Acetaminophen (APAP; N-acetyl-p-aminophenol)–induced liver disease is the leading cause of drug-related liver failure in humans.1 High APAP doses cause: 1) rapid depletion of GSH, 2) saturation of the sulfonation and glucuronidation pathways, and 3) increased production (and longer half-life) of the toxic reactive intermediate, N-acetyl-p-benzoquinone imine (NAPQI), which consequently leads to cellular damage and death.2 Liver injury associated with high doses of APAP shows a marked dose threshold because of the protective action of hepatic GSH.3 A decreased capacity to metabolize APAP by sulfonation could put more pressure on the existing GSH pool and hence lower the APAP dose threshold for liver injury. Sulfate conjugation (sulfonation), a major step in the metabolism of APAP in humans, relies on the availability of inorganic sulfate (SO42−) and its universal sulfonate donor 3′-phosphoadenosine 5′-phosphosulfate (PAPS).4 SO42− is obtained directly from the diet and from the oxidation of sulfur containing amino acids, cysteine, and methionine.4 Because of its hydrophilicity, SO42− needs to enter cells via plasma membrane sulfate transporters.4 We cloned a kidney sulfate transporter, NaS1, which is expressed on the brush border membrane of renal proximal tubular epithelial cells, where it is involved in the first step of SO42− reabsorption.5 Recently, we generated NaS1-null (Nas1−/−) mice, which have very low (≈0.2 mmol/L) serum sulfate levels compared to those (≈1 mmol/L) in wild-type (Nas1+/+) mice.6 In this study, we identified two loss-of-function polymorphisms in the human NaS1 gene and characterized the Nas1−/− mouse showing increased APAP hepatotoxicity, suggesting disturbed NaS1 function produces hyposulfatemia that leads to increased drug-induced liver damage.
Sulfate is required for detoxification of xenobiotics such as acetaminophen (APAP), a leading cause of liver failure in humans. The NaS1 sulfate transporter maintains blood sulfate levels sufficiently high for sulfonation reactions to work effectively for drug detoxification. In the present study, we identified two loss-of-function polymorphisms in the human NaS1 gene and showed the Nas1-null mouse to be hypersensitive to APAP hepatotoxicity. APAP treatment led to increased liver damage and decreased hepatic glutathione levels in the hyposulfatemic Nas1-null mice compared with that in normosulfatemic wild-type mice. Analysis of urinary APAP metabolites revealed a significantly lower ratio of APAP-sulfate to APAP-glucuronide in the Nas1-null mice. These results suggest hyposulfatemia increases sensitivity to APAP-induced hepatotoxicity by decreasing the sulfonation capacity to metabolize APAP. In conclusion, the results of this study highlight the importance of plasma sulfate level as a key modulator of acetaminophen metabolism and suggest that individuals with reduced NaS1 sulfate transporter function would be more sensitive to hepatotoxic agents. (HEPATOLOGY 2006;43: 1241–1247.)
Materials and Methods
Xenopus laevis Oocytes and Transport Measurements.
Mature Xenopus laevis females were purchased from the African Xenopus Facility C.C. (Noordhoek, South Africa). Stage V and VI oocytes from X. laevis were maintained at 18°C in a modified Barth's solution (MBS; 88 mmol/L NaCl, 1 mmol/L KCl, 0.82 mmol/L MgSO4, 0.4 mmol/L CaCl2, 0.33 mmol/L Ca[NO3]2, 2.4 mmol/L NaHCO3, 10 mmol/L HEPES/Tris [pH 7.4], gentamicin sulfate 20 mg/L). Oocytes were injected with either 50 nL of water (control) or 5 ng of hNaSi-1 cRNA using a Nanojet automatic injector (Drummond Scientific Co., Broomall, PA). For cRNA synthesis, the pSPORT:hNaSi-1 plasmid was linearized by NotI digestion; the cDNA was transcribed in vitro using T7 RNA polymerase (Promega), and the resulting capped cRNA was dissolved in H2O before use. Uptake of 35S-sulfate was performed on day 3 postinjection. Briefly, 10 oocytes (individual data points) were washed at room temperature for 2 min in solution A (100 mmol/L NaCl, 2 mmol/L KCl, 1 mmol/L CaCl2, 1 mmol/L MgCl2, 10 mmol/L HEPES/Tris [pH 7.5]), then placed in 100 μL of solution A additionally containing 0.1 mmol/L Na2SO4 and 20 μCi/mL 35SO42− (New England Nuclear Radiochemicals) for 1 hour. The oocytes were washed 3 times with ice-cold solution A, lysed with 1% SDS, dissolved in scintillant (Emulsifier Safe, Canberra Packard), and counted by liquid scintillation spectrometry.
Animals and Treatment.
Male wild-type and Nas1−/− mice were housed at a constant temperature (23°C ± 1°C) with a 12-hour light/12-hour dark cycle. Mice were weaned at 3 weeks of age and were then fed a standard rodent chow (No. AIN93G; Glen Forrest Stockfeeders, Glen Forrest, Western Australia) and water ad libitum. Acetaminophen (APAP; Sigma Aldrich) was dissolved in saline and filter-sterilized. At about 3 months of age, the mice received intraperitoneal injection of either saline (control) or APAP at a dose of 125, 250, or 500 mg/kg (n = 4-8 per group). Mice were sacrificed 2, 4, 5, or 12 hours after administration of the saline or APAP. All experiments conformed to the guidelines of the University of Queensland Animal Ethics Committee.
A modified turbidometric assay was used to measure free sulfate levels.7 Serum was deproteinized with trichloroacetic acid. Then 0.5% barium chloride in 0.01% agarose was added, and the samples were read at 500 nm using a microtiter plate reader (XReadPlus, Tecan). A calibration curve was prepared using potassium sulfate standards. Serum and urine sulfate levels were normalized to total protein levels, which were quantitated using a protein microassay kit (Bio-Rad).
APAP Metabolite Measurement.
Urine was collected 2 hours after APAP treatment and stored at −20°C until needed. Immediately prior to analysis, urine samples were thawed, 10 μL of a 1 mmol/L internal standard (IS; 3-acetamidophenol, Sigma) was added, and the sample was diluted 10-fold with eluent A (0.1% [v/v] aqueous formic acid solution). Each sample was then mixed by vortexing and was centrifuged at 10,000g for 5 minutes before injection. The APAP (Sigma), paracetamol glucuronide (PG; Sigma), and paracetamol sulfate (PS; Toronto Research Chemicals, Inc., North York, Ontario, Canada) in the samples were identified and quantitated (in duplicate) by reference to appropriate calibration curves constructed using standards prepared by adding known amounts of these compounds to urine from untreated animals. The separation of compounds was carried out using an Agilent binary HPLC system and a Cogent phenyl 100 Å (2.1 × 50 mm, 5 μm) HPLC column (Microsolv Technology Corporation, Long Branch, NJ). An API 3000 tandem mass spectrometer equipped with a turbo ion spray (Applied Biosystems, Foster City, CA) was used to detect the separated compounds. Chromatography was performed at ambient temperature using a flow rate of 0.2 mL/min. After injection of each sample/standard (30 μL), analytes were separated with an isocratic elution using 3% eluent A for 15 min. The column was then flushed with 100% eluent B (0.1% [v/v] formic acid solution in acetonitrile) and re-equilibrated for 10 min with 3% eluent A prior to injection of each sample or standard. The mass spectrometric detector was set to multiple reaction monitoring (MRM) mode, where a specific transition of molecular ion → fragment was monitored. The fragmentations of the 152 m/z ion → 110 m/z ion (for APAP and IS), the 232 m/z ion → 152 m/z ion (for PS), and the 328 m/z ion → 152 m/z ion (for PG) were monitored for each chromatographic run. A dwell time of 1 second was used for all transitions. The resolution of both quadrupoles was 1 amu. On the basis of its MRM response, quantitation of each analyte was performed on duplicates of each sample with reference to calibration curves prepared using standard compounds; the internal standard was used to compensate for sample losses during extraction and analysis.
Liver samples (200 mg) were homogenized in 1 mL of cold MES buffer (0.2 mol/L 2-N-morpholino ethanesulfonic acid, 50 mmol/L potassium phosphate [pH 6.0], 1 mmol/L EDTA). Hepatic glutathione (GSH) levels were assayed by using a commercially available kit (Cayman Chemical Company).
Blood samples were centrifuged at 9,000g for 5 minutes at 4°C. Serum alanine aminotransferase (ALT) concentrations were assayed by a commercial pathology laboratory (Sullivan Nicolaides Pathology, Taringa, Queensland, Australia) using a Vitros 250 chemistry analyzer (Ortho-Clinical Diagnostics, New York).
Liver and kidney samples were placed in approximately 50 volumes of 10% buffered formalin and fixed for 3 days before paraffin embedding. Embedded tissue was sectioned at a thickness of 4 μm, hematoxylin and eosin stained, and examined by light microscopy. Liver sections were scored for the severity of degeneration and necrosis using a scale from 0 to 5, as previously described.8
Analysis of RNA.
Total RNA was extracted from liver and kidney using standard procedures (TRIzol reagent; Invitrogen). For RT-PCR, total RNA (2 μg) was reverse-transcribed using Moloney murine leukemia virus reverse transcriptase (Invitrogen). Primers (Table 1) were used to amplify CYP1A2 (546 bp), CYP2E1 (468 bp), CYP3A11 (1,524 bp), GSTPi (314 bp), sat-1 (831 bp), and dtdst (748 bp) cDNA fragments. Each cDNA was coamplified with GAPDH (control; 983 bp) using cycle parameters of 94°C for 1 minute; followed by 15-25 cycles of 94°C for 1 min, 57°C (CYP2E1, GSTPi), 58°C (dtdst), 65°C (CYP1A2) or 68°C (CYP3A11, sat-1) for 1 minute, and 72°C for 1 minute. PCR products in the linear phase of amplification were analyzed by densitometry (Scion Image Beta 4.0.2) and normalized to GAPDH.
|cDNA||Direction||Sequence (5′ to 3′)|
Data obtain from individual animals are expressed as means ± SEM. The statistical significance of differences between Nas1+/+ and Nas1−/− groups was assessed by an unpaired Student's t-test, with P < .05 considered significant. The statistical significance of the differences of Nas1+/+ and Nas1−/− groups with their respective controls (at time 0, or saline treatments) and of the differences between wild-type and mutant NaS1 protein groups for sulfate uptake studies was evaluated using a one-way ANOVA, followed by Duncan's multiple range test, with P < .05 considered significant. Differences in histology scores were analyzed by the Mann-Whitney test, with the significance level set at P < .05.
Results and Discussion
Loss-of-Function Polymorphisms in Human NaS1 Gene.
The open-reading frame of the human NaS1 gene contains a nonsense mutation (R12X) and an amino acid substitution (N174S), with reported allelic frequencies of 0.36% and 26.99%, respectively (http://www.pharmgkb.org/views/index.jsp?objId=PA322&objC1s=Gene&view=VariantPosition). Expressed in Xenopus oocytes,4 R12X abolished all sulfate transport, whereas N174S reduced sulfate transport by more than 60% compared with that in wild-type NaS1 (Fig. 1). N174 residue in NaS1 is conserved in all eukaryotic species, suggesting that N174S disrupts a highly conserved amino acid and is frequently found in the normal human population. To our knowledge, this is the first study to report a loss-of-function polymorphism in NaS1.
Sulfate, APAP Metabolites, and GSH Levels in APAP-Treated Mice.
Because APAP administration has been shown to cause plasma sulfate depletion in humans and rodents,9, 10 we aimed to characterize APAP detoxification in a hyposulfatemic mouse model with reduced NaS1 function.6 After a single dose of APAP (250 mg/kg), no change in serum sulfate levels was observed in Nas1−/− mice, whereas in Nas1+/+ mice serum sulfate levels were significantly reduced, by about 30%, after 2 hours and were restored to basal levels after 12 hours (Fig. 2A). This suggests that APAP cannot further deplete the low serum sulfate concentration in Nas1−/− mice, whereas APAP depletes sulfate (after 2 h) from the serum of wild-type mice, as previously shown in rats.11 After a single APAP dose (250 mg/kg), urinary sulfate levels were decreased after 2 and 12 hours in both Nas1+/+ and Nas1−/− mice (Fig. 2B), suggesting sulfate is being utilized for hepatic APAP detoxification. Analysis of urinary APAP metabolites after 2 hours revealed a lower APAP-sulfate (PS) to APAP-glucuronide (PG) ratio (reduced by ≈83%) in Nas1−/− mice compared with that in Nas1+/+ mice (Fig. 3A). This suggests that the low serum sulfate levels in Nas1−/− mice may have been limiting the supply of sulfate for APAP metabolism. Serum sulfate levels influence the rate of sulfonation reactions, with sulfate depletion leading to decreased PAPS synthesis and availability,12 as well as to sulfate depletion in the liver.11 Because the depletion of hepatic GSH levels and the rate of cellular GSH repletion are critical factors in APAP hepatotoxicity,13, 14 we determined if liver GSH levels were lower in Nas1−/− mice than in Nas1+/+ mice. After a single dose of APAP (250 mg/kg), GSH levels after 2 hours were significantly lower in Nas1−/− mice (reduced by ≈90%) than in wild-type mice (reduced by ≈60%) and were restored to basal levels after 12 hours in both Nas1+/+ and Nas1−/− mice (Fig. 3B), suggesting that the hyposulfatemia in Nas1−/− mice exacerbates GSH depletion after APAP treatment.
Enhanced APAP-Induced Hepatotoxicity in Nas1−/− Mice.
We measured serum alanine aminotransferase (ALT) levels in Nas1+/+ and Nas1−/− mice as an indicator of APAP-induced liver damage. Twelve hours after administration of APAP (250 mg/kg), a more than threefold increase in serum ALT levels was observed in Nas1−/− mice compared to that in Nas1+/+ mice (Fig. 4A), suggesting increased APAP-induced liver injury in the Nas1−/− mice. This was supported by histological findings of increased cellular damage in the Nas1−/− mice, which showed more extensive liver necrosis (Fig. 4C; Table 2) compared to that in the wild-type mice (Fig. 4B; Table 2). Two hours after a single APAP dose of 500 mg/kg, vacuolation in the cell cytoplasm of hepatocytes was observed in the Nas1−/− mice (Fig. 5A), whereas the overall cellular architecture was still intact in the Nas1+/+ mice (Fig. 5B). After 5 hours, the livers of Nas1−/− mice showed hemorrhaging in the lobular areas (Fig. 5C), which was absent in the Nas1+/+ mice (Fig. 5D; Table 2). These data suggest that hyposulfatemic Nas1−/− mice are more sensitive to APAP-induced hepatotoxicity than are their normosulfatemic Nas1+/+ littermates. Because NaS1 mRNA is not expressed in the liver,15 we propose that the observed differences in hepatotoxicity in the Nas1−/− mice is most likely a result of reduced sulfate availability in the liver as a consequence of the hyposulfatemia found in the Nas1−/− mice.6
|250 mg APAP/kg, t = 12 h|
|500 mg APAP/kg, t = 5 h|
Renal Histology and Urinary Protein Levels.
APAP nephrotoxicity has been observed in humans,16 and large doses of APAP (1,200 mg/kg) cause renal injury in mice.17 We performed histological analysis of kidney tissue sections from the Nas1+/+ and Nas1−/− mice 5 hours after administration of APAP (500 mg/kg), but no signs of cellular damage were observed (data not shown). Under these conditions, no significant differences were observed in urinary protein levels of APAP-treated Nas1+/+ (0.76 ± 0.24 mg/mL, n = 4) and Nas1−/− (1.18 ± 0.21 mg/mL, n = 5) mice, as well as in saline-treated Nas1+/+ (1.44 ± 0.01 mg/mL, n = 4) and Nas1−/− (1.04 ± 0.31 mg/mL, n = 5) mice. This suggests that (under these conditions) APAP enhances hepatotoxicity in Nas1−/− mice (compared to Nas1+/+ mice) without causing renal damage. These findings, together with studies showing that N-acetylcysteine protects against hepatotoxicity but not nephrotoxicity,18 suggest that different mechanisms are involved in APAP-induced renal and hepatic injury.
Liver Sulfate Transporter mRNA Levels.
Sulfate conjugation requires a sufficient supply of intracellular SO42−,19 which is dependent on the delivery of sulfate into cells via plasma membrane sulfate transporters.4 In the present study, we tested the effect of APAP on the mRNA expression of two sulfate transporters expressed in the liver: diastrophic dysplasia sulfate transporter (dtdst) and sulfate anion transporter-1 (sat-1).20, 21 No change in dtdst mRNA expression was observed, whereas sat-1 mRNA expression was increased after the administration of APAP in Nas1−/− mice (but not in Nas1+/+ mice), compared to that in saline-treated mice (Fig. 6). The induction of sat-1 mRNA levels in APAP-treated Nas1−/− mice is most likely a compensatory response to low hepatic sulfate levels as a consequence of hyposulfatemia. We have also shown increased hepatic phenol sulfotransferase activity in Nas1−/− mice under basal conditions.6 However, despite elevated sat-1 expression and sulfotransferase activity6 in Nas1−/− mice, it appears these two factors may not be sufficient to compensate for the low blood sulfate levels in order to overcome the increased APAP-induced hepatotoxicity in Nas1−/− mice.
Liver P-450 and GSTPi mRNA Levels.
It has been proposed that the mechanism of APAP hepatotoxicity is initiated through production of NAPQI by cytochrome P-450 metabolism.22 Because P-450 and glutathione S-transferase (GSTPi) gene expression is linked to APAP hepatotoxicity in mice,14, 23 we measured the mRNA levels of CYP1A2, CYP2E1, CYP3A11, and GSTPi in Nas1−/− and Nas1+/+ mice after APAP administration. After 2 hours, liver GSTPi and CYP3A11 mRNA levels increased in a dose-dependent (125, 250, and 500 mg/kg APAP) manner in Nas1−/− mice, whereas they were increased only at the 500 mg/kg APAP dose in Nas1+/+ mice (Fig. 6). CYP2E1 and CYP1A2 mRNA levels were similarly increased in Nas1−/− and Nas1+/+ mice after APAP treatment (Fig. 6). The increased level of CYP3A11 mRNA and liver injury in APAP-treated Nas1−/− mice are consistent with an association of increased CYP3A11 enzyme levels with liver damage in mice.24 The induction of GSTPi mRNA levels would be a compensatory mechanism in response to the depleted GSH levels in Nas1−/− mice, as previously proposed.25 Studies of GSTPi-null mice indicated that GSTPi potentiates APAP-induced toxicity,14 suggesting that elevated GSTPi mRNA levels in Nas1−/− mice may be contributing to their increased hepatotoxicity.
Sulfate and Detoxification of APAP.
Current therapies for APAP overdose include oral or intravenous administration of the sulfhydryl compound N-acetylcysteine, which acts as a GSH precursor and sulfate donor and has been shown to increase serum sulfate levels.26 Zinc sulfate or sodium sulfate treatments also increase serum sulfate levels and have been shown to be beneficial in preventing APAP-induced hepatotoxicity.27, 28 Dietary intake of sulfate may also play a role in APAP metabolism, as shown by the reduced sulfonation capacity and increased APAP-induced hepatotoxicity in rats fed sulfur-deficient diets.29 In humans, the consumption of drinking water containing high sulfate levels has been associated with increased serum sulfate concentrations and was shown to prevent APAP-induced depletion of serum sulfate levels.30 In addition, individuals with low blood sulfate levels have a reduced capacity to sulfonate APAP.31–35 Hyposulfatemia has been found in the neurological diseases Alzheimer's, Parkinson's, motor neuron disease, and autism.33, 34 However, the etiology of hyposulfatemia in these disorders is not yet known. Taken together, these findings underscore the essential role of sulfate in the detoxification of APAP and warrant further studies of APAP toxicity in individuals with loss-of-function polymorphisms in the NaS1 gene. In conclusion, we have shown that NaS1 loss-of-function polymorphisms exist in the human population and that disrupting the NaS1 sulfate transporter produces hyposulfatemia, which increases sensitivity to hepatotoxic drugs such as acetaminophen.
We thank Drs. R. Thier, R. F. Minchin, and M E. McManus at the University of Queensland and Dr. D. J. Jollow at the Medical University of South Carolina for valuable discussions.