ADP-Ribose and oxidative stress activate TRPM8 channel in prostate cancer and kidney cells

Activation of TRPM8 channel through oxidative stress may induce Ca2+ and pro-apoptotic signals in prostate cancer and kidney cells. The aim of this study was to evaluate activation of TRPM8 can increase apoptosis and oxidative stress in the prostate cancer (Du145M8), TRPM8 knock out (Du 145M8KO), transfected (HEK293TM8) and non-transfected human kidney (HEK293) cells. Intracellular Ca2+ responses to TRPM8 activation were increased in the Du145M8 and HEK293TM8 cells from coming cumene hydrogen peroxide (CHPx), menthol, ADP-Ribose (ADPR), but not in the HEK293 and Du 145M8KO cells. The intracellular Ca2+ responses to both ADPR and CHPx were totally inhibited by the thiol cycle antioxidant glutathione, and TRPM8 blockers (N-(p-amylcinnamoyl)anthranilic acid and capsazepine). Apoptosis, Annexin V, mitochondrial membrane depolarization, intracellular ROS, caspase 3 and 9 values were increased through TRPM8 activation in the Du 145M8 but not in the Du 145M8KO and non-transfected HEK293 cells by CHPx and hydrogen peroxide. In conclusion, apoptotic and oxidant effects on the cells were increased activation of TRPM8 by oxidative stress and ADPR. Activation of TRPM8 through oxidative stress and ADPR in the cells could be used as an effective strategy in the treatment of prostate cancer cells.

Oxidative stress occurs during the physiological functions such as phagocyte activity and mitochondrial function. The oxidative stress is controlled by the antioxidants such as glutathione (GSH) and glutathione peroxidase (GSH-Px). GSH as a member of thiol cycle antioxidants endogenously synthesized all mammalian cells and it has several physiological functions such as antioxidant defense, inhibition of prostate cancer and transport of cysteine 1,2 . GSH and N acetyl cysteine (NAC) treatments as a member of thiol redox system, induced transient receptor (TRP) melastatin 2 (TRPM2) and 8 (TRPM8) channel inhibitor roles [3][4][5][6] . ADP-Ribose (ADPR) is synthesized in the nucleus beta nicotinamide adenine dinucleotide by activation CD38 enzyme through hydrogen peroxide (H 2 O 2 ) production 7,8 . The H 2 O 2 has been using for investigation of oxidative stress dependent TRP channel activations such as TRPM2 and TRPV1 [7][8][9] . The TRPM8 channel is activated by cold and menthol 10,11 . However, there is no report ADPR and H 2 O 2 dependent activation of TRPM8 in the prostate cancer and human embryonic kidney cells 293 (HEK293) cells.
Intracellular free calcium ion ([Ca 2+ ] i ) concentration is a major intracellular second messenger factor that regulates many physiological and pathophysiological functions including cell migration 12,13 . Apoptosis, proliferation, differentiation and migration in cells are controlled by the Ca 2+ signaling pathways. Prostate cancers are a most common diagnosis in men. It is also well known that an increase of [Ca 2+ ] i concentration involved in prostate cancer carcinogenesis and in metastasis development 14 . The Ca 2+ passes the cell membranes through different cation channels including TRP channels. As a member of the TRP superfamily, TRPM8 channel, changes in its expression level is involved in the etiology of prostate cancers and it seems to be one of the most promising potential drug target channels in the treatment of prostate cancers 15 . Androgen-dependent expression of TRPM8 increases in both benign prostate hyperplasia and in prostate carcinoma cells 15,16 . Involvement of transmembrane domains-isoforms of TRPM8 in the mitochondria of keratinocyte cells for the regulating [Ca 2+ ] i concentration was recently reported 17 . In addition, an increase of [Ca 2+ ] i concentration through menthol activation of TRPM8 channels in the prostate cancer cells induced increase the rate of mitochondrial oxidative stress, resulting Scientific RepoRts | (2019) 9:4100 | https://doi.org/10.1038/s41598-018-37552-0 apoptosis of the cancer cells 18 . Hence, activation of TRPM8 through oxidative stress may induce pro-apoptotic signals in prostate cancer cells, but it remains unclear. To our knowledge, there is no report on the oxidative stress and ADPR dependent activation of TRPM8 channels in TRPM8 positive androgen insensitive prostate cancer (Du 145 M8 ) and overexpressing human TRPM2 channel HEK293 (HEK293 TM8 ) cells. Therefore, we propose that investigation of the involvement of oxidative stress in the TRPM8 activation might represent two of the mechanisms controlling up-regulation of mitochondrial oxidative stress, apoptosis and [Ca 2+ ] i concentration in the Du 145 M8 and HEK293 TM8 cells.

Results
Oxidative stress activates TRPM8 in the Du 145 M8 cells. As the first step in the current study whether activation of TRPM8 channel is related to oxidative stress (cumene hyroperoxide, CHPx) activator and menthol, the influences of the channel on Ca 2+ fluorescence intensity in the Du 145 cells were investigated by using the activators and inhibitors (thiol cycle antioxidant GSH and TRPM8 channel blocker [N-(p-amylcinnamoyl) anthranilic acid (ACA)]. The confocal microscope images (Fig. 1a) and columns (Fig. 1b) of Ca 2+ fluorescence intensity in Du 145 M8 are presented in Fig. 1. The Ca 2+ fluorescence intensity was increased in the cells by CHPx stimulations. On the other word, the Ca 2+ fluorescence intensity was significantly (p ≤ 0.001) higher in the control + CHPx groups as compared to control. However, the Ca 2+ fluorescence intensity was markedly (p ≤ 0.001) decreased in the control + CHPx + ACA group as compared to the CHPx group by the ACA treatment. This increase in Ca 2+ fluorescence intensity was totally prevented by pretreatment with GSH and the Ca 2+ fluorescence intensity was markedly (p ≤ 0.001) lower in the GSH, GSH + CHPx and GSH + CHPx + ACA than in the control + CHPx and control + CHPx + ACA groups.
Oxidative stress has no TRPM8 activation in the absence of TRPM8 and extracellular Ca 2+ in the Du 145 M8 and Du 145 M8KO cells. After observation of oxidative stress dependent activation of TRPM8 Figure 1. Activation of TRPM8 in the Du 145 M8 cells by oxidative stress. (mean ± SD). The cells were stained with Fluo-3 calcium dye and mean ± SD of fluorescence in 15 mm 2 of cell as arbitrary unit are presented; n = 10-20 independent experiments. In GSH experiments, the cells were pretreated with GSH (10 mM for 2 hours). The cells were extracellularly stimulated by cumene hyroperoxide (CHPx and 1 mM for 5 min) but they were extracellularly inhibited by ACA (25 μM for 10 min). The samples were analyzed by the laser confocal microscopy fitted with a 40× oil objective. The scale bar was 20 µm. Representative images and fluorescence intensities of the CHPx, ACA and GSH effect on the TRPM8 activation in the laser confocal microscope analyses are shown in (a,b) respectively. ( a p ≤ 0.001 versus control. b p ≤ 0.001 versus control + CHPx group. c p ≤ 0.001 versus control + CHPx + ACA group).  18 , showed no detectable TRPM8 response-induced Ca 2+ fluorescence intensity (Fig. 2a,b). Addition of CHPx and menthol in the presence of Ca 2+ led to a significant increase in the Ca 2+ fluorescence intensity in the Du 145 M8KO cells, which was decreased by the addition of ACA, the TRPM2 channel specific inhibitor (Fig. 2a,b). In contrast, CHPx and menthol treatments induced no increase in the Ca 2+ fluorescence intensity level in the absence of Ca 2+ (Fig. 2a,b). Furthermore, the Ca 2+ fluorescence intensity increases were not observed in the absence of TRPM8 in the Du 145 M8KO cells. These results exclude the Ca 2+ release from intracellular organelles such as the endoplasmic reticulum and mitochondria and more importantly, for the first time, demonstrate the existence of a specific mechanism for Ca 2+ influx involving TRPM2 channels.

TRPM2 blocker (ACA) inhibits the ADPR -induced TRPM8 currents in the Du 145 M8KO cells.
ADPR is synthesized in the nucleus beta nicotinamide adenine dinucleotide by activation CD38 enzyme through extracellular H 2 O 2 production 19,20 . As a member of TRP superfamily, TRPM2 channel is activated by ADPR 19,20 but there is no report on the ADPR-induced TRPM8 in cells. Therefore, we firstly tested involvement of ADPR on  Fig. 1a the TRPM8 activation in the Du 145 cells. TRPM8 channel in the patch-clamp experiments was gated in the Du 145 M8 cells by ADPR (1 mM in the patch-pipette), although they were reversibly blocked by ACA and NMDG + (replacement of Na + ) (Fig. 3b). There were no currents in the absence of the TRPM8 agonists (ADPR, CHPx and menthol) and antagonists (and ACA) (Fig. 3a). Treatment of wild type (Du 145 M8 ) cells with the 25 μM ACA as a TRPM2 channel inhibitor, strongly suppressed ADPR-induced current densities (Fig. 3b,c). On the other word, the current densities in the cells were significantly higher in the control + ADPR group compared with the control group (p ≤ 0.001); however, the current density of TRPM8 was significantly (p ≤ 0.001) lower in the control + ADPR + ACA group than in the control + ADPR group (Fig. 3b,f). The H 2 O 2 has been using for investigation of oxidative stress dependent TRP channel activation such as TRPM2 and TRPV1 7,9 . To further investigate the relative contribution of oxidative stress in the TRPM8 activation, the effect of CHPx was studied in the TRPM8 present (Du 145 M8 ) and knockout (Du 145 M8KO ) prostate cancer cells (Fig. 3c-e). In addition, we used specific agonist of TRPM8 (menthol) as positive control records. The current densities in the neurons were increased in CHPx and menthol groups (Fig. 3g), and they were decreased in the CHPx + ACA and menthol + ACA groups by the ACA treatments (p ≤ 0.001) (Fig. 3g). Hence, these effects of CHPx and menthol were partially abolished by ACA.
In patch clamp experiment, we also tested the role of antioxidant GSH and deletion of TRPM8 on the TRPM8 activation in the Du 145 M8 cells. The menthol and CHPx-induced currents were completely blocked in the presence of intracellular GSH (2 mM in the patch pipette) (Fig. 3d) and deletion of TRPM8 (Fig. 3e). The current densities were markedly (p ≤ 0.001) lower in the Du 145 M8 + GSH, Du 145 GSH + menthol, Du 145 M8 + GSH + CHPx and Du 145 M8KO groups than in the Du 145 M8KO + menthol and Du 145 M8KO + CHPx groups.
These results clearly indicated that oxidative stress induced excessive Ca 2+ influx through the TRPM8 channel. However, the oxidative stress-induced TRPM8 currents through ROS production modulation were decreased by treatment with the antioxidant (GSH).

Hydrogen peroxide induces TRPM8-dependent increase of Ca 2+ fluorescence intensity in the HEK293 cells overexpressing human TRPM8 channel (HEK293 TM8 ) cells. After observation of
oxidative stress dependent activation of TRPM8 in the cells, we tested effects of oxidative stress (H 2 O 2 ) on the fluorescence intensity (Figs 4 and 5) through TRPM8 activation in the HEK293 TM8 cells. It is well known that TRPM2 channel is activated within 2-5 minutes in different cell lines by oxidative stress and ADPR 19,20 . Similarly, we observed activation of TRPM8 in HEK293 TM8 cell within 2-5 minutes by ADPR and H 2 O 2 (Fig. 4b). There was no increase in the fluorescence intensity of Ca 2+ in the control HEK293 TM8 cells within 6 minutes (Fig. 4b). The intensity was markedly (p ≤ 0.001) increased in the cell by H 2 O 2 (Fig. 4b,c), although it was decreased TRPM2 channel blocker (ACA). However, HEK293 cells, which do not express TRPM8 channels, showed no detectable TRPM8 response-induced Ca 2+ fluorescence intensity through activation of TRPM8 by the H 2 O 2 stimulation (Fig. 5a,b) and the fluorescence intensity levels did not change in the control, H 2 O 2 and ACA groups, statistically. It is well known that several TRP channels such as TRPM2 and TRPM7 can be activated by oxidative stress 12 . These results in the TRPM8 expressing the HEK293 TM8 cells exclude involvement of oxidative stress dependent activated other TRP channels and more importantly, for the first time, demonstrate the existence of a specific mechanism for oxidative stress-induced Ca 2+ influx involving TRPM8 channels.  (Fig. 6a,b) current density (Fig. 6c,f) through activation of TRPM8 by the H 2 O 2 and ADPR stimulations. Induction of TRPM2 expression using a transfection system, however resulted in decrease ADPR and oxidative stress-sensitive [Ca 2+ ] i concentration (Fig. a,b) and current density (Fig. 6e,f) through ACA treatment. In addition, we observed ADPR dependent activation in the single channel (inside out) patch clamp records (Fig. 6g). However there was no the single channel currents in the absence of ADPR (Fig. 6h). The single channel results exclude the involvement of second messengers for the activation of TRPM8 via oxidative stress and ADPR. On the other word, it is more importantly, for the first time, demonstrate the existence of a specific mechanism as a TRPM2 channel for Ca 2+ involving TRPM8 channels.

Discussion
In the current study, we found that oxidative stress and ADPR treatments could induce the TRPM8 activations resulting in the overload Ca 2+ entry, apoptosis, and mitochondrial oxidative stress. More importantly, we found that GSH could protect the Du 145 M8 prostate cancer cells from oxidative stress-induced apoptosis via maintaining the intracellular Ca 2+ homeostasis as well as down-regulating mitochondrial oxidative stress pathway. The major findings of this study are that TRPM8 channel is separately activated in the prostate cancer cells by ADPR and oxidative stress and its sensitivity enhance to ROS. There is debating evidence obtained from the prostate cancer and human kidney cells, that TRPM8 channel activation is associated with production of oxidative stress 17,18,21 . Indeed, H 2 O 2 stimulation induced functional changes on the TRPM8 in the urothelium cell of elderly subject and human lung epithelial cells, although the changes were reduced by NAC treatments 20 . However, conflicting report is also presented on the subject and the TRPM8 channel was not activated in urothelium bladder cells by 1 mM H 2 O 2 21 . In general, induction of oxidative stress as a mechanism that may contribute to the antitumor induction effect has been gaining acceptance 15 . Most of chemotherapeutic agents induce excessive ROS production for killing the cancer cells 14 . It is well known that an increase in [Ca 2+ ] i concentrations through activation of TRP channels such as TRPM2 and TRPV1 induces an increase of intracellular mitochondrial ROS production 22,23 . However, GSH as a member of thiol cycle antioxidants has been shown to inhibit CHPx-evoked increased in cell viability and decreases in intracellular levels of ROS and apoptosis 13,14,23 . GSH has been also reported to prevent completely ADPR and CHPx-evoked TRPM2 and TRPV1 channel activations 13,14,23 . Thus, the pro-apoptotic effects of oxidative stress in the cancer cells, including prostate cancer cells seem to be dependent on one single mechanism, e.g., the ability of TRPM8 activation to generate oxidative stress. We have recently identified the primary role of menthol dependent, but not oxidative stress TRPM8 activation in the Du 145 cells 18 . GSH and NAC treatments as two members of thiol redox system, induced TRPM2 and TRPM8 channel inhibitor roles through inhibition of oxidative stress in different cell lines [3][4][5][6] . Of interest for the present discussion is the finding that ADPR and CHPx-evoked TRPM8 currents were completely abated by intracellular GSH treatment. These findings imply that oxidative stress directly gates TRPM8, but rather probably exerts this action indirectly via the generation ADPR in DNA damage of nucleus by oxidative stress byproducts that eventually target the channel in the prostate cancer cells, through the direct formation of intracellular ROS 4 .
In the current study, we observed increased levels of apoptosis, caspase 3, caspase 9, mitochondrial membrane depolarization and ROS values through activation of TRPM8 channel in the Du 145 M8 cells, but not Du Du 145 M8KO cells by CHPx and ADPR, although the values were decreased in the cells by the GSH treatment. During the treatment of tumor cells including prostate cancer cells, increase of mitochondrial oxidative stress through activation of TRPM8 channels and mitochondrial dysfunction has been suggested to account in cancer cells the induction of apoptosis 24,25 . Mitochondrial oxidative stress and apoptosis in human epithelial prostate cancer cells were induced by suppression of TRPM8 isoforms 17 , through alterations in mitochondrial membrane depolarization and ATP production 26 , which leads to oxidative phosphorylation through the electron transport chain and hence the formation of JC1 27 . Thus, induction of apoptosis through overload Ca 2+ entry by oxidative stress probably lead to the increase of this toxic protein aggregates inhibiting cancer cell survival. It has been reported that Ca 2+ entered from the cytosol during mitochondrial stress accumulates in the mitochondria and mediates the excessive apoptosis through activation of caspase 3 and 9 18 . ROS generation activates both survival and death signaling, depending upon the intensity of the production process. In turn, TRPM8 activation is increased by the increase of mitochondrial ROS production and then the prostate cancer cells are killed by the TRPM8 channel-induced overproduction of intracellular ROS, apoptosis and Ca 2+ entry.
As a sulfur containing substance, GSH is containing sulfur groups and it is a member of thiol cycles 2,28 . Oxidation of thiol redox system and cysteine groups in cancer cells have the main role in the activation of thiol group containing TRP channels such as TRPA1, TRPM8 and TRPV1. Intracellular cysteine suppression reduced tumor growth in prostate cancer cells 1,29 . In the current study, the GSH treatment inhibited the oxidative stress and ADPR-induced TRPM8 activation through supporting the thiol cycle antioxidants such as GSH and GSH-Px in the cell line. Similarly, the protective role of GSH treatment on the oxaliplatin-induced TRPA1 activation in mouse dorsal root ganglion (DRG) neurons was reported by Materazzi et al. 29 . In addition, it was recently reported that redox-sensitive TRPV1, TRPC1, TRPM2, and TRPM7 channels are inhibited in human hepatoma cell line 30 and rat DRG neurons 5,24 by GSH and N acetyl cysteine.
In conclusion, our data clearly show that oxidative stress and ADPR stimulus increased TRPM8-mediated responses, including an increase of intracellular Ca 2+ and mitochondrial ROS sensitive-apoptosis in the Du 145 M8 and HEK293 TM8 cells. However, these responses were attenuated by the treatment with the ROS scavenger GSH and TRPM8 blockers (ACA and CPZ). All together, these data support the hypothesis that oxidative stress is able to induce functional changes in the prostate cancer cell TRPM8 channel signaling and suggest that the killing the prostate cancer cells is susceptible to oxidative stress, with possible implications for treatment of prostate cancer.

Methods
Cell lines. Human prostate (Du 145 M8 ) cancer cells were purchased from ATCC (Manassas, VA, USA), although HEK293 cells were obtained from the Şap Institute of Agriculture and Animal Ministry of Turkey (Ankara, Turkey). The cells were cultured in a medium consisting of 90% Dulbecco's modified Eagle's medium (DMEM, Invitrogen, Istanbul, Turkey), 10% fetal bovine serum (FBS, Gibco, Istanbul, Turkey), and 100 μg/ml streptomycin + penicillin (100 U/ml) combination (Biochrom, Berlin, Germany) and the appropriate supplements, including 100 μg/ml sodium pyruvate (Sigma-Aldrich, Istanbul, Turkey) as suggested by the supplier in a humidified atmosphere in 5% CO 2 at 37 °C. The cells were tested within 24 hours after plating onto the coverslips. The cells were tested within 24 hours after plating onto the coverslips. Then the cells were counted by using an automatic cell counter (Casy Modell TT, Roche, Germany). In plate reader and patch-clamp analyses, the cells were seeded in 6 flasks at a density of 1 × 10 6 cells per flask (filter cap, sterile, 260 ml, 80 cm²) (Thermo Fischer Sci. Inc., Istanbul Turkey). In confocal microscope analyses, the cells were seeded in 35 mm glass bottom dishes (Mattek Corporation Inc., Ashland, MA, USA).

Transfection of HEK293. Transient transfections of HEK293 cells with the 2 μg cDNAs of human TRPM8
(hTRPM8 and a gift from Dr. Simon Hebeisen, B'SYS GmbH, Witterswil Switzerland) were performed according to the manufacturer's instructions (B'SYS GmbH). For control experiments, 2 μg of wild type TRPM8 empty vector hTRPM8 (C-terminal FLAG tag) plasmid (OriGene Technologies, Istanbul, Turkey) was used for 24 hours using Lipofectamine 2000 (Invitrogen; Istanbul, Turkey. The transfected HEK293 cells (HEK293 TM8 ) seeded on glass coverslips at a suitable dilution and were maintained for 24 h in an incubator at 37 °C and 5% CO 2 . Then, patch-clamp, Fura-2 and laser confocal microscope experiments were carried out with cells visibly positive for EGFP. ] i concentration is expressed as nanomolar (nM) taking a sample every second as previously described 23 .
For imaging Du 145 M8 and Du 145 M8KO cells, the cells were analyzed by using Ca 2+ indicator florescent dye (Fluo-3, Calbiochem, Darmstadt, Germany) in the dark. The Fluo-3 is a single wavelength excitation and emission dye that excited by a 488 nm argon laser from the confocal microscope 32 . The cells were treated with TRPM8 antagonist (ACA and 25 μM) to inhibit Ca 2+ entry before stimulation of TRPM8 (CHPx and 1 mM). Fluorescence emission of the cells was inspected with a plan Apo 40×/0.2 immersion objective on a confocal microscope (LSM 800, Zeiss, Ankara, Turkey) at 515 nm. Intracellular fluorescence intensities of 10 cells were analyzed in the confocal microscope before CHPx stimulations by ZEN program. Ca 2+ concentration (1.2 mM) and content of the extracellular buffer were described in a previous study 11 . Results of a recent study expressed the importance of TRPM8 on the Ca 2+ release from intracellular organelles in the prostate cancer cells 33 . For the clarifying importance of Ca 2+ release from the intracellular organelles through TRPM8 activation we used calcium-free extracellular buffer. In the experiments where calcium-free medium was required, Ca 2+ was omitted and 2 mM of the chelator EGTA was added.
Manufacturers and preparations of the ADPR, CPZ, menthol, and ACA were described in the previous studies 11,18,23 . The CHPx were dissolved in the extracellular buffer with and without Ca 2+ (1.2 mM).
Electrophysiology. Whole-cell voltage clamp recording was taken from the Du 145 M8 , Du 145 M8KO HEK293 TM8 and HEK293 cells (EPC10 patch-clamp set, HEKA, Lamprecht, Germany). We used standard extracellular bath and pipette solutions as described in previous studies 11,18,23 . Holding potential of the patch-clamp analyses in the cell was −60 mV. The current-voltage (I-V) relationships were obtained from voltage ramps from −150 to +150 mV applied over 200 milliseconds. All experiments were performed at room temperature (22 ± 2 °C).
We also performed single cell record experiments in the HEK293 TM8 cell as described in a previous study 19 .
In the whole cell and single cell experiments, TRPM8 was intracellularly gated by ADPR (1 mM), and the channels were extracellularly blocked by ACA (25 μM). In recent studies, we observed inhibitory role of intracellular GSH (2 mM) on the oxidative stress dependent activations of TRPM2 and TRPV1 channels 11,14,23 . Hence, the TRPM8 channels in some path-clamp experiments were treated with the intracellular GSH. The maximal current amplitudes (pA) in the Du 145 and HEK293 cells were divided by the cell capacitance (pF), a measure of the cell surface. Values of current density were expressed as pA/pF in the patch-clamp experiments.
Assay of cell viability. Cells were plated in 48-well plates, incubated after treatment with CHPx (1 mM) and ACA (25 μM). Number of viable cell was determined using the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl tetrazolium bromide colorimetric (MTT) colorimetric assay as described previously 34 . Absorbance in the spectrophotometer (UV-1800) was read at 570 nm. A total of 3 experiments (n = 3) was performed for the cell viability assay. The data are presented as fold-increase over the pretreatment level.
Assay of apoptosis, caspase 3 and 9 activities. For the apoptosis spectrophotometric analysis apoptosis, we used a commercial kit and the analyses were performed according to the instructions provided by Biocolor Ltd. (Northern Ireland) and elsewhere 34 .
The determinations of caspase 3 and 9 activities were based on a method previously reported 35,36 with minor modifications. Caspase 3 (N-acetyl-Asp-Glu-Val-Asp-7-amido-4-methylcoumarin) and 9 (N-acetyl-Leu-Glu-His-Asp-7-amino-4-methylcoumarin) substrates were purchased from Bachem (Bubendorf, Switzerland) and cleavages of the substrates were measured with a microplate reader (Infinite pro200; Tecan Austria GmbH, Groedig, Austria) with excitation wavelength of 360 nm and emission at 460 nm. The data were calculated as fluorescence units/mg protein and presented as fold-increase over the pretreatment level (experimental/control). A total of 3 experiments were performed for the caspase and apoptosis assays.
Detection of intracellular reactive oxygen species (ROS) level. Dihydrorhodamine-123 (DHR 123) as a non-fluorescent and non-charged dye can easily diffuse across membranes 34 USA) at 37 °C in the dark for 30 min. The fluorescence intensity of the oxidized product (Rh123) was measured in the microplate reader (Infinite Pro200). Excitation and emission wavelengths were 488 and 543 nm, respectively 32 . The data are presented as fold-increase over the pretreatment level.
In imaging the ROS production in HEK293 TM8 and HEK293 cells, the intracellular oxidative stress was monitored by DHR123 (514 nm excitation, 570 emission) 36 . After exposed to indicated treatments, they were incubated in culture medium containing 1 μM DHR123 for 30 min at 37 °C in the dark. Cells were washed and maintained with the phosphate buffer before images were captured using a ZEN Program Imaging System. Fluorescence intensity in 15 μm 2 of each cell as arbitrary unit was measured by using ZEN program and analyzed using Image J/Imaris software. The results of JC1 and DHR123 were expressed as the mean fluorescence intensity as arbitrary unit /cell. Measurement of mitochondrial membrane potential (ΔΨm). 5,5′,6,6′-Tetrachloro-1,1′,3,3′tetraethylbenzimidazolylcarbocyanine iodide (JC1, Molecular Probes, Eugene, OR, USA) floresecen dye has been using for measurement of ΔΨm level 36 . Hence, we used the dye in the current study for measurement of ΔΨm level. The green (excitation; 485 nm and emission; 535 nm) and red (excitation; 540 nm and emission; 590 nm) JC1 signals were measured in the cell line as described in a previous study 34 . Fluorescence changes were analyzed using the microplate reader (Infinite Pro200). The data are presented as the fold-increase over the pretreatment level.
In imaging of mitochondrial membrane depolarization, the HEK293 TM8 cells were re-suspended in 0.2 ml of phosphate buffer with calcium and then incubated with JC1 (5 μl) dye solutions for 30 min at 37 °C in the dark. The samples were then analyzed by the laser confocal microscopy. JC1 (505 nm excitation, 535 emission) was excited with a diode laser at 488 nm, an Argon laser at 488 nm 36 . Fluorescence intensity in 15 μm 2 of each cell as arbitrary unit was measured by using ZEN program and analyzed using Image J/Imaris software. The results of JC1 were expressed as the mean fluorescence intensity as arbitrary unit/cell. Annexin V-FITC assay by laser confocal microscope. The protective effects of DTX-induced apoptosis were determined by the laser confocal microscope (LSM-800) using the Annexin V (FITC) dye as described in the manufacturer's guidelines (Santa Cruz). Briefly, the Annexin V apoptosis detection Kit utilizes FITC-conjugated Annexin V protein for detection of cells undergoing apoptosis. Annexin V FITC binds to the membranes of apoptotic cells, displaying a green characteristic staining pattern which was viewed by the laser confocal microscope (LSM-800).
At the end of the H 2 O 2 treatment, the HEK293 TM8 cells were washed twice with phosphate-buffered saline. The cells were re-suspended in 0.2 ml of extracellular buffer and then loaded with Annexin V-FITC (1 μl) for 15 min at room temperature in dark. The samples were then analyzed by the laser confocal microscopy fitted with a 40× oil objective. The fluorescence intensity of each cell as arbitrary unit was measured by using ZEN program and analyzed using Image J/Imaris software. The results of Annexin V-FITC were expressed as the mean fluorescence intensity as arbitrary unit /cell.

Statistical analyses.
All data were represented as means ± standard deviation (SD). Statistical analysis was performed with SPSS Version 18.0 statistic software package (Chicago, Illinois, USA). P value as ≤0.05 was considered to indicate a statistically significant. Presence of significance was once detected by LSD test. Then, comparisons between groups for finding levels of p values were performed with analysis of non-parametric Mann Whitney U test.