Role of the Renal Cysteine Conjugate β-Lyase Pathway in Inhaled Compound A Nephrotoxicity in Rats 

Compound A (99.9% by gas chromatography) was synthesized by Central Glass Co. (Ube City, Japan) and provided by Abbott Laboratories (Abbott Park, IL). Aminooxyacetic acid (AOAA), probenecid, acivicin, and 1-aminobenzotriazole were purchased from Sigma Chemical Co. (St. Louis, MO), as were all other reagents unless specified.

All animal experiments were approved by the Huntington Life Sciences Animal Use Committee in accordance with American Association for Accreditation of Laboratory Animal Care guidelines. Male Fischer 344 rats (220 - 240 g; Charles River, Kingston, NY) were housed in individual stainless steel metabolic cages, provided with food and water ad libitum (except during exposure to compound A), maintained on a 12-h light/dark cycle, and acclimated for at least 1 week before experiments.

Animals (n = 10 per group) underwent 3-h exposure to compound A using a 40-l nose-only exposure apparatus operated at 10 l/min (15 air changes/h), permitting careful control of vapor composition and distribution, temperature, and concentrations of O²and CO², without the use of a CO²absorber. They were placed in individual polycarbonate cylinders with their snouts projecting into a central exposure chamber. Nitrogen containing vaporized compound A was mixed with 40% oxygen/60% nitrogen to achieve desired concentrations of compound A and delivered by side-port to the central chamber. Controls underwent similar treatment except that compound A was omitted. For initial dose - response experiments, target concentrations of compound A were 0, 50, 100, 150, and 200 ppm. For subsequent inhibitor experiments, the target concentration was 105 - 110 ppm. Exposure chamber temperature, humidity, pressure, and concentrations of compound A were determined every 30 min (by gas chromatography with on-line sampling). Additional methodologic details have been published previously. [12] 

For the inhibitor experiments, animals (n = 10 per group) were randomized to one of three groups: inhibitor (control), compound A, or inhibitor plus compound A. Control animals received either probenecid, AOAA, acivicin, or 1-aminobenzotriazole in saline by intraperitoneal injection before air-only exposure. Compound A animals received saline followed by exposure to compound A. Inhibitor plus compound A animals received either probenecid, AOAA, acivicin, or 1-aminobenzotriazole before exposure to compound A. Rats in both compound A groups were exposed together in the same chamber to ensure similar exposures. To assess the role of renal organic anion transport in the nephrotoxicity of compound A, animals were treated with probenecid (0.5 mmol/kg 30 min before exposure and again 10 h after the first injection), a competitive inhibitor of organic anion transport, [15]or saline. To assess the role of renal cysteine conjugate [Greek small letter beta]-lyase in the nephrotoxicity of compound A, animals received AOAA (0.5 mmol/kg 1 h before exposure and 0.25 mmol/kg 10 h after the first injection of AOAA), a competitive inhibitor of renal [Greek small letter beta]-lyase, [29]or saline. To assess the role of compound A - glutathione conjugates conversion to compound A - cysteine conjugates in compound A nephrotoxicity, animals received acivicin (10 mg/kg 1 h before exposure and 5 mg/kg 10 h after the first injection), a noncompetitive and very slowly reversible inhibitor of [Greek small letter gamma]-glutamyl transferase [30]or saline. To assess the role of cytochrome P450 in the nephrotoxicity of compound A, animals received 1-aminobenzotriazole (100 mg/kg, 40 mg/ml in saline adjusted to pH 7 with acetic acid, 2 h before exposure to compound A or air), an isoform-nonselective, irreversible inhibitor of hepatic and renal cytochrome P450, [31]or saline. Additional methodologic details have been presented elsewhere. [14] 

Urine was collected on ice for 24 h after exposure to compound A or air, the volume recorded, and the urine frozen. Animals were killed by CO (²) asphyxiation, and blood was obtained via the aorta. The left kidney was immediately excised, trimmed, cut in a midtransverse plane through cortex and medullary pyramid, and fixed in 10% neutral buffered formalin for embedding, sectioning, and hematoxylin-eosin staining. Blood was centrifuged and serum frozen. Serum urea nitrogen and creatinine, urine glucose, and urine total protein concentrations were measured spectrophotometrically using Sigma assay kits (nos. 66, 555, 115, and 611). Concentrations of [Greek small letter alpha] GST in urine were measured by enzyme immunoassay (Biotrin, Dublin, Ireland). Histologic analysis was performed by a pathologist blinded to the animal treatment, using a system for analysis and scoring of tubular necrosis described previously. [14]The severity of proximal tubular cell necrosis was expressed as the percentage of necrotic versus intact tubular profiles per field of view (10x ocular lenses, 20x objective lens).

All results are expressed as mean +/- SD. Biochemical results were compared by analysis of variance followed by Student-Neuman-Keuls post hoc tests. Data with unequal variances or which were not normally distributed were log transformed or analyzed by nonparametric tests. Necrosis was compared by the Kruskal-Wallis test because of zero variance in controls. Significance was assigned at P < 0.05.

Previous investigations of the nephrotoxicity of inhaled compound A used Wistar or Sprague-Dawley rats, [11–13]and our earlier experiments with intraperitoneally administered compound A used Fischer 344 rats. [14]Preliminary experiments, therefore, characterized the dose - response relationship for the nephrotoxicity of inhaled compound A in Fischer 344 rats. Target concentrations of compound A were 0, 50, 100, 150, and 200 ppm. Delivered concentrations of compound A were 0 +/- 0, 46 +/- 19, 98 +/- 7, 150 +/- 29, and 220 +/- 40 ppm. No tubular cell necrosis was observed in any rat receiving 0 or 46 ppm compound A (Figure 1). The threshold for necrosis was 98 ppm, characterized by 1.6 +/- 0.7% necrotic tubules and 10 of 10 rats affected, with greater necrosis at higher exposure concentrations. The necrosis was localized to proximal tubular cells, within the outer stripe of the outer medulla. The associated tubular lumina were often filled with sloughed cells, debris, and condensed proteinaceous substance, with plugging of most necrotic tubular lumina suggesting minimal if any effective flow. Biochemical evidence of nephrotoxicity included dose-related increases in urine volume and excretion of glucose, protein, and [Greek small letter alpha] GST (a cytosolic enzyme localized in proximal and possibly distal tubular cells, [32,33]used as a sensitive marker of renal tubular cell necrosis [14,33–35];Figure 2). The threshold for biochemical evidence of nephrotoxicity was 98 ppm, associated with increased urine volume and excretion of [Greek small letter alpha] GST. Significant proteinuria and glucosuria occurred at 150 and 220 ppm compound A. Serum creatinine concentrations were unchanged from control at all doses of compound A. Based on the histologic necrosis and biochemical markers, 105 - 110 ppm compound A was targeted in subsequent experiments to attempt 4 - 5% tubular necrosis.

Pretreatment with probenecid significantly diminished the nephrotoxicity of compound A, evidenced by both histology and some biochemical markers. The delivered concentration of compound A was 100 +/- 12 ppm. There was no renal tubular necrosis observed in any rat given probenecid before compound A (Table 1). Pretreatment with probenecid also partially prevented compound A - induced increases in urinary excretion of glucose and [Greek small letter alpha] GST (Figure 3). Serum creatinine and BUN concentrations in both compound A groups were not different from controls (Table 2).

Treatment with AOAA before exposure to compound A diminished compound A - induced increases in proteinuria and glucosuria (Figure 4). Protein and glucose excretion were significantly increased in compound A - treated rats but not those pretreated with AOAA. The delivered concentration of compound A was 109 +/- 29 ppm. Urinary excretion of [Greek small letter alpha] GST was significantly decreased in AOAA-pretreated animals receiving compound A compared with those receiving compound A alone, although the increase after compound A alone did not achieve statistical significance. Pretreatment with AOAA did not alter compound A - related renal tubular cell necrosis, which, however, was less than the 4 - 5% target (Table 1). Serum creatinine and BUN concentrations were unchanged in both groups receiving compound A (Table 2).

Treatment with acivicin preceding administration of compound A markedly exacerbated the nephrotoxicity of compound A, evidenced by both histologic and biochemical markers (Figure 5and Table 1). The delivered concentration of compound A was 104 +/- 9 ppm. Acivicin-pretreated rats had more extensive tubular lesions (5 - 19% necrotic tubules) and significant increases in urine volume and urinary glucose, protein, and [Greek small letter alpha] GST excretion compared with compound A alone. BUN and creatinine concentrations were increased slightly in acivicin/compound A - treated rats compared with compound A alone (Table 2).

Pretreatment with aminobenzotriazole elicited mixed results. Compound A - induced tubular necrosis was significantly reduced in extent by amino-benzotriazole (Table 1); however, the number of animals showing necrosis (9 of 10 vs 10 of 10) was unchanged. Pretreatment with aminobenzotriazole partially diminished glucosuria but not [Greek small letter alpha] GST excretion after administration of compound A (Figure 6). The delivered concentration of compound A in these experiments was 105 +/- 9 ppm. BUN concentration was increased but creatinine concentration decreased in compound A - treated rats.

The effects of inhaled compound A on renal tubular cells in Fischer 344 rats were qualitatively similar to those after intraperitoneal injection. [14,17]Histologically, the renal necrosis after both routes of administration was localized to proximal tubular cells, in the outer stripe of the outer medulla. Sloughing of necrotic cells and plugging of tubular lumina were common findings. Biochemically, both intraperitoneal and inhaled compound A caused diuresis, proteinuria, glucosuria, and increased urinary [Greek small letter alpha] GST excretion, without substantial changes in serum BUN and creatinine concentrations. Histologic and biochemical manifestations of renal tubular necrosis caused by inhaled compound A in Fischer 344 rats were also similar to those in both Wistar and Sprague-Dawley rats. [12,13,36] 

Renal tubular cell necrosis in Fischer 344 rats after inhalation of compound A was also quantitatively similar to that after intraperitoneal injection. [14]In dose-response experiments the threshold for the nephrototoxicity of inhaled compound A was 98 ppm (necrosis, diuresis, [Greek small letter alpha] GST excretion) or 150 ppm (proteinuria, glucosuria). Subsequent experiments at 100 - 109 ppm compound A also caused necrosis and biochemical alterations. Therefore, the nephrotoxicity threshold in Fischer 344 rats after 3 h of inhalation of compound A is 98 - 109 ppm (294 - 327 ppm-h). The tubular necrosis and biochemical alterations at this threshold dose quantitatively approximated those caused by the 0.2-mmol/kg intraperitoneal threshold dose found previously in Fischer 344 rats. [14] 

There is general agreement that the threshold for nephrotoxicity of inhaled compound A in rats is approximately 300 ppm-h. The 98 - 109 ppm threshold for 3-h inhalation exposure in male Fischer 344 rats (294 - 327 ppm-h, 0.4 - 3.4% necrosis, 45 of 49 animals) is consistent with the 114-ppm threshold in Sprague-Dawley rats (342 ppm-h, 10 of 15 animals affected)[12]and the 100-ppm threshold in Wistar rats (300 ppm-h, 8 - 10 of 10 animals affected, 1.0 - 2.5% necrosis). [11]It is also similar to the threshold for longer exposures (300 ppm-h both 6 and 12 h). [36]Neither the current nor a previous [12]investigation observed renal necrosis at <98 ppm compound A for 3 h of exposure (294 ppm-h), nor at <300 ppm-h for longer exposures. [36]The 3-h 50-ppm threshold (150 ppm-h) reported by Gonsowski et al. does not agree with these other results (possibly because the rats were substantially smaller - 124 - 159 g), and only 3 of 10 rats were purportedly affected. [11,36] 

The effects of probenecid, AOAA, and acivicin on the nephrotoxicity of compound A in Fischer 344 rats after inhalation were similar to those after intraperitoneal injection. [14,37]Probenecid completely prevented necrosis and diminished the increases in glucose and [Greek small letter alpha] GST excretion after both inhaled and intraperitoneally administered compound A. Aminooxyacetic acid lessened the increases in glucose, protein, and [Greek small letter alpha] GST excretion, and did not decrease necrosis after both routes of administration. Acivicin worsened necrosis, diuresis, glucosuria, proteinuria, and [Greek small letter alpha] GST excretion, whether compound A was injected or inhaled. These results suggest that the nephrotoxicity of compound A and the consequences of manipulating key transport and enzymatic processes of the renal [Greek small letter beta]-lyase pathway are similar whether compound A is inhaled (which mirrors clinical exposure) or injected intraperitoneally. Although Eger et al. [38]contend that intraperitoneal administration delivers compound A - glutathione conjugates but little compound A to the kidney (supposedly because of portal uptake and hepatic conjugation followed by pulmonary elimination of unchanged compound A), thus confounding our previous investigation, [14]no data support such speculation. Rather, intraperitoneally administered compound A may be absorbed into the systemic circulation via (1) the parietal peritoneum draining into the inferior vena cava, (2) abdominal peritoneal lymphatics draining into the thoracic duct (or diaphragmatic lymph drainage), or (3) the liver surface (with transhepatic absorption into the systemic circulation or hepatic lymph), and into the portal system via the visceral peritoneum. [39]Intraperitoneal administration of extensively metabolized volatile compounds clearly can result in substantial systemic bioavailability. [40,41]Therefore, the independence of compound A toxicity from the route of administration is similar to that found for numerous other haloalkenes, whose toxicity is mediated by glutathione and cysteine conjugate formation and [Greek small letter beta]-lyase-mediated metabolism. [24,26,27,42–47]Intraperitoneal injection is confirmed as a simple, cost-effective model for studying the nephrotoxicity of compound A as originally proposed. [14] 

Numerous haloalkenes cause nephrotoxicity in rats, owing to their metabolism and bioactivation by the cysteine conjugate [Greek small letter beta]-lyase pathway. [14–16]This term collectively represents a pathway involving the concerted participation of several organs and enzyme systems, which has been extensively described. Haloalkenes are initially metabolized to glutathione conjugates (primarily in the liver), which are excreted in bile and cleaved into their corresponding cysteine conjugates, which in turn may be N-acetylated to mercapturic acids. Glutathione conjugates also enter the systemic circulation, as do intestinal cysteine conjugates and mercapturates. Circulating S-conjugates are actively transported into proximal tubular cells by a probenecid-sensitive organic anion transporter. Proximal tubule cells also readily cleave glutathione- into cysteine-conjugates. Intracellularly, cysteine conjugates may undergo [Greek small letter beta]-lyase - mediated bioactivation to highly reactive intermediates, which are the presumed causes of proximal tubular cell necrosis. Both glutathione and cysteine conjugates, when administered to animals, may be nephrotoxic.

The current results are consistent with the hypothesis that renal uptake of compound A S-conjugates and cysteine conjugates metabolism by renal [Greek small letter beta]-lyase mediate, in part, the nephrotoxicity of inhaled compound A. The effects of probenecid suggest participation of the renal tubular cell organic anion transport system in the renal uptake of compound A S-conjugates. The effects of AOAA support the participation of renal [Greek small letter beta]-lyase in compound A - induced nephrotoxicity. Exacerbation of nephrotoxicity by acivicin, for which an explanation is presently not available, is consistent with some previous observations with compound A and hexachlorobutadiene. [14,48,49]A more detailed discussion of similar results, obtained with intraperitoneally administered compound A, has been presented elsewhere. [14]Iyer et al., however, recently reported that acivicin did partially block nephrotoxicity caused by compound A glutathione conjugates. [21] 

The effects of probenecid and AOAA on the nephrotoxicity of compound A after inhalation were less than those after intraperitoneal injection and more variable. [14]This may be explained by the greater variability, and perhaps by the lesser degree of necrosis, achieved after inhaled versus intraperitoneally administered compound A (0.4 - 3.4% vs. 5.5 - 8.1%). Although 4 - 5% tubular necrosis was targeted, this was not consistently achieved. A greater dose and degree of toxicity (such as 150 ppm compound A, with 16 +/- 9% necrosis) was not targeted in the inhibitor experiments for several reasons. First, inhaled compound A experiments aimed to replicate, as closely as possible, our previous investigation using intraperitoneal injection. [14]Therefore, deliberate attempts were made to replicate a threshold injury level as achieved previously (5% necrotic tubules). Second, variability in necrosis at higher (150 ppm) doses of compound A was even greater (6 - 38%) than at lower doses. Third, because concentrations of compound A found clinically (average 8 - 24 ppm and 20 - 32 ppm with soda line and barium hydroxide lime, respectively [3–10]) are less than causing nephrotoxicity in rats, our goal was to use the lowest possible supraclinical concentration that would achieve toxicity (100 - 110 ppm). Fourth, high doses of compound A can cause hepatotoxicity (potentially confounding the results) and nephrotoxicity; however, the former does not occur at the threshold for nephrotoxicity. [11,14]Fifth, at high doses of toxins, particularly parent haloalkenes compared with their S-conjugates, otherwise effective inhibitors may not prevent toxicity. For example, at a dose producing 50% tubular necrosis, 1,1-dichloroethylene nephrotoxicity was not ameliorated by probenecid; nonetheless, toxicity was attributed to S-conjugates and [Greek small letter beta]-lyase-dependent metabolism. [50]Similarly, AOAA completely prevented the nephrotoxicity of the [Greek small letter beta]-lyase-dependent toxin 1,1-dichloro-2,2-difluoroethylene at the threshold dose (150 [micro sign]mole/kg) but had no effect at 600 [micro sign]mole/kg. [46]This was also observed previously with intraperitoneally administered compound A, in which AOAA had lesser inhibitor effects on nephrotoxicity when the dose of compound A was increased from 0.2 to 0.25 mmol/kg. [14]Greater effects of AOAA on the nephrotoxicity of compound A than observed herein might provide more unequivocal evidence of involvement of renal [Greek small letter beta]-lyase. Nevertheless, determination of inhibitors effects over a full range of compound A doses was beyond the scope of the current work.

The current investigation also explored the role of cytochrome P450-catalyzed metabolism of compound A to potentially nephrotoxic inorganic fluoride and pentafluoroacetone. [22,51]Although compound A is selectively metabolized in human liver by P4502E1, [22]the cytochromes P450 that catalyze metabolism in rat liver and kidney are unknown. Inhaled anesthetic agents are metabolized less selectively (i.e., by multiple P450 isoforms) in rat compared with human liver, and this might also apply to compound A. [52,53]Further, the role of hepatic versus renal P450-dependent compound A metabolism in rats is unknown. Therefore, 1-aminobenzotriazole was used to nonselectively inhibit a broad spectrum of both hepatic and renal cytochromes P450, at a dose previously shown to destroy 80% of hepatic and renal P450. [31]Pretreatment with 1-aminobenzotriazole produced equivocal results. Tubular necrosis was reduced in extent but not frequency, and glucosuria but not [Greek small letter alpha] GST excretion was diminished. This does not offer strong evidence to support a substantial role for P450 in the toxicity of compound A in rats and is consistent with previous observations discounting a role for inorganic fluoride. [14]Extrapolation of these results to human exposure to compound A, which occurs in the presence of sevoflurane, is not recommended and further tempered by the finding that P4502E1 is a major P450 isoform in rat kidneys but is not found in human kidneys. [54,55] 

Involvement of the [Greek small letter beta]-lyase pathway in the toxicity of compound A, or any haloalkene, is supported by evidence of (1) formation of the required S-conjugates, (2) S-conjugates bioactivation by the essential enzymes of the pathway, (3) altered toxicity after manipulating the [Greek small letter beta]-lyase pathway, and (4) replication of toxicity by the S-conjugates. Considerable accumulated evidence indicates a role for the cysteine conjugate [Greek small letter beta]-lyase pathway in the metabolism and nephrotoxicity of compound A in rats. Formation in vivo of four compound A - glutathione conjugates (two alkane and two alkene conjugates), their cleavage to the corresponding cysteine conjugates, and N-acetylation to mercapturic acids has been shown. [17–19]Metabolism of compound A - cysteine conjugates, by rat [Greek small letter beta]-lyase in vitro and in vivo, to reactive intermediates has been demonstrated unambiguously. [19,20]Human renal [Greek small letter beta]-lyase also catalyzes the metabolism of compound A - cysteine conjugates. [20]Inhibition of renal transport by probenecid and inhibition of renal [Greek small letter beta]-lyase by AOAA diminished histologic or biochemical evidence of inhaled (Figure 3and Figure 4) and intraperitoneal [14,17,37]nephrotoxicity of compound A. Compound A - cysteine conjugates cause rat proximal tubular toxicity in vitro (E. D. Kharasch and R. A. Zager, unpublished observation, 1997).

Further, administration of alkene and alkane compound A - glutathione conjugates and the alkane compound A - cysteine conjugates to rats in vivo caused dose-dependent nephrotoxicity characterized by proximal tubular necrosis, glucosuria, and proteinuria, which replicated the nephrotoxicity caused by compound A. [21]In addition, acivicin and AOAA partially blocked the nephrotoxicity of compound A - glutathione conjugates. [21]Finally, the [Greek small letter alpha]-methylcysteine - compound A conjugate, an analogue of the compound A - cysteine conjugate that is not a substrate for renal [Greek small letter beta]-lyase, was not nephrotoxic. [21]Together, these observations suggest involvement of glutathione conjugate formation and S-conjugates metabolism by the [Greek small letter beta]-lyase pathway in the nephrotoxicity of compound A. In contrast, results interpreted to show that bioactivation by the [Greek small letter beta]-lyase pathway is not the mechanism of the nephrotoxicity of compound A have been presented, although no alternative mechanism was proposed. [49] 

Involvement of the renal cysteine conjugate [Greek small letter beta]-lyase pathway in the nephrotoxicity of compound A may confer interspecies differences in the effects of compound A. Renal [Greek small letter beta]-lyase activity and [Greek small letter beta]-lyase metabolism of compound A - cysteine conjugates in rats are approximately 8 - 30 times greater than those in human kidneys and higher than in nonhuman primates. [20,26,56]Other contributing interspecies differences in bioactivating or detoxifying [Greek small letter beta]-lyase pathway enzyme activities also may exist, [57]such as N-acetylation and deacetylation of compound A S-conjugates. [15] 

Whether compound A was administered by inhalation or intraperitoneal injection, the histologic and biochemical manifestations of nephrotoxicity in Fischer 344 rats were similar, and selectively inhibiting key components of the [Greek small letter beta]-lyase pathway elicited similar results. Results are consistent with the current hypothesis that renal uptake of compound A S-conjugates and cysteine conjugates metabolism by renal [Greek small letter beta]-lyase mediate, in part, the nephrotoxicity of compound A in rats. Route of administration of compound A does not appear to influence the manifestation or mechanism of the nephrotoxicity of compound A in Fischer 344 rats.