Introduction

Tetrandrine (Tet, Figure 1) is a compound extracted from the root of Stephania tetrandra and is often used for the treatment of rheumatism, inflammation, occlusive cardiovascular disorders and tumors1,2,3,4,5,6,7,8. Some early studies demonstrated that tetrandrine causes hepatotoxicity9,10.

Figure 1
figure 1

Structure of tetrandrine.

PowerPoint slide

Full size image

Mitochondrial dysfunction is involved in many diseases and drug-induced toxicities. In our previous work, we showed that tetrandrine causes hepatotoxicity both in vivo and in vitro, and mitochondrial dysfunction plays an important role in mediating this process. Tetrandrine-induced mitochondrial dysfunction is cyclosporine A-sensitive, suggesting the presence of a mitochondrial permeability transition (MPT)7,9,11. The MPT plays a major role in mitochondrial dysfunction and represents an abrupt increase in the permeability of the mitochondrial inner membrane, resulting in the decrease of cellular ATP content and the dissipation of mitochondrial membrane potential. The MPT can be specifically inhibited by cyclosporine A (CsA)12,13,14.

Another study showed that tetrandrine could protect mitochondria isolated from rat livers against the damage induced by oxidants15. These two inconsistent reports indicated that mitochondrial dysfunction induced by tetrandrine in hepatocytes may be not due to the direct action of tetrandrine on mitochondria, but rather by other mechanisms that indirectly influence tetrandrine-induced mitochondrial damage.

MPT can be induced by endogenous ROS that are generated by the metabolism of drugs, pollutants and chemicals16,17,18. The increase of ROS induced by tetrandrine has been previously reported9,19. A recent study showed that ketoconazole, an inhibitor of CYP3A, protects mice against pulmonary toxicity induced by tetrandrine20. These studies suggested a potential role of CYP450 in ROS generation and MPT induced by tetrandrine. Therefore, we designed the present study to study the role of ROS and CYP450 in tetrandrine-induced mitochondrial dysfunction using rat primary hepatocytes as a model system for hepatotoxicity.

Materials and methods

Reagents and cell culture

Tetrandrine (Sigma Chem Co, St Louis, MO, USA) was dissolved in 50 μL of 0.1 mol/L HCl and diluted with HBS solution (33 mmol/L HEPES, 160.8 mmol/L NaCl, 3.15 mmol/L KCl, 0.7 mmol/L Na2HPO4·12H2O, pH 7.65) to a final concentration of 10 mmol/L.

Hepatocytes were isolated from Sprague-Dawley male rats (180–220 g) by two-step collagenase perfusion as previously described with some modifications21,22,23. Collagenase IV was purchased from Sigma Chem Co (St Louis, MO, USA). Hepatocytes were cultured as a monolayer in Ham's F-12/DMEM (Invitrogen, Carlsbad, CA, USA) (1:1) medium supplemented with 15% fetal bovine serum (PAA Laboratories GmbH, Linz, Austria).

HL-7702 cells, an immortalized hepatocyte cell line derived from a human liver (Cell Bank, SIBS, CAS, Shanghai, China), were cultured in DMEM containing 10% fetal bovine serum (PAA Laboratories GmbH, Linz, Austria), 100 U/mL penicillin and 70 μg/mL streptomycin.

Plasmid construction and transfection

The coding sequences of human CYP2C8 (NM_000770, 96-1568), CYP2D6 (NM_000106, 91-1584) and CYP2E1 (NM_000773, 34-1515) were amplified by high fidelity PCR and cloned into the pcDNA3.1 (–) plasmid (Invitrogen, Carlsbad, CA, USA). The following primers were used for cloning: CYP2C8 (Xho I/Hind III, Forward: 5′-CCCCTCGAGGCCACCATGGAACCTTTTGTGGTCCTG-3′, Reverse: 5′-CCCAAGCTTTCAGACAGGGATGAAGCAGA-3′), CYP2D6 (BamH I/Hind III, Forward: 5′-CGGGATCCGCCACCATGGGGCTAGAAGCACTGGT-3′, Reverse: 5′-CCCAAGCTTCTAGCGGGGCACAGCACAAA-3′) and CYP2E1 (EcoR I/Hind III, Forward: 5′-CCCGAATTCGCCACCATGTCTGCCCTCGGAGTCAC-3′, Reverse: 5′-CCCAAGCTTTCATGAGCGGGGAATGACAC-3′). HL-7702 cells were seeded at 5×105 cells/mL per well in 6-well plates in growth medium containing 10% serum. Cells were grown for 24 h at 37 °C and transfected with plasmids using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The expression of CYP450 protein was determined by Western blotting 48 h after transfection.

Cell viability and ATP bioluminescence assays

Cell viability and ATP bioluminescence reagents were provided in the CellTiter-Glo™ Luminescent Cell Viability Assay kit (Promega, Madison, WI, USA). Intracellular ATP levels were determined according to the manufacturer's protocol. After a 24-h treatment with tetrandrine, the bioluminescence was measured by a fluorimeter (NOVOstar, BMG LABTECH, Offenburg, Germany).

Measurement of the mitochondrial membrane potential

The mitochondrial membrane potential was measured by the accumulation of rhodamine 123 (Sigma Chem Co) according to a previously published method24. After treatment with tetrandrine (25 μmol/L) for either 4 h or 8 h, the hepatocytes were washed twice with PBS and loaded with rhodamine 123 (1 μmol/L) for 20 min. The fluorescence was detected on the NOVOstar at an excitation of 485 nm and emission of 520 nm.

Measurement of ROS generation

DCFH-DA (Sigma Chem Co, St Louis, MO, USA) was used to detect the intracellular generation of ROS by tetrandrine as previously described25. After a 1-h treatment with tetrandrine (25 μmol/L), rat primary hepatocytes were washed twice and resuspended in PBS (1×106 cells/mL). This cell suspension was loaded with DCFH-DA (5 μmol/L) and incubated at 37 °C for 20 min. The fluorescence of DCFH was measured on NOVOstar at an excitation of 485 nm and emission of 520 nm.

Measurement of intracellular GSH content

GSH levels in the hepatocytes were measured according to the method described by Dringen and Hamprecht with some modifications26. Cell homogenates were diluted with 10% trichloroacetic acid and centrifuged at 3000×g for 5 min. The supernatant was mixed with 10 mmol/L 5,5′-dithio-bis (2-nitrobenzoic acid) (DTNB) and the absorbance was measured at 412 nm within 5 min. Reduced pure GSH was used to generate a standard curve.

Measurement of caspase 3 activity

Caspase 3 activity was measured with a caspase 3 assay kit (BD Biosciences Pharmingen, San Diego, CA, USA). Reaction mixtures containing Ac-DEVD-AMC and cell lysates in buffer were incubated for 1 h at 37 °C. AMC liberated from Ac-DEVD-AMC was measured in the NOVOstar at an excitation of 380 nm and emission of 460 nm.

Confocal microscopy and detection of the mitochondrial permeability transition

Cell loading and confocal microscopy were performed as previously described27. Briefly, 5×106 hepatocytes were plated on 40-mm diameter glass coverslips and loaded with 500 nmol/L TMRE for 30 min at 37 °C. After three washes with KRH buffer, the coverslips were mounted on a Leica TCS-SP2 laser scanning confocal microscope. To reduce the influence of room temperature, an increased tetrandrine concentration (50 μmol/L) and shorter incubation time (30 min) were used. Images were collected at the indicated time points after adding tetrandrine in KRH buffer containing 100 nmol/L TMRE. The occurrence of the MPT was defined as loss of bright mitochondrial TMRE staining, which is indicative of mitochondrial depolarization. During the incubation, the hepatocytes of the control group did not show any loss of TMRE staining.

Western blotting

Standard Western blotting analysis was performed using polyclonal antibodies against β-actin, CYP2C8, CYP2D6, and CYP2E1 (Millipore, Billerica, MA, USA).

Statistical analysis

Data were entered into a database and analyzed using SPSS software. Group mean values and standard deviations were calculated. After homogeneic analysis, homogeneous data were analyzed using one-way analysis of variance and a post hoc test of least significant difference. Heterogeneous data were analyzed using t-tests. When P<0.05, the results were considered to be statistically significant.

Results

The protective effects of ROS scavengers on the cytotoxicity induced by tetrandrine in rat hepatocytes

In our previous work, we reported an increase of ROS induced by tetrandrine treatment in rat hepatocytes9. The role of oxidative damage in the hepatotoxicity of tetrandrine was examined by using three ROS scavengers (NAC, VE, and VC), all of which largely attenuated the cytotoxicity of tetrandrine (25 μmol/L) at 8 h (Figure 2).

Figure 2
figure 2

Protective effect of ROS scavengers on the cytotoxicity of tetrandrine (10 and 25 μmol/L) after 8-h treatment in rat primary hepatocytes. Cell viability was evaluated by ATP content. (A) Effect of N-acetyl-cysteine (NAC, 4 mmol/L) on the cytotoxicity of tetrandrine. (B) Effect of vitamin E (VE, 2 mmol/L) on the cytotoxicity of tetrandrine. (C) Effect of vitamin C (VC, 1 mmol/L) on the cytotoxicity of tetrandrine. All of the results are presented as the mean±SD. n=3. cP<0.01 vs 'Tet' group.

PowerPoint slide

Full size image

The protective effects of CYP450 inhibitors on oxidative stress induced by tetrandrine

To analyze the source of the induced ROS, different inhibitors of ROS generation that block several different intracellular ROS-producing systems were applied. The inhibitors used are summarized in Table 1. As shown in Figure 3A and Figure S1, only inhibitors of CYP450 blocked the increase of DCFH fluorescence after a 1-h treatment with tetrandrine.

Table 1 Different enzyme inhibitors on ROS-producing systems.
Full size table
Figure 3
figure 3

Protective effect of CYP450 inhibitors on oxidative stress induced by tetrandrine. (A) Effects of CYP450 inhibitors on ROS production induced by tetrandrine. ROS were detected using the fluorescent probe DCFH-DA. Rat hepatocytes were incubated with 25 μmol/L Tet and the indicated inhibitors for 1 h. The concentration of each inhibitor is listed in Table 1. (B) Effect of SKF on the depletion of GSH induced by tetrandrine (10 and 25 μmol/L). GSH content was measured in rat primary hepatocytes at time points up to 8 h in either the presence or absence of SKF (10 μmol/L). All of the results are presented as the mean±SD. n=4. cP<0.01 vs 'Tet 25 μmol/L' group.

PowerPoint slide

Full size image

Additional assays were performed to examine the effect of the CYP450 inhibitor SKF-525A (SKF) on cellular GSH content after treatment with tetrandrine for either 4 h or 8 h. Depleted GSH levels were attenuated by SKF (Figure 3B). No protection was observed when using inhibitors of other ROS-producing systems.

The CYP450 inhibitor SKF and ROS scavenger VC recovered the mitochondrial dysfunction induced by tetrandrine

TMRE, a cationic fluorophore, accumulates in the mitochondria of cultured hepatocytes in response to a negative mitochondrial membrane potential. During normal aerobic incubation, confocal microscopy reveals the mitochondria as bright red spots, which is consistent with previous results28.

After 30 min of exposure to tetrandrine (50 μmol/L), more than 80 percent of the mitochondria lost TMRE fluorescence, indicating depolarization of mitochondria and the presence of the MPT (Figure 4A1–4A3). We further examined the effect of VC and SKF-525A (SKF) on tetrandrine-induced MPT. When cotreated with VC (1 mmol/L), approximately two-thirds of all of the mitochondria maintained TMRE fluorescence (Figure 4C1–4C3, 4D). In the presence of SKF, approximately 20% of mitochondria had undergone MPT and a change in the distribution of TMRE (Figure 4B1–4B3, 4D).

Figure 4
figure 4

Effect of SKF and VC on the mitochondrial permeability transition induced by tetrandrine. TMRE accumulates into the mitochondria. A loss of the red fluorescence characteristic of TMRE indicates the depolarization of mitochondria and the presence of the mitochondrial permeability transition. Rat hepatocytes were incubated with 50 μmol/L Tet alone (A1–A3), or in combination with either 10 μmol/L SKF (B1–B3) or 1 mmol/L VC (C1–C3) for 30 min. The scale bar is 47.62 μm. (D) Red fluorescence intensity was semi-quantitated from confocal images in three separated experiments. SKF, SKF-525A, a nonspecific inhibitor of CYP450. All of the results are presented as the mean±SD. n=3.

PowerPoint slide

Full size image

Next, we tested the ability of VC and SKF to protect against tetrandrine-induced depletion of ATP and mitochondrial membrane potential (MMP)9. VC and SKF both significantly inhibited tetrandrine-induced ATP and MMP depletion, with SKF exerting a stronger protective effect than VC and CsA (Figure 5A, 5B; Figure S2).

Figure 5
figure 5

Protective effect of SKF and VC on mitochondrial dysfunction induced by tetrandrine. Rat hepatocytes were incubated with 25 μmol/L Tet, 10 μmol/L SKF and 1 mmol/L VC. (A) The effect of VC and SKF on tetrandrine-induced ATP depletion. Samples were taken at the indicated time points and assayed for ATP levels. (B) Effect of VC and SKF on mitochondrial membrane potential depletion induced by tetrandrine. (C) The activity of caspase 3 in the presence of tetrandrine and SKF. Caspase 3 activity was measured in rat primary hepatocytes at time-points up to 8 h. bP<0.05, cP<0.01 vs 'Tet' group at each time point. SKF, SKF-525A, a nonspecific inhibitor of CYP450. All results are presented as the mean±SD. n=3.

PowerPoint slide

Full size image

The increased permeability of the mitochondrial inner membrane allows for the release of cytochrome c into the cytosol, which activates downstream apoptotic factors such as caspase 314. Tetrandrine induced the release of cytochrome c and activated caspase 39. Treatment with SKF showed a marked inhibition of caspase 3 activity induced by tetrandrine (Figure 5C).

Role of CYP2E1 in the oxidative stress and cytotoxicity induced by tetrandrine

Multiple inhibitors (Table 2) for the rat CYP450 isoforms were used to examine which CYP450 isoforms may be involved in the oxidative stress and cytotoxicity induced by tetrandrine. The inhibitors of CYP2C (sulfaphenazole, SFZ), CYP2D (quinine, QUN), and CYP2E1 (diallyl sulfide, DAS) significantly reduced tetrandrine-induced ROS production (Figure 6A & Figure S3A). These inhibitors also attenuated the ATP depletion induced by tetrandrine (Figure 6B and Figure S3B).

Table 2 Inhibitors of CYP450 isoforms.
Full size table
Figure 6
figure 6

Role of CYP2E1 in tetrandrine-induced hepatotoxicity. (A) Effects of the CYP450 isoform inhibitors on the ROS production of tetrandrine. ROS was detected using the fluorescent probe DCFH-DA. Rat hepatocytes were incubated with 25 μmol/L Tet and the indicated inhibitors for 1 h. The concentration of each inhibitor is listed in Table 2. (B) Effect of the CYP450 isoform inhibitors on the cytotoxicity of tetrandrine. Rat hepatocytes were treated with 25 μmol/L Tet for 8 h. The concentration of each inhibitor was listed in Table 2. (C) IC50 values of tetrandrine in HL-7702 cells overexpressing various CYP450 isoforms. Cells were treated with a range of tetrandrine concentrations (0, 0.03, 0.1, 0.3, 1, 3, 10, 30, 100, and 300 μmol/L). The IC50 values were calculated using the Logit method. (D) Caspase 3 activity in HL-7702 cells overexpressing various CYP450 isoforms. Cells were treated with 25 μmol/L Tet for 8 h. (E) The protein levels of CYP450 isoforms were evaluated by Western blotting. All of the results are presented as the mean±SD. n=3. bP<0.05, cP<0.01 vs '25 μmol/L Tet' group.

PowerPoint slide

Full size image

We constructed plasmids containing human CYP2C8, CYP2D6 and CYP2E1 and transiently transfected them individually into HL-7702 cells. Empty pcDNA3.1 (–) plasmid was used as the mock-vehicle group. The IC50 values of tetrandrine in these four groups were compared. Overexpressed CYP2E1 in HL-7702 cells decreased the IC50 of tetrandrine (Figure 6C) and enhanced caspase 3 activity induced by tetrandrine (Figure 6D).

Discussion

In this study, we demonstrated that tetrandrine promotes mitochondrial dysfunction through ROS generated by CYP450, with CYP2E1 as the specific isoform that is primarily responsible for this process. Tetrandrine-induced cytotoxicity could be markedly reduced by ROS scavengers, indicating the major role of oxidative damage in tetrandrine-induced hepatotoxicity (Figure 2). Only CYP450 inhibitors could inhibit the tetrandrine-induced increase of ROS and GSH depletion (Figure 3), which implied the role of CYP450 in ROS production induced by tetrandrine. Additionally, an antioxidant (VC) and a CYP450 inhibitor (SKF-525A) could both prevent tetrandrine-induced mitochondrial dysfunction, although SKF-525A had a stronger protective effect (Figures 4 and 5). Finally, an inhibitor of the CYP2E1 isoform decreased the cytotoxicity and ROS production effects of tetrandrine. Overexpression of CYP2E1 enhanced the toxicity of tetrandrine (Figure 6). Taken together, these results suggest that the ROS from CYP450, specifically CYP2E1, may account for the mitochondrial dysfunction induced by tetrandrine.

The CYP450 family of enzymes are responsible for the metabolism of both exogenous and endogenous materials29. Jin et al showed that inhibition of CYP3A attenuates the pulmonary toxicity induced by tetrandrine in mice, and only the human CYP3A4 and CYP3A5 isoforms are responsible for the reactive metabolites of tetrandrine20. In our study, ketoconazole and other CYP3A inhibitors had no effect on the cytotoxicity of tetrandrine, which may be due to the rapid decrease of CYP3A1/2 expression during the initial 24 h that rat hepatocytes are cultured on collagen30 and the ineffective inhibition of ketoconazole on rat CYP3A1/231.

A number of studies have shown the important role of overexpressed CYP2E1 in oxidative stress32,33,34,35,36,37. In this study, the protein levels of CYP2E1 in rat hepatocytes were well maintained (data not shown). Diallyl sulfide, an inhibitor of CYP2E1, decreased oxidative stress of tetrandrine and significantly attenuated ATP depletion induced by tetrandrine. Furthermore, overexpression of CYP2E1, but not CYP2C8 or CYP2D6, enhanced the toxicity and caspase 3 activity induced by tetrandrine in HL-7702 cells. Together, this strongly suggests a potential synergistic effect between CYP2E1 and tetrandrine in producing ROS. More work is necessary to clarify the mechanism through which CYP2E1 is involved in tetrandrine-induced hepatotoxicity.

In the present study, inhibitors of CYP450 largely decreased the ROS production induced by tetrandrine after 1 h (Figure 3A). However, CYP450 inhibitors did not completely recover the GSH depletion after 8 h (Figure 3B). It is well-known that GSH can be depleted by ROS and reactive metabolites of drugs. In a recent study, generation of the GSH conjugate from O-demethylated tetrandrine by CYP3A4 was observed with liver and lung microsomes. A high concentration of ketoconazole (1 μmol/L, approximately 50-fold higher than the Ki value of ketoconazole on CYP3A4) did not completely inhibit the production of GSH conjugates (35% GSH conjugates remained) (Figure S1 and Figure 9 in Ref 20)20,31, which was indicative of GSH consumption. SKF-525A has a much higher Ki value against CYP3A4 than ketoconazole, indicating less effective inhibition of CYP3A. Together, this suggests that the incomplete recovery of GSH depletion may be due to the persistent consumption of GSH conjugates with the metabolites of tetrandrine, which could not be completely inhibited by proadifen over a longer incubation time (Figure 3B).

In our previous study, daily administration of tetrandrine (57 mg/kg by oral gavage) for 8 d induced obvious liver injury in rats9. Early reports involving tetrandrine administration to rodents (15 mg/kg by single gavage or 7.5 mg/kg by single intravenous injection) showed obvious accumulation in the rodent liver and very slow hepatic clearance. The concentration of tetrandrine in the liver reached approximately 22–32 μmol/L and was maintained for at least 48 h38,39. Converting to an equivalent human dose, the dosage of tetrandrine for humans is approximately 2.38 mg/kg. Clinical doses of tetrandrine range from 2 to 5 mg/kg over a course of approximately 3 months. Taken together, these data from the animal studies indicated the potential accumulation and injury to human livers from chronic administration of tetrandrine. Clinical attention regarding the dosage and treatment course of tetrandrine should be addressed.

In summary, ROS generated by CYP450 play a major role in tetrandrine-induced mitochondrial dysfunction, and CYP2E1 likely plays a more important role in this process. Our data will be helpful in the clinical application of tetrandrine.

Author contribution

Jin REN and Li-kun GONG participated in research design; Xin-ming QI, Ling-ling MIAO, and Yan CAI conducted experiments; Xin-ming QI, Ling-ling MIAO, and Li-kun GONG performed data analysis; Xin-ming QI, Ling-ling MIAO, Li-kun GONG, and Jin REN wrote or contributed to the writing of the manuscript.