The effect of non-genotoxic carcinogens, phenobarbital and clofibrate ...

2 downloads 30 Views 472KB Size Report
The effect of non-genotoxic carcinogens, phenobarbital and clofibrate, on the relationship between reactive oxygen species, antioxidant enzyme expression and ...
Carcinogenesis vol.19 no.10 pp.1715–1722, 1998

The effect of non-genotoxic carcinogens, phenobarbital and clofibrate, on the relationship between reactive oxygen species, antioxidant enzyme expression and apoptosis

Carmen Dı´ez-Ferna´ndez, Nuria Sanz, Alberto M.Alvarez1, Armin Wolf2 and Marı´a Cascales3 Instituto de Bioquı´mica (CSIC-UCM) and 1Centro de Citometrı´a de Flujo, Facultad de Farmacia, Universidad Complutense, Plaza de Ramo´n y Cajal sn, 28040 Madrid, Spain and 2Novartis Pharma AG, Drug Safety Department, CH 4002 Basel, Switzerland 3To

whom correspondence should be addressed Email: [email protected]

Phenobarbital and clofibrate, two non-genotoxic carcinogens, have been investigated regarding the relationship between reactive oxygen species, antioxidant enzyme expression and apoptosis in primary cultures of rat hepatocytes. Low toxicity concentrations, 200 and 100 µg/ml for phenobarbital and clofibrate respectively, were used to examine their effect on spontaneous or transforming growth factor β1 (TGFβ1)-induced apoptosis and on the expression of antioxidant defence enzymes (superoxide dismutases and catalase). The increased incidence of apoptotic nuclei was visualized in TGFβ1-treated cultures with the fluorescent dye Hoechst 33258 and was quantified under all experimental conditions by measurement of the hypodiploid peak in DNA histograms obtained by flow cytometry. Both substances, when added separately to hepatocyte cultures and incubated for 24 and 48 h, significantly diminished spontaneous apoptosis and exhibited a slight suppression of TGFβ1-induced apoptosis. Endogenous peroxide production by hepatocytes increased with TGFβ1, phenobarbital or clofibrate and the increase was greater with phenobarbital and in the presence of TGFβ1 with both drugs. Gene expression of catalase and Mn- and Cu,Zn superoxide dismutases (SOD) was evaluated by northern blot analysis of hepatocytes incubated in the presence of phenobarbital or clofibrate with or without TGFβ1 and the following differences were detected: phenobarbital induced a significant decrease in both dismutases (to 56%, P < 0.05, and 55%, P < 0.05, for Mn- and Cu,Zn-SOD respectively) and a 2-fold increase (P < 0.01) in catalase; clofibrate induced a slight decrease in both SODs and a 4-fold increase (P < 0.05) in catalase; TGFβ1 significantly decreased to 37% (P < 0.05) expression of catalase while not significantly affecting expression of both SODs. We conclude that inhibition of spontaneous apoptosis induced by either phenobarbital or clofibrate is accompanied by increases in the endogenous levels of peroxides and by significant induction of catalase gene expression. Furthermore, the lack of effect of both compounds on TGFβ1-induced apoptosis could be a consequence of the inability of these two compounds to counteract the depressing effect of TGFβ1 on expression of catalase. Abbreviations: DCF, 29,7’-dichlorofluorescein; DCFH-DA, 29,7’-dichlorofluorescein diacetate; DMSO, dimethyl sulphoxide; LDH, lactate dehydrogenase; PI, propidium iodide; PMA, phorbol-12-myristate-13-acetate; PBS, phosphate-buffered saline; ROS, reactive oxygen species; SOD, superoxide dismutase; TGFβ1, transforming growth factor β1. © Oxford University Press

Introduction Under normal culture conditions isolated hepatocytes lose their viability and die rapidly. However, exposure to non-genotoxic carcinogens, such as peroxisome proliferators or tumour promoters, confers stability because of the suppressor effects of these substances on apoptosis (1,2). Elimination of these compounds results in death of hepatocytes which shows that, at least in vitro, the effect induced by these carcinogens is reversible and is not due to genotypic changes (3). It is well known that tumour promoters, such as phorbol esters, inhibit apoptosis through direct activation of protein kinase C and the subsequent stimulation of superoxide generation, which can be inhibited by antioxidants (4–6). Accordingly, it is proposed that superoxide anion concentration plays a key role in regulating cell sensitivity to a potentially lethal signal and provides inducible resistance against apoptosis. However, at present it is generally accepted that reactive oxygen species (ROS) are involved in apoptosis, providing an effector mechanism for the pathway of programmed cell death (7). Furthermore, there is evidence that oxidative stress, induced by overproduction or decreased elimination of superoxide anion, confers a survival advantage on tumour cells over normal counterparts (4) and that the oxidative stress adaptative response involves expression of many genes, which indicates that newly synthesized proteins are required for this adaptation (8,9). Peroxisome proliferators are known to increase the size and number of peroxisomes (10), accompanied by an enhancement in peroxisomal fatty acid β-oxidation and microsomal ω hydroxylation activities. The increased levels of hydrogen peroxide due to enhanced peroxisomal β-oxidation of fatty acids can produce DNA damage and tumour formation (11). Cells possess specific enzymes that act directly as an endogenous defence system. Mn- and Cu,Zn-superoxide dismutases (SODs), located in the mitochondria and cytosol respectively, together with catalase act coordinately, protecting cells against direct or indirect oxidative damage (12,13). It has been shown that apoptosis caused by low doses of H2O2 (14) is prevented by catalase (15) and also that H2O2 is implicated, as a mediator of transforming growth factor β1 (TGFβ1), in growth inhibition (16,17). Thus, it is accepted that reactive oxygen species are involved in the mechanisms that control apoptosis (18) and superoxide anion and hydrogen peroxide can affect apoptosis, playing a key role in regulating lethal signals (4,6). Moreover, suppression of antioxidant enzyme gene expression together with the appearance of oxidative stress is involved in the process of TGFβ1 induction of apoptosis (19). In a previous study, we showed that inhibition of apoptosis induced by TGFβ1 in primary cultures of rat hepatocytes by the phorbol ester phorbol-12-myristate-13-acetate (PMA) (20) was not associated with changes in expression of antioxidant enzymes. In the present study we have tried to find a relationship between these two phenomena, inhibition of apoptosis and induction of gene expression of antioxidant systems, since 1715

C.Dı´ez-Ferna´ndez et al.

Fig. 1. LDH release in primary cultured hepatocytes incubated in the presence of phenobarbital (PB) or clofibrate (CF). Cells (1.53106) were exposed to different concentrations (0–500 µg/ml) of phenobarbital or clofibrate and were incubated at 37°C for 24 h. LDH activity was measured spectrophotometrically at 340 nm in the presence of NADH and pyruvate (24). LDH activity in the medium was expressed as a percentage of total LDH activity present in the cells at the beginning of the incubation. Results are expressed as means 6 SD of four experimental observations.

ROS are involved in both processes. On the basis of previous experiments, the following investigations were carried out: two non-genotoxic carcinogens, phenobarbital and clofibrate, were assayed (selecting low toxicity concentrations to avoid overlap with the apoptotic effect) for their effects on spontaneous and TGFβ1-induced apoptosis in primary cultures of rat hepatocytes. The in vitro effect of phenobarbital and clofibrate was also studied respecting ROS production and expression of antioxidant defence systems, Mn-SOD, Cu-Zn-SOD and catalase. Materials and methods Reagents Tissue culture media were from Biowhittaker (Walhersville, MD). Recombinant human TGFβ1 was from Calbiochem (La Jolla, CA). Standard analytical grade laboratory reagents were obtained from Merck (Darmstadt, Germany). [α-32P]dCTP (3000 Ci/mmol) and the multiprimer DNA-labelling system kit were purchased from Amersham (Aylesbury, UK). Agarose was from Hispanagar (Barcelona, Spain). 29,7’-Dichlorofluorescein diacetate (DCFH-DA) was obtained from Molecular Probes (Eugene, OR) and propidium iodide (PI) was from Sigma (St Louis, MO). Animals Two-month-old male Wistar rats with an average body weight of 180–230 g were used for the cell preparations. All animals received care as outlined in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institute of Health. Rats were supplied with food and water ad libitum and exposed to a 12 h light–dark cycle. Isolation and culture of hepatocytes Hepatocytes were isolated from Wistar rats aged 2 months by a collagenase perfusion method (21,23) and cell viability, determined by trypan blue exclusion, was always .90%. Freshly isolated hepatocytes (1.53106) were seeded in 60315 mm Petri dishes (Becton Dickinson) in 3 ml Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 100 IU penicillin, 50 µg/ml streptomycin, 50 µg/ml gentamycin and 10% fetal calf serum. After 4 h incubation at 37°C in a humidified atmosphere of 5% CO2, the medium was replaced with fresh medium supplemented with 0.1% BSA. Treatment with TGFβ1, phenobarbital and clofibrate Additions of TGFβ1 and non-genotoxic carcinogens were made so that the total volume of the medium was not increased by .2%. Phenobarbital and clofibrate were prepared at 10 mg/ml in dimethyl sulphoxide (DMSO). Aliquots of 2 ng/ml TGFβ1 were added from a 2 µg/ml stock in 4 mM HCl. Control plates were treated with HCl and/or DMSO to a final concentration of 0.1% v/v (20). Cytotoxic assay Cytotoxicity of phenobarbital and clofibrate were measured with concentrations of these two compounds ranging from 0 to 500 µg/ml, using the index of

1716

lactate dehydrogenase (LDH) leakage from damaged hepatocytes, and was expressed as a percentage of total cellular activity (24,25). Flow cytometry assays After incubation with or without TGFβ1 for 24 or 48 h in the presence of the respective compounds, phenobarbital and clofibrate, the medium was removed. Trypsin/EDTA (0.5 ml/plate) was added and the plates were incubated for 1– 2 min until detachment of the cells. For the analysis of DNA content, cells were stained with PI as previously described (26). The emitted fluorescence of the DNA–PI complex was analysed in a FACscan flow cytometer (Becton Dickinson) in the FL2 channel. A double discriminator module was used to distinguish between signals coming from a single nucleus and those products of nuclear aggregation. Data analysis was carried out by means of evaluation of single inputs (104 nuclei/assay). H2O2 production (peroxides) was monitored by flow cytometry using DCFH-DA (27). This dye is a stable non-polar compound that readily diffuses into cells. Once inside the cells, the acetate groups are cleaved from the molecule by intracellular esterases to yield DCFH, which is trapped within the cells. H2O2 or low molecular weight peroxides, produced by the cells in the presence of cellular peroxidases, oxidize DCFH to the highly fluorescent compound 29,7’-dichlorofluorescein (DCF). Thus, fluorescence intensity is proportional to the amount of peroxides produced by the cells. Following incubation, hepatocytes were washed with phosphate-buffered saline (PBS) and immediately detached with trysin/EDTA, then incubated with agitation for 30 min in 2 ml PBS containing 5 µM DCFH-DA at 37°C. The cells were washed twice with PBS to remove extracellular DCFH-DA, followed by analysis in a FACScan flow cytometer (Becton Dickinson) (excitation 488 nm, emission 525 nm). Because identification of non-viable and late apoptotic cells is essential to obtain accurate data, PI (10 µg/ml) was added to each tube 10 min before flow cytometry analysis to ensure that only living and early apoptotic cells were analysed. Analysis of hepatocyte nuclear morphology Primary hepatocyte monolayers were fixed in ice-cold methanol/acetic acid (3:1) for 5 min. For staining with Hoechst 33258, 5 µg/ml H2O of the chromophore was added, dishes were incubated for 10 min and then washed gently in distilled water. Cells were mounted in a solution of 20 mM citric acid, 50 mM disodium orthophosphate and 50% glycerol (pH 5.5) to achieve optimum fluorescence and were examined at a wavelength of excitation of 330–380 nm and emission of 460 nm using an Olympus IMT-2 microscope with fluorescence attachment (28,29). RNA isolation and northern blot analysis of Mn-SOD, Cu,Zn-SOD and catalase mRNAs For RNA isolation 1.53106 hepatocytes were lysed with guanidinium thiocyanate/phenol reagent (30). Total cellular RNA (20 µg) was submitted to northern blot analysis, being electrophoresed on 0.9% agarose gels containing 0.66 M formaldehyde, transferred to GeneScreen™ membranes and cross-linked to the membranes with UV light. Hybridization was in 0.25 mM NaHPO4, pH 7.2, 0.25 M NaCl, 100 µg/ml denatured salmon sperm DNA, 7% SDS and 50% deionized formamide, containing denatured 32P-labelled cDNA (106 c.p.m./ml) for 40 h at 42°C as described (31). cDNA labelling was carried out employing cDNA sequences of catalase, Cu,Zn-SOD and Mn-SOD (32) labelled with [α-32P]dCTP using the multiprimer DNA labelling system kit (Amersham). The resulting membranes were subjected to autoradiography for 1–3 days. Relative densities of the hybridization signals were determined by densitometric scanning of the autoradiograms in a laser densitometer (Molecular Dynamics, Sunnyvale, CA). Finally, the filters were hybridized with an 18S rRNA probe for RNA normalization. Analysis of the northern blot was performed three times. The variability in the measurement of fold increase in mRNA, after quantification by scanning densitometry from the filters, was ø15%. Statistical analysis The results were expressed as the means 6 SD of four experimental observations (four animals). Differences between groups were analysed by ANOVA followed by the Snedecor F(α 5 0.05) statistic. Student’s t-test was performed for statistical evaluation of (a) all values against control (no additions); (b) values obtained in the presence of TGFβ1 against TGFβ1; (c) comparisons between phenobarbital and clofibrate. P , 0.05 was considered as the level of significance.

Results Measurement of cytotoxicity of phenobarbital and clofibrate on primary hepatocyte cultures As a marker of cytotoxicity, LDH release was determined in hepatocytes incubated for 24 h in the presence of increased

ROS, antioxidant enzyme expression and apoptosis

Fig. 2. DNA histograms and scatter profiles of rat hepatocytes stained with the DNA-intercalating dye PI (26). Fluorescence of stained DNA was detected with the FL2 channel and plotted against the number of cells. |–M1–| defines the percentage of hypodiploid cells ,2N as an index of apoptosis. The experimental conditions were: cells (1.53106) were incubated for 24 or 48 h with or without TGFβ1 (2 ng/ml) in the presence of phenobarbital (PB) (200 µg/ml) or clofibrate (CF) (100 mg/ml). (A) Representative histograms of four independent experiments. (B) The hypodiploid (,2C) peaks were quantified and the results, expressed as a percentage of total hepatocytye population, are the means 6 SD of experimental observations from four rats. Student’s t-test was performed for statistical evaluations between all values against control (none) (a). P , 0.05 was considered the level of significance.

concentrations of both phenobarbital and clofibrate. According to the toxicity of each product, low toxicity concentrations were selected for further experiments, in order to avoid in vitro overlapping effects on cell death. In this respect 200 and 100 µg/ml were chosen for phenobarbital and clofibrate respectively as the highest concentrations that show very low toxicity. Figure 1 shows the percentages of LDH release in cultures incubated with increasing concentrations (0–500 µg/ml) of each compound. Parameters of apoptosis in primary hepatocyte cultures On the basis of a recently described investigation of the effect of PMA on TGFβ1-induced apoptosis (20), we performed a series of experiments to study the effects of phenobarbital and clofibrate on spontaneous and TGFβ1-induced apoptosis in hepatocyte cultures at 24 h incubation. Apoptosis was visualized by treatment of the hepatocyte cultures at the end of the incubation period with Hoechst 32258 and by fluorescence microscopy and quantified by flow cytometry prior to treatment of samples with PI (26). Figure 2A shows DNA multiploid histograms obtained by flow cytometry in which the fluorescence of stained DNA was detected with the FL2 channel and plotted against the number of cells. |–M1–| defines the percentage of hypodiploid cells ,2N as an apoptosis index. The experimental conditions were as follows: cells were incubated for 24 or 48 h with or

without TGFβ1 (2 ng/ml) in the presence of the appropriate concentrations of either phenobarbital (200 µg/ml) or clofibrate (100 µg/ml). Analysis of the cells showed the percentage of the hepatocyte population that underwent apoptosis, by quantifying the hypodiploid peak. These percentage values are shown in Figure 2B. At 24 h incubation the following results were obtained: apoptosis of cultured hepatocytes was 10 6 1.6 and 21.5 6 2.6% for control (spontaneous) and TGFβ1treated hepatocytes respectively. These values show that under our conditions TGFβ1 increased spontaneous apoptosis of hepatocytes in culture 2-fold. Either phenobarbital or clofibrate, when added to the culture medium at the appropriate concentrations, significantly diminished spontaneous apoptosis (to 55%, P , 0.05 and 26%, P , 0.05 respectively). However, only a slight inhibitory effect of both compounds on TGFβ1-induced apoptosis was detected (to 80 and 74% respectively). At 48 h incubation spontaneous apoptosis was significantly diminished to 36 and 30% by phenobarbital and clofibrate respectively, while the inhibitory effect of both compounds on TGFβ1induced apoptosis was to 87 and 78% respectively. Figure 3A–D shows photographs obtained by fluorescence microscopy in which apoptosis was visualized by observing the morphology of cultured hepatocytes incubated for 24 h in the absence (Figure 3A) or presence (Figure 3B) of TGFβ1 and exposed to TGFβ1 and phenobarbital (tumour promoter) 1717

C.Dı´ez-Ferna´ndez et al.

Fig. 3. Morphology of cultured rat hepatocytes following incubation for 24 h. Nuclei were stained in situ with the DNA-binding fluorochrome Hoechst 33258 and were visualized by fluorescence microscopy. (A) Control (no additions): (left) phase contrast image; (right) fluorescent nuclei stained with Hoechst 33258 (3400). (B) Cells were incubated with TGFβ1 (2 ng/ml). (C) Cells were incubated for 24 h in the presence of 2 ng/ml TGFβ1 and 200 µg/ml phenobarbital (3400). (D) Cells were incubated for 24 h in the presence of 2 ng/ml TGFβ1 and 100 µg/ml clofibrate (3400). Staining with Hoechst 33258 shows that in treated cells nuclei appear condensed or fragmented (apoptotic morphology).

(Figure 3C) or TGFβ1 plus clofibrate (peroxisome proliferator) (Figure 3D). Nuclei were stained in situ with the DNA-binding fluorochrome Hoechst 33258, a useful dye for revealing the morphology of nuclear chromatin (28,29). With this technique it was possible to visualize those nuclei containing a disintegrated chromatin structure. Figure 3A corresponds to a hepatocyte culture incubated for 24 h with no additions (control). On the left side of Figure 3A is shown the corresponding phase contrast image and on the right side is shown fluorescent nuclei stained with Hoechst 33258 (3400). Figure 3B corresponds to a hepatocyte culture incubated in the presence of TGFβ1 (2 ng/ml) for 24 h (3400). In dead apoptotic cells the condensed chromatin was dispersed into multiple small nuclear fragments or appeared condensed. Figure 3C corresponds to a hepatocyte culture incubated in the presence of TGFβ1 (2 ng/ml) and phenobarbital (200 µg/ml) for 24 h (3400). Figure 3D corre1718

sponds to a hepatocyte culture incubated in the presence of TGFβ1 (2 ng/ml) and clofibrate (100 µg/ml) for 24 h (3400). Intracellular concentration of peroxides in hepatocyte cultures incubated with either phenobarbital or clofibrate with or without TGFβ1 Figure 4 shows the intracellular concentration of peroxides measured by flow cytometry of the fluorescence emitted due to DCFH oxidation. Figure 4A shows histograms in which the fluorescence of DCF, detected with the FL1-H channel, is plotted against the number of cells. |–M1–| defines the peak of peroxides (intense fluorescence). Figure 4B shows the quantification in arbitrary units of the |–M1–| peak of Figure 4A. These results show that both phenobarbital and clofibrate increased endogenous levels of peroxides, but the increase was higher (146% versus control, P , 0.05) in cells cultured

ROS, antioxidant enzyme expression and apoptosis

an internal control (Figure 5B). The effects of phenobarbital, when added to hepatocyte cultures (200 µg/ml) and incubated for 24 h, were an increase of 2-fold in expression of catalase and a decrease to 50% in expression of both Mn-SOD and Cu,Zn-SOD. The effects of clofibrate (100 µg/ml) on hepatocyte cultures incubated for 24 h were a 4-fold increase in catalase expression associated with slight decreases in Cu,Zn-SOD and Mn-SOD expression. TGFβ1 by itself when added to hepatocyte cultures and incubated for 24 h affected the expression of SOD only slightly (76 and 112% versus control for Mn-SOD and Cu,Zn-SOD respectively), but sharply decreased expression of catalase, from 4.01 6 0.4 to 1.5 6 0.1 arbitrary units (37% versus control, P , 0.05). When phenobarbital and clofibrate were incubated in the presence of TGFβ1 the results obtained were a sharp decrease in gene expression of catalase, from 8.5 6 0.9 to 2.5 6 0.3 arbitrary units (29%, P , 0.05) in the case of phenobarbital and from 16.7 6 1.7 to 3.0 6 0.3 (17%, P , 0.05) in the case of clofibrate. Although in experiments with incubation with TGFβ1, addition of phenobarbital and clofibrate significantly enhanced expression of catalase (Figure 5B), from 1.5 6 0.1 to 2.5 6 0.2 arbitrary units (166%, P , 0.001) and from 1.5 6 0.1 to 3.0 6 0.3 arbitrary units (200%, P , 0.001) respectively, the level of catalase mRNA in hepatocyte cultures incubated without TGFβ1 was not reached. Discussion

Fig. 4. Levels of peroxides in hepatocyte cultures incubated in the presence of phenobarbital (PB) (200 µg/ml) or clofibrate (CF) (100 µg/ml) with or without TGFβ1 (2 ng/ml). Analysis of intracellular peroxides in hepatocyte cultures by measuring DCFH fluorescence by flow cytometry. (A) A representative histogram of the four independent experiments. (B) Results, expressed as arbitrary units, are the means 6 SD of experimental observations from four rats. Student’s t-test was performed for statistical evaluations between (a) all values against control (none); and (b) values obtained in the presence of TGFβ1 versus TGFβ1. P , 0.05 was considered the level of significance.

in the presence of phenobarbital than in the presence clofibrate (116% versus control). When cells were incubated in the presence of TGFβ1 the levels of peroxides were slightly higher when compared with the control (120%) and when either phenobarbital or clofibrate was added to the culture medium with TGFβ1 the increase was significantly enhanced in both cases (286% versus control, P , 0.05, and 182% versus control, P , 0.05, for phenobarbital and clofibrate respectively). Antioxidant gene expression in rat hepatocyte cultures incubated in the presence of the non-genotoxic carcinogens phenobarbital and clofibrate The results shown in Figure 5 relate to antioxidant enzyme gene expression of Mn-SOD, Cu,Zn-SOD and catalase in cultured rat hepatocytes when incubated in the presence of either phenobarbital or clofibrate with or without TGBβ1. The respective mRNAs were analysed by the northern blot technique (Figure 5A) and the bands were quantified by laser densitometry using hybridization with an 18S rRNA probe as

Apoptosis occurs at a negligible rate in normal liver, but various physiological conditions, diseases and xenobiotics can promote this form of cell death. Apoptosis in the liver also occurs after removal of mitogenic tumour promoters (33). A variety of compounds cause apoptosis in isolated rat hepatocyte cultures (16,34–36). Among them, TGFβ1 is an inducer of apoptosis in liver cell cultures (33) and inhibits DNA synthesis in regenerating liver (38–40). The mechanisms by which TGFβ1 induces cell death are still only partly understood; Kayanoki (19) determined that TGFβ1 produces a decrease in the mRNA of antioxidant enzymes such as Mn- and Cu,ZnSOD and catalase, indicating that inhibition of protein synthesis is involved in TGFβ1-induced apoptosis. In our results we found that TGFβ1 sharply decreased catalase expression, which led us to propose that this lack of antioxidant defence may be involved, either as a cause or as an effect, in the apoptogenic activity of TGFβ1. However, in a recent paper (41), the effect of inhibiting protein systhesis with cycloheximide on apoptosis induced by TGFβ1 was studied, which suggested that TGFβ1 induced synthesis of proteins required for GSH depletion and apoptosis. Depletion of GSH has been reported to be involved in TGFβ1 induction of apoptosis (42). On the other hand, TGFβ1 inhibits cell proliferation by inducing G1 phase arrest (43). The mechanism is by decreasing the activity of cdk4, by maintaining Rb protein in the unphosphorylated, growth suppressor state. The growth suppressor capacity of Rb depends on its ability to interact with and inhibit release of E2F. Therefore, control of E2F activity by Rb may be an important target of the TGFβ1 growth inhibitor signal (44). ROS could be derived from non-mitochondrial sources, such as microsomal cytochrome P450 and peroxisomal β-oxidation. Phenobarbital has been studied as an epigenetic liver tumour promoter in mice, since it induces microsomal P450 enzymes, stimulates hepatocyte cell proliferation, inhibits apoptosis and is not DNA reactive (41). After cessation of phenobarbital 1719

C.Dı´ez-Ferna´ndez et al.

Fig. 5. Northern blot analysis of Mn-SOD, Cu,Zn-SOD and catalase mRNAs in rat hepatocyte cultures incubated in the presence of phenobarbital (PB) or clofibrate (CF) with or without TGFβ1. (A) Total RNA prepared from rat hepatocytes following incubation with 200 µg/ml phenobarbital or 100 µg/ml clofibrate in the presence or absence of TGFβ1 (2 ng/ml) for 24 h. Northern blot analysis was carried out using rat Mn-SOD cDNA, rat Cu,Zn-SOD cDNA and rat catalase cDNA (32). 18S rRNA probe was used as internal standard for RNA normalization. Lane 1, control without any addition; lanes 2 and 5, the respective controls (DMSO) and TGFβ1; lanes 3 and 6, the respective samples incubated with phenobarbital and with phenobarbital plus TGFβ1; lanes 4 and 7, the respective samples incubated with clofibrate and with clofibrate plus TGFβ1. (B) Quantification of all mRNAs, following correction with 18S rRNA by densitometer scanning. Results, expressed as arbitrary units, are the means 6 SD of experimental observations from three rats. Student’s t-test was performed for statistical evaluations between (a) all values against control (none); (b) values obtained in the presence of TGFβ1 versus TGFβ1; and (c) comparisons between phenobarbital and clofibrate. P , 0.05 was considered the level of significance.

treatment, reversion to normal size involves apoptosis (33,45,46). Clofibrate, a peroxisomal proliferator, plays a role in inhibiting apoptosis and is an inducer of catalase (47). These two non-genotoxic carcinogens, whose concentrations were carefully selected to find the highest concentration that did not exhibit significant toxicity, were used in the present study to find an association between their inhibitory effect on apoptosis and gene expression of antioxidant defence systems. Our experiments show that both phenobarbital and clofibrate significantly decreased spontaneous hepatocyte apoptosis in culture and that both compounds were unable to significantly suppress TGFβ1-induced apoptosis. Although Bayly et al. (28) reported that apoptosis elicited in rat hepatocyte cultures by TGFβ1 was suppressed (40% reduction) by the tumour promoter and peroxisome proliferator nafenopin (using a 50 µM dose), Wo¨rner and Schrenk (1) found, under approximately the same experimental conditions, that the tumour promoters 2,3,7,8-tetrachlorodibenzo-p-dioxin and phenobarbital did not suppress TGFβ1-induced apoptosis. Peroxide generation under the experimental conditions of the present study was analysed with the fluorescent probe DCFH, which provided us with direct evidence of significant 1720

changes observed in the case of phenobarbital and clofibrate and, although slight increases were observed due to TGFβ1 itself, marked increases were detected when both phenobarbital and clofibrate were incubated together with this cytokine. In the present paper, we report direct evidence for increased generation of peroxides in cultured hepatocytes induced by both phenobarbital and clofibrate, which should drive apoptosis, and at the same time increased expression of catalase, which should inhibit apoptosis. This discrepancy led us to discuss how, up to now, attempts to consistently detect a clear requirement for ROS in apoptosis have not been convincing. Slater et al. (48) proposed that an accelerated efflux of GSH rather than any intracellular oxidation could be responsible for inducing apoptosis, since in the absence of any increase in oxidants, oxidative damage to proteins and lipids would result in apoptosis. Thus, we propose that the increases in peroxides are not the major cause of apoptosis. Outstanding changes were detected in gene expression of antioxidant enzymes. In this respect, phenobarbital decreased both SODs to 50% and increased catalase (~2-fold), while clofibrate did not apparently affect the level of mRNA of either SOD but markedly increased, .4-fold, that of catalase.

ROS, antioxidant enzyme expression and apoptosis

However, when hepatocytes were incubated with TGFβ1, no significant change was observed in expression of either SOD, while catalase expression was decreased (37% of control activity). Phenobarbital and clofibrate, when incubated in the presence of TGFβ1, produced significant increases (versus TGFβ1 alone) in catalase expression, but the values were significantly lowered when compared with their respective controls (phenobarbital or clofibrate). Catalase is present in the peroxisomal matrix of virtually all peroxisomes. Under our conditions, clofibrate, as a peroxisome proliferator, increased catalase expression .4-fold. As peroxisomal peroxidases generate hydrogen peroxide, the increased expression of catalase induced by clofibrate could be the consequence of an adaptative response. In the case of phenobarbital, the induction of catalase, .2-fold, is also adaptative, since the cytochrome P450 system is involved in hydrogen peroxide generation. The mechanism by which TGFβ1 suppresses catalase expression could be due to inhibition of the adaptative response against peroxide generation and neither phenobarbital nor clofibrate when incubated with TGFβ1 was able to counteract this inhibitory effect. We can also interpret these results on the basis that genes encoding for enzymes and factors involved in the response to oxidative stress could participate in the same way. The oxidative stress adaptative response involves expression of many genes, indicating that de novo synthesized proteins are required for this adaptation (8,9). The equilibrium between catalase (or other peroxidases) and SODs is extremely important for cells. A failure in removal of H2O2 produces a deleterious generation of hydroxyl radical, catalyzed by SOD (49). It has been also reported that at very low H2O2 exposures (3–15 µM H2O2) significant stimulation of cell growth was observed, while extremely high concentrations (1 mM H2O2) produced necrosis. Adaptation to oxidative stress seems to be an important mechanism by which mammalian cells can cope with the fluctuating levels of oxidants (9). The changes in catalase activity and gene expression in response to TGFβ1, phenobarbital and clofibrate could be an effect rather than a cause. However, speculation that catalase can afford protection against apoptosis is based on the fact that oxidative stress contributes signalling rather than terminal events in the cell death programme. ROS show some of the characteristics of second messengers and there is good evidence that ROS can promote transcription factor activity and regulate expression of a variety of genes (18). Catalase has been reported to protect against many different forms of apoptosis (50,51): addition of catalase significantly reduces nuclear fragmentation induced by a polyamine analogue (52); clofibrate, through induction of catalase, partially protected selenium-deficient cells from apoptosis after exposure to exogenous H2O2 (19). Finally, in the present study, three important conclusions should be emphasized: (i) catalase expression is remarkably increased by phenobarbital and clofibrate; (ii) catalase expression parallels the suppressive effect on apoptosis induced by phenobarbital and clofibrate and possibly involved in it; and (iii) the apoptogenic effect of TGFβ1 is associated with the inhibitory effect of this cytokine on catalase activity. Acknowledgements We thank Dr J.L.Tilly for generously providing us with the catalase, Cu,ZnSOD and Mn-SOD cDNA probes, Dolores Velasco for her technical assistance and Professor Erik Lundin for help in preparation of this manuscript. This study was supported by grant 95/0032/01 from the Fondo de Investigaciones

Sanitarias and PM 96/0010 from the Direccio´n General de Investigacio´n Cientı´fica y Te´cnica.

References 1. Wo¨rner,W. and Schrenk,D. (1996) Influence of liver tumor promoters on apoptosis in rat hepatocytes induced by 2-acetylaminofluorene, ultraviolet light or transforming growth factor beta 1. Cancer Res., 56, 1272–1278. 2. Mills,J.J., Jirtle,R.L. and Boyer, I.J. (1995) Mechanisms of liver tumor promotion. In Jirtke,R.L. (ed.) Liver Regeneration and Carcinogenesis. Academic Press, New York, NY, pp. 199–226. 3. Oberhammer,F. and Roberts,R.A. (1994) Apoptosis: a widespread process involved in liver adaptation and carcinogenesis. In Arias,I.M., Boyer,J.L., Fausto,N., Jakoby,W.B., Schachter,D.A. and Shafritz,D.A (eds) The Liver Biology and Pathobiology, 3rd edn. Raven Press, New York, NY, pp. 1547–1556. 4. Cerutti,P.A. (1985) Prooxidant states and tumor promotion. Science, 227, 375–381. 5. Cross,A.R. and Jones,O.T. (1991) Enzymic mechanisms of superoxide production. Biochim. Biophys. Acta, 1057, 281–298. 6. Watson,F., Robinson,J. and Edwards,S.W. (1991) Protein kinase Cdependent and -independent activation of the NADPH oxidase of human neutrophils. J. Biol. Chem., 266, 7432–7439. 7. Buttke,T.M. and Sandstrom,P.A. (1994) Oxidative stress as a mediator of apoptosis. Immunol. Today, 15, 7–10. 8. Crawford,D.R. and Davies,K.J.A. (1994) Adaptative response and oxidative stress. Environ. Health Perspect., 102 (suppl. 10), 25–28. 9. Wiese,A.G., Pacifici,R.E. and Davies,K.J.A. (1995) Transient adaptation to oxidative stress in mammalian cells. Arch. Biochem. Biophys., 318, 231–240. 10. Moody,D.E., Reddy,J.K., Lake,B.G., Popp,J.A. and Reese,D.H. (1991) Peroxisome proliferator and non-genotoxic carcinogens: commentary on a symposium. Fundam. Appl. Toxicol., 16, 233–248. 11. Lock,E.A., Mitchell,A.M. and Elcombe,C.R. (1989) Biochemical mechanisms of induction of hepatic peroxisome proliferation. Annu. Rev. Pharmacol. Toxicol., 29, 145–163. 12. Hassan,H.M. (1988) Biosynthesis and regulation of superoxide dismutases. Free Radical Biol. Med., 5, 377–385. 13. Amstad,P., Peskin,A., Shah,G., Mirault,M.-E., Moret,R., Zbinden,I. and Cerutti,P. (1991) The balance between Cu-Zn-superoxide dismutase and catalase affects the sensitivity of mouse epidermal cells to oxidative stress. Biochemistry, 30, 9305–9313. 14. Lennon,S.V., Martin,S.J. and Cotter,T.G. (1991) Dose-dependent induction of apoptosis in human tumour cell lines by widely diversing stimuli. Cell Proliferat., 24, 203–214. 15. Sandstrom,P.A. and Buttke,T.M. (1993) Autocrine production of extracellular catalase prevents apoptosis of the human CEM T-cell line in serum-free medium. Proc. Natl Acad. Sci. USA, 89, 4708–4712. 16. Shinagawa,M., Yoshioka,K., Kakumu,S., Wakita,T., Ishigawa,T., Itoh,Y. and Takayanagi,M. (1991) Apoptosis in cultured rat hepatocytes: the effects of tumor necrosis factor-α and interferon-γ. J. Pathol., 165, 247–253. 17. Kayanoki,Y., Fujii,J., Islam,K.N., Suzuki,K., Kawata,S., Matsuzawa,Y. and Taniguchi,N. (1996) The protective role of glutathione peroxidase in apoptosis induced by reactive oxygen species. J. Biochem. Tokyo, 119, 817–822. 18. Clement,M.V. and Stamenkovic,I. (1996) Superoxide anion is a natural inhibitor of Fas-mediated cell death. EMBO J., 15, 216–225. 19. Kayanoki,Y., Fujii,J., Suzuki,K., Kawata,S., Matsuzawa,Y. and Taniguchi,N. (1994) Suppression of antioxidative enzyme expression by transforming growth factor-β1 in rat hepatocytes. J. Biol. Chem., 269, 15488–15492. 20. Dı´ez-Ferna´ndez,C., Sanz,N., Wolf,A. and Cascales,M. (1998) Effect of phorbol ester (PMA) on antioxidant enzyme expression in TGFβ1 induced apoptosis in primary cultures of hepatocytes. Biofactors, 8, 65–71. 21. Cascales,M., Martı´n-Sanz,P., Alvarez,A., Sanchez-Pe´rez,M., Dı´ezFerna´ndez,C. and Bosca´,L. (1992) Isoenzymes of carbohydrate metabolism in primary cultures of hepatocytes from thioacetamide-induced rat liver necrosis. Response to growth factors. Hepatology, 16, 232–240. 22. Cascales,M., Alvarez,A., Gasco´,P., Ferna´ndez-Simon,L., Sanz,N. and Bosca´,L. (1994) Cocaine-induced liver injury in mice elicits specific changes in DNA ploidy and induces programmed cell death of hepatocytes. Hepatology, 20, 992–1001. 23. Seglen,O. (1993) Isolation of hepatocytes by collagenase perfusion. In Tyson,C.A. and Frazier,J.M. (eds) Methods in Toxicology. Academic Press, New York, NY, Vol. 1A, pp. 231–261. 24. Vassault,A. (1987) Lactate dehydrogenase. UV-methods with pyruvate and

1721

C.Dı´ez-Ferna´ndez et al. NADH. In Bermeyer,H.U. (ed.) Methods of Enzymatic Analysis. Verlag Chemie, Weinheim, Germany, Vol. 3, pp. 118–133. 25. Wolf,A., Trendelenburg,C.-F., Dı´ez-Ferna´ndez,C., Prieto,P., Houy,S., Trommer,W.E. and Cordier,A. (1997) Cyclosporine A-induced oxidative stress in rat hepatocytes. J. Pharmacol. Exp. Ther., 280, 1328–1334. 26. Vindelov,L.L., Christensen,I.J. and Nissen,N.I. (1983) A detergent trypsin method for the preparation of nuclei for flow cytometry. Cytometry, 3, 323–327. 27. Bass,D., Parce,J.W., DeChatelet,L.R., Szejda,P., Seeds,M.C. and Thomas,M. (1983) Flow cytometric studies of oxidative product formation by neutrophils: a graded response to membrane stimulation. J. Immunol., 130, 1910–1917. 28. Bayly,A.C, Roberts,R.A. and Dive,C. (1994) Suppression of liver cell apoptosis in vitro by the non-genotoxic hepatocarcinogen and peroxisome proliferator nafenopin. J. Cell Biol., 125, 197–203. 29. Gressner,A.M., Polzar,B., Lahme,B. and Mannherz,H.G. (1996) Induction of rat liver parenchymal cell apoptosis by hepatic miofibroblasts via transforming growth factor β. Hepatology, 23, 571–581. 30. Chomczynski,K. and Sacchi,N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction. Anal. Biochem., 162, 156–159. 31. Amasino,R.M, (1986) Acceleration of nucleic acid hybridization rate by polyethylene glycol. Anal. Biochem., 152, 304–307. 32. Tilly,J.L. and Tilly,K.I. (1995) Inhibitors of oxidative stress mimic the ability of follicle-stimulating hormone to suppress apoptosis in cultures of rat ovarian follicles. Endocrinology, 136, 242–252. 33. Bursch,W., Lauer,B., Timmerman-Trosiener,I., Barthel,G., Schuppler,J. and Schulte-Herman,R. (1984) Controlled cell death (apoptosis) of normal and putative preneoplastic cells in rat liver following withdrawal of tumor promoters. Carcinogenesis, 5, 453–458. 34. Boe,R., Gjertsen,B.T., Wintermeyr,O.K., Houge,G., Lanotti,M. and Doeskeland,S.O. (1991) The protein phosphatase inhibitor okadaic acid induces morphological changes typical of apoptosis in mammalian cells. Exp. Cell Res., 95, 237–246. 35. Sanchez,V., Lucas,M., Sanz,A. and Goberna,R. (1992) Decreased protein kinase C activity is associated with programmed cell death (apoptosis) in freshly isolated hepatocytes. Biosci. Rep., 12, 199–206. 36. Oberhammer,F., Fritsch,G., Schmied,M., Pavelka,M., Printz,D., Purchio, T., Lassmann,H. and Schulte-Hermann,R. (1993) Condensation of the chromatin at the membrane of an apoptotic nucleus is not associated with activation of an endonuclease. J. Cell Sci., 104, 317–320. 37. Bursch,W., Oberhammer,F., Jirtle,R.L., Askari,M., Sedivy,R., GraslKraupp,B., Purchio,A.F. and Schulte-Herman,R. (1993) Transforming growth factor β1 as a signal for induction of cell death by apoptosis. Br. J. Cancer, 67, 531–536. 38. Carr,B.I., Hayashi,I., Branum,E.L. and Moses,H.L. (1986) Inhibition of DNA synthesis in rat hepatocytes by platelet derived TGF-β1. Cancer Res., 46, 2330–2334. 39. Russell,W.E., Coffey,R.J., Quellette,A.J. and Moses,H.L. (1988) TGF reversibly inhibits the early proliferative response to partial hepatectomy. Proc. Natl Acad. Sci. USA, 85, 5126–5130. 40. Carr,B.I., Huang,T.H., Itakura,K., Noel,M. and Marceau,N. (1989) TGFbeta gene transcription in normal and neoplastic liver growth. J. Cell. Biochem., 39, 477–487. 41. Sa´nchez,A., Alvarez,A.M., Benito,M. and Fabregat,I. (1997) Cycloheximide prevents apoptosis, reactive oxygen species production and glutathione depletion induced by transforming growth factor β in fetal hepatocytes in primary culture. Hepatology, 26, 935–943. 42. van den Dobbelsteen,D.J., Stefan,C., Nobel,I., Schlegel,J., Cotgreave,I.A., Orrenius,S. and Slater,A.F.G. (1996) Rapid and specific efflux of reduced glutathione during apoptosis induced by anti-Fas/Apo-1 antibody. J. Biol. Chem., 271, 15420–15427. 43. Hannon,G.J. and Beah,D. (1994) p15ink4B is a potential effector of TGFβ-induced cell cycle arrest. Nature, 371, 257–261. 44. Schwarz,J.K., Bassing,C.H., Kovesdi,I., Datto,M.B., Blazing,M., George,S., Wang,X.-F. and Nevins,J.R. (1995) Expression of E2F1 transcription factor overcomes type β transforming growth factor-mediated growth suppression. Proc. Natl Acad. Sci. USA, 92, 483–487. 45. Sarraf,C.E., Horgan,M., Edwards,R.J. and Alison,M.R. (1997) Reversal of phenobarbital-induced hyperplasia and hypertrophy in the liver of lpr mice. Int. J. Exp. Pathol., 78, 49–56. 46. Whysner,J., Ross,P.M. and Williams,G.M. (1996) Phenobarbital, mechanistic data and risk assessment: enzyme induction, enhanced cell proliferation and tumor promotion. Pharmacol. Ther., 71, 153–191. 47. Stangl,H., Kovacs,W., Bock,P. and Kremser,K. (1995) Differential induction of peroxisomal enzymes by hypolipidemics in human (HepG2)

1722

and rat (MH1C1) hepatoma cell lines. Eur. J. Clin. Chem. Clin. Biochem., 33, 775–783. 48. Slater,A.F.G., Stefan,C., Nobel,I., van den Dobbelsteen,D.J. and Orrenius,S. (1996) Intracellular redox changes during apoptosis. Death Differentiat., 3, 57–62. 49. Amstad,P., Moret,R. and Cerutti,P. (1994) The balance between Cu,Znsuperoxide dismutase and catalase affects the sensitivity of mouse epidermal cells to oxidative stress. Biochem. J., 30, 9305–9313. 50. Furuke,K., Sasava,T., Ueda-Taniguchi,Y., Yamauchi,A., Inamoto,T., Yamaoka,Y., Masutani,H. and Yodoi,J. (1997) Pretreatment with catalase markedly inhibited 13CRA-induced apoptosis. Cancer Res., 57, 4916– 4923. 51. Xu,Y., Nguyen,Q., Lo,D.C. and Czaja,M.J. (1997) Cell death induced by c-myc expression was inhibited by the addition of catalase. J. Cell Physiol., 170, 192–199. 52. Ha,H.C., Woster,P.M., Yager,J.D. and Casero,R.A.Jr (1997) The role of polyamine catabolism in polyamine analogue-induced programmed cell death. Proc. Natl Acad. Sci. USA, 94, 11557–11562. Received on April 6, 1998; revised on June 2, 1998; accepted on June 15, 1998