CYP2S1 and CYP2W1 Mediate 2-(3,4-Dimethoxyphenyl)-5-Fluorobenzothiazole (GW-610, NSC 721648) Sensitivity in Breast and Colorectal Cancer Cells Free

Previously, we have described the potent and selective antitumor activity of 2-(4-aminophenyl)benzothiazoles in vitro (1, 2) and in vivo (3–5). Sensitivity to these agents is clustered within breast [irrespective of estrogen receptor (ER) status] and ovarian human tumor models and correlates with cytochrome P450 1A1 (CYP1A1) inducibility within these tumors (6, 7). As aryl hydrocarbon receptor (AhR) ligands, aminophenylbenzothiazoles act as both inducer and substrate for CYP1A1-catalyzed bioactivation to electrophilic species in sensitive phenotypes (8). A study comparing efficacy of 2-(4-amino-3-methylphenyl)-5-fluorobenzothiazole (5F-203; Fig. 1A), lysyl amide (Phortress), and doxorubicin in mammary xenografts revealed Phortress to be equiactive against 6 xenograft models, outperforming doxorubicin in 1 model (5). Significant in vivo antitumor activity has also been seen in IGROV-1 ovarian xenografts. Worldwide, 1.38 million women were diagnosed with breast cancer in 2008. Although the 10-year survival rate is greater than 75%, 458,000 women were killed by this disease in the same year. Ovarian cancer represents 2% of all cancers diagnosed and has a 5-year survival rate of less than 40%. Thus, discovery and development of novel agents to treat these malignancies are required. Phortress is currently undergoing clinical evaluation under the auspices of Cancer Research UK.

Subsequent synthesis of a series of dimethoxyphenylbenzothiazoles led to identification of a novel antitumor pharmacophore bearing oxygen substituents. The lead compound of this series, 2-(3,4-dimethoxyphenyl)-5-fluorobenzothiazole (GW-610; NSC 721648; Fig. 1A), elicits strikingly selective in vitro growth-inhibitory activity after 48-hour exposure to GW-610; exquisite sensitivity was observed in colon as well as certain non–small-cell lung, breast, ovarian, and renal cancer cell lines in the NCI human-derived 60 cell line panel (GI₅₀ values <100 nmol/L; Supplementary Fig. S1). Thus, the spectrum of activity of this bis-methyl ether extends to human cell lines derived from intractable colorectal carcinomas (CRC). A potent AhR ligand (K_i = 6.8 nmol/L), GW-610 is sequestered exclusively and rapidly by sensitive breast and CRC cells only (9, 10).

CRC is the third most common cancer worldwide and caused more than 600,000 deaths globally in 2008. Successful surgery underlies the 5-year survival rate statistics which hover around 50% for both sexes. Thus, for inoperable CRC, discovery of novel chemotherapeutic agents is imperative. It has recently been reported that novel AhR-inducible cytochrome P450 2W1 (CYP2W1) is overexpressed in human colon cancer tissue (11, 12). CYP2W1 expression is tumor-specific and regulated by DNA methylation status, that is, hypomethylation in the body of CYP2W1 results in enhanced protein expression. Edler and colleagues reported that 36% CRCs expressed high CYP2W1 protein levels, which correlated with worse clinical outcome (P = 0.007; ref. 13). Thus, expression of CYP2W1 is an independent prognostic factor in patients with stage II or stage III CRC. CYP2S1 expression is found in epithelial cells that are targets for carcinogen exposure including epithelia of the respiratory and gastrointestinal tracts (common sites of tumorigenesis; refs. 14–16). Indeed, strong CYP2S1 expression has been detected in tumors of epithelial origin (including CRC and lung; refs. 14, 17, 18).

In this article, we describe experiments undertaken to elucidate the roles, if any, of novel AhR-inducible CYP isoforms 2S1 and 2W1 in biotransformation of GW-610 and 5F-203, in 4 human-derived carcinoma cell lines-ER⁺ MCF-7, ER⁻ MDA-MB-468 breast, and HCC2998, KM12 CRC cells. The role of CYP1A1 in benzothiazole bioactivation has been compared.

MDA-MB-468 and MCF-7 human breast cancer cell lines and MRC-5 human embryonic lung fibroblast cells were purchased from the American Type Culture Collection. KM12 and HCC2998 CRC cell lines were kind gifts of the National Cancer Institute, Bethesda, MD. All cells were expanded on receipt and tested for mycoplasma using MycoAlert Mycoplasma Detection Assay (Lonza) before banking in liquid nitrogen. All cells were maintained in RPMI-1640 medium containing 10% FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich) at 37°C, 5% CO₂. Cells underwent passage in culture for less than a month and no further authentication was done by the authors.

Concentration–response curves and IC₅₀ values were determined using the MTT cell viability assay as described previously (8, 19, 20). Cells were seeded in 96-well plates for 24 hours at a density of 2 × 10³ cells per well. Concentrated (100 mmol/L) solutions of 5F-203 and GW-610 were prepared in dimethyl sulfoxide. Cell viability after 72 hours of drug exposure was determined by cell-mediated MTT reduction. Cell growth and drug activity was determined by measuring absorbance at 570 nm using a TECAN Infinite F200 plate reader (Männedorf).

Total RNA from cells was extracted using the Qiagen RNA Isolation Kit (Qiagen) and first-strand cDNA was synthesized using High Capacity RNA-to-cDNA Master Mix (Applied Biosystems) according to the manufacturer's instructions. Gene expression levels were measured by real-time quantitative PCR (qPCR) using the FastStart Universal SYBR Green Master Reagent (Roche) and a Biorad iQ5 real-time PCR detector system (Bio-Rad). Data analysis was carried out using Biorad iQ5 Optical System Software V1.0. The specific forward and reverse primer sequences are shown in Supplementary Table S1. The conditions for all qPCR reactions were as follows: 3 minutes at 94°C followed by 40 seconds at 94°C, 40 seconds at 60°C, and 25 seconds at 72°C for 40 cycles. The expression data was normalized against GAPDH as housekeeping gene.

Protein lysates from cells were extracted in ice-cold lysis buffer (0.75% NP-40, 1 mmol/L DTT, and protease inhibitors in PBS). Total protein (50 μg) was subjected to SDS-PAGE followed by immunoblotting. Primary polyclonal antibodies against CYP1A1 (ab3568; 1:1,000), CYP2S1 (ab69650; 1:1,000), and CYP2W1 (ab76666; 1:100) were obtained from Abcam. Mouse monoclonal antibody against α-actin (clone C-2; 1:250) was obtained from Santa Cruz Biotechnology.

Lentiviral short-hairpin RNA (shRNA) constructs targeting CYP1A1, CYP2S1, and CYP2W1 were purchased from Sigma-Aldrich. High-titer lentiviral stocks were generated by cotransfection with packaging plasmids psPAX2 (Addgene plasmid 12260) and envelope plasmids pMD2.G (Addgene plasmid 12259) into HEK-293T cells as reported previously (19–23). The shRNA target sequences for CYP1A1, CYP2S1, and CYP2W1 are shown in Supplementary Table S2. Stable pools were generated by transduction of 2 independent lentiviral shRNA constructs targeting each CYP isoform in the breast cancer and CRC cells, followed by brief drug selection with puromycin.

To show the selective antitumor activity of 5F-203 and GW-610, we compared their antiproliferative activities in breast cancer cells (MDA-MB-468 and MCF-7), CRC cells (KM12 and HCC2998), and normal lung fibroblasts (MRC-5). As shown in Fig. 1A, 5F-203 elicits activity against MDA-MB-468 and MCF-7 breast cancer cell lines (IC₅₀ < 1 μmol/L) whereas CRC cells, KM12 and HCC2998, were relatively resistant to 5F-203 (IC₅₀ > 10 μmol/L; Table 1). Similarly, GW-610 also shows potent antitumor activity against the 2 breast cancer cell lines. However, unlike 5F-203, GW-610 also potently inhibits the growth of KM12 and HCC2998 CRC cells (IC₅₀ values < 0.1 μmol/L). These results are consistent with the NCI-60 cell lines screen, which showed selectivity of 5F-203 and GW-610 against different tumor types (Supplementary Figs. S1 and 2). Notably, MRC-5 embryonic lung fibroblasts are inherently resistant to 5F-203 and GW-610, demonstrating the selective growth-inhibitory activity of benzothiazole agents. Estimated IC₅₀ values of 5F-203 and GW-610 in various cell lines are shown in Table 1.

Previous studies have shown that CYP1A1 is expressed in breast cancer cell lines and is essential for bioactivation of 5F-203 to active metabolites that are capable of causing DNA damage (6, 8, 24). However, the expression of the newly identified extrahepatic CYP2S1 and CYP2W1 in breast cancer and CRC cell lines was not investigated. Using highly sensitive real-time qPCR and chemiluminescence immunoblotting, we showed that CYP1A1, CYP2S1, and CYP2W1 mRNA and protein are expressed in breast cancer (MDA-MB-468 and MCF-7) and CRC (KM12 and HCC2998) cells but not in MRC-5 lung fibroblast (data not shown). The highest expression of CYP1A1 and CYP2W1 was observed in MDA-MB-468 and MCF-7 cells, respectively (Fig. 1B and C). Interestingly, high expression of CYP2S1 was observed mainly in the CRC cells, whereas only very low CYP2S1 protein levels are present in breast cancer cells (Fig. 1B and C). When basal expression of these CYP isoforms was compared with 5F-203 and GW-610 cellular sensitivity, no direct correlation was observed. These results suggest that, although the CYP isoforms might play an important role in the bioactivation of the compounds, their endogenous levels might not be a good predictive marker for treatment response.

Potent AhR ligands, 5F-203 and GW-610 have been shown to activate AhR and induce CYP1A1 expression in MCF-7 cells (25, 26). Recently, human CYP2S1 and CYP2W1 isoforms have also been shown to be transcriptionally regulated by AhR (14, 27, 28). As such, we investigated whether 5F-203 and GW-610 are able to induce CYP2S1 and CYP2W1 expression. To test this hypothesis, MDA-MB-468, MCF-7, KM12, and HCC2998 cells were treated with 100 nmol/L and 1 μmol/L 5F-203 or GW-610 for 48 hours and their mRNA and protein isolated for qPCR and immunoblotting analyses, respectively.

Induction of CYP1A1 mRNA and protein was observed in all cell lines on treatment of cells with 5F-203 and GW-610, including KM12 and HCC2998 CRC cells which are inherently resistant to 5F-203 (Fig. 2A). CYP2S1 and CYP2W1 mRNA and protein expression were also induced in MDA-MB-468 and MCF-7 breast cancer cells following 5F-203 and GW-610 treatment (Figs. 2B and C and 3). In stark contrast, no induction of CYP2S1 or CYP2W1 was observed in KM12 and HCC2998 CRC cells. In fact, CYP2S1 and CYP2W1 expression decreased following treatment of HCC2998 cells with 5F-203 (Fig. 2B and C). These results show that 5F-203 and GW-610 are capable of inducing CYP1A1 in both breast cancer and CRC cells, whereas CYP2S1 and CYP2W1 were induced only in breast cancer cells. Thus, the differential regulation of CYP1A1, CYP2S1, and CYP2W1 in breast cancer and CRC cells might play an important role in the regulation of 5F-203 and GW-610 antitumor activity.

To further investigate the role of CYP1A1, CYP2S1, and CYP2W1 in the regulation of 5F-203 and GW-610 antitumor activities, we generated stable cell lines after depletion of individual CYP isoforms. Cellular sensitivity to 5F-203 and GW-610 was compared with corresponding isogenic vector control cells. The mRNA and protein levels of CYP1A1, CYP2S1, and CYP2W1 in MDA-MB-468, MCF-7, KM12, and HCC2998 following gene knockdown were confirmed by qPCR and immunoblotting, respectively (Fig. 4; Supplementary Fig. S3).

Next, we tested whether knockdown of CYP1A1, CYP2S1, or CYP2W1 affected cellular sensitivity to 5F-203 and GW-610. As expected, depletion of CYP1A1 markedly reduced the sensitivity of cancer cells to 5F-203 and GW-610, suggesting that CYP1A1 is required for benzothiazole bioactivation (Fig. 5 and Table 2).

In contrast to CYP1A1 knockdown, depletion of CYP2S1 increased the sensitivity of MDA-MB-468, KM12, and HCC2998 to 5F-203 and GW-610 (Fig. 6 and Table 2). This sensitization, however, was not observed in MCF-7 cells, which expressed only low CYP2S1 levels (Fig. 1C). Notably, depletion of CYP2S1 in KM12 and HCC2998 cells, which are inherently resistant to 5F-203, caused a marked increase in sensitivity to 5F-203, suggesting that CYP2S1 might be involved in deactivation of benzothiazoles.

Although depletion of CYP1A1 or CYP2S1 affected the sensitivity of both breast and CRC cells to 5F-203 and GW-610, CYP2W1 depletion impacted only GW-610 sensitivity in CRC cells. As shown in Fig. 7 and Table 2, knockdown of CYP2W1 reduced sensitivity to GW-610 in HCC2998 and KM12 more than 10- to 100-fold, respectively. However, no such effect was observed in CYP2W1-depleted breast cancer cells following treatment with GW-610. Moreover, there was no change in sensitivity of breast cancer or CRC cells to 5F-203 following CYP2W1 depletion (Fig. 7). These results suggest that CYP2W1 specifically mediates GW-610 sensitivity in CRC cells and might explain the distinctive selectivity of GW-610 toward CRC cells.

Both 5F-203 and GW-610 are representative molecules of 2 series of antitumor agents distinct from all mechanistic classes of chemotherapeutic agents in current clinical use (8, 24, 29, 30). Exquisitely potent and highly selective 5F-203 antitumor activity has been observed in certain human tumor-derived models in vitro (including breast, renal, and ovarian cell lines) and in vivo against breast and ovarian xenograft models (1, 5, 31). GW-610, on the other hand, exhibits a broader antitumor spectrum to include CRC and non–small-cell lung cancer, in addition to breast, renal, and ovarian cancer cell lines (25).

Because 5F-203 and GW-610 are potent ligands for AhR and have been postulated to be metabolically activated by CYP1A1 leading to antitumor activities, we investigated whether other AhR-inducible cytochrome (CYP) P450 isoforms might be mediating the sensitivity of cancer cells to benzothiazole compounds. Using sensitive qPCR and immunoblotting, we showed that the 5F-203 and GW-610 sensitivity in breast cancer and CRC cells does not correlate with CYP1A1 and CYP2W1 basal expression (Fig. 1A; Table 1). Interestingly, the 2 CRC cell lines, KM12 and HCC2998, which were resistant to 5F-203 (but sensitive to GW-610) expressed high basal level of CYP2S1, suggesting that CYP2S1 might play a role in 5F-203 or GW-610 resistance (Fig. 1B and C).

To further understand the mechanism of action of 5F-203 and GW-610, we investigated the inducibility of CYP1A1, CYP2S1, and CYP2W1 following 48-hour exposure of cells to benzothiazole analogues. As expected, CYP1A1 was strongly induced in mammary carcinoma cells following treatment with 5F-203 or GW-610 (Figs. 2A and 3). Surprisingly, we also observed strong CYP1A1 induction in KM12 and HCC2998 CRC cells which are inherently resistant to 5F-203. These results suggest that 5F-203 might be activated in CRC cells but its activity is suppressed. Tumor-specific CYP1B1, for example, has been shown to metabolize 2-(4-amino-3-methylphenyl)benzothiazole to hydroxylated biotransformation products devoid of activity (6). CYP2S1 (highly expressed in CRC cells)-catalyzed biotransformation represents a plausible mechanism of metabolic deactivation (as discussed below). Interestingly, both CYP2S1 and CYP2W1 were also induced by 5F-203 and GW-610 (Figs. 2B and C and 3). Although AhR-inducible, unlike CYP1A1, CYP2S1, and CYP2W1 further induction was observed only in breast, and not CRC, cancer cells.

To investigate the functional roles of CYP1A1, CYP2S1, and CYP2W1 in mediating sensitivity of cancer cells to 5F-203 and GW-610, we conducted gene knockdown by lentiviral shRNA targeting each specific CYP P450 isoforms. As expected, CYP1A1, which was postulated to metabolically activate benzothiazoles, is required for 5F-203 and GW-610 activity. Depletion of CYP1A1 from breast and CRC cells abrogated the antitumor activity of 5F-203 and GW-610 (Figs. 4 and 5). This result is consistent with a previous study which showed that cotreatment with a CYP1A1 inhibitor, resveratrol, and 5F-203 dramatically decreased 5F-203–induced CYP1A1 expression, correspondingly diminishing 5F-203 antitumor activity. Together, these results provide direct evidence that CYP1A1 induction is crucial for antitumor activity of benzothiazoles.

Similarly, gene knockdown was also carried out to investigate the role of CYP2S1 and CYP2W1 in 5F-203 and GW-610 activity. Surprisingly, depletion of CYP2S1 dramatically increased sensitivity to 5F-203 and GW-610 (∼10-fold) in MDA-MB-468 breast, KM12, and HCC2998 CRC cells. MCF-7 cells, which express very low CYP2S1 levels, defied this trend (Fig. 6). Interestingly, the 2 CRC cell lines which expressed high endogenous level of CYP2S1 and are inherently resistant to 5F-203 became sensitive to this agent after depletion of CYP2S1. These results suggest that CYP2S1 metabolically inactivates 5F-203 and GW-610 to an as yet unidentified secondary metabolite. Reports show that human CYP2S1 plays an important role in detoxification of many toxic substrates to less toxic metabolites including benzo[a]pyrene, 7,12-dimethylbenz[a]anthracene, aflatoxin B1, naphthalene, and styrene (32). Whether CYP2S1 is targeting inactivation of parental 5F-203 and GW-610 or their CYP1A1-bioactivated electrophilic species remains to be determined. However, on the basis of the observed induction of CYP1A1 in 5F-203–resistant CRC cells and sensitization of these cells following CYP2S1 depletion, it is likely that CYP2S1 metabolically inactivates the electrophilic species generated from benzothiazoles.

Knockdown of CYP2W1, on the other hand, significantly reduced the sensitivity of CRC cells, but not breast cancer cells, to GW-610. Interestingly, depletion of CYP2W1 did not affect the sensitivity of breast and CRC cells to 5F-203, suggesting that CYP2W1 is important for activation of GW-610 in CRC cells only. Thus, these results provide an alternative mechanism for GW-610 activation distinct from 5F-203.

In conclusion, we have showed that 5F-203 and GW-610 are capable of inducing CYP2S1 and CYP2W1 in breast cancer cells. Knockdown of CYP2S1 sensitized both breast cancer and CRC cells to 5F-203 and GW-610, suggesting that CYP2S1 is able to catalyze deactivation of these agents. Knockdown of CYP2W1 markedly reduced GW-610 sensitivity in CRC cells only, implicating an important role for CYP2W1 in bioactivation of GW-610 in CRC cells. Because of the exclusive metabolic activation of GW-610 by CYP2W1, which has recently been shown to be a tumor-specific drug target in CRC, GW-610 offers tremendous potential for translation into a therapeutic agent for CRC.

M.F.G. Stevens and T.D. Bradshaw are listed as inventors on patents covering the intellectual property of 5F-203 and GW-610. M.F.G. Stevens and T.D. Bradshaw are consultants and shareholders of Pharminox Ltd., which has a commercial interest in antitumor benzothiazoles.

The authors would like to acknowledge Dr Geoff Wells (School of Pharmacy, University of London, UK) for his original synthesis of GW-610. This is part 32 of the series on “Antitumour 2-(4-aminophenyl)benzothiazoles.”

This work was financially supported by research grants from the International Medical University (B.S. Tan, K.H. Tiong, H.L. Choo, and C.-O. Leong) and by the International Medical University BMedSci Research Training Program (A. Muruhadas and N. Randhawa).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.