Cytoprotective effects of (E)-N-(2-(3, 5-dimethoxystyryl) phenyl) furan-2-carboxamide (BK3C231) against 4-nitroquinoline 1-oxide-induced damage in CCD-18Co human colon fibroblast cells

Stilbenes are a group of chemicals characterized with the presence of 1,2-diphenylethylene. Previously, our group has demonstrated that synthesized (E)-N-(2-(3, 5-dimethoxystyryl) phenyl) furan-2-carboxamide (BK3C231) possesses potential chemopreventive activity specifically inducing NAD(P)H:quinone oxidoreductase 1 (NQO1) protein expression and activity. In this study, the cytoprotective effects of BK3C231 on cellular DNA and mitochondria were investigated in normal human colon fibroblast, CCD-18Co cells. The cells were pretreated with BK3C231 prior to exposure to the carcinogen 4-nitroquinoline 1-oxide (4NQO). BK3C231 was able to inhibit 4NQO-induced cytotoxicity. Cells treated with 4NQO alone caused high level of DNA and mitochondrial damages. However, pretreatment with BK3C231 protected against these damages by reducing DNA strand breaks and micronucleus formation as well as decreasing losses of mitochondrial membrane potential (ΔΨm) and cardiolipin. Interestingly, our study has demonstrated that nitrosative stress instead of oxidative stress was involved in 4NQO-induced DNA and mitochondrial damages. Inhibition of 4NQO-induced nitrosative stress by BK3C231 was observed through a decrease in nitric oxide (NO) level and an increase in glutathione (GSH) level. These new findings elucidate the cytoprotective potential of BK3C231 in human colon fibroblast CCD-18Co cell model which warrants further investigation into its chemopreventive role.


Introduction
Cancer-related mortality has increased tremendously and is expected to further increase despite emerging medical improvements [1]. The global incidence of cancer is estimated to have risen in 2018 with colorectal cancer being the third most commonly diagnosed cancer and is ranked second in terms of mortality due to poor prognosis worldwide [2]. In Malaysia, cancer is the third most common cause of death after cardiovascular diseases and respiratory diseases. According to Malaysia National Cancer Registry (MNCR) Report 2007-2011, colorectal cancer is the second most common cancer [3]. Advances in costly surgical and medical therapies for primary and metastatic colorectal cancer have had limited impact on cure rates and long-term survival [4,5]. The shift of momentum towards chemoprevention is the result of several observations. These include local recurrences after surgery, treatment-induced long-term complications, chemotherapyinduced adverse effects, as resistance to chemotherapy due to multidrug resistance phenotypes and tumour heterogeneity [6][7][8].
Chemopreventive approaches have effectively decreased cancer incidence rates such as for lung cancer and cervical cancer [2]. The cancer chemoprevention approach exploits nontoxic natural or synthetic pharmacological agents to prevent, block or reverse the multistep processes of carcinogenesis [9,10]. Chemopreventive agents inhibit the invasive development of cancer by affecting the three defined stages of carcinogenesis namely initiation, promotion and progression which are induced by carcinogens through genetic and epigenetic mechanisms [11,12].
Exposure of cells to carcinogens causes DNA mutation and leads to accumulation of additional genetic changes through sustained cell proliferation. This rapid and irreversible process is known as tumour initiation, the first stage of carcinogenesis. Tumour promotion, which is referred to as the lengthy and reversible second stage of carcinogenesis, involves the selective clonal expansion of initiated cells to produce preneoplastic lesions which enables additional mutations to accumulate. The final stage of carcinogenesis, tumour progression, involves neoplastic transformation after accumulating chromosomal aberrations and karyotypic instability resulting in metastatic malignancy [13,14].
Altered cellular redox status and disrupted oxidative homeostasis play key roles in cancer development by enhancing DNA damage and modifying key cellular processes such as cell proliferation and apoptosis [15]. Oxidative/nitrosative stress is the result of disequilibrium between reactive oxygen species (ROS)/reactive nitrogen species (RNS) and antioxidants [16]. If oxidative/nitrosative stress persists, this may lead to modification of signal transduction and gene expression, which in turn may lead to mutation, transformation and progression of cancer [17,18].
Stilbenes are produced in the skin, seeds, leaves and sapwood of a wide variety of plant species including dicotyledon angiosperms such as grapevine (Vitis vinifera), peanut (Arachis hypogaea) and Japanese knotweed (Fallopia Japonica); monocotyledons like sorghum (Sorghum bicolor) and gymnosperms such as several Pinus and Picea species [19][20][21]. They are a well-known class of naturally occurring phytochemicals acting as antifungal phytoalexins, providing protection against UV light exposure and also involved in bacterial root nodulation and coloration [19,[22][23][24]. These compounds bear the core structure of 1,2-diphenylethylene in which two benzene rings are separated by an ethanyl or ethenyl bridge [25].
Despite being known as plant defense compounds, stilbenes have an enormous diversity of effects on biological and cellular processes applicable to human health, particularly in chemoprevention. Resveratrol, as the biosynthetic precursor of most oligostilbenoids, has been known to possess a myriad of biological activities such as anticancer, antioxidant, anti-aging, antimicrobial, cardioprotection, anti-diabetes, anti-obesity, and anti-inflammation [26][27][28][29][30][31][32][33]. However, low water solubility and poor bioavailability are the major setbacks to the exploitation of these biological activities [34,35].
Our group has previously demonstrated that synthetic stilbene BK3C231 (Fig 1) potently induced antioxidant gene NQO1 as a detoxifying mechanism in human embryonic hepatocytes, WRL-68 cells [36]. Therefore in this study, we propose to elucidate the cytoprotective effects of BK3C231 using normal human colon fibroblast CCD-18Co cells. We anticipate this study will build a strong base and accelerate the development of BK3C231 as a potential drug for chemoprevention.

MTT cytotoxicity assay
CCD-18Co cells were seeded in 96-well microplate (Nest Biotechnology, Jiangsu, China) at the concentration of 5 x 10 4 cells/mL in a volume of 200 μL per well. The seeded cells were incubated under 5% CO 2 at 37˚C for 24 hours prior to respective compound treatments at different timepoints. After incubation, 20 μL of Thiazolyl Blue Tetrazolium Bromide (MTT; Sigma-Aldrich, St. Louis, MO, USA) solution (5mg/mL in PBS) was added to the treated cells and further incubated for 4 hours at 37˚C. Subsequently, the total medium in each well was discarded and the crystalline formazan was solubilised using 200 μL DMSO. For complete dissolution, the plate was incubated for 15 minutes followed with gentle shaking for 5 minutes. The cytotoxicity of BK3C231 and 4NQO was assessed by measuring the absorbance of each well at 570 nm using iMark TM microplate reader (Bio-Rad Laboratories, Hercules, CA, USA). Mean absorbance for each compound concentration was expressed as a percentage of vehicle control absorbance and plotted versus compound concentration. The inhibitory concentration that kills 50% of cell population (IC 50 ) represents the compound concentration that reduced the mean absorbance at 570 nm to 50% of those in the vehicle control wells. [dx.doi.org/10.17504/ protocols.io.bdp6i5re]

Alkaline comet assay
Seeded cells (5 x 10 4 cells/mL) in 6-well plate (Nest Biotechnology, Jiangsu, China) were pretreated with BK3C231 at 6.25 μM, 12.5 μM, 25 μM and 50 μM for 2 hours prior to 4NQO treatment at 1 μM for 1 hour. Following incubation, detached cells in the medium were collected and added back to trypsinised cells. Then, the cell suspension was transferred to the tube for centrifugation (450 x g/5 minutes at 4˚C). The supernatant was removed and pellet was washed with Ca 2+ -and Mg 2+ -free PBS and re-centrifuged. The pellets left at the bottom were mixed thoroughly with 80 μl of 0.6% w/v LMA (Sigma-Aldrich, St. Louis, MO, USA). The mixture was then pipetted onto the hardened 0.6% w/v NMA (Sigma-Aldrich, St. Louis, MO, USA) as the first layer gel on the slide. Cover slips were placed to spread the mixture and slides were left on ice for LMA to solidify. Following removal of the cover slips, the embedded cells were lysed in a lysis buffer containing 2.5M NaCl (Merck Milipore, Burlington, MA, USA), 1 mM Na 2 EDTA (Sigma-Aldrich, St. Louis, MO, USA), 10 mM Tris (Bio-Rad Laboratories, Hercules, CA, USA) and 1% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) overnight at 4˚C. After lysis, the slides were soaked in electrophoresis buffer solution for 20 minutes for DNA unwinding before electrophoresis at 300 mA, 25V for 20 minutes. Subsequently, the slides were rinsed with neutralising buffer for 5 minutes and stained with 30 μL of 50 μg/mL ethidium bromide (EtBr; Sigma-Aldrich, St. Louis, MO, USA) solution. Slides were left overnight at 4˚C before analyzing with Olympus BX51 fluorescence microscope (Tokyo, Japan) equipped with 590 nm filter. DNA damage scoring was performed on 50 cells per slide whereby tail moment representing the product of tail length and fraction of total DNA in tail was quantified using Comet Score TM software (TriTek Corp, Sumerduck, VA, USA). [dx.doi.org/10.17504/ protocols.io.bdqgi5tw]

Cytokinesis-block micronucleus (CBMN) assay
Seeded cells (5 x 10 4 cells/mL) in 6-well plate were pretreated with BK3C231 at 6.25 μM, 12.5 μM, 25 μM and 50 μM for 2 hours prior to 4NQO treatment at 1 μM for 2 hours. After incubation, cells were treated with 4.5 μg/mL Cytochalasin B (Sigma-Aldrich, St. Louis, MO, USA) for 24 hours to block cytokinesis. The cells were then harvested and centrifuged (450 x g/5 minutes at 4˚C). The supernatant was removed and pellet was resuspended with 300 μL of 0.075M KCl solution for 5 minutes. The cells were then fixed with Carnoy's solution consisting of acetic acid (Sigma-Aldrich, St. Louis, MO, USA) and methanol (HmbG Chemicals, Hamburg, Germany) prepared at the ratio of 1:3 and spreaded on glass slides which were placed on a slide warmer. The slides were dried overnight and stained with 30 μL of 20 μg/mL acridine orange (AO; Sigma-Aldrich, St. Louis, MO, USA) prior to fluorescence microscopic observation. The number of viable mononucleated, binucleated and multinucleated cells per 500 cells were scored to derive Nuclear Division Index (NDI) and frequency of micronucleus in 1,000 binucleated cells was measured. [dx.doi.org/10.17504/protocols.io.bdqhi5t6]

Mitochondrial membrane potential (ΔCm), mitochondrial mass and ROS assessment
The treated cells (5 x 10 4 cells/mL) were collected by centrifugation (450 x g/5 minutes at 4˚C). The supernatant was discarded and pellet was resuspended with 1 mL fresh prewarmed FBSfree MEM with addition of 1 μL of 50 μM tetramethylrhodamine ethyl ester (TMRE; Thermo Fisher Scientific, Waltham, MA, USA), 5 mM nonyl acridine orange (NAO; Sigma-Aldrich, St. Louis, MO, USA), 10 mM hydroethidine (HE; Thermo Fisher Scientific, Waltham, MA, USA) or 10 mM 2',7'-dichlorodihydrofluorescein diacetate (DCFH-DA; Thermo Fisher Scientific, Waltham, MA, USA). The cells stained with TMRE or NAO were incubated for 15 minutes at 37˚C whereas cells stained with HE and DCFH-DA were incubated for 30 minutes at 37˚C in the dark. After incubation, the cells were centrifuged (450 x g/5 minutes at 4˚C) and pellet was washed with 1 mL chilled PBS solution. The supernatant was discarded and 500 μL of chilled PBS was used to resuspend the pellets. The stained cell suspension was transferred to flow tubes and analyzed using FACSCanto II Flow Cytometer (BD Biosciences, San Jose, CA, USA

Intracellular Nitric Oxide (NO) assessment using BD Pharmingen™ orange nitric oxide (NO) probe staining
The seeded cells (5 x 10 4 cells/mL) were pre-stained with 1 μL Orange NO probe (BD Biosciences, San Jose, CA, USA) per 500 μL cell suspension for 30 minutes. The cells were then pretreated with BK3C231 at 50 μM and positive control, resveratrol (Sigma-Aldrich, St. Louis, MO, USA) at 25 μM for 2 hours, 4 hours, 6 hours, 12 hours and 24 hours prior to 4NQO treatment at 1 μM for 1 hour. The stained and treated cells were centrifuged (450 x g/5 minutes at 4˚C) and pellet was washed with 1 mL chilled PBS solution. The supernatant was discarded and 500 μL of chilled PBS was used to resuspend the pellets. The stained cell suspension was transferred to flow tubes and analyzed using FACSCanto II Flow Cytometer (BD Biosciences, San Jose, CA, USA). [dx.doi.org/10.17504/protocols.io.bdqni5ve]

Extracellular Nitric Oxide (NO) assessment using griess reagent
CCD-18Co cells were seeded in culture dish (60 x 15 mm) at the concentration of 5 x 10 4 cells/ mL. The seeded cells were incubated under 5% CO 2 at 37˚C for 24 hours. The cells were then pretreated with BK3C231 at 50 μM and positive control, resveratrol at 25 μM for 2 hours, 4 hours, 6 hours, 12 hours and 24 hours prior to 4NQO treatment at 1 μM for 1 hour. Subsequently, 100 μL of culture medium from each sample was collected and mixed with the same volume of Griess reagent (1% sulfanilamide in 5% phosphoric acid and 0.1% N-(1-naphthyl) ethylenediamine (NNED) hydrochloride in distilled water, Merck Milipore, Burlington, MA, USA) in 96-well microplate. Absorbance of the mixture in each well was determined at 570 nm using iMark TM microplate reader (Bio-Rad Laboratories, Hercules, CA, USA). The concentration of nitrite accumulated in the culture was determined in comparison to the sodium nitrite standards. [dx.doi.org/10.17504/protocols.io.bdqpi5vn]

Glutathione (GSH) assessment using ellman's reagent
The treated cells (5 x 10 4 cells/mL) were detached, collected and centrifuged (450 x g/5 minutes at 4˚C). The supernatant was discarded and pellet was resuspended in 100 μL ice-cold lysis buffer (50 mM K 2 HPO 4 , 1 mM EDTA, pH 6.5 and 0.1% v/v Triton X-100, Sigma-Aldrich, St. Louis, MO, USA). The cells were incubated on ice for 15 minutes with gentle tapping from time to time. The crude lysates were cleared by centrifugation (10000 x g/15 minutes at 4˚C). At this point, the lysates were used immediately or stored at -80˚C for a day or two. Then, 50 μl of lysates and GSH standards (two fold dilution from 1.

Statistical analysis
The data are expressed as the mean ± standard error of mean (S.E.M.) from at least three independent experiments. The statistical significance was evaluated using one-way ANOVA with the Tukey post hoc test used to assess the significance of differences between multiple treatment groups. Differences were considered statistically significant with a probability level of p<0.05.

Cytotoxic assessment of BK3C231 and 4NQO
The non-cytotoxic concentrations of BK3C231 and 4NQO were determined using MTT cytotoxicity assay. BK3C231 did not show evidence of cytotoxicity up to 50 μM treatment, however an IC 50 value of 99 μM was observed (Fig 2A). Therefore a series of BK3C231 concentrations ranging from 6.25 μM till 50 μM was used for subsequent experiments. On the other hand, 4NQO treatment exerted no cytotoxicity at 1 hour. However, reduction in cell viability was significant with IC 50 values observed starting from 2 hours till 24 hours (Fig 2B). Hence, 4NQO concentration at 1 μM was selected to induce genotoxicity and mitochondrial toxicity in subsequent experiments as used by previous studies as well [37,38]. Interestingly, in comparison to 4NQO-treated cells whereby cell viability greatly reduced especially at higher concentrations, BK3C231 was able to suppress 4NQO-induced cytotoxicity by increasing cell viability up to 8-fold with no IC 50 value observed (Fig 2C).

BK3C231 protection against 4NQO-induced DNA microlesions
Significant DNA damage as indicated by the comet tail which represents DNA strand breaks can be observed in cells treated only with 4NQO. Untreated control cells and BK3C231-treated cells showed intact round nuclear DNA and no DNA strand break was observed at all treated concentrations (Fig 3A). There was also a decrease in comet tail in cells pretreated with BK3C231 when compared with that of cells treated only with 4NQO ( Fig 3B). This was further confirmed by quantification of tail moments obtained from comet scoring. Tail moment increased significantly up to 48-fold in 4NQO-treated cells at 28.79 ± 1.02 (p<0.05) over control and BK3C231-treated cells ranging from 0.59 ± 0.11 to 0.68 ± 0.06 (Fig 3C). On the other hand, BK3C231 pretreatment showed a 0.8-fold decrease of 4NQO-induced DNA strand breaks in a concentration-dependent manner, significantly at 50 μM with a tail moment value of 7.21 ± 0.34 (p<0.05) (Fig 3D).

Inhibition of 4NQO-induced DNA macrolesions by BK3C231
The protective role of BK3C231 against 4NQO-induced micronucleus formation was assessed using CBMN assay (Fig 4A). In untreated control cells, a micronucleus frequency level as low as 0.23 ± 0.03 was observed. Cells treated with 4NQO significantly demonstrated up to 93-fold increase in frequency of micronucleus in binucleated cells at 21.56 ± 1.36 (p<0.05). However, pretreatment of cells with BK3C231 was shown to cause a maximum of 0.8 fold decrease of 4NQO-induced micronucleus formation in a concentration-dependent manner, significantly at 25 μM with a frequency level of 6.58 ± 0.52 and 50 μM with a frequency level of 3.80 ± 0.47 (p<0.05) (Fig 4B). In addition, the NDI values measured in control, 4NQO-treated cells and BK3C231-treated cells were 1.78 ± 0.01, 1.68 ± 0.03 and 1.79 ± 0.01 respectively. As for cells pretreated with BK3C231 prior to 4NQO induction, the average NDI value measured was 1.72 ± 0.01 (Table 1). All NDI values obtained in this assay indicated normal cell proliferation [39].

Cytoprotective role of BK3C231 in 4NQO-induced mitochondrial damages
The cytoprotective role of BK3C231 was further investigated at the mitochondrial level through flow cytometric assessment of ΔCm loss using TMRE staining. Significant loss of ΔCm (p<0.05) as indicated by a 1.2-fold increase of TMRE-negative cells from 15.63 ± 1.09% in control cells to 34.77 ± 1.29% in 4NQO-treated cells was observed. However, BK3C231 pretreatment was shown to reduce the amount of TMRE-negative cells significantly to 22.13 ± 2.51% (p<0.05) at 50 μM, thereby protecting the cells against 4NQO-induced ΔCm loss (Fig 5A). Positive control, resveratrol also decreased the amount of TMRE-negative cells significantly to 24.07 ± 0.22% (p<0.05) at 25 μM.
In a bid to further establish the protective role of BK3C231 in mitochondria, cardiolipin level was assessed through flow cytometric analysis using NAO staining. Our study demonstrated significant cardiolipin loss (p<0.05) as indicated by a 2.8-fold increase of NAO-negative cells from 5.07 ± 0.52% in control cells to 19.33 ± 0.94% in 4NQO-treated cells. However, BK3C231 pretreatment was shown to induce up to 0.4-fold decrease of 4NQO-induced cardiolipin loss in a concentration-dependent manner (Fig 5B). Positive control, resveratrol also reduced the amount of TMRE-negative cells significantly to 12.83 ± 0.13% (p<0.05) at 25 μM.

4NQO-induced DNA and mitochondrial damages independent of ROS production
Flow cytometric assessment of intracellular ROS namely superoxide and hydrogen peroxide levels using HE and DCFH-DA staining was performed to determine the role of ROS in 4NQO-induced DNA and mitochondrial damages. Interestingly, as shown in Fig 6A and 6B, there were no inductions of superoxide and hydrogen peroxide levels in 4NQO-treated cells as compared to control cells. Hydroquinone (HQ), which was used as positive control, had significantly increased ROS level in CCD-18Co cells (p<0.05). This suggested that ROS was not involved in DNA and mitochondrial damages caused by 4NQO, also leading to indication that neither cytoplasm nor mitochondria played a role in ROS production in 4NQO-treated cells. In addition, BK3C231 demonstrated potential as antioxidant by inhibiting HQ-induced ROS production.

Inhibition of 4NQO-induced nitrosative stress by BK3C231
Intracellular nitric oxide (NO) level was assessed using BD Pharmingen™ Orange NO Probe staining whereas extracellular NO level was assessed using Griess assay to determine the involvement of RNS in 4NQO-induced DNA and mitochondrial damages. Our study demonstrated a significant 0.98-fold increase of intracellular NO level, 15.7 ± 0.19% and 2.4-fold increase of extracellular NO level, 5.15 ± 0.17 μM (p<0.05) in 4NQO-treated cells over control cells, at 7.63 ± 0.19% and 1.51 ± 0.26 μM respectively, thereby demonstrating the involvement of NO in 4NQO-induced DNA and mitochondrial damages. Moreover, both BK3C231 and positive control, resveratrol significantly inhibited 4NQO-induced NO production from as early as 2 hours up till 24 hours of pretreatment (p<0.05) and were able to restore NO level back to basal level at 24 hours of pretreatment (Fig 7A and 7B).
In addition to that, antioxidant GSH level was assessed using Ellman's reagent. 4NQOtreated cells showed a reduced GSH level at 194.70 ± 23.83 nmol/mg as compared to untreated control cells at 245.96 ± 12.44 nmol/mg (Fig 7C). Overall, the simultaneous increase in NO level and decrease in GSH level by 4NQO further confirmed the involvement of nitrosative stress in 4NQO-induced DNA and mitochondrial damages. However, no induction of GSH level was observed in cells pretreated with BK3C231 for 2 hours, 4 hours, 6 hours and 12 hours. BK3C231 was only able to significantly increase GSH level, 313.97 ± 27.83 nmol/mg (p<0.05) at 24 hours of pretreatment as compared to that of 4NQO-treated cells (Fig 7C). This

PLOS ONE
suggested that BK3C231 inhibited 4NQO-induced nitrosative stress through early reduction of NO production and late induction of GSH level in CCD-18Co cells.

Discussion
Epidemiological studies have shown that consumption of fruits particularly rich in stilbenes led to a reduced risk of colorectal cancer, which is one of the most commonly diagnosed cancers worldwide [40,41]. Furthermore, cytoprotection of DNA and mitochondrial function

PLOS ONE
limits the occurrence of cancer. Since DNA is the repository of hereditary material and genetic information in every living cell, the maintenance of its stability is pivotal as unrepaired DNA damage caused by diverse assaults from the environment, nutrition and natural cellular processes lead to cancer [42,43]. As for mitochondria, impairments and alterations of mitochondrial structure and functions, including morphology and redox potential, are associated with cancer transformation and have been frequently reported in human cancers [44][45][46]. In agreement with this, our study showed that BK3C231 was able to inhibit 4NQO-induced cytotoxicity as well as protect against DNA damage and mitochondrial dysfunction in the normal human colon fibroblast CCD-18Co cell line.
Firstly, we sought to understand the carcinogenic actions of 4NQO. Studies have demonstrated that 4NQO elicited carcinogenicity through its proximate carcinogenic metabolite namely 4-hydroxyaminoquinoline 1-oxide (4HAQO) produced by the enzymatic four-electron reduction of 4NQO's nitro group [47,48]. Being a potent chemical carcinogen and as a UV-mimetic agent, 4NQO is often used as positive control in various genotoxicity studies due to its well characterized metabolic processes [49]. As a result of a study by Brüsehafer et al. [50], it was reported that 4NQO predominantly induces mutagenicity more than clastogenicity and that the latter depends on cell types, our study has proved that 4NQO is a significant cause of DNA damage via DNA strand breaks and chromosomal damage via micronucleus formation. Our study was also in agreement with previous studies which demonstrated that 4NQO caused damage to mitochondrial membrane as characterized by loss of mitrochondrial membrane potential (ΔCm) and cardiolipin [51].
As 4HAQO's carcinogenic effect is mainly based on DNA adduct formation, our study investigated 4NQO's other carcinogenic mechanism of action through generation of ROS and RNS and its involvement in the cytoprotective role of BK3C231 [52][53][54]. Interestingly, our study which revealed no superoxide and hydrogen peroxide production by 4NQO at 1 μM for 1 hour in CCD-18Co cells contradicts the study by Arima et al. [37] who reported ROS formation in human primary skin fibroblast by 4NQO using the same treatment concentration and timepoint. The discrepancy is likely due to the difference in the origin of fibroblast used. Hence, our study is the first to elucidate such findings on 4NQO mechanism which has never been shown in other studies thus far.
In addition, our study demonstrated an increased NO level and a depleted GSH level by 4NQO. This is possibly due to formation of 4NQO-GSH conjugates leading to generation of nitrite, a stable end product of NO, which inactivated γ-glutamylcysteine synthase and therefore suppressed intracellular synthesis of GSH [37,[54][55][56]. Our data was also in agreement with previous studies that NO could be the main culprit in 4NQO-induced DNA and mitochondrial damages in CCD-18Co cells as NO has been demonstrated to induce genotoxicity and damage to mitochondria via multiple mechanisms directly or indirectly [57,58]. Moreover, the concurrent increase in NO level and decrease in GSH level postulates the occurrence of nitrosative stress which may contribute to 4NQO-induced DNA and mitochondrial damage [59,60].
More importantly, BK3C231 was shown in our study to protect against 4NQO-induced DNA and mitochondrial damages by decreasing DNA strand breaks and micronucleus formation as well as reducing loss of mitochondrial membrane potential (ΔCm) and cardiolipin (Fig  8). Our study further revealed that BK3C231 exerted these cytoprotective effects in CCD-18Co cells by suppressing 4NQO-induced nitrosative stress through reduction in NO level and late upregulation of GSH level. Furthermore, inhibition of HQ-induced ROS production by BK3C231 as demonstrated in this study corroborates the potential of BK3C231 as an antioxidant. The role of stilbene derivatives as potential antioxidants has been demonstrated in several studies involving for example resveratrol, a well-known stilbenoid, which attenuates nitrosative stress in the small intestine of rats [61]. Piceatannol and isorhapontigenin, which are natural occurring stilbenes, have also been demonstrated to scavenge NO and nitrogen dioxide (NO 2 ) radicals as well as increasing GSH/GSSG ratio [62,63]. Therefore, we have included resveratrol, which is under the same stilbene family as BK3C231, as the positive control in this study which supported the BK3C231 findings and were consistent with previous studies which had reported the protective effects of resveratrol on mitochondrial function as well as its role in NO suppression and GSH induction [64][65][66][67][68].
Nitrosative stress is one of the most critical factors in multi-stage carcinogenesis with NO playing an important role in tumour biology and overproduction of NO can promote tumour growth [69]. Kee et al. [36] has reported the chemopreventive activity of BK3C231 involving upregulation of the detoxifying enzyme NQO1 due to the presence of methoxy and furan carboxamide groups. Therefore, it is possible that the presence and substitution pattern in relation to the methoxy group enables BK3C231 to act as an NO scavenger.
The Keap1-Nrf2 signaling pathway may also be involved in the depletion of NO level by BK3C231. The upstream Keap1-Nrf2 signaling pathway, which is a major regulator of phase II detoxification and cytoprotective genes, is postulated to be involved through upregulation of detoxifying enzymes which may lead to NO suppression [70]. Stilbene derivatives particularly resveratrol play a significant role in the activation of Nrf2-related gene transcription which induces expression of cytoprotective enzymes such as NQO1, glutathione S-transferase (GST), glutamate-cysteine ligase catalytic subunit (GCLC) and heme oxygenase-1 (HO-1) thus leading to protection against cancer [71]. Moreover, the increase in GSH as observed in this study can be attributed to the redox-sensitive Nrf2 activation [72,73].
Cytoprotection plays an important role in chemoprevention by suppressing the initiation stage of carcinogenesis. Cytoprotective mechanisms in response to DNA and mitochondrial damages by key causes of malignant transformation such as electrophile, oxidative stress and nitrosative stress represent a target for chemopreventive agent [74]. Our study has demonstrated the cytoprotective role of BK3C231, therefore it warrants further investigation of the chemopreventive role of BK3C231 in the Keap1-Nrf2 pathway which serves as both an important target thus provide significant protection of normal cells against carcinogenesis.

Conclusion
In conclusion, this study has provided a good insight into 4NQO-induced carcinogenicity in CCD-18Co cells. The demonstration of BK3C231 as a cytoprotective agent also served as a stepping stone for further elucidation of its chemopreventive potential against both genetic and epigenetic bases of cancer development. Based on our current findings in this study, we aim to reduce the gap between understanding molecular mechanism occurring in cancer carcinogenesis and instigating successful adoption of chemoprevention using BK3C231.