HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 290/3/L607    most recent 00302.2005v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Merker, M. P. Articles by Krenz, G. S. Search for Related Content PubMed PubMed Citation Articles by Merker, M. P. Articles by Krenz, G. S.

Influence of pulmonary arterial endothelial cells on quinone redox status: effect of hyperoxia-induced NAD(P)H:quinone oxidoreductase 1

Departments of ¹Anesthesiology, ²Pharmacology and Toxicology, and ³Pulmonary Medicine, Medical College of Wisconsin, Milwaukee; Departments of ⁴Biomedical Engineering and ⁵Mathematics, Statistics and Computer Science, Marquette University, Milwaukee; and ⁶Zablocki Veterans Affairs Medical Center, Milwaukee, Wisconsin

Submitted 13 July 2005 ; accepted in final form 19 October 2005

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
The objective of this study was to examine the impact of chronic hyperoxic exposure (95% O₂ for 48 h) on intact bovine pulmonary arterial endothelial cell redox metabolism of 2,3,5,6-tetramethyl-1,4-benzoquinone (duroquinone, DQ). DQ or durohydroquinone (DQH₂) was added to normoxic or hyperoxia-exposed cells in air-saturated medium, and the medium DQ concentrations were measured over 30 min. DQ disappeared from the medium when DQ was added and appeared in the medium when DQH₂ was added, such that after 15 min, a steady-state DQ concentration was approached that was 4.5 times lower for the hyperoxia-exposed than the normoxic cells. The rate of DQ-mediated reduction of the cell membrane-impermeant redox indicator, potassium ferricyanide [Fe(CN)], was also approximately twofold faster for the hyperoxia-exposed cells. Inhibitor studies and mathematical modeling suggested that in both normoxic and hyperoxia-exposed cells, NAD(P)H:quinone oxidoreductase 1 (NQO1) was the dominant DQ reductase and mitochondrial electron transport complex III the dominant DQH₂ oxidase involved and that the difference between the net effects of the cells on DQ redox status could be attributed primarily to a twofold increase in the maximum NQO1-mediated DQ reduction rate in the hyperoxia-exposed cells. Accordingly, NQO1 protein and total activity were higher in hyperoxia-exposed than normoxic cell cytosolic fractions. One outcome for hyperoxia-exposed cells was enhanced protection from cell-mediated DQ redox cycling. This study demonstrates that exposure to chronic hyperoxia increases the capacity of pulmonary arterial endothelial cells to reduce DQ to DQH₂ via a hyperoxia-induced increase in NQO1 protein and total activity.

duroquinone; pulmonary endothelial cells; mathematical modeling

Of the quinone reductases identified in various cell types, including pulmonary endothelial cells, among the most thoroughly studied has been NAD(P)H:quinone oxidoreductase 1 (NQO1) (3, 12, 32, 37, 44, 57, 64). NQO1 is a cytoplasmic NAD(P)H quinone oxidoreductase that carries out two-electron quinone reduction, at least one function of which is to decrease quinone availability to one-electron quinone reductases, thereby avoiding semiquinone toxicity resulting from redox cycling and/or covalent modification of cellular macromolecules (12, 14, 22, 23, 36, 38). Studies carried out in cultured pulmonary endothelial cells and the isolated intact perfused lung established a role for NQO1 in generation of the two-electron reduction product durohydroquinone (DQH₂) when DQ was added to the extracellular medium or lung perfusate (2, 3, 44). The amount of DQH₂ produced was limited primarily by an opposing DQH₂ oxidation reaction catalyzed by complex III of the mitochondrial electron transport chain. Also observed in the pulmonary endothelial cells was a one-electron DQ reduction pathway that initiated intracellular DQ redox cycling.

The NQO1 gene is among the phase II and antioxidant enzyme genes that contain the antioxidant response element (ARE) (31, 32). NQO1 gene expression has been reported to be induced after exposure of various tissues and cells to a range of stimuli, e.g., reactive oxygen species (ROS), hyperoxia, xenobiotics, electrophilic and other reactive and redox cycling compounds including quinones, and ionizing radiation (31, 32, 55). In endothelial cells in culture, laminar flow, which also stimulates ROS production, activates ARE-driven gene expression of antioxidant components including NQO1 (16, 30). These observations raise the question of the impact of oxidative stress on pulmonary endothelial cell NQO1 activity and its contribution to the net effect of the cells on DQ redox status.

Hyperoxia has often been chosen as a model of pulmonary oxidative stress because although it is the most common treatment for respiratory failure, it is also injurious to the lung, with endothelial cells being a prime target (19, 47). Hyperoxia-induced ROS generation has been demonstrated in both the lung and pulmonary endothelial cells in culture and is implicated as the mechanism underlying its effects (24, 59). Various studies have evaluated the roles of an array of redox enzymes and other constituents, including the classic antioxidant enzymes (e.g., catalase, superoxide dismutase) in the pathogenesis of, or protection from, the response of pulmonary endothelial cells to hyperoxia (28, 35, 53, 68). Whereas certain redox enzymes identified in these and other studies can catalyze quinone redox reactions and/or influence quinone metabolism via reactions involving scavenging or generation of ROS, the impact of chronic hyperoxia on pulmonary endothelial cell quinone metabolism and/or NQO1 activity has not been studied (11, 12, 26, 27, 71). However, our previous observations of increased DQ reduction on passage through the lungs of hyperoxia-exposed rats (85% O₂ for 21 days), with a concomitant increase in total lung NQO1 protein and activity, suggested a contribution of hyperoxia-induced pulmonary endothelial NQO1 (2). In addition, Whitney and Frank (69) observed an increase in total NQO1 activity in neonatal and late gestational rodent lungs in response to chronic hyperoxia, and in studies of the protective effect of nuclear factor E2 p45-related factor 2 (NRF2)-mediated ARE-driven gene expression in chronically hyperoxic mice, Cho et al. (17, 18) demonstrated an effect on various antioxidant and other proteins, including an increase in NQO1 gene expression, total activity, and protein.

The present study was carried out to evaluate the influence of bovine pulmonary arterial endothelial cells exposed to chronic hyperoxia (95% O₂ for 48 h) on the redox status of DQ in the extracellular medium. If any effect was found, the further objective was to determine the relative contributions of hyperoxia-induced changes in total NQO1 activity and/or other redox pathways involved in the net effect of the cells on DQ. The results of this study are expected to give insight into the utility of quinones as probes of cellular redox function, to supply additional information regarding the response of pulmonary endothelial cells to chronic hyperoxia, and to elucidate the potential for chronic oxidative stress to alter pulmonary endothelial cell-mediated redox metabolism of blood-borne quinones.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
DQ, potassium hexocyanoferrate (III) [K₃Fe(CN), hereafter referred to as ferricyanide or Fe(CN)], HEPES, and other reagents, unless otherwise specified, were purchased from Sigma Chemical (St. Louis, MO). RPMI 1640 tissue culture medium, fetal bovine serum, and polyacrylamide gel electrophoresis supplies were from Invitrogen (Carlsbad, CA). Biosilon microcarrier beads were from Nunc (Roskilde, Denmark). Protein determinations were performed using the Bio-Rad protein assay reagent (Bio-Rad Laboratories, Hercules, CA). DQH₂ was prepared by reduction of DQ with potassium borohydride as described in Ref. 13. The NQO1 inhibitor ES936, human recombinant NQO1, and antibody to NQO1 were the kind gifts of Dr. David Siegel (School of Pharmacy, Univ. of Colorado Health Sciences Center, Denver, CO).

Endothelial cell culture. Bovine pulmonary arterial endothelial cells were isolated from segments of calf pulmonary artery obtained from a local slaughterhouse, and cells between passages 4 and 20 were cultured on Biosilon microcarrier beads (mean diameter 230 µm; surface area 255 cm²/g beads) in magnetic stirrer bottles (Techne, Burlington, NJ) containing RPMI 1640 medium supplemented with 20% fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin, and 30 mg/ml L-glutamine as previously described (45, 46).

The hyperoxic exposure was accomplished by connecting a gas tank filled with 95% O₂ and 5% CO₂ to one of the side arms of the culture stirrer bottles by a length of tubing that was fitted through an opening in the back wall of the incubator. The other side arm was fitted with tubing that exited the incubator opening and was immersed in water to a depth of 3–5 cm. The latter was to allow for visualization of gas flow through the bottle via bubble formation in the water. The gas flow rate was 5 ml/min over a period of 48–52 h, designated hereafter as 48 h. The normoxic cells were exposed in the same manner to a gas mixture containing 15% O₂, 5% CO₂, balance N₂, the latter to simulate intrapulmonary PO₂. After 24 h of gassing, a sample of culture medium was removed from six each of the normoxic and hyperoxia-exposed cultures to measure PO₂, PCO₂, and pH. In mmHg, the means ± SE normoxic and hyperoxic medium PO₂ levels were 85.5 ± 2.7 and 548.1 ± 15.5, respectively, with PCO₂ levels of 36.1 ± 0.32 and 36.7 ± 1.1, respectively. The pH values obtained for the normoxic and hyperoxic culture media were 7.28 ± 0.0 and 7.28 ± 0.1, respectively.

Protocols for measuring DQ reduction and DQH₂ oxidation by intact cells. Approximately 0.3 ml of packed volume of cell-coated beads (for amounts of cell protein per unit surface area, see Measurements of cell viability and protein content) were aliquoted into 55 x 10 x 10-mm acrylic spectrophotometric cuvettes (Sarstedt, Newton, NC). After the cell-coated beads had settled, they were washed three consecutive times by resuspension in 3 ml of room air saturated Hanks' balanced salt solution (HBSS) containing 5.5 mM glucose and 10 mM HEPES (HBSS/HEPES), pH 7.4, allowing the beads to settle between each wash. The HBSS/HEPES was the experimental medium in all experiments that follow.

The cell-coated beads were then resuspended in room air-saturated HBSS/HEPES containing DQ or DQH₂ (50 µM) and allowed to settle below the level of the spectrophotometer light path. The absorption spectrum of the medium was measured between 250 and 350 nm using a Model DU 7400 spectrophotometer (Beckman Instruments). The capped cuvettes were placed on a Nutator mixer in a 37°C incubator, and periodically, the mixing was stopped and the cell-coated beads were allowed to settle to measure the medium absorbance spectrum. At the end of the incubation period, the medium was removed from the cells, and H₂O₂ (0.1 mM, final concentration) and peroxidase (1.48 U/ml, final concentration) were added to the cell-free medium to oxidize any DQH₂ present to DQ. The absorbance spectrum was measured again to determine the total concentration of DQ present, and the difference between the DQ concentration before and after the oxidation procedure was used to determine the DQH₂ concentration that was in the medium surrounding the cells at the end of the 30-min incubation period. The concentration of DQ in the extracellular medium was calculated as described previously (2) from the absorbance values at 265 nm using the extinction coefficient of 21.64 mM^–1·cm^–1. In other experiments with this protocol, dicumarol (25 µM) or ES936 (0.5 µM) was added to the medium along with the DQ, and potassium cyanide (2 mM; KCN) was added along with DQH₂. The same protocol was used in a previous study with the cuvettes only, without cells present, to control for nonspecific association of DQ with the cuvettes (44).

DQ-mediated ferricyanide reduction was measured using a similar protocol. After washing the cell-coated beads free of the culture medium as described above, we added HBSS/HEPES containing DQ (1–50 µM) and 600 µM ferricyanide, or 600 µM ferricyanide only, and the absorbance of the ferricyanide [Fe(CN)] in the medium was measured at 421 nm. The concentration of the ferricyanide reduction product, ferrocyanide [Fe(CN)], produced was calculated from the decrease in ferricyanide concentration, determined using an extinction coefficient of 1.0 mM^–1·cm^–1. The ferricyanide protocol included experiments in which KCN (2 mM), dicumarol (25 µM), or dicumarol plus KCN (25 µM + 2 mM, respectively) were added to the medium. The background rate of cell-mediated ferricyanide reduction, which was relatively low in the absence of DQ, was measured under each experimental condition with no DQ added to the medium and subtracted from the measurements made in the presence of DQ, as described for the data in Fig. 2. The stoichiometry of DQH₂-mediated ferricyanide reduction to ferrocyanide, determined by the addition of 0.16, 0.32, or 0.48 µmol of DQH₂ to 2 µmol ferricyanide and measuring ferrocyanide production, was 1.00:1.89 ± 0.17 (means ± SD), which was not significantly different from the expected ratio of 1:2 (P = 0.1; Mann-Whitney test), providing a basis for representing the reaction as second order in ferricyanide in Eq. 3.

View larger version (24K): [in this window] [in a new window]   Fig. 2. DQ-mediated ferricyanide reduction by normoxic and hyperoxia-exposed cells. Shown is the ferrocyanide concentration generated vs. time after addition of ferricyanide (600 µM) and DQ (50 µM) to the normoxic or hyperoxia-exposed cells. Also shown is the ferrocyanide concentration vs. time when ferricyanide was incubated with cells in the absence of DQ. The latter was subtracted as background from data obtained in the presence of DQ. Experiments were carried out as sets of normoxic and hyperoxia-exposed cells, with each set from the same cell passage studied on the same day. The surface areas of cell-coated microcarrier beads (means ± SE in units of cm2) used for the normoxic and hyperoxia-exposed cell studies in the presence of DQ were 50.6 ± 1.9 and 51.3 ± 2.3, respectively, and 55.1 ± 3.6 and 45.6 ± 4.3, respectively, in the absence of DQ. The symbols represent the data (means ± SE) from n = 16 each of the normoxic and hyperoxia-exposed cell studies with DQ and n = 4 each without DQ. The solid lines represent linear regression fits to the ferrocyanide vs. time data.

NQO1 immunoblots. Cell-coated beads (0.2 ml packed volume of settled beads) were washed three times by resuspension and settling in HBSS/HEPES and mixed with 2 ml of 25 mM Tris·HCl containing 250 mM sucrose and 40 µl of Protease Inhibitor Cocktail III (Sigma), pH 7.4, and sonication and centrifugation were carried out as described above to obtain a cytosol-enriched fraction. Cytosol fraction proteins (3 or 6 µg) or purified recombinant human NQO1 protein (0.25 ng) were mixed with Nu-PAGE LDS sample buffer, the samples were loaded onto 4–12% gradient Nu-PAGE Bis-Tris polyacrylamide gels, and electrophoresis was carried out in MES-SDS running buffer (Invitrogen) as previously described (3). The proteins were transferred to nitrocellulose membranes, which were incubated for 1 h in Tris-buffered saline containing 0.1% Tween 20 and 2% bovine serum albumin, the latter as a blocking agent. The membranes were then incubated sequentially in a 1:10 dilution of -NQO1 MAb from IgG₁-secreting hybridomas (1:1 mixture of clones A180 and MAb B771; gift of Dr. David Siegel), a 1:7,500 dilution of goat -mouse IgG-horseradish peroxidase (Jackson ImmunoResearch Laboratories), and the Supersignal West Pico Chemiluminescent substrate (Pierce Biotechnology). The signal was captured on CL-Xposure Film (Pierce Biotechnology), and the band intensities were measured by laser densitometry. As a control, nonspecific mouse IgG₁ (Jackson ImmunoResearch Laboratories) was substituted for the -NQO1 MAb.

Native gel zymography. Polyacrylamide gels were prepared by mixing 6.6 ml of a 40% acrylamide/Bis stock solution (Bio-Rad Laboratories), 14.1 ml 1 M Tris·HCl (pH 8.8), 375 µl detergent solution (10% Triton X-100 + 10% sodium deoxycholate in H₂O), 6 ml 50% sucrose in H₂O, 9.6 ml H₂O, 940 µl 50 mg ammonium persulfate/ml H₂O, and 10 µl N,N,N',N'-tetramethyl-ethylenediamine (TEMED). The final concentrations of polyacrylamide, sodium deoxycholate, and Triton X-100 were 7%, 0.1%, and 0.1%, respectively. A sufficient volume of gel was poured into a minigel cassette such that the gel reached to within 1 cm of the top opening of the cassette. A stacking gel containing 6.3 ml 0.375 M Tris·HCl, pH 6.6, 1.9 ml of a 40% acrylamide/Bis stock solution, 190 µl detergent solution (10% Triton X-100 + 10% sodium deoxycholate in H₂O), 9.1 ml H₂O, 1.5 ml ammonium persulfate solution, and 7.5 µl TEMED was overlaid onto the separating gel.

To prepare a cell extract for zymography, 0.25 ml of packed volume of settled cell-coated beads were washed by resuspension and settling three times in HBSS/HEPES. After the final wash, the HBSS/HEPES was removed with a fine-tipped pipette tip, and 400 µl of a solution containing 1 ml Native TrisGly 2x Sample Buffer (Invitrogen), 20 µl Triton X-100, 20 mg sodium deoxycholate, and 20 µl Protease Inhibitor Cocktail Set III (Sigma Chemical) were added to the cell-coated beads such that the final concentrations of Triton X-100 and sodium deoxycholate were 2% each. The cells were solubilized by mixing on a Nutator mixer at room temperature for 90 min, after which the solubilized cell mixture was removed from the beads and centrifuged at 10,000 g for 30 min at 4°C. The supernatants (20 µl containing 60 µg of protein of the normoxic and 66 µg of protein of hyperoxia-exposed cell lysates, respectively) were applied to the polyacrylamide gels. The NQO1 standard was composed of 50 ng of human recombinant NQO1 (gift of Dr. David Siegel) mixed with 20 µl of the normoxic cell lysate containing 60 µg of protein, the latter to serve as an indicator of the relative electrophoretic mobilities of the standard and cell-derived reductase staining regions of the gels. Electrophoresis was carried out on ice for 2 h and 15 min at 125 V in running buffer containing 100 ml Tris-Glycine Native Running Buffer (Invitrogen), 900 ml H₂O, 5 ml Triton X-100, and 5 g sodium deoxycholate.

A modification of a stain described for NADPH diaphorase was used (40). After electrophoresis, the gel was incubated for 18 h at room temperature in 10 ml 150 mM Tris·HCl buffer, pH 8.2, containing 0.01% Tween 20, 5 µM FAD, 0.14 mM of the NQO1 substrate DCPIP, 0.24 mM NADPH, and 0.34 mM MTT, with or without 10 µM dicumarol. At the end of the staining period, gels were rinsed in water and then 50% ethanol and were then photographed.

To subject stained bands to immunoblotting for detection of NQO1, a putative normoxic cell-derived NQO1 staining band that comigrated with the standard (human recombinant NQO1) and the standard band were excised from the gel. To elute the proteins from the gel pieces, they were minced and incubated in Invitrogen Nu-PAGE LDS sample buffer overnight at room temperature. The eluates were loaded onto a polyacrylamide gel, and electrophoresis and immunoblotting were carried out as described above.

Oximetry. Cellular oxygen consumption was measured using a Yellow Springs Instruments Model 5300 biological oxygen monitor (YSI Instruments, Yellow Springs, OH) as previously described (44). Cell-coated beads in 3 ml air-saturated HBSS/HEPES containing KCN (2 mM) with or without dicumarol (25 µM) were placed in sealed magnetically stirred chambers at 37°C. Additions of DQ and ferricyanide to the chambers were made using a syringe (Hamilton, Reno, NV).

Measurements of cell viability and protein content. As a measure of cell viability, lactate dehydrogenase (LDH) activity in the medium was determined for each cell sample at the end of the experiments. The experimental medium was removed from the settled cell-coated beads, and the cells were lysed by the addition of 3 ml H₂O and sonication (3 pulses of 15 s each with the power set to 35% of maximum using a model 16-850 Virsonic Cell Disrupter). The protein content and LDH activity of each cell lysate was measured using the Bio-Rad protein assay and the method of Wroblewski and LaDue (70), respectively. The %LDH release for all of the normoxic and hyperoxia-exposed cell samples in the kinetic studies in Figs. 1–6 was 2.4 ± 0.2 (means ± SE; n = 119) and 2.4 ± 0.2 (means ± SE; n = 119), respectively, with no difference detected between the two cell types (P > 0.05, t-test). To evaluate the potential effects on cell viability of the different experimental conditions, that is, with and without inhibitors, ferricyanide, etc., we considered the condition in which ferricyanide, DQ, KCN, and dicumarol were present together in the cell medium (Fig. 6), chosen because it was anticipated to be most likely to exert a toxic effect on the cells. The %LDH release for this condition was 2.7 ± 0.5 (means ± SE; n = 9) for the normoxic and 3.0 ± 1.6 (means ± SE; n = 9) for the hyperoxia-exposed cells, with no difference detected between these values and those for all the conditions combined above (P > 0.05, t-test).

View larger version (18K): [in this window] [in a new window]   Fig. 1. Duroquinone (DQ) concentrations in the medium surrounding normoxic (A) and hyperoxia-exposed (C) cells during incubation with 50µM DQ () or durohydroquinone (DQH2; ). The DQ concentrations were determined spectrophotometrically as described in MATERIALS AND METHODS. Experiments were carried out as sets of normoxic and hyperoxia-exposed cells, with each set from the same cell passage studied on the same day. B and D: after 30 min of incubation, the medium was removed from the normoxic or hyperoxia-exposed cells, respectively, and peroxidase and H2O2 were added to the cell-free medium. The reappearance of DQ, indicated by the arrows, reflects the DQ generated via the peroxidase-catalyzed oxidation of DQH2, and thus provides a measure of the DQH2 concentration that was in the medium surrounding the cells at 30 min in A and C, respectively. The surface areas of cell-coated microcarrier beads (means ± SE in units of cm2) used for the normoxic cell studies in A were 50.2 ± 3.4 and 53.6 ± 2.4, for DQ and DQH2 additions, respectively, and for the hyperoxia-exposed studies in C were 52.0 ± 2.1 and 48.1 ± 3.4, for the DQ and DQH2 additions, respectively. The symbols represent the data (means ± SE) for n = 9 each for the DQ additions and n = 10 each for the DQH2 additions for the normoxic and hyperoxia-exposed cells. The lines represent the model fit to the data, as described in Data analysis.

View larger version (29K): [in this window] [in a new window]   Fig. 6. Inhibitor effects on normoxic and hyperoxia-exposed cell DQ-mediated ferricyanide reduction rates. The bars represent the DQ-mediated ferricyanide reduction rates (means ± SE) in the presence of DQ (50 µM), with no inhibitor additions or with dicumarol (25 µM), KCN (2 mM), or dicumarol + KCN (25 µM and 2 mM, respectively). The reduction rates were obtained by linear regression of the ferrocyanide concentration vs. time curves, and data obtained without DQ were subtracted as background from the data obtained with DQ, as in Fig. 2. The surface area of cell-coated microcarrier beads (means ± SE in units of cm2) used for normoxic and hyperoxia-exposed cells, respectively, were as follows: no additions, 51.4 ± 1.6 and 51.7 ± 2.0; dicumarol, 54.1 ± 2.2 and 50.9 ± 1.7; KCN, 54.8 ± 1.8 and 53.2 ± 2.0; dicumarol + KCN, 51.8 ± 1.8 and 53.0 ± 2.4. The number of experiments carried out with no additions for normoxic and hyperoxia-exposed cells was n = 20 each, which includes 16 from the 50 µM concentration data in Fig. 2. For the other conditions, the number of experiments for the normoxic and hyperoxia-exposed cells was n = 9 each with dicumarol, n = 11 each with KCN, and n = 9 each with dicumarol + KCN. *Significantly different from normoxic cells with no inhibitor addition; significantly different from hyperoxia-exposed cells with no inhibitor addition; significantly different from corresponding treatment in normoxic cells (t-test, P < 0.05).

View larger version (15K): [in this window] [in a new window]   Fig. 9. Effect of DQ on KCN-insensitive oxygen consumption by normoxic (A) and hyperoxia-exposed (B) cells in the absence (solid lines) or presence (dotted lines) of dicumarol. The initial recordings are after the addition of KCN (2 mM; solid line) or KCN + dicumarol (2 mM and 25 µM, respectively, dotted lines) to the oximeter chambers. DQ (50 µM) was added to the oximeter chambers at the times indicated by the arrows. Ferricyanide was added only to the chambers containing DQ without dicumarol (solid lines) at the times indicated by the arrows. Ferricyanide, DQ, or DQH2 alone did not stimulate oxygen consumption in the absence of the cells, and DQH2 alone did not stimulate oxygen consumption in the presence of cells [data not shown, (44)]. The reaction mixtures contained 49.3 and 51.3 cm2 cell-coated bead surface area for the normoxic cells with DQ or DQ + dicumarol, respectively, and 61.6 and 54.3 cm2 for the hyperoxia-exposed cells with DQ or DQ + dicumarol, respectively. The µg of protein per cm2 surface area were 29.9 and 29.7 for the normoxic and hyperoxia-exposed cells, respectively.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Figure 1, A and C, shows the DQ concentrations in the medium surrounding normoxic and hyperoxia-exposed cells, respectively, after addition of 50 µM DQ or DQH₂. For both the normoxic and hyperoxia-exposed cells, the DQ concentration in the medium decreased with time when DQ was added and increased with time when DQH₂ was added, the latter of which could not be accounted for by the relatively slow rate of DQH₂ autooxidation determined in a previous study (44). By 15 min, the medium DQ concentrations approached steady states that were independent of whether DQ or DQH₂ was initially added but were 4.5 times lower in the medium surrounding the hyperoxia-exposed cells than the normoxic cells. As shown in Fig. 1, B and D, the DQH₂ concentrations in the medium removed from the cells at the end of the 30-min incubation were 3.5 times higher in the medium surrounding the hyperoxia-exposed than the normoxic cells and were also reasonably independent of whether DQ or DQH₂ was initially added.

In Fig. 1, recoveries of DQ or DQH₂ in the cell medium at the end of the 30-min incubations were less than the initial measured concentrations. Previous studies showed that DQ, but not DQH₂, associates nonspecifically in a first-order manner with the spectrophotometric cuvettes (44). Taking this binding into consideration, on average >98% of the DQ or DQH₂ initially measured in the medium could be accounted for as DQ plus DQH₂ at the end of the 30-min incubation periods (see Data analysis and Ref. 44).

The results of the studies in Fig. 1, as well as a previous study (44), implied that the cells mediated both DQ reduction and DQH₂ oxidation. A difference in the rates of either or both reactions may have accounted for the different net effects of normoxic and hyperoxia-exposed cells on extracellular DQ redox status observed in Fig. 1. To begin to distinguish between these possibilities, the DQ reduction rates were measured using the cell membrane-impermeant electron acceptor ferricyanide, which acts as a secondary electron acceptor for DQH₂, as previously described (44). Because DQH₂ is cell membrane permeant and DQH₂-mediated ferricyanide reduction to ferrocyanide occurs virtually instantaneously, ferricyanide acts as an extracellular sink for DQH₂ produced by the cells. Thus ferricyanide minimizes the effect of cell-mediated DQH₂ oxidation on the extracellular DQH₂ appearance rate, allowing for estimation of the DQ reduction rate.

Figure 2 shows that ferricyanide reduction to ferrocyanide by normoxic or hyperoxia-exposed cells proceeded relatively slowly in the absence of DQ, but when DQ (50 µM) was present, the ferricyanide reduction rates increased, with the DQ-mediated ferricyanide reduction following zero-order kinetics. The hyperoxia-exposed cell DQ-mediated ferricyanide reduction rate, obtained by linear regression of the ferrocyanide vs. time data, was approximately twofold faster than that of normoxic cells. The cellular capacity for DQ reduction was also higher for the hyperoxia-exposed than the normoxic cells, as shown by the DQ-mediated ferricyanide reduction rates obtained over a range of DQ concentrations (Fig. 3).

View larger version (26K): [in this window] [in a new window]   Fig. 3. Effect of DQ concentration on DQ-mediated ferricyanide reduction rates in normoxic and hyperoxia exposed cells. The DQ-mediated ferricyanide reduction rates were measured in the extracellular medium of normoxic and hyperoxia-exposed cells after addition of 600 µM ferricyanide and concentrations of DQ ranging from 1 to 50 µM. The reactions followed zero-order kinetics, as in Fig. 2, at all concentrations studied. The mean data obtained in the absence of DQ (n = 3 each for the normoxic and hyperoxia-exposed cells) were subtracted from the data obtained with DQ and the DQ-mediated ferricyanide reduction rates obtained by linear regression of the ferrocyanide concentration vs. time curves (as in Fig. 2). The symbols represent the DQ-mediated ferricyanide reduction rates (means ± SE) for 6 normoxic and 6 hyperoxia-exposed cell dose-response curves, for a total of 33 data points for each cell type. The dashed lines represent the model (Eq. 9) fit to the data allowing for separate Km1 for the normoxic and hyperoxia-exposed cell data. The solid lines represent the Eq. 9 fit to data assuming single Km1 for the 2 cell types. The surface areas of cell-coated microcarrier beads (means ± SE in units of cm2) used for the normoxic and hyperoxia-exposed cell studies were 53.0 ± 1.7 and 53.7 ± 1.9, respectively. *Significantly different from normoxic cell DQ-mediated ferricyanide reduction rate obtained at the same DQ concentration (P < 0.05; t-test).

View larger version (18K): [in this window] [in a new window]   Fig. 4. DQ concentrations in the medium surrounding normoxic (A) or hyperoxia-exposed (C) cells incubated with 50 µM DQ (closed symbols) or DQH2 (open symbols) in the presence of the NAD(P)H:quinone oxidoreductase 1 (NQO1) inhibitors dicumarol (25 µM) or ES936 (0.5 µM). DQ concentrations were determined spectrophotometrically as described in MATERIALS AND METHODS. B and D: as in Fig. 1, after 30 min of incubation with DQ or DQH2, the medium was removed from the cells, and peroxidase and H2O2 were added to the cell free medium. The fact that there was essentially no difference in the medium DQ concentration before and after addition of the peroxidase and H2O2 reveals that no detectable DQH2 was present in the medium surrounding the cells at the end of the 30-min incubations in A and C. The surface areas of cell-coated microcarrier beads (means ± SE in units of cm2) used for the normoxic and hyperoxia-exposed cell studies were 46.7 ± 1.7 and 45.6 ± 2.8, respectively. The symbols represent the data (means ± SE) for n = 4 each for the normoxic and hyperoxia-exposed cells with DQ (closed symbols) and n = 5 each for DQH2 (open symbols), respectively. For ES936, 1 study each was carried out for normoxic and hyperoxia-exposed cells with DQ and 1 each with DQH2. The lines represent the model simulation carried out as described in Data analysis.

View larger version (17K): [in this window] [in a new window]   Fig. 5. Inhibition of normoxic and hyperoxia-exposed cell mediated DQH2 oxidation by the mitochondrial electron transport inhibitor KCN. A: concentration of DQ in the medium surrounding normoxic or hyperoxia-exposed cells after incubation with 50 µM DQH2 in the presence of 2 mM KCN. In B, peroxidase and H2O2 were added to the medium removed from the cells after the 30-min incubation period. The increase in DQ concentration, as indicated by the arrows, reflects the amount of DQH2 in the medium that was surrounding the cells at the end of the incubation in A. The surface area of cell-coated microcarrier beads (means ± SE in units of cm2) used for the normoxic and hyperoxia-exposed cell studies were 45.0 ± 4.0 and 43.9 ± 6.3, respectively. The data represent the means ± SE for 3 studies each for the normoxic and hyperoxia-exposed cells.

The effects of the NQO1 inhibitors implied involvement of NQO1 in DQ reduction and as a factor contributing to the differences between normoxic and hyperoxia-exposed cell-mediated DQ reduction kinetics. Figure 7 and Table 1 show that NQO1 protein levels and total activity, respectively, were higher in cytosol-enriched fractions of hyperoxia-exposed than normoxic cells.

View larger version (19K): [in this window] [in a new window]   Fig. 7. Immunoblots of normoxic and hyperoxia-exposed cell NQO1. The image is a composite of 5 immunoblots obtained using 5 different sets of normoxic (N) and hyperoxia-exposed (H) cell cultures. The same amount of cytosolic protein was loaded onto the gels for each normoxic and hyperoxia-exposed set (6 µg for pairs in A and B and 3 µg of protein for pairs in C–E). The mean intensity of the NQO1 bands was 13 ± 6 (means ± SE) times greater for the hyperoxia-exposed than the normoxic cells for the 5 cell pairs studied. There were no detectable corresponding bands in the same region of the blots when control, nonspecific IgG1 was used in the place of -NQO1 antibody. Human recombinant NQO1 was the standard.

View larger version (57K): [in this window] [in a new window]   Fig. 8. In-gel zymography of normoxic and hyperoxia-exposed cell NQO1. A: normoxic (N) and hyperoxia-exposed (H) cell extracts containing 60 and 66 µg of protein, respectively, were separated by nondenaturing gel electrophoresis and stained in a solution containing the NQO1 substrate 2,6-dichlorophenolindophenol, NADPH as the electron donor, and the tetrazolium dye MTT as the precipitating redox indicator, as described in MATERIALS AND METHODS. The arrow designates the band that was enriched when the NQO1 standard (50 ng of protein) was mixed with the normoxic cell extract before electrophoresis (NQO1 standard + N). The NQO1 standard + N, N, and H samples were run in duplicate on the same gel, the gel was cut in half before staining, and the second half, shown in B, was stained in the same staining solution except that dicumarol (10 µM) was also included. C: NQO1 immunoblot obtained using eluates from an excised NQO1 standard and a normoxic cell zymography band, the latter selected based on comigration with the NQO1 standard in the zymography, as indicated by the arrow in A.

Data analysis. A kinetic model was developed to evaluate the contributions of cellular redox processes to extracellular redox status of DQ incubated with cells (Figs. 1–4). The model includes two regions representing the cells and the medium, with volumes V_c and V_m, respectively. The model allows for NQO1-mediated two-electron DQ reduction to DQH₂ (Eq. 1) and mitochondrial electron transport complex III-mediated DQH₂ oxidation to DQ (Eq. 2). In addition to these intracellular reactions, the model allows for processes occurring in the medium not attributable to cells, including the reaction of cell-generated DQH₂ with ferricyanide [Fe(CN)] to produce ferrocyanide [Fe(CN)] (Eq. 3) and nonspecific interactions of DQ with the spectrophotometric cuvettes (Eq. 4) (44)

(1)

(2)

(3)

(4)

where RH and R⁺ are the reduced and oxidized forms, respectively, of intracellular electron donor(s), H⁺ is the hydrogen ion, and C represents cuvette sites for DQ interactions.

The dicumarol-sensitive (i.e., NQO1-mediated) reduction of DQ to DQH₂ and the KCN-sensitive (i.e., mitochondrial electron transport complex III-mediated) oxidation of DQH₂ to DQ are assumed to follow Michaelis-Menten kinetics, where for each reaction, V_max and K_m represent the apparent maximum rate and Michaelis-Menten constant, respectively (44). The other reactions are assumed to follow the law of mass action and to proceed with a rate constant k_i in the forward direction and, if reversible within the time course of the study, with a rate constant k_–i in the reverse direction, for i = 1, 2 (44).

DQ and DQH₂ were previously demonstrated to have free access to the regions representing the cells and the medium (3), i.e., the concentrations of DQ and DQH₂ at the intracellular reaction sites are essentially in equilibrium with their respective medium concentrations. With the use of mass balance and mass action, the temporal variations in the concentrations of the various species in medium or cell region are described in Eqs. 5–8.

(5)

(6)

(7)

(8)

where [DQ](t), [DQH₂](t), and [Fe(CN)](t) are the respective medium concentrations of DQ, DQH₂, and Fe(CN) at time t, and [DQ–C](t) is the concentration of DQ bound to cuvette at time t and ₂ = k₂[C] (44). Note that in deriving Eqs. 5 and 6, we assumed that V_m>>V_c, that [H⁺] and [RH] are constant during a given sample collection period. Sufficient head space was present in the spectrophotometric cuvettes that [O₂](t) may be assumed constant during the time course of the experiments and is fixed at 230 µM, which is the product of O₂ partial pressure and medium O₂ solubility.

Because DQH₂ is highly cell membrane permeant and the reaction between the ferricyanide and DQH₂ is instantaneous on the study time scale, ferricyanide acts as a sink for virtually all the cell-generated DQH₂ (44). Thus the contribution of cell-mediated DQH₂ oxidation is assumed to be eliminated when ferricyanide is present, and the ferricyanide reduction rate in the presence of DQ is given in Eq. 9, derived from Eqs. 5–8, as described in Ref. 44

(9)

where [DQ]₀ is the initial DQ concentration in the cell medium.

The parameters descriptive of DQ interactions with the cuvette in the absence of cells, ₂ = 0.028 min^–1 and k_–2 = 0.051 min^–1, were previously obtained from DQ concentration vs. time experiments carried out in the absence of cells (44). The model parameters to be estimated were the apparent maximum rates of NQO1-mediated DQ reduction to DQH₂, mitochondrial electron transport complex III-mediated DQH₂ oxidation to DQ, and their corresponding Michaelis-Menten constants. The steps for fitting the kinetic model to the data were analogous to those used previously to evaluate DQ redox reaction kinetics mediated by normal endothelial cells (44). First, to obtain the V_max1 and K_m1 for the NQO1-mediated reduction pathway, Eq. 9 was fit to the DQ-mediated Fe(CN) reduction rate vs. DQ concentration data in Fig. 3 (dashed line). The estimated V_max1 values were 0.32 and 0.73 nmol per min^–1 per cm² for the normoxic and hyperoxia-exposed cells, respectively, and the values for K_m1 were 1.8 and 3.9 µM for the normoxic and hyperoxia-exposed cells, respectively.

As can be appreciated in Fig. 3, the reduction rates for both normoxic and hyperoxia-exposed cells followed zero-order kinetics at DQ concentrations of >10 µM, the concentration range of the majority of the DQ concentration vs. time data in Figs. 1 and 4. Thus, since most of the data in this range would not be sensitive to K_m1, the model was fit to the data in Fig. 3 again, this time with the K_m1 for both conditions constrained to be the same (solid line). As seen in Fig. 3, the model fits under the different assumptions are reasonably similar, most strikingly in the key zero-order concentration range (i.e., >10 µM), again emphasizing the insensitivity of reduction rates in this range to K_m1. The estimated common K_m1 was 3.4 µM and the estimated V_max1 with the common K_m1 were 0.35 and 0.70 nmol per min^–1 per cm² for NQO1-mediated DQ reduction by the normoxic and hyperoxia-exposed cells, respectively. Comparison of these V_max1 values to those estimated from the model fit when the K_m1 were unconstrained, i.e., 0.32 and 0.73 nmol per min^–1 per cm² for the normoxic and hyperoxia-exposed cells, respectively, makes the point once again regarding the insensitivity of the majority of the data pertinent to this study to K_m1.

The results shown in Fig. 4 suggested that the cell-mediated DQH₂ oxidation capacity was not affected by the hyperoxic exposure. This hypothesis was evaluated by fitting the model, represented by Eqs. 5–7, to the DQ concentration vs. time data, and the DQH₂ data at the 30-min time point, for the normoxic and hyperoxia-exposed cell data in the absence of inhibitors (Fig. 1, A–D). The fitting was carried out under the assumption that the DQH₂ oxidation parameters (V_max2 and K_m2) were the same for the normoxic and hyperoxia-exposed cells, and by setting the V_max1 to 0.35 and 0.70 nmol per min^–1 per cm², respectively, scaled to the average cell surface areas, and the K_m1 to 3.4 µM for both the normoxic and hyperoxia-exposed cells. As shown in Fig. 1, the model fit to the data was reasonably consistent with the hypothesis, implying that the difference between the impact of the normoxic and hyperoxia-exposed cells on extracellular DQ redox status could be attributed predominately to the hyperoxia-induced increase in the maximum NQO1-mediated DQ reduction rate (V_max1), with no apparent need to account for a change in DQH₂ oxidation parameters. The resulting estimated values for the DQH₂ oxidation parameters, V_max2 and K_m2, were 0.62 nmol per min^–1 per cm² and 7.4 µM, respectively.

To further evaluate these estimates for the DQH₂ oxidation parameters, the kinetic data obtained in the presence of dicumarol were simulated under the assumption that the inhibitor eliminates DQ reduction to DQH₂, but not DQH₂ oxidation. The solid lines in Fig. 4 represent the model solution with V_max1 set to zero, with V_max2 and K_m2 set to the estimated values, and V_max2 scaled to the average cell surface areas. Because, in the absence of DQ reduction, the only model contribution to DQ disappearance from the medium is DQ cuvette binding, the model simulation is the same for the normoxic and hyperoxia-exposed cell data. As can be seen in Fig. 4, this simulation provides a reasonable fit to both data sets. This result is consistent with the hypothesis that differences between the normoxic and hyperoxia-exposed cell kinetic data could be attributed primarily to an increase in NQO1 activity in the hyperoxia-exposed cells, which makes no contribution to the data when dicumarol is present, and, therefore, the remaining DQ disappearance from the medium over the experimental time course in the presence of dicumarol can be accounted for by cuvette binding.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
The results demonstrate that the impact of intact normoxic control and hyperoxia-exposed bovine pulmonary arterial endothelial cells on the extracellular redox status of DQ could be attributed to both DQ reduction and DQH₂ oxidation, but that the balance between the two redox processes, and thus the net effect, was different for the normoxic and hyperoxia-exposed cells. The difference was attributable predominately to a hyperoxia-induced increase in DQ reduction capacity rather than an effect on DQH₂ oxidation capacity. The consequence was a quasi-steady-state extracellular [DQH₂]:[DQ] concentration ratio after 30 min of incubation with either DQ or DQH₂ that was at least 10 times higher for the hyperoxia-exposed than for the normoxic cells.

The dominant DQ reductase in the normoxic and hyperoxia-exposed cells was NQO1, as revealed by the effects of the competitive and suicide NQO1 inhibitors, dicumarol and ES936, respectively. For both the normoxic and hyperoxia-exposed cells, the inhibitors blocked DQ reduction to DQH₂ and decreased the DQ-mediated ferricyanide reduction rates. The demonstration of a hyperoxia-induced increase in NQO1 protein and total activity provided support for NQO1 induction as an explanation for the increased DQ reduction rate in the hyperoxia-exposed cells. That DQ acts as an electron acceptor for NQO1 is perhaps not that surprising since it has been used in studies of the isolated enzyme (1). The present results demonstrate that this propensity is also reflected in the net effect of intact pulmonary endothelial cells on DQ, where other potentially competing quinone reductases also coexist (58, 60), and confirm our observations in previous studies of normoxic control pulmonary endothelial cells (44).

The results further suggested that the predominant pulmonary endothelial DQH₂ oxidase in both the normoxic and hyperoxia-exposed intact cells was mitochondrial electron transport complex III, consistent with our previous observations of normoxic cells (44). This was inferred from the effect of KCN to block DQH₂ oxidation, wherein inhibition of complex IV promotes complex III reduction, closing the latter for DQH₂ oxidation. When the impact of DQ reduction was masked with dicumarol, the DQ appearance rates following incubation with DQH₂ were similar for the normoxic and hyperoxia-exposed cells, suggesting no appreciable difference in DQH₂ oxidation kinetics between the two cell types. The ability of DQH₂ to act as an electron donor to complex III in isolated mitochondria and submitochondrial particles has long been appreciated (9, 15).

The preceding qualitative interpretation of the data was evaluated using a mathematical modeling approach that provided both a means to test the hypotheses regarding the roles of the redox processes involved as well as estimates for kinetic parameter values descriptive of these processes. The result was that the differences between the kinetics of normoxic and hyperoxia-exposed cell-mediated DQ redox reactions could be explained predominately by a twofold increase in the maximum rate of DQ reduction via NQO1 (V_max1), from 0.35 to 0.70 nmol per min^–1 per cm², respectively, with no apparent effect on DQH₂ oxidation (V_max2). The implication is that NQO1 induction resulting from the hyperoxic exposure, observed as increases in cytosolic NQO1 protein and total activity, provides an explanation for the increase in V_max1.

The kinetic model analysis also revealed an apparent difference in K_m1 for normoxic and hyperoxia-exposed cell DQ-mediated ferricyanide reduction. This result was anticipated by the observation that at DQ concentrations below 10 µM, the reduction rates were similar for the two cell types, whereas they diverged at the higher concentrations (>10 µM). One implication is that had the studies been confined to only the lower DQ concentrations, the impact of hyperoxia on DQ reduction kinetics and the role of NQO1 would not have been appreciated using the described experimental protocols. Different studies and kinetic model analyses would be needed to rigorously evaluate this apparent difference in K_m1. This was not deemed necessary for the current study, wherein the goal was to expose the impact of the oxidative stress. The key point is that the majority of data collected represented DQ concentrations above 10 µM, where DQ concentration has little effect on the NQO1-mediated DQ reduction rate (i.e., zero-order kinetics).

Quinone-stimulated oxygen consumption is generally interpreted as a signature of semiquinone production and its subsequent redox cycling, to which the prooxidant effect of DQ in various cells and tissues has been attributed (48, 49, 51, 52, 67). For both normoxic and hyperoxia-exposed cells, DQ-stimulated KCN-insensitive oxygen consumption returned to baseline when NQO1 converted all the DQ to DQH₂, depleting the DQ source for one-electron reduction. This occurred more rapidly in the hyperoxia-exposed cells because they had higher NQO1 activity and were thus, relatively speaking, better protected from semiquinone generation and redox cycling. That the effect was conferred by the NQO1 is demonstrated by its loss in the presence of dicumarol, eliminating the antioxidant advantage of the hyperoxia-exposed cells. Such competition for quinone reduction via one-electron reduction pathways is a hallmark of NQO1 antioxidant function (6, 12, 20, 23, 36, 57).

The studies of DQ-mediated ferricyanide reduction in Fig. 6 revealed redox processes other than NQO1 itself that contributed to the net effect of the cells on DQ in the presence of KCN, dicumarol, or KCN plus dicumarol. For both the normoxic and hyperoxia-exposed cells, KCN increased the DQ-mediated ferricyanide reduction rate. When dicumarol was also present, the KCN-stimulated component of DQ-mediated ferricyanide reduction that was observed in normoxic cells was nearly undetectable in hyperoxia-exposed cells. For the hyperoxia-exposed cells, the implication would be that DQ-mediated ferricyanide reduction was via NQO1 under all conditions studied. The stimulatory effect of KCN could then be explained as an NQO1 dependence on the increased electron donor supply resulting from suppression of mitochondrial electron chain activity (44). This would be analogous to the dependence of transplasma membrane electron transport-mediated thiazine reduction rate on the intracellular [NADH]:[NAD⁺] ratio in these cells (43).

We previously observed the concomitant KCN-insensitive DQ-stimulated oxygen consumption and KCN- and dicumarol-insensitive DQ-mediated ferricyanide reduction in normoxic cells. This was explained by an intracellular one-electron DQ reductase, the activity of which also appeared to depend on the increased electron donor supply caused by KCN blockade of mitochondrial electron transport activity (44). The semiquinone produced could account for both the oxygen consumption via redox cycling and production of sufficient DQH₂ via semiquinone disproportionation to explain the observed ferricyanide reduction (44). This explanation is consistent with the normoxic cell data in the present study. However, a question remains regarding the mechanism underlying the persistence of the oxygen consumption signature of semiquinone production unaccompanied by DQH₂ appearance (measured as DQ-mediated ferricyanide reduction) in the medium of hyperoxia-exposed cells incubated with DQ in the presence of KCN and dicumarol. Whether the differences between the normoxic and hyperoxia-exposed cells in this regard can be explained by a change in expression or activity of a one-electron reductase and/or an effect of hyperoxia on another redox process or processes involved in DQ fate was not addressed by the present experimental design. The key point is that when the dominating roles of NQO1 and complex III are masked and/or there are changes in cell redox status that influence NQO1 activity, additional effects of the hyperoxic exposure on underlying, nondominant redox processes that can potentially contribute to the net effect of the cells on DQ are revealed. Their identity and role in quinone metabolism under various physiological and pathophysiological conditions merit further study.

NQO1 is a phase II antioxidant enzyme implicated in defense against oxidative stress whose expression is transcriptionally regulated by NRF2 via the ARE (17, 18, 31, 32, 54). Precedent for NQO1 induction in endothelial cells via this pathway is provided by studies of NRF2-dependent expression of ARE-driven genes, including NQO1, in human endothelial cells exposed to laminar flow (16, 30), an effect that may be mediated by increased ROS generation via NAD(P)H oxidase (21). To our knowledge, the present study is the first to demonstrate an effect of hyperoxia to increase NQO1 protein and total activity in pulmonary endothelial cells. However, the potential for this pulmonary endothelial response was implied by certain previous observations of the impact of chronic hyperoxia on NQO1 in mouse and rat lung. In studies that provided evidence for a role of NRF2 in protection from hyperoxic lung injury in mice, Cho et al. (17, 18) demonstrated an effect of hyperoxia (>95% O₂) on an array of NRF2-dependent lung antioxidant and phase II enzymes, including NQO1; whole lung NQO1 protein levels, total activity, and gene expression increased after 48–72 h of hyperoxic exposure of wild-type (Nrf2^+/+) compared with Nrf2^–/– (knockout) adult mice. On the other hand, in gene expression profiling studies of the early pulmonary response to hyperoxia using a different mouse strain, Perkowski et al. (54) did not detect a change in lung NQO1 gene expression during the first 48 h of exposure to >95% O₂, whereas Whitney and Frank (69) observed a hyperoxia-induced increase in total lung NQO1 activity in neonatal and late gestational, but not adult, rats and mice. Our previous studies involved rats exposed to sublethal hyperoxia, which provides a model of hyperoxic adaptation wherein the animals develop tolerance to the otherwise lethal effects of 100% O₂ after exposure to 85% O₂ for 5–7 days. We observed an increase in whole lung NQO1 protein levels and total activity and a concomitant increase in dicumarol-sensitive DQ reduction on passage through the pulmonary circulation after 21 days of hyperoxia, with no detectable effect on NQO1 or DQ reduction capacity after the first 48 h of hyperoxia (2). Whether the variations in time course and responsiveness of different experimental models are due to differences in animal species or strains, oxygen concentrations, exposure times, or other methodological factors is not known. Nevertheless, the effect of hyperoxia to induce lung NQO1 under certain conditions is reflected in the present study as an increase in bovine pulmonary arterial cell NQO1 protein and total activity. In this context, the results are consistent with a contribution of hyperoxia-induced pulmonary endothelial NQO1 activity to the higher DQ reduction capacity we observed previously in the hyperoxic rat lungs (2).

Our previous studies of DQ reduction in the lungs of rats exposed to the sublethal hyperoxia (85% O₂) allowed for estimation of Michaelis-Menten kinetic parameters for redox processes contributing to DQ fate as a basis of comparison with the present study (2). The estimated V_max for NQO1-mediated DQ reduction for the normoxic cells (0.35 nmol per min^–1 per cm²) was very similar to that estimated for control rat lungs (0.43 nmol per min^–1 per cm²), and the estimated K_m for both were 1.8 µM. Also reflected in the rat lungs after a 21-day exposure to 85% O₂ was an increase in V_max for NQO1-mediated DQ reduction from 0.43 to 0.90 nmol per min^–1 per cm², which is qualitatively consistent with the twofold increase, from 0.35 to 0.70 nmol per min^–1 per cm², seen in the endothelial cells exposed to 95% O₂ for 48 h in the present study (2). Although a role for NQO1 in defense of, or adaptation to, hyperoxic lung and pulmonary endothelial injury has not been identified (69), the fact that it has been implicated in protection from toxicity of exogenous and endogenous quinones and certain anticancer drugs (6, 12, 20, 23, 36, 51, 57), maintenance of cellular redox status (25, 39), regeneration of endogenous antioxidants (7, 57, 63), and superoxide scavenging (65) suggests a diversity of protective functions, which may provide at least part of the explanation for the fact that elevated expression of the classic antioxidant enzymes alone (i.e., superoxide dismutase, catalase, glutathione peroxidase) in transgenic animals does not confer full protection from hyperoxic lung injury (29).

The pulmonary endothelium encounters blood-borne quinoids including redox cycling metabolites of cigarette smoke, antioxidant polyphenolic compounds, redox cycling endogenous quinone metabolites of dopamine and other catecholamines and estrogen, and pharmacological, chemotherapeutic, xenobiotic, and dietary quinones (11). The results of our studies demonstrate that contact of quinones with the pulmonary endothelium can affect their redox status in the extracellular medium or lung perfusate, which for the pulmonary endothelium is equivalent to the blood. Thus, depending on the quinone properties, this endothelial function has the potential to influence quinone bioactivity within the pulmonary endothelium itself, the lung tissue, pulmonary vasculature, and downstream vessels and organs. The observations also emphasize the importance of considering the influence of oxidative stress on the balance between endothelial redox pathways that contribute to quinone fate. One implication is that for spent or oxidized forms of plasma antioxidants or coantioxidants that are NQO1 substrates and freely cell membrane permeant in the oxidized and reduced forms, the pulmonary endothelium may contribute to their regeneration. Thus under conditions of oxidative stress wherein NQO1 activity is elevated, the capacity to carry out this function may be enhanced. This concept may also be important when considering the utility of quinone-based therapeutics (15, 33, 50), wherein the very pathological conditions these compounds are intended to treat may also affect pulmonary endothelial redox pathways contributing to their redox status in the plasma and hence their biological effects in the lung and in downstream vessels and organs.

	GRANTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
This study was supported by National Heart, Lung, and Blood Institute R01 Grants HL-65537 and HL-24349 and the Department of Veterans Affairs.

	ACKNOWLEDGMENTS

 
The authors thank Dr. David Siegel for generous contributions of human recombinant NQO1, antibody to NQO1, ES936, and for helpful discussions.

	FOOTNOTES

 

Address for reprint requests and other correspondence: M. P. Merker, VA Medical Center, Research Service 151, Milwaukee, WI 53295 (e-mail: mmerker{at}mcw.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

	REFERENCES

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 

1. Anusevicius Z, Sarlauskas J, and Cenas N. Two-electron reduction of quinones by rat liver NAD(P)H:quinone oxidoreductase: quantitative structure-activity relationships. Arch Biochem Biophys 404: 254–262, 2002.[CrossRef]