2′-Benzoyloxycinnamaldehyde (BCA), one of the derivatives of 2′-hydroxycinnamaldehyde (HCA) isolated from the bark of Cinnamomum cassia, induces apoptosis in human cancer cells. We found that BCA induces stronger antiproliferative effects in K-ras-transformed cells (RK3E-ras) than in isogenic non-transformed cells (RK3E). Treatment of RK3E-ras with BCA resulted in increased ROS generation and depletion of intracellular glutathione, whereas BCA-treated RK3E showed no significant increase in the ROS level with concurrent increase in intracellular glutathione (GSH). Thiol antioxidants recovered cell proliferation inhibition caused by BCA in both cell lines, while non-thiol antioxidants failed to recover cell death. BCA decreased metallothionein (MT) expression in RK3E-ras, while inducing remarkable MT expression in RK3E. The increase of intracellular GSH in RK3E is partially caused by differential induction of γ-glutamylcysteine synthetase (γ-GCS) due to BCA treatment. To evaluate the upstream pathway for differential expression of γ-GCS and MT, we analyzed early DJ-1 (PARK7) and NF-E2 p45-related factor 2 (Nrf2) changes after BCA treatment. In RK3E, DJ-1 expression considerably increased for 3 h with concurrent induction of Nrf2, whereas in RK3E-ras cells BCA decreased these protein levels. Based on these findings, it seems that the therapeutic selectivity of BCA in RK3E-ras results from decreased thiol antioxidants via decreased DJ-1 and Nrf2 expression. (Cancer Sci 2011; 102: 212–218)
2′-Benzoyloxycinnamaldehyde (BCA) (Fig. 1A) was origi-nally synthesized as a derivative of 2-hydroxycinnamaldehyde (HCA, Fig. 1B),(1) a natural compound isolated from the bark of Cinnamomum cassia Blume.(2) These compounds as antitumor agents induce apoptosis via reactive oxygen species (ROS) generation, caspase-3 activation(3) and the inhibition of proteasome activity.(4) BCA significantly blocked tumor growth in a nude mouse assay without bodyweight loss, making it a good drug candidate for cancer therapy.(3) It was also reported that the hepatic tumor volume and the total number of tumors decreased in BCA-treated mice compared with control H-ras12V transgenic mice.(5) In a rat oral tumor model induced by RK3E-ras-Fluc cells, direct injection of HCA into tumor significantly inhibited the growth of tumor mass.(6) Histological analysis also showed that HCA decreased tumor cell proliferation and induced apoptosis in a rat tumor.(6) When establishing a rat tumor model for the oral antitumorigenic test, we found that HCA and BCA had a greater antiproliferation effect in K-ras-transformed RK3E cells (RK3E-ras) than in isogenic normal K-ras cells (RK3E). We therefore tried to discern the underlying antitumor mechanism of BCA via the differential antiproliferative effect in these two cells.
K-ras is generally considered one of the most frequently mutated genes in a wide variety of human cancers. The highest incidences of K-ras mutation are found in: adenocarcinomas of the pancreas (90%), colon (50%) and lung (30%); in thyroid tumors (50%); and in myeloid leukemia (30%).(7) Therefore, using isogenic human cancer cells with or without the mutant K-ras gene could be a valuable strategy in drug screening for selective toxicity towards cancer cells.(8) A previous study showed that the intrinsic oxidative stress associated with K-ras-transformed oncogenic transformation may cause cancer cells that are highly dependent on their antioxidant systems to maintain redox balance.(9) In contrast, such stress is less likely in normal cells due to their low basal ROS level. This biochemical difference between normal and cancer cells may thus constitute a strategy for modulating cellular ROS to selectively kill cancer cells.(10) Hypothesizing that the ROS stress associated with oncogenic transformation would make cells highly dependent on their antioxidant systems to remove the damaging effect of ROS, we therefore compared intracellular ROS with major cellular antioxidant systems such as glutathione (GSH) and metallothionein (MT) levels, which are important in protecting against oxidative-stress-induced injury due to BCA treatment.
Nrf2 is generally known to regulate cytoprotective genes such as γ-glutamylcysteine synthetase (γ-GCS) and MT that contain an antioxidant response element (ARE) in their promoters. Another important mechanism for GSH regulation could be via the DJ-1 protein, which improves cell survival from oxidative damage by increasing cellular GSH levels.(11) Therefore, we evaluated whether BCA-induced differential antiproliferation in RK3E and RK3E-ras cells has been related with DJ-1 and Nrf2 pathways.
Materials and Methods
Chemicals and antibodies. The ROS was measured with 2′,7′-dichlorodihydro-fluorescein diacetate (DCFH2-DA) from Fluka (Buchs, Switzerland). Specific electron transport complex inhibitors such as rotenone, thenoyltrifluoroacetone (TTFA), antimycin A and potassium cyanide (KCN), and specific GSH synthesis inhibitor buthionine sulfoximine (BSO) were from Sigma–Aldrich (St Louis, MO, USA). Several antioxidants including glutathione-monoethyl ester (GSH-MEE), N-acetyl-L-cysteine (NAC), cysteine-HCl, butylated hydroxytoluene (BHT) and pyruvate were from Sigma–Aldrich. DJ-1 monoclonal antibody was from Cell Signaling Technology (Danvers, MA, USA). γ-GCS (polyclonal, catalytic heavy chain subunit) antibody was from Santa Cruz (Santa Cruz, CA, USA), and MT (monoclonal, isotypes 1 and 2) antibody was from Dako (Carpinteria, CA, USA). Nrf2 monoclonal antibody was from Abcam (Cambridge, CA, USA).
Cell culture. Rat kidney RK3E and K-ras-transformed RK3E-ras cell lines have been described previously.(12) Cells were grown in DMEM (GibcoBRL, Carlsbad, CA, USA), supplemented with 10% FBS, 100 units/mL penicillin and 100 μg/mL streptomycin (GibcoBRL). Cells were maintained at 37°C in a 5% CO2 humidified atmosphere.
Cell proliferation assay and morphological change. Cells (5000 cells/well) were seeded into 96-well plates. After 24 h, the cells were pretreated with BSO (20 μM), GSH-MEE (5 mM), NAC (7.5 mM), cysteine-HCl (1.85 mM), BHT (50 μM) or pyruvate (2 μM) for 30 min before the addition of BCA or Me2SO. After incubation for 24 h, cell proliferation was evaluated by performing 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoli-um bromide (MTT) assays. Cellular morphology was investigated using phase contrast microscopy (Olympus BX 51, Tokyo, Japan).
Cell cycle analysis. To analyze the DNA content, 24-h-cultured cells were trypsinized, washed twice with PBS and fixed with ice cold 70% ethanol overnight. Fixed cells were washed twice with PBS containing 1% FBS. Collected cells were resuspended in PBS (1 × 105 cells/100 μL) and treated with 100 μg/mL RNase A at 37°C for 30 min. Propidium iodide was then added to a final concentration of 50 μg/mL for DNA staining, and the fixed cells were analyzed by flow cytometry using FACScalibur (BD Biosciences, San Jose, CA, USA).
Measurement of intracellular ROS. Intracellular ROS accumulation was detected using the H2O2-sensitive fluorescent dye DCFH2-DA. Cells were pretreated with rotenone (2.0 μM), TTFA (0.25 mM), antimycin A (0.25 μM), KCN (0.25 mM) or NAC (7.5 mM) in PBS containing 5 μM of DCFH2-DA for 30 min before the addition of 20 μM of BCA or Me2SO in serum-free DMEM medium. After incubation for 6 h, cells were washed with PBS and intracellular ROS was analyzed by flow cytometry.
Glutathione assay. Cells were seeded in 100 mm culture dishes and treated with BCA for 24 h. After washing with PBS, trypsinized cells were resuspended in 5 mL of PBS. One hundred microliters of the resuspended cells were used for protein quantification. Cells were centrifuged at 600g and treated with three volumes of 5% 5-sulfosalicylic acid (SSA). After lyses by two cycles of freezing and thawing with liquid nitrogen, the lysates were centrifuged at 10 000g for 10 min to collect supernatant. Total GSH concentration (reduced GSH + oxidized GSH) was determined by Glutathione assay kit (Sigma–Aldrich) according to the manufacturer’s instructions. The concentrations of sample glutathione were calculated from the standard curve of serial diluted standard glutathione.
Western blot analyses. Cells were washed twice with cold PBS, after which PRO-PREP protein extraction solution (Intron, Daejon, South Korea) was added. Cell lysates were centrifuged to remove cellular debris, and protein concentrations were estimated using the Coomassie protein assay reagent (Thermo Scientific, Rockford, IL, USA). Forty micrograms of protein samples were electrophoresed on 10–12% SDS-PAGE gel. Proteins were then transferred to nitrocellulose membranes, which were blocked in 5% skim milk in TBS (25 mM Tris base and 150 mM NaCl) for 2 h at room temperature, and then incubated with 0.2 μg/mL of primary antibody overnight at 4°C. Horseradish peroxidase conjugated secondary antibody was used at 1:5000 dilutions for 1 h at room temperature and then washed three times for 10 min each in TBST (TBS and 0.1% Tween 20). The target proteins were detected with ECL detection reagents and the relative intensity of the bands was analyzed by Image-J software (NIH, Bethesda, MD, USA).
Statistical analysis. The differences in mean values among groups were evaluated and the values expressed as the means ± SD. All statistical calculations were carried out using SPSS 14 (SPSS Inc, Chicago, IL, USA) software.
Inhibition of cell proliferation and cell cycle. Cells were treated with various concentrations of BCA for 24 h and then a MTT assay was performed. BCA inhibited cell growth dose-dependently in both cell lines (Fig. 2A). However, BCA induced antiproliferation more effectively in RK3E-ras (IC50 value, 9.7 μM) when compared with RK3E (IC50 value, >30 μM). We analyzed the cell cycle distribution under BCA-induced growth inhibitory conditions using flow cytometry. When RK3E-ras was treated with BCA for 24 h, the cell cycle was arrested at the G2/M phase depending on BCA concentrations (Fig. 2B). G2/M-arrested cells were increased from 22.7% in 0.1% Me2SO-treated control to 40.7% in BCA-treated (30 μM) RK3E-ras cells. The cell population in the sub-G1 phase also increased from 3.2% in the control to 30.9% in the BCA-treated RK3E-ras cells. On the other hand, under the same experimental conditions, BCA had no significant effect on cell cycle phase distribution or on the sub-G1 cell population in RK3E.
Intracellular ROS measurements. Intracellular ROS production was measured by DCF fluorescence assay following BCA treatment. In RK3E and RK3E-ras cells, the basal ROS levels were similar at time 0 of the BCA treatment (33.7, and 39.8 ± SD%, respectively). A time-dependent increase of the percentage of cells with an elevated ROS level after treatment with 20 μM of BCA was observed in RK3E-ras (Fig. 3A). However, in RK3E, the percentage of cells with an elevated ROS level increased at 30 min of BCA treatment and then gradually decreased, without declining below the basal level. After treatment for 24 h, 79.8% of the RK3E-ras cells were positive for DCF fluorescence, while only 37.1% of RK3E cells showed DCF fluorescence.
To determine whether the ROS produced after BCA treatment were of mitochondrial origin, we pretreated the cells with specific inhibitors for mitochondrial electron transport chain complexes (ETC.) such as rotenone (inhibitor of NADH dehydrogenase, ETC. I), TTFA (inhibitor of succinate dehydrogenase, ETC. II), antimycin A (inhibitor of cytochrome c reductase, ETC. III) and KCN (inhibitor of cytochrome c oxidase, ETC. IV). When we pretreated cells with rotenone followed by BCA treatment, the ROS generation was significantly reduced only in RK3E-ras (Fig. 3B,C). On the other hand, KCN and NAC inhibited the ROS generation in both cells. Interestingly, in RK3E pretreated with rotenone, TTFA and antimycin, there was a marked increase in the ROS level when compared with cells treated with only BCA (Fig. 3C). Intracellular ROS generation in the presence of diphenyleneiodonium (DPI, an inhibitor of NADPH oxidase) or carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazon (CCCP, a mitochondrial uncoupler) were also evaluated, but there was no significant difference in the intracellular ROS levels following BCA treatment in both cells (data not shown).
The effect of specific ETC inhibitors for antiproliferation by BCA. To further determine whether specific inhibitors for the ROS generation can recover BCA-induced cell death, the effects of pretreatment with sub-toxic doses of rotenone, KCN and NAC were investigated using MTT assay. Unexpectedly, rotenone and KCN did not recover the inhibited cell proliferation caused by the BCA treatment, although NAC restored cell proliferation almost completely in both cells (Fig. 4).
Glutathione and MT expression changes by BCA. Because pretreatment with NAC, a precursor of glutathione, efficiently blocked BCA-induced cytotoxicity, we analyzed changes in the total GSH level in these two cells. Cells were treated with 20 μM of BCA and then the intracellular GSH concentration in nM/mg protein was determined. In addition, dose-dependent effects of BCA on GSH levels were evaluated after 24 h in both cells. Only in RK3E did GSH levels increase significantly in concentration- and time-dependent manners (Fig. 5A). However, in RK3E-ras there was no significant increase in GSH concentration at all doses of BCA. Furthermore, intracellular GSH decreased significantly at 3 and 6 h of BCA treatment and then increased gradually to the basal level.
We also analyzed the expression level of MT protein, which is a thiol-rich antioxidant protein, after 24 h of BCA treatment. The MT protein expression increased remarkably in RK3E but markedly decreased in RK3E-ras after BCA treatment (Fig. 5B).
Effect of BSO on BCA-induced cell death. To further demonstrate the important role of GSH in the survival of RK3E, we used BSO, an inhibitor of GSH synthesis, to determine whether this agent can sensitize RK3E against BCA. A single treatment of BSO caused strong GSH depletion (Fig. 6A) and minor cytotoxicity (Fig. 6B) in both cells. However, only RK3E was sensitized to BCA-induced antiproliferation through BSO pretreatment.
Specific recovery of BCA-induced cell death by thiol anti-oxidants. To further clarify the contribution of intracellular ROS and GSH to BCA-induced antiproliferation, various cell-permeable thiol antioxidants (GSH-MEE, NAC, cysteine) and non-thiol antioxidants (pyruvate, BHT) were pretreated in sub-toxic doses before the BCA treatment. BCA-induced cell shrinkage (Fig. 7A) and cell death (Fig. 7B) were prevented by thiol antioxidant pretreatments. Interestingly, cell-permeable non-thiol antioxidants failed to recover BCA-induced cell shrinkage and antiproliferation.
γ-GCS expression increase in RK3E after BCA treatment. To further confirm the differential increase of GSH only in RK3E, we investigated the effect of BCA on protein expression changes in γ-GCS, a rate-limiting step in de novo GSH synthesis. Twenty micro molar of BCA induced a remarkable increase of γ-GCS expression in RK3E (Fig. 8). Under the same experimental conditions, γ-GCS expression was decreased for 6 h and then recovered to a near-basal level in RK3E-ras.
DJ-1 and Nrf2 expression increases in RK3E after BCA treatment. To evaluate the mechanism for differential expression of γ-GCS and MT in these two cells, we investigated DJ-1 and Nrf2 expression changes after 20 μM of BCA treatment. In RK3E, BCA increased DJ-1 expression remarkably after only 30 min with a concurrent increase in Nrf2 expression (Fig. 9) On the contrary, in RK3E-ras, DJ-1 protein expression decreased for 3 h and then reached a near-basal level, and the Nrf2 expression level decreased continuously after 20 μM of BCA treatment.
Many anticancer agents can induce intracellular ROS to kill cancer cells. We have previously shown that BCA-induced apoptosis is completely blocked by pretreatment with GSH or NAC, concluding that intracellular ROS could be an important factor in apoptosis induction.(3) In the present study, we compared ROS production to BCA treatment to determine how BCA causes more effective cell death in K-ras-transformed cells than in normal K-ras cells. An increase in BCA-induced ROS in RK3E occurred after only 30 min and gradually decreased in a time-dependent manner, whereas BCA continuously increased the ROS generation in RK3-ras (Fig. 3A). We determined the effects of specific ETC. inhibitors on the ROS generation. Rotenone and/or KCN have been identified to inhibit the ROS generation caused by BCA in RK3E and RK3E-ras (Fig. 3C). Unexpectedly, rotenone and KCN failed to recover the antiproliferative effect of BCA, and only NAC restored cell proliferation completely in both cells (Fig. 4).
Many drug resistant cells are known to often show very low levels of ROS, usually due to high intracellular GSH levels and enhanced activities in antioxidant enzymes. A recent study showed that the electrophilic nature of cinnamaldehydes makes them potentially reactive with nucleophiles such as sulfhydryl groups in proteins and GSH.(13) Therefore, we measured the intracellular GSH levels upon BCA treatment. Interestingly, RK3E showed a continuous increase in intracellular GSH for 24 h, whereas the intracellular GSH level decreased early at 3 h in RK3E-ras, reaching a basal state at 16 h (Fig. 5A). Furthermore, pretreatment with BSO, an inhibitor of GSH synthesis, sensitized only RK3E to BCA (Fig. 6B). We tried to determine whether oxidative stress induced by hydrogen peroxide could also cause differential GSH depletion only in RK3E-ras. Hydrogen peroxide (1–100 mM) did cause oxidative stress, inducing concentration-dependent GSH depletion but in both cell lines (data not shown), suggesting that selective GSH depletion only in K-ras-transformed cells is specific to BCA.
A recent study showing BCA as an effective antitumor agent against hepatocellular carcinoma induced by the oncogenic H-ras12V gene,(14) confirmed that the MT gene was more than halved by BCA treatment. Metallothionein, which contains the highest amount of thiol groups within the cytoplasm, was initially identified as an antioxidant. Mechanisms underlying its function may include the maintenance of GSH levels.(15) Furthermore, MT thiol groups are able to bind several cytotoxic agents.(16) Comparing MT expression in both cells upon BCA treatment, we found the basal level of MT expression in RK3E to be lower than in RK3E-ras, while MT expression of RK3E exceeded that of RK3E-ras upon BCA treatment for 24 h (Fig. 5B).
These results led us to speculate that GSH and MT decreases in RK3E-ras can increase intracellular ROS, but the increased ROS is not critical for BCA-induced antiproliferation in RK3E-ras. To further verify the contribution of the intracellular ROS generation and GSH depletion to BCA-induced apoptosis, various thiol- and non-thiol antioxidants were pretreated with BCA. Thiol-antioxidants completely blocked apoptosis in both cells (Fig. 7). However, pretreatment with other non-thiol antioxidants were ineffective in preventing cell death. In addition to BCA’s differential antiproliferation in K-ras-transformed cells through intracellular GSH depletion, these data support the theory that intracellular thiol status and redox balance could be crucial in BCA-induced cell death. Many cancer therapeutic drugs function by depleting the intracellular thiol buffer system and thus disrupting the redox balance.(17,18) What is exciting about these studies is that GSH depletion occurs only in tumor cells enhancing antitumor cytotoxicity without increasing toxicity to normal tissue.(19) It is well known that intracellular redox status plays an important role in apoptosis, and as the most abundant non-protein antioxidant in the cell, intracellular GSH depletion often occurs at the onset of apoptosis.(20)
From the results of the present study, it can be suggested that the intracellular status of GSH affects toxicity of BCA in RK3E-ras. That is, increased levels of thiol antioxidants such as GSH and MT in RK3E are thus one mechanism of its resistance to BCA, as cytoplasmic binding of these antioxidants with BCA prevents the active molecules from reaching their target. However, as shown in the chromatogram of BCA mixed with GSH in vitro,(13) not all BCA compounds have been changed to BCA-GSH adduct even though there are enough GSH molecules for a reaction. These data partially resolve the inconsistency between intracellular GSH and the cell viability of RK3E after BCA treatment (Fig. 6). Furthermore, more specific mechanisms for therapeutic selectivity could be deduced from the activation process of ras by farnesylation of their cys residues. It is known that ras protein is farnesylated on the carboxy-terminal cysteine for activation, and that HCA inhibits farnesyl transferase in an enzymatic assay.(2) Therefore, further studies are needed to uncover whether BCA can inhibit ras activation via its reactivity with the sulfhydryl groups of the cys residues.
Endogenous GSH is synthesized from its constituent amino acids and γ-GCS is the rate-limiting step in de novo GSH synthesis. While anticancer drugs such as cisplatin and doxorubicin have been widely used to treat various tumors, Yao et al.(21) found γ-GCS expression to be elevated in cisplatin-resistant cancer cells. In the present study, BCA increased the γ-GCS protein level only in RK3E, with a remarkable decrease in RK3E-ras (Fig. 8). Therefore, the differential γ-GCS expression level between these two cells might be a key reason for why BCA causes significantly less cell death in RK3E.
Intracellular GSH upregulation could be via the DJ-1 protein, which improves cell survival from oxidative damage.(11) Interestingly, DJ-1 showed a cooperative transforming activity with H-ras in NIH3T3 cells.(22) Although the exact physiological role of DJ-1 is presently unclear, DJ-1 may mediate its antioxidant effects by Nrf2 pathway activation, which induces its target genes such as γ-GCS and MT.(23,24) In the present study, BCA increased DJ-1 and Nrf2 expression after only 30 min in RK3E (Fig. 9), possibly rendering upregulation of intracellular GSH via increased γ-GCS and MT expression. On the contrary, BCA decreased the expression of these proteins in RK3E-ras.
One of our further studies would be to confirm the differential effect of BCA in other normal/K-ras-mutated cells. We analyzed this effect in MCF-7 (normal K-ras) and MDA-MB-231 (K-ras-mutant) breast cancer cells. As expected and confirming our current results, BCA treatment induced stronger antiproliferation in MDA-MB-231 (IC50 28 μM) than in MCF-7 cells (IC50 > 60 μM) (data not shown).
Using isogenic cell lines, we showed that K-ras-transformed cells are more sensitive to BCA, which selectively decreases GSH and MT expression via downregulation of DJ-1. Increased levels of thiol antioxidants in RK3E are thus one mechanism of resistance to BCA (Fig. 10). More studies are necessary to directly evaluate in vivo BCA-GSH adduct formation in both cells. Furthermore, the differential regulation of DJ-1 expression and the role of DJ-1 in inactivating K-ras-mediated signals upon BCA treatment need to be elucidated.
This research was supported by the Basic Science Research Program through the National Research Foundation (NRF) of Korea, funded by the Ministry of Education, Science and Technology (KRF-2008-E00027) (S.H.H.), and a grant from the Center for Biological Modulators of the 21st Century Frontier Research Program (CBM1-B500-001-1-00) (S.H.H., B.M.K.).