A novel lung slice system with compromised antioxidant defenses.

In order to facilitate the study of oxidative stress in lung tissue, rat lung slices with impaired antioxidant defenses were prepared and used. Incubation of lung slices with the antineoplastic agent 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) (100 microM) in an amino acid-rich medium for 45 min produced a near-maximal (approximately 85%), irreversible inhibition of glutathione reductase, accompanied by only a modest (approximately 15%) decrease in pulmonary nonprotein sulfhydryls (NPSH) and no alteration in intracellular ATP, NADP+, and NADPH levels. The amounts of NADP(H), ATP, and NPSH were stable over a 4-hr incubation period following the removal from BCNU. The viability of the system was further evaluated by measuring the rate of evolution of 14CO2 from D-[14C(U)]-glucose. The rates of evolution were almost identical in the compromised system when compared with control slices over a 4-hr time period. By using slices with compromised oxidative defenses, preliminary results have been obtained with paraquat, nitrofurantoin, and 2,3-dimethoxy-1,4-naphthoquinone.


Introduction
An important type of lung injury is produced by toxicants that induce an oxidative stress. Such a mechanism has been implicated in the pulmonary toxicity of the herbicide paraquat and the antimicrobial agent nitrofurantoin (1). Both compounds undergo a one-electron reduction catalyzed by microsomal NADPH-cytochrome P-450 reductase to form radical species. In the presence of molecular oxygen, the radical is oxidized with the generation of superoxide anion and reforms the parent molecule that can be rereduced forming a futile redox cycle. Compounds that redox cycle are characterized by a disproportionately greater consumption of oxygen, as compared with the concentration of the compound present (2,3). Active oxygen species so formed may induce toxicity by a number of different mechanisms including DNA-strand breakage, lipid peroxidation, and enzyme inactivation (4).
Despite extensive investigation, the mechanism of paraquat toxicity is not established, but the two major hypotheses propose an involvement of lipid peroxidation resulting from the production of active oxygen species and the depletion of the pyridine nucleotide, NADPH (5-?). These two mechanisms need not be mutually exclusive (8).
In order to combat the potential toxicity of active oxygen species, all cells possess antioxidant defenses (2).
An important component of these is the glutathione reductase-glutathione peroxidase system. Glutathione reductase maintains intracellular glutathione, predominantly t 95%) in the reduced form (GSH), by reducing oxidized glutathione (GSSG), formed by the peroxidasecatalyzed decomposition of hydrogen peroxide at the expense of NADPH (9).
The role of thiol status in paraquat toxicity is unclear. No thiol depletion is seen in rats dosed with paraquat (10), and sulfhydryl pretreatments do not ameliorate lethality (11). However, exogenous GSH protects alveolar Type II cells against loss of viability (12). Nitrofurantoin has been shown to cause a reduction in GSH levels in the isolated, perfused rabbit lung (13).
BCNU has been employed in a variety of studies in isolated hepatocytes for the study of oxidative stress (16,1?) and BCNU pretreatment potentiates hyperoxia-induced pulmonary fibrosis in mice in vivo (18). However, the maximal inhibition of glutathione reductase obtainable in vivo (-60%) is restricted by the dose-limiting toxicity of the compound. The aim of the present study was to establish a model system with compromised oxidative defenses in order to study oxidative stress in the lung. Specifically, we wished to develop lung slices with a relatively specific inhibition of glutathione reductase.

Materials and Methods
All chemicals, unless otherwise stated, were obtained from Sigma Chemical Co. (London, UK) and were of the highest grade purity available. BCNU was obtained from Bristol-Myers Pharmaceuticals (Uxbridge, UK).
Paraquat dichloride was a generous gift of ICI plc (Macclesfield, Cheshire, UK). 2,3-Dimethoxy-1,4naphthoquinone was prepared as previously described (19). MEM amino acids were obtained from Gibco Ltd. Male Wistar rats (160-250 g) were obtained from Olac Ltd. (UK). Lungs were perfused with Krebs Ringer phosphate (KRP) buffer with glucose, pH 7.4, and slices were prepared as previously described (20). Compromised slices were incubated with BCNU (100 AM) in KRPenriched with amino acids. The amino acid-rich medium was prepared using MEM amino acid solution and appropriate additions of L-serine, L-cysteine, L-glycine, L-glutamine, and L-methionine (final concentration 0.2 mM), according to the specifications of Seglen (21). Slices were incubated for 45 min in a shaking water bath at 37°C. Slices were washed with KRP to remove BCNU and amino acids and used for biochemical measurements and incubations. Control slices were treated similarly except BCNU was omitted, and the slices were incubated in KRP in the absence of amino acids, as normal thiol levels were maintained in these slices under these conditions. Lung slices were homogenized using a Polytron homogenizer for 30 sec. Samples for assay of NPSH and ATP were homogenized in 6.5% TCA/5 mM EDTA, whereas those for glutathione reductase were homogenized in potassium phosphate buffer, pH 7.4. Samples for NADPH assay were prepared in 0.1 M sodium hydroxide/0.1 M nicotinamide. For NADP+, slices were homogenized in 0.5 M perchloric acid.
Glutathione reductase was measured using the spectrophotometric method of Carlberg and Mannervik (22). Nonprotein sulfhydryls were assayed using the method of Hissin and Hilf (23). This fluorometric assay is not specific for GSH. However, since GSH constitutes approximately 85% of lung thiols, NPSH measurements reflect GSH levels. ATP was assayed using a bioluminescence technique as previously described (24). Viability was evaluated using D-[14C(U)]-glucose (1 ,uCi/center-well flask) (20). Samples were counted on a LKB-1216 Rackbeta liquid scintillation counter. Pyridine nucleotides were assayed using an established spectrophotometric method (25).

BCNU and Pulmonary Glutathione Reductase
Isolation of lung slices in an amino acid-rich medium with BCNU (100 PzM), as described in "Materials and Methods," resulted in a marked inhibition of glutathione reductase activity (80-85%), which remained constant over the 4-hr incubation period (Fig. 1).

BCNU and Other Biochemical Indices
In order to assess the effects of BCNU treatment on other biochemical markers related to oxidative stress and cell viability, NPSH, ATP, NADP+, NADPH, and glucose oxidation were measured. A small but significant difference (p < 0.05), which was maintained throughout the 4-hr incubation, in NPSH levels, was observed following incubation with BCNU (Fig. 2a). Little or no differences were observed in the levels of ATP (Fg 2a) and NADP(H) (Fig. 2b) following incubation with BCNU. Treatment of lung slices with BCNU had no effect on D-['4C(U)]-glucose oxidation, in contrast to slices exposed to the mitochondrial inhibitor rotenone (Fig. 3).

Control and Compromised Slices
Paraquat (100 and 500 IzM) did not cause depletion of NPSH in control slices (Fig. 4). However, some depletion was seen in compromised lung slices (Fig. 4).   Discussion BCNU has been employed in studies with isolated hepatocytes in order to study oxidative stress. Such pretreatment potentiates the toxicities of a number of compounds, including menadione, adriamycin, and diquat (17,26) that are proposed to be toxic by a mechanism involving oxidative stress. We wished to develop a lung slice model with compromised oxidative defenses, and therefore selected BCNU in an analogous manner to the earlier studies with hepatocytes.
Lung slices, incubated with BCNU, possessed a severely inhibited glutathione reductase (Fig, 1). Preincubation of slices with BCNU (100 1M) and amino acids for 45 min resulted in 80 to 85% inactivation, compared with glutathione reductase activities in control slices. This degree of inhibition is similar to that obtained with isolated hepatocytes (16) and greater than that obtained in lung tissue in vivo with the maximum nonlethal dose of BCNU (-60%) (27).
In our studies, inhibition of glutathione reductase with BCNU was time-and concentration-dependent (data not shown). Glutathione reductase activity observed in control slices was in good agreement with literature values (27). Inhibition of greater than 85% was achieved with either higher concentrations of BCNU or longer incubation times (results not shown), but these conditions resulted in a more marked loss of NPSH levels, compared to the small decrease observed (Fig. 2a). Thus, we have described a simple, rapid (preparation time -60 min) method for obtaining rat lung slices with a markedly inhibited glutathione reductase activity accompanied by a minimal decrease in NPSH, normal ATP (Fig. 2a), normal NADP(H) levels (Fig. 2b), and unchanged glucose oxidation (Fig. 3). The small NPSH depletion observed is unlikely to have a significant effect on the biochemical functions of the slices.
Preliminary results with compromised slices showed an apparent paraquat-induced NPSH depletion in contrast to control slices (Fg 4). Similar results were observed in hepatocytes with the structurally related herbicide, diquat (28), and a role for GSH (and glutathione reductase) was implicated in diquat toxicity. These results suggested the possibility of a similar role for glutathione and reductase in response to paraquat in our studies. However, the magnitude of the NPSH change is small and its relevance to toxicity is still unclear.
The potentiation of thiol depletion with nitrofurantoin and 2,3-dimethoxy-1,4-naphthoquinone was observed in compromised slices (data not shown), and the significance of these effects is under investigation. In summary, the development of a novel lung slice model with compromised oxidative defenses is described. The method has the advantages of speed and simplicity, and cell-to-cell interactions are maintained. We believe that the use of this model should increase our understanding of oxidative stress in the lung Simon Hardwick is in receipt of an SERC CASE Studentship. We thank M. Fagg for typing the manuscript and D. King for preparing the figures.