Effects of schisandrin B pretreatment on tumor necrosis factor-α induced apoptosis and Hsp70 expression in mouse liver

Abstract

Tumor necrosis factor-α (TNFα) could cause apoptosis in hepatic tissue of d-galactosamine sensitized mice, as evidenced by the increase in the extent of DNA fragmentation. The hepatic apoptosis induced by TNFα was associated with hepatocellular damage as assessed by plasma alanine aminotransferase activity. Schisandrin B (Sch B) pretreatment at daily doses ranging from 0.5 to 2 mmol/kg for 3 days caused a dose-dependent protection against TNFα-induced apoptosis in mice. The hepatoprotection was accompanied by a parallel reduction in the extent of hepatocellular damage. The same Sch B pretreatment regimens increased hepatic Hsp70 level in a dose-dependent manner. The relevance of Sch B–induced increase in Hsp70 expression to the prevention of TNFα-triggered hepatic apoptosis remains to be elucidated.

INTRODUCTION

Tumor necrosis factor-α (TNFα) is a cytokine mainly produced by activated macrophages. The early interest in TNFα was emerged from the finding of its abilities to cause necrosis of Meth A sarcomas in vivo and to selectively kill transformed and neoplastic cell lines in vitro (Carswell et al 1975). In addition to its antitumor action, TNFα also plays an important role in many pathophysiological conditions, such as those related to inflammation and immunomodulatory function (Grunfeld and Feingold 1991; Bonavida 1992; Beutler 1999), as well as ischemia- reperfusion injury (Donnahoo et al 1999; Ferrari 1999). Recent studies have shown that TNFα could induce apoptosis in vivo in mouse liver that had been sensitized with transcriptional inhibitors such as d-galactosamine (d-gal) (Leist et al 1994).

Heat shock proteins (Hsps) are a family of inducible and constitutively expressed gene products that collectively function to maintain cellular protein conformation during stressful conditions. The synthesis of Hsps is induced by a variety of mild stress, including oxidants, heat, hypoxia, and low pH, all of which can affect protein conformation (Welch 1992). The synthesis of Hsps allows cells to adapt to gradual changes in their environment and to survive in otherwise lethal conditions (Jaattela 1999). Among various Hsps, the inducible Hsp70 plays an important role as molecular chaperon in the folding of newly synthesized, unassembled, or unfolded proteins. The overexpression of Hsp70 has been shown to protect the cell against toxicity induced by TNFα (Nishimura et al 1997; Ahn et al 1999).

Schisandrin B (Sch B, Fig 1) is the most abundant and active dibenzocyclooctadiene derivative isolated from the fruit of Schisandra chinensis, a traditional Chinese herb clinically used for the treatment of viral and chemical hepatitis (Li 1991). Previous studies in our laboratory have demonstrated the ability of Sch B to protect against free radical–mediated hepatocellular damage (Ip et al 1995). The hepatoprotective action of Sch B pretreatment was associated with the enhancement in hepatic glutathione antioxidant status (Ip et al 1995), particularly in the mitochondrion (Ip and Ko 1996). In the present study, the effects of Sch B pretreatment on TNFα-induced apoptosis and Hsp70 expression in hepatic tissue were examined in mice in an effort to further explore the molecular mechanism involved in the hepatoprotective action of Sch B.

Fig 1. Chemical structure of Sch B. Sch B was purified from the petroleum ether extract of Fructus schisandrae by silica gel column chromatography as previously described (Ko et al 1995). The chemical structure of Sch B was confirmed by comparing the silica gel TLC and spectral characteristics (¹H- and ¹³C-NMR and mass spectra) with authentic standard obtained from the Institute of Materia Medica, Chinese Academy of Sciences, Beijing. The purity of the compounds, as assessed by HPLC, was found to be higher than 95% (w/w)

RESULTS AND DISCUSSION

Effect of Sch B pretreatment on TNFα-induced apoptosis in hepatic tissue

As shown in Figure 2A, 6 hours after an intravenous injection of mouse recombinant TNFα at 5 μg/kg immediately following the intraperitoneal administration of d-gal at 700 mg/kg in mice, apoptosis was observed in hepatic tissue, as evidenced by an increase (6-fold) in the extent of DNA fragmentation, a molecular marker of apoptosis. The extent of hepatic apoptosis was further increased (12-fold) 8 hours following the TNFα treatment. The hepatic apoptosis induced by TNFα was associated with hepatocellular damage, as indicated by a parallel increase in plasma alanine aminotransferase (ALT) activity (Fig 2B). Pretreating mice with Sch B at daily doses ranging from 0.5 to 2 mmol/kg for 3 days caused a dose-dependent protection against TNFα-induced apoptosis and hepatocellular damage in hepatic tissue, as indicated by the suppression of DNA fragmentation and the decrease in plasma ALT activity, respectively. It has been suggested that hepatic apoptosis induced by TNFα in d-gal-sensitized mice is an early and possibly causal event leading to experimental liver failure (Leist et al 1995). Although the precise molecular mechanism of apoptosis is not fully understood, the cytotoxicity of TNFα has been attributed to various cellular actions, such as the activation of phospholipase A₂ (Johns and Webb 1998; Liu and McHowat 1998), the release of ceramide (Higuchi et al 1996), the activation of endogenous endonucleases (Fernandez et al 1995), and the fragmentation of DNA caused by reactive oxygen species generated from the mitochondrion (Mertens et al 1995). Apoptosis has first been considered as a process not involving the changes in mitochondrial function. However, results obtained from recent studies have strongly suggested that the loss of mitochondrial membrane integrity is a critical event (or the point of no return) leading to the programmed cell death (Hirsch et al 1998). Free radical scavengers (Egawa et al 1994; Fernandez-Checa et al 1997) and overexpression of superoxide dismutase in the mitochondrion have been shown to inhibit TNFα-induced cytotoxicity (Kim and Kim 1999). Our finding of the hepatoprotective effect of Sch B pretreatment against TNFα-induced apoptosis may also be related to the ability of Sch B to enhance hepatic mitochondrial glutathione redox status as previously reported (Ip et al 1996, 1997).

Fig 2. Time course of Sch B pretreatment on TNFα-induced hepatic apoptosis and hepatocellular damage in mice: The effect of Sch B pretreatment. Female Balb/c mice (20–22 g) were randomly assigned into groups of 5 individuals. In the pretreatment groups, animals were treated intragastrically with Sch B (suspended in olive oil, 4% [w/v]) for 3 days at daily doses ranging from 0.5 to 2 mmol/ kg. Control animals were administered with the olive oil. Twenty-four hours after the last dosing, animals will be treated intraperitoneally with d-gal (700 mg/kg) followed by an intravenous injection of mouse recombinant TNFα at 5 μg/kg. Animals were sacrificed 4, 6, or 8 hours after the intoxication. (A) The extent of DNA fragmentation, a molecular marker of apoptosis, was quantitated by measuring hepatic cytosolic oligonucleosome-bound DNA using Cell Death Detection Kit (Roche, Germany) and expressed in percentage of untreated control (ie, non-Sch B and non-TNFα). (B) The extent of hepatocellular damage was assessed by measuring plasma alanine aminotransferase (ALT) activity using an assay kit from Sigma. Data were analyzed by one-way ANOVA followed by Duncan's multiple range test in order to detect intergroup differences. Significant difference was determined when P < 0.05. Values given are the mean ± SEM, with n = 5. * Significantly different from the untreated control. # Significantly different from the TNFα-treated control at the same time period

Effect of Sch B treatment on Hsp70 expression

The major heat shock protein, Hsp70, has been shown to protect against TNFα-mediated cellular damage in various in vitro cell cultures (Jaattela 1995; Jaattela et al 1998; Meng et al 1999; Schroeder et al 1999). As an approach to investigating the mechanism involved in the protection afforded by Sch B pretreatment against TNFα-induced hepatic apoptosis, the hepatic level of Hsp70 was measured by Western blotting analysis in mice treated with increasing doses of Sch B. As shown in Figure 3, treating mice with Sch B at a daily dose of 1 mmol/kg for 1 or 2 days did not significantly increase the expression of Hsp70. However, the Hsp70 level was greatly increased by 95% when mice was treated with Sch B for 3 consecutive days. Treating mice with Sch B at daily doses ranging from 0.5 to 2 mmol/kg for 3 days caused a dose- dependent increase of Hsp70 level in hepatic tissue (Fig 4). The same dosage regimen of Sch B did not produce significant changes in the levels of Hsp25, Hsp60, or Hsp90 (data not shown).

Fig 3. Time course of hepatic Hsp70 expression in Sch B–pretreated mice. (A) Female Balb/c mice (20–22 g) were randomly assigned into groups of 3 individuals. In the pretreatment groups, animals were treated intragastrically with Sch B (suspended in olive oil) at a daily dose of 1 mmol/kg for 1 day (lane 3), 2 days (lane 4), and 3 days (lane 5), respectively. Control animals (lane 2) were administered with the olive oil only. Animals were sacrificed 24 hours after the last dosing. Hsp70 standard (lane 1) was purchased from StressGen (Victoria, BC, Canada). Hepatic cytosolic fraction were separated by SDS-PAGE with 7.5% gel using Mini-PROTEAN II Electrophoresis Cell (Bio-Rad, USA). Separated proteins were transferred electrophoretically onto the nitrocellulose paper using Mini Trans-Blot Electrophoretic Transfer Cell (Bio-Rad). The nitrocellulose membrane was probed with anti-Hsp70 antibody purchased from StressGen. Subsequently, the Hsp70 band was visualized using Immun-Bot Colorimetric Assay Kit (Bio-Rad). The intensity of the immunostained band was quantitated by an imaging densitometer (GS-670, Bio-Rad). (B) Hepatic Hsp70 level in Sch B–treated mice. Values given are the mean ± SEM, with n = 3. * Significantly different from the control

Fig 4. Dose response of Sch B on hepatic Hsp70 expression in mice. (A) Mice were treated intragastrically with Sch B at a daily dose of 0.5 (lane 3), 1.0 (lane 4), and 2.0 (lane 5) mmol/kg for 3 days, respectively. Control animals (lane 2) were administered with the olive oil only. Hsp70 level in hepatic cytosolic fractions and Hsp70 standard (lane 1) were detected by Western blotting as described in Figure 3. (B) Hepatic Hsp70 level in Sch B–treated mice. Values given are the mean ± SEM, with n = 3. * Significantly different from the control

Given the complexity and diversity of pathways leading to apoptosis, the mitochondrion has been suggested to be a convergent site for numerous signal transduction pathways, beyond which cell death is irreversibly committed (Petit et al 1997). In this connection, the mitochondrial antioxidant status has been shown to play a determinant role in preventing TNFα-induced cytotoxicity (Fernandez-Checa et al 1997; Kim and Kim 1999). It is unlikely that Hsp70, being a cytosolic protein, can produce a direct antioxidant action in the mitochondrion. Instead, Hsp70 may prevent TNFα toxicity by blocking the signal transduction to the mitochondrion or regulating mitochondrial protein synthesis through its molecular chaperon activities. The apoptotic process involves several components, including mitochondrial membrane proteins of the bcl-2 family, a family of cysteine proteinases (caspases), and stress-activated kinases; the inactivation of these components can completely prevent the loss of cell viability (Gabai et al 1998). It has been demonstrated that elevated levels of Hsp70 inhibited a signal transduction pathway leading to programmed cell death by preventing the activation of stress-induced activation of Jun N-terminal kinase (JNK), a very early stage of the apoptotic process (Gabai et al 1997). In addition, stress- induced activation of another stress-kinase, p38 (HOG1), was also blocked when the level of Hsp70 was increased. The activation of caspases, a crucial step in stress-induced apoptosis, was also inhibited in Hsp70-expressing cells (Mosser et al 1997; Buzzard et al 1998). It has been demonstrated that S-nitroso-N-acetylpenicillamine pretreatment, which could increase the expression of hepatic Hsp70 but not the activities of antioxidant enzymes, was able to prevent TNFα-induced liver damage (Kim et al 1997). The important role of Hsp70 was further supported by the observation that blocking of Hsp70 expression by an antisense oligonucleotide could abrogate the hepatoprotective action (Kim et al 1997).

Some pharmacological agents, such as nonsteroidal anti-inflammatory drugs, have been shown to induce heat shock gene activation (Amici et al 1995). It has been shown that long-term administration of aspirin increased the expression of Hsp70 in the gastric mucosa and prevented the mucosal damage induced by a higher dose of aspirin (Jin et al 1999). Since Sch B treatment can significantly increase the expression of Hsp70 in the liver and conceivably in gastric mucosa too, the possible therapeutical application of Sch B in the prevention of mucosal injury is therefore worthy of further investigation. In fact, the activation of heat shock response has been suggested to be a broad-based treatment for a range of diseases, including stroke, neurodegenerative diseases, myocardial diseases, as well as tissue damage and trauma (Morimoto and Santoro 1998). In this regard, by virtues of its naturally occurring origin and long history of use, Sch B may serve as a promising therapeutic agent for the induction of adaptive cytoprotection.

In conclusion, Sch B treatment protected against TNFα- induced hepatic apoptosis and hepatocellular damage in d-gal sensitized mice in a dose-dependent manner. The same Sch B pretreatment regimens produced a dose-dependent increase in hepatic Hsp70 level. Further investigation is under way to define the molecular mechanism involved in the Sch B-induced increase in Hsp70 expression and its relevance to the prevention of TNFα-induced apoptosis.