Melatonin modulates red-ox state and decreases viability of rat pancreatic stellate cells

In this work we have studied the effects of pharmacological concentrations of melatonin (1 µM–1 mM) on pancreatic stellate cells (PSC). Cell viability was analyzed by AlamarBlue test. Production of reactive oxygen species (ROS) was monitored following CM-H2DCFDA and MitoSOX Red-derived fluorescence. Total protein carbonyls and lipid peroxidation were analyzed by HPLC and spectrophotometric methods respectively. Mitochondrial membrane potential (ψm) was monitored by TMRM-derived fluorescence. Reduced (GSH) and oxidized (GSSG) levels of glutathione were determined by fluorescence techniques. Quantitative reverse transcription-polymerase chain reaction was employed to detect the expression of Nrf2-regulated antioxidant enzymes. Determination of SOD activity and total antioxidant capacity (TAC) were carried out by colorimetric methods, whereas expression of SOD was analyzed by Western blotting and RT-qPCR. The results show that melatonin decreased PSC viability in a concentration-dependent manner. Melatonin evoked a concentration-dependent increase in ROS production in the mitochondria and in the cytosol. Oxidation of proteins was detected in the presence of melatonin, whereas lipids oxidation was not observed. Depolarization of ψm was noted with 1 mM melatonin. A decrease in the GSH/GSSG ratio was observed, that depended on the concentration of melatonin used. A concentration-dependent increase in the expression of the antioxidant enzymes catalytic subunit of glutamate-cysteine ligase, catalase, NAD(P)H-quinone oxidoreductase 1 and heme oxygenase-1 was detected in cells incubated with melatonin. Finally, decreases in the expression and in the activity of superoxide dismutase were observed. We conclude that pharmacological concentrations melatonin modify the redox state of PSC, which might decrease cellular viability.

www.nature.com/scientificreports www.nature.com/scientificreports/ Detection of protein Carbonyls (Allysine). Cells were incubated during 1 h with stimuli and, thereafter, were lysed for analysis. Detection of protein carbonyls was performed according to the methods described by Villaverde et al. 21 . In brief, five hundred µL of each sample were treated with cold 10% trichloroacetic acid (TCA) solution. After centrifugation (600 × g for 5 min at 4 °C) the supernatants were removed and the pellets were sequentially incubated with a solution containing 0.5 mL 250 mM 2-(N-morpholino) ethanesulfonic acid (MES) buffer pH 6.0 containing 1 mM diethylenetriaminepentaacetic acid (DTPA), a solution containing 0.5 mL 50 mM ABA in 250 mM MES buffer pH 6.0 and a solution containing 0.25 mL 100 mM NaBH 3 CN in 250 mM MES buffer pH 6.0. Next, samples were treated with a cold 50% TCA solution and centrifuged at (1200 × g for 10 min). The pellets were then washed twice with 10% TCA and diethyl ether-ethanol (1:1). Finally, the pellet was treated with 6 M HCl and kept in an oven at 110 °C for 18 h until completion of hydrolysis. Thereafter, the samples were dried in vacuo and the generated residue was reconstituted with 200 µL of milliQ water and filtered for HPLC analysis using a Shimadzu 'Prominence' HPLC apparatus (Shimadzu Corporation, Japan). The elutes were monitored with excitation and emission wavelengths set at 283 and 350 nm, respectively. Standards (0.1 μL) were run and analysed under the same conditions. The nmol of allysine per mg of protein were calculated. Results are expressed as percentage ± SEM (n) with respect to non-stimulated cells, where n is the number of independent experiments.
Analysis of thiobarbituric-reactive substances. Cells were incubated during 1 h with stimuli and, thereafter, were lysed for analysis. Malondialdehyde (MDA) and other thiobarbituric-reactive substances (TBARS) were measured, by adding 500 µL thiobarbituric acid (0.02 M) and 500 µL trichloroacetic acid (10%) to 200 µL of a sample from each treatment. Next, the mixture was incubated for 20 min at 90 °C. After cooling, a 5 min centrifugation at 600 × g was made and the absorbance of supernatant was measured at 532 nm employing a plate reader (VariosKan Lux 3020-205, Thermo Sci., Vantaa, Finland). The mg/L of TBARS in each sample were calculated. Results are expressed as percentage ± SEM (n) with respect to non-stimulated cells, where n is the number of independent experiments.
Determination of mitochondrial membrane potential. Changes in mitochondrial membrane potential (ψ m ) were recorded using the dye TMRM as described previously 22 . Cells were incubated during 1 h in the presence of stimuli. A decrease in TMRM fluorescence reflects depolarization of ψ m . Fluorescence was measured employing a spectrofluorimeter (VariosKan Lux 3020-205, Thermo Sci., Vantaa, Finland). The experiments were carried out employing batches of cells obtained from different preparations. The increase of fluorescence with respect to non-stimulated cells was calculated and expressed in percentage as the mean ± SEM (n) (n is the number of experiments).
Determination of glutathione levels. The changes in the levels of reduced (GSH) and oxidized (GSSG) glutathione were determined using methods described previously 18 . Cells were incubated during 4 h with the different stimuli assayed. A spectrofluorimeter (Tecan Infinite M200, Grödig, Austria) was employed to detect GSH or GSSG at 350 nm/420 nm (excitation/emission) respectively. For quantification, standard curves of GSH and GSSG were used. Normalization was carried out based on the total protein concentration in each sample 23 . A standard curve was prepared using bovine serum albumin. The experiments were carried out employing batches of cells obtained from different preparations.
Determination of SoD activity. This procedure was carried out using a commercially available kit from BioVision. Stimuli were added to the cells and were incubated during 1 h. Thereafter SOD activity was determined following the manufacturer's directions. The sensitive SOD assay kit utilizes WST-1 that produces a water-soluble formazan dye upon reduction with superoxide anion.
The activity of SOD can be determined by a colorimetric method. Absorbance at 450 nm of the samples was measured employing a spectrofluorimeter (VariosKan Lux 3020-205, Thermo Sci., Vantaa, Finland). The experiments were carried out employing batches of cells obtained from different preparations. Results show the mean www.nature.com/scientificreports www.nature.com/scientificreports/ change of absorbance expressed in percentage ± SEM (n) with respect to non-stimulated cells, where n is the number of independent experiments. Determination of total antioxidant capacity. Total antioxidant capacity (TAC) was determined using a commercially available kit from BioVision, following manufacturer's directions. Absorbance at 570 nm of the sample was measured employing a plate reader (CLARIOstar Plus, BMG Labtech., C-Viral, Madrid, Spain). Results show the mean change of absorbance expressed in percentage ± SEM (n) with respect to non-stimulated cells, where n is the number of independent experiments. Western blotting analysis. Western blotting was performed using previously described methods 14 . Cells in culture were incubated in the presence of different stimuli during 1 h and lysed. Bradford's method was used for quantification of the protein content of lysates 23 . Protein lysates (12 µg/lane) of each sample were separated by SDS-PAGE, using 10% polyacrylamide gels, and were transferred to nitrocellulose membranes. Specific primary and the corresponding IgG-HRP conjugated secondary antibody were used for detection of proteins. Quantification of the intensity of the bands which appear was performed using the software ImageJ (http://imagej. nih.gov/ij/). The experiments were carried out employing batches of cells obtained from different preparations. Values are expressed as the mean ± SEM of normalized values expressed as % vs control (non-stimulated) cells.
Statistical analysis. Statistical analysis of data was performed by one-way analysis of variance (ANOVA) followed by Tukey post hoc test, and only P values < 0.05 were considered statistically significant. For individual comparisons and statistics between individual treatments we employed the Student's t test, and only P values <0.05 were considered statistically significant.

Results
Effects of melatonin on cell viability. It has been suggested that melatonin modulates cell viability of different cellular types 9,14,25,26 , including PSC 17,18 . At this point it was of interest to corroborate the effect of melatonin on cell viability. Thus, PSC were incubated in the absence (non-treated cells) or in the presence of 1 mM, 100 µM, 10 µM or 1 µM melatonin, and cell viability was evaluated after 48 h of culture. The viability of cells that had been incubated in the presence of melatonin was compared with that of non-treated cells.
Cell viability dropped in the presence of 10 µM to 1 mM of melatonin ( Fig. 1). A maximal effect was noted with 1 mM melatonin. Separate batches of cells were treated with 1 µM thapsigargin (Tps), which served as control for cell death 27 . In the presence of Tps a strong decrease in cell viability was observed.
Effect of melatonin on cellular oxidative state. It has been suggested the melatonin may exert a pro-oxidant action that could underlie its antiproliferative actions 28 . To study this possibility we analyzed the effect of melatonin on ROS production. For this purpose PSC were loaded with the ROS-sensitive fluorescent dyes CM-H 2 DCFDA or MitoSOX Red. Thereafter, cells were incubated during 1 h with melatonin (1 mM, 100 µM, 10 µM or 1 µM). The compound evoked a concentration-dependent increase in ROS production both in the cytosol and in the mitochondria. Hydrogen peroxide (100 µM) was used as a control of oxidation. For this purpose the oxidant was added to the cells, which were then incubated during 1 h. In the presence of hydrogen peroxide a statistically significant increase in dye-derived fluorescence was observed, reflecting an increase in oxidation ( Fig. 2A,B). Increases of cellular calcium (Ca 2+ ) have been related with ROS generation and with pancreatic disease 22,29 . In a former work we have shown that melatonin induces mobilization of Ca 2+ in PSC 17 . In order to check whether ROS generation in response to melatonin was dependent on Ca 2+ , we performed a series of experiments in which PSC were challenged in the absence of extracellular Ca 2+ (medium containing 0.5 mM EGTA). Under these conditions ROS production evoked by melatonin did not differ from that observed in the presence of Ca 2+ (Fig. 2C,D).
In order to investigate whether the increase in ROS production was accompanied by lipid and/or protein oxidation, the effect of melatonin on protein carbonyl levels and on TBARS were assayed. For this purpose, cells were incubated during 1 h in the presence of melatonin (1 mM, 100 µM, 10 µM or 1 µM). H 2 O 2 (100 µM) was used as control. We observed a concentration-dependent increase in the total protein carbonyls content in cells treated with melatonin in comparison with that noted in non-stimulated cells. A maximal effect was observed in response to 1 mM melatonin (Fig. 3A). However, no statistically significant changes were detected in the levels of TBARS (Fig. 3B). Treatment of cells with H 2 O 2 (100 µM) induced statistically significant increases in both total protein carbonyls and TBARS (Fig. 3A,B).
Effect of melatonin on mitochondrial membrane potential. It has been suggested that oxidative stress and changes in ψ m are closely related 30 . In order to analyze whether melatonin induces changes in ψ m , we performed a series of experiments in which PSC were loaded with the mitochondria-specific voltage-sensitive dye TMRM. The cells were then incubated during 1 h in the presence of melatonin (1 mM, 100 µM, 10 µM or 1 µM). We could only observe a statistically significant decrease in ψ m in cells treated with 1 mM melatonin. No detectable changes in ψ m were noted in response to the other concentrations of melatonin employed. As a control, different batches of cells were incubated in the presence of the mitochondrial uncoupler CCCP 22,31 . In the presence of CCCP (100 nM) a statistically significant decrease in ψ m was detected (Fig. 4). www.nature.com/scientificreports www.nature.com/scientificreports/ Effect of melatonin on glutathione levels. Glutathione represents a major antioxidant defense against oxidative stress 32 . Because we had observed ROS production in the presence of the melatonin, it was of interest to test its effect on the glutathione system in PSC. Therefore, cells were incubated during 4 h in the presence of melatonin (1 mM, 100 µM, 10 µM or 1 µM) and the levels of GSH and GSSG were analyzed. We observed a concentration-dependent decrease in GSH/GSSG ratio in cells treated with melatonin in comparison with that noted in non-stimulated cells. A maximal effect was observed in response to 1 mM or 100 µM melatonin. A slight decrease in GSH/GSSG ratio was observed in response to 10 µM melatonin, which was not statistically significant. Whereas we did not detect changes in GSH/GSSG ratio in cells treated with 1 µM melatonin (Fig. 5A).
Effect of melatonin on Nrf2-dependent antioxidant enzymes. Nrf2 is a transcription factor that enhances the expression of a multitude of antioxidant and phase II enzymes, which regulate redox homeostasis 33 . The results shown above indicate that melatonin induces changes in the redox status of PSC. Therefore, we decided to study whether melatonin could stimulate the transcriptional activation of certain antioxidant enzymes through the activation of Nrf2. For this purpose PSC were incubated during 4 h in the presence of melatonin (1 mM, 100 µM, 10 µM or 1 µM) and RT-qPCR of the relative mRNA abundance was performed. Melatonin evoked statistically significant increases in the expression of GCLc, CAT, NQO1 and HO-1 (Fig. 5B-D). As a control, cells were incubated in the presence of H 2 O 2 (100 µM), a known Nrf2 activator 34 . The oxidant increased the expression of all four antioxidant enzymes studied. www.nature.com/scientificreports www.nature.com/scientificreports/ Effect of melatonin on superoxide dismutase. Superoxide dismutases (SOD) catalyze the dismutation of superoxide anion (O2 − ) to H 2 O 2 , which is then catalyzed to innocuous O 2 and H 2 O by glutathione peroxidase and catalase. Thus, SOD is involved in the defense system against ROS 35 . Several classes of SOD have been identified: Cu/Zn SOD (SOD1), which is localized in cytosol, and MnSOD (SOD2), which is localized in mitochondria 36,37 . We were intereseted in analyzing whether melatonin exerted any affect on SOD. Thus, PSC were incubated during 1 h with the compound (1 mM, 100 µM, 10 µM or 1 µM) and SOD activity was then analyzed. In the presence of melatonin a concentration-dependent decrease in SOD activity was observed (Fig. 6).
We further analyzed the effect of melatonin on SOD and decided to study the protein levels of the enzyme by Western blotting. The results show that PSC that had been incubated with melatonin exhibited lower levels of both SOD1 and SOD2, compared with non-treated cells. The stronger decrease of protein expression was noted for SOD1 ( Fig. 7A-D).
Additional studies were carried out to confirm the effect of melatonin on SOD expression. PSC were incubated during 1 h in the presence of melatonin (1 mM, 100 µM, 10 µM or 1 µM) and RT-qPCR of the relative mRNA abundance of SOD1 and SOD2 were performed. In cells treated with melatonin, statistically significant decreases in the mRNA of both proteins were observed (Fig. 7E-F).
Effect of melatonin on the total antioxidant capacity. We additionally evaluated the TAC of PSC. As shown in Fig. 8, the TAC of cells incubated in the presence of melatonin was decreased in comparison with that noted in non-stimulated cells (incubated in the absence of melatonin). The effect did not depend on the concentration of melatonin used. Incubation of PSC with the oxidant H 2 O 2 (100 µM) evoked a statistically significant decrease in TAC compared with non-stimulated cells. These results confirm that melatonin induces changes in the oxidative state of PSC.

Discussion
It is well known that tumors undergo adaptive responses that lead to resistance and accelerated repopulation. This allows them to overcome doses of radiation and chemotherapy. Resistance can occur following different adaptive responses, which are due to the nature of the tumor cells or to the release of factors by immune cells as well as to participation of other cell types present in the tumor microenvironment 9 . In this line a major contributing factor is the characteristic extensive stromal or fibrotic reaction found in tumors 2 .
In some cancer cells, melatonin itself induces apoptosis 10,13,14 or aids sensitizing cancer cells to therapy [38][39][40][41] . In addition, previous results of our laboratory showed that melatonin modulates viability of PSC. This is of relevance because PSC have been pointed out as major players in stromal formation within tumors 17,18 . Therefore melatonin is emerging as a potential tool in the treatment of cancer.
In this study, we provide further evidences that support a potential role for melatonin in the regulation of PSC proliferation by setting-up a prooxidant environment within the cells, which decreases their viability. The oxidative conditions that we have observed might be based on ROS production together with a decrease in TAC of the cells. The latter might have a basis on a reduction of glutathione levels and a decrease in SOD activity. As a whole, the results that we have obtained can be considered relevant bearing in mind that PSC play major roles in fibrosis developed in pancreatic diseases. www.nature.com/scientificreports www.nature.com/scientificreports/ The drop in PSC viability that we have observed confirms previous studies of our laboratory 17,18 . Interestingly, a decrease in the proliferation of this cellular type would be a helpful maneuver that could help in diminishing the fibrosis present in the pancreas under pathological conditions, especially in tumors.
Maintenance of adequate cellular red-ox equilibrium is critical for cell function and viability 42 . Conversely to the protective role of melatonin against oxidative stress 43 the compound can also exhibit prooxidant effects, which have been related with a cytotoxic effect 28 . The analysis of the results that we have obtained showed that melatonin induced ROS production. The generation of ROS could be detected in both the cytosol and the mitochondria, but the contribution of Ca 2+ was negligible. Participation of mitochondria in ROS generation has been demonstrated 44,45 . These results are in agreement with previous findings of our laboratory, which showed that ROS www.nature.com/scientificreports www.nature.com/scientificreports/ production was increased in PSC treated with melatonin 17,18 , and confirm the hypothesis of putative prooxidant actions of melatonin in this cellular type. The present research was conducted in order to further investigate other possible points of action of melatonin to exert its prooxidant effects that could explain its actions on PSC viability.
Our results additionally show that melatonin treatment might be accompanied by oxidation of certain cellular structures. This could be reflected by the increase in the oxidation of cellular proteins that we have noted; however, we could not detect changes in the oxidation of lipids (TBARS). From these observations we could assume that melatonin might differentially affect lipids and proteins within the cell. Besides, it could be possible that certain proteins are more prone to oxidation that lipids upon melatonin treatment. Therefore, melatonin effects on protein redox state could lead to the modulation of metabolic pathways regulated by such proteins, which are activated/inactivated due to changes in their oxidative state.
In addition, impairment of mitochondria leads to ROS generation 22,46 . Our results also show that ψ m decreased in the presence of melatonin. At this point we could hypothesize that melatonin might affect mitochondrial physiology in PSC. In fact, different studies have suggested that melatonin alters mitochondrial physiology which is related with cell death 9,14 . Moreover, a few studies using cultured cells found that melatonin stimulated ROS generation at pharmacological concentrations (micro-molar to milli-molar range) in several tumor and non-tumor cells; thus, melatonin functioned as a conditional pro-oxidant 47 .
Additional evidences for a disruption by melatonin of the redox balance in PSC derives from the experiments directed to analyze its effect on glutathione. The glutathione system is a major tool used in the defense against damage caused by ROS. A defeat of antioxidant systems, like a decrease in the GSH content, can lead cells to fault in the control ROS production and, therefore, can induce cell damage and death 48 . Our results show that, in the presence of melatonin the ratio GSH/GSSG decreased. This action depended on the concentration of melatonin used. Higher effects were found at 100 µM and 1 mM of the indole, whereas no detectable changes were noted in cells treated with 1 µM melatonin. The decline in GSH/GSSG ratio that we have noted points towards an increase in oxidized glutathione. This observation might reflect a pro-oxidant action of melatonin. In other words, the decrease in the availability of reduced glutathione could be related with the increase in ROS generation evoked by melatonin. These results are in agreement with previous observations of our laboratory, obtained in human PSC, in which we showed that melatonin evoked concentration-dependent changes in glutathione oxidation 18 . Interestingly, it could be feasible that melatonin might exert the same effects in human cells as those noted in murine cells, thus providing putative beneficial actions of the compound on human health as expected from the results obtained in studies carried out on animal cells.
In another set of experiments we have detected an increase in the expression of the Nrf2-regulated antioxidant enzymes GClc, CAT, HO-1 and NQO1. Specifically, GCLc is involved in glutathione synthesis 49 . Nrf2 is required for systemic protection against redox-mediated injury. Under oxidative conditions the Keap1-ARE (antioxidant response element) pathway is activated via the upregulation of Nrf2 50 . Melatonin activates this pathway to induce protective antioxidant actions 24,51 . In our study, the prooxidant conditions evoked by melatonin might activate the Nrf2-regulated pathway in an attempt to counteract the pro-oxidative state that we have observed.
SOD is another enzyme with pivotal role in cellular antioxidant defence 52 . Our results show that SOD activity is decreased in the presence of melatonin. This effect could be explained by a diminished expression of both SOD1 and SOD2, whith a higher effect on SOD1. Our results further suggest that melatonin regulates SOD at the translation level. To our knowledge, this is the first time to show that melatonin decreases the expression of SOD. Findings of other researchers show that melatonin either increases SOD expression 53,54 or does not induces changes in the levels of these proteins 55 .  Interestinlgy, melatonin exerts prooxidant effects 28 (Sanchez-Sanchez et al., 2011). A SOD activity under a certain level could lead to diminished antioxidant pretection of the cell that, if is not counteracted by other antioxidant defenses, might lead to prooxidant conditions that could compromise cell function and viability. As a consequence, and taking also into account the effects on glutathione that we have mentioned above, the TAC of the cells should be expected to decrease, as we have observed. Therefore, our results point out that melatonin modulates pivotal points of the cellular antioxidant machinery and leads to prooxidant conditions that could drive the mechanisms involved in PSC viability and/or proliferation. In fact, we have shown previously that melatonin induced changes in the phosphorylation state of members of the mitogen-activated protein kinases family, which are involved in cell proliferation and survival. This resulted in a decrease in cell viability 17 .
The concentrations of melatonin that we have employed are not physiological and fairly fall within the pharmacological range 56 . However, pharmacological concentrations of melatonin have been used in a plethora of studies directed to the study of disease [57][58][59] , including studies carried out our laboratory 14,17,24,60,61 .
In conclusion, we present evidences that stand out melatonin as a compound with the ability to regulate PSC physiology. Despite the protective role that melatonin exerts in a wide variety of cellular types, here we show that the compound induces pro-oxidative conditions that might have consequences on cell viability. It is noteworthy to bear in mind that the actions of melatonin on cellular physiology might be cell-and context-dependent. Contribution of stellate cells to survival and development of transformed epithelia within the pancreas has been documented 62,63 . Thus, strategies directed to controlling the growth of fibrotic tissue within tumors might be challenging in the treatment of cancer 2 . In this line, our results suggest a probable mechanism by which melatonin modulates fibrosis within the pancreas. Therefore, melatonin could be considered a hopeful aid in the therapy of pancreatic cancer.