Signaling Modulation of Bile Salt-Induced Necrosis
http://www.100md.com
《毒物学科学杂志》
ABSTRACT
Hydrophobic bile salts induce either necrosis or apoptosis depending on the severity of the injury caused by them. Since bile salt-induced apoptosis is influenced by Ca2+- and protein kinase-signaling pathways, and both necrosis and apoptosis share common initiating mechanisms, we analyzed whether these signaling cascades also influence bile salt-induced necrosis in isolated rat hepatocytes. Taurochenodeoxycholate (TCDC, 0.25–1.50 mM, 2 h) reduced, in a dose-dependent manner, the percentage of viable hepatocytes, and increased the release of the cytosolic enzyme, lactate dehydrogenase (LDH) and alanine aminotransferase (ALAT), and that of the plasma membrane enzyme, alkaline phosphatase (AP). The PKC inhibitors, H7 (100 μM) and chelerythrine (2.5 μM), both prevented significantly TCDC-induced necrosis. On the contrary, the PKA activator, dibutyryl-cAMP, exacerbated TCDC-induced cell damage in a dose-dependent manner; this effect was more likely due to cAMP-mediated PKA activation, as the PKA inhibitor, KT5720 (1 μM), counteracted this effect. Instead, the intracellular Ca2+ chelator, BAPTA/AM (20 μM), was without effect. TCDC (1 mM) increased lipid peroxidation from 0.7 ± 0.2 to 7.5 ± 0.9 nmol of malondialdehyde per mg of protein, p < 0.001; the addition of the free radical scavenger, diphenyl-p-phenylendiamine, completely blocked this increase and prevented significantly TCDC-induced necrosis. PKC inhibition induced only a slight attenuation of TCDC-induced lipid peroxidation. Possible mechanisms accounting for the modulatory effect of signal transduction pathways on TCDC-induced necrosis, including signaling influence on TCDC transport events and TCDC-induced oxidative stress, are discussed.
Key Words: cytosolic calcium; hepatocyte; hydrophobic bile salt; necrosis; oxidative stress; protein kinase.
INTRODUCTION
Under physiological conditions, bile salts (BSs) reach very low concentration both in hepatic tissue and in subcellular organelles, thanks to the high capability of the hepatocytes to extrude BSs into bile. Therefore, hepatocellular membranes are not exposed to high BS concentrations; otherwise, BSs would induce extensive structural and functional damage (Strange, 1981; Yousef et al., 1987).
Hydrophobic BSs induce either necrosis or apoptosis, depending on the severity of the injury (Benz et al., 1998). These mechanisms of cell death may initiate or aggravate the original hepatocellular damage in cholestatic liver diseases, which results in hepatocellular retention of BSs, along with other normal bile constituents.
BS-induced necrosis, as indicated by hepatocellular enzyme release and morphologic signs of membrane destruction, occurs at high BS concentrations, capable to surpass the critical micellar concentration (Benz et al., 1998); this is a prerequisite for conjugated BSs to exert their detergent effects, by actively incorporating membrane cholesterol and phospholipids into the micellar hydrophobic core (Coleman, 1987). In contrast, apoptosis is induced by low BS concentrations (Benz et al., 1998); this process involves vesicular trafficking of Fas death receptor from the cytosol to the cell membrane, their further oligomerization, and the initiation of the caspase-dependent death-signaling pathway (Sodeman et al., 2000). Considering the differential conditions at which BSs induce both cell death mechanisms, it seems likely that necrosis is the main mechanism of BS-induced cell death in severe cholestasis, whereas apoptosis would be predominant in less severe cholestatic conditions (Benz et al., 1998).
Several experimental studies where BSs were administered at high doses either into the whole rat (Drew and Priestly, 1979), the isolated perfused rat liver (Baumgartner et al., 1992; Yousef et al., 1987), isolated hepatocytes (Scholmerich et al., 1984), or membrane fractions (Scholmerich et al., 1984) have shown that these compounds induce extensive membrane damage, as assessed by the release of membrane lipids, intracellular protein and, in the first two cases, bile secretory failure. In these studies, BS efficiency to induce membrane damage was shown to depend critically on BS hydrophobicity; lipophilic BSs are highly cytotoxic, whereas hydrophilic BSs have low, if any, cytotoxic effect (Scholmerich et al., 1984).
The necrotic damage induced by hydrophobic BSs also depends on their capability to induce oxidative stress and membrane lipid peroxidation (Sokol et al., 1993, 1998, 2001). BSs have toxic effects on mitochondria by inducing formation of mitochondrial permeability transition (MPT) pores; this leads to collapse of the mitochondrial inner transmembrane potential, rupture of the outer membrane, blockage of the mitochondrial respiratory chain and, eventually, leakage of electrons with formation of reactive oxygen species (ROS) (Botla et al., 1995). Both oxidative stress and MPT formation interact positively with each other, as ROS favor MPT formation, and MPT induces further ROS formation, by impairing of mitochondrial respiration (Sokol et al., 2001). Mitochondrial dysfunction seems to be a common event in both BS-induced necrosis and apoptosis (Lemasters et al., 2002).
BSs are able to evocate a number of signal-transduction cascades, including the Ca2+-dependent and the PKC-dependent signaling pathways. On the other hand, changes in signaling balance modify transport events involved in the hepatic handling of BSs, thus affecting their steady-state cytosolic concentrations (for a review, see Bouscarel et al., 1999). It is therefore not surprising that some BS-deleterious effects are associated, at least in part, to signal-transduction imbalance. Indeed, glycochenodeoxycolate-induced apoptosis is associated with PKC activation (Jones et al., 1997). In addition to promote pro-apoptotic signaling pathways, BSs activate signaling cascade of opposite nature. For example, BSs activate the antiapoptotic signaling molecule, phosphoinositide-3-kinase (PI3K) (Rust et al., 2000). Similarly, cAMP prevents BS-induced apoptosis in a PKA- and a PI3K-dependent manner (Webster et al., 2002).
The recognition that both apoptosis and necrosis share common mechanisms of induction and that signaling pathways are involved in BS-induced apoptosis prompted us to assess the participation of signal-transduction cascades in BS-induced necrosis. For this purpose, we studied, in isolated rat hepatocytes, whether activation and/or inhibition of PKA-, PKC-, and Ca2+-dependent signaling cascades, three pathways known to influence BS-induced apoptosis, can modulate BS-induced necrotic damage as well. Since oxidative stress is thought to be a main mechanism of BS-induced hepatotoxicity (Sokol et al., 1993, 1998, 2001), we also ascertained whether these signal-transduction pathways influence the capability of BSs to induce lipid peroxidation, a key event in oxidative stress-induced damage.
MATERIALS AND METHODS
Materials. Collagenase type IV (from Clostridium histolyticum), bovine serum albumin (BSA) fraction V, trypan blue, sodium taurochenodeoxycholate (TCDC), N6,2'-O-dibutyryladenosine 3':5'-cyclic monophosphate sodium salt (dibutyryl-cAMP, DB-cAMP), digitonin, ethylene glycol-bis(2-aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA), and Triton X-100 were purchased from Sigma Chemical Co. (St. Louis, MO). KT5270, 1-(5-isoquinolinylsulfonyl)-2-methyl-piperazine (H7), chelerythrine chloride (Che) and 1,2-bis-(o-aminophenoxy)-ethene-N,N,N',N'-tetraacetate tetrakis-(acetomethyl ester) (BAPTA/AM) were from Alexis Co. (Bingham, Nottingham, U.K.). N,N'-diphenyl-p-phenylendiamine (DPPD) and fura-2 pentakis(acetomethyl) ester (Fura-2/AM) were obtained from Fluka Chemika (Steihem, Switzerland). 2',7'-Dichlorofluorescin diacetate (DCF-DA) was from Molecular Probes (Eugene, OR). All other reagents were of analytical grade.
Animals. Wistar male rats 120–130 days of age (300–350 g) were used throughout. Before the experiments, the animals were maintained on a standard diet (Purina Laboratory Rodent Chow 5001, Purina Mills, Inc., St. Louis, MO) and water ad libitum, and housed in a temperature- (21–23°C) and humidity- (45–50%) controlled room, under a constant 12 h-light, 12 h-dark cycle. All animals received humane care according to the criteria outlined in the ‘Guide for Care and Use of Laboratory Animals’ (National Institutes of Health publication 25–28, revised 1996).
Isolation of hepatocytes. Hepatocytes were isolated from livers by the collagenase perfusion technique, using a modification of the method of Berry and Friend (1969). Briefly, under sodium pentobarbital anesthesia (50 mg/kg body wt, ip), heparin was administered in the inferior vena cava (1500 U/kg of body weight), and a 14G catheter (Abbocath-T, Venisystem, Abbocath Ireland Ltd., Sligo, Ireland) was introduced in the portal vein. This was followed by a non-recirculant, portal perfusion of the liver for 10 min with a Ca2+-free, oxygenated (95% O2/5% CO2) Hanks' solution, pH = 7.47–7.50, supplemented with HEPES (3 g/l) and EGTA (0.24 g/l). The livers were perfused for a further 5-min period with the same solution without EGTA, which was supplemented with 1 mM MgSO4, 2.5 mM CaCl2 and collagenase type IV (4300 U/l). Finally, the livers were removed, and the cells isolated by mechanical dissociation by gently stirring with a glass stick for 3–4 min. Hepatocytes were further purified from non-parenchymal cells by low-speed centrifugation (50 x g, 2 min), followed by three consecutive washings in oxygenated Hanks' solution containing 2.5 mM CaCl2 and 5 mM Tris. The resulting preparation yielded 400–600 x 106 hepatocytes per liver of high viability (>90%), as assessed by the trypan blue exclusion test (Baur et al., 1975).
Treatments. Hepatocytes were resuspended in Krebs-Ringer-HEPES buffer, pH = 7.4, supplemented with 0.5% D-glucose and 3% BSA, to reach a final density of 2.5 x 105 cells/ml (unless otherwise indicated). The suspension was kept on ice no longer than 30 min before use. Four ml of this suspension were incubated without or with the hydrophobic BS, TCDC (0.25, 0.50, 1.00, and 1.50 mM) for 2 h in 20 ml in plastic beakers, immersed in a Dubnoff water bath at 37°C, under an atmosphere of 95% O2/5% CO2; TCDC was used as a tool, since it was shown to induce a dose- and time-dependent necrotoxic effect to hepatocytes, as apparent from loss of cell viability and leakage of cytosolic enzymes (Ohiwa et al., 1993; Sokol et al., 1993). The selection of the concentrations and the time of exposure of TCDC was based upon a previous study by Ohiwa et al. (1993), which showed that necrotoxic changes occur in hepatocytes at TCDC concentrations higher than 0.1 mM, and at exposure periods longer than 1 h.
The effect of pre-incubation of the hepatocytes with a number of signaling modulators was studied to ascertain the respective roles of PKA-, PKC-, and the Ca2+-dependent signal pathways in the necrotic effect of TCDC. The compounds tested, their biological effects, their final concentrations and the volume and kind of vehicle used for delivery are also indicated in Table 1. Hepatocytes were pre-incubated with the signaling modulators for 15 min, and then exposed to increasing TCDC concentrations for a further 2-h period. The signaling modulators were kept in the incubation medium throughout TCDC exposure.
In a separate set of experiments, we sought to determine whether addition of an antioxidant before exposure of TCDC attenuates TCDC-induced necrosis. For this purpose, hepatocytes were pretreated for 15 min with the antioxidant agent, DPPD (50 μM), before adding TCDC.
Analytical Methods
Assessment of hepatocellular integrity. At the end of the incubation period with TCDC, aliquots of hepatocytes were removed to assess cell viability, leakage of the cytosolic enzymes, lactate dehydrogenase (LDH) and alanine aminotransferase (ALAT), as well as the release of the plasma membrane-associated protein, alkaline phosphatase (AP).
Hepatocyte viability was assessed by the trypan blue exclusion test (Baur et al., 1975). For this purpose, 5 μl of cell suspension were added to 150 μl of trypan blue (1.3 g/l), dissolved in HEPES-supplemented Hanks' solution. Viability was calculated as the percentage of hepatocytes able to exclude the dye from their cell bodies, referred to the values recorded in control cells not exposed to TCDC.
Impairment of barrier properties of the hepatocellular plasma membrane is a chief event in cellular necrosis. To evaluate plasma membrane integrity, leakage of the cytosolic enzymes, LDH (EC 1.1.1.27) and ALAT (EC 2.6.1.2), into the incubation medium was assessed. These enzymes were determined spectrophotometrically in the incubation medium (Perkin Elmer UV/Vis Spectrometer Lambda2S, überlingen, Germany), by measuring the rate of NADH consumption at 340 nm using commercial, kinetic kits (Wiener Lab., Rosario, Argentina).
The capability of BSs to impair hepatocellular integrity is associated with their ability to remove membrane lipids, thus releasing plasma membrane-associated proteins into the incubation medium. We evaluated this process by studying the release of the plasma membrane protein, AP (EC 3.1.3.1), assessed by measuring the rate of the AP-catalyzed conversion of p-nitrophenyl phosphate to p-nitrophenol, using a commercial, kinetic kit (Wiener Lab., Rosario, Argentina).
Correction of the inhibition of these enzyme activities by the TCDC present in the reaction medium where enzyme activities had been assessed was carried out. For this purpose, a rat serum sample previously subjected to assessment of LDH, ALAT, and AP activity was used as an internal standard, by adding it into the reaction medium after the enzyme activity in the cell incubation medium had been measured. Serum sample addition increases abruptly the rate of NADH consumption (for LDH and ALAT) or p-nitrophenol apparition (for AP), as it becomes proportional to the sum of the enzyme activities of both extracellular medium and serum. The apparent enzyme activity of the serum sample in the reaction medium subjected to TCDC-induced inhibition () can be therefore calculated as the difference between the slope of NADH consumption (or p-nitrophenol apparition) before and after the serum sample is added into the reaction medium. Inhibition of the activity of the exogenously added enzymes (I) can be then calculated as
where AS is the initially measured enzyme activity in the serum sample, before adding it to the reaction medium. Actual enzyme activity in the extracellular medium (AECM) was then calculated by correcting the initially recorded activity () by the inhibition of the activity of the exogenously added enzyme as follows:
The activity of the enzymes released into the incubation medium was expressed as the percentage of total enzyme cell activity, to minimize influence of interindividual differences in enzyme cellular content. For this purpose, aliquots of the cellular suspension were treated with Triton X-100 (0.1% v/v), followed by centrifugation at 9000 x g for 2 min. Treatment with this tensioactive compound induces release of the total enzyme cellular content, both by dissolving membrane components and by releasing cytosolic enzymes due to loss of membrane barrier integrity.
Assessment of lipid peroxidation. ROS production in the presence of TCDC was assessed by measuring generation of lipid peroxidation products, by a modification of the thiobarbituric acid-reactive substances (TBARS) method (Buege and Aust, 1978). Briefly, 0.2 ml of a cell suspension containing 106 cells/ml were added to 0.5 ml of trichloroacetic acid (10% w/v) and 50 μl of the antioxidant, DPPD (60 μM). The resulting supernatant was then added to 1 ml of thiobarbituric acid (0.7% w/v), and heated in a water bath to 100°C for 15 min. After cooling and centrifugation (1000 x g for 10 min), absorbance was measured at 532 nm. A standard curve using 1,1,3,3-tetramethoxypropane, which is converted mol for mol into malondialdehyde (MDA), was routinely run. Protein content in the aliquots of cell suspension used for the assay was measured by the method of Lowry et al. (1951). TBARS were then expressed as nmol of MDA equivalents per mg of proteins.
Measurement of intracellular Ca2+ concentration ([Ca2+]i). The effect of the pre-treatment with the intracellular Ca2+ chelator, BAPTA/AM (20 μM, 15 min), on TCDC (1 mM)-induced increase in [Ca2+]i was assessed 15 min after the administration of the BS, using Fura-2/AM as a probe. For this purpose, 2 x 106 hepatocytes were resuspended at 37°C in 3 ml of a PBS buffer solution (pH = 7.4), containing 3 mM CaCl2, and then supplemented with 10 μM Fura-2/AM. Fluorescence intensities (F) were measured by using alternating excitation of 340 and 380 nm, and a fluorescence emission wavelength of 510 nm (3 nm bandwidth), using a spectrofluorometer Shimadzu RF-5301 PC. [Ca2+]i was calculated from the 340 nm/380 nm Fura-2/AM fluorescence intensity ratio (R), according to the following equation (Grynkiewicz et al., 1985):
where Kd is the dissociation constant of the complex, Fura-2/Ca2+ (135 nM), Rmax and Rmin are R values measured sequentially by addition of 100 μg/ml digitonin to the Fura-2-loaded cells before and after chelating Ca2+ with 5 mM EGTA/Tris solution (pH 8.7), respectively.
Statistical analysis. The results were expressed as mean ± SE. When requirements for parametric analysis were met, a Student's unpaired t-test was used for comparison between two groups; comparisons between groups that did not meet this criterion were made by using the Mann-Whitney's rank sum test. The Kruskal-Wallis' test (one-way ANOVA by ranks) was used when more than two groups were compared, followed by the Dunn's multiple-comparison, post hoc test for pairwise comparisons, if ANOVA reached any statistical significance among groups. P values lower than 0.05 were judged to be significant.
RESULTS
Effect of PKC Inhibitors on TCDC-Induced Hepatocellular Damage
The incubation of freshly isolated hepatocytes for 2 h in the absence of TCDC administration led to a very small decrease in hepatocellular viability (–5 ± 1%) and to a slight increase in the activity in the incubation medium of LDH, ALAT, and AP (+3 ± 1%, +5 ± 2%, and +4 ± 1%, respectively). In contrast, TCDC exposure resulted in a far more severe, dose-dependent decline in cells viability (Fig. 1). This decline was associated with a gradual increase in the release of LDH, ALAT, and AP, suggesting a dose-dependent impairment in plasma membrane integrity.
Activation of PKC by BSs has been well documented (Bouscarel et al., 1999). Therefore, we ascertained whether PKC activation plays a role in BS-induced toxicity. Figure 1 shows the effect of pretreatment of isolated hepatocytes with H7, a preferential PKC inhibitor which exhibits a slight PKA inhibitory effect as well. This inhibitor prevented partially the disruption of plasma membrane integrity induced by TCDC from a dose of the BS of 0.5 mM onwards, as visualized by the improvement in cell viability and a reduction in the release of the enzymes, LDH, ALAT, and AP.
To confirm whether the toxic effect induced by TCDC actually involves activation of the PKC-dependent pathways, we pretreated hepatocytes with the more specific PKC inhibitor, Che. Like H7, this compound partially attenuated the hepatocellular damage, as revealed by the improvement in cell viability and a reduction in the TCDC-induced release of LDH, ALAT, and AP into the incubation medium (Fig. 2). The protective effect of Che reached a significant difference in all the parameters of cell integrity evaluated at the TCDC concentration of 1 mM, although cell viability showed an improvement in a wider range (0.25–1 mM), with a tendency towards protection with the remaining parameters evaluated. Lack of involvement of PKA inhibition as an artifact in the evaluation of the protective effect of H7 was further confirmed by using the specific PKA inhibitor, KT5720, which failed to affect per se TCDC-induced necrosis (data not shown).
Effect of DB-cAMP on TCDC-Induced Hepatocellular Damage
DB-cAMP, a freely diffusible cAMP analogue which renders the second messenger, cAMP, by non-selective, esterase-mediated hydrolysis, is a selective PKA activator. We used this tool to ascertain whether activation of PKA-dependent pathways modulates BS-induced necrosis. As shown in Figure 3, DB-cAMP, administered at the concentration range of 0.05–1 mM, exacerbated in a concentration-dependent manner the hepatocellular injury induced by TCDC, as revealed by both a more marked decrease in cell viability and an exacerbation of the release of LDH, ALAT, and AP.
Since, apart from activating PKA, DB-cAMP evokes cytosolic Ca2+ elevations (Staddon and Hansford, 1986), we verified whether the exacerbation of the necrotoxic effect of TCDC depends on PKA activation, by using the specific PKA inhibitor, KT5720. As shown in Figure 4, the deleterious effect induced by DB-cAMP (0.05 mM) was dependent on PKA activation, since KT5720 attenuated significantly its harmful effect.
Role of Cytosolic Ca2+ Elevations in TCDC-Induced Hepatocellular Damage
BSs act as Ca2+ ionophores and induce both Ca2+ influx from the extracellular medium and mobilization of Ca2+ from intracellular stores (Bouscarel et al., 1999). To ascertain the role of Ca2+ elevations in BS-induced necrosis, we compared the capability of TCDC to induce hepatocellular damage in the presence or in the absence of the intracellular Ca2+-chelating agent, BAPTA/AM. As shown in Table 2, TCDC induced a 86% increase in [Ca2+]i. BAPTA/AM significantly reduced basal [Ca2+]i by 69%, and completely prevented the increase in this concentration induced by TCDC, maintaining Ca2+ levels at values even lower than controls, despite the presence of TCDC in the incubation medium. In spite of this, BAPTA/AM did not protect against TCDC necrotoxic effect at any of the BS concentrations tested
Role of Oxidative Stress in TCDC-Induced Hepatocellular Damage
Evidence has been provided for the involvement of oxidative stress in BS-induced hepatocellular death (Sokol et al., 1993, 2001). We tested in our experimental setting the contribution of this pathomechanism by assessing the protective effect of the antioxidant, DPPD, on TCDC toxicity.
As depicted in Table 3, TCDC, at the dose of 1 mM, increased more than one order of magnitude lipid peroxidation levels, as assessed by MDA generation. Pretreatment with the ROS scavenger, DPPD (50 μM), completely blocked this increase. As shown in Figure 6, DPPD prevented significantly the drop of cell viability and the release induced by TCDC of the three hepatocellular enzymes studied only at the higher concentration studied (1.5 mM), although a tendency towards protection was apparent at the TCDC concentration of1 mM, which did not reach statistical significance for LDH and ALAT.
Effect of PKC/PKA Modulators on TCDC-Induced Lipid Peroxidation
Since both PKC and PKA activation affected BS-induced hepatocellular damage, and this effect was partially dependent on BS-induced ROS formation (see above), we evaluated whether PKC and PKA modulators affected BS-induced lipid peroxidation.
As can be seen in Table 3, 1 mM TCDC increased more than one order of magnitude lipid peroxidation levels, as assessed by MDA generation. This increase was slightly, but significantly, counteracted by the PKC specific inhibitors, Che and staurosporine (SP); TCDC induced only a 8.1- and 8.6-fold increase in MDA generation in the presence of Che and SP treatment, respectively, as compared with the 10-fold increase for TCDC alone. On the contrary, the PKA activator, DB-cAMP, was without effect on this parameter.
DISCUSSION
The mechanisms by which BSs exert their hepatotoxic effect have not been completely elucidated. However, several lines of evidence suggest that ROS are generated and participate actively in its pathogenesis. Sokol et al. showed an association between BS-induced hepatotoxicity and the generation of ROS both in isolated rat hepatocytes (Sokol et al., 1993) and in the intact rat (Sokol et al., 1998), since BS-induced hepatotoxicity in both models was significantly attenuated by antioxidants; these previous findings were confirmed in this study by using the ROS scavenger, DPPD (see Fig. 6). These data, together with other works emphasizing the role of hydrophobic BSs as mitochondrial toxins due to their MPT-inducing properties (Botla et al., 1995), support a key role of oxidative stress in the pathogenesis of BS-induced necrosis. However, our observation that DPPD did not completely counteract the diminution in hepatocellular viability and the release of cytosolic and plasma membrane enzymes despite its full prevention of TCDC-induced lipid peroxidation, clearly suggests that pathomechanisms other than oxidative stress are involved. Although a full prevention of TCDC-induced necrotic damage had been previously reported to occur in the isolated rat hepatocytes treated with the antioxidant, D--tocopheryl succinate (Sokol et al., 1993), the TCDC concentration employed (0.2 mM) was lower than those used in this study (0.25–1.5 mM). Therefore, other mechanisms operating at higher concentrations may have been overlooked in that study.
A likely mechanism for this additional damage is the detergent action of lipophilic BSs on plasma membranes, a contention supported by their well-recognized tensioactive properties, derived from their amphoteric structure. Our results showing a progressive release into the incubation medium by increasing concentrations of TCDC of the plasma membrane-constitutive protein, AP, support this possibility. This enzyme binds to the plasma membrane via a glycan-phosphatidylinositol anchor, which interacts strongly with plasma membrane fatty acids (Low, 1987). At TCDC concentrations higher than its critical micellar concentration (4 μM), like that employed in this study (250–1500 μM), AP incorporates into TCDC micelles, which favors its stability and solubility in the extracellular aqueous medium (Coleman, 1987). Although we cannot rule out a contribution of AP from cells other than hepatocytes present in the cell preparation (e.g., cholangiocytes), this is likely to be negligible, as our isolation procedure yields hepatocyte preparations with high (>95%) purity (Berry and Friend, 1969).
A cross talk exists between both oxidative stress and signal-transduction pathways (Kamata and Hirata, 1999). Since BSs induce ROS generation (Sokol et al., 1993, 2001), it is not surprising that BS-induced hepatocellular damage is influenced by the cellular signaling status. In line with this view, previous studies carried out in primary hepatocyte cultures showed that apoptosis induced by glycochenodeoxycholate is counteracted by PKC inhibitors (Jones et al., 1997), suggesting that PKC-dependent signaling pathways play a key role in hydrophobic BS-induced apoptosis. Taking into account the existence of common mechanisms between BS-induced apoptosis and necrosis (e.g., MPT formation, oxidative stress), it is possible to infer a similar protective effect of PKC inhibitors on TCDC-induced necrosis. Our results agree with this view. H7, a preferential PKC inhibitor (although it can inhibit in certain extent PKA as well) prevented partially the necrotoxic damage induced by TCDC (see Fig. 1). Furthermore, the specific PKA inhibitor, KT5720, was without effect, suggesting that H7 protective effect depended exclusively on its capability to block PKC activity. This was supported further using the specific PKC inhibitor, Che, which mimicked H7 protective effect. The mechanisms by which PKC inhibition protects against BS-induced necrosis can be multifactorial in nature. Our results showing here that PKC inhibitors prevented partially TCDC-induced ROS formation suggest that mitochondrial ROS production is facilitated somewhat by PKC activation. In line with this observation, lipid peroxidation induced to isolated rat hepatocytes by the oxidizing compound, tert-butyl hydroperoxide (tBOOH) (von Ruecker et al., 1989), or by the heavy metal, cooper (Mudassar et al., 1992), was prevented by the PKC inhibitor, H7, and exacerbated by PKC activators. Furthermore, H7 attenuates tBOOH-induced LDH leakage (Mudassar et al., 1992).
The preventive effect of PKC inhibitors on TCDC-induced lipid peroxidation is, however, rather marginal, suggesting that other mechanisms must be involved. For example, PKC inhibition may favor TCDC efflux into the extracellular medium by stimulating the exocytic discharge of vesicles containing BSs, as PKC inhibits hepatocellular vesicular trafficking (Zegers and Hoekstra, 1998); this mechanism is thought to play a key role in BS overcharging conditions, like that occurring in our experimental setting (Erlinger, 1990). Furthermore, we have shown that PKC inhibitors blocked, and PKC activators stimulated, vesicle-mediated trafficking of vesicle-containing BS transporters towards the apical hepatocellular pole (Roma et al., 2000). In concordance with this, a study in isolated rat perfused liver showed that H7-induced PKC inhibition increased biliary excretion of TCDC at a concentration in the perfusate within the range used in this study (1 mM) (Nakazawa et al., 1996).
Hydrophobic BSs induce elevation of [Ca2+]i by an inositol (1,4,5)triphosphate-independent mechanism (Combettes et al., 1988). Conceptually, Ca2+ elevations can activate different Ca2+-dependent proteases, phospholipases and endonucleases, with the consequent hepatocellular damage. Furthermore, Ca2+-elevations lead to activation of Ca2+-dependent PKC isoforms, which may be involved in TCDC-induced damage as well (see above). Therefore, we analyzed here whether the Ca2+-chelating agent, BAPTA/AM, has any beneficial effect against TCDC-induced hepatocellular necrosis. Despite this compound completely prevented TCDC-induced elevations in [Ca2+]i, the capability of TCDC to induce hepatocellular damage was not attenuated (see Fig. 5). This result, however, should not be conclusively interpreted to indicate that intracellular Ca2+ plays no role in BS-induced cytotoxicity. The predominance of other deleterious mechanisms not influenced by Ca2+ levels, e.g., the detergent properties of TCDC on cellular membranes, may have masked its contribution, particularly shortly after TCDC injury. Indeed, TCDC induced, in the perfused rat liver model, an early (4 min), transient increase in LDH release, followed by a subsequent time- and dose-dependent elevation in this parameter; only the first peak was significantly suppressed by pretreatment with the Ca2+- channel blocker, Ni2+ (Hasegawa et al., 2003). It is therefore possible that the protective effects of intracellular Ca2+ chelation have been overlooked in our model, which evaluate events occurring later in the necrotic process. Nevertheless, Ca2+ elevations are not a prerequisite for some forms of hepatocellular necrosis to occur, like that following ATP depletion due to metabolic inhibition (Nieminen et al., 1988).
The second messenger, cAMP, an endogenous activator of the PKA-dependent signaling pathway, was shown to have protective, dose-dependent effects in several models of hepatotoxicity (Kasai et al., 1996). Although its hepatoprotective mechanism/s have not been completely elucidated, its stabilizing effect on intracellular membrane is probably involved (Ignarro et al., 1973). Furthermore, cAMP inhibits BS-induced apoptosis by blocking caspase activation and cytochrome c release (Webster et al., 2002). Paradoxically, cAMP exacerbated rather than prevented TCDC-induced necrosis in a dose-dependent fashion (see Fig. 3). Although how this signaling molecule aggravates TCDC-induced damage remains elusive, changes in TCDC hepatocellular bioavailability may be involved. cAMP was shown to stimulate the Na+-dependent BS uptake by the basolateral transporter, Na+-taurocholate cotransporting polypeptide (ntcp), in the isolated rat hepatocyte model. This was attributed to the capability of cAMP to hyperpolarize the plasma membrane via PKA-mediated Na+/K+-ATPase phosphorilation (Edmondson et al., 1985), and by stimulation of the PKA-dependent translocation of ntcp from an endosomal compartment to the sinusoidal membrane (Webster and Anwer, 1999). Based upon these previous observations, and our own results showing the PKA dependency of DB-cAMP-induced exacerbation of TCDC-induced necrosis (see Fig. 4), we proposed that putative protective effects of cAMP could have been masked by the simultaneous increase in TCDC intracellular concentration due to enhanced uptake. Our results are in apparent contradiction to a previous study showing a protective effect of cAMP-permeant analogues on TCDC-induced necrotoxicity in hepatocytes cultured overnight (Ohiwa et al., 1993). The explanation for these varied results may depend on the different experimental conditions employed. Whereas freshly isolated hepatocytes like those used in our study maintain an intact capability to take up BSs, this function decreases significantly under culture conditions (Follmann et al., 1990). Indeed, uptake of the model bile salt, taurocholate, decreases to approximately half of the value recorded in freshly isolated hepatocytes during a culture period compatible to that used by Ohiwa et al. (1993). Therefore, our approach more clearly reflects the situation of the hepatocytes in situ, at least in terms of bile salt uptake.
In summary, the present findings clearly show that modulation of PKC- and PKA-dependent signaling pathways can modify the capability of hydrophobic BSs to induce hepatocellular necrotoxicity; whereas PKC inhibition attenuates BS-induced necrosis, PKA activation exacerbates this harmful effect. This seems to occur either or both by modulating differentially the intracellular availability of these endogenous, harmful compounds or by affecting the pathomechanisms involved in their deleterious effects. Although it is not possible at this point to establish whether these data have relevance to the situation in vivo, our results encourage future application of signaling molecules in the prevention/cure of hepatopathies occurring with elevated hepatocellular levels of endogenous BSs.
ACKNOWLEDGMENTS
This work was funded by CONICET, Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT, PICT 05-08669), Beca ‘Ramón Carrillo-Arturo O?ativia 2003,’ from Ministerio de Salud de la Nación, and Fundación Antorchas (Subsidio de Emergencia para ‘Ex-Beneficiarios’ 14264/40).
REFERENCES
Baumgartner, U., Scholmerich, J., Leible, P., and Farthmann, E. H. (1992). Cholestasis, metabolism and biliary lipid secretion during perfusion of rat liver with different bile salts. Biochim. Biophys. Acta 1125, 142–149.
Drew, R., and Priestly, B. G. (1979). Choleretic and cholestatic effects of infused bile salts in the rat. Experientia 35, 809–811.
Zegers, M. M., and Hoekstra, D. (1998). Mechanisms and functional features of polarized membrane traffic in epithelial and hepatic cells. Biochem. J. 336(Pt. 2), 257–269.(Mariana Borgognone, Leona)
Hydrophobic bile salts induce either necrosis or apoptosis depending on the severity of the injury caused by them. Since bile salt-induced apoptosis is influenced by Ca2+- and protein kinase-signaling pathways, and both necrosis and apoptosis share common initiating mechanisms, we analyzed whether these signaling cascades also influence bile salt-induced necrosis in isolated rat hepatocytes. Taurochenodeoxycholate (TCDC, 0.25–1.50 mM, 2 h) reduced, in a dose-dependent manner, the percentage of viable hepatocytes, and increased the release of the cytosolic enzyme, lactate dehydrogenase (LDH) and alanine aminotransferase (ALAT), and that of the plasma membrane enzyme, alkaline phosphatase (AP). The PKC inhibitors, H7 (100 μM) and chelerythrine (2.5 μM), both prevented significantly TCDC-induced necrosis. On the contrary, the PKA activator, dibutyryl-cAMP, exacerbated TCDC-induced cell damage in a dose-dependent manner; this effect was more likely due to cAMP-mediated PKA activation, as the PKA inhibitor, KT5720 (1 μM), counteracted this effect. Instead, the intracellular Ca2+ chelator, BAPTA/AM (20 μM), was without effect. TCDC (1 mM) increased lipid peroxidation from 0.7 ± 0.2 to 7.5 ± 0.9 nmol of malondialdehyde per mg of protein, p < 0.001; the addition of the free radical scavenger, diphenyl-p-phenylendiamine, completely blocked this increase and prevented significantly TCDC-induced necrosis. PKC inhibition induced only a slight attenuation of TCDC-induced lipid peroxidation. Possible mechanisms accounting for the modulatory effect of signal transduction pathways on TCDC-induced necrosis, including signaling influence on TCDC transport events and TCDC-induced oxidative stress, are discussed.
Key Words: cytosolic calcium; hepatocyte; hydrophobic bile salt; necrosis; oxidative stress; protein kinase.
INTRODUCTION
Under physiological conditions, bile salts (BSs) reach very low concentration both in hepatic tissue and in subcellular organelles, thanks to the high capability of the hepatocytes to extrude BSs into bile. Therefore, hepatocellular membranes are not exposed to high BS concentrations; otherwise, BSs would induce extensive structural and functional damage (Strange, 1981; Yousef et al., 1987).
Hydrophobic BSs induce either necrosis or apoptosis, depending on the severity of the injury (Benz et al., 1998). These mechanisms of cell death may initiate or aggravate the original hepatocellular damage in cholestatic liver diseases, which results in hepatocellular retention of BSs, along with other normal bile constituents.
BS-induced necrosis, as indicated by hepatocellular enzyme release and morphologic signs of membrane destruction, occurs at high BS concentrations, capable to surpass the critical micellar concentration (Benz et al., 1998); this is a prerequisite for conjugated BSs to exert their detergent effects, by actively incorporating membrane cholesterol and phospholipids into the micellar hydrophobic core (Coleman, 1987). In contrast, apoptosis is induced by low BS concentrations (Benz et al., 1998); this process involves vesicular trafficking of Fas death receptor from the cytosol to the cell membrane, their further oligomerization, and the initiation of the caspase-dependent death-signaling pathway (Sodeman et al., 2000). Considering the differential conditions at which BSs induce both cell death mechanisms, it seems likely that necrosis is the main mechanism of BS-induced cell death in severe cholestasis, whereas apoptosis would be predominant in less severe cholestatic conditions (Benz et al., 1998).
Several experimental studies where BSs were administered at high doses either into the whole rat (Drew and Priestly, 1979), the isolated perfused rat liver (Baumgartner et al., 1992; Yousef et al., 1987), isolated hepatocytes (Scholmerich et al., 1984), or membrane fractions (Scholmerich et al., 1984) have shown that these compounds induce extensive membrane damage, as assessed by the release of membrane lipids, intracellular protein and, in the first two cases, bile secretory failure. In these studies, BS efficiency to induce membrane damage was shown to depend critically on BS hydrophobicity; lipophilic BSs are highly cytotoxic, whereas hydrophilic BSs have low, if any, cytotoxic effect (Scholmerich et al., 1984).
The necrotic damage induced by hydrophobic BSs also depends on their capability to induce oxidative stress and membrane lipid peroxidation (Sokol et al., 1993, 1998, 2001). BSs have toxic effects on mitochondria by inducing formation of mitochondrial permeability transition (MPT) pores; this leads to collapse of the mitochondrial inner transmembrane potential, rupture of the outer membrane, blockage of the mitochondrial respiratory chain and, eventually, leakage of electrons with formation of reactive oxygen species (ROS) (Botla et al., 1995). Both oxidative stress and MPT formation interact positively with each other, as ROS favor MPT formation, and MPT induces further ROS formation, by impairing of mitochondrial respiration (Sokol et al., 2001). Mitochondrial dysfunction seems to be a common event in both BS-induced necrosis and apoptosis (Lemasters et al., 2002).
BSs are able to evocate a number of signal-transduction cascades, including the Ca2+-dependent and the PKC-dependent signaling pathways. On the other hand, changes in signaling balance modify transport events involved in the hepatic handling of BSs, thus affecting their steady-state cytosolic concentrations (for a review, see Bouscarel et al., 1999). It is therefore not surprising that some BS-deleterious effects are associated, at least in part, to signal-transduction imbalance. Indeed, glycochenodeoxycolate-induced apoptosis is associated with PKC activation (Jones et al., 1997). In addition to promote pro-apoptotic signaling pathways, BSs activate signaling cascade of opposite nature. For example, BSs activate the antiapoptotic signaling molecule, phosphoinositide-3-kinase (PI3K) (Rust et al., 2000). Similarly, cAMP prevents BS-induced apoptosis in a PKA- and a PI3K-dependent manner (Webster et al., 2002).
The recognition that both apoptosis and necrosis share common mechanisms of induction and that signaling pathways are involved in BS-induced apoptosis prompted us to assess the participation of signal-transduction cascades in BS-induced necrosis. For this purpose, we studied, in isolated rat hepatocytes, whether activation and/or inhibition of PKA-, PKC-, and Ca2+-dependent signaling cascades, three pathways known to influence BS-induced apoptosis, can modulate BS-induced necrotic damage as well. Since oxidative stress is thought to be a main mechanism of BS-induced hepatotoxicity (Sokol et al., 1993, 1998, 2001), we also ascertained whether these signal-transduction pathways influence the capability of BSs to induce lipid peroxidation, a key event in oxidative stress-induced damage.
MATERIALS AND METHODS
Materials. Collagenase type IV (from Clostridium histolyticum), bovine serum albumin (BSA) fraction V, trypan blue, sodium taurochenodeoxycholate (TCDC), N6,2'-O-dibutyryladenosine 3':5'-cyclic monophosphate sodium salt (dibutyryl-cAMP, DB-cAMP), digitonin, ethylene glycol-bis(2-aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA), and Triton X-100 were purchased from Sigma Chemical Co. (St. Louis, MO). KT5270, 1-(5-isoquinolinylsulfonyl)-2-methyl-piperazine (H7), chelerythrine chloride (Che) and 1,2-bis-(o-aminophenoxy)-ethene-N,N,N',N'-tetraacetate tetrakis-(acetomethyl ester) (BAPTA/AM) were from Alexis Co. (Bingham, Nottingham, U.K.). N,N'-diphenyl-p-phenylendiamine (DPPD) and fura-2 pentakis(acetomethyl) ester (Fura-2/AM) were obtained from Fluka Chemika (Steihem, Switzerland). 2',7'-Dichlorofluorescin diacetate (DCF-DA) was from Molecular Probes (Eugene, OR). All other reagents were of analytical grade.
Animals. Wistar male rats 120–130 days of age (300–350 g) were used throughout. Before the experiments, the animals were maintained on a standard diet (Purina Laboratory Rodent Chow 5001, Purina Mills, Inc., St. Louis, MO) and water ad libitum, and housed in a temperature- (21–23°C) and humidity- (45–50%) controlled room, under a constant 12 h-light, 12 h-dark cycle. All animals received humane care according to the criteria outlined in the ‘Guide for Care and Use of Laboratory Animals’ (National Institutes of Health publication 25–28, revised 1996).
Isolation of hepatocytes. Hepatocytes were isolated from livers by the collagenase perfusion technique, using a modification of the method of Berry and Friend (1969). Briefly, under sodium pentobarbital anesthesia (50 mg/kg body wt, ip), heparin was administered in the inferior vena cava (1500 U/kg of body weight), and a 14G catheter (Abbocath-T, Venisystem, Abbocath Ireland Ltd., Sligo, Ireland) was introduced in the portal vein. This was followed by a non-recirculant, portal perfusion of the liver for 10 min with a Ca2+-free, oxygenated (95% O2/5% CO2) Hanks' solution, pH = 7.47–7.50, supplemented with HEPES (3 g/l) and EGTA (0.24 g/l). The livers were perfused for a further 5-min period with the same solution without EGTA, which was supplemented with 1 mM MgSO4, 2.5 mM CaCl2 and collagenase type IV (4300 U/l). Finally, the livers were removed, and the cells isolated by mechanical dissociation by gently stirring with a glass stick for 3–4 min. Hepatocytes were further purified from non-parenchymal cells by low-speed centrifugation (50 x g, 2 min), followed by three consecutive washings in oxygenated Hanks' solution containing 2.5 mM CaCl2 and 5 mM Tris. The resulting preparation yielded 400–600 x 106 hepatocytes per liver of high viability (>90%), as assessed by the trypan blue exclusion test (Baur et al., 1975).
Treatments. Hepatocytes were resuspended in Krebs-Ringer-HEPES buffer, pH = 7.4, supplemented with 0.5% D-glucose and 3% BSA, to reach a final density of 2.5 x 105 cells/ml (unless otherwise indicated). The suspension was kept on ice no longer than 30 min before use. Four ml of this suspension were incubated without or with the hydrophobic BS, TCDC (0.25, 0.50, 1.00, and 1.50 mM) for 2 h in 20 ml in plastic beakers, immersed in a Dubnoff water bath at 37°C, under an atmosphere of 95% O2/5% CO2; TCDC was used as a tool, since it was shown to induce a dose- and time-dependent necrotoxic effect to hepatocytes, as apparent from loss of cell viability and leakage of cytosolic enzymes (Ohiwa et al., 1993; Sokol et al., 1993). The selection of the concentrations and the time of exposure of TCDC was based upon a previous study by Ohiwa et al. (1993), which showed that necrotoxic changes occur in hepatocytes at TCDC concentrations higher than 0.1 mM, and at exposure periods longer than 1 h.
The effect of pre-incubation of the hepatocytes with a number of signaling modulators was studied to ascertain the respective roles of PKA-, PKC-, and the Ca2+-dependent signal pathways in the necrotic effect of TCDC. The compounds tested, their biological effects, their final concentrations and the volume and kind of vehicle used for delivery are also indicated in Table 1. Hepatocytes were pre-incubated with the signaling modulators for 15 min, and then exposed to increasing TCDC concentrations for a further 2-h period. The signaling modulators were kept in the incubation medium throughout TCDC exposure.
In a separate set of experiments, we sought to determine whether addition of an antioxidant before exposure of TCDC attenuates TCDC-induced necrosis. For this purpose, hepatocytes were pretreated for 15 min with the antioxidant agent, DPPD (50 μM), before adding TCDC.
Analytical Methods
Assessment of hepatocellular integrity. At the end of the incubation period with TCDC, aliquots of hepatocytes were removed to assess cell viability, leakage of the cytosolic enzymes, lactate dehydrogenase (LDH) and alanine aminotransferase (ALAT), as well as the release of the plasma membrane-associated protein, alkaline phosphatase (AP).
Hepatocyte viability was assessed by the trypan blue exclusion test (Baur et al., 1975). For this purpose, 5 μl of cell suspension were added to 150 μl of trypan blue (1.3 g/l), dissolved in HEPES-supplemented Hanks' solution. Viability was calculated as the percentage of hepatocytes able to exclude the dye from their cell bodies, referred to the values recorded in control cells not exposed to TCDC.
Impairment of barrier properties of the hepatocellular plasma membrane is a chief event in cellular necrosis. To evaluate plasma membrane integrity, leakage of the cytosolic enzymes, LDH (EC 1.1.1.27) and ALAT (EC 2.6.1.2), into the incubation medium was assessed. These enzymes were determined spectrophotometrically in the incubation medium (Perkin Elmer UV/Vis Spectrometer Lambda2S, überlingen, Germany), by measuring the rate of NADH consumption at 340 nm using commercial, kinetic kits (Wiener Lab., Rosario, Argentina).
The capability of BSs to impair hepatocellular integrity is associated with their ability to remove membrane lipids, thus releasing plasma membrane-associated proteins into the incubation medium. We evaluated this process by studying the release of the plasma membrane protein, AP (EC 3.1.3.1), assessed by measuring the rate of the AP-catalyzed conversion of p-nitrophenyl phosphate to p-nitrophenol, using a commercial, kinetic kit (Wiener Lab., Rosario, Argentina).
Correction of the inhibition of these enzyme activities by the TCDC present in the reaction medium where enzyme activities had been assessed was carried out. For this purpose, a rat serum sample previously subjected to assessment of LDH, ALAT, and AP activity was used as an internal standard, by adding it into the reaction medium after the enzyme activity in the cell incubation medium had been measured. Serum sample addition increases abruptly the rate of NADH consumption (for LDH and ALAT) or p-nitrophenol apparition (for AP), as it becomes proportional to the sum of the enzyme activities of both extracellular medium and serum. The apparent enzyme activity of the serum sample in the reaction medium subjected to TCDC-induced inhibition () can be therefore calculated as the difference between the slope of NADH consumption (or p-nitrophenol apparition) before and after the serum sample is added into the reaction medium. Inhibition of the activity of the exogenously added enzymes (I) can be then calculated as
where AS is the initially measured enzyme activity in the serum sample, before adding it to the reaction medium. Actual enzyme activity in the extracellular medium (AECM) was then calculated by correcting the initially recorded activity () by the inhibition of the activity of the exogenously added enzyme as follows:
The activity of the enzymes released into the incubation medium was expressed as the percentage of total enzyme cell activity, to minimize influence of interindividual differences in enzyme cellular content. For this purpose, aliquots of the cellular suspension were treated with Triton X-100 (0.1% v/v), followed by centrifugation at 9000 x g for 2 min. Treatment with this tensioactive compound induces release of the total enzyme cellular content, both by dissolving membrane components and by releasing cytosolic enzymes due to loss of membrane barrier integrity.
Assessment of lipid peroxidation. ROS production in the presence of TCDC was assessed by measuring generation of lipid peroxidation products, by a modification of the thiobarbituric acid-reactive substances (TBARS) method (Buege and Aust, 1978). Briefly, 0.2 ml of a cell suspension containing 106 cells/ml were added to 0.5 ml of trichloroacetic acid (10% w/v) and 50 μl of the antioxidant, DPPD (60 μM). The resulting supernatant was then added to 1 ml of thiobarbituric acid (0.7% w/v), and heated in a water bath to 100°C for 15 min. After cooling and centrifugation (1000 x g for 10 min), absorbance was measured at 532 nm. A standard curve using 1,1,3,3-tetramethoxypropane, which is converted mol for mol into malondialdehyde (MDA), was routinely run. Protein content in the aliquots of cell suspension used for the assay was measured by the method of Lowry et al. (1951). TBARS were then expressed as nmol of MDA equivalents per mg of proteins.
Measurement of intracellular Ca2+ concentration ([Ca2+]i). The effect of the pre-treatment with the intracellular Ca2+ chelator, BAPTA/AM (20 μM, 15 min), on TCDC (1 mM)-induced increase in [Ca2+]i was assessed 15 min after the administration of the BS, using Fura-2/AM as a probe. For this purpose, 2 x 106 hepatocytes were resuspended at 37°C in 3 ml of a PBS buffer solution (pH = 7.4), containing 3 mM CaCl2, and then supplemented with 10 μM Fura-2/AM. Fluorescence intensities (F) were measured by using alternating excitation of 340 and 380 nm, and a fluorescence emission wavelength of 510 nm (3 nm bandwidth), using a spectrofluorometer Shimadzu RF-5301 PC. [Ca2+]i was calculated from the 340 nm/380 nm Fura-2/AM fluorescence intensity ratio (R), according to the following equation (Grynkiewicz et al., 1985):
where Kd is the dissociation constant of the complex, Fura-2/Ca2+ (135 nM), Rmax and Rmin are R values measured sequentially by addition of 100 μg/ml digitonin to the Fura-2-loaded cells before and after chelating Ca2+ with 5 mM EGTA/Tris solution (pH 8.7), respectively.
Statistical analysis. The results were expressed as mean ± SE. When requirements for parametric analysis were met, a Student's unpaired t-test was used for comparison between two groups; comparisons between groups that did not meet this criterion were made by using the Mann-Whitney's rank sum test. The Kruskal-Wallis' test (one-way ANOVA by ranks) was used when more than two groups were compared, followed by the Dunn's multiple-comparison, post hoc test for pairwise comparisons, if ANOVA reached any statistical significance among groups. P values lower than 0.05 were judged to be significant.
RESULTS
Effect of PKC Inhibitors on TCDC-Induced Hepatocellular Damage
The incubation of freshly isolated hepatocytes for 2 h in the absence of TCDC administration led to a very small decrease in hepatocellular viability (–5 ± 1%) and to a slight increase in the activity in the incubation medium of LDH, ALAT, and AP (+3 ± 1%, +5 ± 2%, and +4 ± 1%, respectively). In contrast, TCDC exposure resulted in a far more severe, dose-dependent decline in cells viability (Fig. 1). This decline was associated with a gradual increase in the release of LDH, ALAT, and AP, suggesting a dose-dependent impairment in plasma membrane integrity.
Activation of PKC by BSs has been well documented (Bouscarel et al., 1999). Therefore, we ascertained whether PKC activation plays a role in BS-induced toxicity. Figure 1 shows the effect of pretreatment of isolated hepatocytes with H7, a preferential PKC inhibitor which exhibits a slight PKA inhibitory effect as well. This inhibitor prevented partially the disruption of plasma membrane integrity induced by TCDC from a dose of the BS of 0.5 mM onwards, as visualized by the improvement in cell viability and a reduction in the release of the enzymes, LDH, ALAT, and AP.
To confirm whether the toxic effect induced by TCDC actually involves activation of the PKC-dependent pathways, we pretreated hepatocytes with the more specific PKC inhibitor, Che. Like H7, this compound partially attenuated the hepatocellular damage, as revealed by the improvement in cell viability and a reduction in the TCDC-induced release of LDH, ALAT, and AP into the incubation medium (Fig. 2). The protective effect of Che reached a significant difference in all the parameters of cell integrity evaluated at the TCDC concentration of 1 mM, although cell viability showed an improvement in a wider range (0.25–1 mM), with a tendency towards protection with the remaining parameters evaluated. Lack of involvement of PKA inhibition as an artifact in the evaluation of the protective effect of H7 was further confirmed by using the specific PKA inhibitor, KT5720, which failed to affect per se TCDC-induced necrosis (data not shown).
Effect of DB-cAMP on TCDC-Induced Hepatocellular Damage
DB-cAMP, a freely diffusible cAMP analogue which renders the second messenger, cAMP, by non-selective, esterase-mediated hydrolysis, is a selective PKA activator. We used this tool to ascertain whether activation of PKA-dependent pathways modulates BS-induced necrosis. As shown in Figure 3, DB-cAMP, administered at the concentration range of 0.05–1 mM, exacerbated in a concentration-dependent manner the hepatocellular injury induced by TCDC, as revealed by both a more marked decrease in cell viability and an exacerbation of the release of LDH, ALAT, and AP.
Since, apart from activating PKA, DB-cAMP evokes cytosolic Ca2+ elevations (Staddon and Hansford, 1986), we verified whether the exacerbation of the necrotoxic effect of TCDC depends on PKA activation, by using the specific PKA inhibitor, KT5720. As shown in Figure 4, the deleterious effect induced by DB-cAMP (0.05 mM) was dependent on PKA activation, since KT5720 attenuated significantly its harmful effect.
Role of Cytosolic Ca2+ Elevations in TCDC-Induced Hepatocellular Damage
BSs act as Ca2+ ionophores and induce both Ca2+ influx from the extracellular medium and mobilization of Ca2+ from intracellular stores (Bouscarel et al., 1999). To ascertain the role of Ca2+ elevations in BS-induced necrosis, we compared the capability of TCDC to induce hepatocellular damage in the presence or in the absence of the intracellular Ca2+-chelating agent, BAPTA/AM. As shown in Table 2, TCDC induced a 86% increase in [Ca2+]i. BAPTA/AM significantly reduced basal [Ca2+]i by 69%, and completely prevented the increase in this concentration induced by TCDC, maintaining Ca2+ levels at values even lower than controls, despite the presence of TCDC in the incubation medium. In spite of this, BAPTA/AM did not protect against TCDC necrotoxic effect at any of the BS concentrations tested
Role of Oxidative Stress in TCDC-Induced Hepatocellular Damage
Evidence has been provided for the involvement of oxidative stress in BS-induced hepatocellular death (Sokol et al., 1993, 2001). We tested in our experimental setting the contribution of this pathomechanism by assessing the protective effect of the antioxidant, DPPD, on TCDC toxicity.
As depicted in Table 3, TCDC, at the dose of 1 mM, increased more than one order of magnitude lipid peroxidation levels, as assessed by MDA generation. Pretreatment with the ROS scavenger, DPPD (50 μM), completely blocked this increase. As shown in Figure 6, DPPD prevented significantly the drop of cell viability and the release induced by TCDC of the three hepatocellular enzymes studied only at the higher concentration studied (1.5 mM), although a tendency towards protection was apparent at the TCDC concentration of1 mM, which did not reach statistical significance for LDH and ALAT.
Effect of PKC/PKA Modulators on TCDC-Induced Lipid Peroxidation
Since both PKC and PKA activation affected BS-induced hepatocellular damage, and this effect was partially dependent on BS-induced ROS formation (see above), we evaluated whether PKC and PKA modulators affected BS-induced lipid peroxidation.
As can be seen in Table 3, 1 mM TCDC increased more than one order of magnitude lipid peroxidation levels, as assessed by MDA generation. This increase was slightly, but significantly, counteracted by the PKC specific inhibitors, Che and staurosporine (SP); TCDC induced only a 8.1- and 8.6-fold increase in MDA generation in the presence of Che and SP treatment, respectively, as compared with the 10-fold increase for TCDC alone. On the contrary, the PKA activator, DB-cAMP, was without effect on this parameter.
DISCUSSION
The mechanisms by which BSs exert their hepatotoxic effect have not been completely elucidated. However, several lines of evidence suggest that ROS are generated and participate actively in its pathogenesis. Sokol et al. showed an association between BS-induced hepatotoxicity and the generation of ROS both in isolated rat hepatocytes (Sokol et al., 1993) and in the intact rat (Sokol et al., 1998), since BS-induced hepatotoxicity in both models was significantly attenuated by antioxidants; these previous findings were confirmed in this study by using the ROS scavenger, DPPD (see Fig. 6). These data, together with other works emphasizing the role of hydrophobic BSs as mitochondrial toxins due to their MPT-inducing properties (Botla et al., 1995), support a key role of oxidative stress in the pathogenesis of BS-induced necrosis. However, our observation that DPPD did not completely counteract the diminution in hepatocellular viability and the release of cytosolic and plasma membrane enzymes despite its full prevention of TCDC-induced lipid peroxidation, clearly suggests that pathomechanisms other than oxidative stress are involved. Although a full prevention of TCDC-induced necrotic damage had been previously reported to occur in the isolated rat hepatocytes treated with the antioxidant, D--tocopheryl succinate (Sokol et al., 1993), the TCDC concentration employed (0.2 mM) was lower than those used in this study (0.25–1.5 mM). Therefore, other mechanisms operating at higher concentrations may have been overlooked in that study.
A likely mechanism for this additional damage is the detergent action of lipophilic BSs on plasma membranes, a contention supported by their well-recognized tensioactive properties, derived from their amphoteric structure. Our results showing a progressive release into the incubation medium by increasing concentrations of TCDC of the plasma membrane-constitutive protein, AP, support this possibility. This enzyme binds to the plasma membrane via a glycan-phosphatidylinositol anchor, which interacts strongly with plasma membrane fatty acids (Low, 1987). At TCDC concentrations higher than its critical micellar concentration (4 μM), like that employed in this study (250–1500 μM), AP incorporates into TCDC micelles, which favors its stability and solubility in the extracellular aqueous medium (Coleman, 1987). Although we cannot rule out a contribution of AP from cells other than hepatocytes present in the cell preparation (e.g., cholangiocytes), this is likely to be negligible, as our isolation procedure yields hepatocyte preparations with high (>95%) purity (Berry and Friend, 1969).
A cross talk exists between both oxidative stress and signal-transduction pathways (Kamata and Hirata, 1999). Since BSs induce ROS generation (Sokol et al., 1993, 2001), it is not surprising that BS-induced hepatocellular damage is influenced by the cellular signaling status. In line with this view, previous studies carried out in primary hepatocyte cultures showed that apoptosis induced by glycochenodeoxycholate is counteracted by PKC inhibitors (Jones et al., 1997), suggesting that PKC-dependent signaling pathways play a key role in hydrophobic BS-induced apoptosis. Taking into account the existence of common mechanisms between BS-induced apoptosis and necrosis (e.g., MPT formation, oxidative stress), it is possible to infer a similar protective effect of PKC inhibitors on TCDC-induced necrosis. Our results agree with this view. H7, a preferential PKC inhibitor (although it can inhibit in certain extent PKA as well) prevented partially the necrotoxic damage induced by TCDC (see Fig. 1). Furthermore, the specific PKA inhibitor, KT5720, was without effect, suggesting that H7 protective effect depended exclusively on its capability to block PKC activity. This was supported further using the specific PKC inhibitor, Che, which mimicked H7 protective effect. The mechanisms by which PKC inhibition protects against BS-induced necrosis can be multifactorial in nature. Our results showing here that PKC inhibitors prevented partially TCDC-induced ROS formation suggest that mitochondrial ROS production is facilitated somewhat by PKC activation. In line with this observation, lipid peroxidation induced to isolated rat hepatocytes by the oxidizing compound, tert-butyl hydroperoxide (tBOOH) (von Ruecker et al., 1989), or by the heavy metal, cooper (Mudassar et al., 1992), was prevented by the PKC inhibitor, H7, and exacerbated by PKC activators. Furthermore, H7 attenuates tBOOH-induced LDH leakage (Mudassar et al., 1992).
The preventive effect of PKC inhibitors on TCDC-induced lipid peroxidation is, however, rather marginal, suggesting that other mechanisms must be involved. For example, PKC inhibition may favor TCDC efflux into the extracellular medium by stimulating the exocytic discharge of vesicles containing BSs, as PKC inhibits hepatocellular vesicular trafficking (Zegers and Hoekstra, 1998); this mechanism is thought to play a key role in BS overcharging conditions, like that occurring in our experimental setting (Erlinger, 1990). Furthermore, we have shown that PKC inhibitors blocked, and PKC activators stimulated, vesicle-mediated trafficking of vesicle-containing BS transporters towards the apical hepatocellular pole (Roma et al., 2000). In concordance with this, a study in isolated rat perfused liver showed that H7-induced PKC inhibition increased biliary excretion of TCDC at a concentration in the perfusate within the range used in this study (1 mM) (Nakazawa et al., 1996).
Hydrophobic BSs induce elevation of [Ca2+]i by an inositol (1,4,5)triphosphate-independent mechanism (Combettes et al., 1988). Conceptually, Ca2+ elevations can activate different Ca2+-dependent proteases, phospholipases and endonucleases, with the consequent hepatocellular damage. Furthermore, Ca2+-elevations lead to activation of Ca2+-dependent PKC isoforms, which may be involved in TCDC-induced damage as well (see above). Therefore, we analyzed here whether the Ca2+-chelating agent, BAPTA/AM, has any beneficial effect against TCDC-induced hepatocellular necrosis. Despite this compound completely prevented TCDC-induced elevations in [Ca2+]i, the capability of TCDC to induce hepatocellular damage was not attenuated (see Fig. 5). This result, however, should not be conclusively interpreted to indicate that intracellular Ca2+ plays no role in BS-induced cytotoxicity. The predominance of other deleterious mechanisms not influenced by Ca2+ levels, e.g., the detergent properties of TCDC on cellular membranes, may have masked its contribution, particularly shortly after TCDC injury. Indeed, TCDC induced, in the perfused rat liver model, an early (4 min), transient increase in LDH release, followed by a subsequent time- and dose-dependent elevation in this parameter; only the first peak was significantly suppressed by pretreatment with the Ca2+- channel blocker, Ni2+ (Hasegawa et al., 2003). It is therefore possible that the protective effects of intracellular Ca2+ chelation have been overlooked in our model, which evaluate events occurring later in the necrotic process. Nevertheless, Ca2+ elevations are not a prerequisite for some forms of hepatocellular necrosis to occur, like that following ATP depletion due to metabolic inhibition (Nieminen et al., 1988).
The second messenger, cAMP, an endogenous activator of the PKA-dependent signaling pathway, was shown to have protective, dose-dependent effects in several models of hepatotoxicity (Kasai et al., 1996). Although its hepatoprotective mechanism/s have not been completely elucidated, its stabilizing effect on intracellular membrane is probably involved (Ignarro et al., 1973). Furthermore, cAMP inhibits BS-induced apoptosis by blocking caspase activation and cytochrome c release (Webster et al., 2002). Paradoxically, cAMP exacerbated rather than prevented TCDC-induced necrosis in a dose-dependent fashion (see Fig. 3). Although how this signaling molecule aggravates TCDC-induced damage remains elusive, changes in TCDC hepatocellular bioavailability may be involved. cAMP was shown to stimulate the Na+-dependent BS uptake by the basolateral transporter, Na+-taurocholate cotransporting polypeptide (ntcp), in the isolated rat hepatocyte model. This was attributed to the capability of cAMP to hyperpolarize the plasma membrane via PKA-mediated Na+/K+-ATPase phosphorilation (Edmondson et al., 1985), and by stimulation of the PKA-dependent translocation of ntcp from an endosomal compartment to the sinusoidal membrane (Webster and Anwer, 1999). Based upon these previous observations, and our own results showing the PKA dependency of DB-cAMP-induced exacerbation of TCDC-induced necrosis (see Fig. 4), we proposed that putative protective effects of cAMP could have been masked by the simultaneous increase in TCDC intracellular concentration due to enhanced uptake. Our results are in apparent contradiction to a previous study showing a protective effect of cAMP-permeant analogues on TCDC-induced necrotoxicity in hepatocytes cultured overnight (Ohiwa et al., 1993). The explanation for these varied results may depend on the different experimental conditions employed. Whereas freshly isolated hepatocytes like those used in our study maintain an intact capability to take up BSs, this function decreases significantly under culture conditions (Follmann et al., 1990). Indeed, uptake of the model bile salt, taurocholate, decreases to approximately half of the value recorded in freshly isolated hepatocytes during a culture period compatible to that used by Ohiwa et al. (1993). Therefore, our approach more clearly reflects the situation of the hepatocytes in situ, at least in terms of bile salt uptake.
In summary, the present findings clearly show that modulation of PKC- and PKA-dependent signaling pathways can modify the capability of hydrophobic BSs to induce hepatocellular necrotoxicity; whereas PKC inhibition attenuates BS-induced necrosis, PKA activation exacerbates this harmful effect. This seems to occur either or both by modulating differentially the intracellular availability of these endogenous, harmful compounds or by affecting the pathomechanisms involved in their deleterious effects. Although it is not possible at this point to establish whether these data have relevance to the situation in vivo, our results encourage future application of signaling molecules in the prevention/cure of hepatopathies occurring with elevated hepatocellular levels of endogenous BSs.
ACKNOWLEDGMENTS
This work was funded by CONICET, Agencia Nacional de Promoción Científica y Tecnológica (ANPCyT, PICT 05-08669), Beca ‘Ramón Carrillo-Arturo O?ativia 2003,’ from Ministerio de Salud de la Nación, and Fundación Antorchas (Subsidio de Emergencia para ‘Ex-Beneficiarios’ 14264/40).
REFERENCES
Baumgartner, U., Scholmerich, J., Leible, P., and Farthmann, E. H. (1992). Cholestasis, metabolism and biliary lipid secretion during perfusion of rat liver with different bile salts. Biochim. Biophys. Acta 1125, 142–149.
Drew, R., and Priestly, B. G. (1979). Choleretic and cholestatic effects of infused bile salts in the rat. Experientia 35, 809–811.
Zegers, M. M., and Hoekstra, D. (1998). Mechanisms and functional features of polarized membrane traffic in epithelial and hepatic cells. Biochem. J. 336(Pt. 2), 257–269.(Mariana Borgognone, Leona)