Combination of atorvastatin and celecoxib ... - Wiley Online Library

Int. J. Cancer: 122, 2115–2124 (2008) ' 2008 Wiley-Liss, Inc.

Combination of atorvastatin and celecoxib synergistically induces cell cycle arrest and apoptosis in colon cancer cells Hang Xiao1, Qiang Zhang1, Yong Lin2,3, Bandaru S. Reddy1,3 and Chung S. Yang1,3* 1 Susan Lehman Cullman Laboratory for Cancer Research, Department of Chemical Biology, Ernest Mario School of Pharmacy, Rutgers, The State University of New Jersey, Piscataway, NJ 2 Department of Biostatistics, School of Public Health, University of Medicine and Dentistry of New Jersey, Piscataway, NJ 3 Cancer Institute of New Jersey, New Brunswick, NJ Previous studies in animal models have shown enhanced efficacy of a combined treatment of statins and Nonsteroidal anti-inflammatory drugs against colorectal cancer development. In our study, we investigated the combinational effects of atorvastatin and celecoxib in 2 human colon cancer cell lines HCT116 and HT29. Celecoxib moderately inhibited the growth of both cell lines with a similar IC50 of 40–50 lM, whereas atorvastatin showed stronger growth inhibitory effect in HCT116 cells than in HT29 cells (IC50 of 5–8 lM vs. 30–35 lM) after treatment for 48–72 hr. The combination of these 2 agents produced strong synergistic actions, as determined by isobologram analysis. Flow cytometry analysis indicated that the combination treatment for 24 hr caused extensive cell cycle arrest in G0/G1 phase; whereas at 48 hr or longer, apoptosis was induced significantly. The effects produced by the combination were much stronger than that by atorvastatin or celecoxib alone. Our results further demonstrated that the combinational effects of atorvastatin/celecoxib were associated with increased levels of p21Cip1/Waf1, p27Kip1, and phospho-JNK; decreased levels of phospho-AKT and hyper-phosphorylated Rb; and activation of caspase cascade. Atorvastatin/celecoxib combination also selectively modified membrane localization of small Gproteins, such as RhoA, RhoB and RhoC, which may contribute to the anti-cancer effects. Taken together, the results demonstrated a strong synergy between the actions of atorvastatin and celecoxib in growth inhibition and killing of human colon cancer cells. The present work suggests the possible therapeutic application of this combination and provides leads for mechanistic and biomarker investigations in clinical trials. ' 2008 Wiley-Liss, Inc. Key words: atorvastatin; celecoxib; combination; synergy; Rho; colorectal cancer

Colorectal cancer is a leading cause of cancer death in the Western countries,1 and its incidence is increasing globally as the dietary patterns and life styles of the industrialized Western countries are adopted worldwide.2,3 The relatively slow progression of adenomatous polyps to colorectal cancer offers a great opportunity for colorectal cancer chemoprevention and treatment of premalignant lesions. Nonsteroidal anti-inflammatory drugs (NSAIDs) have been a subject of intensive investigations for their anticarcinogenic effect on colorectal cancer. Celecoxib is a selective cyclooxyhenase-2 (COX-2) inhibitor believed to have fewer side effects than nonselective traditional NSAIDs.4,5 The recent placebo-controlled Adenoma Prevention with Celecoxib (APC) and Prevention of Colorectal Sporadic Adenomatous Polyps (PreSAP) trials have provided convincing evidence that the use of celecoxib among patients with colonic adenomas reduces the risk of metachronous adenomas.6,7 However, the relative high dose required for the observed chemopreventive effect may discourage the singular use of celecoxib on a long-term basis, because of possibly increased risk in cardiovascular events.8,9 Statins are small-molecule inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, which are used as cholesterol lowering drugs. The outstanding efficacy in cardiovascular disease prevention and the relative safety of the statins have resulted in their widespread use and the recent conversion from prescription to over-the-counter drug in the United Kingdom.10 In 2005, atorvastatin (Lipitor) was the second most commonly prescribed drugs in the United States with over 63 million prescripPublication of the International Union Against Cancer

tions written (Drug fact from www.Rxlist.com). Although results from many preclinical studies have demonstrated potential anticancer effect of statins against different types of cancers, especially colorectal cancer, the clinical and epidemiological studies have been inconclusive.10–12 There is a growing body of evidence suggesting that combination of low dose of cancer chemopreventive and therapeutic agents with different modes of action may produce synergistic effects on efficacy and minimize possible side effects associated with high dose administration. In cell line studies, NSAIDs or statins have shown synergistic effects in combination with other therapeutics, such as PPARg ligands, inhibitors of the EGFR family, TRAIL receptor ligands, cisplatin and doxorubicin.13,14 Some of these observations have been confirmed in animal models.13,14 Recently, combinations of NSAIDs and statins have also been subjects of several in vivo studies for intestinal cancer. In an azoxymethane (AOM)-induced colon carcinogenesis model, combination of lovastatin and sulindac produced stronger inhibition on colonic aberrant crypt foci (ACF) than did either agent alone.15 Using the same AOM-rat model, Reddy et al. showed that combination of low dose of atorvastatin and celecoxib decreased adenocarcinoma multiplicity by 90%, and this low-dose combination was more effective than a high dose of either agent alone.16 In another study using the APCMin/1 mouse model, combination of low dose of atorvastatin and celecoxib suppressed colon polyps completely and small intestinal polyps by more than 86%, and the combination significantly increased chemopreventive efficacy of atorvastatin and celecoxib.17 The aforementioned promising findings have suggested the possible use of the combination of statins and NSAIDs in human for colorectal cancer chemoprevention and treatment. However, the mode of interaction between the 2 agents is poorly understood. In our study, we used 2 human cancer cell lines to demonstrate the synergistic interaction between atorvastatin and celecoxib, and elucidate the mechanisms underlying the synergy. Material and methods Cell culture and drug treatment Human colorectal cancer cell lines HCT116 and HT29 were obtained from American Type Cell Collection (ATCC, Manassas, Abbreviations: ACF, aberrant crypt foci; AKT, protein kinase B; AOM, azoxymethane; COX, cyclooxygenase; EGFR, epidermal growth factor receptor; FPP, farnesyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; HMG-CoA, 3-hydroxy-3-methylglutaryl coenzyme A; JNK, c-Jun NH2-terminal kinase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NSAIDs, nonsteroidal anti-inflammatory drugs; TRAIL, TNF-related apoptosis-inducing ligand. Grant sponsor: NIH; Grant numbers: CA56673, CA120915, CA37663; Grant sponsor: Cancer Center; Grant number: CA72720; Grant sponsor: NIEHS Center; Grant number: ES05022. *Correspondence to: Department of Chemical Biology, Ernest Mario School of Pharmacy, Rutgers, The State University of New Jersey, 164 Frelinghuysen Road, Piscataway, NJ 08854, USA. Fax: 1732-445-0687. E-mail: [email protected] Received 30 May 2007; Accepted after revision 19 October 2007 DOI 10.1002/ijc.23315 Published online 2 January 2008 in Wiley InterScience (www.interscience. wiley.com).

2116

XIAO ET AL.

VA), and were maintained in McCoy’s 5A media (ATCC, Manassas, VA) supplemented with 10% heat inactivated FBS (Mediatech, Herndon, VA), 100 units/mL of penicillin, and 0.1 mg/mL of streptomycin (Sigma-Aldrich) at 37°C with 5% CO2 and 95% air. Human intestinal epithelial cell line INT407 were obtained from American Type Cell Collection (ATCC, Manassas, VA), and were maintained in RPMI1640 (Mediatech, Herndon, VA) supplemented with 10% heat inactivated FBS (Mediatech, Herndon, VA), 100 units/mL of penicillin, and 0.1 mg/mL of streptomycin (Sigma-Aldrich) at 37°C with 5% CO2 and 95% air. Cells were kept sub-confluent and media were changed every other day. All cells used were between 3 and 30 passages. Atorvastatin and celecoxib were from LKT Laboratories (St. Paul, MN). DMSO was used as the vehicle to deliver drugs, and the final concentration of DMSO in all experiments was 0.1%. Cell viability assay HCT116 (2,000 cells/well), HT29 (5,000 cells/well) and INT407 (2,000 cells/well) cells were seeded in 96-well plates. After 24 hr, cells were treated with serial concentrations of atorvastatin, celecoxib and their combinations in 200 lL of serum complete media. At 24, 48, and 72 hr after treatments, cells were subject to 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Media were replaced by 100 lL fresh media containing 0.5 mg/mL of MTT (Sigma-Aldrich). After 1 hr incubation at 37°C, MTT-containing media were removed and the reduced formazan dye was solubilized by adding 100 lL of DMSO to each well. After gentle mixing, the absorbance was monitored at 550 nm using a plate reader (TECAN, Phenix Research Products, Candler, NC). Analyses of synergy The analyses were based on the Chou and Talalay’s empirical method18 with modifications. It was assumed that the dose response model follows log [E/(1 2 E)] 5 a(log d 2 log Dm), which is a linear regression model with the response log [E/(1 2 E)] and the regressor log (d). This model is used for Drug 1, Drug 2 and the combinations of the 2 drugs with fixed ratio of the doses of the 2 drugs. E is fraction of cell survived, d is the dose applied, Dm is the median effective dose of a drug (IC50), and a is a slope parameter. On the basis of this regression model, median effect plots were constructed using data from MTT assay. Suppose that the combination (d1, d2) elicits the same effect x as Drug 1 alone at dose level Dx,1, and Drug 2 alone at dose Dx,2, then the interaction index 5 d1/Dx,1 1 d2/Dx,2 (Dx,1 and Dx,2 were calculated from median effect models). The interaction index was used to determine additivity, synergy, or antagonism of the combination at dose (d1, d2) depending on interaction index 51, 1, respectively. The delta method was used to calculate the variance of the interaction index, which is given by var(interaction index) 5 var(Dx,1)(d12/Dx,14) 1 var(Dx,2)(d22/Dx,24). Data were analyzed by R program. Cell cycle analyses HCT116 (8 3 104 cells/well) and HT29 (1 3 105 cells/well) cells were seeded in 6-well plates. After 24-hr incubation for attachment, cells were treated with different concentrations of atorvastatin, celecoxib and their combinations. After another 24 or 48 hr, media were collected and combined with adherent cells that were detached by brief trypsinization (0.25% trypsin-EDTA; Sigma-Aldrich). Cell pellets were washed with 1 mL of ice-cold PBS and then resuspended in 0.5 mL of 70% ethanol overnight. After centrifugation (1,500g, 3 min), the supernatant was removed and cells were incubated with 0.5 mL of PBS containing 50 lg RNAse (Sigma-Aldrich) and 5 lg propidium iodine (SigmaAldrich) for 30 min at room temperature. Single-cell suspension was generated by gentle pipetting. Cell cycle was analyzed using a Beckman Coulter flow cytometer (FC500) at the analytical cyto-

metry core facility on our campus, and data were processed using AXP acquisition and analysis software. Detection of apoptosis Apoptotic cells were quantified by Annexin V/PI double staining assay. Annexin V/PI staining was done using apoptotic detection kit (Zymed, South San Francisco, CA) following manufacture’s instruction. Cells were gently detached by brief trypsinization (any floating cells were also collected), and then washed with ice cold PBS. Cells (7 3 104) were then suspended in 200 lL binding buffer containing Annexin V, and incubated for 10 min at room temperature. After centrifugation (1,500g, 1 min), cell pellet was resuspended in 200 lL binding buffer containing 5 lg/mL propidium iodide. Early apoptotic cells were detected as Annexin V positive/PI negative using a Beckman Coulter flow cytometer (FC500), and data were processed using AXP acquisition and analysis software. Membrane preparation Cell culture media were collected and spun down to recover any floating cells. Adherent cells were washed with ice-cold PBS, and then collected with cell-scrapers in the presence of buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 1 mM EGTA, 0.1 mM dithiothreitol, and protease inhibitor cocktail (Sigma-Aldrich) at 1 to 100 dilution).19 After combining with floating cells, cells were sonicated on ice 4 times at 5 sec each, and then centrifuged at 25,000g for 1 hr. The pellets were solublized with lysis buffer (cell signaling, Beverly, MA) supplemented with protease inhibitor cocktail (Sigma-Aldrich) (1:100) on ice for 30 min after thorough mixing. Membrane fraction was collected as supernatant after centrifugation at 10,000g for 10 min. Immunoblotting Cells were washed with ice-cold PBS, collected with cell-scrapers. The cells were combined with floating cells, if any, and incubated on ice for 10 min in lysis buffer (cell signaling, Beverly, MA) supplemented with cocktails of protease inhibitor (1:100), phosphotase inhibitor 1 (1:100) and phosphotase inhibitor 2 (1:100) (Sigma-Aldrich). Cell suspensions were then subject to sonication (5 sec, 3 times). After further incubation of 20 min, followed by centrifugation at 10,000g for 10 min, supernatants were collected. Proteins were quantified by BCATM protein assay kit (Pierce Biotechnology, Rockford, IL), and 20–50 lg of proteins were resolved by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane. After blocking, proteins of interest were probed using different antibodies at manufacturer’s recommended concentrations, and then visualized using Odyssey system (LI-COR, NE) after incubation with suitable IR-antibodies. Antibodies for AKT, phospho-AKT (Ser473), PDK1, phosphoPDK1 (Ser241), phospho-PTEN (Ser380), PI3K P85, phosphoPI3K p85 (Tyr458)p55 (Tyr199), JNK, phospho-JNK (Thr183/ Tyr185), c-Jun, phospho-c-Jun (Ser73), hyperphospho-Rb, Rb, cleaved caspase-3 (Asp175), caspase-8, caspase-9, and p27Kip1 were from Cell Signaling (Beverly, MA). Antibody for RhoA (26C4), RhoB (119), RhoC (C-16), ICAM-1 (G-5), PTEN, cleaved PARP and K-Ras (F234) were from Santa Cruz Biotechnology (Santa Cruz, CA). Antibody for p21Cip1/Waf1 and b-actin were from Upstate (Lake Placid, NY) and Sigma-Aldrich, respectively. Statistical analysis All data were presented as mean 6 S.D. Student’s t-test (Excel) was used to test the mean difference between 2 groups. Analyses of variance (ANOVA) model (http://faculty.vassar.edu/lowry/ anova1u.html) was used for the comparison of the differences among more than 2 groups. A 1% significant level was used for all the tests.

SYNERGY BETWEEN ATORVASTATIN AND CELECOXIB

2117

FIGURE 1 – Growth inhibitory effects of atorvastatin, celecoxib, and their combinations. Human colon cancer cells HCT116 and HT29 were treated with serial concentrations of atorvastatin (a) and celecoxib (b) for 24, 48, or 72 hr, and cell viability was measured by MTT assay. Combined treatment with atorvastatin and celecoxib potentiated the growth inhibitory effects after 24 hr of drug exposure (c). Data were shown as mean 6 SD (n 5 6), and viable cells of controls were used as 100%. *p < 0.01 (n 5 3) by Student’s t-test as compared to treatments by atorvastatin or celecoxib alone. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Results Time- and dose-dependent growth inhibition of HCT116 and HT29 cells by atorvastatin and celecoxib As shown in Figures 1a and 1b, treatment of atorvastatin and celecoxib produced time- and dose-dependent growth inhibition in both cell lines. HCT116 cells were more sensitive (IC50: 5–8 lM at 48–72 hr) to atorvastatin than HT29 cells (IC50: 30–35 lM at 48–72 hr), whereas, both cell lines responded similarly to celecoxib treatment (IC50: 40–50 lM at 48–72 hr). We next examined the effect of treatment with a combination of atorvastatin and celecoxib on these 2 cell lines. The combined treatment showed drastically increased growth-inhibitory effects on both cell lines as compared to the treatments with atorvastatin or celecoxib alone

(Fig. 1c). For example, in HT29 cells, the combined treatment with atorvastatin (20 lM) 1 celecoxib (25 lM) inhibited cell growth by 49% at 24 hr, whereas, only marginal or no inhibitory effects were produced by treatments with atorvastatin or celecoxib alone. Similarly, in HCT116 cells, treatment with atorvastatin (4 lM) 1 celecoxib (25 lM) for 24 hr produced much stronger growth inhibition than did the treatments with each agent alone. Synergistic interaction between atorvastatin and celecoxib The promising combinational effect observed above warranted further investigation on whether the enhanced growth inhibitory effect was additive or synergistic. We used the Chou and Talalay’s method for two-drug interaction18 to study the combination

2118

XIAO ET AL.

FIGURE 2 – Median effect and interaction index plots of atorvastatin, celecoxib, and their combinations in HCT116 (a), HT29 (b) and INT407 (c) cells. Cells were treated with atorvastatin or celecoxib alone or in combination at serial concentrations for 48 hr in 96-well plates. Cell viability was measured by MTT assay, and median effect plots and interaction index plots were constructed using Chou and Talalay’s method.18 Synergy was defined as interaction index lower than 1.0. Error bar in interaction index plots represented 95% confidence interval and were calculated using delta method (n 5 6). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

between atorvastatin and celecoxib. Based on IC50 of atorvastatin and celecoxib, we examined combination of atorvastatin/celecoxib at ratio of 1:6.25 and 1:1.25 in HCT116 and HT29 cells, respectively. The concentrations of atorvastatin in combined treatments tested were 0.125, 0.25, 0.5, 1.0, 2.0, 3.0, and 4.0 lM in HCT116 cells, and 2.5, 5.0, 7.5, 10.0, 12.5, 15.0, 17.5, and 20.0 lM in HT29 cells. The concentrations of celecoxib in combined treatments tested were correspondingly 6.25 and 1.25-fold higher than those of atorvastatin in HCT116 and HT29 cells, respectively. After 24, 48 and 72 hr of exposure to the series of treatments, growth inhibitory effects were measured by MTT assay, and results were analyzed in median effect plot using the linear regression model described in Method section.

The results from 48 hr treatment experiment were shown in Figure 2 as examples, and the results from 24 and 72 hr treatment experiments showed similar patterns. The median effect plots (Fig. 2) demonstrated that the linear regression model exploited herein well fitted the dose-response relationship of atorvastatin, celecoxib and their combinations in the concentration ranges used in the experiments. On the basis of the media effect plots, the interaction indexes of each combined dose pair of atorvastatin/celecoxib were calculated as described in the Methods section. The interaction index-effect plots (Figs. 2a and 2b) demonstrated that all combined dose pairs at the ratios tested produced interaction index lower than 1.0 for both HCT116 and HT29 cells. Delta analysis of variance of interaction index indicated that all results were statistically sig-


2119

FIGURE 3 – Effects of atorvastatin, celecoxib and their combinations on cell cycle. Cells were treated in 6-well plates for 24 hr before cell cycle analyses by flow cytometry. Combined treatment with atorvastatin and celecoxib potentiated G0/G1 arrest in HCT116 cell (a) and in HT29 cells (b). Western blots demonstrated the treatment effects on p21Cip1/Waf1, p27Kip1 hyper-phosphorylated Rb, and nonphosphorylated Rb (c). b-Actin served as an equal loading control. The results were representative of 3 experiments. *p < 0.01 (n 5 3) by Student’s t-test as compared to treatments by atorvastatin or celecoxib alone. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley. com.]

nificant with 95% confidence. We also tested atorvastatin/celecoxib combination in human intestinal epithelial INT407 cells. INT407 cells were treated with the same combination treatments as employed in HCT116 cells, and MTT assay results were analyzed by the isobologram method as described earlier. As shown in Figure 2c, the atorvastatin/celecoxib combination did not produce synergy at atorvastatin doses of 0.25, 0.5, 1, and 2 lM, and only showed synergy at atorvastatin doses of 3 and 4 lM. Taken together, our results demonstrated a synergistic interaction between atorvastatin and celecoxib in growth inhibition of both HCT116 and HT29 human colon cancer cell lines, and these 2 cancer cell lines were more sensitive to the synergy produced by atorvastatin and celecoxib, compared to the normal human intestinal epithelial INT407 cells. Combined treatment potentiates cell cycle arrest in G0/G1 phase caused by atorvastatin or celecoxib We studied mechanisms underlying the synergy between atorvastatin and celecoxib in growth inhibition in colon cancer cells.

Cell cycle analyses by flow cytometry were performed on both HCT116 and HT29 cells after 24 hr of drug exposure. In HCT116 cells, atorvastatin (4 lM) or celecoxib (25 lM) alone marginally increased cell population in the G0/G1 phase by 2.8% or 4.6%, respectively, as compared to control; however, treatment with a combination of atorvastatin and celecoxib caused a drastic increase in G0/G1 phase population by 21.7% (Fig. 3a). Similarly in HT29 cells, the combined treatment caused a much more extensive cell cycle arrest in G0/G1 phase than either agent alone at the same concentrations (Fig. 3b). These results demonstrated that atorvastatin and celecoxib acted synergistically to arrest colon cancer cells in G0/G1 phase. Cyclin dependent kinase (CDK) inhibitors, such as p21Cip1/Waf1 and p27Kip1, are negative regulators of cell cycle progression. They inhibit CDK activities, which causes hypo-phosphorylation of Rb, and in turn, the inhibition leads to cell cycle arrest. We analyzed the effects of the combined treatments on key cell cycle regulating proteins. It was found that, after 24 hr drug treatment when substantial G0/G1 phase arrest was seen (as described earlier),

2120

XIAO ET AL.

FIGURE 4 – Effects of treatments with atorvastatin and celecoxib on apoptosis. Cells were treated for 24 or 48 hr, and apoptosis was assayed by flow cytometry after Annexin-V/PI co-staining. (a) Annexin V/PI intensity dot plots of HCT116 cells showed significantly increased dot intensity in J4 region after combined treatment with atorvastatin/celecoxib for 48 hr. Cells located in J4 region (Annexin-V positive, PI negative) are considered as early apoptotic. (b) Quantification of early apoptosis after 48 hr of treatment with atorvastatin, celecoxib and their combination in HCT116 cells. *p < 0.01 (n 5 3) by Student’s t-test as compared to other groups. (c) Western blot demonstrated that combined treatments with atorvastatin/celecoxib for 48 hr activated caspase cascade. b-Actin served as an equal loading control. The results were representative of 3 experiments. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

atorvastatin and celecoxib in combination caused upregulation of p21Cip1/Waf1 and p27Kip1, decreased hyper-phosphorylated form of Rb, and increased non-phosphorylated form of Rb in both cell lines (Fig. 3c). Treatment with atorvastatin or celecoxib alone also decreased hyper-phosphorylation of Rb, but with much lesser extent as compared to the combined treatment. The effect of atorvastatin/celecoxib combination on key cell cycle regulating proteins was consistent with its effect on cell cycle in both cell lines. Our results showed that synergistic action of atorvastatin/cele-

coxib combination on G0/G1 phase cell cycle arrest in HCT116 and HT29 cells was through their synergistic action on upregulation of p21Cip1/Waf1, p27Kip1 and subsequent decrease of hyperphospho-Rb. Prolonged combined treatment induces apoptosis Results from growth inhibition assay showed that longer drug exposure period led to stronger inhibitory effects. Cell cycle


analyses on samples after longer treatment period (e.g., 48 hr) showed significant increases of cell population in sub-G0/G1 phase, which suggested possible apoptosis caused by drug exposure (data not shown). On the basis of these observations, we determined whether prolonged treatment induced apoptosis in HCT116 and HT29 cells. After drug treatment for 24 or 48 hr, cells were subject to Annexin V-FITC/PI co-staining assay. Annexin V positive and PI negative cells (J4 in Fig. 4a) were considered as early apoptotic, and annexin V/PI double positive cells (J2 in Fig. 4a) were categorized as late apoptotic and necrotic. At 24 hr, treatment with atorvastatin or celecoxib alone or in combination did not increase apoptotic cell population compared to controls in either HCT116 or HT29 cells (data not shown). However, at 48 hr, combined treatment with atorvastatin (4 lM) and celecoxib (25 lM) significantly increased dot intensity of J4 region compared to control (Fig. 4a). Quantification of dot intensity in J4 region showed that the combined treatment with atorvastatin and celecoxib induced a 7-fold increase in early apoptotic population over the control in HCT116 cells (Fig. 4b). In contrast, atorvastatin or celecoxib alone at same concentrations did not cause any appreciable change in early apoptotic population. In HT29 cells, similar results were observed on treatment with atorvastatin (20 lM) or celecoxib (25 lM) alone or in combination (data not shown). To further confirm the apoptosis-inducing effect of the combined treatment, Western blot analyses were performed to analyze caspase-3 activation (cleaved caspase-3 at 17/19 kDa), a hallmark of apoptosis. As shown in Figure 4c, there was no significant caspase-3 activation in all cells treated with either combination or single agents for 24 hr. However, caspase-3 was significantly activated in cells after prolonged exposure (48 hr) to the combined treatment with atorvastatin and celecoxib. After the same treatment period, there was no or little caspase-3 activation detected in cells treated with either agent alone. Moreover, our results also showed that at 48 hr the combination treatments significantly decreased full-length caspase-8 and full-length caspase-9 levels, and increased cleaved caspase-8, cleaved caspase-9, and cleaved PARP levels. The pattern of caspase activation by the combined treatment is consistent with Annexin V/PI co-staining analyses that showed apoptosis-inducing effect of the combined treatment after drug exposure of 48 hr. The results suggested that the synergistic action between atorvastatin and celecoxib on growth inhibition after prolonged exposure time (i.e., 48 hr) was due to apoptosis caused by the combined treatment of these 2 agents.

Combined treatment modifies membrane association of small G-proteins Many line of evidences have shown that the cancer preventive effects of statins were dependent on inhibition of prenylation of small G-proteins, primarily Rho proteins.10 Here we examined the effect of atorvastatin and celecoxib on membrane localization of important small G-proteins, K-Ras, RhoA, RhoB, and RhoC, in HCT116 and HT29 cells. Treatment with atorvastatin for 24 and 48 hr significantly decreased membrane bound RhoA, and its combination with celecoxib enhanced the effect (Fig. 5). In contrast, the total RhoA protein was increased by atorvastatin and again the effect was enhanced by celecoxib. The responses of RhoC to these treatments were similar to that of RhoA, although the signals were weaker and the changes were not as clear. Membrane bound RhoB, however, was significantly increased by treatment with atorvastatin alone or its combination with celecoxib. As to total RhoB, treatment of atorvastatin increased its expression, and celecoxib potentiated the increase, especially at 24 hr (Fig. 5). These results showed that atorvastatin modified membrane localization of Rho proteins, namely RhoA, RhoB, and RhoC, and upregulated their cytosolic expression in HCT116 and HT29 cells possibly through a feed back mechanism. It was also observed that combination of

2121

atorvastatin with celecoxib potentiated atorvastatin action on Rho protein membrane localization and expression. As shown in Figure 5, combined treatment with atorvastatin/ celecoxib decreased membrane bound K-Ras moderately at 24 hr in both cell lines, but not at 48 hr in HCT116 cells. Moreover, treatment with atorvastatin alone or in combination with celecoxib for 48 hr significantly increased (2-fold) membrane bound K-Ras in HT29 cells. The results indicated that K-Ras might not be the primary target of combined treatment with atorvastatin/celecoxib in cell lines tested herein, because membrane bound K-Ras was not downregulated after 48 hr of drug exposure when most significant growth inhibition and apoptosis have been already observed.

FIGURE 5 – Effects of atorvastatin and celecoxib on membrane localization of small G-proteins in HCT116 cells (a) and HT29 Cells (b). Cells were treated for 24 or 48 hr, and cellular protein were prepared as total protein (T) or fractionated into membrane (M) and cytosol fractions as described in Methods. Western blot showed that atorvastatin and celecoxib combinations modify membrane localization of Rho proteins. The results were representative of 3 experiments. ICAM-1, a membrane bound protein, served as an equal loading control for membrane bound (M) proteins, and b-actin served as an equal loading control for total proteins. (c) Equal amount of membrane and cytosol protein were subject to Western blot for ICAM-1, the results showed exclusive enrichment of ICAM-1 in membrane fraction.

2122

XIAO ET AL.

FIGURE 6 – Effects of atorvastatin and celecoxib on AKT and JNK pathways. (a) The phosphorylated and total protein levels of Akt, PDK1, PI3K, PTEN, JNK, and c-Jun were determined by Western blot analyses. Cells were treated for 24 or 48 hr, and b-actin served as an equal loading control. The results were representative of 3 experiments. (b) and (c) JNK inhibitor SP600125 abolished G0/G1 cell cycle arrest caused by combined treatment with atorvastatin and celecoxib. (b) HCT116 cells and (c) HT29 cells were treated with atorvastatin, celecoxib, and their combination either in the presence or absence of SP600125 for 24 hr before cell cycle analyses by flow cytometry. The percentage of G0/G1 phase cell cycle arrest was expressed in mean 6 S.D. (error bar). Different notation indicates statistical difference with p < 0.01 (n 5 3) analyzed by one-way ANOVA. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Combined treatment causes AKT inactivation We analyzed the status of AKT phosphorylation in colon cancer cells after treatment with atorvastatin, celecoxib, or their combination for 24 or 48 hr when significant G0/G1 phase cell cycle arrest and apoptosis could be seen (as described earlier). The combination treatments significantly decreased phosphorylation of AKT in both cell lines, while treatments with atorvastatin or celecoxib caused marginal or no effect (Fig. 6a). Consistent with these results, it was found that the combination treatments for 48 hr decreased phosphorylation levels of PDK1 and PI3K, and also decreased protein levels of PDK1 and PI3K p85. Moreover, phosphorylation of PTEN at Ser380, which is believed to inhibit PTEN activity, was inhibited by the combination treatments without significant change in total PTEN protein levels in both HCT116 and HT29 cells at 48 hr. All of these effects were not seen or of much less extent in the cells treated with atorvastatin or celecoxib alone. These results indicated that atorvastatin and celecoxib acted synergistically in inhibiting AKT phosphorylation in both HCT116 and HT29 cells.

Combined treatment induces c-Jun NH2-terminal kinase (JNK) activation and its blockade reverses G0/G1 phase cell cycle arrest by the combined treatment We examined the effect of atorvastatin/celecoxib combination on the JNK pathway and its possible relationship to cell cycle arrest and apoptosis. Atorvastatin activated JNK at 24 and 48 hr, and the activation can be seen with concentrations of 4 lM in HCT116 cells. In HT29 cells, the ability of atorvastatin to activate JNK was much weaker than in HCT116 cells (Fig. 6a). Celecoxib did not activate JNK with concentration up to 75 lM (data not shown). However, the combination with celecoxib enhanced JNK activation caused by atorvastatin in both cell lines (Fig. 6a). Consistent with the results from JNK activation, atorvastatin increased phosphorylation of c-Jun, an important downstream effector of JNK, at both 24 and 48 hr, and this upregulation was also potentiated by the presence of celecoxib (Fig. 6a). Treatment with atorvastatin and its combination with celecoxib also upregulated c-Jun protein expression in both cells. To study the role of JNK activation in cell cycle arrest and apoptosis caused by treatment with the combination of atorvastatin


and celecoxib, we used JNK inhibitor SP600125 to determine if it can alter the effects of the treatment. Co-incubation of SP600125 with atorvastatin/celecoxib combination for 24 hr completely abolished cell cycle arrest in G0/G1 phase caused by combined treatment in HCT116 cells (Fig. 6b). Similarly in HT29 cells, SP600125 was able to release cell cycle arrest by the combined treatment (Fig. 6c). These results suggested that JNK activation played a key role in G0/G1 phase cell cycle arrest caused by the combined treatment. We also examined the effect of SP600125 on combined treatment-induced apoptosis, and the results from Annexin V/PI co-staining assay showed that SP600125 did not rescue HCT116 or HT29 cells from apoptosis after 48 hr co-incubation with atorvastatin/celecoxib (data not shown). Discussion Our analyses using isobologram for two-drug interaction showed that combination of atorvastatin and celecoxib synergistically inhibited growth of human colon cancer cells. We further demonstrated that atorvastatin/celecoxib combination greatly arrested human colon cancer cells in G0/G1 phase, and this effect was much stronger than those caused by atorvastatin or celecoxib individually. Moreover, prolonged treatment with the combination induced significant apoptosis that could not be seen with atorvastatin or celecoxib treatment individually. The strong synergy between atorvastatin and celecoxib on cell cycle arrest and apoptosis is of great importance for potential increase of chemoprevention and treatment efficacy by this combination in humans for colorectal cancer. The increased efficacy could also reduce celecoxib dose to minimize possible detrimental side effects, while receiving beneficial effect of atorvastatin in cardiovascular health.10 Statins inhibit HMG-CoA reductase, the rate-limiting enzyme of mevalonate synthesis pathway. This in turn inhibits formation of downstream lipid isoprenoids, such as farnesyl pyrophosphate (FPP) and geranylgeranyl pyrophosphate (GGPP) that are essential for prenylation, membrane localization, and subsequent activation of Rho proteins. Our results indicated that atorvastatin significantly decreased membrane-bound (potentially active) RhoA and RhoC, and this decrease was further potentiated by the combination with celecoxib, especially for RhoA. RhoA and RhoC have been found overexpressed in various human cancer including colon cancer, and they are associated with cell cycle progression and increased invasiveness and metastasis of tumor cells.20 Active RhoA can negatively regulate p21Cip1/Waf1 possibly through affecting phosphorylation of SP1 transcription factors.21 RhoA also downregulates p27Kip1 by decreasing transcriptional efficiency of p27Kip1 mRNA and increasing p27Kip1 degradation.13 Our results demonstrated that synergistic action of atorvastatin/ celecoxib combination in the upregulation of p21Cip1/Waf1 and p27Kip1 and decrease of hyper-phosphorylated Rb occurred concomitantly with the decrease of membrane bound RhoA. This suggested that decreased membrane bound RhoA likely contributed to the cell cycle arrest observed after treatment with atorvastatin and its combination with celecoxib in colon cancer cells. Treatment with statins depletes both FPP and GGPP, and it is expected to see decreased membrane localization of all Rho proteins by atorvastatin. It is unclear why the membrane bound RhoB was increased by atorvastatin and its combination with celecoxib in contrast to the response of RhoA and RhoC. What may account for this phenomenon is the fact that RhoB distinguishes itself from RhoA and RhoC by its unique pattern of prenylation, subcellular localization, and functions. RhoB can be either farnesylated or geranylgeranylated, while RhoA and RhoC can be only geranylgeranylated.22 RhoB is localized in early endosomes and prelysomal compartment with a much more rapid turn over rate, whereas other Rho proteins are localized in plasma membrane and relatively stable.23 Although RhoA, RhoB, and RhoC share identical effector domains, the unique properties of RhoB result in its distinct physiological functions. In contrast to the oncogenic RhoA

2123

and RhoC, preclinical studies have suggested suppressive role of RhoB in cancer, and RhoB expression was found markedly decreased in human lung, brain, and head and neck cancer tissues through cancer progression.24 RhoB was suggested also to play an important role in apoptosis induced by DNA damaging agents.22 The increase of membrane bound RhoB observed in our study may further contribute to the growth inhibitory and cell killing effect of atorvastatin/celecoxib combination in colon cancer cells. Additional studies are needed to illustrate the molecular basis for the atorvastatin/celecoxib-induced changes on Rho proteins and their roles in cancer cell growth inhibition and apoptosis. In our study, celecoxib did not inhibit phosphorylation of AKT at 25 lM, at which concentration significant growth inhibition and cell cycle arrest were seen. This result is in contrast to the proposal that AKT inactivation is a mechanism for celecoxib-induced growth inhibition and apoptosis in human colon cancer cells including HCT116 and HT29.25,26 The proposal was based on the observation that 100 lM celecoxib inhibited AKT phosphorylation.25 We observed that celecoxib decreased phospho-AKT levels at 75 lM in HT29 cells (date not shown), but the concentration was too high to be relevant to the effects caused by celecoxib in our combination studies. We showed that co-treatment with celecoxib and atorvastatin almost abolished the phosphorylation of AKT in both HCT116 and HT29 cells. Our data also demonstrated that inhibition of AKT activation by the combination treatment was associated with modulation on other key components of the AKT pathway, including PDK1, PI3K, and PTEN. AKT signaling is one of the important cell survival pathways, and it protects cells from apoptosis and promotes cell survival by several different mechanisms.27–30 Through blocking activation of AKT, the atorvastatin/celecoxib combination may inhibit cell proliferation and enhance apoptosis, which may be crucial for the synergy observed herein. Statins have been shown to activate JNK in cells during induction of apoptosis.31,32 We also observed that atorvastatin activated JNK pathway, and celecoxib synergistically enhanced this activity even though celecoxib did not affect JNK by itself. Through the use of JNK inhibitor SP600125, we demonstrated that the synergistic action of atorvastatin/celecoxib combination on JNK pathway is associated with cell cycle arrest. It is surprising that this inhibitor did not affect apoptosis induced by this combination. Other mechanisms, such as AKT inactivation and RhoB activation, may play more important roles than JNK activation in inducing apoptosis after atorvastatin/celecoxib combination treatment. It needs further investigation to understand the role of JNK activation in the synergy caused by atorvastatin/celecoxib combination. We also examined the effect of this combination on the NF-jB pathway that plays important roles in carcinogenesis. However, our results did not show that the combination treatments inhibited the NF-jB pathway at 24 or 48 hr in either HCT116 or HT29 cells (data not shown). Although the drug concentrations used herein for mechanistic study were higher than the concentrations found in human plasma after taking clinical dose of atorvastatin or celecoxib, we believe that the strong synergy observed between these 2 drugs is of important physiological relevance because: (i) the presently observed cell cycle arrest and apoptosis, caused by the combined treatment, are in agreement with the results from recent in vivo study showing that combined oral administration of atorvastatin and celecoxib significantly decreased proliferation and increased apoptosis in AOMinduced colonic adenocarcinoma in F344 rats16; (ii) our isobologram analyses results demonstrated that all dose pairs tested showed strong synergy on growth inhibition of colon cancer cells, even in the case of atorvastatin (0.125 lM)/celecoxib (0.78 lM) combination (interaction index: 0.21) in HCT116 cells, and these concentrations are achievable in human plasma in clinical settings (The reported plasma Cmax of atorvastatin included its active metabolites that may contribute to the anti-cancer effects due to their cholesterol-lowering activity)33,34; (iii) colon tissue may be exposed to higher drug concentrations due to direct contact with ingested drugs

2124

XIAO ET AL.

in intestinal content, and/or excreted drugs from bile; (iv) the effective drug concentrations used in cell culture studies are generally higher than the plasma concentrations in animals or humans after drug treatments, and this is possibly due to much shorter drug exposure time in cell culture system (hours or a few days vs. weeks or months of drug therapy in clinics).35 In summary, we demonstrated that atorvastatin and celecoxib produced a strong synergy in growth inhibition of human colon

cancer cells, and this was associated with their synergistic actions on cell cycle arrest and apoptosis. We proposed that selective modification of membrane bound Rho proteins, AKT inactivation, and JNK activation likely contributed to the synergistic interaction. Our results identified potential targets of atorvastatin/celecoxib combination in colorectal cancer chemoprevention and therapy, and provided important leads for mechanistic and biomarker studies in clinical trials.

References 1. 2. 3. 4. 5.

6.

7.

8.

9. 10. 11. 12. 13. 14.

15.

16.

17.

18.

Jemal A, Murray T, Ward E, Samuels A, Tiwari RC, Ghafoor A, Feuer EJ, Thun MJ. Cancer statistics, 2005. CA Cancer J Clin 2005;55:10–30. Ji BT, Devesa SS, Chow WH, Jin F, Gao YT. Colorectal cancer incidence trends by subsite in urban Shanghai, 1972–1994. Cancer Epidemiol Biomarkers Prev 1998;7:661–6. Yiu HY, Whittemore AS, Shibata A. Increasing colorectal cancer incidence rates in Japan. Int J Cancer 2004;109:777–81. Goldenberg MM. Celecoxib, a selective cyclooxygenase-2 inhibitor for the treatment of rheumatoid arthritis and osteoarthritis. Clin Ther 1999;21:1497–513; discussion 1427–98. Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, Makuch R, Eisen G, Agrawal NM, Stenson WF, Burr AM, Zhao WW, et al. Gastrointestinal toxicity with celecoxib vs. nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Longterm Arthritis Safety Study. JAMA 2000;284: 1247–55. Bertagnolli MM, Eagle CJ, Zauber AG, Redston M, Solomon SD, Kim K, Tang J, Rosenstein RB, Wittes J, Corle D, Hess TM, Woloj GM, et al. Celecoxib for the prevention of sporadic colorectal adenomas. N Engl J Med 2006;355:873–84. Arber N, Eagle CJ, Spicak J, Racz I, Dite P, Hajer J, Zavoral M, Lechuga MJ, Gerletti P, Tang J, Rosenstein RB, Macdonald K, et al. Celecoxib for the prevention of colorectal adenomatous polyps. N Engl J Med 2006;355:885–95. Solomon SD, McMurray JJ, Pfeffer MA, Wittes J, Fowler R, Finn P, Anderson WF, Zauber A, Hawk E, Bertagnolli M. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med 2005;352:1071–80. Psaty BM, Potter JD. Risks and benefits of celecoxib to prevent recurrent adenomas. N Engl J Med 2006;355:950–2. Demierre MF, Higgins PD, Gruber SB, Hawk E, Lippman SM. Statins and cancer prevention. Nat Rev Cancer 2005;5:930–42. Friis S, Olsen JH. Statin use and cancer risk: an epidemiologic review. Cancer Invest 2006;24:413–24. Jacobs EJ, Rodriguez C, Brady KA, Connell CJ, Thun MJ, Calle EE. Cholesterol-lowering drugs and colorectal cancer incidence in a large United States cohort. J Natl Cancer Inst 2006;98:69–72. Graaf MR, Richel DJ, van Noorden CJ, Guchelaar HJ. Effects of statins and farnesyltransferase inhibitors on the development and progression of cancer. Cancer Treat Rev 2004;30: 609–41. Jalving M, Koornstra JJ, De Jong S, De Vries EG, Kleibeuker JH. Review article: the potential of combinational regimen with non-steroidal anti-inflammatory drugs in the chemoprevention of colorectal cancer. Aliment Pharmacol Ther 2005;21:321–39. Agarwal B, Rao CV, Bhendwal S, Ramey WR, Shirin H, Reddy BS, Holt PR. Lovastatin augments sulindac-induced apoptosis in colon cancer cells and potentiates chemopreventive effects of sulindac. Gastroenterology 1999;117:838–47. Reddy BS, Wang CX, Kong AN, Khor TO, Zheng X, Steele VE, Kopelovich L, Rao CV. Prevention of azoxymethane-induced colon cancer by combination of low doses of atorvastatin, aspirin, and celecoxib in F 344 rats. Cancer Res 2006;66:4542–6. Swamy MV, Patlolla JM, Steele VE, Kopelovich L, Reddy BS, Rao CV. Chemoprevention of familial adenomatous polyposis by low doses of atorvastatin and celecoxib given individually and in combination to APCMin mice. Cancer Res 2006;66:7370–7. Chou TC, Talalay P. Quantitative analysis of dose–effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzym Regul 1984;22:27–55.

19. Park HJ, Kong D, Iruela-Arispe L, Begley U, Tang D, Galper JB. 3Hydroxy-3-methylglutaryl coenzyme A reductase inhibitors interfere with angiogenesis by inhibiting the geranylgeranylation of RhoA. Circ Res 2002;91:143–50. 20. Sahai E, Marshall CJ. RHO-GTPases and cancer. Nat Rev Cancer 2002;2:133–42. 21. Adnane J, Bizouarn FA, Qian Y, Hamilton AD, Sebti SM. p21(WAF1/CIP1) is upregulated by the geranylgeranyltransferase I inhibitor GGTI-298 through a transforming growth factor b- and Sp1responsive element: involvement of the small GTPase rhoA. Mol Cell Biol 1998;18:6962–70. 22. Prendergast GC. Actin’ up: rhoB in cancer and apoptosis. Nat Rev Cancer 2001;1:162–8. 23. Chen Z, Sun J, Pradines A, Favre G, Adnane J, Sebti SM. Both farnesylated and geranylgeranylated RhoB inhibit malignant transformation and suppress human tumor growth in nude mice. J Biol Chem 2000;275:17974–8. 24. Huang M, Prendergast GC. RhoB in cancer suppression. Histol Histopathol 2006;21:213–18. 25. Arico S, Pattingre S, Bauvy C, Gane P, Barbat A, Codogno P, OgierDenis E. Celecoxib induces apoptosis by inhibiting 3-phosphoinositide-dependent protein kinase-1 activity in the human colon cancer HT-29 cell line. J Biol Chem 2002;277:27613–21. 26. Yamazaki R, Kusunoki N, Matsuzaki T, Hashimoto S, Kawai S. Selective cyclooxygenase-2 inhibitors show a differential ability to inhibit proliferation and induce apoptosis of colon adenocarcinoma cells. FEBS Lett 2002;531:278–84. 27. Cardone MH, Roy N, Stennicke HR, Salvesen GS, Franke TF, Stanbridge E, Frisch S, Reed JC. Regulation of cell death protease caspase-9 by phosphorylation. Science 1998;282:1318–21. 28. Ozes ON, Mayo LD, Gustin JA, Pfeffer SR, Pfeffer LM, Donner DB. NF-jB activation by tumour necrosis factor requires the Akt serinethreonine kinase. Nature 1999;401:82–5. 29. Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y, Greenberg ME. Akt phosphorylation of BAD couples survival signals to the cellintrinsic death machinery. Cell 1997;91:231–41. 30. Brunet A, Bonni A, Zigmond MJ, Lin MZ, Juo P, Hu LS, Anderson MJ, Arden KC, Blenis J, Greenberg ME. Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell 1999;96:857–68. 31. Kaneta S, Satoh K, Kano S, Kanda M, Ichihara K. All hydrophobic HMG-CoA reductase inhibitors induce apoptotic death in rat pulmonary vein endothelial cells. Atherosclerosis 2003;170: 237–43. 32. Koyuturk M, Ersoz M, Altiok N. Simvastatin induces proliferation inhibition and apoptosis in C6 glioma cells via c-jun N-terminal kinase. Neurosci Lett 2004;370:212–17. 33. Stern RH, Yang BB, Hounslow NJ, MacMahon M, Abel RB, Olson SC. Pharmacodynamics and pharmacokinetic-pharmacodynamic relationships of atorvastatin, an HMG-CoA reductase inhibitor. J Clin Pharmacol 2000;40:616–23. 34. McAdam BF, Catella-Lawson F, Mardini IA, Kapoor S, Lawson JA, FitzGerald GA. Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA 1999;96:272–7. 35. Maier TJ, Janssen A, Schmidt R, Geisslinger G, Grosch S. Targeting the b-catenin/APC pathway: a novel mechanism to explain the cyclooxygenase-2-independent anticarcinogenic effects of celecoxib in human colon carcinoma cells. FASEB J 2005;19: 1353–5.