INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Hypoxic pulmonary vasoconstriction (HPV) is considered essential in regulating the distribution of pulmonary blood flow, permitting the lung to maintain a pertinent match between ventilation and blood flow (1). If HPV has deteriorated, efficiency of gas exchange in the lung is worsened, resulting in significant hypoxemia because of an augmented blood flow to hypoxic regions. Impaired HPV occurs in various lung diseases, including acute respiratory distress syndrome (2), severe pneumonia (6, 7), diffuse granulomatous lung disease (8), and oxidant lung injury induced by exposure to a high-O₂ environment (9, 10). This problem could, therefore, become clinically relevant in patients ventilated with high inspired fractions of oxygen. An excess of reactive oxygen species (ROS), generally encountered when the lung is exposed to a hyperoxic environment, has been suggested to diminish responsiveness of the pulmonary vasculature to various stimuli, especially to alveolar hypoxia (11, 12). In a summary of experimental findings reported by that time, Gurtner and Burke-Wolin (12) suggested that ROS may act as a double-edged sword for modifying vascular reactivity in the pulmonary circulation. ROS stimulate production of vasoconstrictive arachidonate mediators in the presence of endothelial cells. On the other hand, ROS may cause vasodilation through either direct activation of guanylate cyclase or interference with signal transduction within smooth muscle cells. Furthermore, Archer and colleagues (13) demonstrated that ROS may dilate pulmonary vessels by opening the potassium channels (K⁺ channels) located on the smooth muscle cell membrane. Meanwhile, several investigators, including ourselves (1), confirmed that increased production of vasodilating prostaglandins (such as prostacyclin, PGI₂) is also an important cause of blunted HPV in ROS-associated lung injury. Although nitric oxide, NO, has vasodilating properties comparable to those of PGI₂, little attention has been paid to the importance of NO in blunted HPV in oxidant-injured lungs. The current study was undertaken to address the following issues: (1) whether prolonged exposure of the lung to hyperoxia diminishes HPV in the pulmonary microcirculation; (2) if so, what kinds of pulmonary microvessels lose reactivity to hypoxic stimulation; (3) whether hyperoxia-exposed lungs enhance the production of vasodilating substances mediated via cyclooxygenase (COX) and NO synthase (NOS); (4) if so, what types of COX and NOS are responsible for the increased genesis of vasodilators. We attempted to shed light on potential effects of not only constitutive forms of cyclooxygenase (COX-1) and NO synthase (eNOS) but also their inducible forms (i.e., COX-2 and iNOS) on impaired HPV in hyperoxia-exposed lungs.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolated Perfused Rat Lungs

Changes in mean pulmonary arterial pressure (pa), microvascular diameter, erythrocyte velocity (V_r), and flow direction in the acinar microvessels were analyzed under hypoxic stimulation in isolated perfused rat lungs harvested from animals exposed either to normoxic or hyperoxic environments. Preparation of isolated perfused lungs has been described in detail previously (16). Briefly, the animals were anesthetized with 50 mg/kg pentobarbital sodium administered into the peritoneal cavity. Tracheotomy was performed and the lungs were ventilated with room air to adjust the tidal volume to 10 ml/kg and the respiratory rate to 60 breaths/min. A median sternotomy was performed and the chest was opened widely. A few minutes after 1,000 U/kg heparin had been administered into the left ventricle, blood was expelled by heart puncture, leading to rapid exsanguination and death. Thereafter, cannulas were inserted into the left atrium and main pulmonary artery and secured with strings. A ligature was also placed around the aorta to prevent loss of perfusate into the systemic circulation. The isolated lung was then fixed in the supine position on the stage of a microscope and perfused at a constant flow rate of 12 ml/min and recirculated with a roller pump (rotor 1500N; Taitec, Tokyo, Japan). Krebs-Henseleit solution with 3% bovine serum albumin was used as the perfusate, in which the hematocrit (Ht) was adjusted to 6.0 ± 0.5% by adding fresh blood obtained from donor rats. To avoid movement caused by artificial ventilation, which would hamper precise observations of blood cell kinetics in the pulmonary microcirculation, the trachea was ligated at the end-inspiratory position and gas exchange was maintained with an extracorporeal membrane oxygenator (ECMO) (Merasilox-S; Senko, Tokyo, Japan) connected with the isolated lung. A gas mixture containing 21% O₂ and 5% CO₂ flowed into the ECMO, allowing adjustment of the perfusate PO₂, PCO₂, and pH to 142 ± 6 mm Hg, 37 ± 2 mm Hg, and 7.41 ± 0.03, respectively. A warmed and humidified gas mixture containing the same composition of gases as used for the ECMO was supplied continuously to the lung surface to maintain a temperature of 37 ± 0.2° C and to avoid desiccation of the lung surface.

After stable pa had been attained, the gas flowing into the ECMO and that blown onto the lung surface were simultaneously switched from a mixture containing 21% O₂ to one containing 2% O₂ for 15 min to elicit HPV, resulting in perfusate PO₂, PCO₂, and pH being maintained at 34 ± 3 mm Hg, 40 ± 2 mm Hg, and 7.37 ± 0.02, respectively. Changes in pa from baseline during hypoxic gas breathing were used as the measure of overall HPV, i.e., vasoconstriction occurring in all the pulmonary vessels, including intraacinar microvessels and large extraacinar vessels. The pa was continuously monitored with force displacement of pressure transducers (TP-400T; Nihon Kohden, Tokyo, Japan). Perfusate PO₂, PCO₂, and pH were measured with the electrodes (model 1306; Instrumentation Laboratory, Lexington, MA).

Erythrocyte Velocity and Architecture of Pulmonary Microvessels in the Acini

Erythrocytes were stained with fluorescein isothiocyanate (FITC) (Sigma, St. Louis, MO). Fresh rat blood was centrifuged at 1,000 rpm for 5 min and the buffy coats were discarded. The same procedure was repeated three times in the presence of phosphate-buffered saline (PBS), reducing leukocyte contamination to 0.1% or less. The packed erythrocyte solution thus prepared was diluted with PBS, and FITC was added to give a final concentration of 0.1 mg/ml. This solution was then incubated at 37° C for 30 min and washed three times in PBS. One milliliter of FITC-labeled erythrocyte solution was administered to the perfusion circuit, allowing examination of erythrocyte velocity, V_r, and flow direction in the pulmonary microcirculation.

To obtain images fit for precise estimation of events in pulmonary microvessels, we used a real-time confocal luminescence microscope with 1,000-fold greater resolution velocity than a conventional confocal scanning optical microscope. Our system has been described in detail elsewhere (17). The fluorescent emission from the specimen was imaged onto a high-sensitivity charge-coupled device (CCD) camera with an image intensifier (EktaPro intensified imager VSG; Kodak, San Diego, CA). By incorporating an excitation wavelength of 488 nm emitted from a low-power air-cooled argon (Ar) ion laser (532-BSA04; Omnichrome, Chino, CA) with appropriate fluoresceins, the present confocal units allowed us to obtain apparently instantaneous images at 1,000 frames/s. The final magnifying power of our system reached ×484, with the ×20 objective, on the video screen. The resulting field of view was 210 × 210 µm, corresponding roughly to the diameter of a single pulmonary microvessel adjacent to the terminal bronchiole (19), resulting in images that could creditably detect events occurring in a single acinus. We recorded confocal images at a rate of 250 frames/s in terms of a high-speed video analysis system (EktaPro 1000 processor; Kodak) coupled with an image-intensified CCD camera (EktaPro intensified imager VSG; Kodak). All images were monitored on color video television (PVM-1444Q; Sony, Tokyo, Japan) and stored in a video cassette recorder (SVQ-260; Sony). In replaying the videotapes, which were photographed with the high-speed video analysis system, at the normal video rate, we estimated the FITC- erythrocyte velocity by measuring the distance traveled between two or more successive video frames. One frame at replay indicates events during a 4-ms interval. In addition, measurements of erythrocyte flow direction permitted a reliable discrimination between precapillary arterioles and postcapillary venules. The microvessel from which FITC- erythrocytes entered the capillary networks was defined as the arteriole, while the microvessel into which FITC-erythrocytes flowed from the capillary networks was taken to be the venule.

Precise determinations of diameter and architecture of microvessels in which erythrocyte behavior was examined were made by adding 200 µl of 5% FITC-dextran with a molecular weight of 145,000 (Sigma) to the reservoir. We estimated vessel diameter by processing a confocal video image with a digital image-analyzing system (Quadra 840AV/Image 1.58; Apple, Cupertino, CA). Hypoxia-evoked diameter changes in microvessels, including arterioles, venules, and capillaries, were used as a direct measure of HPV in the acinar microcirculation.

Experimental Protocols

Male pathogen-free Sprague-Dawley rats (8 wk old) weighing 250-300 g (n = 154) were housed in either a normoxic environment (21% O₂) or a hyperoxic environment (90% O₂) for 48 h. Animals exposed to normoxia and hyperoxia were defined as the N group and H group, respectively. On preparing isolated lungs from these animals, we examined 2% O₂-elicited changes in overall HPV, microcirculatory diameter, and erythrocyte velocity under conditions with and without various agents that would inhibit either COX or NOS. N and H groups were divided into six subgroups, respectively, on the basis of the agents applied.

1. C condition (n = 12 for the N group; n = 9 for the H group): No agent was administered.
2. I condition (n = 7 for the N group; n = 8 for the H group): Indomethacin was used to inhibit constitutive as well as inducible forms of cyclooxygenase (COX-1 and COX-2). The concentration of indomethacin in the perfusate was maintained at 20 µM.
3. NS condition (n = 5 for the N group; n = 8 for the H group): N-(2-cyclohexyloxy-4-nitrophenyl)methane sulfonamide (NS-398; Taisho, Tokyo, Japan) was applied as the specific agent to inhibit predominantly inducible COX-2. After the preliminary experimental results, we adjusted the perfusate concentration of NS-398 to 1 µM. The specificity of NS-398 in suppressing COX-2 function has been extensively discussed by Futaki and coworkers (20, 21).
4. 4. LN condition (n = 6 for the N group; n = 7 for the H group): A constitutive form of NO synthase (eNOS) and an inducible form of NO synthase (iNOS) were concomitantly inhibited with N-nitro- L-arginine methyl ester (L-NAME), whose concentration in the perfusate was maintained at 100 µM.
5. 5. AM condition (n = 8 for the N group; n = 6 for the H group): Aminoguanidine (perfusate concentration, 4 mM) was administered as a specific inhibitor against iNOS (22, 23).
6. 6. ILN condition (n = 6 for the N group; n = 7 for the H group): Measurements were made in the presence of both indomethacin and L-NAME to inhibit simultaneously COX-1, COX-2, eNOS, as well as iNOS.

PGI₂ and NO-associated Metabolites in the Perfusate

We measured perfusate concentrations of PGI₂ and NO-related metabolites before the introduction of hypoxic stimulation in the N and H groups.

A 2-ml perfusate sample collected in a tube containing indomethacin and EDTA was centrifuged and frozen at 80° C until extraction. At the time of extraction, 1 ml of the sample was extracted into medium containing 4 ml of ethyl acetate. The ethyl acetate layer was transferred to a second tube, and evaporated to dryness with pure N₂ gas. Thereafter, the dried extract was reconstituted in the assay buffer. The concentration of immunoreactive 6-keto-prostaglandin F₁ (6-keto-PGF₁), a stable metabolite of PGI₂, was examined by enzyme-linked immunosorbent assay (ELISA) (EIA kit; Cayman Chemical, Ann Arbor, MI).

A 2-ml perfusate sample was centrifuged and frozen at 80° C immediately after collection. As a measure of NO production in the lung, we measured the total concentration of end products of NO metabolism, NO₂ and NO₃, in the perfusate by the method of Green and colleagues (24). To deproteinize the perfusate sample, 100 µl of 35% sulfosalicyclic acid was added to a 500-µl sample. These samples were mixed by vortexing every 5 min and allowed to react for 30 min at room temperature. They were then centrifuged at 10,000 × g for 15 min. Two hundred microliters of supernatant, 300 µl of 5% NH₄Cl, and 60 µl of 5% NaOH were combined for analysis. The prepared sample was passed through a high-pressure Teflon column packed with fine particles of copper-plated cadmium metal allowing reduction of NO₃ to NO₂. The effluent was mixed at 60° C with the reagent containing one part 1% sulfanilamide in distilled water plus one part 0.1% N-(1-naphthyl)-ethylenediamine in 5% concentrated phosphoric acid, and the color of the product yielded by the diazotization reaction was examined at 0° C with a spectrophotometer (US 501; Unisoku, Osaka, Japan) at an absorbance wavelength of 546 nm. The standard curve for estimating the concentration of NO-related metabolites was constructed from freshly prepared sodium nitrite solution. The efficiency in reducing NO₃ to NO₂ was estimated by passing 10 µM sodium nitrate solution through the copper-plated cadmium column.

Statistical Evaluation

Statistical differences in results were generally judged by one-way ANOVA followed by multiple comparison analysis by Scheffe examination. Comparison of increments in pa and of microvascular diameter changes observed in the N and H groups under various experimental conditions was made by applying the unpaired t test. Changes in pa, microvascular diameter, and erythrocyte velocity in respective microvessels before and after hypoxic stimulation were examined by the paired t test. Values are presented as means ± SD, and p < 0.05 was considered to be significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of Hypoxia on Overall HPV and Intraacinar Microvessel Diameter Changes in the Absence of Medication

Increments of pa by hypoxic stimulation in the absence of medications were lower for the H group than those observed for the N group (Figure 1). Baseline diameters of intraacinar arterioles in the N group averaged 24.9 µm and were reduced by 2.6 µm on hypoxic stimulation (Figures 2 and 3). Although the arteriolar diameters in the H group before hypoxic stimulation were not different from those of the N group, hyperoxia-exposed arterioles were not constricted during hypoxic gas breathing (Figures 2 and 3). In the N and H groups, venule diameters in the acini were comparable to those of the arterioles. Hypoxic stimulation did not alter venular or capillary diameters in either the N or H group (Figure 3).

View larger version (13K): [in this window] [in a new window]   Figure 1.   Increments in mean pulmonary arterial pressure (pa) during hypoxic stimulation in normoxia-treated lungs (N group; n = 12) and hyperoxia-exposed lungs (H group; n = 9) under medication-free conditions. #Significantly lower than the values obtained in the N group.

View larger version (149K): [in this window] [in a new window]   Figure 2.   Confocal views of intraacinar arterioles before and after hypoxic stimulation in the N and H groups. (A and B) N group before and after hypoxia, respectively. (C and D) H group before and after hypoxia, respectively. a = arteriole; c = capillary; alv = alveolus. Original magnification: ×20.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Microvascular diameter changes in the N group (n = 12) and H group (n = 9) with no medication. *Smaller than zero (i.e., active vasoconstriction); #Significantly differing from the values obtained for the N group.

Effects of COX Inhibition on Overall HPV and Intraacinar Microvessel Diameter Changes

Under medication-free conditions, perfusate concentrations of 6-keto-PGF₁ were 310 pg/ml in the N group, whereas those in the H group averaged 616 pg/ml, a twofold greater difference (Figure 4). Administration of indomethacin or NS-398 significantly reduced perfusate 6-keto-PGF₁ concentrations in the H group (Figure 4). These values were comparable to those of the N group obtained under medication-free conditions.

View larger version (18K): [in this window] [in a new window]   Figure 4.   Perfusate concentrations of vasodilating substances in the N and H groups. (A) 6-keto-PGF1; (B) NO-related metabolites. C = control condition (absence of medications; n = 7 each for N and H groups); I = indomethacin, used for inhibiting COX-1 and COX-2 (H group, n = 6); NS = NS-398, administered for inhibition of COX-2 (H group, n = 6); LN = eNOS and iNOS, simultaneously suppressed with L-NAME (H group, n = 6); AM = aminoguanidine (only iNOS was inhibited with this substance; H group, n = 5). #Significantly greater than the values of the N group. $Significantly smaller than the values obtained for the H group under no medication.

In the N group, administration of either indomethacin or NS-398 exerted no influence on pa increments (Table 1) and microvascular diameter changes in comparison with that without medications (Figure 5).

View larger version (13K): [in this window] [in a new window]   Figure 5.   Effects of inhibiting COX-1 and/or COX-2 on hypoxia-elicited intraacinar microvessel diameter changes in lungs exposed to normoxic (N group) or hyperoxic environment (H group). (A) Precapillary arterioles; (B) postcapillary venules. C = control condition (n = 12 for N group, n = 9 for H group); I = indomethacin (n = 7 for N group, n = 8 for H group); NS = NS-398 (n = 5 for N group, n = 8 for H group). Data of the C condition are the same as given in Figure 3, but are again presented here for the sake of clear comparison. *Active vasoconstriction; #Significantly differing from the values of the corresponding N group; $Significantly differing from the C condition; +Significantly different from the I condition.

Although the hypoxia-induced pa increase in the H group was not altered by treatment with indomethacin (Table 1), hypoxia-elicited arteriolar constriction was restored, i.e., the decrease in the arteriolar diameter after hypoxic stimulation was 2.2 µm (Figure 5). In responding to hypoxic gas, venules as well as capillaries in the H group did not display constriction in the presence of indomethacin (Figure 5). NS-398 exerted little influence on overall pressor response (Table 1) and constriction of arterioles, venules, and capillaries in the H group during hypoxic stimulation (Figure 5).

Effects of NOS Inhibition on Overall HPV and Intraacinar Microvessel Diameter Changes

In the absence of medications, concentrations of NO-related metabolites in the perfusate averaged 8.8 µM for the N group and 14.2 µM for the H group, with a significant difference between the two values (Figure 4). NO production in the hyperoxia-exposed lungs was suppressed by treatment with L-NAME and aminoguanidine, respectively (Figure 4). Neither L-NAME nor aminoguanidine influenced perfusate concentrations of 6-keto-PGF₁ in the H group experiments.

In the N group, L-NAME markedly augmented hypoxia- induced overall pressor response (Table 1), but did not alter hypoxia-elicited diameter changes in any microvessels in comparison with that obtained under medication-free conditions (Figure 6). Meanwhile, aminoguanidine had no effects on either overall pressor response (Table 1) or microvascular diameter changes in intact lungs during hypoxic stimulation (Figure 6).

View larger version (13K): [in this window] [in a new window]   Figure 6.   Effects of inhibiting eNOS and/or iNOS on hypoxia-evoked intraacinar microvessel diameter changes in normoxia-exposed (N group) and hyperoxia-exposed lungs (H group). (A) Precapillary arterioles; (B) postcapillary venules. C = control condition (n = 12 for N group, n = 9 for H group); LN = L-NAME (n = 6 for N group, n = 7 for H group); AM = aminoguanidine (n = 8 for N group, n = 6 for H group). C-condition data are identical to those shown in Figure 3. *Active vasoconstriction; #Significantly differing from the values of the corresponding N group; $Significantly differing from the C condition; +Significantly differing from the values obtained under the LN condition.

In the H group, L-NAME enhanced hypoxia-evoked pa increments (Table 1). Although mean values of pa increase caused by hypoxic stimulation deviated by 4.5 mm Hg between the N and H groups, the difference did not attain significance owing to a large dispersion of the measured data (p = 0.1, Figure 7). In parallel, L-NAME restored the arteriolar constriction, but did not have any influence on diameter changes in venules and capillaries during hypoxic gas breathing (Figure 6). Although aminoguanidine did not alter pa increments (Table 1), it perceptibly improved arteriolar constriction of hyperoxia-exposed lungs on hypoxic stimulation (Figure 6). However, the recovery of arteriolar constriction with aminoguanidine was less than that obtained with L-NAME (Figure 6). Aminoguanidine exerted no influence on the vasoconstriction of venules and capillaries in the H group (Figure 6).

View larger version (15K): [in this window] [in a new window]   Figure 7.   Effects of simultaneous administration of indomethacin and L-NAME on pulmonary hemodynamics in lungs exposed to normoxia (N group) and hyperoxia (H group). (A) Overall pressor increments after hypoxic stimulation. C-condition data are the same as those presented in Figure 1. (B) Arteriolar diameter changes on hypoxic stimulation. Data for the C, I, and LN conditions are, respectively, identical to those given in Figures 3, 5, and 6. C = control (n = 12 for N group, n = 9 for H group); I = indomethacin (n = 7 for N group, n = 8 for H group); LN = L-NAME (n = 6 for N group, n = 7 for H group); ILN = concomitant administration of indomethacin and L-NAME (n = 6 for N group, n = 7 for H group). #Significantly differing from the corresponding N group values; $Significantly differing from the C condition; +Significantly greater than the values observed for the I condition; *Active vasoconstriction.

Effects of Simultaneous Inhibition of COX and NOS on Overall HPV and Intraacinar Microvessel Diameter Changes

In both the N and H groups, inhibition of COXs and NOSs by the concurrent administration of indomethacin and L-NAME did not augment pa changes after hypoxic gas breathing, in comparison with administration of L-NAME alone (Figure 7). Although, as in the case of administering L-NAME alone, mean values of hypoxia-evoked pa changes in the presence of both indomethacin and L-NAME diverged by 7.5 mm Hg between the N and H groups, the difference was not significant (p = 0.08, Figure 7). When compared with the microvascular results with either indomethacin or L-NAME alone, additional improvement in arteriolar contraction on hypoxic stimulation was not investigated, in both groups, by administering indomethacin and L-NAME concomitantly (Figure 7).

Effects of Hypoxia on Erythrocyte Velocity in Intraacinar Microvessels

Since there were no differences in results among various medications in mean erythrocyte velocity, V_r, in respective microvessels, all such V_r values were lumped together (Table 2). V_r values in each microvasculature were not different between the N group and the H group. Arteriolar V_r was lower than venular V_r but higher than capillary V_r in both groups. There were no variations in V_r in each microvessel on hypoxic stimulation in either the H group or the N group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Critique of Methods

The issue of whether intraacinar microvessels could be constricted has been largely unresolved. This may be attributed to the technical difficulties in analyzing events in such small vessels. We have, therefore, applied a real-time confocal laser scanning luminescence optical microscope, developed for the precise discrimination of individual microvessels in the highly complex acinar microcirculation (17).

Although hypoxic stimulation reduced the arteriolar diameter by approximately 10% in intact lungs, it neither increased nor decreased erythrocyte velocity in any of the microvessels studied (Figure 2 and Table 2). These results may indicate that a decrease in total cross-sectional area of intraacinar arterioles produced by a reduction in their diameters (original diameter; 20-30 µm) is not accompanied by a significant alteration in erythrocyte velocity. These findings are inconsistent with those of Shirai and coworkers (25), who demonstrated that blood flow velocity in intact feline arteries approximately 400 µm in diameter was reduced on hypoxic stimulation. Although the reason for the discrepancy between the two studies is not evident, perhaps it is due to differences in animal species studied and vascular size analyzed, i.e., Shirai and colleagues (25) measured blood flow velocity in extraacinar arteries that were 20-fold larger than those measured in our study. Since the total cross-sectional area of intraacinar arterioles is larger than that of extraacinar arteries because of the increased branching number in the former, the relative extent of decreasing total vascular area on hypoxic stimulation is expected to be much less for arterioles. Therefore, the effects of reduced cross-sectional area on flow velocity may be concealed at the intraacinar arteriolar level.

Attenuated Hypoxia-induced Vasoconstriction in Hyperoxia-exposed Lungs

Although normoxia-exposed intact lungs exhibited a distinct increase in pa associated with intraacinar arteriolar constriction on hypoxic stimulation, hyperoxia-exposed lungs showed much less change in pa and less disappearance of arteriolar constriction (Figures 1 and 3), indicating an attenuated HPV in hyperoxia-treated lungs. These findings suggest that the intraacinar arteriole would be one of the portions responsible for impaired HPV in hyperoxia-exposed lungs. Our findings are consistent with those reported by Troug and coworkers (10), who, by using pa increments as a simple measure of HPV, showed that overall HPV was impaired in hyperoxia-injured lungs.

Although the issue of whether intraacinar arterioles have sufficient quantities of cells capable of vasoconstriction has been controversial (26, 27), our results are in accordance with the findings of Davies and colleagues (27), who demonstrated that intraacinar arterioles with a diameter greater than 10 µm have cells with contractile properties.

Importance of COX-related Substances for Attenuated HPV in Hyperoxia-exposed Lungs

Inhibition of COXs by indomethacin had no influence on pa increments and microvascular diameters during hypoxic stimulation in normoxia-treated lungs (Table 1 and Figure 5), suggesting that COX-related vasodilating PGs would play no role in modifying intra- and extraacinar HPV in intact lungs.

Although simultaneous inhibition of COX-1 and COX-2 by indomethacin did not alter hypoxia-induced pa changes in lungs exposed to hyperoxia, it allowed perfusate 6-keto-PGF₁ concentrations to decrease and intraacinar arterioles to regain the contractility to hypoxic stimulation. On the other hand, inhibition of COX-2 by NS-398 improved neither overall HPV nor arteriolar constriction during hypoxic stimulation. Taken together, these findings may indicate that constitutive COX-1 expression is augmented at least along intraacinar arteriolar walls in hyperoxia-treated lungs and is an important factor in depressing arteriolar constriction caused by hypoxia. Enhanced COX-1 expression, however, may not be a decisive factor for impairment of overall HPV.

ROS, such as superoxide and hydrogen peroxide, were reported to be potent inducers of COX-2 gene expression in a variety of cells or tissues (28). Lee and coworkers (29) demonstrated that exposure to prolonged hyperoxia resulted in an enhanced prostaglandin synthesis in perinatal rat lung cells, suggesting that hyperoxia exposure could induce COX-2, probably mediated via ROS. Supporting their findings (28, 29), we showed that prostaglandin synthesis in hyperoxia-treated lungs was decreased with NS-398, a specific COX-2 inhibitor (Figure 4). Although COX-2 was shown to be upregulated in lung cells exposed to hyperoxia, its role in modifying hypoxia-elicited vasoconstriction in the pulmonary circulation is not clear, because the COX-2 inhibitor did not improve either hypoxia-evoked overall pressor changes or microvascular contraction in acini exposed to hyperoxia.

Importance of NOS-related Substances for Attenuated HPV in Hyperoxia-exposed Lungs

Although overall hypoxia-evoked pressor changes were augmented by inhibiting NOSs with L-NAME in both normoxia- and hyperoxia-treated lungs (Table 1 and Figure 6), arteriolar response in the presence of L-NAME qualitatively differed between the two groups. L-NAME did not alter hypoxia-elicited arteriolar constriction in intact acini, but enhanced that in acini exposed to hyperoxia. Inhibition of iNOS by aminoguanidine partially restored arteriolar contraction in association with no alteration in the overall pressor response during hypoxic stimulation in hyperoxia-exposed lungs. These findings may indicate that the attenuated arteriolar constriction on hypoxic stimulation in hyperoxia-exposed lungs is, in part, attributable to an excessive amount of NO yielded through both eNOS and iNOS. By considering the results of microvascular constriction together with those of overall pressor changes, we could infer that upregulation of iNOS in hyperoxia-treated lungs is confined to intraacinar regions and is not crucial in the depression of the overall HPV. Instead, eNOS expression may be enhanced along not only intraacinar arterioles but also extraacinar arteries and function as an important factor in blunting the overall HPV in hyperoxia-exposed lungs. Our findings are consistent with those of Liao and colleagues (30), who demonstrated that exposure of bovine pulmonary artery endothelial cells to 95% O₂ produced an increase in the eNOS mRNA level.

In comparison with results obtained with inhibiting NOS alone, simultaneous suppression of COX and NOS in hyperoxia-exposed lungs did not cause a further improvement in either overall pressor changes or vasoconstriction of intraacinar arterioles during hypoxic stimulation (Figure 7). Salvemini and colleagues (31) reported that NO enhances COX activity through a mechanism independent of cGMP and leads to an NO-mediated increase in the production of proinflammatory prostaglandins in Escherichia coli lipopolysaccharide (LPS)- treated rats. However, our results did not support their finding, i.e., neither L-NAME nor aminoguanidine influenced 6-keto-PGF₁ concentrations in the perfusate, suggesting that the mechanism proposed by Salvemini and coworkers (31) cannot explain our experimental results. The plasma NO concentrations in LPS-treated rats (Salvemini and colleagues [31]) were more than 100 µM, whereas those in hyperoxia-exposed rats (our study) were about 15 µM. Salvemini and coworkers (31) demonstrated in addition that plasma 6-keto-PGF₁ concentrations in LPS-treated rats were not reduced by the presence of either L-NAME or aminoguanidine at times when plasma NO concentrations were less than 60 µM.

Hypoxia-evoked overall pressor changes in the presence of L-NAME (or L-NAME concomitant with indomethacin) diverged about twofold between normoxia- and hyperoxia-treated lungs, although the difference did not reach the statistical significance (Figure 7). We thought this might be caused by a large scatter of the measured data, such that we might not eliminate the possibility of a pathway, independent of prostaglandins and NO, also that plays a role in impairing overall HPV in lungs exposed to hyperoxia. Archer and colleagues (13) reported that ROS open the K⁺ channels by oxidizing the sulfhydryl groups, leading to vasodilation due to hyperpolarization of smooth muscle cells. It is, therefore, possible that an attenuated overall HPV in lungs treated with hyperoxia is, in part, regulated by ROS-associated opening of K⁺ channels. This, however, requires further clarification.

Our study of hyperoxia-exposed lungs yielded the following conclusions: (1) A 48-h exposure to hyperoxia attenuates overall pressor changes of pulmonary circulation and arteriolar constriction responding to hypoxic stimulation; (2) attenuated overall pressor response to hypoxia may be preferentially mediated by excessive NO yielded via upregulated eNOS; (3) depressed arteriolar constriction may be associated not only with vasoactive prostaglandins produced by enhanced COX-1 but also with NO yielded by upregulated eNOS and iNOS; (4) COX-2 may not be a decisive enzyme causing attenuated hypoxia-evoked arteriolar constriction; and (5) erythrocyte velocity may be of little value in understanding intraacinar HPV.