Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

para-Aminophenol (PAP)² causes selective injury to rat renal proximal tubules (Newton et al., 1982; Tarloff et al., 1989) and is proposed to account, at least in part, for the nephrotoxicity of acetaminophen (APAP; Newton et al., 1985; Mugford and Tarloff, 1997). PAP cytotoxicity may involve enzymatic or nonenzymatic oxidation of PAP to reactive intermediates. Consistent with the notion of direct cytotoxicity, PAP is genotoxic (Hayward and Lavin, 1985; Eiche et al., 1990), nephrotoxic in the isolated perfused kidney (Davis et al., 1983), and cytotoxic to isolated rat renal tubular cells (Klos et al., 1992; Lash et al., 1993) and rabbit proximal tubule suspensions (Lock et al., 1993). However, in situ oxidation to reactive intermediates is inconsistent with the observed selectivity of PAP toxicity, because PAP damages renal but not hepatic cells. Other evidence suggests the involvement of hepatic metabolites of PAP in cytotoxicity. For example, cannulation of the bile duct to prevent hepatic metabolites of PAP from undergoing enterohepatic recirculation partially protected rats against PAP-induced nephrotoxicity (Gartland et al., 1990). Furthermore, pretreating rats with buthionine sulfoximine to deplete GSH protected rats from PAP nephrotoxicity (Gartland et al., 1990) suggesting that nephrotoxic metabolites of PAP generated in liver may include GSH conjugates.

Although glutathione conjugates of PAP (PAP-GSHs) are nephrotoxic in vivo (Fowler et al., 1991, 1994) and cytotoxic in vitro (Klos et al., 1992), it is unclear as to whether sufficient amounts of these conjugates are formed to account for PAP nephrotoxicity. In a previous study examining biliary metabolites of PAP, acetaminophen-O-glucuronide (APAP-Gluc) was identified as the major PAP metabolite appearing in bile with only trace amounts of PAP-GSHs detected (Klos et al., 1992). Therefore, these studies were designed to examine the metabolism of PAP in rat hepatocytes. The results indicate that PAP-GSHs account for over half of the metabolites identified in hepatocytes and quantitatively may contribute to PAP nephrotoxicity.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. All chemicals were purchased from commercial sources as analytical or reagent grade and were used without any additional treatment. 1-Aminobenzotriazole (ABT) was obtained from Dr. Bruce Mico (Hoffman-LaRoche, Inc., Nutley, NJ).

Animals and Hepatocyte Incubations. Male Sprague-Dawley rats (200-225 g) were purchased from Ace Animals, Inc. (Boyertown, PA). All experiments were performed in accordance with the guidelines and practices established by the Institutional Animal Care and Use Committee at the University of the Sciences in Philadelphia. Hepatocytes were prepared using a two-step perfusion and collagenase digestion procedure as previously described (Moldeus et al., 1978; Nyarko and Harvison, 1995). Trypan blue exclusion was used to determine viability. The viability of freshly isolated hepatocytes was 91.8 ± 0.8% (n = 3).

Identification and Quantitation of Metabolites.

HPLC analysis Deproteinized samples from hepatocyte incubations were analyzed by HPLC or HPLC/mass spectrometry (MS). HPLC analysis was performed using a system consisting of an LDC ConstaMetric 3000 pump, a SpectroMonitor 3100 detector set at 254 nm, and an LDC Analytical D-2500 Computing Integrator (Milton Roy Co., Riviera Beach, FL). An Alltech Econosil C₁₈ column, 250 × 4.6 mm, 10 µm, was used. The mobile phase consisted of ACN/H₂O (25:75, v/v) at a flow rate of 1 ml/min or ACN/H₂O (8.5:91.5, v/v) with 0.1 M ammonium acetate (NH₄Ac) at pH 4.7 and a flow rate of 1 ml/min. The injection volume was 20 µl. Quantitation of PAP metabolites was based on the assumption that PAP and metabolites would have about the same UV absorptivity, as has been demonstrated for APAP and its metabolites (Howie et al., 1977).

HPLC/thermospray (TSP)/MS analysis. For analyzing synthetic compounds and metabolites from hepatocytes, we used an HPLC/TSP/MS system consisting of a Waters 600 MS pump, a Waters 900 MS diode array detector, and a Delsi Nermag R30-10 triple quadrupole mass spectrometer (Delsi Nermag Instruments, Argentueil, France) equipped with a Vestec 740C TSP interface (Vestec, Houston, TX). We used a Beckman ODS (250 × 4.6 mm) column, a mobile phase composed of ACN/H₂O (8.5:91.5, v/v) with 0.1 M NH₄Ac at pH 7.5, and flow rate of 1 ml/min. The temperature of the tip heater was 240°C with the source block temperature at 250°C. The scan range of the mass spectrometer was 100 to 1500 mass to charge (m/z).

Gas chromatography (GC)/MS analysis. GC/MS analysis was performed using a GC/MS system consisting of a Hewlett-Packard 5890 series II gas chromatograph, a 5971A mass selective detector, a capillary column of cross-linked polydimethyl silicone (12 m × 0.2 mm i.d. × 0.33-µm film thickness), and a Vectra OS/65 personal computer with Microsoft Chemstation software. The GC column temperature gradient was programmed at 70°C for 2 min, raised to 220°C at 20°C/min and held at 220°C for 2 min. Electron impact ionization mass spectra were generated using 70-eV bombardment with a source temperature at 250°C. The temperatures of the injection port and MS interface were 200 and 250°C, respectively. The mass scan range was 75 to 400 m/z.

NMR ( ¹H-NMR) analysis. ¹H-NMR analyses were performed using a Varian EM390 spectrometer (Varian Associates, Palo Alto, CA). Tetramethylsilane or (trimethylsilyl)propionic-2,2,3,3-d₄ acid sodium salt was used as the internal standard and CDCl₃, d₆-dimethyl sulfoxide, or D₂O was used as the solvent.

Flow injection analysis electrospray ionization mass spectrometry (FIA/ESI/MS). FIA/ESI/MS was done using a Perkin Elmer Sciex API III Biomolecular Mass Analyzer (Sciex, Ontario, Canada) equipped with a Sciex ESI source.

Synthesis of Proposed PAP Metabolites.

4-Aminophenyl-O-glucuronide (PAP-Gluc) PAP-Gluc was synthesized using a method for synthesis of analogous compounds (Focella et al., 1972) with minor modifications. 4-Nitrophenyl glucuronide (0.11 g, 0.33 mmol) was dissolved in 1 ml of 10% Na₂CO₃, and 85% Na₂S₂O₄ (0.14 g, 0.70 mmol) was added at 45°C in small portions with stirring. The reaction mixture was heated to 95°C and held for 30 min. After cooling to room temperature, the reaction mixture was lyophilized to obtain a white powder, melting point 250-260°C (decomposed).

4-Aminophenyl sulfate potassium salt (PAP-O-SO₃K). 4-Nitrophenyl sulfate potassium salt (1.28 g, 5.0 mmol) was dissolved in 8 ml of 10% K₂CO₃, and 85% Na₂S₂O₄ (2.6 g, 12.7 mmol) was added at 45°C in small portions with stirring. The reaction mixture was kept at 95°C for 1 h. After cooling, filtering, and drying, 0.92 g of white crystals were obtained (melting point 265-270°C, decomposed).

4-Hydroxyformanilide (FPAP). We used a method for formylating aniline (Fieser and Jones, 1955) with minor modifications. PAP (5.5 g, 0.05 mol) was mixed with formic acid (5.0 g) and 30 ml of toluene. The mixture was distilled slowly until 10 ml of solvent remained. The residue was washed with 10 ml of petroleum ether and filtered to yield 5.9 g of a grayish-white powder, melting point 128-130°C. Yield: 85.3%.

p-Benzoquinoneimine (PBQI). We followed a previously reported method (Eckert et al., 1990) with minor modifications. PAP (0.22 g, 2 mmol) was dissolved in 40 ml of ACN, and anhydrous magnesium sulfate (1 g, 8.3 mmol) was added. The mixture was stirred for 3 min, lead dioxide (5.0 g, 20 mmol) was added, and the reaction was carried out in the dark for 15 min. Filtration yielded a yellow solution that was kept in the dark for use in preparation of PAP-thioethers.

PAP-S-glutathione conjugates (PAP-GSH). A previously reported method (Eckert et al., 1990) was used with modifications. PBQI solution was mixed with a solution containing GSH (0.6 g, 2.0 mmol), sodium acetate (0.5 g, 6.1 mmol), and sodium hydroxide (0.1 g, 2.5 mmol) in 10 ml of water (pH 11). The mixture was stirred in the dark at room temperature for 10 min. After completion of the reaction, the top layer containing unreacted PBQI and PAP was discarded, and the bottom layer containing PAP-GSHs was mixed with an equal amount of ACN to precipitate the solid products. After filtration, the cake was suspended in 20 ml of methanol, filtered, and dried to yield 0.37 g of a greenish-yellow powder, melting point 62-64°C (started to decompose at about 100°C).

PAP-cysteine conjugates (PAP-Cys). A solution of PBQI was mixed with a solution containing cysteine (0.24 g, 2 mmol), sodium acetate (0.5 g, 6.1 mmol), and sodium hydroxide (0.05 g, 1.3 mmol) in 10 ml of water (pH 11). The mixture was stirred in the dark for 10 min. The top layer containing unreacted PBQI and PAP was discarded, and the bottom layer was mixed with 20 ml of 90% methanol. The mixture was filtered and the cake was suspended in 20 ml of methanol. After filtration and drying, 0.2 g of a brown solid was obtained (melting point 234-238°C with decomposition).

PAP-S-(N-acetylcysteine) (PAP-NACys) conjugates. A solution of PBQI was mixed with a solution containing N-acetylcysteine (0.32 g, 2 mmol), sodium acetate (0.5 g, 6.1 mmol), and sodium hydroxide (0.05 g, 1.3 mmol) in 10 ml of water (pH 11). The mixture was stirred in the dark for 10 min. The solvent was evaporated under a reduced pressure to 20% of the original volume. After filtration, the cake was washed with 2 ml of water to yield 0.14 g of a brownish-black powder that started to melt at 50°C with decomposition.

Data Analysis and Statistical Analyses. All results are expressed as mean ± S.E. of a minimum of three determinations. Data were analyzed by one-way ANOVA followed by a Student-Newman-Keuls test when appropriate. The probability level for statistical significance was P < .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Identification of PAP Metabolites by HPLC/MS and GC/MS. We analyzed synthesized postulated metabolites of PAP using HPLC/TSP/MS; results are summarized in Table 1. We observed molecular ions in the spectra for PAP-Gluc, PAP-NACys, PAP, FPAP, and APAP. We were unable to detect molecular ions of PAP-Cys using HPLC/TSP/MS in positive ion mode but were able to detect molecular ions using negative ion mode. There were no molecular ions for PAP-GSH, APAP-Gluc, or GSH conjugates of APAP (APAP-GSH), most likely because these compounds underwent thermal degradation during HPLC/TSP/MS analysis.

Mass Balance of PAP and Metabolites. We assumed that equal molar concentrations of PAP and its metabolites would have approximately the same UV absorptivity at 254 nm. We tested this assumption by calculating the total recovery of PAP and metabolites obtained from hepatocyte samples as determined by HPLC analysis. The results are listed in Table 2, where PAP indicates hepatocytes incubated with PAP alone and PAP+ABT indicates hepatocytes preincubated with ABT. The calculated recovery of PAP and metabolites correlated well with the original concentration of PAP (2.3 mM). There were no significant differences among time intervals or between treatments. The results support the assumption that PAP and its metabolites have approximately equal UV absorbance regardless of the ratio of PAP to its metabolites.

Viability of Rat Hepatocytes. There were no significant differences between viability of hepatocytes incubated in the presence or absence of PAP. Viability was determined by trypan blue exclusion. Initial viability for control and PAP-treated hepatocytes was 91.8 ± 0.85%. Viability in hepatocytes incubated in the absence of PAP declined to 75.4 ± 4.11% after 5 h whereas viability in hepatocytes incubated in the presence of PAP declined to 74.9 ± 6.01%.

1

1

1

1

PAP Metabolism in Rat Hepatocytes. PAP was extensively metabolized by rat hepatocytes and less than 5% of the original PAP remained at the end of the incubation. As shown in Fig. 1, PAP metabolism was not significantly altered by preincubating hepatocytes with ABT. Metabolism appeared to follow first-order kinetics with apparent rate constants of 6.81 ± 0.95 × 10¹ h¹ and 5.49 ± 1.74 × 10¹ h¹ and half-lives of 1.02 and 1.26 h for control and ABT-pretreated hepatocytes, respectively.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Disappearance of PAP from the incubation medium when hepatocytes (5 × 106 cells/ml) were incubated in Krebs-Henseleit buffer in the presence () or absence () of ABT, a suicide substrate inhibitor of cytochromes P450. Symbols indicate means ± S.E. of three determinations. Means were not significantly different at any time point (P > .05). Inset: Natural logarithmic decline of PAP over 5-h incubation. Linear regression analysis lines are included where the solid line represents hepatocytes incubated with PAP and the dashed line represents hepatocytes incubated with PAP + ABT.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Formation of PAP-GSHs by rat hepatocytes (5 × 106 cells/ml) incubated with 2.3 mM PAP in the presence () or absence () of ABT. Symbols indicate means ± S.E. of three determinations. Means were not significantly different at any time point (P > .05).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Formation of PAP-NACys conjugates by rat hepatocytes (5 × 106 cells/ml) incubated with 2.3 mM PAP in the presence () or absence () of ABT. Symbols indicate means ± S.E. of three determinations. Means were not significantly different at any time point (P > .05).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Analysis of PAP metabolites formed by hepatocytes. A, MS analysis of PAP-GSHs formed by hepatocytes. No molecular ions were detected. B, MS analysis of PAP-NACys conjugates formed by hepatocytes. No molecular ions were detected.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Formation of PAP-Gluc by rat hepatocytes (5 × 106 cells/ml) incubated with 2.3 mM PAP in the presence () or absence () of ABT. Symbols indicate means ± S.E. of three determinations. Means were not significantly different at any time point (P > .05).

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Formation of APAP by rat hepatocytes (5 × 106 cells/ml) incubated with 2.3 mM PAP in the presence () or absence () of ABT. Symbols indicate means ± S.E. of three determinations. Means were not significantly different at any time point (P > .05).

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Analysis of PAP metabolites formed by hepatocytes. A, MS analysis of PAP-Gluc conjugates formed by rat hepatocytes. The molecular ion was detected at m/z 286 and peaks matched those of synthesized PAP-Gluc conjugates. B, MS analysis of APAP formed by rat hepatocytes. The molecular ion was detected at m/z 152 and peaks matched those of authentic APAP. C, MS analysis of FPAP formed by rat hepatocytes. The molecular ion was detected at m/z 138 and matched that of synthetic FPAP (see Table 1).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We found that rat hepatocytes readily converted PAP to two major metabolites, PAP-GSH and PAP-NACys conjugates. Minor metabolites included PAP-Gluc, APAP, FPAP, APAP-Gluc, and APAP-GSH conjugates (Fig. 8). A small amount (<5%) of PAP remained unreacted after a 5-h incubation with hepatocytes. PAP-GSH, PAP-Gluc, APAP, APAP-Gluc, and APAP-GSH conjugates have been observed previously as PAP metabolites (Eckert, 1988; Temellini et al., 1991; Klos et al., 1992) whereas PAP-Cys, PAP-NACys, and FPAP are novel metabolites.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Summary of PAP and metabolites observed during incubation of rat hepatocytes (5 × 106 cells/ml) with 2.3 mM PAP. Percentages below compound labels refer to the fraction of PAP identified as each metabolite at the end of a 5-h incubation.

PAP-GSH conjugates are reported to be equitoxic or up to 4-fold more toxic than PAP when administered in vivo (Fowler et al., 1994) or when incubated with renal tubular cells (Fowler et al., 1991; Klos et al., 1992). However, bile contained only trace amounts of PAP-GSH whereas the major metabolite in bile was PAP-Gluc (Klos et al., 1992), raising questions concerning the role of PAP-GSHs in nephrotoxicity. Our data show that hepatocytes converted about 50% of PAP to PAP-GSHs, supporting the idea that PAP-GSHs may mediate the nephrotoxicity of PAP. In addition, we observed depletion of both renal and hepatic GSH after PAP administration in vivo (Shao and Tarloff, 1996), consistent with the concept that PAP undergoes GSH conjugation in liver. The differences observed between our results and those of Klos and coworkers (1992) may relate to differences between Sprague-Dawley and Wistar rats and to differences in experimental procedures.

We anticipated that sulfate conjugates of PAP would be formed by hepatocytes and prepared synthetic sulfate conjugates for mass spectral comparison. However, we did not find peaks corresponding to proposed sulfate conjugates and we were able to account for virtually all of the PAP originally present in our incubations. Thus, we suggest that in our hands, rat hepatocytes do not form sulfate conjugates from PAP. The absence of sulfate conjugates is in contrast to a previous report in which detectable amounts of a sulfate conjugate of PAP were identified during incubations with hepatocytes from Wistar rats (Evelo et al., 1984). In contrast to our results with hepatocytes, we found appreciable quantities of PAP-O-SO₃H in urine of rats treated with PAP (Z. Yan and J. B. Tarloff, unpublished observations). Currently, we are unable to reconcile the discrepancy between absence of PAP-O-SO₃H from hepatocytes and presence in rat urine. It is possible that sulfate was depleted from hepatocytes during incubation and the buffer contained insufficient sulfur to replace that lost during incubation. However, Moldeus (1978) measured sulfate conjugates of APAP using Krebs-Henseleit buffer, the same buffer used in our studies, suggesting that the amount of sulfur in the buffer was sufficient to support sulfation of APAP.

PAP-NACys conjugates may arise from additional metabolism of PAP-GSH. The liver contains -glutamyl transpeptidase, the first enzyme involved in the degradation of GSH, although at lower levels than kidney. However, we did not observe appreciable quantities of PAP-Cys-Gly or PAP-Cys, thioethers that would be expected during PAP-GSH degradation. It is equally possible that cysteine may serve as an alternate scavenger for reactive PAP intermediates, and the PAP-Cys conjugates formed may undergo additional metabolism by N-acetyltransferases present in hepatocytes. Certainly, N-acetylation occurs in hepatocytes because APAP is a metabolite of PAP. The concentration of cysteine in liver is about 0.2 µmol/g whereas the concentration of GSH in liver is about 6 µmol/g (Potter and Tran, 1993). Thus, it is theoretically possible that cysteine may serve as an alternate acceptor for radical intermediates generated from PAP.

The pattern of PAP-Gluc formation was interesting in that low concentrations were present during the first 3 h of the incubation and the amount of PAP-Gluc increased appreciably during the last 2 h. Although we did not measure concentrations of GSH or cysteine in our hepatocytes, it is possible that formation of PAP-Gluc increased in response to decreases in hepatocyte thiol concentrations.

FPAP was a unique metabolite of PAP. The results of our experiments support the postulated structure for FPAP but the mechanism of formation is not clear. Yoshida and coworkers observed p-chloroformanilide as a metabolite of chloroaniline formed during metabolism of p-chloronitrobenzene (Yoshida et al., 1991, 1992). They postulated that p-chloroformanilide was a product of thermal degradation of an acidic p-chloroaniline metabolite. S-Triazo[3,4]phthalazine is a minor metabolite of hydralazine that may be formed through formylation and cyclization reactions (Noda et al., 1979; Lacagnin et al., 1986). Alternatively, FPAP may be formed through nonenzymatic combination of PAP with carbon dioxide. Carbon dioxide may combine nonenzymatically with primary and secondary aliphatic amines to form carbamic acids (Straub et al., 1988; Tremine et al., 1989). In turn, carbamic acids may form the corresponding carbamoyl-O-glucuronides via UDP-glucuronosyltransferase. Carbamic acids and carbamoyl-O-glucuronides are sensitive to pH and temperature and are readily degraded. Carbon dioxide is a natural component of biological systems as well as a portion of the gas mixture (O₂/CO₂) used during hepatocyte incubations, and it is possible the FPAP was formed by degradation of PAP-carbamic acid formed nonenzymatically through combination of PAP with carbon dioxide.

Importantly, hepatocyte viability was unaltered with PAP in the incubation medium. This observation is consistent with the known cytotoxicity of PAP for renal tubular epithelial cells (Newton et al., 1982; Evelo et al., 1984) and is in contrast to the notion of autoxidation contributing to PAP cytotoxicity. The lack of toxicity toward rat hepatocytes may be due to different pathways of metabolism in renal tubular cells or enhanced detoxification processes in hepatocytes. However, lack of toxicity in hepatocytes is at variance with the notion that PAP-GSHs are cytotoxic. Because hepatocytes were exposed to significant amounts of PAP-GSH, it is reasonable to expect that some cytotoxicity should occur. Potential explanations for lack of toxicity due to PAP-GSHs include lower concentrations of -glutamyl transpeptidase in liver (J. B. Tarloff and S. C. Ring, unpublished observations) and dilution of PAP-GSHs in the incubation medium, possibly preventing these conjugates from reaching toxic concentrations in hepatocytes.

Oxidation of PAP, either enzymatic or nonenzymatic, occurs before conjugation. Several enzymes may mediate oxidative metabolism of PAP, including cytochromes P450 and prostaglandin H synthase (Calder et al., 1979; Josephy et al., 1983). Our data suggest that cytochromes P450 are not involved in PAP metabolism because inclusion of ABT, a suicide substrate inhibitor of cytochrome P450, did not alter the extent or pathways of PAP metabolism. These observations are consistent with our previous results in renal tubules, wherein ABT failed to attenuate cytotoxicity due to hepatic bioactivation of PAP (Shao et al., 1997). The role of cytochromes P450 in PAP-induced nephrotoxicity is unclear. Calder and coworkers (1979) suggested that cytochromes P450 were not involved in PAP nephrotoxicity after observing inconsistent effects with inhibitors and inducers of this enzyme system. In contrast, Klos and coworkers found that inhibition of cytochromes P450 by SKF525A reduced the cytotoxicity of PAP in isolated proximal tubule cells from Wistar rats (Klos et al., 1992). It is possible that PAP is metabolized differently in liver and kidney and that hepatic metabolism of PAP does not require cytochromes P450 whereas renal PAP metabolism is cytochrome P450-dependent.

In conclusion, we found that hepatocytes rapidly metabolized PAP to two major products, PAP-GSH and PAP-NACys. Cytochrome P450-dependent oxidation of PAP was not apparent because a suicide substrate inhibitor of cytochromes P450, ABT, failed to alter the metabolic profile. Quantitatively, PAP-GSH are formed in sufficient amounts to account for the nephrotoxicity of PAP. PAP-GSHs accounted for about 50% of PAP initially present in our incubation medium and PAP-GSHs are at least equitoxic, if not more toxic, than PAP itself. Even at relatively high concentrations (2.3 mM), PAP was not cytotoxic to hepatocytes, possibly due to the rapid and efficient metabolism. These studies lend credence to the idea that PAP-GSHs are nephrotoxic.