JS‐K induces reactive oxygen species‐dependent anti‐cancer effects by targeting mitochondria respiratory chain complexes in gastric cancer

Abstract As a nitric oxide (NO) donor prodrug, JS‐K inhibits cancer cell proliferation, induces the differentiation of human leukaemia cells, and triggers apoptotic cell death in various cancer models. However, the anti‐cancer effect of JS‐K in gastric cancer has not been reported. In this study, we found that JS‐K inhibited the proliferation of gastric cancer cells in vitro and in vivo and triggered mitochondrial apoptosis. Moreover, JS‐K induced a significant accumulation of reactive oxygen species (ROS), and the clearance of ROS by antioxidant reagents reversed JS‐K‐induced toxicity in gastric cancer cells and subcutaneous xenografts. Although JS‐K triggered significant NO release, NO scavenging had no effect on JS‐K‐induced toxicity in vivo and in vitro. Therefore, ROS, but not NO, mediated the anti‐cancer effects of JS‐K in gastric cancer. We also explored the potential mechanism of JS‐K‐induced ROS accumulation and found that JS‐K significantly down‐regulated the core proteins of mitochondria respiratory chain (MRC) complex I and IV, resulting in the reduction of MRC complex I and IV activity and the subsequent ROS production. Moreover, JS‐K inhibited the expression of antioxidant enzymes, including copper‐zinc‐containing superoxide dismutase (SOD1) and catalase, which contributed to the decrease of antioxidant enzymes activity and the subsequent inhibition of ROS clearance. Therefore, JS‐K may target MRC complex I and IV and antioxidant enzymes to exert ROS‐dependent anti‐cancer function, leading to the potential usage of JS‐K in the prevention and treatment of gastric cancer.

and in vivo cancer models. 1,2 JS-K induces differentiation of human acute myeloid leukaemia HL-60 cells, 3 inhibits cell proliferation and triggers caspase-dependent and caspase-independent cell death in various cancer cell lines. [4][5][6][7][8] Moreover, JS-K also inhibits human cancer angiogenesis and metastasis 9,10 and enhances the cytotoxicity of some chemotherapeutic drugs, including cisplatin and arsenic, in drug-resistant cells by increasing their intracellular drug concentration. 7,11 In addition, in the integrated animal, JS-K significantly suppresses the growth of some cancer cells inoculated subcutaneously in mice, including human myeloid leukaemia cells, 3 multiple myeloma cells, 6 prostate cancer cells 3 and non-small-cell lung cancer (NSCLC) cells, 12 demonstrating that JS-K is effective against both solid tumours and blood malignancies. However, it is not clear whether JS-K is effective in killing gastric cancer cells.
As a nitric oxide (NO) donor prodrug, JS-K is activated to release NO upon nucleophilic attack by reduced thiols, such as glutathione (GSH). 13,14 The reaction is catalysed by glutathione-S-transferases (GSTs), and JS-K has been identified as a selective GSTα targeting compound. 3,15 Glutathione-S-transferases are phase II detoxification enzymes and catalyse the conjunction of xenobiotics with cellular reduced GSH; therefore, overexpression of GST in tumour cells allows tumour cells to gain a selective survival advantage over normal cells to chemotherapeutics by enhanced detoxification through GSH conjunction. 16,17 Therefore, up-regulation of GST in tumour cells usually induces multi-drug resistance, and GSTs have been regarded as potent targets for anti-cancer drug design and synthesis. 17,18 The design strategy for JS-K set out to exploit the overexpression of GST in malignant cancer cells compared with that in normal tissue. In line with this concept, JS-K has been reported to F I G U R E 1 JS-K inhibits cell proliferation in different gastric cancer cell lines. A, The effect of JS-K on the proliferation of gastric epithelial cells and gastric cancer cells. GES-1, SGC7901, MGC803 and HGC27 cells were treated with different JS-K concentrations for 48 h, and an MTT assay was used to determine cell viability. Cell survival rates were calculated by normalizing cell survivals in different groups with those in the control group. The IC50 values were calculated by using the GraphPad Prism 7 software. B, JS-K inhibits the clonogenic ability of different gastric cancer cell lines. SGC7901, MGC803 and HGC27 cells were treated with different JS-K concentrations for 48 h and then cultured with medium without JS-K for another 7 d. Cells were stained with crystal violet, and the representative plates of three independent experiments are shown. C, JS-K induces G2-M phase arrest in SGC7901 cells. Cells were treated with JS-K at the indicated concentration for 12 h and then collected to determine the cell cycle phases with flow cytometry selectively kill human multiple myeloma cells but not patient-derived bone marrow stromal cells. 6 Recently, JS-K was reported to induce caspase-dependent apoptosis by facilitating reactive oxygen species (ROS) accumulation, 5,7,12,19 but the specific target and detailed mechanism by which JS-K triggers ROS accumulation is not completely understood. Reactive oxygen species are reactive chemical species containing oxygen and are formed as the by-product of normal oxygen metabolism. 20 The mitochondria respiratory chain (MRC) is the main source of the electrons required for ROS production; therefore, broken of MRC resulted from the inhibition of MRC complex activity promotes ROS production by facilitating electron escape from the MRC. 21,22 In normal cells, ROS are cleared by antioxidant enzymes, including superoxide dismutase (SOD), catalase and GSH peroxidase; therefore, the equilibrium between ROS production and clearance maintains ROS at low levels in normal cells. [21][22][23][24] However, suppression of the activity of MRC complexes and antioxidant enzymes promotes ROS production and inhibits ROS clearance, leading to ROS accumulation.
High ROS levels usually cause damage to lipids, protein and DNA, and then induce oxidative stress or cell death; therefore, many kinds of chemotherapeutic drugs have been reported to kill cancer cells through facilitating ROS accumulation. 25,26 F I G U R E 2 JS-K induces caspase-dependent apoptosis in SGC7901 cells. A, JS-K induced apoptosis of SGC7901 cells in a dose-dependent manner. Cells were treated with JS-K at the indicated concentration for 24 h, and cell death was measured with flow cytometry. *P < 0.05. **P < 0.01. B, The effect of different caspase inhibitors on JS-K-induced cell death. SGC7901 cells were treated with JS-K in the presence or absence of Z-VAD (50 μmol/L), Z-LEHD (50 μmol/L) or Z-DEVD (50 μmol/L) for 24 h, and cell death was measured with flow cytometry. **P < 0.01. C. JS-K induces PARP, caspase 9 and caspase 3 cleavage. SGC7901 cells were treated with JS-K for the indicated time, and Western blotting was used to detect PARP, caspase 9 and caspase 3 cleavage. Actin was used as a loading control. D, JS-K promotes caspase 9 and caspase 3 activation. SGC7901 cells were treated with JS-K for 12 h and then harvested to measure the caspase 3 and caspase 9 activities with specific assay kits. More than three independent experiments were performed for each group, and the relative caspase activities were calculated by normalizing the caspase activities of all groups with the activities in a negative control group. *P < 0.05. **P < 0.01 Gastric cancer is a leading cause of malignant death worldwide 27 ; therefore, developing effective anti-cancer reagents and exploring the molecular mechanisms of chemotherapeutics are fundamental for gastric cancer prevention and treatment. Therefore, the present study was designed to evaluate the effect of JS-K on gastric cancer cells.
We found that JS-K inhibited gastric cancer cells proliferation by blocking the cell cycle in the G2-M phase and suppressed the growth of gastric cancer subcutaneous xenografts. Although JS-K induced significant ROS accumulation and NO release, the clearance of ROS, but not of NO, reversed JS-K-induced apoptosis and inhibition of gastric xenograft tumour growth; therefore, ROS accumulation is essential for the JS-K-induced toxic effect in gastric cancer. This mechanistic study on JS-K-induced ROS accumulation demonstrated that JS-K significantly down-regulated the core proteins of MRC complex I and IV, which contributed to the reduction of MRC complex I and IV activity and the subsequent ROS production. In addition, JS-K also suppressed the expression of antioxidant enzymes, including copper-zinc-containing superoxide dismutase (SOD1) and catalase, resulting in the decrease of SOD1 and catalase activity and the subsequent inhibition of ROS clearance. Therefore, our study identified a potent target and molecular mechanism for JS-K in mediating ROS-dependent anti-neoplastic effects in gastric cancer.

| Clonogenic survival assay
One thousand SGC7901 cells were plated per well in 6-well plates for 48 hours and followed by treatment with JS-K at the indicated concentrations for another 48 hours. The cells were washed, fresh culture fluid was added, and colonies were stained with crystal violet 7 days later.

| Apoptosis analysis
Cells were collected by trypsinization and stained with annexin V-FITC and propidium iodide provided in the Apoptosis Assay Kit F I G U R E 3 JS-K-induced cytotoxicity was mediated by reactive oxygen species (ROS) accumulation but not nitric oxide (NO) release. A, JS-K induces ROS accumulation and NO release in a dose-dependent manner. SGC7901 cells were treated with JS-K at the indicated concentration for 3 h and then harvested to measure the ROS and NO levels with flow cytometry. Three independent experiments were performed for each group. The relative ROS or NO levels were calculated by normalizing the ROS or NO levels in all the groups with those in a control group. *P < 0.05. **P < 0.01. B, JS-K-induced cytotoxicity was reversed by a ROS clearance reagent but not a NO scavenger. SGC7901 cells were treated with JS-K in the presence or absence of carboxy-PTIO (100 μmol/L), N-acetyl-L-cysteine (NAC) (500 μmol/L) and Z-VAD (50 μmol/L) for 24 h, and cell survival was measured with flow cytometry. **P < 0.01. C, NAC, but not carboxy-PTIO, inhibited the PARP, caspase 3 and caspase 9 cleavage induced by JS-K. SGC7901 cells were treated with JS-K in the presence or absence of NAC, carboxy-PTIO or Z-VAD for 12 h, and Western blot analysis was used to detect PARP, caspase 3 and caspase 9 cleavage. Actin was used as a loading control. D, NAC suppresses JS-K-induced caspase 3 and caspase 9 activation. SGC7901 cells were treated with JS-K in the presence or absence of NAC or Z-VAD for 12 h and then harvested to measure caspase 9 and caspase 3 activities with specific assay kits. The relative caspase activities were calculated by normalizing the caspase activities of all the groups with the activities of a negative control group. *P < 0.05. E, The effect of NAC on ROS accumulation and nitric oxide release induced by JS-K. SGC7901 cells were treated with JS-K in the presence or absence of NAC for 12 h and then harvested to measure the ROS or nitric oxide levels with flow cytometry. The relative ROS and nitric oxide levels were calculated by normalizing the level of ROS and nitric oxide in all the groups with those in a negative control group. **P < 0.01 (Beyotime Institute of Biotechnology, Haimen, China), and then analysed with flow cytometry (FACS Calibur) and the Cell Quest software (FACS Calibur). More than 10 000 cells were analysed for each measurement. More than three independent experiments were performed in each group, and the representative measurements are shown.

| Caspase activity assays
The activities of caspase 3 or caspase 9 were measured using the   F I G U R E 4 JS-K induces reactive oxygen species-dependent cytotoxicity by activating the mitochondria apoptosis pathway. A, NAC inhibits the depolarization of mitochondria induced by JS-K. SGC7901 cells were treated with JS-K in the presence or absence of NAC for 24 h, stained with JC-1 and analysed with flow cytometry. The JC-1 red/green fluorescence intensity ratio was normalized by comparing the data with the control group and is represented as relative mitochondrial membrane potential. Each experiment was performed in triplicate, and the representative measurements are shown. **P < 0.01. B, JS-K induces the cytoplasmic release of Cytochrome c (Cyto-C). SGC7901 cells were treated with 30 μmol/L JS-K for the indicated time, and then harvested to isolate mitochondria and cytoplasm. Western blot analysis was used to detect the Cyto-C levels in the mitochondria and cytoplasm. GAPDH and VADC were used as loading controls. C, The effect of NAC and carboxy-PTIO on JS-K-induced Cyto-C release. SGC7901 cells were treated with JS-K in the presence or absence of NAC or carboxy-PTIO for 12 h and then harvested to isolate mitochondria and cytoplasm. Western blot analysis was used to detect Cyto-C levels in the mitochondria and cytoplasm. Actin and VADC were used as loading controls. D, The effect of Cyto-C knockdown on JS-K-induced cell death. SGC7901 cells were transfected with Cyto-C siRNA and a negative control siRNA for 24 h and then treated with or without JS-K for another 24 h. Cells were stained with annexin V-FITC and PI and analysed by flow cytometry. Western blot analysis was used to evaluate the efficiency of Cyto-C knockdown. *P < 0.05. E, JS-K-induced apoptosis was inhibited by ectopic expression of Bcl-2 or Bcl-xL. SGC7901 cells were transfected with Bcl-2, Bcl-xL or negative control plasmids for 24 h, and then treated with 30 μmol/L JS-K for another 24 h. Cells were stained with annexin V-FITC and PI, and then analysed by flow cytometry. Western blot analysis was used to determine the Bcl-2 and Bcl-xL levels. GAPDH was used as a loading control. **P < 0.01 The siRNA target sequence against Cyto-C is: 5ʹ-actcttacacagccgccaata-3ʹ. and anti-β-actin (A5441) (Sigma-Aldrich).

| Ectopic expression of Bcl-2 and Bcl-xL
The plasmids expressing Bcl-2 or Bcl-xL and the empty negative control plasmid were purchased from Genechem (Shanghai, China). Plasmid transfections were performed using the Chemifect transfection reagent (Fengrui Biology) according to the manufacturer's protocol.
Briefly, SGC7901 cells were seeded in 6-well plates for 24 hours to reach 50%-70% confluence, and then the transfection complex consisting of plasmid and Chemifect transfection reagent was added into the cell culture medium. After 48 hours, the ectopic expression efficiency was evaluated by Western blot.

| Histopathology
Tissues were immediately collected from killed mice and fixed in Mice were administered with JS-K (3 mg/kg) once every 2 d for 30 d, and blood was obtained before being killed to isolate serum. The levels of alanine transaminase, aspartate transaminase and creatinine in serum were measured using specific assay kits. After blood collection, the mice were killed to isolate the liver and kidney. The tissues were cut into slices and then stained with haematoxylin and eosin. Four mice in each group were examined independently, and representative images are shown (×200) ZHAO ET AL.
| 2497 microscope slides and then stained with haematoxylin and eosin (H&E) for analyses. Representative images were captured using identical settings in a Leica DM2500 optical microscope.

| Activity measurement of ALT, AST and creatinine
Blood obtained from the mice was placed at room temperature for 60 minutes to clot and then centrifuged at 3000 g for 10 minutes at

| Statistical analysis
GraphPad Prism 7 software was used to analyse the data and construct statistical graphs. Statistical significance was analysed using The tumour tissues isolated from the mice used in Figure 5C were minced and lysed, and Western blot assays were used to determine the protein levels of Ndfus4 and COX2 in the tumour tissue lysates. GAPDH was used as a loading control. F, JS-K decreases SOD1 and catalase activity in SGC7901 cells. Cells were treated with JS-K at the indicated concentration for 12 h and collected to measure SOD1 and catalase activity using specific assay kits. The activity in JS-K-treated groups was normalized to that in the control group. *P < 0.05, **P < 0.01. G, JS-K decreases SOD1 and catalase activity in gastric tumour tissues. The lysates of isolated gastric tumour tissues ( Figure 5A) were used to measure the activities of SOD1 and catalase using specific assay kits, and the relative activities were normalized by comparing the activities of SOD1 and catalase in all groups with those in a negative control group. **P < 0.01. H, JS-K down-regulates SOD1 and catalase in SGC7901 cells. Cells were treated with JS-K at the indicated concentrations for 12 h and then harvested to detect the protein levels of SOD1 and catalase with Western blot analysis. Actin was used as a loading control. I, JS-K administration suppresses SOD1 and catalase expression in gastric tumour tissues. Tumour tissues were isolated from the negative control and JS-K-treated mice used in Figure 5C and minced and lysed to determine SOD1 and catalase protein levels using Western blot. Actin was used as a loading control ANOVA or unpaired t tests and defined as *P < 0.05 or **P < 0.01.
All the experiments were repeated at least three times, and the data are expressed as the mean ± SD from representative experiments.

| JS-K induces caspase-dependent apoptosis in gastric cancer cells
JS-K has been reported to induce caspase-dependent and independent cell death in various cancer cell lines 4,6,8,28 ; therefore, we next detected whether JS-K induces apoptosis in gastric cancer cell lines.
As shown in Figure 2A,B, JS-K induced SGC7901 cell death in a dose-dependent manner, and Z-VAD, a pan-caspase inhibitor, almost completely inhibited JS-K-induced cell death, suggesting that JS-K induced caspase-dependent apoptosis. In addition, we found that JS-K-induced cell death was significantly inhibited by Z-LEHD-FMK and Z-DEVD-FMK, the caspase 9 and caspase 3 inhibitors, respectively, indicating that the activation of caspase 9/3 is essential for JS-K-induced apoptosis.
To further explore the mechanism of JS-K-induced cell death, we detected the effect of JS-K on caspase signalling pathway activation in SGC7901 cells. As shown in Figure 2C, significant cleavage of caspase 9, caspase 3 and its substrate protein, poly (ADP-ribose) polymerase (PARP), were detected in SGC7901 cells following JS-K stimulation, indicating that JS-K initiated caspase signalling pathway activation in gastric cancer cells. Moreover, we also found that the activities of caspase 9 and caspase 3 were significantly increased in SGC7901 cells following JS-K administration ( Figure 4D) Moreover, we found that Cyto-C knockdown by using siRNA protected SGC7901 cells from JS-K-induced apoptosis ( Figure 4D), indicating that the cytoplasmic releasing of Cyto-C is essential for JS-Ktriggered apoptosis. In addition, we found that ectopic expression of Bcl-2 and Bcl-xL, which are two proteins that protect against mitochondrial apoptosis, significantly protected SGC7901 cells from JS-K-induced cell death ( Figure 4E), further confirming that JS-K-induced cell death occurs through mitochondria apoptotic pathway activation.
In conclusion, JS-K induced ROS-dependent mitochondria apoptotic cell death of gastric cancer cells by promoting mitochondria depolarization and subsequent cytoplasmic release of Cyto-C.

| In vivo anti-tumour effects of JS-K on gastric cancer cells
As JS-K has been identified to exhibit significant in vitro cytotoxic effects on gastric cancer cell lines, we next determined its in vivo anti-tumour effects on gastric cancer cells. First, we established a gastric cancer mouse model by subcutaneously inoculating SGC7901 The model of JS-K-induced anti-tumour activity in gastric cancer. As a lead anti-cancer drug compound, JS-K significantly suppressed the expression of the core proteins of mitochondria respiratory chain (MRC) complex I and IV, resulting in the reduction of MRC complex I and IV activity and the subsequent reactive oxygen species (ROS) production.
In addition, JS-K down-regulated SOD1 and catalase, which facilitated the reduction of SOD1 and catalase reducing activity and promoted the inhibition of ROS clearance. The aberrant ROS then induces mitochondria depolarization, caspase signalling pathway activation and subsequent apoptotic cell death. Therefore, MRC complex I and IV or antioxidant enzymes act as novel targets for JS-K in mediating ROS-dependent anti-cancer activity in gastric cancer cells into the right flank region of nude mice. JS-K was administered through the tail vein to evaluate its anti-cancer effects. As shown in Figure 5A, the tumours were smaller in the JS-K-administered mice than those in the negative control group, indicating that JS-K suppressed gastric cancer xenograft growth. Moreover, the average tumour weight was also lower in the JS-K-administered group than that in the negative group, further confirming the in vivo anti-cancer effect of JS-K against gastric cancer cells. As our data demonstrated that ROS accumulation contributed to JS-K-induced cytotoxicity in gastric cancer cells and that aberrant ROS usually induces oxidative stress in vivo, we next determined the oxidative state in a gastric cancer xenograft model. As malondialdehyde (MDA) is the product of lipid oxidation and is usually used to represent the oxidative stress level, we measured MDA levels in the gastric tumour tissues.
As shown in Figure 5B, In addition, no significant pathological changes were observed in liver and kidney tissues isolated from the negative control mice and the JS-K-treated mice ( Figure 5F), confirming that JS-K (3 mg/kg) had no significant toxic effect on mouse livers and kidneys.
Collectively, JS-K significantly induced the growth inhibition of gastric cancer cells and oxidative stress in vivo.

| JS-K down-regulates MRC complex I and IV core proteins, as well as the antioxidant enzymes
Although ROS accumulation mediated the anti-cancer effect in gastric cancer cells in vitro and in vivo, the involved mechanism is not clear. Therefore, we next explored the mechanism for JS-K-induced ROS accumulation. Reduction of MRC complex activity promotes electron escape to produce ROS; therefore, we determined the effects of JS-K on MRC complex activity. Figure 6A shows mice. 3,6,29 Therefore, JS-K exhibits significant anti-cancer activity in both solid tumours and blood malignancies. However, the role of JS-K in the treatment of orthotopic tumours is controversial because JS-K inhibited the growth of human hepatoma JM-1 cells implanted intrahepatically in nude rats 2 ; however, it did not result in tumour growth retardation or extend survival within tracranial U87 rat glioma. 30 JS-K has been reported to reduce the growth of U87 cells inoculated subcutaneously bilaterally into the flank of nude rats, 13 and the dosage of JS-K administration in the treatment of U87 cell subcutaneous xenografts (4 μmol/L/kg) is higher than that used in the treatment of orthotopic U87 rat glioma tumours (3.5 μmol/L/ kg). 13,30 Therefore, low dosages of JS-K administration may be not enough to suppress the growth of U87 orthotopic glioma. Moreover, the effective dosage of JS-K administration used in the previous reports for the treatment of subcutaneous xenografts ranged from 4 to 75 μmol/L/kg, 3,9,12,13 and our results also demonstrate that 1.5 mg/kg (4 μmol/L/kg) JS-K administration efficiently inhibited the growth of gastric cancer subcutaneous xenografts, further confirming that the failure of JS-K treatment in orthotopic U87 rat glioma may be attributed to the low dosage of JS-K that was administered.
Therefore, our results and the previous studies all demonstrate that the appropriate dosage of JS-K administration efficiently inhibits the growth of many kinds of cancer xenografts, including gastric cancer.
In this study, we found that JS-K induced ROS accumulation in gastric cancer cells and oxidative stress in gastric tumour tissues.
Moreover, NAC significantly reversed JS-K-induced ROS accumulation and the subsequent cytotoxicity. Therefore, ROS are essential for JS-K to exert its anti-tumour function in gastric cancer. In line with our observation, JS-K also induces ROS accumulation in prostate, bladder and NSCLC cells, and the aberrant ROS trigger caspase-dependent apoptosis by promoting mitochondrial depolarization and subsequent caspase pathway activation. 5,7,12 Even in blood tumour cells, NAC protects leukaemia HL-60 and U937 cells from JS-K-induced apoptosis. 4,19 Therefore, ROS accumulation plays a key role in mediating JS-K-induced cytotoxicity in solid tumours as well as blood malignancies. Reactive oxygen species accumulation usually leads to oxidative stress and DNA damage, which then activate c-Jun N-terminal kinase (JNK) and initiate apoptotic cell death. 26,31 Consistent with this concept, ROS accumulation and the subsequent JNK activation mediate JS-K-induced cell death and growth arrest in human hepatoma, 32 NSCLC 33 and multiple myeloma cells. 6,34 As many kinds of chemotherapeutic drugs, including arsenic trioxide and cisplatin, also exert their anti-cancer activities by facilitating ROS accumulation, 11 JS-K is a promising anti-cancer drug with a similar mechanism.
JS-K is a NO donor prodrug, which releases high levels of NO upon enzymatic activation by GSTα. 1,35 As NO has been reported to induce apoptosis or differentiation in human leukaemia cells, 36 it is possible for NO to mediate JS-K-induced cytotoxicity. In line with the postulation, JS-K-induced cytotoxic effects were abolished by cobalamin (Vitamin B12), a NO scavenger, in human multiple myeloma cells. 6 Moreover, carboxy-PITC, another NO scavenger, was found to reverse JS-K-induced growth inhibition of hepatoma carcinoma JM-1 cells that were intrahepatically implanted into nude rats. 2 Therefore, NO release has been identified as another important mechanism for JS-K in killing tumours. In this study, we found that JS-K significantly induced NO release and ROS accumulation simultaneously in gastric cancer cells; however, JS-K-induced cell death and tumour growth inhibition were blocked by NAC but not carboxy-PTIO. In addition, we found that NAC almost completely suppressed ROS accumulation but had no effect on NO release.
Therefore, high NO levels are not enough to initiate gastric cancer cell death in response to JS-K stimulation, and ROS played a critical role in mediating the JS-K-induced anti-cancer effects in gastric cancer cells. Consistent with our results, the NO scavenger, Vitamin 12, has no significant effect on the ability for JS-K to kill the erythroleukemia SFFV-MEL cells. 37 Furthermore, low-dose JS-K induces cell death, NO release and ROS accumulation in sensitive NSCLC cells, but only NO release in resistant NSCLC cells. 12 Therefore, it is easy to postulate that both NO release and ROS accumulation are required to mediate JS-K-induced cytotoxicity. Nitric oxide has been reported to be an antioxidant molecule that can be oxidized by ROS into nitrite and then trigger cell death in prostate and bladder cancer cells. 5,7,38 Therefore, it is natural to hypothesis that NO release and ROS accumulation may work together to exert anti-cancer effects in an integrated manner. Nitric oxide alone has been found to induce differentiation and apoptosis in human leukaemia cells 36 ; therefore, we cannot exclude the possibility that NO mediates the anti-cancer effects of JS-K in some NO-sensitive cancer cells in the absence of ROS accumulation.
Although ROS accumulation has been reported to mediate JS-Kinduced anti-cancer effects in different cancer cell lines, 4,5,7,12,19 the target and involved mechanism for JS-K in promoting ROS accumulation are not completely clear. Reactive oxygen species accumulation is usually derived from a disrupted balance between ROS production and clearance. 24 The MRC is well known to be the main source of free electron, and inhibition of MRC complex activity leads to the MRC disruption, which then facilitates electron escape to form ROS. 39,40 Therefore, inhibition of MRC complex activation increases ROS production, 39,40 which has been identified as the main mechanism for ROS accumulation and ROS-dependent cytotoxicity induced by some anti-cancer compounds, including celastrol and rotenone. 41,42 In this study, we found that JS-K induced a significant decrease in MRC complex I and IV activity in vitro and in vivo, which promoted ROS production and increased the cellular ROS levels.
Therefore, MRC complex I and IV might be novel targets for JS-K in mediating ROS accumulation. In addition, we also found that JS-K decreased the activity of antioxidant enzymes, including SOD1 and catalase, which may facilitate ROS accumulation by inhibiting ROS clearance. Consistent with our results, JS-K down-regulated SOD1 expression in human leukaemia HL-60 cells in a previous study. 43 The disruption of reductase antioxidant systems, including SOD1, catalase or GSH peroxides, is well known to contribute to ROS-dependent cytotoxicity for some chemotherapeutic agents. 24,26 Therefore, the suppression of antioxidant enzyme activation might be another mechanism by which JS-K induces ROS accumulation. In addition, we also explored the potential mechanism for JS-K-induced reduction of MRC complex I and IV and antioxidant enzyme activity.
We found that JS-K significantly down-regulated MRC complex I and IV core proteins and antioxidant enzymes SOD1 and catalase, which may result in the reduction of activity of MRC complex I and IV and antioxidant enzymes and the promotion of ROS accumulation.
Therefore, suppressing the expression of MRC complex I and IV core proteins and antioxidant proteins may be the major mechanism for JS-K to induce ROS accumulation. Furthermore, as a NO donor prodrug, JS-K reacts with GSH to produce NO upon GSTα metabolism; therefore, JS-K was identified to decrease cellular GSH levels in different cancer cells. 15,19 GSH depletion is known to facilitate ROS accumulation by suppressing ROS clearance; thus, we cannot exclude the possibility that JS-K increases cellular ROS levels by inducing GSH depletion. Although these mechanisms might be cell type specific or selective, they are not mutually exclusive and work together to promote ROS accumulation induced by JS-K in an integrated manner.
In summary, for the first time, we evaluated the anti-cancer effects of JS-K in gastric cancer in this study and found that JS-K inhibited gastric cancer cell growth in vitro and in vivo. In addition, JS-K induced the activation of the mitochondria apoptotic pathway and subsequent cell death by facilitating ROS accumulation but not NO release. We also explored the potential mechanism for JS-K-induced ROS accumulation and found that JS-K significantly down-regulated MRC complex I and IV core proteins and antioxidant enzymes, which resulted in the reduction of activity of MRC complex I and IV and antioxidant enzyme and then facilitated ROS production and clearance inhibition. These results raised the possibility that JS-K targets the antioxidant enzymes and MRC complex I and IV to exert its anti-tumour function ( Figure 7). Therefore, our study determined the ROS-dependent anti-cancer effect of JS-K in gastric cancer and identified novel targets and the involved mechanism for JS-K-induced ROS accumulation, leading to the development of JS-K in cancer prevention and treatment.