Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The bioactivated xenobiotic naphthalene (NA) is a widespread environmental contaminant with significant human exposures occurring both in the workplace and through lifestyle choices (i.e., cigarette smoking). Previous studies have established that parenteral administration of NA causes bronchiolar epithelial necrosis in mice and hamsters and cytotoxicity in the olfactory epithelium of rats, hamsters, and mice (Plopper et al., 1992). Nonciliated or "Clara" cells in bronchiolar epithelium of mice are particularly susceptible to NA injury (Plopper et al., 1992). Cytotoxic injury in mice, rats, and hamsters is associated with the presence of high rates of cytochrome P450 metabolism (Buckpitt et al., 1992). Cytochrome P-450 2F2 is particularly efficient in metabolizing NA to an epoxide (Shultz et al., 1999), and this protein appears to be responsible for the catalytic generation of the toxic NA metabolite. An ortholog (CYP2F1) with approximately 80% sequence homology has been identified in humans (Nhamburo et al., 1990) and human lung microsomes metabolize NA (Buckpitt and Bahnson, 1986). Although NA metabolism rates in microsomes from whole lung homogenates of humans appear lower than in mice, susceptibility of human airway epithelial cells to injury is unknown.

Despite the widespread prevalence of this compound, the effects of repeated exposures on the lungs are not well understood. Repeated systemic exposure to NA results in Clara cells, which are refractory, or tolerant, to further injury (O'Brien et al., 1989). A similar effect also exists with the bioactivated cytotoxicants 4-ipomeanol (Boyd et al., 1981) and coumarin (Born et al., 1999). Even at doses approaching the LD₅₀, little or no injury occurs in tolerant animals. The mechanism for this dramatic decrease in susceptibility is not well understood.

Several studies have speculated that tolerance results from a decreased capability for bioactivation caused by down-regulation of cytochrome P450 monooxygenases. For example, a recent study has demonstrated that hepatic tolerance to carbon tetrachloride (CCl₄) is due to decreased cytochrome P450 2E1 activity (Wong et al., 1998). Previous studies of NA-tolerant mice found that the rate of microsomal NA metabolism had decreased significantly. However when NA-tolerant mice were allowed to recuperate for 96 h after the last of seven NA injections, whole lung microsomal activity remained significantly depressed, whereas susceptibility to NA injury had returned (O'Brien et al., 1989). Additionally when the rates of NA activation were measured in isolated airways at saturating substrate concentrations, no significant differences were evident between control and tolerant mice at the primary target site of injury, the terminal bronchiole (Lakritz et al., 1996). Taken together, these studies indicate a mismatch between rates of NA bioactivation and shifts in susceptibility to injury and suggest that decreases in bioactivation may not be the critical determinant in the development of NA tolerance.

Conjugation with GSH is a major pathway for the detoxification of reactive metabolites of bioactivated cytotoxicants (Deneke and Fanberg, 1989). Mouse lung microsomes metabolize NA to GSH conjugates (Buckpitt et al., 1987). NA exposure depletes lung and Clara cell GSH in a dose/concentration-dependent fashion (Warren et al., 1982; West et al., 2000). Toxicity is augmented by pretreatment of mice with diethyl maleate, an agent that depletes cellular GSH (Warren et al., 1982). Because GSH appears to be a frontline defense for bioactivated cytotoxicants such as NA, it seems reasonable that increases in GSH synthesis may result in NA tolerance.

This study was designed to test the hypothesis that increases in airway GSH resynthesis play a critical role in NA-induced tolerance in Clara cells. To test this hypothesis, we asked the following questions: 1) Are alterations in steady state GSH coordinated with alterations in susceptibility to NA injury in three different cases; after one NA exposure (acute), after seven NA exposures (tolerant), and 96 h after the last of seven exposures (recuperation)? 2) Is the development of tolerance related to increases in the catalytic activity of the rate limiting step in GSH synthesis, -glutamylcysteine synthetase (-GCS)? and 3) Does blocking the resynthesis of GSH cause an increase in susceptibility to NA injury in tolerant mice?

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. NA was purchased from Fisher (Fairlawn, NJ). DL- Buthionine-[S,R]-sulfoximine (99.0% purity; BSO) was purchased from Sigma Chemical Co. (St. Louis, MO). Waymouth's medium was obtained from Life Technologies laboratories (Grand Island, NY) All fixatives and embedding reagents were purchased from Electron Microscopy Sciences (Fort Washington, PA) All other solvents were reagent grade or better.

Animals and Treatment. Male Swiss-Webster mice were purchased from Charles River Breeding Laboratories (Wilmington, MA). Animals were allowed free access to food and water and were housed in an AAALAC accredited facility in HEPA-filtered cage racks at the University of California, Davis, for at least 5 days before use in an experiment. There were four experimental protocols in which animals were administered seven repeated daily injections (i.p.) of NA (0 or 200 mg/kg) in corn oil (5 µl/g b.wt.). The goal of the first experiment was to determine whether increases in GSH corresponded with a loss of susceptibility to NA injury. NA was administered for 7 days; 24 h later, mice were either challenged with 300 mg/kg NA or sacrificed for GSH analysis. The goal of the second experiment was to determine whether the cessation of NA administration resulted in increased susceptibility and decreased target site GSH levels. NA was administered for 7 days, followed by a 96-h recuperation period, after which mice were challenged with 300 mg/kg NA or sacrificed for GSH analysis. The goal of the third experiment was to determine whether the development of tolerance is related to increases in the catalytic activity of the rate-limiting step in GSH synthesis, -GCS? Mice received NA for up to 7 days. Airways were isolated from lungs by microdissection for the preparation of supernatant from homogenates for analysis of -GCS activity. The goal of the fourth experiment was to determine whether blocking the resynthesis of GSH increases the susceptibility to NA injury in tolerant mice. Mice received NA for 7 days followed 24 h later by a dose of BSO (0 or 800 mg/kg) in combination with NA (0 or 200 mg/kg). Mice were sacrificed 3 h later and lungs were processed for high-resolution histopathology. Additional groups of mice were exposed to one dose of NA (200 or 300 mg/kg) without previous exposure to corn oil or NA, depending on the experiment. To reduce effects from diurnal GSH fluctuations, NA was administered each day between 8:00 and 11:00 AM. Animals were sacrificed with an overdose of pentobarbital sodium.

GSH Determination. To measure changes in GSH levels within target cell populations, specific airway sites were isolated by microdissection as previously described (Plopper et al., 1991). Samples of trachea, lobar bronchus, proximal bronchi, distal bronchi, and terminal bronchiolar airways were isolated from tolerant (n = 13), 96-h recuperated (n = 10), and control (n = 23) mice. Briefly, lungs were inflated with 1% agarose in Waymouth's media deficient of sulfur-containing amino acids. After microdissection, airway samples were placed in 200 mM methane sulfonic acid solution containing 5 mM diethylenetetrapentaacetic acid and snap frozen for later analysis. Defrosted samples were sonicated with a Heat systems ultrasonic processor and centrifuged (20 min at 16,000g) to remove the insoluble protein fraction.

-GCS Activity. Measurement of -GCS activity in microdissected airways was performed by a modification of methods described previously (Liu et al., 1998). Briefly, 200 mg/kg NA was administered to mice daily for up to 7 days. Twenty-four hours after each of the NA injections, mice (both treated and control) were sacrificed for measurement of airway -GCS enzyme activity. In a separate experiment, mice were administered NA for 7 days, then allowed to recuperate for 96 h before being sacrificed for analysis of -GCS enzyme activity. Distal (terminal to minor daughter) and proximal (major daughter to hilus) airways were isolated by microdissection after inflation via tracheal cannula with 1% agarose with Waymouth's medium deficient in sulfur-containing amino acids. Microdissected airways were placed in an ice-cold lysis buffer containing 0.1 M Tris (pH 8.2), 150 mM KCl, 20 mM MgCl₂, 2 mM EDTA, 2 µg/ml leupeptin, 2 µg/ml aprotinin, and 50 µg/ml phenylmethylsulfonyl fluoride. Airways were then homogenized in a small volume Polytron from Brinkman Instruments Co. (Westbury, NY). After homogenization, samples were centrifuged at 12,000g for 20 min. To remove glycine from the soluble fraction, the supernatant was placed in a Microcon-YM3 microconcentrator (Millipore Co., Milford, MA) and centrifuged for 30 min at 13,000g. The supernatant was washed and recentrifuged at the same speed with two additional 300-µl volumes of lysis buffer. Protein content was determined by the micro-Bradford method (Bradford, 1976) and samples were frozen at 80°C until analysis. Samples prepared fresh for enzyme analysis were run both immediately after isolation and after freezing to confirm that freezing samples did not affect enzyme activity.

BSO Treatment. To inhibit the resynthesis of GSH, mice were treated with BSO. Previous studies, in liver and kidney, have demonstrated that treatment with BSO causes an initial depletion of GSH over 6 h followed by a rebound due to the feedback inhibition of -GCS (Griffith and Meister, 1979). Several preliminary experiments were conducted to determine the optimal timing for the administration of the BSO and NA. At the target site of injury, the terminal bronchiole, BSO depleted GSH approximately 60% in tolerant mice. No adverse effects were observed when BSO was administered alone to mice. In the course of these experiments, several animals from both tolerant (n = 5) and control groups (n = 4) died when treated with the combination of 800 mg/kg BSO and 300 mg/kg NA. Accordingly, the challenge dose of NA was reduced (from 300 to 200 mg/kg). Mortality was still high in groups of corn oil-treated animals (n = 6) when administered 800 mg/kg BSO and 200 mg/kg NA 6 h after treatment. The following protocol was used for all mice reported here. One day after receiving the last of seven repeated NA injections (200 mg/kg), mice were given BSO (0 or 800 mg/kg) then challenged 1 h later with NA (0 or 200 mg/kg). Three hours after administration of the NA challenge dose, mice were sacrificed for histopathological assessment. At least three animals per group were used in this experiment, which was conducted twice.

High-Resolution Histopathology. All lungs were prepared for histopathological assessment by inflation via a tracheal cannula with 1% glutaraldehyde/1% paraformaldehyde in 0.1 M cacodylate buffer (335 mOsm) for 1 h at 30 cm H₂O pressure (Plopper, 1990). The entire fixed middle (cardiac) lobe was postfixed with osmium tetraoxide and incubated overnight in uranyl acetate. The postfixed tissue was embedded in Araldite-502 (Electron Microscopy Sciences), and embedded tissue was then grossly sectioned parallel to the long axis of the mainstem bronchi. Sections (0.5-µm) were cut with glass knives using a Zeiss Microm HM340E microtome and stained with 1% toluidine blue (Electron Microscopy Sciences). Representative images were selected for histopathological illustrations. Slides were imaged with a 330 CCD Dage camera on a Ziess Axiakop MC80 microscope using Scion 1.59 imaging software.

Statistics. All data are reported as the mean ± S.D. Differences between groups were determined by one-way ANOVA, and post hoc analysis was performed by the Bonferroni/Dunn significance test using Statview (SAS Institute, Cary, NC). The significance level is indicated in the figure legend.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

GSH and Susceptibility of Tolerant Mice. After seven daily injections of NA (200 mg/kg), steady-state GSH levels in the conducting airways, except the trachea, of tolerant mice were greater than in airways of corn oil-treated mice (Fig. 1). This was not the case in the liver, where GSH levels remained unchanged from control (control = 8.4 ± 3.1, tolerant = 10.6 ± 3.7 S.D.; GSH expressed as nanomoles of GSH per milligram of protein). The lobar bronchus, proximal bronchi, and distal bronchi of tolerant mice had significantly increased levels of GSH (160, 157, and 154%, respectively) compared with control. Changes in GSH status were most pronounced in the terminal airways, where GSH had increased 2.7-fold above that observed in corn oil-treated mice. The changes in airway compartment GSH levels correlated with the development of NA tolerance. Little or no injury was observed in airways of tolerant mice administered a NA challenge dose (300 mg/kg) (Fig. 2B). This was in contrast to control animals, where GSH levels were lower in terminal airways (Fig. 1) and where the NA challenge dose caused severe bronchiolar epithelial necrosis (Fig. 2A). The bronchiolar epithelium of mice in this group had extensive Clara cell swelling and necrosis, with most cells of the terminal airways exfoliating into the lumen in clusters. The injury pattern was not restricted to the terminal bronchiole and extended into adjacent, more proximal airways (data not shown).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   GSH levels in airways microdissected from mice treated with NA (Tolerant) or corn oil carrier (Control) for 7 days and from mice that have been allowed a 4-day recuperation period from NA tolerance (Recuperation). Values are expressed as nanomoles of GSH per milligram of protein, mean ± 1 S.D. *P < .05; **P < .01; ***P < ,001; represent significant differences between control (n = 23) and tolerant mice (n = 13). P < .01 represents significant difference between tolerant mice (n = 13) and those allowed a 4-day recuperation period (n = 10).

View larger version (50K): [in this window] [in a new window]   Fig. 2.   Histopathologic comparison of bronchioles of mice challenged with 300 mg/kg NA. A, mice received corn oil (× 7) followed by NA challenge. B, mice received 200 mg/kg NA (× 7) followed by NA challenge. C, mice allowed to recuperate for 4 days after repeated NA (200 mg/kg) (× 7) injections then challenged with NA. Mag Bar in large images represents 10 µm. Bracketed areas in inset outline site of high-magnification image.

GSH and Susceptibility of Recuperated Tolerant Mice. When mice received seven daily NA injections and then were allowed to recuperate for 96 h, GSH levels were lower than in tolerant mice without a recuperation period (Fig. 1). Airway GSH levels in the tolerant mice allowed a recuperation period were not significantly different from controls, but compared with tolerant mice were significantly lower at the site of greatest epithelial injury, the terminal airway. These changes also correlated with changes in the susceptibility to NA injury. When administered a challenge dose of NA (300 mg/kg), tolerant animals that had recuperated for 96 h after the last of seven NA injections also had severe bronchiolar epithelial necrosis (Fig. 2C). Mice having a recuperation period after the NA tolerance exposure protocol had injury limited to the most distal airways. In general, the injury appeared to be less severe than that of control animals challenged with the same dose of NA (300 mg/kg). Necrotic Clara cells appeared to be swollen and contained large cytoplasmic vacuoles, with a few cells exfoliating from the basement membrane.

Effect of Repeated NA Exposure on -GCS Activity. Cysteinylglycine, GSH, and -glutamylcysteine were separated (Fig. 3A) by a modification of a previously published procedure (Lakritz et al., 1997), which did not require derivatization of the sulfur-containing peptides. An analytical column with a higher carbon load and mobile phase with both decreased pH and acetonitrile content provided baseline separation. Detector responses were linear from 10 to 1000 ng (r² = 0.996; Fig. 3B). The analyte peak in the samples was easily identifiable. The linearity of the reaction was confirmed by running a time course up to 45 min and by measuring linearity at different protein concentrations. The reaction was linear until 30 min where, at high concentrations of protein, the product formation leveled off (data not shown). Diluting the protein concentration by 50% resulted in product formation of 50% of original values (Fig. 3C). The formation of the product was ATP-dependent; the absence of ATP completely abolished -glutamylcysteine formation (data not shown). Subsequently, all samples were run under conditions that maintained linearity with respect to time and protein concentration.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   HPLC profile, linear regression, and linearity with protein concentration of -GCS enzyme assay. A, separation of cysteinylglycine (cys-gly), GSH, and -glutamylcysteine (glu-cys) by HPLC. Separation conditions are as in Materials and Methods. B, linear regression of peak area (detector response) versus nanograms of -glutamylcysteine injected. Regression was linear (r2 = 0.996) between 10 and 1000 ng of the standard. C, effect of diluting protein samples on the formation of -glu-cys product. Dividing protein concentration by 50% resulted in 50% of the activity of the same samples.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Rates of -GCS activity in microdissected airways of mice treated with corn oil (control) or after one (24 h), two (48 h), three (72 h), four (96 h), or seven (7 day) daily injections of NA and after a 4-day recuperation period from NA tolerance (Recuperated). Values are expressed as nanomoles of -glutamylcysteine formed per milligram of protein per minute ± 1 S.D. ***P < .001 represents significant differences between control (n = 15) and tolerant (n = 5/timepoint). P < .001 represents significant differences between tolerant (n = 5/time point) and recuperated (n = 10).

Effect of BSO on Control Animals. The terminal airways of mice receiving repeated corn oil injections were lined by a simple cuboidal epithelium (Fig. 5A). The majority of the cells were nonciliated. Many of these nonciliated, or Clara cells, had apical projections into the airway lumen. Three hours after control mice were challenged with 200 mg/kg NA (without BSO exposure), Clara cells were swollen (Fig. 5E), formed large clear cytoplasmic vacuoles and had apical blebs that are segregated from the basal half of the cell. The injury pattern was similar in control mice treated with BSO before challenge with NA (Fig. 5F). Again, Clara cells in this group had swollen apical blebs and, additionally, had more cytoplasmic vacuoles. In this group, large clusters of degenerating cells had exfoliated and are underlain by a layer of squamated ciliated cells. When imaged by scanning electron microscopy, the exfoliating epithelium appeared as a continuous uniform sheet lifting off into the airway lumen (data not shown). No signs of toxicity were detected in control mice treated with BSO alone without NA (Fig. 5B).

View larger version (83K): [in this window] [in a new window]   Fig. 5.   Histopathologic comparison of terminal airway epithelium from mice receiving repeated injections of corn oil (CO), NA, and/or pretreated with BSO (0 or 800 mg/kg), and challenged with NA (0 or 200 mg/kg). A, CO × 7; B, CO × 7 + BSO; C, NA × 7; D, NA × 7 pretreated with BSO; E, CO × 7 + NA; F, CO × 7 + BSO + NA; G, NA × 7 + NA; H, NA × 7 + BSO + NA. Bar in (A) represents 10 um. **, apical blebs characteristic of injured Clara cells.

Effect of BSO on Tolerant Animals. Twenty-four hours after the last of seven daily injections of NA, mice were divided into four groups, administered BSO (0 or 800 mg/kg), then further subdivided and challenged with NA (0 or 200 mg/kg). The epithelium of mice receiving repeated daily injections of NA appeared similar to control with the overwhelming majority of terminal airways appearing as a simple cuboidal epithelium (Fig. 5C). Tolerant animals challenged with a 200 mg/kg NA dose (Fig. 5G) showed no detectable morphological differences from their unchallenged counterparts (Fig. 5C). In contrast, 3 h after administration of the NA challenge dose, tolerant animals treated with BSO had severe airway injury (Fig. 5H) similar to challenged nontolerant mice (Fig. 5E). Clara cells were exfoliated and had swollen apical membrane blebs, which appeared to bulge into the airway lumen. Overall, the airway epithelium from these mice was similar to NA-challenged nontolerant mice. No signs of toxicity were detected in tolerant mice treated with BSO alone without NA (Fig. 5D).

BSO Treatment: Comparison of Control and Tolerant. When airway epithelium of mice challenged with 200 mg/kg NA was compared between control (repeated corn oil injection) and tolerant (repeated NA injection), corn oil-treated mice exhibited significant injury (Fig. 5E), whereas tolerant mice had no signs of Clara cell cytotoxicity (Fig. 5G). However, comparison with the tolerant mice pretreated with BSO revealed differences. Tolerant mice pretreated with BSO, then challenged with NA (Fig. 5H), had injury that appeared similar in extent to that of corn oil control mice challenged with NA without BSO pretreatment (Fig. 5E). Comparing the groups challenged with NA revealed that both have Clara cells blebbing with a few cells exfoliating from the basement membrane. In summary, when mice were challenged with NA, the severity of injury appeared to be greatest in control mice pretreated with BSO (Fig. 5F), less severe, but substantial, in both nonpretreated control (Fig. 5E) and BSO-pretreated tolerant mice (Fig. 5H). Finally, Clara cell cytotoxicity was not apparent in non-BSO-pretreated tolerant mice challenged with 200 mg/kg NA (Fig. 5G).

View larger version (84K): [in this window] [in a new window]   Fig. 6.   Hyperplastic/dysplastic nodules. High-resolution light histopathology of mouse terminal airway receiving repeated NA injections. Bar in large image represents 10 µm. Arrow within the inset indicates location of high magnification image.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The objective of this study was to establish whether development of tolerance to bioactivated xenobiotics, such as NA, is dependent on increases in the levels of GSH or in increased ability to generate this tripeptide in target sites for injury. Four experimental approaches were used to test this hypothesis. First, measurement of steady-state GSH levels in specific sites of both target and nontarget airways showed that repeated exposure to NA increases GSH at target sites that were resistant to acute NA injury. Second, target sites in tolerant mice that had been allowed to recuperate for 96 h were again susceptible to NA injury, and this corresponded with significantly lower GSH levels. Third, the activity of -GCS, the rate-limiting enzyme in GSH biosynthesis is elevated during the development of tolerance. Fourth, treatment with BSO, a specific inhibitor of -GCS, eliminates the ability of Clara cells of tolerant mice to resist cytotoxic injury by NA.

For the xenobiotics examined to date, the development of tolerance to environmental toxicants coordinates closely with shifts in GSH status within lung compartments that develop resistance to injury. Significant regional shifts in the antioxidant potential of the lung also are present in the development of tolerance to ozone (Plopper et al., 1994), including significant shifts in GSH in distal airways of rats and rhesus monkeys after long-term exposures (90 days and 20 months) (Duan et al., 1996). Similar responses occur in rats after repeated sublethal exposure to 85% oxygen, where nonprotein sulfhydryl levels increased to approximately 200% above control (Kimball et al., 1976). Our study demonstrates similar changes in airways that become tolerant to NA injury. GSH levels of terminal bronchioles, the most susceptible site in nontolerant animals, were 270% that of control in tolerant mice (Fig. 1). In addition, after seven repeated injections of NA, the biosynthetic capability for GSH resynthesis increased to greater than 225% that of control (Fig. 4). We noted that differences in GSH level were not always coordinated with differences in -GCS activity. In Fig. 1 we measured the content of the actual tripeptide, which under in vivo conditions is subject to feedback inhibition, differences in local cysteine metabolism, as well as the cycle of synthesis, conjugation, transport, and degradation. Figure 4 is focused on the maximal activity of -GCS, which when assayed as a semipurified protein, was not limited by any of the above mentioned factors. Additionally, it is important to point out that the distal airway preparations used in the -GCS assay contained up to eight terminal airways, whereas measurement of GSH was from a single defined region of the airways. More protein was required for the measurement of the -GCS activity than for the measurement of GSH.

Regardless of sampling technique, these results indicate that reduction in the metabolic activation of NA is not fully responsible for the development of tolerance. Although tolerant animals have decreased levels of cytochrome P450-mediated NA activation in whole lung microsomal preparations, changes in the rates of NA metabolism do not correlate in all cases with sensitivity (O'Brien et al., 1989). Tolerant mice are resensitized to NA injury when they are allowed to recuperate for up to 96 h, yet P450 activity is depressed to 60% of control (O'Brien et al., 1989). The current study shows that this mismatch between susceptibility to cytotoxicants and bioactivation potential can be explained by differences in the detoxification potential between mice that are tolerant and those that have had a 96-h recuperation period.

Finally, using compounds that block intracellular GSH synthesis, we were able to confirm that shifts in the detoxification potential played a key role in the development of tolerance. Several studies have demonstrated the importance of GSH in cellular protection from lung cytotoxicants. Normal mice and rats pretreated with diethyl maleate, a GSH depletor, are more susceptible to injury by a wide range of bioactivated xenobiotics including 4-ipomeanol, -napthylthiourea, dichloroethylene, and NA (Boyd and Neal, 1976; Boyd et al., 1981; Warren et al., 1982; Okine et al., 1985). Using BSO, we reached the same conclusion regarding the susceptibility in normal mice. A previous study indicates that the development of tolerance to ozone is altered by concurrent administration of BSO (Sun et al., 1988) with the production of pulmonary fibrosis. Our experiments demonstrate that BSO pretreatment not only augments NA toxicity in previously untreated control mice, but more importantly for this study, BSO pretreatment eliminates the resistance to injury in tolerant mice.

Based on the three experimental approaches used in our studies, we conclude that the mechanism of NA tolerance is partly dependent on increases in steady-state intracellular GSH pool. Our results, taken together with the previous studies of NA tolerance (O'Brien et al., 1989; Lakritz et al., 1996), indicate that tolerance results from a shift in the balance between bioactivation and detoxification. Our studies demonstrate that when mice were allowed a 96-h recuperation period after seven NA administrations, susceptibility to NA did not return to the same level as control. When challenged with 300 mg/kg NA, control mice exhibited injury that began in the distal airways and extended into the proximal airways, whereas, in contrast, injury was limited to the distal airways in mice that had recuperated for 96 h from tolerance. Although bronchiolar epithelial necrosis was observed in tolerant mice having recuperated for 96 h, it did not appear as severe as the injury in control animals. The finding that both GSH levels and the activity of -GCS are similar in controls and in tolerant mice allowed to recuperate for 96 h suggests that differences in susceptibility to challenge may be related to the depressed levels of NA activation noted in these mice (O'Brien et al., 1989).

This study raises several fundamental questions concerning the development of tolerance to bioactivated toxicants. First, is the development of tolerance to pulmonary cytotoxicants a lung-specific mechanism or does the liver play a significant role by clearing the parent compound through first pass metabolism? In our study of tolerant animals, neither liver nor blood GSH levels were increased significantly, nor were there any signs of hepatotoxicity in tolerant mice challenged with NA. If the liver of tolerant animals was responsible for the clearance of additional amounts of the challenge dose of NA, we argue that the liver would be susceptible to the additional quantities of the obligate intermediate, NA oxide, because hepatic GSH levels are the same in tolerant animals. These data suggest that tolerance is not related to changes in the hepatic clearance of the parent compound. This view is further supported by ongoing studies showing that tolerant mice are resistant to injury by repeated inhalation exposures to NA. Second, does the mechanism we have demonstrated for NA tolerance represent a general defense mechanism for bioactivated cytotoxicants? Recent studies of tolerance demonstrating that repeated administration of coumarin reduced the susceptibility to acute NA injury in mice (Born et al., 1999) support this hypothesis. These results combined with the knowledge that GSH is critical in the defense of Clara cells to single acute doses of 4-ipomeanol, dichloroethylene, -napthylthiourea, as well as NA, strengthen the conclusion that the development of resistance to bioactivated xenobiotics is critically dependent on elevation of cellular thiol status. Further studies will be required to determine at what point Clara cells become resistant to injury after repeated exposures to bioactivated cytotoxicants, and if the resistant cells retain normal Clara cell phenotypic characteristics while only shifting the level of detoxification enzymes such as -GCS. Our initial data suggest that Clara cells become tolerant without major phenotypic changes.

Based on the data from this study, we conclude that the development of resistance to acute injury from bioactivated lung cytotoxicants, such as NA, is critically dependent on increases in steady-state GSH pools and that the mechanism for increasing these pools includes an up-regulation in the catalytic activity of -GCS, the rate-limiting step of GSH biosynthesis.