at 3 and 24 h postincubation with sulindac sulfide. Whereas COX-2 induction is modulated by drug dose at 3 h, a high level of sustained induction is observed at 24 h, irrespective of dose. ERK and p38 MAPK activation accompany COX-2 induction at both time points without appreciable changes in total ERK or p38. Phospho-p38 MAPK activation is dose dependent at both time points. Figure 3. Pharmacologic inhibitors of MEK/ERK (PD98059) and p38 MAPK (SB203580) inhibit sulindac sulfide– induced ERK and p38 activation, as well as COX-2 expression. A and B, HCA-7 cells were serum starved for 24 h, then treated with the indicated concentrations of PD98059 or SB203580 for 35 min. Cells were then incubated with sulindac sulfide at the indicated concentrations for 15 min before harvesting. Cell lysates were analyzed by immunoblotting using ERK or p38 antibodies to detect phosphorylated and unphosphorylated proteins. PD98059 (A) and SB203580 (B) dose-dependently inhibited sulindac sulfide– induced ERK and p38 phosphorylation, respectively. C, both PD98059 and SB203589 were found to inhibit sulindac sulfide– induced COX-2 expression in a dose-dependent manner. b-actin, control for protein loading. MAPKs Mediate Sulindac Sulfide– Induced COX-2 Expression 54 Mol Cancer Ther 2005;4(1). January 2005 on July 12, 2017. © 2005 American Association for Cancer Research. mct.aacrjournals.org Downloaded from enzymatic activity and prostaglandin synthesis, the inhibitory effect of COX-2 upon apoptotic susceptibility is prostaglandin independent. These in vitro data suggest that COX-2 protein expression may contribute to tumor cell resistance to apoptotic doses of NSAIDs and accordingly, may reduce the effectiveness of these drugs for colorectal cancer prevention and treatment. Pharmacologic Inhibitors of ERK and p38 Activation Potentiate Sulindac Sulfide ^ Induced Growth Inhibition and Apoptosis NSAIDs, including sulindac sulfide, inhibit the growth of cultured human colon cancer cells mainly through induction of apoptosis (4, 5). However, the signaling pathways involved in NSAID-induced apoptosis remain unclear. We determined the role of ERK and p38 MAPK activation in mediating sulindac sulfide–induced growth inhibition and apoptosis induction in HCA-7 cells. The effect of sulindac sulfide (0–160 Amol/L) upon HCA-7 cell viability was determined by the 3-(4,5-dimethylthiazol-2yl)-2, 5-diphenyltetrazolium bromide assay. Next, cells were incubated with the MEK1/ERK inhibitor PD098059 or the p38 inhibitor SB203580 alone or in combination with sulindac sulfide (80 Amol/L) for 48 hours (Fig. 5). Whereas PD98059 (2.5–10 Amol/L) did not reduce cell viability, SB203580 (2.5–10 Amol/L) reduced the viable cell population by 10% and this effect was maximal at the lowest dose studied. At a dose up to 80 Amol/L, sulindac sulfide alone did not appreciably alter cell viability; however, its combination with PD98059 or SB203580 resulted in a significant dose-dependent reduction in cell viability compared with either drug alone (Fig. 5). Evidence indicates that both ERK and p38 can mediate extracellular stimulus-induced apoptosis (24, 35, 36), although these MAPKs have been shown to exert both antiapoptotic and proapoptotic effects depending Figure 4. COX-2 antisense oligonucleotides inhibit endogenous and stimulated COX-2 expression and increase the apoptotic susceptibility of HCA-7 cells to sulindac sulfide. A, transfection of HCA-7 cells with COX-2 antisense oligodeoxynucleotide (COX-2 AS ODN ) was shown to suppress constitutive and sulindac sulfide– induced COX-2 protein expression. B, HCA-7 cells transfected with COX-2 AS ODN show a significant increase in their apoptotic susceptibility to sulindac sulfide relative to untransfected cells. Apoptosis was determined by the TUNEL assay. Points, means of triplicate observations; bars, FSD. Figure 5. Inhibition of MAPKs potentiates the reduction in cell viability induced by sulindac sulfide. A, sulindac sulfide (4) treatment (48 h) reduced HCA-7 cell viability in a dose-dependent manner, as shown using the MTT assay. B and C, cells were pretreated (20 min) with PD98059 (o) or SB203580 ( w ) at the indicated concentrations alone and in combination with sulindac sulfide (80 Amol/L; .,y) for 48 h. Neither PD98059 nor sulindac sulfide reduced cell viability alone (B); however, their coadministration produced a significant reduction in cell viability. SB203580 reduced cell viability as a single agent and this effect was significantly augmented by the addition of sulindac sulfide (C). Points, means for triplicate experiments; bars, SD. Molecular Cancer Therapeutics 55 Mol Cancer Ther 2005;4(1). January 2005 on July 12, 2017. © 2005 American Association for Cancer Research. mct.aacrjournals.org Downloaded from upon the cellular context (19, 37–40). HCA-7 cells were pretreated with increasing concentrations of PD98059 or SB203580 alone or in combination with sulindac sulfide (120 Amol/L, 48 hours). By themselves, these inhibitors exerted only a modest proapoptotic effect as shown by the TUNEL assay (Fig. 6). Pretreatment of cells with PD98059 or SB203580 augmented sulindac sulfide–induced apoptosis f3-fold compared with that produced by sulindac sulfide alone (Fig. 6). PD98059 was a more potent modulator of sulindac sulfide–induced apoptosis than was SB203580. These data indicate that sulindac sulfide– induced activation of ERK and p38 MAPKs promote the survival of HCA-7 cells. As shown here, selective inhibitors of these MAPKs potentiate sulindac sulfide– induced growth inhibition and apoptosis induction.