3,4-Methylenedioxymethamphetamine (“Ecstasy”) Stimulates the Expression of a1(I) Procollagen mRNA in Hepatic Stellate Cells

3,4-Methylenedioxymethamphetamine (“Ecstasy”) Stimulates the Expression of a1(I) Procollagen mRNA in Hepatic Stellate Cells M. Varela-Rey*, C. Montiel-Duarte*, G. Beitia†, E. Cenarruzabeitia†, and M. J. Iraburu*1. *Department of Biochemistry, University of Navarra, Pamplona, Spain. †Department of Pharmacology, University of Navarra, Pamplona, Spain 1 To whom correspondence should be addressed at Departamento de Bioquímica, Universidad de Navarra, C/ Irunlarrea s/n. 31080 Pamplona, Navarra, Spain. Fax: 34-48- 425619. E-mail: [email protected]. ABSTRACT 3,4-Methylenedioxymethamphetamine, MDMA (“Ecstasy”), has been previously shown to produce cell necrosis and fibrosis in the liver. Our aim was to study the ecect of MDMA on the type I collagen production by a cell line of hepatic stellate cells (HSC), the cell type mainly responsible for collagen synthesis in the liver. We demonstrated that MDMA increases a1(I) procollagen mRNA levels and that this increase correlates with glutathione depletion and enhanced hydrogen peroxide production by HSC. Pre-treatment with either glutathione monoethyl ester or deferoxamine prevents the MDMA-induced a1(I) procollagen mRNA expression, indicating oxidative stress to be a mediator of this ecect. Lipid peroxidation was not detected in MDMA-treated cells and therefore does not seem to be involved in the profibrogenic action of MDMA on HSC. Key Words: MDMA; type I collagen; hepatic stellate cells; glutathione; liver fibrosis. INTRODUCTION MDMA (3,4-methylenedioxymethamphetamine, “Ecstasy”) is a synthetic derivative of amphetamine that has been reported to produce dicerent systemic and organ-specific ecects, including serotonergic neurotoxicity, hyperthermia, cardiac arrythmias, convulsions, hepatotoxicity and fulminant liver and renal failure (1, 2). Most of the studies carried out to elucidate the molecular mechanisms of MDMA action have been performed on neurons. They have shown that several ecects of MDMA are not directly caused by this substance but by some of its metabolic derivatives, such as semiquinone or quinone species which react easily with nucleophiles like glutathione (3–5). There is clinical evidence pointing to the liver as one of the target organs for MDMA. Thus, MDMA has been described to cause cell necrosis (1, 6, 7) as well as accelerated fibrosis (8) in the liver. However, little is known about the molecular mechanisms responsible for these ecects. Hepatic fibrosis is characterized by an overproduction of extracellular matrix proteins, especially type I collagen, that are mainly synthesized by hepatic stellate cells (HSC). Lipid peroxidation and oxidative stress have been shown to be involved in the induction and/or development of liver fibrosis, as well as in the increased collagen production by HSC, both in animal models and in cell culture systems (9–12). Although MDMA can induce oxidative stress and lipid peroxidation in neurons, these ecects have been explained as a consequence of neuronspecific alterations such as dopamine deamination (13) and there is no evidence of their involvement in the hepatotoxicity of MDMA. In this study the ecect of MDMA on the type I collagen mRNA levels of a HSC cell line is analyzed. We demonstrated MDMA to exert a pro fibrogenic ecect that is mediated by oxidative stress in the absence of lipid peroxidation. MATERIALS AND METHODS Reagents. MDMA-HCl was a gift from the “Audiencia Provincial de Navarra”. Glutathione monoethyl ester was a gift from Dr. Mato. Deferoxamine was from Sigma (St. Louis, MO). 5- 6- chloromethyl-29,79-dichlorohydrofluorescein diacetate (CM-H2DCFDA) was from Molecular Probes (Eugene, OR). Cell culture reagents were from Gibco BRL. Cell culture and materials. All the experiments were performed using the HSC line CFSC-2G that has a similar phenotype to that of freshly isolated HSC (14). Cells were cultured in MEM supplemented with 10% bovine fetal serum and non-essential amino acids for 36 hours after which the medium was replaced for a serum-free medium. Treatments were carried out 12 hours later. Unless otherwise indicated, HSC were treated with 1.0 mM MDMA for 24 hours. In some experiments cells were pre-treated for 30 min. with either 2.0 mM glutathione monoethyl ester or 0.1 mM deferoxamine. RNA extraction and Northern blot analysis. Total RNA was extracted as described by Chomzcynski et al (15). Northern blot assays with 32P-labeled probes for a1(I) procollagen (COL-I) and b-actine. Autoradiographic signals were quantitated by scanning densitometry. Measurement of intracellular GSH levels. The intracellular levels of glutathione (reduced form) were determined by the method of Hissin and Hilf (16). HSC were cultured and treated as described above. After treatment, HSC were scraped and resuspended in MEM (1x106 cell/ml). An aliquot of the deproteinized cell suspension (50 ml) was mixed with 2.1 ml of 200 mM sodium phosphate bucer, pH 8.0, containing 5 mM EDTA. Then 100 ml of a solution of o-phthaldialdehyde (1 mg/ml in methanol) was added, and 15 min later the intensity of fluorescence was determined (excitation 350 nm, emission 420 nm). Measurement of lipid peroxidation. Lipid peroxidation was determined by assay for thiobarbituric acid reactive substances (TBARS) at 535 nm, as described by Buegue and Aust (17). HSC were scraped and resuspended in PBS (1x106 cell/ml) and aliquots (500 µl) were precipitated with 1.0 ml of 10% trichloroacetic acid (TCA) and centrifuged for 20 s at 6000 rpm. Aliquots (1.0 ml) of the supernatants were added to an equal volume of 1% thiobarbituric acid and the mixture was heated for 10 min in a boiling water bath and allowed to cool down. The absorbance at 535 nm was determined. TBARS are represented as malondialdehyde equivalents, calculated using an extintion coecicient of 1.6 x 105 M-1 cm-1. Measurement of hydrogen peroxide production. Hydrogen peroxide levels were measured using the fluorescent probe CM-H2DCFDA (18). For these experiments HSC were grown in MEM without phenol red. For time-course studies HSC plated to sub-confluence in 12 well plates, treated for dicerent times with 1.0mM MDMA, and then incubated for 20 minutes with 5 µM CM-H2DCFDA at room temperature. Fluorescence was analyzed in a Cytofluor 2350 (excitacion at 485 nm, emission at 530 nm). Flow cytometry studies were performed with HSC previously treated with 1.0 mM MDMA for 4 hours, trypsinized and resuspended on MEM without phenol red. After adding 10 µM CM-H2DCFDA the cell suspension was maintained at room temperature for 20 minutes and analyzed in a FACScan (Becton Dickinson) at 530 nm (excitation at 488 nm). Statistical analysis. Data were analyzed using the Kruskal-Wallis test to determine dicerences between all independent groups. When significant dicerences were obtained (p<0.05), dicerences between two groups were tested using the Mann-Withney U test. RESULTS MDMA Induces COL-I mRNA Levels in HSC The first experiments were aimed to determine the ecect of MDMA on the a1(I) procollagen (COL-I) mRNA levels of the HSC cell line CSFC-2G. Cell cultures were exposed to increasing concentrations of MDMA, ranging from 0.1 to 1.0 mM. Cell viability decreased significantly when higher concentrations of MDMA were used (3.0–5.0 mM). As illustrated in Fig. 1, COL-I mRNA levels increased as a result of MDMA treatment in a dose-dependent fashion, 1.0 mM being the dose with highest ecect. Time-course experiments were then carried out using 1.0 mM MDMA. As Fig. 2 shows, the statistically significant ecect of MDMA on the COL-I mRNA levels was obtained 24 h after treatment. MDMA Induces Oxidative Stress in HSC To determine whether oxidative stress was involved in the induction of the expression of COL-I mRNA in HSC by MDMA, we investigated the ecect of MDMA on dicerent biochemical parameters related to the intracellular redox state. HSC were treated for 4, 8 and 24 hours with 1.0 mM MDMA. At each time point, lipid peroxidation, GSH content and hydrogen peroxide production were determined. The extent of total lipid peroxidation of HSC, measured as thiobarbituric reactive substances (TBARS), was not acected by exposure of the cells to MDMA at any of the time points (Table I). However, the intracellular GSH content significantly decreased as a consequence of treatment with MDMA, being 20% lower in comparison to the control cells at all the time points (Table II). The determination of hydrogen peroxide production by HSC using CM-H2DCFDA as a fluorogenic probe, showed that MDMA produced a transient increase in the intracellular levels of hydrogen peroxide 4 hours after treatment (Fig. 3A). This ecect was confirmed by flow cytometry using the same probe in HSC treated with 1.0 mM MDMA for 4 h (Fig. 3B). Antioxidants Prevent the E@ect of MDMA on COL-I mRNA Levels Since the COL-I mRNA increase caused by MDMA correlates with a decrease in the GSH levels of HSC, the relationship between both ecects was studied in experiments using two dicerent antioxidants: glutathione monoethyl ester and deferoxamine. Glutathione monoethyl ester acts as a membrane-soluble form of glutathione that can very ecectively restore the intracellular levels of GSH. Deferoxamine has been extensively used as an inhibitor of the Fenton reaction, which is the main source of a very reactive chemical species, the hydroxyl radical (OH•). HSC were pre-treated with either glutathione monoethyl ester (2.0 mM) or deferoxamine (0.1 mM) for 30 min before adding MDMA (1.0 mM for 24 h) to the cell cultures. Norhern blot analysis of the COL-I mRNA levels in control and treated cells, showed that both pre-treatments were able to prevent the COL-I mRNA expression induced by MDMA (Fig. 4). Glutathione monoethyl ester, but not deferoxamine, had an inhibitory ecect on the basal levels of COL-I mRNA that was restored to control levels by treatment with MDMA. DISCUSSION Hepatic stellate cells (HSC) are the main cell type involved in the increased synthesis of type I collagen characteristic of the fibrotic liver. There are two lines of evidence indicating oxidative stress to be a possible mediator of liver fibrosis. One is based on animal models and cell culture studies using ethanol or Cl4C as fibrogenic agents, indicating lipid peroxidation as the common link to hepatic fibrosis (9, 12). The other is based on the fact that the main pro-fibrogenic cytokine involved in the development of fibrosis, namely TGF-b, produces an increase in the intracellular levels of reactive oxygen intermediates such as hydrogen peroxide (19, 20). Furthermore, hydrogen peroxide has been recently shown to induce a1(I) procollagen (COL-I) mRNA levels due to a higher rate of transcriptional activity in the same HSC cell line used in the present work (21). In the present paper, we show MDMA to exert a pro-fibrogenic ecect on HSC, significantly increasing the COL-I mRNA levels (Fig. 1). Interestingly, lipid peroxidation does not seem to be involved in this ecect, as we did not find any changes in levels of MDA after treatment with MDMA (Table I). However, the induction of COL-I mRNA did correlate with a depletion of the intracellular GSH levels caused by MDMA on HSC (Table II). Some MDMA metabolic derivatives have been shown to react with glutathione (3, 4) and more recently a depletion of intracellular GSH levels in hepatocytes exposed to MDMA has been reported (22). Glutathione is the most abundant intracellular thiolcontaining small molecule and is involved in the regulation of hydrogen peroxide levels through glutathione peroxidase. Therefore, a depletion of GSH could generate an increase of hydrogen peroxide, which is a precursor of more reactive species such as the hydroxyl radical (OH•). As shown in Fig. 3, MDMA treatment of HSC produced a transient increase in hydrogen peroxide production. The relationship between the GSH depletion and the increase on COL-I mRNA levels in HSC treated with MDMA is further demonstrated by the inhibition produced by pre-treatment with glutathione monoethyl ester and deferoxamine (Fig. 4). Although additional studies would be needed to establish which reactive oxygen intermediate is responsible for the induction of COL-I mRNA levels by MDMA, the fact that deferoxamine inhibits this ecect suggests the hydoxyl radical (OH•) as a candidate. The Fenton reaction is the main source of OH• and its inhibition by deferoxamine, while not necessarily acecting the hydrogen peroxide content, would diminish the OH• production. Some previous reports using pharmacological modifications to deplete the GSH content of HSC have failed to find a correlation between collagen production and the glutathione status of the cells (23). However, our results suggest that under certain conditions, such as exposure to MDMA, glutathione may be important in the regulation of collagen production by HSC. Taken together, our data provide evidence of a fibrogenic ecect caused by MDMA on HSC which is mediated by oxidative stress and that could account for some clinical observations in MDMA users, such as hepatotoxicity and liver fibrosis. New studies using in vivo models should be carried out to determine the physiologic relevance of our findings. REFERENCES 1. Henry, J. A., Jecreys, K. J., and Dawling, S. (1992) Lancet 340, 384–387. 2. Milroy, C. M., Clark, J. C., and Forrest, A. R. (1996) J. Clin. Pathol. 49, 149–153. 3. Tucker, G. T., Lennard, M. S., Ellis, S. W., Woods, H. F., Cho, A. K., Lin, L. Y., Hiratsuka, A., Schmitz, D. A., and Chu, T. Y. (1994) Biochem. Pharmacol. 47, 1151–1156. 4. Hiramatsu, M., Kumagai, Y., Unger, S. E., and Cho, A. K. (1990) J. Pharmacol. Exp. Ther. 254, 521–527. 5. Miller, R. T., Lau, S. S., and Monks, T. (1997) J. Eur. J. Pharmacol. 323, 173–180. 6. Miller, R. T., Lau, S. S., and Monks, T. J. (1996) Chem. Res. Toxicol. 9, 457–465. 7. Ellis, A. J., Wendon, J. A., Portmann, B., and Williams, R. (1996) Gut 38, 545–548. 8. Khakoo, S. I., Coles, C. J., Armstrong, J. S., and Barry, R. E. (1995) J. Clin. Gastroenterol. 20, 244–247. 9. Bedossa, P., Houglum, K., Trautwein, C., Holstege, A., and Chojkier, M. (1994) Hepatology 19, 1262–1271. 10. Svegliati-Baroni, G., D’Ambrosio, L., Ferretti, G., Casini, A., Di-Sario, A., Salzano, R., Ridolfi, F., Saccomanno, S., Jezequel, A. M., and Benedetti, A. (1998) Hepatology 27, 720– 726. 11. Lee, K. S., Buck, M., Houglum, K., and Chojkier, M. (1995) J. Clin. Invest. 96, 2461–2468. 12. Parola, M., Pinzani, M., Casini, A., Albano, E., Poli, G., Gentilini, A., Gentilini, P., and Dianzani, M. U. (1993) Biochem. Biophys. Res. Commun. 194, 1044–1050. 13. Sprague, J. E., and Nichols, D. E. (1995) J. Pharmacol. Exp. Ther. 273, 667–6673. 14. Greenwel, P., Rubin, J., Schwartz, M., Hertzberg, E. L., and Rojkind, M. (1993) Lab. Invest. 69, 210–216. 15. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 16. Hissim, P. J., and Hilf, R. A. (1976) Anal. Biochem. 74, 214–226. 17. Buege, J. A., and Aust, S. D. (1978) Methods Enzymol. 52, 302– 310. 18. Bass, D. A., Parce, J. W., Dechatelet, L. R., Szejda, P., Seeds, M. C., and Thomas M. (1983) J. Immunol. 130, 1911–1917. 19. Ohba M., Shibanuma, M., Kuroki, T., and Nose, K. (1994) J. Cell. Biol. 126, 1079–1088. 20. Hong, Y., H., and Peng, H., B., La Fata, V., and Liao, J. K. (1997) J. Immunol. 159, 2418– 2423. 21. Garcıa-Trevijano, E. R., Iraburu, M. J., Fontana, L., Dominguez- Rosales, J. A., Auster, A., Covarrubias-Pinedo, A., and Rojkind, M. (1999) Hepatology 29, 960–970. 22. Beitia, G., Cobreros, A., Sainz, L., and Cenarruzabeitia, E. (1999) In press. 23. Maher, J. J., and Neuschwander-Tetri, B. A. (1997) Hepatology 26, 618–623. Figure 1. FIG. 1. EAect of MDMA on the expression of a1(I) procollagen (COL-I) mRNA. HSC were treated with 0.1, 0.5, and 1.0 mM MDMA for 24 h. Northern blot analysis was performed with 10 µg of total RNA. (A) COL-I mRNA fold increase in response to MDMA. Each bar represents the mean 6 SD of at least quadruplicate experiments (*p<0.05, vs control). Values were corrected for loading dicerences after hybridization with a cDNA probe for b-actine. (B) Representative Northern blot showing the autoradiographic signals for COL-I and b-actine. Figure 2. FIG. 2. Time-course analysis of COL-I mRNA levels of HSC treated with 1.0 mM MDMA. Northern blot analysis was performed with 10 mg of total RNA extracted at the various time points indicated in the figure. Each bar represents the mean +/- SD of at least quadruplicate experiments (**p<0.01, vs control). Values were corrected for loading dicerences after hybridization with a cDNA probe for b-actine. Figure 3: FIG. 3. EAect of MDMA on hydrogen peroxide production. (A)Time course analysis of hydrogen peroxide levels of HSC treated with MDMA. HSC were exposed for 4, 8, and 24 h to 1.0 mM MDMA, and hydrogen peroxide levels determined by fluorimetry, using CMH2DCFDA as a probe. Each bar represents the mean 6 SD of fluorescence fold change compared to controls of at least quadruplicate experiments (*p<0.05 vs control). (B) Analysis by flow cytometry of the hydrogen peroxide levels in HSC treated for 4 h with 1.0 mM MDMA compared to control (untreated) HSC. CM-H2DCFDA was used as a fluorogenic probe as described in Materials and Methods.