Antioxidant Defense Mechanisms in Cultured Pleural Mesothelial Cells V. L. Kinnula, J. I. Everitt, J. B. Mangum, L.-Y. Chang, and J. D. Crapo Duke University Medical Center, Durham, North Carolina; Department of Pulmonary Medicine, University of Helsinki, Helsinki, Finland; and Chemical Industry Institute of Toxicology, Research Triangle Park, North Carolina

The role of different antioxidant pathways in cultured rat pleural mesothelial cells was studied by exposing the cells to various hydrogen peroxide (H202) concentrations and by measuring H202 cell cytotoxicity and the capacity of the cells to scavenge H202 • The antioxidant enzymes, glutathione peroxidase, glutathione reductase, glucose-e-phosphate dehydrogenase, and catalase were analyzed biochemically. Catalase and CuZn superoxide dismutase were localized by immunocytochemistry. To enable investigation of the glutathione redox cycle and catalase pathways, glutathione reductase was inactivated with 1,3-bis(2chloroethyl)-l-nitrosourea (BCNU) and catalase was inactivated with aminotriazole. When the cells were exposed to a low, sublethal (0.030 mM) H202 concentration, glutathione reductase but not catalase inactivation resulted in a decreased capacity to remove H202 from the extracellular medium. When the cells were exposed to a high (0.25 mM) H202 concentration, H20z-scavenging capacity decreased remarkably when catalase was inactivated. When the cells were exposed to 0.1 to 0.5 mM H202 , cell cytotoxicity (lactate dehydrogenase release) increased significantly if glutathione reductase was inactivated; catalase inactivation resulted in a significant cytotoxicity only at high (~ 0.25 mM) H202 concentrations. Immunocytochemical studies showed that the cells, both in situ and in vitro, contained low amounts of catalase. This suggests that the results of the catalase-inhibition studies are probably not due to a change in the characteristics of the cells in culture. 3-Aminobenzamide is a compound that is known to prevent NAD depletion through inhibition of poly(ADP-ribose) polymerase during oxidant stress. When intact cells were treated with different antioxidants and exposed to 0.5 mM H20z, both catalase and 3-aminobenzamide protected the cells completely. These findings suggest that the glutathione redox cycle is a major source of protection of mesothelial cells against low levels of oxidant stress and that catalase becomes more significant in protecting these cells against severe oxidant stress.

The role played by reactive oxygen species in various pleural diseases is still uncertain. Several studies have suggested that reactive oxygen species may playa central role in the inflammation, cellular injury, and/or carcinogenic responses that result after asbestos exposure (1, 2). Lung inflammation is associated with recruitment of macrophages and neutrophils (3), which can be stimulated to produce high amounts of reactive oxygen species by numerous inflammatory mediators in vitro (4). Similar events are presumed to occur in the pleura, since in vitro asbestosexposed mesothelial cells have been shown to produce (Received in original form August 30, 1991 and in final form January 6, 1992) Address correspondence to: James D. Crapo, M. D., Duke University Medical Center, P.O. Box 3177, Durham, NC 27710. Abbreviations: aminotriazole, ATZ; 1,3-bis(2-chloroethyl)-nitrosourea, BCNU; buthionine sulfoximine, BSO; CuZn superoxide dismutase, CuZn SOD; 5,5'-dithiobis(2-nitrobenzoic acid), urNB; fetal bovine serum, FBS; Hanks' balanced salt solution, HBSS; hydrogen peroxide, H 20 2; lactate dehydrogenase, LDH. Am. J. Respir. CeU Mol. Bioi. \bl. 7. pp. 95-103, 1992

chemotactic factors for neutrophils (5) and asbestos exposure has been observed to result in an inflammatory reaction on the surface of the mesothelium in vivo (6). Several in vitro studies have shown that hydrogen peroxide (H20 2) is one of the more important reactive oxygen species produced by phagocytic cells, and that cytotoxicity of phagocytes can be mediated by H202 in endothelial cells (7), epithelial cells (8), fibroblasts (9), and certain tumor cells (10). For this reason, H20z was selected as an agent for studying oxidant injury in mesothelial cells in this study. The antioxidant characteristics of mesothelial cells are unknown, although the mesothelium is thought to be an important target site of oxidant stress under conditions such as fiber-induced inflammatory disease. Numerous earlier studies have focused on the antioxidant pathways in parenchymal lung cells and have observed that the importance of antioxidant pathways differs profoundly according to lung cell type. Endothelial cells are sensitive to injury by oxidant stress in vivo (11). Cultured endothelial cells have been shown to be protected against oxidant stress mainly through the glutathione redox cycle (12). Alveolar epithelial type II cells, which are more oxidant resistant than type I epithelial cells

96

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992

or endothelial cells, contain relatively high antioxidant enzyme levels including having a high catalase activity (13). Presently, it is not known how sensitive mesothelial cells are to oxidant injury, or which antioxidant pathways play a critical role in scavenging reactive oxygen species to protect these cells against oxidant stress. The resistance against oxidant stress can be related either to oxidant characteristics of the cells themselves, such as reactive oxygen species generation or to antioxidant pathways. Currently, it is not known whether mesothelial cells release reactive oxygen species into the extracellular space, although this would have important implications for the pathogenesis of most pleural diseases. The present study was designed to address these questions using cultured rat pleural mesothelial cells.

Materials and Methods Mesothelial Cell Isolation and Culture Mesothelial cells were isolated from male Fischer 344 rats as described previously (14). Isolated cells were suspended in complete growth medium (FC 10) consisting of Ham's F-12 medium supplemented with 10% heat-inactivated fetal bovine serum (FBS) (GIBCO, Grand Island, NY) and containing 100 U/ml penicillin, 100 I-'g/ml streptomycin, and 0.25 I-'g/ml fungizone (GIBCO). Cells were cultured using F-lO medium as described previously (14), and cell passage numbers between 12 and 15 were used. For the biochemical experiments, the culture medium was washed 3 times with Hepes-buffered Hanks' balanced salt solution, pH 7.4 (HBSS). Ultrastructurally, mesothelial cells cultured as described above were found to have tight junctions, desmosomes, pinocytotic vacuoles, and microvilli similar to those previously described in mesothelial cells, and they expressed vimentin as well as acidic and basic cytokeratins using both Northern blot and immunocytochemical procedures (14). Inactivation of Antioxidant Enzymes and Glutathione Depletion Washed cells in HBSS were treated with 30 mM aminotriazole (ATZ) for 60 min to inactivate catalase (15), or with 100 I-'g/ml 1,3-bis(2-chloroethyl)-I-nitrosourea (BCND) for 15 min to inactivate glutathione reductase (16). In order to deplete glutathione, gamma glutamyl cysteine synthetase was inactivated with 0.1 to 1 mM DL-buthionine-S,R-sulfoximine (BSO) for 18 h (17). Culture medium containing 10% FBS was changed to 1% FBS for the 18-h incubation period. Earlier studies have shown that ATZ is specific for catalase inactivation (15). BCNU specifically inactivates glutathione reductase (12, 16, 17). Although BCNU has been shown to cross-link DNA and lead to cell death when used in higher doses, neither BCNU, BSO, or ATZ at the doses used in these experiments was cytotoxic to our cultured mesothelial cells as measured by following lactate dehydrogenase (LDH) release. The doses were chosen for maximal inhibition of glutathione reductase, gamma glutamyl cysteine synthetase, or catalase. Assay of Antioxidant Enzymes To conduct the antioxidant enzyme measurements, cultured mesothelial cells were scraped into cold HBSS, centrifuged, suspended in a small volume of HBSS, and sonicated. Cata-

lase activity was determined polarographically by following oxygen production using a Clark electrode fitted into a stirred chamber. The reaction was started by adding the cells into the buffer containing 0.5 mM HzOz equilibrated with nitrogen. The activity was expressed as nmol O. produced/min/mg protein. Glutathione peroxidase activity was assayed by following the NADPH oxidation rate in the presence of t-butylhydroperoxide, glutathione, and glutathione reductase (18), and glutathione reductase activity was assayed by following the oxidation of NADPH in the presence of oxidized glutathione (18). Glucose-6-phosphate dehydrogenase was quantitated spectrophotometrically (19) by following the reduction of NADP. Immunocytochemical Localization of Catalase In order to compare mesothelial cells in culture with mesothelial cells in situ, passaged mesothelial cells and those on the parietal pleura lining the chest wall of rats were used. Cultured cells and portions of parietal pleura were fixed using 2 % paraformaldehyde plus 0.2 % glutaraldehyde and then transferred to 2 % paraformaldehyde. Cells were fixed while in culture flasks and processed as described previously (20). Parietal pleura was fixed by injecting the initial fixative into the pleural cavity. After 1 h, the chest wall was cut out and transferred to the second fixative. Pieces of the chest wall containing the parietal pleura and cell pellets were embedded in LR White resin according to the methods of Newman and Jasani (21). A plastic embedding media was needed to facilitate the orientation of chest wall specimens in a way that sections of the block would show mesothelial cells lining the parietal pleura. In addition, a large number of mesothelial cells can be evaluated from plastic sections. Ultrathin sections of cultured mesothelial cell pellets and of the parietal pleura as well as sections of alveolar type II cells taken from lung inflation fixed using the same fixatives were double labeled with antisera against bovine catalase and rat CuZn superoxide dismutase (CuZn SOD) according to Slot and Geuze (22). For greater sensitivity of immunolabeling of catalase, cryoultrathin sections of the pellet of cultured mesothelial cells were also prepared and labeled with rabbit antibovine catalase antiserum. Removal of H20 2 from the Extracellular Medium Mesothelial cells in HBSS were exposed to 0.030 mM H20 Z and incubated for 30 min at 37° C. HzOz was assayed from aliquots drawn during the incubation period. In part of the experiments, antioxidant enzymes were inactivated by prior incubation with specific inhibitors. The inhibitors included ATZ, BCNU, BSO, and ATZ + BCND. These cells were then washed 3 times and exposed to H20 2 • The HzOz concentration was assayed spectroflurometrically using 3-methoxy-4 hydroxy phenylacetic acid and horseradish peroxidase by a method modified from Ruch and co-workers (23). HZ0 2 concentrations were determined by spectrophotometry using its extinction coefficient 40 at 240 nm. Cell Injury Protocols In preliminary experiments, cells were exposed to 0.030 to 1 mM H 20 Z and LDH release during a 3-h period was followed. Based on these experiments, control or pretreated cells were exposed either to 0.030, 0.1, 0.25, or 0.5 mM

Kinnula, Everitt, Mangum et al.: Antioxidant Defense Mechanisms in Cultured Pleural Mesothelial Cells

H20 2 • Washed cells in HBSS containing one of the abovementioned H20 2 concentrations were incubated at 37° C for different time points and assayed for LDH release. Experiments were conducted using control cells and cells that had been pretreated with ATZ, BCND, BSO, or ATZ + BCND as described above. Control cells that were not treated with H20 2 were incubated with ATZ, BCND, BSO, or ATZ + BCND. After the preincubation, cells were washed 3 times before exposure to H20z• Based on preliminary glutathione assays, 0.1 mM BSO was used. Cytotoxicity was expressed as percent LDH release. Cell Protection with Exogenous Antioxidants To investigate the role of different antioxidants, mesothelial cells were pretreated for 5 min with 2 mM 3-aminobenzamide, 2 mM glutathione, 500 D catalase, 2 mM deferoxamine, and 2 mM allopurinoL These cells in HBSS containing the above-mentioned antioxidant concentrations were exposed to 0.5 mM H 20 2 for 3 h, and LDH release from these cells was assayed. Extracellular H z02 Release Extracellular H20 2 release from the cells was assayed by incubating washed cells in HBSS at 37° C for 30 to 120 min. H20 2 from the extracellular medium was assayed as described above. Other Biochemical Analyses Glutathione was assayed from control and BSO-treated cells using continuous reduction of 5,5'-dithiobis(2-nitrobenzoic acid) (abbreviated DTNB) (24). LDH from the extracellular medium and scraped cells was measured spectrophotometrically by following NADH oxidation (25). Proteins were assayed by the method of Lowry and associates (26). Statistical Analyses All data are expressed as mean ± SEM. Two groups were compared using a two-tailed Student's t test, and multiple groups were compared using variance analysis and Scheffe's post hoc test. P value < 0.05 was considered to be significant.

Results Antioxidant Enzyme Activities and Glutathione Concentration in Intact and Pretreated Cells In cultured pleural mesothelial cells, catalase activity was 10.1 ± 2.4 (n = 8) nmol O, produced/min/mg protein, glutathione reductase 6.7 ± 1.4 (n = 4) mD/mg protein, glutathione peroxidase 2.6 ± 0.2 (n = 3) mD/mg protein, and glucose-6-phosphate dehydrogenase 4.6 ± 1.8 (n = 4) mUlmg protein. Catalase activity in ATZ-treated cells was 12% compared with the control cells; it was not changed with BCND or BSo. Glutathione reductase activity decreased 90 % with BCND and was not affected with ATZ or BSo. Intracellular glutathione concentration was 54 ± 5.5 nmol/mg protein (n = 3); it decreased 85 % with 0.1 mM BSO and 95% with 1 mM BSO (Figure 1). In further experiments, 0.1 mM BSO was used. H 20 2-induced Cytotoxicity in Intact Cells When the cells were exposed for 3 h to different exogenous H20 2 concentrations, 0.5 but not 0.1 mM H20 2 was toxic using LDH release as an assay of cytotoxicity (Figure 2). In further experiments that focused on the importance of different antioxidant pathways during sublethal and lethal oxidant stress, 0.03, 0.1, 0.25, and 0.5 mM H20z concentrations were used. Removal of H 20z from the Extracellular Medium Cells were first exposed to 0.030 mM H20 2 , which concentration is sublethal to the cells at least when short incubation times are used. Cells scavenged H 20 2 in a nonlinear fashion during the 30-min course of the exposure (Figure 3, control curves). If the cells were pretreated with ATZ to inactivate catalase or with BSO to deplete intracellular glutathione, H20 2 scavenging did not differ from controls (Figures 3A and 3B). When glutathione reductase was inactivated with BCND, H2 0 2 scavenging was decreased significantly (Figure 3C). The difference compared with control cells became significant after 10 min, possibly related to intracellular glutathione reserves, which could not be restored in glutathione reductase-inactivated cells. Because catalase inhibition had

100

100

z -- 1\

t

80

....

60

0

::E:

80

.,41

60

lIII

~

::E:

40

0 0.0

1mMH202

•

CP

t/)

20

97

~.

0.2

••

•

c ....

40

•

20

O.1mMH202

0 0.4

0.6

0.8

1.0

850, mM

Figure 1. Effect of BSO treatment on glutathione depletion in mesothelial cells. Cells were pretreated for 18 h with 0.1, 0.5, or 1 mM BSo. All BSO concentrations decreased intracellular glutathione concentration significantly compared with the control value. * P < 0.05. Mean ± SEM, n = 3.

1

2 Time, hours

3

Figure 2. Effect of exogenous H202 on LDH release from mesothelial cells. Valuesare means ± SEM from three to eight separate experiments that were conducted in duplicate. After 2 and 3 h, 0.5 and I mM H202 resulted in a significant LDH release. * P < 0.05.

98

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 71992

A

100 80

~ z,

-

CD

60

~

0'" 40

'"

:I:

20 .r--o;;;::~

ATZ

__ Control

0 0

B

10

20 Time, Min

30

40

100 80

~

-

CD ~

60

0'" 40

:I:'"

20

r--o;;;::~

Control

0 0

C

__ BSO

10

20 Time, Min

30

40

100 80

!

-

CD

60

~

0'" 40

0----...10

:I:'"

BCNU

20 Control

0 0

10

20

30

40

Time, Min

Figure 3. Detoxification of a low exogenous concentration (0.030 mM) of H20 2 by control (open circles) and pretreated (closed circles) mesothelial cells. (A) Cells pretreated with 30 mM aminotriazole (ATZ) for 60 min to inactivate catalase. (B) Cells pretreated with 0.1 mM BSO for 18 h to deplete intracellular glutathione. (C) Cells pretreated with 100 ltg/mI BCND for 15 min to inactivate glutathione reductase. Figure 3C also shows the results (open squares) for cells pretreated with ATZ (30 mM, 60 min) plus BCND (100 ltg/mI, 15 min). Pretreated cells were washed 3 times before being exposed to H20 2 • H20 2 was assayed from aliquots taken from the incubation medium. Valuesare means ± SEM from three to four separate experiments conducted in duplicate. * P < 0.05 compared with control.

still left when both catalase and glutathione reductase were inactivated using ATZ + BCND. The remaining H2 0 2-scavenging capacity may be due to the small residual activities of the antioxidants. In addition, a portion of this remaining H20 2-scavenging capacity may represent conversion of H 20 2 to other reactive metabolites such as OR· by reactions not involving catalase or peroxidase, e.g., by Fenton reactions. Cells were next exposed to a high cytotoxic H20 2 concentration (0.25 mM), and the same experiments were repeated. In these experiments, catalase inactivation with ATZ resulted in a remarkable decrease in the cells' capacity to scavenge R 20 2 from the extracellular medium. This change appeared to be even more remarkable than the effect in glutathione reductase-inactivated cells (Figure 4). H20 2-scavenging measurements were also conducted using BSO-treated cells (85 % depletion of intracellular glutathione). Surprisingly, H20 2 detoxification by BSO-pretreated cells did not differ significantly from controls during the 90-min incubation time (data not shown). Possible reasons for the absence of an effect using BSO include (1) an intact glutathione redox cycle, (2) having a minimal but adequate glutathione concentration left, and (3) the difference between the intracellular and extracellular kinetics of H20 2 metabolism under these experimental conditions. Cell Injury Experiments To investigate the protective role of different antioxidant pathwaysduring oxidant stress, the cells were exposed to 0.1, 0.25, and 0.5 mM H20 2 and cytotoxicity was investigated by following LDH release. In these experiments, intact cells or cells that had been treated with ATZ, BCND, or ATZ + BCND were used. When 0.1 mM H20 2 was used, catalase inactivation with ATZ did not increase cytotoxicity. However, when glutathione reductase was inactivated with BCND, cytotoxicity was enhanced significantly (Figure 5). LDH release was still more enhanced in cells that were treated with ATZ + BCND compared with cells treated with BCND alone (Figure 5). When 0.5 mM H20 2 was used, inhibition of either catalase or glutathione reductase increased cytotoxicity remarkably (Figure 6). When both catalase and glutathione reductase were inactivated, LDH release after

100

~

i C'I

80 60 40

A12

0

C'I

::c no effect on scavenging H 20 2 , the cells were next exposed to BCND and to ATZ + BCND using the same (0.03 mM) H20 2 concentration. These experiments showed that when both catalase and glutathione reductase were inactivated at the same time, scavenging of extracellular H2 0 2 was decreased significantly (P < 0.05) compared with controls and slightly but not significantly (P < 0.1, n = 3) compared with glutathione reductase inhibition alone (Figure 3C). A remarkable part of the cellular H 20 2-scavenging capacity was

20

BCNU Control

20

40

60

80

100 120

Time, min Figure 4. Detoxification of a high exogenous concentration (0.250 mM) of H20Z by control and pretreated cells. ATZ and BCND pretreatments were similar to the experiments shown in Figure 3. HzOz was assayed from aliquots taken from the incubation medium. Values are means from two separate experiments each conducted in duplicate.

Kinnula, Everitt, Mangum et al.: Antioxidant Defense Mechanisms in Cultured Pleural Mesothelial Cells

100

100 ATZ

+

BCNU

+ H2

0

2

0

* BCNU

--

80

+ H2

0Z

Q)

60

Q)

C/)

+

HzOz

*

60

CI:l

Q)

Q)

Q)

::I: 0

+ H2 02

AlZ

0~

CI:l

....

BCNU

80

--

0~

en

99

(j)

....

40

::I: 0

40

...J

...J

20

20 ATZ

+

Ha Oz

0.1 mM HzOz Control

o

1 234 Time, hours

Figure 5. Effect of 0.1 mM H20 2 on LDH release from control cells and cells pretreated with ATZ, BCND, or ATZ + BCND. Washed cells were exposed to 0.1 mM H20 2 , and cytotoxicity was assayed as LDH release during a 4-h incubation. Values are means ± SEM from four separate experiments that were each conducted in duplicate. All treatmentgroups were studied separately with independent H20 2 and control groups. The H20 2 and control groups did not differ and were therefore combined for this figure. All statistical comparisons were done using the matched and not the pooled H20 2 and control groups. * P < 0.05 compared with both control and H202-treated cells. After 2, 3, and 4 h of incubation, ATZ + BCND-treated cells differed significantly from BCNDtreated cells and BCND-treated cells differed significantly from ATZ-treated cells.

2 h of incubation was greater than it was after the inactivation of either catalase or glutathione reductase alone (Figure 6). The different H20 2 dose response on cytotoxicity using ATZ-treated cells (Figures 5 and 6) suggests that catalase has an obvious protective effect during severe oxidant stress (0.5 mM H20 2) but that during mild oxidant stress (0.1 mM H20 2) either catalase is not important or the small catalase activity remaining after ATZ treatment is still sufficient. Glutathione Depletion and Cell Cytotoxicity When 85 % of glutathione had been depleted with BSO, 0.5 mM H20 2 resulted in a significant cytotoxicity in 3 h compared with cells treated only with H 20 2 (Figure 7A). However, 0.1 mM H20 2 resulted only in a 25 % LD H release from BSO-treated cells (Figure 7B), whereas it resulted in a 76% LDH release from BCNU-treated cells in corresponding experimental conditions (see Figure 6). These results suggest the importance of the glutathione redox cycle in comparison to glutathione depletion alone in protecting mesothelial cells during oxidant exposure. Immunocytochemical Studies EM immunocytochemical studies were performed to further assess the role of catalase in mesothelial cells. Cultured

Control

o

1

2

3

Time, hours Figure 6. Effect of 0.5 mM H20 2 on LDH release from control cells and cells pretreated with ATZ, BCND, or ATZ + BCND. Values are means ± SEM from four separate experiments each done in duplicate. All treatment groups were studied separately with independent H20 2 and control groups. The H20 2 and control groups did not differ and were thereforecombined for this figure. All statistical comparisons were done using the matched and not the pooled H202 and control groups. * P < 0.05 compared with both control and H20 2-treated cells. At 2 h, ATZ + BCND-treated cells had significantly greater LDH release than cells treated with either ATZ or BCND.

mesothelial cells exhibited very low catalase activity in comparison to alveolar type II epithelial cells. Because it is possible that catalase is decreased during cell culture, the concentration of catalase in mesothelial cells both in situ and in culture was qualitatively compared using immunocytochemistry. In contrast to alveolar type II cells (Figure SA), both in situ (Figure 8B) and cultured mesothelial cells (Figure 8C) showed scarcely any immunolabeling for catalase. Only one labeled peroxisome was found in 40 parietal mesothelial cell profiles examined. No labeled peroxisome was observed from 50 profiles of cultured mesothelial cells. Nonimmune serum, used as a control, showed no specific labeling and virtually no background labeling. These results suggest that biochemical studies showing low catalase in cultured mesothelial cells are not describing artifacts of cell culture. Cells were also immunolabeled for CuZn SOD. Parietal mesothelial cells exhibited a labeling density of this enzyme that was similar to alveolar type II cells. Cultured mesothelial cells, on the other hand, had a reduced cellular content of CuZn SOD (Figure 8). Cryoultrathin sections were also used to evaluate the presence of catalase in cultured mesothelial cells because this technique gives a higher labeling efficiency. Catalase labeling was found over peroxisomes on cryosections of cultured mesothelial cells (Figure 9), although labeled peroxisomes were rarely found in these cells. The results from immunocytochemical studies are in agreement

100

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992

A 100

z

80

.,

60

CD

BSO+G.5mMH202

III

.!!

CD ~

:J:

c

...J

40 20 0 0

1

Discussion

2 3 Time, hours

B 100

l

80

., CD

60

'ii

40

III CD ~

BSO+H202

H202

:J: C

...J

20 Control

0 0.0

0.2

H20 2 Release H202 release from mesothelial cells was determined as the amount of H202 released to the extracellular medium during a 30- to 150-min incubation period. Mesothelial cells released H202 into the extracellular medium at a constant rate (0.014 nmol/min/mg protein) for at least 150 min. H20 2 release could not be stimulated by a l-h incubation with 200 nM phorbol myristate acetate. Phorbol myristate acetate is known to stimulate H20 2 release from phagocytic cells by activating protein kinase C.

0.4 mMH202

Figure 7. The effect of BSO pretreatment on LDH release during oxidant stress. (A) Cells pretreated for 18 h with 0.1 mM BSO and exposed for 1 to 3 h to 0.5 mM H202 • (B) Cells pretreated for 18 h with 0.1 mM BSO and then exposed for 3 h to 0.1, 0.25, or 0.5 mM H202 • Values are means ± SEM from three separate experiments each conducted in duplicate (A) or from two separate experiments each in duplicate (B). * P < 0.05 compared with H202 •

with our biochemical studies which show a relatively low, but detectable, catalase activity in these cells. These observations indicate that antioxidant enzymes may decrease when mesothelial cells are cultured, although catalase content is low both in situ and in culture. Protection by Exogenously Added Antioxidants In the present study, 3-aminobenzamide (2 mM), catalase (500 U), glutathione (2 mM), deferoxamine (2 mM), and allopurinol (1 mM) were used. Aminobenzamide inactivates poly(ADP-ribose) polymerase and prevents NAD depletion (19). Previous studies have shown that one of the early sequence of events in cell injury during oxidant stress involves DNA strand breaks with subsequent activation of poly(ADPribose) polymerase and consumption of NAD (19, 20). As expected, in the present study extracellular catalase prevented cytotoxicity caused by exogenously added H20 2 • Cytotoxicity could be completely prevented also by 3-aminobenzamide but not with other antioxidants (Figure 10). These results suggest the importance of poly(ADP-ribose) polymerase inhibition in preventing mesothelial cell injury during oxidant exposure.

Cultured pleural mesothelial cells were found to be relatively resistant to exogenously produced oxidant stress. Using LDH release as a measure of cell cytotoxicity, these results showed that 0.1 mM H20 2 was not toxic during a 4-h incubation period. It has to be emphasized that LDH release is not sensitive for assessment of minimal cell injury in shortterm experiments. Spragg has shown that endothelial cells undergo rapid induction of DNA strand breaks after exposure to 50 ~M H20 2 ; DNA strand breaks were apparent within 30 min of exposure, whereas cell lysis or death measured by LDH or 51Cr release only became apparent 2 h after exposure to > 0.1 mM H20 2 (27, 28). Based on LDH release, the present results suggest that the resistance of endothelial cells (27, 28) and mesothelial cells against in vitro induced oxidant stress is similar. The specific activities of antioxidant enzymes do not correlate with the relative importance of different antioxidant pathways. When cultured cells are used, enzyme activities must be interpreted with caution because culture conditions and cell passage number may influence the cellular characteristics. This occurs with epithelial type II cells during the first days in culture when catalase, glutathione peroxidase, and glutathione reductase decrease significantly at the same time these cells become less differentiated (29,30). To assess whether or not antioxidant enzymes were being changed by the cell isolation and culture process, catalase and CuZnSOD were localized by immunocytochemistry in situ. Catalase labeling in mesothelial cells was low compared with epithelial type II cells, and labeling density did not change during cell culture. Low catalase labeling and low catalase activity in mesothelial cells suggest that pathways other than catalase may be critical in removing toxic, reactive oxygen species during oxidant stress by these cells. Because mesothelial cells contain low catalase levels, it is likely that the glutathione redox cycle plays a critical role in detoxifying reactive oxygen species. The importance of the glutathione redox cycle as a potent antioxidant defense system has been previously established in a variety of tumor cells (10) and in endothelial cells (12). However, different cell types appear to differ profoundly with regard to the importance of antioxidant pathways, which also suggests that these differences may be independent of cell culture techniques. Furthermore, earlier studies have shown that not only cell type but also the degree of oxidant stress may influence which pathway plays the major antioxidant role (31). In this study, inhibition of intracellular catalase did not decrease the H20 rscavenging capacity in low, sublethal H 20 2 concentrations, whereas glutathione reductase inhibi-

Kinnula, Everitt, Mangum et al.: Antioxidant Defense Mechanisms in Cultured Pleural Mesothelial Cells

Figure 8. Immunolabeling of both CuZn superoxide dismutase (CuZn SOD) and catalase on LR White ultrathin sections of (A) alveolar epithelial type II cells, (B) mesothelial cells of the parietal pleural, and (C) cultured mesothelial cells. CuZn SOD was labeled with rabbit anti-rat CuZn SOD antisera and 15-nm gold particles (arrowheads). Catalase was labeled with rabbit anti-bovine catalase and 6-nm gold particles (arrows). Ultrathin sections of all the specimens were processed at the same time with identical procedures. CuZn SOD was present in the cytosol and nucleus (Nu) of all three types of cells studied. The labeling densities of CuZn SOD in alveolar type II cells and parietal mesothelial cells are similar. Cultured mesothelial cells have a lower labeling density of CuZn SOD. Catalase was found in peroxisomes (p) in alveolar type II cells. Labeled peroxisomes were not found in the two types of mesothelial cell specimens. m = mitochondria; lb = lamellar bodies; ga = Golgi apparatus. Bar = 0.5 /-lm.

Figure 9. Immunolabeling of catalase on a cryoultrathin section of a cultured mesothelial celL Catalase was labeled with rabbit anti-bovine catalase antisera and 9-nm gold particles. Peroxisomes (p) containing catalase, although very rare, can be found in the cultured mesothelial cells (arrow). m = mitochondria. Bar = 0.5 Jlm.

101

102

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 7 1992

70

---s.

r--------~---------____,

60 50

Q)

~ 40

Q)

CD 30 0:

620 ...J

10

o

L-l_--'--

(jO~\(O\ or.,,-o"- e~1-~~\o~~\~\~se \~\~\O~e oi-~~\~e 'AO'O

...~\,.

Go\\)

Q

e\e(

~,i\~o\

~\O

~.r

Figure 10. The effect of different antioxidants on mesothelial cell protection against an oxidant stress caused by exposure to 0.5 mM HzOz for 3 h. Results shown in solid bars were all following exposure to HzO z. The antioxidants used were: 500 U catalase, 2 mM 3-aminobenzamide, 2 mM glutathione, 2 mM deferoxamine, and 1 mM allopurinol. The antioxidants were added 5 min before the cells were exposed to HzO z. Values are means ± SEM from four separate experiments each done in duplicate. * P < 0.05 compared with HzO z. Glutathione, deferoxamine, and allopurinol also differed significantly from the controls.

tion resulted in a significant decrease in HzOz removal. A concentration of 0.1 mM HzOz was toxic to these cells only if glutathione reductase was inactivated. However, when higher (0.25 to 0.5 mM) levels of HzOz were used, catalase inactivation resulted in significant cytotoxicity. The superiority of the glutathione redox cycle in scavenging HzOz under low HzOz concentrations and an increasing importance of catalase at higher HzOz concentrations has been reported previously (12, 31). Possible reasons for this phenomenon include the relatively high K; of catalase for HzOz (32) or that the ultimate cellular damage may be caused not by HzOz only, but by a secondary hydroperoxide product that can be metabolized by the glutathione redox cycle (33). The results from the present study suggest that mesothelial cells are protected against low levels of exogenous oxidant stress through the glutathione redox cycle but that both glutathione reductase and catalase may have significant roles during severe oxidant exposure. Previous studies with endothelial cells have shown that they can be protected against oxidant stress by 3-aminobenzamide, which prevents an early event in oxidant stressNAD depletion by inhibiting poly(ADP-ribose) polymerase (27,34). The present study suggests that this also applies to mesothelial cells. 3-Aminobenzamide protected these cells completely, while glutathione provided only partial protection. Glutathione may not be freely diffusible through the plasma membrane (35), which may explain the partial effect obtained with this compound. Deferoxamine, an iron chelator, protects endothelial cells against oxidant stress by interfering with the Fenton reaction (28). In the present study, cells could be partially protected with this compound. The protective effect of deferoxamine may be critical not only in

protecting cells against HzOz-induced oxidant stress but also during asbestos exposure. Asbestos is associated with increased lipid peroxidation in vitro, which can be prevented by deferoxamine (36). However, it has to be pointed out that the protective effect of deferoxamine in vitro may be influenced by minimal amounts of Fe2+ in the tissue culture media or incubation solutions. Although the exact mechanism of mesothelial cell lysis during oxidant stress is unclear, the above results suggest that NAD depletion, an early event in oxidant stress, makes these cells vulnerable to the irreversible oxidant injury. Intact cells can release reactive oxygen species as a defense mechanism against bacteria and foreign particles. This mechanism of reactive oxygen species generation may be associated with cell resistance. Mesothelial cells released HzOz into the extracellular medium at the very low rate of 0.014 nmol/min/mg protein. Unstimulated freshly isolated alveolar macrophages release HZ02 at a rate of 3.5 nmol/ min/mg protein (23), and this can be stimulated several-fold by a number of inflammatory agents and protein kinase C activators (10). Freshly isolated and cultured epithelial type II cells release H20 2 at a rate of 0.7 nmol/min/mg and cultured endothelial cells at a rate of 0.06 nmol/min/mg protein (23). It is possible that cell culture may change some oxidant/antioxidant characteristics of mesothelial cells and that the cell response may differ significantly in vitro and in vivo. Although extrapolation of these results to the situation in vivo must be done with caution, these results suggest that release of reactive oxygen species into the extracellular spaces is not a significant factor in mesothelial cells. In summary, the present study found that pleural mesothelial cells scavenge low exogenous H20 2 concentrations mainly by the glutathione redox pathway, which also protects these cells against mild oxidant stress. However, both the glutathione redox cycle and catalase appeared to scavenge extracellularly added H20 2 and protect these cells when high, toxic H20 2 concentrations were used. Intact mesothelial cells could be partly protected against extracellularly added H20 2 by exogenous antioxidants, but only catalase and 3-aminobenzamide, which prevents NAD depletion, protected these cells completely in short-term in vitro experimental conditions. Acknowledgments: This study was supported by Program Project Grant POI HL31992 and Grant ROl HL42609 from the National Institutes of Health and by a grant to the Chemical Industry Institute of Toxicology from the Thermal Insulation Manufacturers Association (TIMA). Dr. Kinnula was partly supported by the Anti-Tuberculosis Association of Finland.

References 1. Mossman, B. T., J. P. Marsh, and M. A. Shatos. 1986. Alteration of superoxide dismutase (SOD) activity in tracheal epithelial cells by asbestos and inhibition of cytotoxicity by antioxidants. IAb. Invest. 54:204-212. 2. Mossman, B. T., J. P. Marsh, A. Sesko et al. 1990. Inhibition of lung injury, inflammation and interstitial pulmonary fibrosis by polyethylene glycol-conjugated catalase in a rapid inhalation model of asbestosis. Am. Rev. Respir. Dis. /41:1266-1271. 3. Barry, B. E., and J. D. Crapo. 1984. Patterns of accumulation of platelets and neutrophils in rat lungs during exposure to 100% and 85% oxygen. Am. Rev. Respir. Dis. 132:549-555. 4. Ward, P. A., R. E. Duque, M. C. Sulavik, and K. J. Johnson. 1983. In vitro and in vivo stimulation of rat neutrophils and alveolar macrophages by immune complexes. Am. J. Pathol. 110:297-309. 5. Antony, V. B., C. L. Owen, and K. J. Hadley. 1989. Pleural mesothelial cells stimulated by asbestos release chemotactic activity for neutrophils in vitro. Am. Rev. Respir. Dis. 139:199-206.

Kinnula, Everitt, Mangum et al.: Antioxidant Defense Mechanisms in Cultured Pleural Mesothelial Cells

6. Goodglick, 1.. A., and A. B. Kane. 1990. Cytotoxicity of long and short crocidolite asbestos fibers in vitro and in vivo. Cancer Res. 50:51535163. 7. Weiss, S. J., J. Young, F. LoBuglio, A. Slivka, and N. F. Nimeh. 1981. Role of hydrogen peroxide in neutrophil-mediated destruction of cultured endothelial cells. J. Clin. Invest. 68:714-721. 8. Suttorp, N., and 1.. M. Simon. 1982. Lung cell oxidant injury: enhancement of polymorphonuclear leucocyte-mediated cytotoxicity in lung cells exposed to sustained in vitro hyperoxia. J. Clin. Invest. 70:342-350. 9. Simon, R. H., C. H. Scoggin, and D. Patterson. 1981. Hydrogen peroxide causes the fatal injury to human fibroblasts exposed to oxygen radicals. J. Bioi. Chem. 256:7181-7186.

10. Nathan, C. F., B. A. Arrick, H. W. Murray, N. M. DeSantis, and Z. A. Cohn. 1980. Tumor cell antioxidant defenses. Inhibition of the glutathione redox cycle enhances macrophage-mediated cytolysis. J. Exp.

Med. 153:766-782. II. Crapo, J. D., B. E. Barry, H. A. Foscue , andR. J. Shelburne. 1980. Structural and biochemical changes in rat lungs occurring during exposures to lethal and adaptive doses of oxygen. Am. Rev. Respir. Dis. 122: 123-143. 12. Suttorp, N., W. Toephfer, and 1.. Roka. 1986. Antioxidant defense mechanisms of endothelial cells: glutathione redox cycle versus catalase. Am. J. Physiol. 25 I (Cell Physiol.):C671-C680. 13. Simon, R., 1. A. Edwards, M. M. Reza, and R. G. Kunkel. 1991. Injury of rat pulmonary alveolar epithelial cells by H,O,: dependence of phenotype and catalase. Am. J. Physiol. 26O(Lung Cell Mol. Physiol. 4): 1.318-1.325. 14. Bermudez, E., J. Everitt, and C. Walker. 1990. Expression of growth factor and growth factor receptor RNA in rat pleural mesothelial cells in culture. Exp. Cell Res. 190:91-98. 15. Margoliash, E., A. Novogrodsky, and A. Schejter. 1960. Irreversible reaction of 3-arnino 1:2,4-triazole and related inhibitors with the protein of catalase. Biochem. J. 74:339-348. 16. Babson, J. R., and D. J. Reed. 1978. Inactivation of glutathione reductase by 2-chloroethyl nitrosourea-derived isocyanates. Biochem. Biophys.

Res. Commun. 83:754-762. 17. Griffith, 0. W., and A. Meister. 1979. Potent and specific inhibition of 18. 19. 20. 21.

glutathione synthesis by buthionine S-R-butylhomocysteine sulfoximine. J. Bioi. Chem. 254:7558-7560. Beutler, E. 1975. Glutathione peroxidase. In Red Cell Metabolism: A Manual of Biochemical Methods. Grune & Stratton, New York. 71-73. Lohr, G. W., and H. D. Waller. 1974. Assay of glucose-fi-phosphate dehydrogenase.In Methods of Enzymatic Analysis. H. U. Bergmeyer, editor. Academic Press, New York. 636-641. Kinnula, V.I.., A. R. Whorton, L.-Y. Chang, and J. D. Crapo. 1992. Regulation of hydrogen peroxide generation in cultured endothelial cells. Am. J. Respir. Cell Mol. Bioi. 6:175-182. Newman, G. R., and B. Jasani. 1984. Post-embedding immunoenzyme techniques. In Immunolabeling for Electron Microscopy. J. M. Polak and 1. M. Varndell, editors. Elsevier Science Publishers, New York. 53-70.

103

22. Slot, J. W., and H. J. Geuze. 1984. Gold markers for single and double immunolabeling of ultrathin cryosections. In Immunolabeling for Electron Microscopy. J. M. Polak and 1. M. Varndell, editors. Elsevier Science Publishers, New York. 129-142. 23. Ruch, W., P. H. Cooper, and M. Baggiolini. 1983. Assay of H,O, production by macrophages and neutrophils with homovanillic acid and horseradish peroxidase. J. Immunol. Methods 63:347-357. 24. Theodorus, P., M. Akerboom, and H. Sies, 1981. Assay of glutathione, glutathione disulfide and glutathione mixed sulfides in biological samples.

Methods Enzymol. 77:373-381. 25. Bergmeyer, H. U. 1974. Lactate dehydrogenase assay with pyruvate and NADH.In Methods of Enzymatic Analysis. Vol. 2. H. U. Bergmeyer, editor. Academic Press, Orlando, FL. 574-579. 26. Lowry, 0. H., N. R. Rosebrough, A. 1.. Farr, and R. J. Randall. 1951. Protein measurement with the folin phenol reagent. J. Bioi. Chem. 193:265-275. 27. Spragg, R. G. 1991. DNA strand break formation following exposure of bovine pulmonary artery and aortic endothelial cells to reactive oxygen products. Am. J. Respir. Cell Mol. Bioi. 4:4-10. 28. Varani, J., S. H. Phan, U. S. Ryan, and P. A. Ward. 1990. H,O, mediated cytotoxicity of rat pulmonary endothelial cells. Changes in adenosine triphosphate and purine products and effects of protective interventions. Lab. Invest. 90:683-689.

29. Panus, P. C., S. Matalon, and B. Freeman. 1989. Responses of type II pneumocyte antioxidant enzymes to normoxic and hyperoxic culture. In

Vitro Cell Dev. Bioi. 1:221-229. 30. Kinnula, V.I.., 1.. Y. Chang, J. 1. Everitt, and J, D. Crapo. 1992. Oxidants and antioxidants in alveolar epithelial type II cells: in situ, freshly isolated and cultured cells. Am. J. Physiol. 262(Lung Cell Mol. Physiol. 6):1.69-I.,77 . 31. Giblin, F. J., J. R. Reddan, 1.. Schrimscher, D. C. Dziedzic, and V. N. Reddy. 1990. The relative roles of the glutathione redox cycle and catalase in the detoxification of H 20 2 by cultured rabbit lens cells. Exp. Eye

Res. 50:795-804. 32. Chance, B., H. Sies, and A. Boveris. 1979. Hydroperoxide metabolism in mammalian organs. Physiol. Rev. 59:527-590. 33. Fluhe, 1.. 1979. Glutathione peroxidase brought into focus. In Free Radicals in Biology. Vol. 5. W. A. Pryor, editor. Academic Press, New York. 223-254. 34. Thies, R. 1.., and A. P. Autor. 1991. Reactive oxygen injury to cultured pulmonary artery endothelial cells. Mediation by poly (ADP-ribose) polymerase activation causing NAD depletion and altered energy balance.

Arch. Biochem. Biophys. 286:353-363. 35. Meister, A., and M. E. Anderson. 1983. Glutathione. Annu. Rev. Bio-

chem. 52:711-760. 36. Weitzman, S. A., and A. B. Weitberg. 1985. Asbestos catalyzed lipid peroxidation and its inhibition by deferoxamine. Biochem. J. 225: 259-262.