Novel Dithiolethione-Modified Nonsteroidal Anti-Inflammatory Drugs in Human Hepatoma Hepg2 and Colon LS180 Cells Sara E

Cancer Therapy: Preclinical

Novel Dithiolethione-Modified Nonsteroidal Anti-Inflammatory Drugs in Human Hepatoma HepG2 and Colon LS180 Cells Sara E. Bass,1Pawel Sienkiewicz,2 ChristopherJ. MacDonald,2 RobertY.S. Cheng,2 Anna Sparatore,3 Piero Del Soldato,4 David D. Roberts,5 Terry W. Moody,6 David A. Wink,7 and Grace ChaoYeh2

Abstract Purpose: Nonsteroidal anti-inflammatory drugs (NSAID) are promising chemopreventive agents against colon and other cancers. However, the molecular basis mediated by NSAIDs for chemoprevention has not been fully elucidated. Environmental carcinogens induce DNA mutation and cellular transformation; therefore, we examined the effect of NSAIDs on carcinogenesis mediated by the aryl hydrocarbon receptor signaling pathway. In this study, we investigated the activities of a new class of NSAIDs containingdithiolethione moieties (S-NSAID) on both arms of carcinogenesis. Experimental Design: We investigated the effects of the S-NSAIDs, S-diclofenac and S-sulindac, on carcinogen activation and detoxification mechanisms in human hepatoma HepG2 and human colonic adenocarcinoma LS180 cells. Results: We found that S-diclofenac and S-sulindac inhibited the activity and expression of the carcinogen activating enzymes, cytochromes P-450 (CYP) CYP1A1, CYP1B1, and CYP1A2. Inhi- bition was mediated by transcriptional regulation of the aryl hydrocarbon receptor (AhR) pathway. The S-NSAIDs down-regulated carcinogen-induced expression of CYP1A1 heterogeneous nuclear RNA, a measure of transcription rate. Both compounds blocked carcinogen-activated AhR from bindingto the xenobiotic responsive element as shown by chromatin immunoprecipitation. S-diclofenac and S-sulindac inhibited carcinogen-induced CYP enzyme activity through direct inhibition as well as through decreased transcriptional activation of the AhR. S-sulindac induced expression of several carcinogen detoxification enzymes of the glutathione cycle includ- ingglutathione S-transferase A2, glutamate cysteine ligase catalytic subunit, glutamate cysteine ligase modifier subunit, and glutathione reductase. Conclusions: These results indicate that S-diclofenac and S-sulindac may serve as effective chemoprevention agents by favorably balancing the equation of carcinogen activation and detoxification mechanisms.

The use of nonsteroidal anti-inflammatory drugs (NSAID) has tive NSAIDs, such as aspirin, diclofenac, ibuprofen, and been associated with a reduced incidence of colorectal, breast, naproxen, is frequently associated with gastrointestinal toxicity esophageal, stomach, prostate, bladder, ovary, and lung cancers due to the chronic suppression of cyclooxygenase-1and (1). However, chronic treatment with conventional nonselec- subsequent inhibition of mucin secretion in the gut (2). The new generation of NSAIDs that target cyclooxygenase-2, such as rofecoxib and celecoxib, although effective as chemopreventive Authors’Affiliations: 1Basic Research Program, ScienceApplications International 2 agents, have been recently shown to cause an increased Corporation-Frederick, Inc., and Laboratory of Metabolism, National Cancer incidence of stroke and myocardial infarctions (3). Institute-Frederick, Frederick, Maryland; 3Istituto di Chimica Farmaceutica, University of Milan and 4Sulfidris, Milan, Italy; 5Laboratory of Pathology, 6Office of A new series of novel redox-modified NSAIDs has been the Director, and 7Radiation Biology Branch, Center for Cancer Research, National synthesized that may reduce these side effects. These compounds Cancer Institute, NIH, Bethesda, Maryland are conventional NSAIDs linked to a gaseous transmitter. The Received 7/17/08; revised 11/13/08; accepted 12/2/08; published OnlineFirst most extensively studied of these compounds are the nitric 3/10/09. Grant support: This Research was supported (in part) by the Intramural Research oxide–releasing NSAIDs, which were first described in 1994 (4) Program of the NIH, National Cancer Institute, Center for Cancer Research and and have because been evaluated in several clinical trials in National Cancer Institute, NIH, under contract N01-CO-12400. which their gastric sparing properties have been well-established The costs of publication of this article were defrayed in part by the payment of page (5, 6). Additional studies have shown nitric oxide–NSAIDs to be advertisement charges. This article must therefore be hereby marked in accordance promising chemopreventive agents (7). Although nitric oxide with 18 U.S.C. Section 1734 solely to indicate this fact. Note:The contentof thispublicationdoes notnecessarily reflect the views orpolicies is the best characterized of the gaseous transmitters, carbon of the Department of Health and Human Services, nor does mention of trade names, monoxide and hydrogen sulfide (H2S) have also been recog- commercialproducts, ororganizations imply endorsement by the U.S. Government. nized for their ability to affect key physiologic functions (8). Requests for reprints: Grace C. Yeh, Building31, Room 3A11, 31 Center Drive, In particular, H2S has been shown to exhibit complex but poorly Center for Cancer Research, National Cancer Institute, Bethesda, MD 20892- 2440. Phone: 301-451-8296; Fax: 301-496-0775; E-mail: [email protected]. understood roles in inflammation with evidence for proinflam- F 2009 American Association for Cancer Research. matory as well as anti-inflammatory activities. H2S has also been doi:10.1158/1078-0432.CCR-08-1870 reported to reduce NSAID-induced gastric mucosal injury

Clin Cancer Res 2009;15(6) March 15, 2009 1964 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Modulation of Phase I and II Enzymes by S-NSAIDs

There is additional evidence that S-NSAIDs may also show a Translational Relevance potential usefulness in cancer treatment. Parent compounds, diclofenac and sulindac, have been investigated for their Occupational and environmental exposure to polycyclic potential chemopreventive properties and have been found aromatic hydrocarbons (PAH) has been suggested to pro- particularly useful for prevention and treatment in in vitro and voke inflammatory disorders and cancer. The molecular in vivo models of colon cancer in addition to other cancer types pathway/mechanisms invoked by PAH-mediated inflam- (16, 17). Recent work on ACS 15 and 18 has also shown mation remain to be clarified. Our article contributed to the antiangiogenic activities of the compounds in models of tumor- identification of both the molecular pathway and regulation driven angiogenesis (14). Our laboratory has shown the ability of PAH metabolism in both human hepatoma cells, which is of sulindac as well as the dithiolethione 5-(p-methoxyphenyl)- the active site for carcinogenesis, and colon adenocarcino- 1,2-dithiole-3-thione (ADT)8, to regulate the aryl hydrocarbon ma cells, which is the first exposure site for environmental receptor (AhR) pathway and phase I enzymes (18). The AhR is xenobiotics. the key transcriptional controller of carcinogen metabolizing The dithiolethione-NSAIDs in our study are newly syn- pathways involving the genes that encode phase I and phase II thesized drugs, which conjugate the dithiolethione moiety enzymes. Inhibiting the activation of procarcinogens by phase I with NSAIDs. We found the novel compounds not only enzymes through suppression of the AhR signaling pathway inhibited the PAH activation of phase 1enzymes induced has been suggested as an important mechanism for cancer by PAHs but also increased the detoxification of phase 2 prevention (19, 20). Based on the above data, we hypothesized enzymes.We are the first to identify the dual function prop- that the novel compounds, ACS 15 and ACS 18, could have erty of S-NSAIDs, which modulated carcinogen toxicity potential usefulness as chemoprotective compounds through mediated by the aryl hydrocarbon receptor as a molecular regulation of the AhR pathway and associated enzymes. We mechanism/pathway toward chemoprotection against en- show that both ACS 15 and 18 blocked carcinogen-induced vironmental xenobiotics. phase I enzyme activity in vitro, and that this activity was regulated through inhibition of AhR-mediated transcription. In addition, ACS 18 increased the expression of several phase II without affecting the functional suppression of prostaglandins genes. This dual action of reduced carcinogen activation (9). Additionally, an H2S-releasing derivative of mesalamine simultaneous with increased carcinogen detoxification supports reduced visceral sensitivity and pain perception in a rat model of the view that the S-NSAID model may be a valuable tool for irritable bowel syndrome (10). Finally, H2S may also play a role chemoprotection against environmental carcinogens. in cardio protection (11). These observations, as well as the fact that many NSAIDs reduce endogenous H2S formation (9), which may contribute to their tendency to cause gastric damage, led to the hypothesis that dithiolethione-modified NSAIDs Materials and Methods (S-NSAID) may serve as valuable chemoprotective drugs. The newly synthesized, novel compounds, S-diclofenac Materials. HepG2 human hepatocellular carcinoma and LS180 and S-sulindac, are composed of an H2S-releasing dithiole- human colorectal adenocarcinoma cell lines were from American Type thione moiety, ADT-OH covalently conjugated with the parent Culture Collection. RPMI 1640, Eagle’s MEM with Earle’s balanced salt NSAID by an ester linkage (Fig. 1). The chemical synthesis of solution, glutamine, fetal bovine serum, PBS, Tris-HCl buffer, trypsin/ S-diclofenac, 2-[(2,6-dichlorophenyl)amino]benzeneacetic acid EDTA, and TRIzol were from Invitrogen. Protease inhibitor tablets were from Roche. 2,3,5,7-Tetrachlorodibenzo-p-dioxin (TCDD) was from 4-(3H-1,2,dithiol-3-thione-5-yl)phenyl ester and S-sulindac, the Midwest Research Institute. The Omniscript kit was from Qiagen. (Z)-5-Fluoro-2-methyl-1-[4-(methylsulfinyl)phenyl]-methy- Recombinant CYP enzymes were from BD Biosciences. ACS 15 and ACS H H lene]-1 -indene-3-acetic acid 4-(5-thioxo-5 -[1,2]dithiol- 18 were from Sulfidris. All other chemicals except where noted were 3-yl)-phenyl ester, which are designated as ACS 15 and ACS from Sigma. 18, respectively, has been previously reported (12, 13). Recent Cell culture. HepG2 cells were grown in RPMI 1640, and LS180 work on ACS 15 has shown promising results. Unlike the cells were grown in Eagle’s MEM with Earle’s balanced salt solution. parent NSAID, ACS 15 did not effect blood pressure or heart Both media were supplemented with 10% fetal bovine serum and rate in rat models (12) and in rabbit models, has shown anti- 2 mmol/L L-glutamine. Cells were passed weekly with 0.05% trypsin/ ischemic activities for cardio protection (14). In addition, ACS 0.53 mmol/L EDTA. All experiments were carried out on confluent cells j 15 was much better tolerated by the gastrointestinal tract of at 37 C in humidified atmosphere with 5% CO2 except where mentioned. ACS 15 and 18 were dissolved in DMSO; final DMSO rats, and enhanced the anti-inflammatory effect of diclofenac concentration was V0.1% in all experiments. by interfering with transcriptional mechanisms necessary for Reverse transcription-PCR. Confluent cells in six-well plates were the inflammatory response (12). The same compound as ACS treated with DMSO (control) or the test drugs dissolved in DMSO. Total 15, although coded differently, ATB-337 has also shown to RNA was isolated using TRIzol as directed. cDNA was synthesized from induce >90% less gastrointestinal damage than diclofenac in 4 Ag of total RNA using random primers and the Omniscript kit rats. ATB-337 also did not induce leukocyte adherence, and was according to manufacturer’s instructions. Real-time PCR was done more potent at reducing paw edema (15). These observations in a master mix containing 12.5 AL Applied Biosystem’s TaqMan evidence the view that the S-NSAID model, in providing the Universal PCR Master Mix, 8.75 AL DEPC water, 0.1 Ag cDNA, and A benefit of small quantities of a dithiolethione moiety released 1.25 L FAM-labeled TaqMan Gene Expression Assay primer/probe over time, as well as the benefit of the parent NSAID, will sets for CYP1A1 (Hs00153120), CYP1A2 (Hs00167927), CYP1B1 provide an overall effect of enhanced anti-inflammatory activity with reduced gastrotoxicity and adverse cardiovascular effects. 8 Yeh, unpublished data.

www.aacrjournals.org 1965 Clin Cancer Res 2009;15(6) March 15, 2009 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Cancer Therapy: Preclinical

the manufacturer’s protocol. Real-time PCR for the xenobiotic responsive element (XRE) of CYP1A1 was conducted with primers and conditions described by Hestermann and Brown (24) on a Bio-Rad iCycler Real Time Detection System. Results were calculated as described above and normalized to input sample DNA. Assay of CYP enzyme activity. The ability of ACS 15 and 18 to affect CYP enzyme activity was evaluated in intact cells by measurement of ethoxyresorufin-O-deethylase (EROD) activity as previously described (18). Recombinant CYP1A1, CYP1A2, andCYP1B1 activity. Recombinant CYP1A1 Supersomes (0.75 Amol/L), CYP1A2 Supersomes (2.5 Amol/L), or CYP1B1 Supersomes (3.5 Amol/L) were incubated with 0.4 Amol/L ethoxyresorufin and indicated concentrations of ACS 15 and 18 in a final volume of 100 AL of PBS (pH 7.2). The reaction was initiated by the addition of 500 Amol/L NADPH. The reaction mixture was transferred to a 96-well plate, and EROD activity was determined as described above. Enzyme kinetics were analyzed using the Michaelis-

Menten model. Maximal velocities (Vmax) and dissociation constant paramters (Km) were calculated (Table 1) using the Michaelis-Menten equation: V = Vmax (S)/[Km + (S)]. Statistical analysis. Statistical analyses were done with StatView and JMP Statistical Analysis Software (SAS Institute). Differences between group mean values were determined by a one-factor ANOVA, followed by Fisher protected least significant difference post hoc analysis for pairwise comparison of means. Statistical analyses for the recombinant enzyme activity were done using the nlme function in S-PLUS version 7 for Windows (InsightfulCorporation, 2005).

Results

For the concentration ranges used in these studies, 0 to 100 Amol/L ACS 15 and 18, we found the compounds to be nontoxic in both cell lines as measured by the cell viability assay measuring reduction of the tetrazolium salt 3,(4,5- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (data Fig. 1. Chemical structures of dithiolethione, ADT-OH, ACS 15, and ACS 18. not shown). ACS 15 and18 inhibit carcinogen-inducedCYPmRNA (Hs00164383), GSTA2 (Hs00747232mH), glutamate cysteine ligase expression. TCDD (250 pmol) caused a 388-, 33.5-, and 52- catalytic (Hs00155249), glutamate cysteine ligase modifier fold increase in CYP1A1, CYP1A2, and CYP1B1 mRNA (Hs00157694), GR (Hs00167317), and glucose-6-phosphate dehydro- expression in HepG2 cells and a 99.3-, 92.6-, and 9-fold genase (Hs00166169) on a Bio-Rad iCycler Real Time Detection increase in LS180 cells compared with DMSO controls. System. Amplification conditions were as follows: 1cycle of 50 jC for 2 Dimethylbenzanthracene (DMBA; 1 Amol/L) induced CYP1A1, j j j min, 95 C for 10 min, then 50 cycles of 95 C for 15 s, 60 C for 1min, CYP1A2, and CYP1B1 mRNA expression in HepG2 cells by 5-, j C and ended with 1cycle of 60 C for 1min. Threshold cycles ( T) were 2- and 8.4-fold, and by 6.6-, 1.3-, and 2.5-fold in LS180 cells. determined using Bio-Rad’s iCycler 3.0a version software, allowing These inductions were inhibited in a concentration-dependent baseline to be set automatically by the software. Relative changes in -DDC gene expression were calculated using the 2 T equation. Results were normalized to 18S RNA levels. Transcription of CYP1A1. The effect of ACS 15 and 18 on the Table 1. Maximal velocities (V max)and transcription of the CYP1A1 gene was evaluated by measuring the level dissociation constant parameters (Km)forthe of heterogeneous nuclear RNA by real-time PCR, as described by Michaelis-Menten curves analyzed for ACS 15 and Elferink and Reiners (21) and modified by Guigal et al. (22). This assay ACS 18 has been well-characterized as a valid substitute for nuclear run-on experiments as a measure of transcription rates. Sequences for Experiment V max K m hnCYP1A1 forward and reverse primers were CTTGGACCTCTTTG- 0 Amol/L ACS15 + recCYP1A1 4.4 324.4 GAGCTG and TGACTGTGTCAAACCCTGGA, respectively. Amplifica- j 0.05 Amol/L ACS15 + recCYP1A1 3.5 463.2 tion conditions were 15 min at 95 C, followed by 45 cycles of 15 s at 0.5 Amol/L ACS15 + recCYP1A1 1.9 1,231.1 j j j hnCYP1A1 94 C, 30 s at 60 C, and 30 s at 72 C. The level of was 10 Amol/L ACS15 + recCYP1A1 3.9 1,085.9 normalized to the level of 18S RNA expression. 50 Amol/L ACS15 + recCYP1A1 6.3 3,716.3 Chromatin immunoprecipitation assay. Cells were incubated with 0 Amol/L ACS18 + recCYP1A1 6.6 416.2 DMSO (0.06%), 10 nmol/L TCDD alone, 50 Amol/L ACS 15 or 18 0.05 Amol/L ACS18 + recCYP1A1 6.4 419.8 alone, or a combination of TCDD and the ACS compounds as indicated 1 Amol/L ACS18 + recCYP1A1 7.1 776.9 for 90 min at 37jC. The chromatin immunoprecipitation (ChIP) assay 10 Amol/L ACS18 + recCYP1A1 6.4 1,110.7 was conducted as previously described (23). ChIP DNA was purified for 50 Amol/L ACS18 + recCYP1A1 6.3 3,717.0 PCR by using QIAquick PCR Purification kit from Qiagen according to

Clin Cancer Res 2009;15(6) March 15, 2009 1966 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Modulation of Phase I and II Enzymes by S-NSAIDs

Fig. 2. Effect of ACS 15 on carcinogen- induced CYP1A1, CYP1A2,and CYP1B1 mRNA expression. HepG2 (A, B,andC)or LS180 (D, E ,andF) cells were treated with 250 pmol TCDD or 1 Amol/L DMBA with or without increasingconcentrations of ACS 15 as indicated for 24 h. Levels of mRNA were measured by semiquantitative PCR, and normalized to 18S RNA. Columns, mean (n = 3); bars, SE. *, P < 0.001.

manner by cotreatment with ACS 15 (Fig. 2A-F). Results were binding of the activated AhR to the CYP1A1 XRE enhancer similar for ACS 18 (data not shown). sequence was measured. We used primers and conditions as ACS15and18decreasetherateofCYP1A1 gene transcription. designed and tested by Hestermann and Brown (25). These The transcription rate of the CYP1A1 gene was measured by primers amplify a region of the XRE 784 to 1156 bp upstream quantifying the level of heterogeneous nuclear RNA, or newly of the CYP1A1 transcriptional start site. It is known that both transcribed RNA, by reverse transcription-PCR. Incubation of HepG2 the TCDD-responsive CYP1B1 and CYP1A2 also contain XRE cells with TCDD (250 pmol; Fig. 3A) or DMBA (1 Amol/L; Fig. 3B) binding sites upstream of their transcriptional start sites. The caused a 67.6- and 12-fold increase in hnCYP1A1 expression over CYP1A1 XRE binds to the CYP1B1 gene in gel shift assays and DMSO control levels, respectively. Cotreatment with ACS 15 AhR binding to this region is confirmed by supershift (26). abrogated the stimulatory effects of both TCDD and DMBA. Results Because the CYP1A1 and CYP1A2 genes are positioned in a for ACS 18 were similar (data not shown). head-to-head orientation, they share a common 5¶ upstream ACS 15 and18 inhibit AhR activation. The transcription of region and thus may share regulatory components. Multiple CYP1A1, CYP1A2, and CYP1B1 is primarily regulated by the studies have shown the TCDD-responsive CYP1A1 enhancer AhR. The AhR is a cytosolic protein that, when activated by a also controls CYP1A2 expression (27, 28). The induction of ligand, i.e., polycyclic aromatic hydrocarbons [PAH; TCDD, CYP1A1 expression is therefore widely studied as a model for DMBA, and benzo(a)pyrene (BP)], translocates to the nucleus AhR activity. and binds to its protein partner, the aryl hydrocarbon nuclear Treatment of HepG2 cells with TCDD (10 nmol) caused a translocator (25). Together, this heterodimer forms a transcrip- 6.2-fold induction of activated AhR (Fig. 3C). Cotreatment tion factor that binds to a specific sequence of nucleotides, the with ACS 15 and 18 inhibited the level of activated AhR to XRE, located in the promoter region of a number of genes 2.6- and 0.32-fold, respectively, indicating that ACS 15 and involved in xenobiotic metabolism, including CYPs. Because 18 directly affect AhR activation. The ChIP assay is somewhat ACS 15 and 18 inhibit carcinogen-induced CYP1A1 transcrip- less sensitive than other assays done in this study, and tion, we investigated whether this activity was mediated by AhR therefore, a higher concentration of TCDD (10 nmol versus activation. To test this, the ChIP assay was used, in which 250 pmol) was required to produce statistically significant

www.aacrjournals.org 1967 Clin Cancer Res 2009;15(6) March 15, 2009 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Cancer Therapy: Preclinical results. DMBA and BP are far less potent ligands of the AhR derivatives, ACS 15 and 18, on phase II gene expression. No than TCDD and, therefore, could not be examined in significant increases were observed by ACS 15. We identified this assay. several phase II conjugation enzymes whose mRNA expression ACS 15 and18 inhibit PAH-inducedERODactivities. AhR- was up-regulated by ACS 18. ACS 18 (50 Amol/L) increased the mediated enzymes are critical to the detoxification of PAHs, a mRNA expression of GSTA2, glutamate cysteine ligase catalytic widespread class of environmental pollutants. EROD activity subunit, glutamate cysteine ligase modifier subunit, GR, and is a measure of the cellular capacity for enzymatic activation glucose-6 phosphate dehydrogenase by 3.6-, 1.9-, 1.6-, 1.6-, of PAHs with regard to the carcinogen activating enzymes, and 1.5-fold respectively over 18S RNA control as measured by CYP1A1, CYP1A2, and CYP1B1. Although EROD measures the reverse transcription-PCR (Fig. 6). combined activity of the three CYP isozymes, the 1A1 form has a much higher specific EROD activity than 1A2 or 1B1 (29). We investigated the effects of ACS 15 and 18, as well as their parent compounds, on EROD activity and PAH-induced EROD activity in both HepG2 and LS180 cell lines. Sulindac induced EROD activity by 4.45 F 0.09 pmoles/minute/well (Fig. 4A), as we have previously reported (23), and ACS 18 caused a slight induction of EROD activity (0.46 F 0.01 pmoles/minute/well). TCDD-induced EROD activity was not inhibited by 50 Amol/L of either parent compound, sulindac, diclofenac; nor by NaHS, an H2S releasing compound. These results were duplicated for DMBA and BP-induced EROD activity (data not shown). Only the dithiolethione-modified compounds, ACS 15 and ACS 18, inhibited PAH-induced EROD activity. Further investigation of the ACS compounds on EROD activity show both compounds suppressed TCDD, DMBA, and BP-induced EROD activity in both cell lines (Fig. 4B-E). Inhibition by ACS 15 and 18 was concentration- dependent, and averaged, interpolated IC50s against TCDD, DMBA, and BP-induced EROD for ACS 15 were 38, 16, and 50 Amol/L, respectively, and for ACS 18 were 22, 8.5, and 100 Amol/L. ACS 15 and18 directlyinhibit recombinant CYP enzyme activity. We investigated CYP enzyme inhibition by ACS 15 and 18 using recombinant CYP1A1, CYP1B1, and CYP1A2 supersomes to test whether the inhibition was due to a direct effect on the CYP enzymes. At low concentrations, ACS 15 caused a concentration-dependent decrease in CYP activity with interpolated IC50 values for CYP1A1, CYP1B1, and CYP1A2 of 0.05, 2.4, and 1.75 Amol/L, respectively (Fig. 5A). ACS 18 also caused concentration-dependent decreases in CYP activity (Fig. 5B) with interpolated IC50 values for CYP1A1, CYP1B1, and CYP1A2 of 0.04, 0.23, and 3.1 Amol/L, respectively. Kinetic analysis using the Michaelis-Menten model revealed an increase in Km with increasing ACS 15 and 18 concentrations, whereas Vmax remained unchanged (Table 1). This indicated CYP1A1 inhibition by both ACS 15 and 18 was competitive in nature. Residual plot analysis revealed a good fit by the data to the models. For ease in visual representation, results are displayed by the Lineweaver Burk plots (Fig. 5C and D). ACS 18 induces phase II enzyme mRNA expression. The induction of phase II enzymes has been emphasized as being a critical step for the inactivation of chemical carcinogens (30). Other dithiolethiones such as ADT, 3H-1,2-dithiole-3-thione, Fig. 3. The effect of ACS15 on the transcription of CYP1A1 was measured by and 5-(2-pyrazinyl)-4-methyl-1,2-dithiole-3-thione (oltipraz) quantifyingthe expression of hnCYP1A1RNA by reverse transcription-PCR have shown chemoprotective effects through the induction of in HepG2 cells. Cells were treated with 250 pmol TCDD (A)or1Amol/L DMBA (B) for 24 h, and levels of heterogeneous nuclear RNA were measured by reverse phase II enzymes such as NAD(P)H:quinone oxidoreductase, transcription-PCR and normalized to glyceraldehyde-3-phosphate dehydrogenase glutathione S-transferase, and glutathione reductase (GR; mRNA. Columns, mean (n = 3); bars, SE. C, effect of ACS 15 and 18 on the refs. 31, 32). ADT-OH is released from S-diclofenac or S-sulindac XRE-bindingactivity of the AhR. HepG2 cells were incubated for 90 min with in vivo DMSO or10 nmolTCDD with or without 50 Amol/L ACS 15 orACS18. Cellular after hydrolytic breakdown of the parent molecule (12). proteins were cross-linked and isolated by ChIP, and XRE mRNA was measured by We further investigated the effect of the dithiolethione real-time PCR and normalized to 18SRNA. Columns, mean (n = 6); bars, SE.

Clin Cancer Res 2009;15(6) March 15, 2009 1968 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Modulation of Phase I and II Enzymes by S-NSAIDs

Fig. 4. A, confluent HepG2 cells were treated with 250 pmol TCDD, 50 Amol/L ACS, or parent compound alone or in combination for 24 h before enzyme activity was measured by EROD. B to E, concentration dependent inhibition of carcinogen-induced EROD activity byACS 15 and 18. Confluent HepG2 (B and D)or LS180 (C and E) were treated with 250 pmol TCDD, or 1 Amol/L DMBA or BP either alone or in combination with 50 Amol/L ACS 15 (B and C)orACS18(D and E)for 24 h before measuringEROD activity. Results are shown as EROD activity relative to control. Columns and points, mean (n = 4); bars, SE.

Discussion enzyme activity, e.g., salicylamide (33). We have also shown that sulindac and its metabolites modulate both phase I and Beginning in the early 1800s, with the isolation of salicin phase II gene expression (18, 23). from willow bark, NSAIDs have become important therapeutic In this study, we used three PAH: DMBA, TCDD, and BP. agents for the treatment of pain and inflammation. Since the DMBA is a well-established model compound that induces 1970s, research has increasingly suggested that NSAIDs may be mammary gland tumors in rodents (34). TCDD is a major effective chemopreventive agents against colorectal and other environmental pollutant, generated by a variety of industrial gastrointestinal tumors. NSAIDs including sulindac, celecoxib, processes that accumulates in fatty tissues of exposed animals. and aspirin have been shown to be among the most effective Humans’ primary source of exposure is through food intake, types of preventive agents against colorectal cancer; however, specifically the consumption of fish, meat, and dairy products. the mechanism of action has not been fully understood (1). BP is an environmental contaminant found in coal tar, auto Our laboratory has identified NSAIDs that modulate the exhaust, tobacco smoke, and charbroiled foods. PAHs bind to activity of the AhR and affect CYP gene transcription and the ligand-dependent AhR, initiating its activation and the

www.aacrjournals.org 1969 Clin Cancer Res 2009;15(6) March 15, 2009 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Cancer Therapy: Preclinical transcription of genes under its control. TCDD is the most (12–14). We undertook the current study to evaluate the effects potent ligand of the AhR. The PAHs are procarcinogens that are of ACS 15 and 18 on the AhR signaling pathway and activated to highly reactive metabolites by the phase I enzymes, carcinogen-metabolizing enzyme gene expression in vitro.We CYP1A1, CYP1A2, and CYP1B1. The reactive metabolites can investigated the effects of ACS 15 and 18 on phase I and phase II enter the nucleus of cells and form adducts with DNA causing enzymes in HepG2 human hepatoma and LS180 human colon damage and mutations that may initiate carcinogenesis. adenocarcinoma cells. HepG2 cells represent an aspect of liver Alternatively, the epoxide products generated by CYP metabo- metabolism for carcinogens, and both HepG2 and LS180 cells lism may be further detoxified by pathways involving phase II have been extensively used in studies of PAH metabolism and enzymes including GSTA2, glutamate cysteine ligase catalytic have a well-characterized AhR pathway (36, 37). subunit, glutamate cysteine ligase modifier, GR, and glucose-6 We examined the effects of ACS 15 and 18 on PAH-induced phosphate dehydrogenase. These enzymes mediate the bio- mRNA expression of CYP1A1, CYP1A2, and CYP1B1 in HepG2 transformation of xenobiotics into more hydrophilic and easily and LS180 cells. We showed that both ACS 15 and 18 inhibited excretable products via the conjugation of glutathione to the TCDD-induced CYP mRNA expression in a concentration- xenobiotic. dependent manner (Fig. 2). We showed the inhibitory effect Sulindac and diclofenac have been shown to inhibit PAH- on CYP mRNA was mediated at the transcriptional level. induced tumorigenesis in animal models, including DMBA- The observed decrease in CYP1A1 heterogeneous nuclear RNA induced mammary cancer and azoxymethane-induced colon levels confirmed the decreased transcription rate, as heteroge- cancer (17, 35). In response to the previously documented neous nuclear RNA levels are not affected by posttranscriptional chemopreventive potential of sulindac and diclofenac, which processing, transport, or stability. Cotreatment of ACS 15 or 18 is complicated by undesirable gastrointestinal side effects, the with TCDD also blocked the binding of TCDD-activated AhR to novel compounds ACS 15 and ACS 18 were generated. These the CYP1A1 enhancer sequence as shown by ChIP assay (Fig. 3C). ACS compounds have shown the ability to modify angiogenesis In agreement with our previous study that showed that and to decrease gastrointestinal and cardiovascular toxicity sulindac is likely a weak agonist of the AhR (18), Fig. 4A

Fig. 5. Effect of ACS 15 and 18 on recombinant CYP enzyme activity. EROD activity of 0.75 Amol/L CYP1A1supersomes, 2.5 Amol/L CYP1A2supersomes, or 3.5 Amol/L CYP1B1supersomes was measured in the presence of DMSO (control) or the indicated concentrations of (A)ACS15or(B) ACS 18. Points, mean (n = 4); bars, SE. EROD activity of 0.75 Amol/L CYP1A1supersomes was measured at the indicated concentrations of ethoxyresorufin (substrate) with increasingconcentrations of ( C)ACS15and (D) ACS18. Results are represented by Lineweaver Burk plot, n =4.

Clin Cancer Res 2009;15(6) March 15, 2009 1970 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Modulation of Phase I and II Enzymes by S-NSAIDs

in 3- to 4-fold increases in hepatic cytosolic glutathione S-transferase activities and mRNA levels for some a, A, and k isoforms. The protective efficacy of oltipraz has been shown in animal studies and is currently under investigation in human clinical trials (39, 40). ADT is a dithiolethione that is thought to exert its chemoprotective effects by increasing levels of reduced glutathione (41) as well as inducing the phase II enzymes glutathione S-transferase, QR, and GR (31). Significant levels of cytotoxic protection were achieved with 25 Amol/L of ADT from increases in the phase II enzymes catalase and GR that were <2-fold (41). Dithiolethiones as a class have been recognized as being potent inducers of enzymes related to the maintenance of glutathione sources (42). We investigated the newly synthetic drugs, the dithiolethione-conjugated NSAIDs, ACS 15 and 18, on the expression of phase II genes related to the detoxification Fig. 6. Effect of ACS 15 and 18 on phase II enzyme expression. HepG2 cells were treated for 24 h with 10 or 50 Amol/L ACS 18 and the mRNA expression levels for pathway regulated by glutathione cycles. ACS 18 treatment GSTA2, glutamate cysteine ligase catalytic subunit, glutamate cysteine ligase resulted in 1.5- to 3.5-fold increases in mRNA expression of modifier, GR, and glucose-6 phosphate dehydrogenase were measured by real-time PCR and normalized to 18S RNA. Columns, mean (n = 3); bars, SE. GSTA2, glutamate cysteine ligase catalytic subunit, glutamate cysteine ligase modifier, GR, and glucose-6 phosphate dehydrogenase (Fig. 6). A coordinated increase in these shows induction of EROD activity by both sulindac as well as enzymes would increase the overall availability of glutathione ACS 18 in HepG2 cells. In contrast, ACS 15 and 18 also for conjugation of activated carcinogens, thereby aiding in showed the ability to directly inhibit CYP enzymes with the their detoxification through increased water-solubility and strongest inhibitory effect shown against CYP1A1 enzyme excretion from the body. activity (Fig. 5A and B). Analysis of the kinetics of H2S has been shown to protect cells from oxidative stress recombinant CYP1A1 enzyme inhibition by Hanes-Woolf plot mainly by increasing the production of glutathione (43). ADT showed that the inhibition was competitive, i.e., Km values for has also been shown to be a powerful antioxidant with the the substrate were increased in the presence of ACS 15 and 18, ability to scavenge free radicals, inhibit lipid peroxidation, and whereas the Vmax remained constant. Both ACS 15 and 18 induce catalase activity (41, 44). The mutagenic effects caused also inhibited DMBA and TCDD-induced EROD activity in by TCDD have been attributed to its capability to generate intact cells, in a concentration-dependent manner, indicating reactive oxygen species, lipid peroxidation, and DNA damage the ability of the novel compounds to enter the intact cell and (45, 46). The activation of TCDD is mediated by the AhR interact with the enzyme (Fig. 4B-E). The ability to inhibit signaling pathway (46). We hypothesize that the inhibitory PAH-induced CYP enzyme activity was unique to ACS 15 and effect of the ACS compounds on the AhR, as well as the 18; neither parent compound (sulindac nor diclofenac), nor induction of glutathione-related phase II enzymes, in conjunc- other H2S-releasing compounds, e.g., NaHS, could mimic the tion with the previously documented anti-inflammatory effects inhibitory effect (Fig. 4A). Bhatia et al. (38) showed a similar (12), will contribute to protective effects of ACS 15 and 18 effect between ACS 15 and NaHS against lung inflammation against TCDD-induced oxidative stress. in which ACS 15 showed protective effects, whereas other AhR activation has been associated with an increase in H2S-releasing compounds could not. The authors concluded cyclooxygenase-2 and chronic inflammation leading to in- the effects may be contributed by the slow H2S-releasing creased cancer risk (47, 48). The backbone of the ACS property of ACS 15 (and ACS 18), which may favor an anti- compounds is the parent NSAID whose primary mechanism inflammatory role and other unknown unique activities. In of action is the inhibition of cyclooxygenases. summary, we show for the first time that the dithiolethione- Additionally, in this study, we have shown ACS 15 and ACS modified compounds, ACS 15 and ACS 18, are potent 18 to be potent inhibitors of the AhR signaling pathway that inhibitors of the AhR pathway that block signal transduction simultaneously block carcinogen-activating phase I enzyme initiated by PAHs. This mechanism of action is distinct and activity and expression while enhancing carcinogen detoxifying different from their parent compounds. In particular, in phase II enzymes in vitro. Our results support the model of S- contrast to our previous work on sulindac that showed its NSAIDs as promising chemopreventive agents and identify the chemoprotective mechanism to be through activation of the molecular mechanisms mediated by ACS 15 and 18. ACS 15 and AhR as the first step in the detoxification process, ACS 18 was 18 serve as important agents for prevention at the initiation shown to inhibit the AhR. Both mechanisms can result in stage of carcinogenesis induced by environmental carcinogens. ultimate chemoprotection, especially when coupled with an Animal studies are needed to further confirm the mechanism of induction in phase II enzymes, which we have also shown. action toward chemoprevention in vivo. Oltipraz and related dithiolethiones are an important class of chemopreventive agents. The chemopreventive effect of dithiolethiones has been well-investigated in association with Disclosure of Potential Conflicts of Interest their ability to induce phase II gene expression through the P. Del Soldato and A. Sparatore are shareholders of Sulfridis, Milan, Italy. The Kelch ECH Associating Protein 1-nuclear factor E2 p45-related company has patents on reagents used in this study. The other authors disclosed factor 2 pathway. The administration of oltipraz in rats results no potential conflicts of interest.

www.aacrjournals.org 1971 Clin Cancer Res 2009;15(6) March 15, 2009 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Cancer Therapy: Preclinical

References 1. Ruegg C, Zaric J, Stupp R. Non steroidal anti-inflam- sulindac maintenance treatment. Advantages of a 33. MacDonald CJ, Ciolino HP,Yeh GC.The drugsalicy- matory drugs and COX-2 inhibitors as anti-cancer low-dose nonsteroidal anti-inflammatory drugregi- lamide is an antagonist of the aryl hydrocarbon recep- therapeutics: hypes, hopes and reality. Ann Med men in reversingadenomas exceeding33 months. tor that inhibits signal transduction induced by 2003;35:476 ^ 87. Dis Colon Rectum 1995;38:813^ 30. 2,3,7,8-tetrachlorodibenzo-p-dioxin. Cancer Res 2. Hawkey CJ. Non-steroidal anti-inflammatory drug 17. Wargovich MJ, Jimenez A, McKee K, et al. Efficacy 2004;64:429 ^ 34. gastropathy: causes and treatment. Scand J Gastro- of potential chemopreventive agents on rat colon ab- 34. Huggins C, Grand LC, Brillantes FP. Mammary can- enterol Suppl 1996;220:124 ^ 7. errant crypt formation and progression. Carcinogene- cer induced by a single feeding of polymucular 3. Mukherjee D, Nissen SE,Topol EJ. Risk of cardiovas- sis 2000;21:1149 ^ 55. hydrocarbons, and its suppression. Nature 1961;189: cular events associated with selective COX-2 inhibi- 18. Ciolino HP, MacDonald CJ, Memon OS, Bass SE, 204 ^7. tors. JAMA 2001;286:954 ^ 9. Yeh GC. Sulindac regulates the aryl hydrocarbon 35.Thompson HJ, JiangC, LuJ, et al. Sulfone metabo- 4.Wallace JL, Reuter B, Cicala C, McKnight W, Grisham receptor-mediated expression of Phase 1 metabolic lite of sulindac inhibits mammary carcinogenesis. Can- MB, Cirino G. Novel nonsteroidal anti-inflammatory enzymes in vivo and in vitro. Carcinogenesis 2006; cer Res 1997;57:267 ^ 71. drugderivatives with markedly reduced ulcerogenic 27:1586 ^ 92. 36.TakahashiY,ItohS,ShimojimaT,KamatakiT.Charac- properties in the rat. Gastroenterology 1994;107: 19. WattenbergLW. Chemoprevention of cancer. Can- terization of Ah receptor promoter in human liver cell 17 3 ^ 9 . cer Res 1985;45:1 ^ 8. line, HepG2. Pharmacogenetics 1994;4:219^ 22. 5. Hawkey CJ, JonesJI, Atherton CT, et al. Gastrointes- 20. YangCS, Smith TJ, HongJY. Cytochrome P-450 37. Harper PA, Prokipcak RD, Bush LE, Golas CL, Okey tinal safety of AZD3582, a cyclooxygenase inhibiting enzymes as targets for chemoprevention against AB. Detection and characterization of the Ah receptor nitric oxide donator: proof of concept study in chemical carcinogenesis and toxicity: opportunities for 2,3,7,8-tetrachlorodibenzo-p-dioxin in the human humans. Gut 2003;52:1537 ^ 42. and limitations. Cancer Res 1994;54:1982^ 6s. colon adenocarcinoma cell line LS180. Arch Biochem 6. Fiorucci S, Santucci L,Wallace JL, et al. Interaction of 21. Elferink CJ, Reiners JJ, Jr. Quantitative RT-PCR on Biophys 1991;290:27 ^ 36. a selective cyclooxygenase-2 inhibitor with aspirin CYP1A1heterogeneous nuclear RNA: a surrogate for 38. Bhatia M, Sidhapuriwala JN, Sparatore A, Moore and NO-releasingaspirin in the human gastric mucosa. the in vitro transcription run-on assay. Biotechniques PK. Treatment with H2s-releasingdiclofenac protects Proc Natl Acad Sci U S A 2003;100:10937 ^ 41. 19 9 6;20 :470 ^ 7. mice against acute pancreatitis-associated lung injury. 7. Konturek PC, Kania J, Burnat G, Hahn EG. NO-releas- 22. Guigal N, Seree E, Nguyen QB, Charvet B, Desobry Shock 2008;29:84 ^ 8. ingaspirin exerts strongergrowth inhibitory effect on A, Barra Y. Serum induces a transcriptional activation 39. Liu YL, Roebuck BD, Yager JD, Groopman JD, Barrett’s adenocarcinoma cells than traditional aspirin. of CYP1A1gene in HepG2 independently of the AhR Kensler TW. Protection by 5-(2-pyrazinyl)-4-methyl- J Physiol Pharmacol 2006;57 Suppl 12:15 ^ 24. pathway. Life Sci 2001;68:2141 ^ 50. 1,2-dithiol-3-thione (oltipraz) against the hepatotoxic- 8.WangR.Two’s company, three’s a crowd: can H2S be 23. Ciolino HP, Bass SE, MacDonald CJ, ChengRY, ity of aflatoxin B1in the rat. Toxicol Appl Pharmacol the third endogenous gaseous transmitter? FASEB J Yeh GC. Sulindac and its metabolites induce carcino- 1988;93:442^51. 2002;16:1792 ^ 8. gen metabolizing enzymes in human colon cancer 40. Gupta E, Olopade OI, Ratain MJ, et al. Pharma- 9. Fiorucci S, Distrutti E, Cirino G, Wallace JL. The cells. Int J Cancer 2008;122:990 ^ 8. cokinetics and pharmacodynamics of oltipraz as a emerging roles of hydrogen sulfide in the gastroin- 24. Hestermann EV, Brown M. Agonist and chemopre- chemopreventive agent. Clin Cancer Res 1995;1: testinal tract and liver. Gastroenterology 2006;131: ventative ligands induce differential transcriptional co- 113 3 ^ 8 . 259 ^ 71. factor recruitment by aryl hydrocarbon receptor. Mol 41. Khanna S, Sen CK, Roy S, Christen MO, Packer L. 10. Distrutti E, Sediari L, Mencarelli A, et al. 5-Amino- Cell Biol 2003;23:7920 ^ 5. Protective effects of anethole dithiolethione against 2-hydroxybenzoic acid 4-(5-thioxo-5H-[1,2 ^ 3yl)- 25. Denison MS, Nagy SR. Activation of the aryl hydro- oxidative stress-induced cytotoxicity in human Jurkat phenyl ester (ATB-429), a hydrogen sulfide-releasing carbon receptor by structurally diverse exogenous and Tcells. Biochem Pharmacol 1998;56:61 ^ 9. derivative of mesalamine, exerts antinociceptive endogenous chemicals. Annu Rev Pharmacol Toxicol 42. Kwak MK, Egner PA, Dolan PM, et al. Role of phase effects in a model of postinflammatory hypersensitivi- 2003;43:309^34. 2 enzyme induction in chemoprotection by dithiole- ty. J Pharmacol ExpTher 2006;319:447 ^58. 26. TsuchiyaY, Nakajima M,Yokoi T. Critical enhancer re- thiones. Mutat Res 2001;480 ^ 481:305 ^ 15. 11. Pan TT, FengZN, Lee SW, Moore PK, Bian JS. En- gion to which AhR/ARNT and Sp1 bind in the human 43. Yan SK, ChangT, Wang H, Wu L, Wang R, Meng dogenous hydrogen sulfide contributes to the cardio- CYP1B1gene. J Biochem (Tokyo) 2003;133:583 ^ 92. QH. Effects of hydrogen sulfide on homocysteine- protection by metabolic inhibition preconditioningin 27. JiangZ, DaltonTP,Jin L, et al.Toward the evaluation induced oxidative stress in vascular smooth muscle the rat ventricular myocytes. J Mol Cell Cardiol 2006; of function in genetic variability: characterizing human cells. Biochem Biophys Res Commun 2006;351: 40:119^30. SNP frequencies and establishingBAC-transgenic 485^91. 12. Li L, Rossoni G, Sparatore A, Lee LC, Del Soldato P, mice carryingthe human CYP1A1_CYP1A2 locus. 44. Sen CK,Traber KE, Packer L. Inhibition of NF-nBac- Moore PK. Anti-inflammatory and gastrointestinal Hum Mutat 2005;25:196 ^ 206. tivation in human T-cell lines by anetholdithiolthione. effects of a novel diclofenac derivative. Free Radic Biol 28. Ueda R, Iketaki H, Nagata K, et al. A common regu- Biochem Biophys Res Commun 1996;218:148 ^ 53. Med 2007;42:706 ^ 19. latory region functions bidirectionally in transcriptional 45. Stohs SJ. Oxidative stress induced by 2,3,7,8- 13. IsenbergJS, Jia Y, Field L, et al. Modulation of an- activation of the human CYP1A1and CYP1A2 genes. tetrachlorodibenzo-p-dioxin (TCDD). Free Radic Biol giogenesis by dithiolethione-modified NSAIDs and Mol Pharmacol 2006;69:1924^ 30. Med 1990;9:79 ^ 90. valproic acid. Br J Pharmacol 2007;151:63 ^ 72. 29. Doostdar H, Burke MD, Mayer RT. Bioflavonoids: 46. Alsharif NZ, Lawson T, Stohs SJ. Oxidative stress 14. Rossoni G, Sparatore A, Tazzari V, Manfredi B, Del selective substrates and inhibitors for cytochrome induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin is Soldato P, Berti F. The hydrogen sulphide-releasing P450 CYP1A and CYP1B1. Toxicology 2000;144: mediated by the aryl hydrocarbon (Ah) receptor com- derivative of diclofenac protects against ischaemia- 31 ^ 8. plex. Toxicology 1994;92:39 ^ 51. reperfusion injury in the isolated rabbit heart. Br J 30. Talalay P, Fahey JW. Phytochemicals from crucifer- 47. Vogel CF, Li W, Sciullo E, et al. Pathogenesis of aryl Pharmacol 2007. ous plants protect against cancer by modulating car- hydrocarbon receptor-mediated development of lym- 15. Wallace JL, Caliendo G, Santagada V, Cirino G, cinogen metabolism. J Nutr 2001;131:3027 ^ 33S. phoma is associated with increased cyclooxygenase- FiorucciS.Gastrointestinalsafetyandanti-inflammato- 31. Ansher SS, Dolan P, BuedingE. Biochemical 2 expression. Am J Pathol 2007;171:1538 ^ 48. ryeffects ofahydrogensulfide-releasing diclofenacde- effects of dithiolthiones. Food Chem Toxicol 1986; 48. Martey CA, Baglole CJ, Gasiewicz TA, Sime PJ, rivative in the rat. Gastroenterology 2007;132:261 ^ 71. 24:405 ^ 15. Phipps RP.The aryl hydrocarbon receptor is a regula- 16. Winde G, Schmid KW, Schlegel W, Fischer R, 32. Kwak MK, Ramos-Gomez M, Wakabayashi N, tor of cigarette smoke induction of the cyclooxyge- Osswald H, Bunte H. Complete reversion and preven- Kensler TW. Chemoprevention by 1,2-dithiole-3-thio- nase and prostaglandin pathways in human lung tion of rectal adenomas in colectomized patients with nes through induction of NQO1 and other phase 2 fibroblasts.AmJPhysiolLungCellMolPhysiol familial adenomatous polyposis by rectal low-dose enzymes. Methods Enzymol 2004;382:414^ 23. 2005;289:L391 ^9.

Clin Cancer Res 2009;15(6) March 15, 2009 1972 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Novel Dithiolethione-Modified Nonsteroidal Anti-Inflammatory Drugs in Human Hepatoma HepG2 and Colon LS180 Cells

Sara E. Bass, Pawel Sienkiewicz, Christopher J. MacDonald, et al.

Clin Cancer Res 2009;15:1964-1972.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/15/6/1964

Cited articles This article cites 47 articles, 9 of which you can access for free at: http://clincancerres.aacrjournals.org/content/15/6/1964.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/15/6/1964.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/15/6/1964. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.