Oxidative Stress-Dependent Synergistic Antiproliferation, Apoptosis, and DNA Damage of Ultraviolet-C and Coral-Derived Sinularin Combined Treatment for Oral Cancer Cells

Simple Summary Combined treatments with low side effects enhance anticancer applications. This study focusses on validating the potential synergistic antiproliferation of the combined treatment of ultraviolet-C and the coral-derived compound sinularin (UVC/sinularin) in oral cancer cells. This study confirms that UVC/sinularin synergistically and selectively inhibits oral cancer cell proliferation with low cytotoxicity on normal oral cells. The mechanisms involve the enhanced cellular and mitochondrial oxidative stress that cause apoptosis, DNA damage, and mitochondrial dysfunction in oral cancer cells. Abstract Combined treatment is increasingly used to improve cancer therapy. Non-ionizing radiation ultraviolet-C (UVC) and sinularin, a coral Sinularia flexibilis-derived cembranolide, were separately reported to provide an antiproliferation function to some kinds of cancer cells. However, an antiproliferation function using the combined treatment of UVC/sinularin has not been investigated as yet. This study aimed to examine the combined antiproliferation function and explore the combination of UVC/sinularin in oral cancer cells compared to normal oral cells. Regarding cell viability, UVC/sinularin displays the synergistic and selective killing of two oral cancer cell lines, but remains non-effective for normal oral cell lines compared to treatments in terms of MTS and ATP assays. In tests using the flow cytometry, luminescence, and Western blotting methods, UVC/sinularin-treated oral cancer cells exhibited higher reactive oxygen species production, mitochondrial superoxide generation, mitochondrial membrane potential destruction, annexin V, pan-caspase, caspase 3/7, and cleaved-poly (ADP-ribose) polymerase expressions than that in normal oral cells. Accordingly, oxidative stress and apoptosis are highly induced in a combined UVC/sinularin treatment. Moreover, UVC/sinularin treatment provides higher G2/M arrest and γH2AX/8-hydroxyl-2′deoxyguanosine-detected DNA damages in oral cancer cells than in the separate treatments. A pretreatment can revert all of these changes of UVC/sinularin treatment with the antioxidant N-acetylcysteine. Taken together, UVC/sinularin acting upon oral cancer cells exhibits a synergistic and selective antiproliferation ability involving oxidative stress-dependent apoptosis and cellular DNA damage with low toxic side effects on normal oral cells.


Introduction
Radiation and chemotherapy are the routine remedies for oral cancer therapy [1]. The problems of radio-and chemo-resistance, and their associated side effects, may weaken their anticancer applications in oral cancer therapies [2][3][4]. Combined treatment with natural products provides an improved strategy to overcome radio-and chemo-resistance or side effects in cancer therapy [5,6]. For the complex etiology and targeting of cancer, combinations of anticancer drugs, natural products and radiation are increasingly used to improve the therapeutic efficiency for oral cancer [7,8].
Although X-rays are commonly used in radiotherapy, anticancer studies using nonionizing radiation are increasing. For example, ultraviolet-C (UVC) exhibited an antiproliferation effect to pancreatic [9] and colon [10] cancer cells. A UVC-generator is cheaper and more portable than an X-ray machine [11]. UVC irradiation may be effortlessly operated in the clinic, rather than the medical center. UVC exposure shows translational potential to inhibit the proliferation of glioblastoma multiforme cancer cells [12]. The combined treatments of clinical drugs [13,14], chemical agents [15], or natural products [16,17] with UVC irradiation may improve the antiproliferation of several types of cancer cells. Moreover, UVC irradiation inhibits tumor growth in an animal model without apparent side effects [11,18]. Therefore, the discovery of the drug amendments applied as UVC radiosensitizers may effectively improve oral cancer therapy.
The selective killing effect of sinularin was firstly reported in our previous works on oral [24] and breast [25] cancer cells, and it was subsequently mentioned in renal cancer cells [26]. These studies reported that sinularin showed low cytotoxicity to non-malignant cells of oral (HGF-1), breast (M10), and renal (HRCEpic) origins [24][25][26], suggesting that sinularin shows low side effects on normal cells.
Currently, the effects of a combined treatment using UVC and sinularin (UVC/sinularin) in oral cancer therapy have not been reported. We hypothesized that UVC/sinularin exhibited the synergistic and selective killing of oral cancer cells. In order to validate the hypothesis, the survival, cell cycle distribution, apoptotic change, cellular oxidative stress imbalance, and induced DNA damage following the separate and combined treatments (UVC and/or sinularin) were compared between cancer and normal types of oral cells. For the above endpoints, we employed ATP analysis, flow cytometry, and Western blotting. The therapeutic goal of this study is to examine the potential anticancer application for UVC/sinularin combined treatment in the example of gingival and tongue cancer cells.

Cell Lines
The human gingival and tongue cancer (Ca9-22 and CAL 27) cell lines and normal (HGF-1) oral cell lines were from JCRB Cell Bank and ATCC, located in Osaka, Japan, and Manassas, VA, USA, respectively. The cell culture materials were purchased from Gibco (Grand Island, NY, USA). All of the cells were cultured with a medium supplemented with 10% fetal bovine serum and antibiotics in an incubator with 5% CO 2 and humidified 37 • C atmospheres. The media for the oral cancer and normal oral cells were 3:2 and 4:1 mixtures of Dulbecco's Modified Eagle Medium (DMEM) and F12 [27].

UVC Irradiation, Sinularin Preparation, Chemicals and Cell Treatments
Following the removal of the culture medium, the human oral cells were placed in a laminar flow and irradiated with the built-in UVC (254 nm) lamp at the dose rate of 2 J/m 2 /sec [16] for the required periods, i.e., 5 sec for 10 J/m 2 (for CAL 27 cells) and 6 s for 12 J/m 2 (for Ca9-22 and HGF-1 cells). The non-UVC irradiated control cells followed the same procedure except for the UVC irradiation.
The sinularin was prepared by the standard purification processes using cultured soft coral S. manaarensis, as previously mentioned [20], and dissolved in DMSO. The 1 H-NMR, 13 C-NMR spectra and the HPLC profile are provided in Figures S1-S3 as evidence for the purity of the sinularin.

Survival Analyses by MTS and ATP Assays
The cell survival was determined by both MTS (Promega Corporation, Madison, WI, USA) [33] and ATP assays (PerkinElmer Life Sciences, Boston, MA, USA) [34]. Briefly, the MTS substrate was mixed with the medium to a final concentration of 58.1 µg/mL at 37 • C for 1 h in the dark for the MTS assay. Subsequently, it was detected by the Epoch™ Microplate Spectrophotometer (Bio Tek; Winooski, USA). For the ATP assay, 100 µL of the cell lysate was mixed with 12.5 µL of the reaction substrate (D-luciferin and luciferase) at room temperature for 5 min in the dark. Subsequently, it was detected by a CentroLIApc LB 962 Microplate Luminometer (Berthold Technologies GmbH & Co., Bad Wildbad, Germany). The synergy between the UVC/sinularin was calculated as previously described [35], i.e., the synergy (α) = the survival fraction for the single treatment (UVC) × the survival fraction for the single treatment (sinularin)/the survival fraction for the combined treatment (UVC/sinularin). The additive, synergistic, or antagonistic effects results were: α = 1, > 1, and < 1, respectively.

Cell Cycle Detection
For the cell cycle assay, Ca9-22 and CAL 27 cells were harvested with trypsin, fixed with 70% ethanol, washed with PBS, and stained with Biotium DNA dye (Hayward, CA, USA), i.e., 7-aminoactinmycin D (7AAD) at the final concentration of 10 µg/mL (37 • C, 30 min) [36]. Subsequently, the cell cycle phase distribution for the stained cells was
For the apoptosis signaling, the apoptosis-related activities were examined by Caspase-Glo ® 3/7 commercial product (Promega; Madison, WI, USA) [15] and Western blotting. Briefly, a luminogenic caspase-3/7 tetrapeptide substrate (DEVD) and buffer were equally mixed and incubated at 37 • C for 30 min in the dark. Subsequently, this mixture was applied to a 96-well plate for the measuring of the caspase 3/7 activity of the drug-treated cells. For the Western blotting, Cell Signalling Technology cleaved poly (ADP-ribose) polymerase (c-PARP) antibody (Danvers, MA, USA) (1:1000 dilution) was used to detect apoptosis, and Sigma-Aldrich β-actin was included for the detection of the loading control [38]. Figure S4: Original western blotting figures.

ROS Flow Cytometry
The ROS specific dye 2 ,7 -dichlorodihydrofluorescein diacetate was obtained from Sigma-Aldrich. Following the cell harvesting and PBS washing, the cell treatment condition for this ROS dye was applied at 10 µM (37 • C, 30 min) in the dark, was reacted in proportion to cellular ROS amount, and became a fluorescence chemical reaction for the flow cytometry and FlowJo analysis [28]. The cells (Ca9-22, CAL 27, and HGF-1) were resuspended in PBS before the flow cytometry.

Mitochondrial Superoxide (MitoSOX) Flow Cytometry
Following the cell harvesting and PBS washing, the cells were incubated with MitoSOX Red (Invitrogen, Eugene, OR, USA) (50 nM, 37 • C, 30 min) in the dark. MitoSOX can react in proportion to the MitoSOX amount, and can generate fluorescence for flow cytometry and FlowJo analysis [25]. The cells (Ca9-22, CAL 27, and HGF-1) were resuspended in PBS before the flow cytometry.

Mitochondrial Membrane Potential (MitoMP) Flow Cytometry
Following the cell harvesting and PBS washing, the cells were incubated with DiOC 2 (3) (Invitrogen) (5 nM, 37 • C, 30 min) in the dark. DiOC 2 (3) can primarily accumulate in mitochondria in proportion to MitoMP, and it generates the fluorescence detected by flow cytometry and FlowJo analysis [39]. The cells (Ca9-22, CAL 27, and HGF-1) were resuspended in PBS before the flow cytometry.

γH2AX Flow Cytometry
Following the cell harvesting and PBS washing, the DNA double-strand break damage was determined by antibody-associated flow cytometry [48]. Briefly, the cells (Ca9-22 and CAL 27) were fixed and subsequently processed with Santa Cruz Biotechnology antibody for γH2AX (Santa Cruz, CA, USA) (1:500 dilution). Its matched Cell Signaling Technology antibody was marked with Alexa Fluor 488 (1:10,000 dilution) and 7AAD staining (5 µg/mL) for the flow cytometry and FlowJo analysis. The cells were resuspended in PBS before the flow cytometry.

Statistics
Except for the Western blotting results (Student's t-test), the one-way ANOVA evaluation and Tukey HSD post hoc examination (JMP software) were used for the statistical analysis for multi-comparisons. The treatments marked without overlapping characters differ significantly. The example to illustrate the meaning of the significance is shown at the end of the figure legend in Figure 1. In the reference of the housekeeping gene GAPDH [43,44] (F: CCTCAAC-TACATGGTTTACATGTTCC and R: CAAATGAGCCCCAGCCTTCT), their relative mRNA expressions (log2) were measured by the 2 −ΔΔCt method [47].

γH2AX Flow Cytometry
Following the cell harvesting and PBS washing, the DNA double-strand break damage was determined by antibody-associated flow cytometry [48]. Briefly, the cells (Ca9-22 and CAL 27) were fixed and subsequently processed with Santa Cruz Biotechnology antibody for γH2AX (Santa Cruz, CA, USA) (1:500 dilution). Its matched Cell Signaling Technology antibody was marked with Alexa Fluor 488 (1:10,000 dilution) and 7AAD staining (5 μg/mL) for the flow cytometry and FlowJo analysis. The cells were resuspended in PBS before the flow cytometry.

Statistics
Except for the Western blotting results (Student's t-test), the one-way ANOVA evaluation and Tukey HSD post hoc examination (JMP software) were used for the statistical analysis for multi-comparisons. The treatments marked without overlapping characters differ significantly. The example to illustrate the meaning of the significance is shown at the end of the figure legend in Figure 1. repeated characters (a to f) differ significantly (p < 0.05). In the example of Ca9-22 cells ( Figure 1B), the viabilities between the control (black) and NAC (gray) for the treatments of UVC (e vs. c), sinularin (d vs. a), and UVC/sinularin (f vs. c) showing different characters indicate significant differences. In comparison, the viabilities between the control and NAC for the control treatments, showing the same character (b), indicate non-significant differences. Similarly, the control (black) viabilities for the treatments of UVC (e vs. b), sinularin (d vs. b), and UVC/sinularin (f vs. b) showing different characters indicate significant differences. The data were plotted as the mean ± SD (n = 3).
For the 48 h ATP assay ( Figure 1B), UVC/sinularin-treated oral cancer cells showed lower viability for 16.1% and 15.9% than UVC or sinularin alone in oral cancer Ca9-22 (56.0% or 60.1%) and CAL 27 (60.7% or 68.7%) cells. However, UVC and/or sinularin for normal HGF-1 cells only showed low cytotoxicity (around 80% viability). Moreover, the oxidative stress involvement in modulating the cell viability was estimated by the presence of antioxidant NAC. NAC rescues the antiproliferation of sinularin and UVC/sinularin treatments in both oral cancer cell lines, although NAC shows different responses to UVC between oral cancer cells. NAC also rescues the mild antiproliferation to normal oral cells.

UVC/Sinularin Combined Treatment of Oral Cancer Cells Highly Induces Cell Cycle Disturbance
The oral cell cycle progression pattern following the four types of treatments-the control, UVC, sinularin, and UVC/sinularin-is displayed for Ca9-22 and CAL 27 cells ( Figure 2A). For these oral cancer cells, the UVC/sinularin treatment exhibited more sub-G1 (%) and G2/M (%), and it shows lower G1 (%) than the separate treatments (UVC or sinularin) as well as the control ( Figure 2B).
Moreover, the involvement of an oxidative stress-modulating effect on the cell cycle disturbance on Ca9-22 and CAL 27 cells was estimated by the presence of antioxidant NAC ( Figure 2A). NAC suppresses the subG1 and G2/M inductions and G1 reduction to the UVC/sinularin treatments in both oral cancer cell lines ( Figure 2B).

UVC/Sinularin Combined Treatment of Oral Cancer Cells Highly Induces Annexin V-, Caspase-and Western Blotting-Detected Apoptosis
The possibility of apoptosis for the UVC/sinularin treatments in Ca9-22 and CAL 27 cells showing higher subG1 ( Figure 2) than others was further tested by annexin V/7AAD, pan-caspase, Cas 3/7, and Western blotting analyses, as follows. In the annexin V/7AAD assay, after 48 h of drug treatment ( Figure 3A), UVC/sinularin in oral cancer Ca9-22 and CAL 27 cells demonstrated higher annexin V (+) (%) than the separate treatments and control ( Figure 3B). However, the UVC and/or sinularin treatments displayed only mild annexin V (+) (%) to normal HGF-1 cells. Cancers 2021, 13, x 7 of 20

UVC/Sinularin Combined Treatment of Oral Cancer Cells Highly Induces Annexin V-, Caspase-and Western Blotting-Detected Apoptosis
The possibility of apoptosis for the UVC/sinularin treatments in Ca9-22 and CAL 27 cells showing higher subG1 ( Figure 2) than others was further tested by annexin V/7AAD, pan-caspase, Cas 3/7, and Western blotting analyses, as follows. In the annexin V/7AAD assay, after 48 h of drug treatment ( Figure 3A), UVC/sinularin in oral cancer Ca9-22 and CAL 27 cells demonstrated higher annexin V (+) (%) than the separate treatments and control ( Figure 3B). However, the UVC and/or sinularin treatments displayed only mild annexin V (+) (%) to normal HGF-1 cells. Moreover, the oxidative stress involvement in modulating the apoptosis in terms of annexin V expression was estimated by the presence of NAC ( Figure 3). NAC rescues the annexin V-detected apoptosis of sinularin and UVC/sinularin treatments in both oral cancer cell lines. NAC also rescues the mild annexin V-detected apoptosis to normal oral cells.
For the Cas 3/7 assay, UVC/sinularin for oral cancer Ca9-22 and CAL 27 cells also activated more Cas 3/7 activity (~4 folds) than UVC, sinularin, and the control. Still, the Cas 3/7 activity for the separate and combined treatments was similar in normal HGF-1 cells (1 fold) ( Figure 4C). In order to further validate the roles of oxidative stress and apoptosis, their acting inhibitors, such as NAC and ZVAD, were pretreated. For oral cancer cells, both NAC and ZVAD preincubations suppressed the pan-caspase activity for UVC/sinularin to a similar degree, while ZVAD suppressed more Cas 3/7 activity for UVC/sinularin than NAC preincubation.
For the Western blotting ( Figure 4D Moreover, the oxidative stress involvement in modulating the apoptosis in terms of annexin V expression was estimated by the presence of NAC (Figure 3). NAC rescues the annexin V-detected apoptosis of sinularin and UVC/sinularin treatments in both oral cancer cell lines. NAC also rescues the mild annexin V-detected apoptosis to normal oral cells.

UVC/Sinularin Combined Treatment of Oral Cancer Cells Highly Induces ROS Production
In the ROS assay ( Figure 5A), UVC/sinularin exhibited higher ROS (+) (%) than UVC, sinularin, and the control in two oral cancer cell lines (Ca9-22 and CAL 27) ( Figure 5B). However, the ROS changes following UVC and/or sinularin for normal HGF-1 cells maintained basal levels. Moreover, the oxidative stress involvement to modulate the ROS generation was estimated by the presence of NAC ( Figure 5). NAC reverted the ROS induction of sinularin and UVC/sinularin treatments in both oral cancer cell lines.

UVC/Sinularin Combined Treatment of Oral Cancer Cells Highly Induces MitoSOX Generation
In the MitoSOX assay ( Figure 6A), UVC/sinularin exhibited higher MitoSOX (+) (%) than UVC, sinularin and the control in two oral cancer cell lines (Ca9-22 and CAL 27) ( Figure 6B). However, the MitoSOX changes following UVC and/or sinularin for normal oral HGF-1 cells maintained basal levels. Moreover, the oxidative stress involvement to modulate the MitoSOX change was estimated by the presence of NAC ( Figure 6). NAC reverted the MitoSOX induction of sinularin and UVC/sinularin treatments in both oral cancer cell lines.

UVC/Sinularin Combined Treatment of Oral Cancer Cells Highly Induces MitoSOX Generation
In the MitoSOX assay ( Figure 6A), UVC/sinularin exhibited higher MitoSOX (+) (%) than UVC, sinularin and the control in two oral cancer cell lines (Ca9-22 and CAL 27) ( Figure 6B). However, the MitoSOX changes following UVC and/or sinularin for normal oral HGF-1 cells maintained basal levels. Moreover, the oxidative stress involvement to modulate the MitoSOX change was estimated by the presence of NAC ( Figure 6). NAC reverted the MitoSOX induction of sinularin and UVC/sinularin treatments in both oral cancer cell lines. . For the multi-comparisons, the treatments marked without repeated characters (a to e) differ significantly (p < 0.05). The data were plotted as the mean ± SD (n = 3).

Oral Cancer Cells Following UVC/Sinularin Combined Treatment Highly Induce MitoMP Destruction
In the MitoMP assay ( Figure 7A), UVC/sinularin exhibited higher MitoMP (-) (%) than UVC, sinularin, and the control in two oral cancer cell lines (Ca9-22 and CAL 27) ( Figure 7B). However, the MitoMP changes following UVC and/or sinularin for normal HGF-1 cells maintained basal levels. Moreover, the oxidative stress involvement to modulate the MitoMP change was estimated by the presence of NAC (Figure 7). NAC reverted the MitoMP destruction of sinularin and UVC/sinularin treatments in both oral cancer cell lines.

Oral Cancer Cells Following UVC/Sinularin Combined Treatment Highly Induce MitoMP Destruction
In the MitoMP assay ( Figure 7A), UVC/sinularin exhibited higher MitoMP (-) (%) than UVC, sinularin, and the control in two oral cancer cell lines (Ca9-22 and CAL 27) ( Figure 7B). However, the MitoMP changes following UVC and/or sinularin for normal HGF-1 cells maintained basal levels. Moreover, the oxidative stress involvement to modulate the MitoMP change was estimated by the presence of NAC (Figure 7). NAC reverted the MitoMP destruction of sinularin and UVC/sinularin treatments in both oral cancer cell lines.

UVC/Sinularin Combined Treatment of Oral Cancer Cells Downregulates Antioxidant Gene Expressions
Because oxidative stresses were confirmed (Figures 5-7), the source of the oxidative stress needs to be investigated. Cellular antioxidant signaling may regulate oxidative stress [49]. Accordingly, the gene expressions of antioxidant genes [42,45,46] such as NFE2L2, SOD1, TXN, GSR, CAT, and GPX1 were evaluated by qRT-PCR following UVC and/or sinularin treatments in oral cancer cells (Ca9-22 and CAL 27) (Figure 8). In general, most of the tested antioxidant gene expressions in these two cell lines were downregulated by separate treatments (UVC or sinularin alone). All of the test antioxidant gene expressions were downregulated by UVC/sinularin. Moreover, the antioxidant gene expressions were lower in the UVC/sinularin treatment than in the separate treatments. For the multi-comparisons, the treatments marked without repeated characters (a to f) differ significantly (p < 0.05). The data were plotted as the mean ± SD (n = 3).

UVC/Sinularin Combined Treatment of Oral Cancer Cells Downregulates Antioxidant Gene Expressions
Because oxidative stresses were confirmed ( Figures 5-7), the source of the oxidative stress needs to be investigated. Cellular antioxidant signaling may regulate oxidative stress [49]. Accordingly, the gene expressions of antioxidant genes [42,45,46] such as NFE2L2, SOD1, TXN, GSR, CAT, and GPX1 were evaluated by qRT-PCR following UVC and/or sinularin treatments in oral cancer cells (Ca9-22 and CAL 27) (Figure 8). In general, most of the tested antioxidant gene expressions in these two cell lines were downregulated by separate treatments (UVC or sinularin alone). All of the test antioxidant gene expressions were downregulated by UVC/sinularin. Moreover, the antioxidant gene expressions were lower in the UVC/sinularin treatment than in the separate treatments. . For the multi-comparisons, the treatments marked without repeated characters (a to f) differ significantly (p < 0.05). The data were plotted as the mean ± SD (n = 3).

Discussion
Sinularin exhibits the selective killing of cancer cells, and it shows low cytotoxicity to several non-malignant cells [24][25][26], which may demonstrate that sinularin has a low side effect potential for normal cells. A combined treatment with natural products may alleviate radio-and chemo-resistance or side effects in cancer therapy [5,6]. Recently, several potential UVC sensitizers for the antiproliferation of oral cancer cells were reported [13,[15][16][17]. Accordingly, the current study evaluated the antiproliferation effect of the UVC/sinularin combined treatment of oral cancer cells. For the MTS and ATP assays, the UVC/sinularin treatment of two cell lines for oral cancer showed high antiproliferation, but normal oral cells were only weakly affected. The detailed mechanism of the synergistic antiproliferation of UVC/sinularin is discussed as follows.

UVC/Sinularin Combined Treatment Synergistically and Selectively Kills Oral Cancer Cells
Several studies reported that several clinical drugs and natural products enhanced the antiproliferation of cancer cells following UVC irradiation. For example, clinical agents such as cisplatin treatment (10 μM) and UVC irradiation (10 J/m 2 ) show synergistic suppression for proliferation to colorectal cancer cells [13]. Similarly, several kinds of combined treatments involving UVC and anticancer agents have been reported to provide the synergistic suppression of the proliferation of oral cancer cells, such as methanol extract

Discussion
Sinularin exhibits the selective killing of cancer cells, and it shows low cytotoxicity to several non-malignant cells [24][25][26], which may demonstrate that sinularin has a low side effect potential for normal cells. A combined treatment with natural products may alleviate radio-and chemo-resistance or side effects in cancer therapy [5,6]. Recently, several potential UVC sensitizers for the antiproliferation of oral cancer cells were reported [13,[15][16][17]. Accordingly, the current study evaluated the antiproliferation effect of the UVC/sinularin combined treatment of oral cancer cells. For the MTS and ATP assays, the UVC/sinularin treatment of two cell lines for oral cancer showed high antiproliferation, but normal oral cells were only weakly affected. The detailed mechanism of the synergistic antiproliferation of UVC/sinularin is discussed as follows.

UVC/Sinularin Combined Treatment Synergistically and Selectively Kills Oral Cancer Cells
Several studies reported that several clinical drugs and natural products enhanced the antiproliferation of cancer cells following UVC irradiation. For example, clinical agents such as cisplatin treatment (10 µM) and UVC irradiation (10 J/m 2 ) show synergistic suppression for proliferation to colorectal cancer cells [13]. Similarly, several kinds of combined treatments involving UVC and anticancer agents have been reported to provide the synergistic suppression of the proliferation of oral cancer cells, such as methanol extract of Cryptocarya concinna/UVC [16], ethyl acetate Nepenthes extract/UVC [17], and sulfonyl chromen-4-ones/UVC [15]. Therefore, the drug discovery of UVC radiosensitizers has the potential to improve oral cancer therapy.
Several synthetic agents and natural products have been developed to sensitize radiation effects upon cancer cells [16,50,51]. An ideal radiosensitizer can enhance the radiotherapy efficiency against tumors and is harmless to normal tissues. It is still necessary to identify effective radiosensitizers in oral cancer therapy.
Sinularin is a potential natural product with selective antiproliferation for oral [24] and breast [25] cancer cells. The drug safety of sinularin was demonstrated to provide high viability to normal breast and oral cells at 24 h of treatment [24,25] and to normal oral cells upon more prolonged exposure (48 h) (Figure 1). Moreover, normal oral cells exhibit high viability following UVC/sinularin combined treatment at 48 h (Figure 1), suggesting that such UVC/sinularin combined treatment is tolerated by normal oral cells. UVC/sinularin showed synergistic effects for antiproliferation against oral cancer cells, with the evidence of both MTS and ATP assays at 48 h. Therefore, oral cancer cells following UVC/sinularin treatment exhibit both synergistic and selective killing effects against cytotoxicity to normal cells.

Oral Cancer Cells Following UVC/Sinularin Combined Treatment Induce Higher Oxidative Stress Than the Separate Treatments
For oral cancer cells, UVC irradiation doses ranging from 12 to 14 J/m 2 have a potential for ROS production [16,17,35] (Figure 5), and sinularin at 23.5 [24], 2 and 3 µM ( Figure 5) also induces ROS production. Following the UVC/sinularin treatment, oral cancer cells generate several oxidative stress inductions, such as ROS production, MitoSOX generation, and MitoMP destruction. In contrast, these oxidative stress inductions of UVC/sinularin are minor in normal oral cells. These findings suggest that oral cancer cells following UVC/sinularin treatment selectively induce higher oxidative stress than normal oral cells.
Antioxidant gene expression and oxidative stress interplay in their relationship [52,53]. For example, resveratrol was reported to activate antioxidant signaling to suppress oxidative stress [54]. In contrast, downregulating antioxidant signaling-such as NFE2L2 [55,56], SOD1 [57], and CAT [58] genes-may trigger oxidative stress. In the current study, antioxidant gene expressions (NFE2L2, SOD1, TXN, GSR, CAT, and GPX1) in oral cancer Ca9-22 and CAL 27 cells were highly downregulated by UVC/sinularin combined treatments compared to the separate treatments ( Figure 8). Therefore, the downregulation of antioxidant signaling contributes to oxidative stress in the UVC/sinularin combined treatment.

UVC/Sinularin Combined Treatment in Oral Cancer Cells Induces Higher DNA Damage and Triggers Apoptosis Compared to the Single Treatment
Excessive oxidative stress can lead to DNA damage [59], and may result in the apoptosis of cancer cells [60]. Because oral cancer cells following UVC/sinularin treatment selectively induce oxidative stress, its inducing abilities for DNA damage and apoptosis are expected. Following the detection of γH2AX and 8-OHdG-detected DNA strand damage and oxidative DNA adduct damages, oral cancer cells following UVC/sinularin treatment showed the higher induction of DNA damage than the UVC or sinularin in separate treatments (Figures 9 and 10).
Moreover, oral cancer cells following UVC/sinularin treatment induce higher apoptosis than UVC or sinularin only. Oral cancer cells following UVC/sinularin trigger more apoptosis than normal oral cells. They are supported by the findings of the increment of the subG1 population and annexin V, and the activation for pan-caspase and Cas 3/7 activities. The Western blotting results further demonstrated that UVC/sinularin overexpresses c-PARP in oral cancer cells compared to UVC or sinularin in separate treatments ( Figure 4D). Still, it shows a low expression of c-PARP in normal oral cells. Therefore, oral cancer cells following UVC/sinularin treatment selectively cause apoptosis compared to normal oral cells.

UVC/Sinularin Combined Treatment in Oral Cancer Cells Induces Higher G2/M Arrest Than Separate Treatments
Several drugs have been reported to induce G2/M arrest associated with apoptosis, such as luteolin for colon cancer cells [61], erianin for osteosarcoma cells [62], and diosgenin for breast cancer cells [63]. UVC irradiation to oral cancer Ca9-22 cells at 24 h recovery time induces G2/M arrest [17]. Similarly, 24 h sinularin treatment induces G2/M arrest for liver [23], breast [25], and oral [24] cancer cells. The current study shows that oral cancer cells following UVC/sinularin at 48 h treatment exhibit moderate G2/M arrest and apoptosis compared to their separate applications. This warrants the detailed investigation of the contribution of G2/M arrest in UVC/sinularin-induced synergistic antiproliferation for oral cancer cells in the future.

The Combined Effects of UVC/Sinularin in Oral Cancer Cells Depend on ROS Regulation
The ROS scavenger NAC can revert the accompanied changes caused by combined UVC/sinularin treatment in oral cancer cells, including antiproliferation (MTS and ATP assays), G2/M arrest, apoptosis enhancement (annexin V, pan-caspase, Cas 3/7, and apoptosis signal), oxidative stress appearance (ROS, MitoSOX, and MitoMP), and DNA strand damage (γH2AX and 8-OHdG). These findings suggest that the synergistic antiproliferation mechanisms of UVC/sinularin in two oral cancer cell lines are mediated by oxidative stress.

Conclusions
Combined treatment is an effective strategy to improve cancer therapy. This study first investigated the enhanced and selective antiproliferation by combined treatment using UVC/sinularin in oral cancer cells. Based on the MTS and ATP assays, oral cancer cells following UVC/sinularin combined treatment show synergistic and selective killing. However, it shows low cytotoxicity acting upon normal oral cells in combined as well as separate treatments. Mechanistically, UVC/sinularin in oral cancer cells show higher ROS production, MitoSOX generation, and MitoMP destruction (indicating oxidative stresses) as well as annexin V, pan-caspase, Cas 3/7, and c-PARP expression (indicating apoptosis) than in normal oral cells. UVC/sinularin also shows higher G2/M arrest and γH2AX/8-OHdG expressions (DNA damages) in oral cancer cells than in single treatments. All of these accompanied changes of UVC/sinularin can be reverted by NAC preincubations, demonstrating that the UVC/sinularin-induced synergistic suppression of the proliferation of oral cancer cells is linked to oxidative stress-associated mechanisms. Therefore, a combined UVC/sinularin treatment for oral cancer cells provides a synergistic and selective antiproliferation function with low toxic side damage acting upon normal oral cells.