- Email: [email protected]
oxidase to oxygen-mediated cell injury
ARCHIVES
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 294, No. 2, May 1, pp. 695-702, 1992
The Contribution of Vascular Endothelial Xanthine Dehydrogenase/Oxidase to Oxygen-Mediated Cell Injury Peter C. Panus,*,$ Sally A. Wright,* Departments Birmingham,
of *Anesthesiology, tBiochemistry, Alabama 352334810
Received July 26, 1991, and in revised form January
Phillip
H. Chumley,*
$Pediatrics,
University
of Alabama at Birminghnm,
8, 1992
The conversion of xanthine dehydrogenase (XDH) to xanthine oxidase (X0) and the reaction of X0-derived partially reduced oxygen species (PROS) have been suggested to be important in diverse mechanisms of tissue pathophysiology, including oxygen toxicity. Bovine aortic endothelial cells expressed variable amounts of XDH and X0 activity in culture. Xanthine dehydrogenase plus xanthine oxidase. specific activity increased in dividing cells, peaked after achieving confluency, and decreased in postconfluent cells. Exposure of BAEC to hyperoxia (95% 0,; 5% COz) for O-48 h caused no change in cell protein or DNA when compared to normoxic controls. Cell XDH + X0 activity decreased 98% after 48 h of 95% O2 exposure and decreased 68% after 48 h normoxia. During hyperoxia, the percentage of cell XDH + X0 in the X0 form increased to lOO%, but was unchanged in air controls. Cell catalase activity was unaffected by hyperoxia and lactate dehydrogenase activity was minimally elevated. Hyperoxia resulted in enhanced cell detachment from monolayers, which increased 112% compared to controls. Release of DNA and preincorporated [S”C]adenine was also used to assess hyperoxic cell injury and did not significantly change in exposed cells. Pretreatment of cells with allopurinol for 1 h inhibited XDH + X0 activity lOO%, which could be reversed after oxidation of cell lysates with potassium ferricyanide (KSFe(CN),). After 48 h of culture in air with allopurinol, cell XDH + X0 activity was enhanced when assayed after reversal of inhibition with K,Fe(CN),, and cell detachment was decreased. In contrast, allopurinol treatment of cells 1 h prior to and during 48 h of hyperoxic exposure did not reduce cell damage. After K3Fe(CN)g oxidation, XDH + X0 activity was undetectable in hyperoxic cell lysates. Thus, X0-derived PROS did not contribute to 1 Permanent address: Department of Biochemistry, Faculty of Medicine, University of the Republic, Montevideo, Uruguay CP11800. ’ To whom correspondence and reprint requests should be addressed at Department of Anesthesiology, University of Alabama at Birmingham, 619 19th St. South, 941 THT, Birmingham, AL 35233-6810. 0003-9661/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Rafael Radi,*Tl and Bruce A. Freeman**t$,2
and $Pharmacology,
cell injury or inactivation of XDH + X0 during hyperoxia. It is concluded that endogenous cell X0 was not a significant source of reactive oxygen species during hyperoxia and contributes only minimally to net cell proo 1992 Academic duction of 0, and HzOz during normoxia. Press, Inc.
Xanthine dehydrogenase (XDH)3 and xanthine oxidase (X0) are isozymes of a molybdo-pterin enzyme, which varies in tissue specific activity from organ to organ and between species (1). Typically, 80~35% of tissue XDH + X0 activity in viuo is in the dehydrogenase form (2, 3). Both XDH and X0 oxidize a variety of purine and pteridine substrates (4). Xanthine dehydrogenase reduces NAD+, while the oxidase form reduces molecular oxygen univalently and divalently, yielding superoxide (0;) and hydrogen peroxide (H202) (5, 6). In recent years, the contribution of X0-derived PROS to postischemic reperfusion injury has generated interest in X0 as a source of PROS in tissues (2,3, 7,8). During hypoxia, purine degradation elevates tissue hypoxanthine and xanthine and conversion of XDH to X0 can sometimes occur (9). Conversion of XDH to X0 may occur by sulfhydryl oxidation of XDH or by limited proteolysis of XDH to an irreversible X0 form (2,lO). Both mechanisms of XDH to X0 conversion can occur during ischemiareperfusion phenomena (2, 3, 8). Upon reperfusion and reintroduction of oxygen, increased 0, and H202 formation by X0 and other tissue sources can occur. Allopurinol, an inhibitor of XDH and X0 activity (ll), or ‘Abbreviations used: BAEC, bovine aortic endothelial cells; H,Oz, hydrogen peroxide; K,Fe(CN)s, potassium ferricyanide; LDH, lactate dehydrogenase; 0; , superoxide; PROS, partially reduced oxygen species; TCA, trichloroacetic acid; XDH + X0, xanthine dehydrogenase + xanthine oxidase; % X0, percentage of total XDH + X0 activity in the oxidase form; HBSS, Hanks’ balanced salt solution; ANOVA, analysis of variance; RGS, Ryan’s growth supplement. 695
696
PANUS
dietary tungsten, an inactive analog of molybdenum which inactivates XDH and X0 (12), often attenuates tissue injury (7,13, 14). Evidence gained from inhibitors of X0 supports a role for this source of 0, and Hz02 in tissue reperfusion injury to liver, kidney, intestine, and lung. In the lung, X0-derived PROS may also mediate pulmonary damage during hyperoxia as well as during hypoxia (2, 14-17). Perfusion of isolated rat lungs from hyperoxic-exposed animals showed enhanced edematous injury, which was attenuated by pretreatment with dietary tungsten, suggesting a role for X0 in hyperoxia-induced lung edema formation and injury (16). Reduced lung X0 activity and serum uric acid was observed after exposure of rats to >98% O2 for 48 h, suggesting oxidant-mediated inactivation of X0 by PROS of undefined origin (15). Also, hyperoxic exposure of rats and cultured bovine pulmonary artery endothelium resulted in the loss of XDH and X0 activity (17). Morphologic examination of lungs from rats exposed to hyperoxia revealed damage to pulmonary endothelium (18), suggesting these cells are key sources of PROS during oxidant stress or are targets of PROS derived from other intravascular sources. It has been proposed that endothelial X0 was a source of PROS which contributed to pulmonary injury and the oxidative inactivation of XDH and X0; however, XOderived oxidants were not causally related to hyperoxic damage (15-17). Herein, changes in endothelial cell XDH and X0 activity during control and hyperoxic conditions were assessed, and the contribution of endothelial cell X0 to oxygen-mediated cell injury was defined. EXPERIMENTAL
PROCEDURES
Endothelial cell culture. Cells were isolated from bovine aorta and cultured in bicarbonate buffered Medium 199 (M-199, GIBCO), supplemented with 5% fetal bovine serum (Hyclone), 5% Ryan’s growth supplement (Dr. Una Ryan, Monsanto Inc., St. Louis, MO), and 1% antibiotics/antimycotic (GIBCO). Cells were grown in T-75 flasks (Costar, Cambridge, MA), kept humidified (37°C) in 95% air, 5% CO2 prior to experimentation, and used between passages 6 to 10. Unless stated otherwise cells were subcultured in a 1:3 ratio, were confluent within 6 to 8 days and medium was replaced every 4 to 7 days after subculturing. Calf serum was obtained from Grand Island Biological Co. (Grand Island, NJ), and fetal calf serum and iron-supplemented calf serum were obtained from Hyclone (Logan, UT). 0, consumption. BAEC were subcultured in a 1:4 ratio and 7 days later total and cytochrome c oxidase independent cell Ox consumption was measured in confluent cells removed from flasks by scraping. Cells were centrifuged at 25Og, 10 min, 4”C, and resuspended in Hanks’ balanced salt solution (HBSS, GIBCO), with 1 mM N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid and 12 mM glucose, pH 7.4. Oxygen consumption was measured in a polarograph with a Clark electrode. Potassium cyanide (25-200 PM), myxothiazole (0.007-0.5 PM), and antimycin A (0.0033-1.0 PM) dose-response studies were performed to determine optimum concentrations for assessing cytochrome c oxidase independent 0s consumption. Hyperoxic exposure. Fresh media was added to cells 30 min prior to O-48 h of hyperoxic (95% 0s. 5% COz) or normoxic (95% air, 5% CO,) exposure in a humidified environment (37’C). At termination of experiments, media was saved for analysis, 5 ml of cold (4°C) HBSS was added to cells, monolayers were scraped and centrifuged (2OOg, 10 min,
ET AL. 4’C), and pellets were resuspended in 1 ml lysing buffer containing 10 mM dithiothreitol and antiproteases intended to inhibit artifactual XDH to X0 conversion (19). Cells were then sonicated (75 W, 10 s), rapidly frozen in liquid nitrogen or -86°C isopentane, and stored at -86°C until analysis. Biochemical analyses. Several markers of cytotoxicity were utilized. Detached cells, assumed to be nonviable, were counted with a Coulter Counter (Model ZM). Release of preincorporated [8-“CJadenine (New England Nuclear) was also utilized as a marker of cytotoxicity (20,21). Cells were cultured 36 h with [8-“Cladenine (4 PM) and washed twice with fresh media prior to initiation of exposures. Aliquots of cell lysates and centrifuged culture media (2OOg, 10 min, 4’C) were analyzed for radioactivity at the end of experiments. Other aliquots of lysates and media from the same cell samples were added to 20% trichloroacetic acid (TCA) 1:l (v/v) and centrifuged (SSOOg,2 min, 25°C). The centrifuged supernatant was assayed for radioactivity to determine the proportion of acid-soluble incorporated counts. Cell and media-associated DNA and LDH activity were also determined (20,22). Preincorporated [8-“Cladenine, LDH, and DNA were reported as a percentage of total original activity in monolayers which was released into the media: % release = [(media activity)/(media
activity + cellular
activity)]
X 100%.
Catalase activity was determined spectrophotometrically (23) and protein was determined by dye binding (24). Cellular XDH and X0 activities were determined as previously described (25) by allopurinol inhibitable oxidation of pterin to fluorescent isoxanthopterin (&.. = 345 nm, X, = 390 nm). Methylene blue was substituted for NAD as an electron acceptor when measuring total XDH + X0 activity. Sample quenching was adjusted for by adding known amounts of isoxanthopterin as an internal standard. The percentage of total XDH + X0 activity in the oxidase form (% X0) was determined by dividing the X0 activity by the total XDH + X0 activity. To determine the stability of LDH and XDH + X0 activity in media, freshly prepared complete culture medium was exposed to hyperoxia and normoxia in the absence of cells. Allopurind and BAEC XDH + X0 activity. In some experiments cells were preincubated with 100 PM allopurinol for 1 h in complete media, prior to exposure to O-48 h of normoxia or hyperoxia. The cells were subsequently harvested as previously described. Nonadhered cells were measured and lysate XDH + X0 activity, protein, and DNA determinations done. Aliquots of lysates were incubated for 10 min with 10 mM potassium ferricyanide (KeFe(CN)a) to reverse binding of the allopurinol metabolite oxypurinol via oxidation of the enzyme molybdop&in center (26). Cell lysates were then chromatographed on Sephadex G-25M PD-10 columns (Pharmacia) and the first 1.0 ml after the void volume was collected and assayed for XDH + X0 activity and protein. In allopurinol inhibition reversal experiments, the cell lysing buffer was modified by decreasing dithiothreitol from 10 to 1 mM; because 10 mM, but not 1 mM, dithiothreitol completely reduced K,Fe(CN), and prohibited reversal of allopurinol inhibition. Statistical analysis. Unless otherwise stated, all data are expressed as mean + standard deviation (X t- SD). Statistical analysis and modeling were performed as previously described (20). Actual data analysis was done using “SAS System, release 6.02.” Data were analyzed via a twoway analysis of variance (ANOVA) and Duncan post hoc groupwise comparison to account for experiment-to-experiment variation within a group, and significance between groups was determined for dependent variables. Significance for ANOVA and Duncan testing was set at For a ~0.05, unless otherwise stated.
RESULTS Cell culture, XDH + X0 actiuity. The XDH + X0 activity of cells in normoxia varied with time after sub-
ENDOGENOUS
z iii s
60
E
50
r
40
XANTHINE
OXIDASE-MEDIATED
ENDOTHELIAL
TABLE
ii
30
+
20
g
10
3 a
0
697
INJURY I
BAEC O2 Consumption in the Absence and Presence of Mitochondrial Respiratory Chain Inhibitors nmol Oz. mini protein’
Variable
22.0 0.4 1.0 6.3
Control Myxothiazole (0.333 pM) Antimycin A (0.033 FM) KCN (200 /.iM)
3
5 9 10 15 24 30 DAYS POST SUBCULTURE
FIG. 1. Changes in total cellular XDH + X0 activity in BAEC during normoxic culture. Total cell XDH + X0 activity was normalized for cell protein (au XDH + XO/mg protein). Peak cell XDH + X0 activity on Day 9 was significantly greater (*) than the activity on any of the other recorded days. The histogram represents the culmination of two separate experiments, in which the cells were subcultured 1 to 4.
culturing (Fig. 1). Specific activity of XDH + X0 increased in preconfluent cells during the first 3 to 7 days after subculturing. The BAEC were 95100% confluent at Day 7, and peak XDH + X0 activity was significantly greater at Day 9 than at any other time measured. Subsequently, the XDH + X0 activity declined in postconfluent cells, even when media was regularly changed, ruling out depletion of key nutrients as a contributory factor. Cell morphology observed by phase contrast microscopy revealed an unchanged cobblestone appearance in cells during the entire study period. Over 30 days, the percentage XDH + X0 in the oxidase form varied, with no consistent pattern. The percentage oxidase at peak specific activity (Day 9 postsubculture) was 53 +_4%. In subsequent studies, cells with maximal XDH + X0 activity (7 to 10 days postsubculture) were utilized. Endothelial cell 0, consumption. For calculation of the contribution of X0 to cell cytochrome c oxidase independent oxygen consumption (Table I), preliminary doseresponse studies of BAEC respiratory chain activity inhibition gave optimum inhibitor concentrations. Myxothiazole (0.333 PM), antimycin A (0.033 PM), and KCN (200 PM) decreased O2 consumption by 98,95, and 71%, respectively. The KCN-resistant O2 consumption was significantly greater than myxothiazole and antimycin Aresistant respiration. Hyperoxia and cellular XDH and X0. BAEC were exposed to O-48 h of normoxia or hyperoxia. Phase contrast microscopy demonstrated no significant differences between hyperoxic and control cell morphology after 48 h of exposure (data not shown), nor was there any significant change in cell monolayer DNA (Table II). Cell-associated protein similarly increased in both air- and 02exposed cells (Table II). Total cell XDH + X0 activity decreased in BAEC exposed to both normoxia and hyperoxia (Fig. 2), with the
* mg n
-+ 6.0 It 0.2’s** + 0.6*,** + 2.2’
16 4 4 8
Note. Inhibitor concentrations represent the minimum utilized, which resulted in maximal inhibition of O2 consumption. * Different from control. ** Different from the KCN group.
loss of XDH + X0 activity in hyperoxia significantly greater than air controls. After 48 h of hyperoxia, only two of nine flasks had detectable cellular XDH + X0 activity. The percentage of total cellular XDH + X0 activity in the X0 form (% X0) increased in Os-exposed BAEC and was significantly different at 48 h from air controls, which remained constant (Fig. 2). While the % X0 in the 48-h hyperoxic-exposed BAEC was greater than controls, X0 activity in hyperoxic cells (0.60 + 1.5 pIJ/ mg protein) was less than air controls (5.4 t 4.2 pU/mg protein). In these experiments, 10 mM dithiothreitol was added to the lysing buffer, to prevent artifactual XDHto-X0 conversion. It was observed that dithiothreitol (O50 mM, 30 min, 37°C) did not alter the percentage of total cellular XDH + X0 activity in the X0 form (data not shown). Thus, changes in X0 activity were not due to inadvertent sulfhydryl oxidation of XDH or disulfide reduction of X0 prior to biochemical analysis. Extracellular release of XDH + X0 activity into the media of 48-h oxygen-exposed BAEC was not detectable. Thus, if XDH and/or X0 was released from BAEC exposed to hyperoxia, enzymatic activity was too dilute in
TABLE
II
Alterations in BAEC DNA and Protein during Exposure to Normoxia or Hyperoxia DNA (wdcm*) Exposure time (h)
Air
0 12 24 48
N.D. N.D. N.D. 2.1 + 0.5
Protein
95% 02 2.0 2.0 2.0 2.0
F + * +
0.5 0.4 0.4 0.4
Air N.D. N.D. N.D. 14 + 1*
(pg/cm*) 95% 02 11 13 13 13
Itr 2 + 2* f 1* rt 1*
Note. Changes in cellular DNA and protein were normalized for surface area (cm*) of the flask. Values represent three separate experiments done in triplicate, and control values at 12 and 24 h were not determined (N.D.). * Different from to control.
698
PANUS
ET AL. TABLE
III
Alterations in Cell Catalase and Lactate Dehydrogenase (LDH) Activities during Normoxia and Hyperoxia
100 00 t 60
3
40
8
20
s
Catalase (U/mg protein) Exposure time (h)
Air
0 12 24 48
N.D. N.D. N.D. 13 If: 3
0
95% 0, (HRS) FIG. 2. Alterations in BAEC XDH and X0 activity during hyperoxic and normoxic exposures. Changes in total cellular XDH + X0 activity were normalized for protein (pU XDH + XO/mg protein), whereas (% OXIDASE) was the amount of cellular XDH/XO activity in the X0 form. The closed circles and closed triangles represent, respectively, the total cell XDH + X0 activity and the % oxidase, in the hyperoxicexposed cells, whereas the open circle and open triangle represent similar 48-h normoxic control values. (*) and (+) denote when different from to and t controls, respectively. Enzymatic activity was determined from BAEC groups previously described in Table II. Values represent three separate experiments done in triplicate (n = 9).
media to be detectable or the enzyme was inactivated. There was X0 activity in culture media used herein (1.53 f 0.65 PU XO/ml, n = 6, with no detectable XDH activity). The X0 activity in the culture medium exposed to hyperoxia and normoxia, in the absence of cells, decreased to undetectable levels during the first 12 h of hyperoxia and after 48 h of normoxia. The source of this X0 activity was Ryan’s growth supplement (65 + 14 pU/ml, n = 6), with activity as high as 113 PU XO/ml observed. Ironsupplemented bovine serum is a major component of RGS, having 213 to 360 PU XO/ml when analyzed from several different lots and sources, and no detectable XDH. In contrast, several different lots of fetal bovine serum showed no detectable XDH or X0 activity. The loss of X0 activity in the cell-free complete medium may be due to autoinactivation by X0-derived PROS. The serumsupplemented cell-free medium contained 6 PM hypoxanthine and 4 PM xanthine, providing sufficient substrate for 0; and HzOz formation. These purine substrates were derived from both the basal medium and serum supplements. Hyperoria and indicators of cell injury. Endothelial cell lactate dehydrogenase activity (per milligram protein) was marginally increased (a < 0.10) after 12 and 24 h hyperoxia when compared to both 0- and 48-h controls (Table III), but returned toward control values after 48 h hyperoxia. Two-way analysis of variance demonstrated significant differences in LDH activities between the various endothelial cell lines utilized for the present experiments. However, the LDH activity within a given cell line was constant at 0 and 48 h normoxia. LDH activity
LDH (mU/mg protein)
95% 02 12 12 12 13
+ + + f
Air
2 3 2 2
N.D. N.D. N.D. 681 f 413
95% 02 587 826 819 767
-t + f f
411 638**** 59a**** 459*
Note. Enzymatic catalase and LDH activities were normalized for cell protein and were determined from the BAEC groups in Table II. Values represent two separate experiments done in triplicate (n = 6). * As previously described in Table II. ** Different (P < 0.10) from t*s control.
in the medium also increased during hyperoxia (Table IV). The media initially contained LDH activity (11 + 0.2 mU/cm2; n = 3) derived from serum supplements. In the absence of cells, exposure of the complete medium to hyperoxia or normoxia for 48 h resulted in a loss of medium LDH activity. Loss of LDH activity in the medium under cell-free conditions during hyperoxia was marginally greater ((u < 0.10) than that in the corresponding media maintained in normoxia. Thus, LDH activity released into the media was normalized for serum-derived LDH activity prior to calculation of release from cells. The LDH released by BAEC during hyperoxia was greater than controls (Table IV), but the percentage of total LDH released by cells into the medium did not increase, due
TABLE
IV
Effects of Normoxia and Hyperoxia and Medium LDH Activity
Time (h)
Exposure
0 12 24 48 48
Air 95% 95% 95% Air
LDH released (mu/cm’) 1.8 2.9 3.3 3.6 1.7
+ f + ? +
1.3 0.6 1.0*,** 1.4*~** 0.7
Cell + LDH released (mu/cm’)
14 14 14 10
9.5 + 9*,** k 8*,** + 8*,** t 5
on Cell
% LDH released (mu/cm’) 27 32 32 29 21
f 20 2121 f 18 f 12 k 14
Note. BAEC were exposed to O-48 h of hyperoxia (95%) or normoxia (Air). Medium LDH activity from cells (column 3) was corrected for LDH activity from medium serum supplements. Medium LDH activity varied with time and was determined in similar flasks under cell-free conditions (to = 11 f 2, ti2 = 10 + 0.8*, tul = 10 f l*, r& = 9 + l*, and control trs = 10 * 2* mu/cm’), representing two separate experiments, n = 5. LDH activities were determined from BAEC reported in Table III, representing two experiments done in triplicate, n = 6. * Different from to control. ** Different from taacontrol.
ENDOGENOIJS
XANTHINE
OXIDASE-MEDIATED
ENDOTHELIAL
699
INJURY
l
P4 g
20
3
16 1P
95% 0, (HRS) FIG. 3. Cell damage during 48-h BAEC hyperoxic exposure. Cell damage (detached cells) is represented as the number of detached cells normalized for surface area of the flask (cm’). The (0) and (A) represent the number of detached cells in BAEC exposed to hyperoxia and normoxia, respectively. (*) and (+) denote when different from to and t& controls, respectively. Cell damage was determined from the BAEC groups previously described in Table II. Values represent two separate experiments done in triplicate (n = 6) (X -t SD).
to the increase in cell LDH during exposure to hyperoxia. When cell LDH activity and that released from cells into the media were added, the total LDH activity in BAEC exposed to hyperoxia was significantly greater than air controls (Table IV). Thus, the increase in cell LDH in hyperoxic BAEC masked the increase in LDH released from cells into the media (Table IV). Release of other markers of cellular damage during the 48-h hyperoxic [8-14C]adenine, exposure, DNA and preincorporated were not significantly elevated when compared to the 48-h normoxia controls. Only detached cell numbers increased significantly in cells exposed to hyperoxia for 48 h (Fig. 3). Allopurinol effects on cell XDH and X0 activity and hyperoxic injury. Loss of cell XDH + X0 activity and XDH-to-X0 conversion paralleled cell damage as indicated by increased numbers of detached cells. Therefore, subsequent experiments examined whether changes in XDH and X0 activity were responsible for cell damage during hyperoxia. BAEC were preincubated with 100 PM allopurinol 1 h prior to exposure to 48 h of normoxia or hyperoxia (Fig. 4; Table V). Allopurinol completely inhibited BAEC XDH + X0 activity and was not cytotoxic, as indicated by an absence of increased cell detachment. Treatment of air-exposed allopurinol-treated t = 0- and t = 48-h cell lysates with K,Fe(CN), increased detectable XDH + X0 activity by 0 and 214%, respectively (Fig. 4). After 48 h air exposure of cells not treated with allopurinol, XDH + X0 activity assayed with or without enzyme oxidation by K,Fe(CN), was reduced by 50%. Hyperoxic exposure for 48 h reduced XDH + X0 activity to undetectable levels. This complete loss of XDH + X0 activity in hyperoxic cells was not prevented by maintaining cells
6
1
4
0
46 AIR
7I
FIG. 4. The BAEC were pretreated in complete media + 100 pM allopurinol for 1 h and subsequently subjected to O-48 h of air (CTL) or hyperoxia (Ox). Total XDH + X0 activity was assayed in both crude cell lysate (m) or after the lysate was treated with K,Fe(CN), and the void volume collected after gel filtration (I@. (*) and (+) denote when values are different from to and ta controls, respectively, without allopurinol. Figure represents two separate experiments done in triplicate, n = 6.
in allopurinol and was not reversed by enzyme reoxidation with K,Fe(CN),, in contrast to control cells maintained in air. In cells maintained in normoxia, prior inhibition of XDH + X0 with allopurinol resulted in a threefold increase in XDH + X0 specific activity after K,Fe(CN), treatment, when compared to untreated 48-h air control cells. Hyperoxic exposure increased cell detachment from the monolayer (Fig. 3 and Table V). Allopurinol inhibition of XDH and X0 caused a minor but significant reduction in cell detachment in monolayers exposed to air for 48 h and was not protective for oxygen-exposed cells (Table V).
TABLE Effects
of Hyperoxia
V
and Allopurinol
on Cell Injury
Detached cell number Time (h) Air Air Hyperoxia
0 48 48
Control 203 + 205 814 _fr 307*
1109 + 287*,*+*
Allopurinol 113 2 120 504 + 229*,**
1216 + 264*,**
Note. Table represents cell damage (detached cells) from BAEC described in Fig. 4. Table represents two separate experiments done in triplicate, n = 6. * Different from to control without allopurinol. ** Different from t@ control without allopurinol. *** Different from the t48 control without allupurinol (a 5 0.10).
700
PANUS
DISCUSSION
The role of xanthine oxidase in oxygen-mediated injury to cultured vascular endothelium was examined because numerous investigations have suggested an important role for X0-derived oxidants in tissue damage occurring secondary to anoxia-reoxygenation, hyperoxia, and neutrophi1 activation (2,3,8,14,17,27). Many of these processes resulted in endothelial damage and increased vascular permeability (14,16, 28). Pretreatment with antioxidant enzymes or compounds which inhibited XDH and X0 activity sometimes prevented endothelial damage and attenuated vascular permeability increases (14,16,28, 29). Formation of the potentially cytotoxic oxidase from XDH may be mediated by a variety of processes, including complement activation, partial proteolysis of XDH by intracellular or neutrophil-derived enzymes, increased sulfhydryl oxidation and glutathione mixed disulfide formation with XDH (2, 8, 27, 30). More extensive oxidation of XDH thiols to sulfenic, sulfinic, or sulfonic acid derivatives by 0, and HzOz (31) may also be expected to yield X0. We compared the effect of hyperoxia on cell LDH, catalase, and XDH + X0 activity. Cell LDH activity increased marginally during the first 24 h of hyperoxic exposure, possibly due to increased glycolytic activity in response to mitochondrial injury (32). The subsequent decrease in cellular LDH activity at 48 h may be due to the predominance of LDH inactivation by PROS (20). It has been previously observed that catalase activity decreases significantly in porcine endothelium exposed to hyperoxia for 48 h (33). A similar oxidant stress to cultured bovine aortic endothelium herein showed catalase activity to be unaffected by hyperoxia. Previous investigations have utilized both detached cells and cellular release of LDH as markers of injury to endothelial cells exposed to 24 h or more of chronic oxidant stress (34,35) with cell detachment a more sensitive marker of cell damage in comparison to LDH release (35). Herein we examined the effects of 48 h of hyperoxia on both cell detachment and cellular release of LDH. Detached cell numbers and LDH release from monolayers increased significantly during the 48-h oxygen exposure to hyperoxia. Release of the cytotoxicity markers, preincorporated [8-14C]adenine and cell DNA, were not significantly elevated in the media when compared to controls (data not shown). The assessment of LDH release from oxidant-injured cells is complicated by a number of underlying phenomena. For example, cell culture medium serum supplements contain not only LDH, but also xanthine oxidase. In cell-free conditions, hyperoxia caused loss of endogenous serum-derived LDH activity, possibly mediated by autooxidation of medium components, since it is known that LDH can be inactivated by PROS (20, 36). Another problem with using LDH as a cytotoxicity marker relates to the increasing endogenous LDH specific
ET AL.
activity of stressed cells, which may not similarly occur under control conditions. We conclude that the instability of cell and medium LDH and its susceptibility to oxidant inactivation make this enzyme an unreliable marker which can lead to misinterpretation of oxidant cell injury. This concept extends to animal models of ischemia-reperfusion injury, where similar oxidant inactivation of creatinine phosphokinase may lead to complications in assessment of extents of myocardial injury and antioxidant protection (37, 38). In our studies, endothelial cell XDH + X0 activity increased after subculture of normoxic cells until confluency and then declined. The cause of this change in activity could not be ascribed to cell injury or inhibition of protein synthesis and warrants further investigation. Allopurinol-mediated inhibition of XDH and X0 depends on oxidation of allopurinol to oxypurinol, which binds to a partially reduced form of the enzyme having the molybdenum trapped at the Mo(IV) level (26). Reversal of allopurinol inhibition of control cell XDH and X0 for 48 h by enzyme oxidation with K,Fe(CN), resulted in recovery of more XDH and X0 activity than was present in parallel cultures of cells maintained in the absence of allopurinol for 48 h. This suggests that X0 may contribute to its own inactivation during normoxia or that allopurinol is having other unrecognized effects on cell metabolism. We observed a general loss of XDH and X0 activity and conversion of XDH to X0 in the hyperoxia-exposed cells, which was not reversible by dithiothreitol. This suggests that the increased cellular X0 activity during hyperoxia was either due to dithiothreitol-resistant sulfhydryl oxidation or partial proteolysis. For cells exposed to 48 h of hyperoxia, there was an almost complete loss of cell XDH and X0 activity, which occurred concurrently with a pathophysiologically insignificant conversion of minimal residual XDH activity to X0. These observations differ from previously reported studies of the contribution of endothelial X0 to free radical injury. First, the percentage XDH + X0 in the oxidase form in our control cells remained consistently near 45%, previously reported to be 9% for bovine pulmonary endothelial X0 (17). The percentage oxidase activity of control endothelial cell XDH reported by these investigators is lower than the partial oxidase activity of pure rat liver XDH or flavin-reconstituted purified XDH (2, 39), raising concern about assay methodology. Also, endothelial cells have been reported to have an XO-dependent extracellular 0; release of -0.5 nM * lo6 cells-‘. 4 h-‘. Assuming 4 X lo6 cells per T-25 flask incubated with 1 ml of cytochrome c solution, this reported rate of cell 0; release would give an undetectable A A550-0.00004 over 4 h (16), even if tungsten were not having other unrecognized effects on cell metabolism. XDH + X0 activity varies between tissues of the same species and between the same tissue from different species. Cultured rat endothelial cells have the greatest XDH
ENDOGENOUS
XANTHINE
OXIDASE-MEDIATED
+ X0 specific activity (27, 30), followed by bovine endothelium (16, 17) (Fig. 4) and finally cultured porcine endothelium, where XDH + X0 activity is undetectable (unpublished observations). At present, XDH + X0 activity has not been reported in cultured human endothelium. The significance of tissue X0 as a source of partially reduced O2 formation during basal or pathologic cell conditions depends upon substrate availability, enzyme specific activity, and the relative contribution of other cellular sources of 0, and HzOz. Myxothiazole-resistant O2 consumption in our cultured bovine endothelial cells was 2.0 nmol * min-’ * mg DNA-‘. Calculation of cytochrome c oxidase independent O2 consumption was based upon myxothiazole-resistant respiration measurements, because this inhibitor does not artifactually enhance complex I and II 0, production, as does antimycin A and CN(40). Additionally, CN- inhibits a portion of cytochrome b-derived 0, production (40) and reacts with diverse cellular hemeproteins and antioxidant enzymes, thus complicating measurement of cytochrome c oxidase independent O:! consumption. Assuming complete reduction of 0, and Hz02 to H,O by antioxidant enzymes during myxothiazole-resistant O2 consumption, a maximum of 8 nmol OZ. rnin~l. mg DNA-’ could have been reduced to 0, or HzOz by cells. This prediction of cell 0; and HaOz production is likely an overestimate, because covalent incorporation of O2 into substrates such as arachidonic acid, which does not include 0;) HzOz, or . OH as intermediates, is simultaneously occurring in endothelial cells. Exposure of cells to 95% O2 should increase cell PROS formation two- to threefold (41-43). The rate of reduction of O2 to 0, and H202 by X0 in these cells would maximally have been 0.29 nmol . mini’ . mg DNA-‘. With the K, of X0 for O2 being 46 KM (44) and assuming that no O2 gradients existed between the media/air interface and the cell surface (45), the increase in X0 activity under saturating O2 conditions would only be 20%, increasing X0-dependent O2 reduction to 0.35 nmol . min-’ . mg DNA-‘. In summary, the maximal contribution of XOderived PROS to the overall cell formation of PROS would be 3.6% during normoxia and less during hyperoxia. Further, the loss of X0 during hyperoxia would diminish the relative contribution of oxidants from this source with time. Thus, other cell sources contribute more substantially to overall cell PROS formation during hyperoxia and can include mitochondria, endoplasmic reticulum, nuclear membranes, and soluble flavoproteins or hemeproteins (41-43,46). These predictive calculations, while only approximate, suggest that X0-derived PROS do not contribute significantly to overall cell PROS formation during hyperoxia and are supported by experimental observations herein as well as by those of Royal1 et al. (47). This does not rule out a potential pathophysiologic role for X0 under other circumstances, however. In summary, hyperoxic endothelial cells lost XDH and X0 activity concurrent with an XDH-to-X0 conversion.
ENDOTHELIAL
INJURY
701
Endothelial cell XDH + X0 activity decreased with early hyperoxic cell damage, but had no causal relationship with hyperoxic cell injury. ACKNOWLEDGMENTS This research was supported by grants from the National and Alabama Affiliate of the American Heart Association, NIH Grant NS24275, and the Council for Tobacco Research. P. C. Panus was supported in part by National Heart, Lung, and Blood Training Grant T32-HL-07553. We also wish to acknowledge the assistance of Joan Wheat and Dr. Dale Parks in HPLC analyses.
REFERENCES 1. Parks, D. A., and Granger, D. N. (1986) Acta Physiol. Stand. 548, 87-99.
2. Engerson, T. D., McKelvey,
T. G., Rhyne, D. B., Boggio, E. B., Snyder, S. J., and Jones, H. P. (1987) J. Clin. Inuest. 79, 1564-
1570. 3. Parks, D. A., Williams,
T. K., and Beckman, J. S. (1988) Am. J. Physiol. 254, G768-G774. 4. Krenitsky, T. A., Spector, T., and Hall, W. W. (1986) Arch. Biochem. Biophys. 247, 108-119. 5. McCord, J. M., and Fridovich, I. (1968) J. Biol. Chem. 243, 5753-
5760. 6. Olson, J. S., Ballow, D. P., Palmer, G., and Massey, V. (1974) J. Biol. Chem. 249, 4350-4362.
7. Parks, D. A., Bulkley,
G. B., Granger, D. N., Hamilton, S. R., and McCord, J. M. (1982) Gastroenterology 82, 9-15. 8. McKelvey, T. G., Hollwarth, M. E., Granger, D. N., Engerson, T. D., Landler, U., and Jones, H. P. (1988) Am. J. Physiol. 264, G753-G760. 9. McCord, J. M. (1985) N. Engl. J. Med. 312, 159-163. 10. Waud, W. R., and Rajagopalan, K. V. (1976) Arch. Biochem. Biophys.
172,365-379. 11. Spector, T. (1977) Biochem. Pharmacol. 26, 355-358. 12. Johnson, J. L., Rajagopalan, K. V., and Cohen, H. .J. (1974) J. Biol. Chem. 249,859-866. 13. Granger, D. N. (1988) Am. J. Physiol. 255, H1269-H1275. 14. Grosso, M. A., Brown, J. M., Viders, D. E., Mulvin, D. W., Banejee, A., Velasco, S. E., Repine, J. E., and Harken, A. H. (1989) J. Surg. Res. 46, 355-360. 15. Elsayed, N. M., and Tierney, D. F. (1989) Arch. Biochem. Biophys.
273,281-286. 16. Rodell, T. C., Cheronis, J. C., Ohnemus, C. L., Piermattei, D. J., and Repine, J. E. (1987) J. Appl. Physiol. 63, 2159-2163. 17. Terada, L. S., Beehler, C. J., Banerjee, A., Brown, J. M., Grosso, M. A., Harken, A. H., McCord, J. M., and Repine, J. E. (1988) J. Appl. Physiol. 65, 2349-2353.
18. Crapo, J. D., Barry, B. E., Fescue, H. A., and Shelburne, J. (1980) Am. Rev. Respir. Dis. 122, 123-143. 19. Panus, P. C., Shearer, J., and Freeman, B. A. (1988) Elcp. Lung Res.
14,959-976. 20. Panus, P. C., Matalon, 21.
22. 23. 24. 25.
S., and Freeman, B. A. (1989) In Vitro Cell Dew. Biol. 25, 821-829. Shirhatti, V., and Krishna, G. (1985) Anal. Biochem. 147,410-418. Labarca, C., and Paigen, K. (1980) Anal. Biochem. 102,3443-3452. Beutler, E. (1975) in Red Cell Metabolism: A Manual of Biochemical Methods, pp. 71-75, Grune & Stratton, New York. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. Beckman, J. S., Parks, D. A., Pearson, J. D., Marshall, P. A., and Freeman, B. A. (1989) Free Radicals Biol. Med. 6, 607-615.
702
PANUS
26. Massey, V., Komai, H., Palmer, G., and Elion, G. B. (1970) J. Biol. Chem. 245,2837-2844. 27. Phan, S. H., Gannon, D. E., Varani, J., Ryan, U. S., and Ward, P. A. (1989) Am. J. Pathol. 134, 1201-1211. 28. Grosso, M. A., Viders, D. E., Brown, J. M., Mulvin, D. W., Miles, R. H., Brentlinger, E. R., Velasco, S. E., Crawford, T. S., Burton, L. K., Repine, J. E., and Harken, A. H. (1989) Surgery 106, 310317. 29. Inauen, W., Payne, D. K., Kvietys, P. R., and Granger, D. N. (1990) Free Radicals Biol. Med. 9, 219-223. 30. FriedI, H. P., Till, G. O., Ryan, U. S., and Ward, P. A. (1989) FASEB J. 3,2512-2518. 31. Radi, R., Bush, K. M., Cosgrove, T. P., and Freeman, B. A. (1991) Arch. Biochem. Biophys. 286, 117-125. 32. Simon, L. M., R&in, T. A., Douglas, W. H., Theodore, J., and Robin, E. D. (1979) J. Appl. Physiol. 47,98-103. 33. Block, E. R., Patel, J. M., and Sheridan, N. P. (1985) J. CeU. Physiol.
122,240-248. 34. Bowman, C. M., Butler, E. N., and Repine, J. E. (1983) Am. Reu. Respir. Dis. 128,469-472. 35. Harlan, J. M., Harker, L. A., Reidy, M. A., Gajdusek, C. M., Schwartz, S. M., and Striker, G. A. (1983) Lab. Znuest. 48, 269-274. 36. Buchanan, J. D., and Armstrong, D. A. (1976) Znt. J. Radiat. Biol. 30,115-127.
ET AL. 37. McCord, J. M., and Russell, W. J. (1988) in Oxy-Radicals lecular Biology and Pathology (Cerutti, P. A., Fridovich, McCord, J. M., Eds.), pp. 27-35, Alan R. Liss, New York.
in MoI., and
38. Omar, B. A., Gad, N. M., Jordan, M. C., Striplin, S. P., Russell, W. J., Downey, J. M., and McCord, J. M. (1990) Free Radicals Biol. Med. 9,465-471. 39. Saito, T., Nishino, 15,390-15,395. 40. Turrens, B&hem.
T., and Massay, V. (1989) J. Biol. Chem. 264,
J. F., Alexandre, A., and Lehninger, Biophys. 237, 408-414.
A. L. (1985) Arch.
41. Turrens, J. F., Freeman, B. A., Levitt, J. G., and Crapo, J. D. (1982) Arch. Biochem. Biophys. 217,401-410. 42. Turrens, J. F., Freeman, B. A., and Crapo, J. D. (1982) Arch. Biochem. Biophys. 217, 411-421. 43. Yusa, T., Crapo, J. D., and Freeman, B. A. (1984) B&him. Acta 798,167-174.
Biophys.
44. Saito, T., and Nishino, T. (1989) J. Biol. Chem. 264,10,015-10,022. 45. Dickman, K. G., and Mandel, c333-c340. 46. Misra, 6965.
H. P., and Fridovich,
L. J. (1989) Am. J. Physiol.
267,
I. (1972) J. Biol. Chem. 247,
6960-
47. Royall, J. A., Gwin, P. D., Parks, D. A., and Freeman, B. A. (1992) Arch. Biochem. Biophys. 294,686-694.
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 294, No. 2, May 1, pp. 695-702, 1992
The Contribution of Vascular Endothelial Xanthine Dehydrogenase/Oxidase to Oxygen-Mediated Cell Injury Peter C. Panus,*,$ Sally A. Wright,* Departments Birmingham,
of *Anesthesiology, tBiochemistry, Alabama 352334810
Received July 26, 1991, and in revised form January
Phillip
H. Chumley,*
$Pediatrics,
University
of Alabama at Birminghnm,
8, 1992
The conversion of xanthine dehydrogenase (XDH) to xanthine oxidase (X0) and the reaction of X0-derived partially reduced oxygen species (PROS) have been suggested to be important in diverse mechanisms of tissue pathophysiology, including oxygen toxicity. Bovine aortic endothelial cells expressed variable amounts of XDH and X0 activity in culture. Xanthine dehydrogenase plus xanthine oxidase. specific activity increased in dividing cells, peaked after achieving confluency, and decreased in postconfluent cells. Exposure of BAEC to hyperoxia (95% 0,; 5% COz) for O-48 h caused no change in cell protein or DNA when compared to normoxic controls. Cell XDH + X0 activity decreased 98% after 48 h of 95% O2 exposure and decreased 68% after 48 h normoxia. During hyperoxia, the percentage of cell XDH + X0 in the X0 form increased to lOO%, but was unchanged in air controls. Cell catalase activity was unaffected by hyperoxia and lactate dehydrogenase activity was minimally elevated. Hyperoxia resulted in enhanced cell detachment from monolayers, which increased 112% compared to controls. Release of DNA and preincorporated [S”C]adenine was also used to assess hyperoxic cell injury and did not significantly change in exposed cells. Pretreatment of cells with allopurinol for 1 h inhibited XDH + X0 activity lOO%, which could be reversed after oxidation of cell lysates with potassium ferricyanide (KSFe(CN),). After 48 h of culture in air with allopurinol, cell XDH + X0 activity was enhanced when assayed after reversal of inhibition with K,Fe(CN),, and cell detachment was decreased. In contrast, allopurinol treatment of cells 1 h prior to and during 48 h of hyperoxic exposure did not reduce cell damage. After K3Fe(CN)g oxidation, XDH + X0 activity was undetectable in hyperoxic cell lysates. Thus, X0-derived PROS did not contribute to 1 Permanent address: Department of Biochemistry, Faculty of Medicine, University of the Republic, Montevideo, Uruguay CP11800. ’ To whom correspondence and reprint requests should be addressed at Department of Anesthesiology, University of Alabama at Birmingham, 619 19th St. South, 941 THT, Birmingham, AL 35233-6810. 0003-9661/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
Rafael Radi,*Tl and Bruce A. Freeman**t$,2
and $Pharmacology,
cell injury or inactivation of XDH + X0 during hyperoxia. It is concluded that endogenous cell X0 was not a significant source of reactive oxygen species during hyperoxia and contributes only minimally to net cell proo 1992 Academic duction of 0, and HzOz during normoxia. Press, Inc.
Xanthine dehydrogenase (XDH)3 and xanthine oxidase (X0) are isozymes of a molybdo-pterin enzyme, which varies in tissue specific activity from organ to organ and between species (1). Typically, 80~35% of tissue XDH + X0 activity in viuo is in the dehydrogenase form (2, 3). Both XDH and X0 oxidize a variety of purine and pteridine substrates (4). Xanthine dehydrogenase reduces NAD+, while the oxidase form reduces molecular oxygen univalently and divalently, yielding superoxide (0;) and hydrogen peroxide (H202) (5, 6). In recent years, the contribution of X0-derived PROS to postischemic reperfusion injury has generated interest in X0 as a source of PROS in tissues (2,3, 7,8). During hypoxia, purine degradation elevates tissue hypoxanthine and xanthine and conversion of XDH to X0 can sometimes occur (9). Conversion of XDH to X0 may occur by sulfhydryl oxidation of XDH or by limited proteolysis of XDH to an irreversible X0 form (2,lO). Both mechanisms of XDH to X0 conversion can occur during ischemiareperfusion phenomena (2, 3, 8). Upon reperfusion and reintroduction of oxygen, increased 0, and H202 formation by X0 and other tissue sources can occur. Allopurinol, an inhibitor of XDH and X0 activity (ll), or ‘Abbreviations used: BAEC, bovine aortic endothelial cells; H,Oz, hydrogen peroxide; K,Fe(CN)s, potassium ferricyanide; LDH, lactate dehydrogenase; 0; , superoxide; PROS, partially reduced oxygen species; TCA, trichloroacetic acid; XDH + X0, xanthine dehydrogenase + xanthine oxidase; % X0, percentage of total XDH + X0 activity in the oxidase form; HBSS, Hanks’ balanced salt solution; ANOVA, analysis of variance; RGS, Ryan’s growth supplement. 695
696
PANUS
dietary tungsten, an inactive analog of molybdenum which inactivates XDH and X0 (12), often attenuates tissue injury (7,13, 14). Evidence gained from inhibitors of X0 supports a role for this source of 0, and Hz02 in tissue reperfusion injury to liver, kidney, intestine, and lung. In the lung, X0-derived PROS may also mediate pulmonary damage during hyperoxia as well as during hypoxia (2, 14-17). Perfusion of isolated rat lungs from hyperoxic-exposed animals showed enhanced edematous injury, which was attenuated by pretreatment with dietary tungsten, suggesting a role for X0 in hyperoxia-induced lung edema formation and injury (16). Reduced lung X0 activity and serum uric acid was observed after exposure of rats to >98% O2 for 48 h, suggesting oxidant-mediated inactivation of X0 by PROS of undefined origin (15). Also, hyperoxic exposure of rats and cultured bovine pulmonary artery endothelium resulted in the loss of XDH and X0 activity (17). Morphologic examination of lungs from rats exposed to hyperoxia revealed damage to pulmonary endothelium (18), suggesting these cells are key sources of PROS during oxidant stress or are targets of PROS derived from other intravascular sources. It has been proposed that endothelial X0 was a source of PROS which contributed to pulmonary injury and the oxidative inactivation of XDH and X0; however, XOderived oxidants were not causally related to hyperoxic damage (15-17). Herein, changes in endothelial cell XDH and X0 activity during control and hyperoxic conditions were assessed, and the contribution of endothelial cell X0 to oxygen-mediated cell injury was defined. EXPERIMENTAL
PROCEDURES
Endothelial cell culture. Cells were isolated from bovine aorta and cultured in bicarbonate buffered Medium 199 (M-199, GIBCO), supplemented with 5% fetal bovine serum (Hyclone), 5% Ryan’s growth supplement (Dr. Una Ryan, Monsanto Inc., St. Louis, MO), and 1% antibiotics/antimycotic (GIBCO). Cells were grown in T-75 flasks (Costar, Cambridge, MA), kept humidified (37°C) in 95% air, 5% CO2 prior to experimentation, and used between passages 6 to 10. Unless stated otherwise cells were subcultured in a 1:3 ratio, were confluent within 6 to 8 days and medium was replaced every 4 to 7 days after subculturing. Calf serum was obtained from Grand Island Biological Co. (Grand Island, NJ), and fetal calf serum and iron-supplemented calf serum were obtained from Hyclone (Logan, UT). 0, consumption. BAEC were subcultured in a 1:4 ratio and 7 days later total and cytochrome c oxidase independent cell Ox consumption was measured in confluent cells removed from flasks by scraping. Cells were centrifuged at 25Og, 10 min, 4”C, and resuspended in Hanks’ balanced salt solution (HBSS, GIBCO), with 1 mM N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid and 12 mM glucose, pH 7.4. Oxygen consumption was measured in a polarograph with a Clark electrode. Potassium cyanide (25-200 PM), myxothiazole (0.007-0.5 PM), and antimycin A (0.0033-1.0 PM) dose-response studies were performed to determine optimum concentrations for assessing cytochrome c oxidase independent 0s consumption. Hyperoxic exposure. Fresh media was added to cells 30 min prior to O-48 h of hyperoxic (95% 0s. 5% COz) or normoxic (95% air, 5% CO,) exposure in a humidified environment (37’C). At termination of experiments, media was saved for analysis, 5 ml of cold (4°C) HBSS was added to cells, monolayers were scraped and centrifuged (2OOg, 10 min,
ET AL. 4’C), and pellets were resuspended in 1 ml lysing buffer containing 10 mM dithiothreitol and antiproteases intended to inhibit artifactual XDH to X0 conversion (19). Cells were then sonicated (75 W, 10 s), rapidly frozen in liquid nitrogen or -86°C isopentane, and stored at -86°C until analysis. Biochemical analyses. Several markers of cytotoxicity were utilized. Detached cells, assumed to be nonviable, were counted with a Coulter Counter (Model ZM). Release of preincorporated [8-“CJadenine (New England Nuclear) was also utilized as a marker of cytotoxicity (20,21). Cells were cultured 36 h with [8-“Cladenine (4 PM) and washed twice with fresh media prior to initiation of exposures. Aliquots of cell lysates and centrifuged culture media (2OOg, 10 min, 4’C) were analyzed for radioactivity at the end of experiments. Other aliquots of lysates and media from the same cell samples were added to 20% trichloroacetic acid (TCA) 1:l (v/v) and centrifuged (SSOOg,2 min, 25°C). The centrifuged supernatant was assayed for radioactivity to determine the proportion of acid-soluble incorporated counts. Cell and media-associated DNA and LDH activity were also determined (20,22). Preincorporated [8-“Cladenine, LDH, and DNA were reported as a percentage of total original activity in monolayers which was released into the media: % release = [(media activity)/(media
activity + cellular
activity)]
X 100%.
Catalase activity was determined spectrophotometrically (23) and protein was determined by dye binding (24). Cellular XDH and X0 activities were determined as previously described (25) by allopurinol inhibitable oxidation of pterin to fluorescent isoxanthopterin (&.. = 345 nm, X, = 390 nm). Methylene blue was substituted for NAD as an electron acceptor when measuring total XDH + X0 activity. Sample quenching was adjusted for by adding known amounts of isoxanthopterin as an internal standard. The percentage of total XDH + X0 activity in the oxidase form (% X0) was determined by dividing the X0 activity by the total XDH + X0 activity. To determine the stability of LDH and XDH + X0 activity in media, freshly prepared complete culture medium was exposed to hyperoxia and normoxia in the absence of cells. Allopurind and BAEC XDH + X0 activity. In some experiments cells were preincubated with 100 PM allopurinol for 1 h in complete media, prior to exposure to O-48 h of normoxia or hyperoxia. The cells were subsequently harvested as previously described. Nonadhered cells were measured and lysate XDH + X0 activity, protein, and DNA determinations done. Aliquots of lysates were incubated for 10 min with 10 mM potassium ferricyanide (KeFe(CN)a) to reverse binding of the allopurinol metabolite oxypurinol via oxidation of the enzyme molybdop&in center (26). Cell lysates were then chromatographed on Sephadex G-25M PD-10 columns (Pharmacia) and the first 1.0 ml after the void volume was collected and assayed for XDH + X0 activity and protein. In allopurinol inhibition reversal experiments, the cell lysing buffer was modified by decreasing dithiothreitol from 10 to 1 mM; because 10 mM, but not 1 mM, dithiothreitol completely reduced K,Fe(CN), and prohibited reversal of allopurinol inhibition. Statistical analysis. Unless otherwise stated, all data are expressed as mean + standard deviation (X t- SD). Statistical analysis and modeling were performed as previously described (20). Actual data analysis was done using “SAS System, release 6.02.” Data were analyzed via a twoway analysis of variance (ANOVA) and Duncan post hoc groupwise comparison to account for experiment-to-experiment variation within a group, and significance between groups was determined for dependent variables. Significance for ANOVA and Duncan testing was set at For a ~0.05, unless otherwise stated.
RESULTS Cell culture, XDH + X0 actiuity. The XDH + X0 activity of cells in normoxia varied with time after sub-
ENDOGENOUS
z iii s
60
E
50
r
40
XANTHINE
OXIDASE-MEDIATED
ENDOTHELIAL
TABLE
ii
30
+
20
g
10
3 a
0
697
INJURY I
BAEC O2 Consumption in the Absence and Presence of Mitochondrial Respiratory Chain Inhibitors nmol Oz. mini protein’
Variable
22.0 0.4 1.0 6.3
Control Myxothiazole (0.333 pM) Antimycin A (0.033 FM) KCN (200 /.iM)
3
5 9 10 15 24 30 DAYS POST SUBCULTURE
FIG. 1. Changes in total cellular XDH + X0 activity in BAEC during normoxic culture. Total cell XDH + X0 activity was normalized for cell protein (au XDH + XO/mg protein). Peak cell XDH + X0 activity on Day 9 was significantly greater (*) than the activity on any of the other recorded days. The histogram represents the culmination of two separate experiments, in which the cells were subcultured 1 to 4.
culturing (Fig. 1). Specific activity of XDH + X0 increased in preconfluent cells during the first 3 to 7 days after subculturing. The BAEC were 95100% confluent at Day 7, and peak XDH + X0 activity was significantly greater at Day 9 than at any other time measured. Subsequently, the XDH + X0 activity declined in postconfluent cells, even when media was regularly changed, ruling out depletion of key nutrients as a contributory factor. Cell morphology observed by phase contrast microscopy revealed an unchanged cobblestone appearance in cells during the entire study period. Over 30 days, the percentage XDH + X0 in the oxidase form varied, with no consistent pattern. The percentage oxidase at peak specific activity (Day 9 postsubculture) was 53 +_4%. In subsequent studies, cells with maximal XDH + X0 activity (7 to 10 days postsubculture) were utilized. Endothelial cell 0, consumption. For calculation of the contribution of X0 to cell cytochrome c oxidase independent oxygen consumption (Table I), preliminary doseresponse studies of BAEC respiratory chain activity inhibition gave optimum inhibitor concentrations. Myxothiazole (0.333 PM), antimycin A (0.033 PM), and KCN (200 PM) decreased O2 consumption by 98,95, and 71%, respectively. The KCN-resistant O2 consumption was significantly greater than myxothiazole and antimycin Aresistant respiration. Hyperoxia and cellular XDH and X0. BAEC were exposed to O-48 h of normoxia or hyperoxia. Phase contrast microscopy demonstrated no significant differences between hyperoxic and control cell morphology after 48 h of exposure (data not shown), nor was there any significant change in cell monolayer DNA (Table II). Cell-associated protein similarly increased in both air- and 02exposed cells (Table II). Total cell XDH + X0 activity decreased in BAEC exposed to both normoxia and hyperoxia (Fig. 2), with the
* mg n
-+ 6.0 It 0.2’s** + 0.6*,** + 2.2’
16 4 4 8
Note. Inhibitor concentrations represent the minimum utilized, which resulted in maximal inhibition of O2 consumption. * Different from control. ** Different from the KCN group.
loss of XDH + X0 activity in hyperoxia significantly greater than air controls. After 48 h of hyperoxia, only two of nine flasks had detectable cellular XDH + X0 activity. The percentage of total cellular XDH + X0 activity in the X0 form (% X0) increased in Os-exposed BAEC and was significantly different at 48 h from air controls, which remained constant (Fig. 2). While the % X0 in the 48-h hyperoxic-exposed BAEC was greater than controls, X0 activity in hyperoxic cells (0.60 + 1.5 pIJ/ mg protein) was less than air controls (5.4 t 4.2 pU/mg protein). In these experiments, 10 mM dithiothreitol was added to the lysing buffer, to prevent artifactual XDHto-X0 conversion. It was observed that dithiothreitol (O50 mM, 30 min, 37°C) did not alter the percentage of total cellular XDH + X0 activity in the X0 form (data not shown). Thus, changes in X0 activity were not due to inadvertent sulfhydryl oxidation of XDH or disulfide reduction of X0 prior to biochemical analysis. Extracellular release of XDH + X0 activity into the media of 48-h oxygen-exposed BAEC was not detectable. Thus, if XDH and/or X0 was released from BAEC exposed to hyperoxia, enzymatic activity was too dilute in
TABLE
II
Alterations in BAEC DNA and Protein during Exposure to Normoxia or Hyperoxia DNA (wdcm*) Exposure time (h)
Air
0 12 24 48
N.D. N.D. N.D. 2.1 + 0.5
Protein
95% 02 2.0 2.0 2.0 2.0
F + * +
0.5 0.4 0.4 0.4
Air N.D. N.D. N.D. 14 + 1*
(pg/cm*) 95% 02 11 13 13 13
Itr 2 + 2* f 1* rt 1*
Note. Changes in cellular DNA and protein were normalized for surface area (cm*) of the flask. Values represent three separate experiments done in triplicate, and control values at 12 and 24 h were not determined (N.D.). * Different from to control.
698
PANUS
ET AL. TABLE
III
Alterations in Cell Catalase and Lactate Dehydrogenase (LDH) Activities during Normoxia and Hyperoxia
100 00 t 60
3
40
8
20
s
Catalase (U/mg protein) Exposure time (h)
Air
0 12 24 48
N.D. N.D. N.D. 13 If: 3
0
95% 0, (HRS) FIG. 2. Alterations in BAEC XDH and X0 activity during hyperoxic and normoxic exposures. Changes in total cellular XDH + X0 activity were normalized for protein (pU XDH + XO/mg protein), whereas (% OXIDASE) was the amount of cellular XDH/XO activity in the X0 form. The closed circles and closed triangles represent, respectively, the total cell XDH + X0 activity and the % oxidase, in the hyperoxicexposed cells, whereas the open circle and open triangle represent similar 48-h normoxic control values. (*) and (+) denote when different from to and t controls, respectively. Enzymatic activity was determined from BAEC groups previously described in Table II. Values represent three separate experiments done in triplicate (n = 9).
media to be detectable or the enzyme was inactivated. There was X0 activity in culture media used herein (1.53 f 0.65 PU XO/ml, n = 6, with no detectable XDH activity). The X0 activity in the culture medium exposed to hyperoxia and normoxia, in the absence of cells, decreased to undetectable levels during the first 12 h of hyperoxia and after 48 h of normoxia. The source of this X0 activity was Ryan’s growth supplement (65 + 14 pU/ml, n = 6), with activity as high as 113 PU XO/ml observed. Ironsupplemented bovine serum is a major component of RGS, having 213 to 360 PU XO/ml when analyzed from several different lots and sources, and no detectable XDH. In contrast, several different lots of fetal bovine serum showed no detectable XDH or X0 activity. The loss of X0 activity in the cell-free complete medium may be due to autoinactivation by X0-derived PROS. The serumsupplemented cell-free medium contained 6 PM hypoxanthine and 4 PM xanthine, providing sufficient substrate for 0; and HzOz formation. These purine substrates were derived from both the basal medium and serum supplements. Hyperoria and indicators of cell injury. Endothelial cell lactate dehydrogenase activity (per milligram protein) was marginally increased (a < 0.10) after 12 and 24 h hyperoxia when compared to both 0- and 48-h controls (Table III), but returned toward control values after 48 h hyperoxia. Two-way analysis of variance demonstrated significant differences in LDH activities between the various endothelial cell lines utilized for the present experiments. However, the LDH activity within a given cell line was constant at 0 and 48 h normoxia. LDH activity
LDH (mU/mg protein)
95% 02 12 12 12 13
+ + + f
Air
2 3 2 2
N.D. N.D. N.D. 681 f 413
95% 02 587 826 819 767
-t + f f
411 638**** 59a**** 459*
Note. Enzymatic catalase and LDH activities were normalized for cell protein and were determined from the BAEC groups in Table II. Values represent two separate experiments done in triplicate (n = 6). * As previously described in Table II. ** Different (P < 0.10) from t*s control.
in the medium also increased during hyperoxia (Table IV). The media initially contained LDH activity (11 + 0.2 mU/cm2; n = 3) derived from serum supplements. In the absence of cells, exposure of the complete medium to hyperoxia or normoxia for 48 h resulted in a loss of medium LDH activity. Loss of LDH activity in the medium under cell-free conditions during hyperoxia was marginally greater ((u < 0.10) than that in the corresponding media maintained in normoxia. Thus, LDH activity released into the media was normalized for serum-derived LDH activity prior to calculation of release from cells. The LDH released by BAEC during hyperoxia was greater than controls (Table IV), but the percentage of total LDH released by cells into the medium did not increase, due
TABLE
IV
Effects of Normoxia and Hyperoxia and Medium LDH Activity
Time (h)
Exposure
0 12 24 48 48
Air 95% 95% 95% Air
LDH released (mu/cm’) 1.8 2.9 3.3 3.6 1.7
+ f + ? +
1.3 0.6 1.0*,** 1.4*~** 0.7
Cell + LDH released (mu/cm’)
14 14 14 10
9.5 + 9*,** k 8*,** + 8*,** t 5
on Cell
% LDH released (mu/cm’) 27 32 32 29 21
f 20 2121 f 18 f 12 k 14
Note. BAEC were exposed to O-48 h of hyperoxia (95%) or normoxia (Air). Medium LDH activity from cells (column 3) was corrected for LDH activity from medium serum supplements. Medium LDH activity varied with time and was determined in similar flasks under cell-free conditions (to = 11 f 2, ti2 = 10 + 0.8*, tul = 10 f l*, r& = 9 + l*, and control trs = 10 * 2* mu/cm’), representing two separate experiments, n = 5. LDH activities were determined from BAEC reported in Table III, representing two experiments done in triplicate, n = 6. * Different from to control. ** Different from taacontrol.
ENDOGENOIJS
XANTHINE
OXIDASE-MEDIATED
ENDOTHELIAL
699
INJURY
l
P4 g
20
3
16 1P
95% 0, (HRS) FIG. 3. Cell damage during 48-h BAEC hyperoxic exposure. Cell damage (detached cells) is represented as the number of detached cells normalized for surface area of the flask (cm’). The (0) and (A) represent the number of detached cells in BAEC exposed to hyperoxia and normoxia, respectively. (*) and (+) denote when different from to and t& controls, respectively. Cell damage was determined from the BAEC groups previously described in Table II. Values represent two separate experiments done in triplicate (n = 6) (X -t SD).
to the increase in cell LDH during exposure to hyperoxia. When cell LDH activity and that released from cells into the media were added, the total LDH activity in BAEC exposed to hyperoxia was significantly greater than air controls (Table IV). Thus, the increase in cell LDH in hyperoxic BAEC masked the increase in LDH released from cells into the media (Table IV). Release of other markers of cellular damage during the 48-h hyperoxic [8-14C]adenine, exposure, DNA and preincorporated were not significantly elevated when compared to the 48-h normoxia controls. Only detached cell numbers increased significantly in cells exposed to hyperoxia for 48 h (Fig. 3). Allopurinol effects on cell XDH and X0 activity and hyperoxic injury. Loss of cell XDH + X0 activity and XDH-to-X0 conversion paralleled cell damage as indicated by increased numbers of detached cells. Therefore, subsequent experiments examined whether changes in XDH and X0 activity were responsible for cell damage during hyperoxia. BAEC were preincubated with 100 PM allopurinol 1 h prior to exposure to 48 h of normoxia or hyperoxia (Fig. 4; Table V). Allopurinol completely inhibited BAEC XDH + X0 activity and was not cytotoxic, as indicated by an absence of increased cell detachment. Treatment of air-exposed allopurinol-treated t = 0- and t = 48-h cell lysates with K,Fe(CN), increased detectable XDH + X0 activity by 0 and 214%, respectively (Fig. 4). After 48 h air exposure of cells not treated with allopurinol, XDH + X0 activity assayed with or without enzyme oxidation by K,Fe(CN), was reduced by 50%. Hyperoxic exposure for 48 h reduced XDH + X0 activity to undetectable levels. This complete loss of XDH + X0 activity in hyperoxic cells was not prevented by maintaining cells
6
1
4
0
46 AIR
7I
FIG. 4. The BAEC were pretreated in complete media + 100 pM allopurinol for 1 h and subsequently subjected to O-48 h of air (CTL) or hyperoxia (Ox). Total XDH + X0 activity was assayed in both crude cell lysate (m) or after the lysate was treated with K,Fe(CN), and the void volume collected after gel filtration (I@. (*) and (+) denote when values are different from to and ta controls, respectively, without allopurinol. Figure represents two separate experiments done in triplicate, n = 6.
in allopurinol and was not reversed by enzyme reoxidation with K,Fe(CN),, in contrast to control cells maintained in air. In cells maintained in normoxia, prior inhibition of XDH + X0 with allopurinol resulted in a threefold increase in XDH + X0 specific activity after K,Fe(CN), treatment, when compared to untreated 48-h air control cells. Hyperoxic exposure increased cell detachment from the monolayer (Fig. 3 and Table V). Allopurinol inhibition of XDH and X0 caused a minor but significant reduction in cell detachment in monolayers exposed to air for 48 h and was not protective for oxygen-exposed cells (Table V).
TABLE Effects
of Hyperoxia
V
and Allopurinol
on Cell Injury
Detached cell number Time (h) Air Air Hyperoxia
0 48 48
Control 203 + 205 814 _fr 307*
1109 + 287*,*+*
Allopurinol 113 2 120 504 + 229*,**
1216 + 264*,**
Note. Table represents cell damage (detached cells) from BAEC described in Fig. 4. Table represents two separate experiments done in triplicate, n = 6. * Different from to control without allopurinol. ** Different from t@ control without allopurinol. *** Different from the t48 control without allupurinol (a 5 0.10).
700
PANUS
DISCUSSION
The role of xanthine oxidase in oxygen-mediated injury to cultured vascular endothelium was examined because numerous investigations have suggested an important role for X0-derived oxidants in tissue damage occurring secondary to anoxia-reoxygenation, hyperoxia, and neutrophi1 activation (2,3,8,14,17,27). Many of these processes resulted in endothelial damage and increased vascular permeability (14,16, 28). Pretreatment with antioxidant enzymes or compounds which inhibited XDH and X0 activity sometimes prevented endothelial damage and attenuated vascular permeability increases (14,16,28, 29). Formation of the potentially cytotoxic oxidase from XDH may be mediated by a variety of processes, including complement activation, partial proteolysis of XDH by intracellular or neutrophil-derived enzymes, increased sulfhydryl oxidation and glutathione mixed disulfide formation with XDH (2, 8, 27, 30). More extensive oxidation of XDH thiols to sulfenic, sulfinic, or sulfonic acid derivatives by 0, and HzOz (31) may also be expected to yield X0. We compared the effect of hyperoxia on cell LDH, catalase, and XDH + X0 activity. Cell LDH activity increased marginally during the first 24 h of hyperoxic exposure, possibly due to increased glycolytic activity in response to mitochondrial injury (32). The subsequent decrease in cellular LDH activity at 48 h may be due to the predominance of LDH inactivation by PROS (20). It has been previously observed that catalase activity decreases significantly in porcine endothelium exposed to hyperoxia for 48 h (33). A similar oxidant stress to cultured bovine aortic endothelium herein showed catalase activity to be unaffected by hyperoxia. Previous investigations have utilized both detached cells and cellular release of LDH as markers of injury to endothelial cells exposed to 24 h or more of chronic oxidant stress (34,35) with cell detachment a more sensitive marker of cell damage in comparison to LDH release (35). Herein we examined the effects of 48 h of hyperoxia on both cell detachment and cellular release of LDH. Detached cell numbers and LDH release from monolayers increased significantly during the 48-h oxygen exposure to hyperoxia. Release of the cytotoxicity markers, preincorporated [8-14C]adenine and cell DNA, were not significantly elevated in the media when compared to controls (data not shown). The assessment of LDH release from oxidant-injured cells is complicated by a number of underlying phenomena. For example, cell culture medium serum supplements contain not only LDH, but also xanthine oxidase. In cell-free conditions, hyperoxia caused loss of endogenous serum-derived LDH activity, possibly mediated by autooxidation of medium components, since it is known that LDH can be inactivated by PROS (20, 36). Another problem with using LDH as a cytotoxicity marker relates to the increasing endogenous LDH specific
ET AL.
activity of stressed cells, which may not similarly occur under control conditions. We conclude that the instability of cell and medium LDH and its susceptibility to oxidant inactivation make this enzyme an unreliable marker which can lead to misinterpretation of oxidant cell injury. This concept extends to animal models of ischemia-reperfusion injury, where similar oxidant inactivation of creatinine phosphokinase may lead to complications in assessment of extents of myocardial injury and antioxidant protection (37, 38). In our studies, endothelial cell XDH + X0 activity increased after subculture of normoxic cells until confluency and then declined. The cause of this change in activity could not be ascribed to cell injury or inhibition of protein synthesis and warrants further investigation. Allopurinol-mediated inhibition of XDH and X0 depends on oxidation of allopurinol to oxypurinol, which binds to a partially reduced form of the enzyme having the molybdenum trapped at the Mo(IV) level (26). Reversal of allopurinol inhibition of control cell XDH and X0 for 48 h by enzyme oxidation with K,Fe(CN), resulted in recovery of more XDH and X0 activity than was present in parallel cultures of cells maintained in the absence of allopurinol for 48 h. This suggests that X0 may contribute to its own inactivation during normoxia or that allopurinol is having other unrecognized effects on cell metabolism. We observed a general loss of XDH and X0 activity and conversion of XDH to X0 in the hyperoxia-exposed cells, which was not reversible by dithiothreitol. This suggests that the increased cellular X0 activity during hyperoxia was either due to dithiothreitol-resistant sulfhydryl oxidation or partial proteolysis. For cells exposed to 48 h of hyperoxia, there was an almost complete loss of cell XDH and X0 activity, which occurred concurrently with a pathophysiologically insignificant conversion of minimal residual XDH activity to X0. These observations differ from previously reported studies of the contribution of endothelial X0 to free radical injury. First, the percentage XDH + X0 in the oxidase form in our control cells remained consistently near 45%, previously reported to be 9% for bovine pulmonary endothelial X0 (17). The percentage oxidase activity of control endothelial cell XDH reported by these investigators is lower than the partial oxidase activity of pure rat liver XDH or flavin-reconstituted purified XDH (2, 39), raising concern about assay methodology. Also, endothelial cells have been reported to have an XO-dependent extracellular 0; release of -0.5 nM * lo6 cells-‘. 4 h-‘. Assuming 4 X lo6 cells per T-25 flask incubated with 1 ml of cytochrome c solution, this reported rate of cell 0; release would give an undetectable A A550-0.00004 over 4 h (16), even if tungsten were not having other unrecognized effects on cell metabolism. XDH + X0 activity varies between tissues of the same species and between the same tissue from different species. Cultured rat endothelial cells have the greatest XDH
ENDOGENOUS
XANTHINE
OXIDASE-MEDIATED
+ X0 specific activity (27, 30), followed by bovine endothelium (16, 17) (Fig. 4) and finally cultured porcine endothelium, where XDH + X0 activity is undetectable (unpublished observations). At present, XDH + X0 activity has not been reported in cultured human endothelium. The significance of tissue X0 as a source of partially reduced O2 formation during basal or pathologic cell conditions depends upon substrate availability, enzyme specific activity, and the relative contribution of other cellular sources of 0, and HzOz. Myxothiazole-resistant O2 consumption in our cultured bovine endothelial cells was 2.0 nmol * min-’ * mg DNA-‘. Calculation of cytochrome c oxidase independent O2 consumption was based upon myxothiazole-resistant respiration measurements, because this inhibitor does not artifactually enhance complex I and II 0, production, as does antimycin A and CN(40). Additionally, CN- inhibits a portion of cytochrome b-derived 0, production (40) and reacts with diverse cellular hemeproteins and antioxidant enzymes, thus complicating measurement of cytochrome c oxidase independent O:! consumption. Assuming complete reduction of 0, and Hz02 to H,O by antioxidant enzymes during myxothiazole-resistant O2 consumption, a maximum of 8 nmol OZ. rnin~l. mg DNA-’ could have been reduced to 0, or HzOz by cells. This prediction of cell 0; and HaOz production is likely an overestimate, because covalent incorporation of O2 into substrates such as arachidonic acid, which does not include 0;) HzOz, or . OH as intermediates, is simultaneously occurring in endothelial cells. Exposure of cells to 95% O2 should increase cell PROS formation two- to threefold (41-43). The rate of reduction of O2 to 0, and H202 by X0 in these cells would maximally have been 0.29 nmol . mini’ . mg DNA-‘. With the K, of X0 for O2 being 46 KM (44) and assuming that no O2 gradients existed between the media/air interface and the cell surface (45), the increase in X0 activity under saturating O2 conditions would only be 20%, increasing X0-dependent O2 reduction to 0.35 nmol . min-’ . mg DNA-‘. In summary, the maximal contribution of XOderived PROS to the overall cell formation of PROS would be 3.6% during normoxia and less during hyperoxia. Further, the loss of X0 during hyperoxia would diminish the relative contribution of oxidants from this source with time. Thus, other cell sources contribute more substantially to overall cell PROS formation during hyperoxia and can include mitochondria, endoplasmic reticulum, nuclear membranes, and soluble flavoproteins or hemeproteins (41-43,46). These predictive calculations, while only approximate, suggest that X0-derived PROS do not contribute significantly to overall cell PROS formation during hyperoxia and are supported by experimental observations herein as well as by those of Royal1 et al. (47). This does not rule out a potential pathophysiologic role for X0 under other circumstances, however. In summary, hyperoxic endothelial cells lost XDH and X0 activity concurrent with an XDH-to-X0 conversion.
ENDOTHELIAL
INJURY
701
Endothelial cell XDH + X0 activity decreased with early hyperoxic cell damage, but had no causal relationship with hyperoxic cell injury. ACKNOWLEDGMENTS This research was supported by grants from the National and Alabama Affiliate of the American Heart Association, NIH Grant NS24275, and the Council for Tobacco Research. P. C. Panus was supported in part by National Heart, Lung, and Blood Training Grant T32-HL-07553. We also wish to acknowledge the assistance of Joan Wheat and Dr. Dale Parks in HPLC analyses.
REFERENCES 1. Parks, D. A., and Granger, D. N. (1986) Acta Physiol. Stand. 548, 87-99.
2. Engerson, T. D., McKelvey,
T. G., Rhyne, D. B., Boggio, E. B., Snyder, S. J., and Jones, H. P. (1987) J. Clin. Inuest. 79, 1564-
1570. 3. Parks, D. A., Williams,
T. K., and Beckman, J. S. (1988) Am. J. Physiol. 254, G768-G774. 4. Krenitsky, T. A., Spector, T., and Hall, W. W. (1986) Arch. Biochem. Biophys. 247, 108-119. 5. McCord, J. M., and Fridovich, I. (1968) J. Biol. Chem. 243, 5753-
5760. 6. Olson, J. S., Ballow, D. P., Palmer, G., and Massey, V. (1974) J. Biol. Chem. 249, 4350-4362.
7. Parks, D. A., Bulkley,
G. B., Granger, D. N., Hamilton, S. R., and McCord, J. M. (1982) Gastroenterology 82, 9-15. 8. McKelvey, T. G., Hollwarth, M. E., Granger, D. N., Engerson, T. D., Landler, U., and Jones, H. P. (1988) Am. J. Physiol. 264, G753-G760. 9. McCord, J. M. (1985) N. Engl. J. Med. 312, 159-163. 10. Waud, W. R., and Rajagopalan, K. V. (1976) Arch. Biochem. Biophys.
172,365-379. 11. Spector, T. (1977) Biochem. Pharmacol. 26, 355-358. 12. Johnson, J. L., Rajagopalan, K. V., and Cohen, H. .J. (1974) J. Biol. Chem. 249,859-866. 13. Granger, D. N. (1988) Am. J. Physiol. 255, H1269-H1275. 14. Grosso, M. A., Brown, J. M., Viders, D. E., Mulvin, D. W., Banejee, A., Velasco, S. E., Repine, J. E., and Harken, A. H. (1989) J. Surg. Res. 46, 355-360. 15. Elsayed, N. M., and Tierney, D. F. (1989) Arch. Biochem. Biophys.
273,281-286. 16. Rodell, T. C., Cheronis, J. C., Ohnemus, C. L., Piermattei, D. J., and Repine, J. E. (1987) J. Appl. Physiol. 63, 2159-2163. 17. Terada, L. S., Beehler, C. J., Banerjee, A., Brown, J. M., Grosso, M. A., Harken, A. H., McCord, J. M., and Repine, J. E. (1988) J. Appl. Physiol. 65, 2349-2353.
18. Crapo, J. D., Barry, B. E., Fescue, H. A., and Shelburne, J. (1980) Am. Rev. Respir. Dis. 122, 123-143. 19. Panus, P. C., Shearer, J., and Freeman, B. A. (1988) Elcp. Lung Res.
14,959-976. 20. Panus, P. C., Matalon, 21.
22. 23. 24. 25.
S., and Freeman, B. A. (1989) In Vitro Cell Dew. Biol. 25, 821-829. Shirhatti, V., and Krishna, G. (1985) Anal. Biochem. 147,410-418. Labarca, C., and Paigen, K. (1980) Anal. Biochem. 102,3443-3452. Beutler, E. (1975) in Red Cell Metabolism: A Manual of Biochemical Methods, pp. 71-75, Grune & Stratton, New York. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. Beckman, J. S., Parks, D. A., Pearson, J. D., Marshall, P. A., and Freeman, B. A. (1989) Free Radicals Biol. Med. 6, 607-615.
702
PANUS
26. Massey, V., Komai, H., Palmer, G., and Elion, G. B. (1970) J. Biol. Chem. 245,2837-2844. 27. Phan, S. H., Gannon, D. E., Varani, J., Ryan, U. S., and Ward, P. A. (1989) Am. J. Pathol. 134, 1201-1211. 28. Grosso, M. A., Viders, D. E., Brown, J. M., Mulvin, D. W., Miles, R. H., Brentlinger, E. R., Velasco, S. E., Crawford, T. S., Burton, L. K., Repine, J. E., and Harken, A. H. (1989) Surgery 106, 310317. 29. Inauen, W., Payne, D. K., Kvietys, P. R., and Granger, D. N. (1990) Free Radicals Biol. Med. 9, 219-223. 30. FriedI, H. P., Till, G. O., Ryan, U. S., and Ward, P. A. (1989) FASEB J. 3,2512-2518. 31. Radi, R., Bush, K. M., Cosgrove, T. P., and Freeman, B. A. (1991) Arch. Biochem. Biophys. 286, 117-125. 32. Simon, L. M., R&in, T. A., Douglas, W. H., Theodore, J., and Robin, E. D. (1979) J. Appl. Physiol. 47,98-103. 33. Block, E. R., Patel, J. M., and Sheridan, N. P. (1985) J. CeU. Physiol.
122,240-248. 34. Bowman, C. M., Butler, E. N., and Repine, J. E. (1983) Am. Reu. Respir. Dis. 128,469-472. 35. Harlan, J. M., Harker, L. A., Reidy, M. A., Gajdusek, C. M., Schwartz, S. M., and Striker, G. A. (1983) Lab. Znuest. 48, 269-274. 36. Buchanan, J. D., and Armstrong, D. A. (1976) Znt. J. Radiat. Biol. 30,115-127.
ET AL. 37. McCord, J. M., and Russell, W. J. (1988) in Oxy-Radicals lecular Biology and Pathology (Cerutti, P. A., Fridovich, McCord, J. M., Eds.), pp. 27-35, Alan R. Liss, New York.
in MoI., and
38. Omar, B. A., Gad, N. M., Jordan, M. C., Striplin, S. P., Russell, W. J., Downey, J. M., and McCord, J. M. (1990) Free Radicals Biol. Med. 9,465-471. 39. Saito, T., Nishino, 15,390-15,395. 40. Turrens, B&hem.
T., and Massay, V. (1989) J. Biol. Chem. 264,
J. F., Alexandre, A., and Lehninger, Biophys. 237, 408-414.
A. L. (1985) Arch.
41. Turrens, J. F., Freeman, B. A., Levitt, J. G., and Crapo, J. D. (1982) Arch. Biochem. Biophys. 217,401-410. 42. Turrens, J. F., Freeman, B. A., and Crapo, J. D. (1982) Arch. Biochem. Biophys. 217, 411-421. 43. Yusa, T., Crapo, J. D., and Freeman, B. A. (1984) B&him. Acta 798,167-174.
Biophys.
44. Saito, T., and Nishino, T. (1989) J. Biol. Chem. 264,10,015-10,022. 45. Dickman, K. G., and Mandel, c333-c340. 46. Misra, 6965.
H. P., and Fridovich,
L. J. (1989) Am. J. Physiol.
267,
I. (1972) J. Biol. Chem. 247,
6960-
47. Royall, J. A., Gwin, P. D., Parks, D. A., and Freeman, B. A. (1992) Arch. Biochem. Biophys. 294,686-694.