Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Hepatotoxicity is one of the adverse reactions of nonsteroidal anti-inflammatory drugs (NSAIDs) in their clinical use (Lewis, 1984; Zimmerman, 1990; Rabinovitz and Van Thiel, 1992; Boelsterli et al., 1995). The toxicity also has been observed experimentally as acute cell injury with freshly isolated hepatocytes and cultured ones (Akesson and Akesson, 1984; Sorensen and Acosta, 1985; Castell et al., 1988; Schmitz et al., 1992). Because biotransformation is an essential step to initiate the hepatotoxicity induced by a number of drugs and chemicals, NSAID toxicity also has been of interest in connection with the involvement of the reactive metabolites. Carboxylic acid-containing NSAIDs were metabolized into a common group of chemically reactive metabolites, acyl glucuronides (Spahn-Langguth and Benet, 1992; Dickinson, 1993), whereas none of the glucuronide has been established to be directly involved in hepatocyte injury. In contrast, a few studies suggested that oxidative metabolism participated in NSAID-induced hepatocyte injury (Kretz-Rommel and Boelsterli, 1993; Jurima-Romet et al., 1994). In line with the involvement of the reactive metabolites, Bort et al. (1999) recently reported that diclofenac metabolism into an N-hydroxymetabolite was responsible for the cytotoxicity in addition to mitochondrial impairment by the parent compound. Thus, the mechanism for hepatocyte injury by NSAID has been extensively studied, whereas a common mechanism to explain the toxicity of all of hepatotoxic NSAIDs has not been provided.

We found that cytotoxic NSAIDs caused decreases in cellular ATP contents before leakage of lactate dehydrogenase (LDH) from the cells (Masubuchi et al., 1998). It also was shown that diphenylamine was one of the common "skeleton" structures of NSAIDs to deplete cellular ATP and leak LDH. In addition, diphenylamine itself, which is not an NSAID, diminished ATP and caused hepatocyte injury. To further elucidate the mechanism for ATP depletion by NSAIDs and diphenylamine, we have studied the effects on mitochondrial oxidative phosphorylation, the major source of cellular ATP, whereas some of the acidic NSAIDs are well known as uncouplers of the oxidative phosphorylation (Mahmud et al., 1996; Mingatto et al., 1996; Petrescu and Tarba, 1997). It was found that diphenylamine was an uncoupler of mitochondrial oxidative phosphorylation as well as mefenamic acid and diclofenac (Masubuchi et al., 1999). Because the potent uncoupling leads to depletion of ATP content at the cellular level, it implies hepatotoxicity of these compounds, whereas the cause-and-effect relation remains to be elucidated.

The purpose of this study was to clarify the role of the uncoupling in cell injury induced by mefenamic acid, diclofenac, and their "skeleton" diphenylamine (Fig. 1) with freshly isolated rat hepatocytes. In addition, we examined the ability of these compounds to induce mitochondrial permeability transition (MPT) and their effects on membrane potential to further characterize the effects on mitochondrial function.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Diphenylamine and its structurally related NSAIDs.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Diphenylamine, mefenamic acid, diclofenac sodium, collagenase (type I), safranine O, fructose, and LDH-UV-Test-Wako were purchased from the Wako Pure Chemical Ind. (Osaka, Japan). Oligomycin was from the Sigma Chemical Co. (St. Louis, MO). ATP sodium salt was from the Oriental Yeast Co., Ltd. (Tokyo, Japan). All other chemicals and solvents were of analytical grade.

Preparation of Liver Mitochondria and Microsomes. Male Wistar rats (2 months old) were obtained from Takasugi Experimental Animals (Saitama, Japan). The animals were housed in air conditioned rooms (25°C under a 12-h light/dark cycle for 1 week before use). Food (commercially available pellet; Oriental Yeast Co., Ltd.) and water were given ad libitum. Liver mitochondria fraction was prepared according to the method described by Schneider and Hogeboom (1950) in a medium containing 0.25 M sucrose, 10 mM Tris-HCl buffer (pH 7.4), and 0.5 mM EDTA. Liver microsomal fractions were prepared according to the method of Omura and Sato (1964). Protein concentrations were assayed by the method of Lowry et al. (1951).

Measurement of Mitochondrial Swelling. Pseudoenergetic mitochondrial swelling was evaluated spectrophotometrically according to Famaey (1973). Reaction mixture containing mitochondria (0.15 mg/ml) and 0.15 M KCl in 30 mM Tris-HCl buffer (pH 7.5) was preincubated at 30°C for 5 min. Absorbance at 520 nm was monitored after adding various concentrations of diphenylamine, mefenamic acid, and diclofenac with a Hitachi U-3200 spectrophotometer. Initial slope for the decrease in the absorbance was measured as a rate of the pseudoenergetic swelling.

Binding of Safranine to Mitochondria. Binding of safranine to energetic mitochondria was monitored spectrophotometrically by evaluating the spectral shift according to Colnna et al. (1973). Reaction mixture containing 200 mM sucrose, 2 µM rotenone, 20 mM KCl, 5 mM succinate, and various concentrations of the test drugs in 10 mM Tris-HCl buffer (pH 7.2) was preincubated at 30°C for 5 min. Safranine (12.5 µM) was added to the mixture and the absorption spectra, in the range between 430 and 580 nm, were monitored.

Preparation of Isolated Rat Hepatocytes. Isolated hepatocytes were prepared from male Wistar rats by the collagenase perfusion method of Moldeus et al. (1978) with modifications. The liver was isolated with perfusion of buffer (pH 7.2) consisting of 121 mM NaCl, 6 mM KCl, 12 mM NaHCO₃, 0.74 mM KH₂PO₄, 0.6 mM MgSO₄, and 5 mM glucose. Then the perfusate was changed to the same buffer described above except for containing 4 mM CaCl₂ and 180 U/ml collagenase at a flow rate of 30 ml/min for 12 to 15 min. The hepatocytes were released from the lobe by gentle agitation, and the cell suspension thus obtained was filtered through a nylon mesh (120) and centrifuged (50g, 2 min). The hepatocytes were resuspended in the Schwarz buffer (pH 7.4), which consists of 137 mM NaCl, 5.2 mM KCl, 3 mM Na₂HPO₄, 0.9 mM MgSO₄, 0.12 mM CaCl₂, 5 mM glucose, and 15 mM HEPES, followed by centrifugation (50g, 2 min). The washing procedure was repeated twice and the hepatocytes were finally suspended in the same buffer. All the preparations used in this study were >85% viable as judged routinely by the trypan blue exclusion test.

Incubation of Hepatocytes with Test Compounds. The freshly isolated hepatocytes suspended in the above-mentioned buffer (2 × 10⁶ cells/ml) were preincubated at 37°C with or without fructose (20 mM). After 30 min, a test compound, diphenylamine, mefenamic acid, or diclofenac (500 µM), which was dissolved in methanol to yield a final concentration of 1%, was added with or without oligomycin (10 µg/ml). Fructose (10 mM) was added again 90 min after the onset of the incubation with the test drug. Aliquots of the suspension were removed from the mixture 60 min after the addition of the drug for assay of cellular ATP content and 180 min after the addition for the assay of LDH leakage.

Assay of LDH Activity. LDH activities in the supernatant obtained by centrifugation (50g, 2 min) of the hepatocyte suspension were assayed with LDH-UV-kit Wako as assessed by oxidation of NADH (Wroblewski and La Due, 1955). Cytotoxicity was expressed as a percentage of the total LDH activity, which was obtained from the cells treated with 0.5% Triton X-100.

Assay of ATP Content. ATP contents were assayed by the HPLC method of Jones (1981) with modifications. The hepatocyte suspension (1 ml) was mixed with 0.5 ml of 3 N HClO₄, followed by addition of 0.25 ml of 6 N KOH and 0.5 ml of 1 M Tris-HCl buffer (pH 7.4). After centrifugation (2000g, 10 min), the supernatant with filtration was applied to HPLC. The HPLC conditions were as follows: column, Inertsil ODS (GL Sciences, Tokyo, Japan); mobile phase, 100 mM potassium phosphate buffer (pH 6.0); flow rate, 1.0 ml/min; and UV-detection, 259 nm.

Diclofenac Metabolism by Liver Microsomes. A 1-ml incubation mixture contained 0.5 mg of liver microsomal protein, 10 mM glyceraldehyde-6-phosphate, 2 U glyceraldehyde-6-phosphate dehydrogenase, 1 mM NADPH, 10 mM MgCl₂, 0.1 mM EDTA, and 25 µM diclofenac in 0.15 M potassium phosphate buffer (pH 7.4). After temperature equilibration without NADPH (37°C, 5 min), incubation was started by adding NADPH, performed for 30 min, and terminated with 1 M ice-cold sodium phosphate buffer (pH 5.0). Unmetabolized diclofenac and its oxidative metabolites were extracted into diethylether, the organic layer was evaporated to dryness, and the residue was dissolved in methanol. The samples thus obtained were subjected to following experiments for mitochondrial respiration.

Measurement of Respiration Rates. The rates of oxygen consumption were measured polarographically with a Clark-type oxygen electrode (Model GU-BMP; Iijima Electronics Mfg. Co., Ltd., Aichi, Japan). Respiration buffer (pH 7.4) contained 225 mM sucrose, 10 mM KCl, 5 mM MgCl₂, 5 mM potassium phosphate, 0.5 mM EDTA, and 20 mM Tris-HCl. Mitochondria (1 mg protein/ml) were preincubated at 30°C in 1.6 ml of respiration buffer containing succinate (5 mM) as a respiration substrate. State 3 and state 4 respiration rates were measured after addition of the samples prepared as described above in the presence (state 3) and after exhaustion (state 4) of ADP (87.5 µM). The respiratory control index was calculated as the ratio of state 3/state 4 respiration (Chance and Williams, 1956).

Statistical Analysis. Statistical significance was calculated by the Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Pseudoenergetic Mitochondrial Swelling. Incubation of mitochondria with diphenylamine as well as mefenamic acid and diclofenac caused time-dependent decreases in absorbance at 420 nm, suggesting that diphenylamine induces pseudoenergetic mitochondrial swelling (Fig. 2). Rates of the pseudoenergetic swelling induced by the compounds revealed similar concentration-dependence, whereas effects of diphenylamine were pronounced compared with those of mefenamic acid and diclofenac (Fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Diphenylamine-induced swelling in rat liver mitochondria. Mitochondrial fractions were incubated without substrate for respiration in the absence (a) or presence of diphenylamine at 10 (b), 25 (c), 50 (d), 100 (e), 250 (f), and 500 µM (g). Absorbance was monitored at 520 nm. Results were from three experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Comparison of concentration dependence of mefenamic acid, diclofenac, and diphenylamine on induction of swelling in rat liver mitochondria. Mitochondrial fractions were incubated without substrate for respiration in the presence of various concentrations of mefenamic acid (), diclofenac () and diphenylamine (). Swelling was evaluated by rate for decrease in absorbance at 520 nm, which were calculated from initial slope for their decreases. Results are means ± S.E. of three different experiments. *P < .05, **P < .01, significantly different from control obtained without drug.

Binding of Safranine to Mitochondria. Diphenylamine caused changes in safranine-binding spectra to mitochondria that was energized by succinate oxidation, i.e., an increase in absorbance and a shift in the wavelength of its maximum absorbance (Fig. 4). This safranine spectral shift indicates the decrease in mitochondrial membrane potentials () and also was caused by mefenamic acid and diclofenac (data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effect of diphenylamine on the absorption spectra of safranine. Mitochondria fractions were incubated with succinate in the absence (a) or presence of diphenylamine at 25 (b), 100 (c), 250 (d), and 500 µM (e). Safranine was added to the medium and absorption spectra were monitored in the range between 430 and 580 nm.

Protection by Fructose and Oligomycin against Hepatocyte Injury. Incubation of hepatocytes with mefenamic acid, diclofenac, and diphenylamine diminished cellular ATP contents, followed by LDH leakage. Fructose, a low K_m substrate for glycolysis, partially protected against ATP depletion by the compounds used (Fig. 5). Further addition of oligomycin, which blocks ATPase, pronounced the protection against ATP depletion by mefenamic acid, resulting in full protection. More pronounced results were obtained for protection against the LDH leakage induced by the compounds, although considerable enzyme leakage was observed in the control incubations for 3 h (Fig. 6). Namely, fructose partially protected against the LDH leakage caused by mefenamic acid, diclofenac, and diphenylamine, and additional effects of oligomycin were observed for all of three compounds, resulting in nearly full protection.

View larger version (48K): [in this window] [in a new window]   Fig. 5.   Effect of fructose and oligomycin on decrease in ATP content in hepatocytes induced by mefenamic acid, diclofenac, and diphenylamine. Freshly isolated rat hepatocytes were incubated with or without fructose (F). After 30 min, a test drug, mefenamic acid (Mef), diclofenac (Dcf), or diphenylamine (Dpa) was added with or without oligomycin (O). ATP contents in the hepatocytes were assayed 60 min after addition of the drug. Results are expressed as a percentage of the initial ATP content in the hepatocytes and are means ± S.E. of three different experiments. *P < .05, **P < .01, significantly different from "Non" obtained without the test drug. P < .05, P < .01, significantly different from "control" obtained with the corresponding drug but without fructose.

View larger version (53K): [in this window] [in a new window]   Fig. 6.   Effect of fructose and oligomycin on hepatocyte injury induced by mefenamic acid, diclofenac, and diphenylamine. Hepatocytes were incubated under the same conditions as described in the legend for Fig. 5. Abbreviations also correspond to those of Fig. 5. Viability of the hepatocytes was assessed by LDH leakage 180 min after the addition of the test drug. Results are expressed as a percentage of the total LDH activity and are means ± S.E. of three different experiments. **P < .01, significantly different from "Non" obtained without the test drug. P < .01, significantly different from "control" obtained with the corresponding drug but without fructose.

Effect of Diclofenac Metabolism on Uncoupling Effect. Table 1 shows the effects of diclofenac and the ether extract from the microsomal mixture after metabolism of diclofenac on oxygen consumption by rat liver mitochondria with succinate as a substrate. Diclofenac stimulated basal oxygen consumption (state 4) and inhibited respiration stimulated with ADP (state 3), resulting in a marked decrease in respiration control index, which are the typical characteristics of uncouplers of mitochondrial oxidative phosphorylation. However, the ether extract from the microsomal mixture after metabolism of diclofenac did not affect the respiration control. Because diclofenac in this sample was eliminated to be less than one-third of the initial (data not shown), the oxidative metabolites thus formed were indicated to lose the ability to uncouple mitochondrial oxidative phosphorylation.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have found that diphenylamine as well as mefenamic acid and diclofenac stimulate basal oxygen consumption (state 4 respiration) and inhibit ADP-stimulated respiration (state 3), suggesting that these compounds are uncouplers of mitochondrial oxidative phosphorylation (Masubuchi et al., 1999). To further characterize the effect of diphenylamine on the mitochondrial membrane, the ability to induced swelling was tested. Incubation of mitochondria with diphenylamine in the absence of a substrate caused a time-dependent decrease in absorbance at 520 nm (Figs. 2 and 3), suggesting that diphenylamine as well as mefenamic acid and diclofenac induce pseudoenergized mitochondrial swelling (Banos and Reyes, 1989; Uyemura et al., 1997). Large-amplitude swelling is a typical phenomenon of MPT, which is due to the opening of high-conductance permeability pores in the mitochondrial inner membrane (Bernardi et al., 1994). The MPT was implicated as a causative factor in lethal cell injury, and was recently noted for the relation to Reye's syndrome (Trost and Lemasters, 1996; Lemasters et al., 1998). Uncouplers of oxidative phosphorylation are also inducers of the MPT. However, because the pore opening causes the uncoupling of oxidative phosphorylation, it is possible to cause apparent uncoupling by the MPT. But it was considered that the uncoupling by diphenylamine did not result as a consequence of the MPT, but the protonophoric ability, because cyclosporin A, a specific inhibitor of the MPT (Broekemeier et al., 1989), did not affect the observed uncoupling (Masubuchi et al., 1999).

The electrochemical gradient, which plays an essential role in the production of ATP, is comprised almost by the membrane potential (). We thus measured the membrane potential according to a classical method, safranine binding to mitochondria (Famaey, 1973). Diphenylamine caused changes in safranine-binding spectra to mitochondria that was energized by succinate oxidation, i.e., an increase in absorbance and a shift in the wavelength of its maximum absorbance (Fig. 4), which also were obtained for mefenamic acid and diclofenac. These spectral shifts indicate the decrease in mitochondrial membrane potentials seem to be due to uncoupling of oxidative phosphorylation (Moreno-Sanchez et al., 1999). Mitochondrial depolarization as well as ATP depletion by the uncoupling is the causative factor to impair hepatocytes (Byrne et al., 1999), suggesting its involvement in cell injury. Moreover, it should be emphasized that diphenylamine has similar properties to structurally related NSAIDs.

Fructose, but not glucose, provides cytoprotection against cell damage associated with mitochondrial poisons and prooxidants (Wu et al., 1990). Because fructose is an excellent substrate for the glycolytic pathway, which results in net ATP production, fructose compensates for deficient mitochondrial ATP production and it was proposed as one of the mechanisms for the cytoprotection. Thus, protective effects of fructose were investigated against the hepatocyte injury along with depletion of the ATP induced by mefenamic acid, diclofenac, and diphenylamine. Fructose partially protected against ATP depletion and subsequent LDH leakage (Figs. 5 and 6). Further addition of oligomycin, which blocks ATPase, pronounced the protection against hepatocyte injury, whereas a marked additional effect of oligomycin on ATP content was observed only against the mefenamic acid-induced decrease. Because uncouplers stimulate ATPase, resulting in enhancement of ATP hydrolysis, its inhibition by oligomycin seems to be effective in compensation of cellular ATP content, which was decreased by the potent uncoupler (Nieminen et al., 1994). The effective cytoprotection by fructose and additional oligomycin suggested that decreases in cellular ATP content, mainly caused by the uncoupling of mitochondrial oxidative phosphorylation, were responsible for acute hepatocyte injury induced by diphenylamine and structurally related NSAIDs.

Hepatocyte injury induced by diclofenac has been discussed in relation to oxidative metabolism (Kretz-Rommel and Boelsterli, 1993; Jurima-Romet et al., 1994). Very recently, a couple of oxidative metabolites were identified as the reactive metabolites of diclofenac (Shen et al., 1999; Tang et al., 1999a,b). In relation to the hepatocyte injury by diclofenac, Bort et al. (1999) reported that N-5-dihydroxydiclofenac was a candidate responsible for cell injury by a futile consumption of NADPH, in addition to the impairment of ATP synthesis by diclofenac itself (Bort et al., 1999). Our previous and present results suggested that the hepatocyte injury by NSAIDs was not "metabolism-dependent" but "structure-dependent" (Masubuchi et al., 1998), i.e., diphenylamine is one of the essential structures to cause hepatocyte injury, which is also necessary to inhibit mitochondrial oxidative phosphorylation. Indeed, it was shown that oxidative metabolism of diclofenac lost the ability to uncouple oxidative phosphorylation, confirming the lack of involvement of the metabolite (Table 1). However, because the consumptive oxidation of NADPH was suggested to disrupt the balance of mitochondrial Ca²⁺ uptake and release, to lead to net increase in mitochondrial Ca²⁺, and to result in opening of the permeability transition pore and the onset of MPT (Byrne et al., 1999), it is possible that N-5-dihydroxydiclofenac contributes to mitochondria impairment at the cellular level. The findings could lead to speculation that not only diclofenac but also diphenylamine and mefenamic acid converted to corresponding N-hydroxylamine metabolites, which provide structural importance of diphenylamine to cause hepatocyte injury, whereas it is only speculative because metabolism of diphenylamine itself has been unknown. Therefore, the contribution of the metabolites to the mitochondrial impairment cannot be excluded and is a further interest to elucidate the molecular mechanism for the structural requirement. Regardless, it is concluded that the depletion of cellular ATP induced by the uncoupling of mitochondrial oxidative phosphorylation plays a crucial role in cytotoxicity of diphenylamine and NSAIDs that have the diphenylamine structure.