Abstract
Prostanoids have been implicated in the regulation of lung vascular tone both under physiological and inflammatory conditions. The conversion of arachidonic acid (AA) to prostaglandin H2 is catalyzed at least by two isoforms of cyclooxygenase, named Cox-1 and Cox-2. Cox-1 is thought to be ubiquitously expressed, enrolled in physiological processes, whereas Cox-2 is mostly assumed to be dynamically regulated, responding to inflammatory conditions. We have recently shown by immunohistochemistry that Cox-2 is constitutively expressed in control rat lungs, with a predominant localization in smooth muscle cells of partially muscular vessels. We now asked whether Cox-2 is basically involved in the physiological regulation of pulmonary vascular tone. Isolated perfused rat lungs were challenged with intravascular bolus application of free AA to elicit thromboxane-related vasoconstrictor responses and to investigate the effects of three different selective Cox-2 inhibitors (NS-398, DUP697, SC-236). AA induced the liberation of prostaglandin I2 and thromboxane A2 into the intravascular space, and it provoked marked pulmonary artery pressure responses and concomitant lung edema formation. All events were dose-dependently inhibited by 1 to 50 μmol/liter NS-398, whereas control vasoconstrictor responses to angiotensin II and the stable thromboxane analogue U46619 were not affected by this agent. Similarly, marked inhibition of the AA elicited pressor response was achieved by 25 μmol/l DUP697 and by 10 μmol/l SC-236. These data suggest a physiological role of Cox-2 rather than Cox-1 in the regulation of vascular tone in rat lungs.
AA metabolites are since long known to be involved in pulmonary vasoregulation (Malik et al., 1985; Holtzman, 1991). Vasodilatory prostanoids are assumed to contribute to the maintenance of low pulmonary vascular tone under physiological conditions. Increased availability of free AA, whether liberated from membrane pools in vascular cells themselves or offered from the intravascular space, provokes vasoconstrictor events largely attributable to thromboxane generation (Seeger et al., 1986; Selig et al., 1986; Grimminger et al., 1995). These vasomotor events have major impact on ventilation-perfusion matching (Walmrathet al., 1993, 1994) and the regulation of microvascular pressure and thus the rate of capillary fluid filtration. In addition, AA metabolites have been implicated in changes in capillary endothelial permeability (Seeger et al., 1986; Townsley et al., 1985; Littner and Lott, 1989; Seeger et al., 1987). Finally, AA itself and its early unstable metabolites such as endoperoxide and leukotriene A4 may be exchanged between circulating cells (platelets, neutrophils, monocytes) and endothelial cells, with impact on both contributors to this cooperative AA metabolism in the lung vasculature (Grimminger et al., 1988, 1990; Mayeuxet al., 1989).
Conversion of AA to prostaglandins and thromboxane uses Cox isoenzymes as first enzymatic step. Cox-1 has been characterized as an ubiquitously expressed isoform in many tissues (DeWitt, 1991;Goppelt-Struebe, 1995), whereas Cox-2 was thought to be predominantly or even exclusively expressed under conditions of inflammation, its upregulation being responsive to a variety of inflammatory mediators (Lee et al., 1992; Feng et al., 1993; Hempelet al., 1994). High levels of prostanoids, present in inflamed tissues, have thus been attributed to highly expressed Cox-2, whereas physiologically generated prostanoids were thought to originate solely through enzymatic activity of constitutive Cox-1 (DeWitt, 1991;Klein et al., 1994). Cox-1-dependent prostanoid formation was implicated in various physiological functions, including vasomotor and bronchomotor regulation, gastric mucosa homeostasis and kidney function and fluid balance (DeWitt, 1991; Klein et al., 1994; Goppelt-Struebe, 1995).
In a recent immunohistochemical study in normal rat lungs we noted both Cox-1 and Cox-2 to be constitutively expressed (Ermert et al., 1998). Cox-1 was localized predominantly to the bronchial epithelial cells, smooth muscle cells of large hilum veins and with lower intensity to alveolar macrophages. Cox-2 was found primarily in vascular smooth muscle cells of the partially muscular vessels and large veins of the hilum and in macrophage- and mast-cell-like cells located in the peribronchial and perivascular connective tissue. We thus hypothesized that Cox-2 rather than Cox-1 might be involved in the regulation of lung vascular tone. To probe this hypothesis in our present study, we used intravascular challenge with free AA to provoke acute vasoconstrictor responses in perfused rat lungs, and investigated the impact of three different Cox-2 inhibitors, NS-398, DUP697 and SC-236, on pressor responses and prostanoid generation. In essence, the data fully support the notion that AA related vasomotion in noninflamed rat lungs proceeds predominantly or even exclusively via constitutively expressed Cox-2 activity. These observation add to the characterization of physiological functions of Cox-2, and they may have implications for antiinflammatory strategies targeting cyclooxygenase inhibition.
Materials and Methods
Reagents.
AA, ASA and U46619 were obtained from Paesel and Lorei AG (Frankfurt, Germany). The Cox-2 inhibitor NS-398 was purchased from Biomol (Hamburg, Germany) and was dissolved in DMSO. Angiotensin II was obtained from Sigma (Deisenhofen, Germany). All other biochemicals were obtained from Merck (Darmstadt, Germany). ELISA kits for the determination of 6-keto PGF1alpha and TxB2 were obtained from Cayman Chemical Company (Ann Arbor, MI). The Cox-2 inhibitor DUP697 was kindly provided from Du Pont Merck Pharmaceutical Co. (Wilmington, DE), the Cox-2 inhibitor SC-236 was a gift from Searle (St. Louis, MO).
Animals.
CD-rats (Sprague Dawley) were obtained from Charles River (Sulzfeld, Germany). All experimental procedure was performed in conformity with the guidelines of the United States National Institutes of Health (Guide for the Care and Use of Laboratory Animals, NIH publication no. 86-23, Revised 1985, US Government Printing Office, Washington DC, 20402-9325).
Lung isolation and perfusion.
The rats (male, body weight 350–400 g) were deeply anesthetized with pentobarbital-Na (100 mg/kg body weight i.p.). After local anesthesia with 2% xylocaine and median incision, the trachea was dissected and a tracheal cannula was immediately inserted. A median laparotomy was performed and subsequently the rats were anticoagulated with 1000 units of heparin. Subsequently mechanical ventilation was started with 4% CO2, 17% O2 and 79% N2 (tidal volume 4 ml, frequency 65/min, endexpiratory pressure 3 cmH2O) using a small animal respirator KTR-4 (Hugo Sachs Elektronic, Germany). After midsternal thoracotomy the right ventricle was incised, a cannula was fixed in the pulmonary artery, and the apex of the heart was cut off to allow pulmonary venous outflow. Simultaneously pulsatile perfusion with buffer solution was started. The buffer contained 2.4 mM CaCl2, 1.3 mM MgCl2, 4.3 mM KCl, 1.1 mM KH2PO4, 125.0 mM NaCl, 25 mM NaHCO3 as well as 13.32 mM glucose (pH ranged between 7.35 and 7.40).
The lungs were carefully excised, avoiding any damage, while being perfused with buffer solution, and placed in a supine position. Next a cannula was fixed through the left ventricle in the left atrium to obtain a closed perfusion circuit without leakage. After extensive rinsing of the vascular bed, the lungs were recirculatingly perfused with a pulsatile flow of 13 ml/min. The alternate use of two separate perfusion circuits, each containing 100 ml, allowed the repetitive exchange of perfusion fluid. Perfusion pressure, ventilation pressure and the weight of the isolated organ were registered continuously. The left atrial pressure was set 2 mm Hg under baseline conditions (0 referenced at the hilum) to guarantee zone III conditions at endexpiration throughout the lung. Lungs selected for the study were those that 1) had a homogenous white appearance without signs of hemostasis or edema formation, 2) had pulmonary artery and ventilation pressures in the normal range and 3) were isogravimetric during a steady state period of 30 min.
Experimental protocol.
Lungs without any drug application were recirculatingly perfused under standard conditions for 2 hr (n = 3). Experiments with drug application were conducted after the scheme depicted in figure1. In control experiments (n = 5) after termination of the steady state period, free AA was admixed to the perfusate at a concentration of 5 μmol/liter, and this challenge was repeated after 30 and after 60 min by adding the AA to the perfusion buffer.
The following concentrations and combinations of inhibitors and challenge were used: 1 mmol/liter ASA and 5 μmol/liter AA (n = 4); NS-398 in different concentrations (1, 5, 10, 25 and 50 μmol/liter) and 5 μmol/liter AA (n = 5, each group); 25 μmol/liter DUP697 and 5 μmol/liter AA (n = 4); 10 μmol/liter SC-236 and 5 μmol/liter AA (n = 4); 30 nmol/liter angiotensin II (n = 4); 25 μmol/liter NS-398 and 30 nmol/liter angiotensin II (n = 4); 2.5 nmol/liter U46619 (n = 5); 25 μmol/liter NS-398 and 2.5 nmol/liter U46619 (n = 4). Further experiments were conducted at Cox-2 inhibitor concentration of 50 μmol/liter (DUP697:n = 3, NS-398: n = 1) and AA application at a concentration of 15 μmol/liter.
The inhibitors were applied 10 min before the first AA application to guarantee a sufficient preincubation period for structural binding and irreversible inhibition of Cox-2 (Copeland et al., 1994). Additional control experiments (n = 3) were conducted with application of corresponding amounts of the solvent DMSO alone. Samples for perfusate analysis were taken before each AA injection (0, 30 and 60 min) and then every 2, 5 and 10 min (fig. 1).
Perfusate analysis.
TxA2 and PGI2were assayed by ELISA from the recirculating buffer fluid as their stable hydrolysis products TxB2 and 6-keto PGF1alpha.
Statistical analysis.
ANOVA and Student’s t test for unpaired data were used to evaluate differences among different groups. A value of P < .05 was considered significant. All data are given as mean ± S.E.M.
Results
In control lungs, no lung weight gain or change in vascular or ventilation pressure was registered over the entire observation period of 2 hr. Application of 5 μmol/liter free AA provoked a reproducible and reversible elevation in PAP (fig. 2). Concomitantly, a slow but steadily increasing weight gain was noted (table 1). In the presence of 1 mmol/liter ASA the pressor response to AA bolus injection was completely abolished. In parallel the AA-induced Δ W was largely suppressed by ASA (<0.2 g, data not presented).
Preapplication of NS-398 markedly reduced the pressor response to AA challenge in a time- and dose-dependent manner (fig. 2). Inhibition appeared to be more prominent for the first as compared to the second and third AA bolus injection. Pretreatment with 50 μmol/liter NS-398 resulted in complete suppression of pressor responses to AA application. Only after the third bolus injection a small elevation of 0.3 mmHg of pulmonary artery pressure became again detectable. DUP697 and SC-236 correspondingly inhibited the vasoconstrictor response to AA challenge. The weight gain evoked by AA application was dose-dependently suppressed following pretreatment with NS-398 (table1). Experiments with application of larger concentration of AA (15 μmol/liter) to overcome Cox-2-inhibition and probe possible activity of Cox-1 showed no pressor response, but weight gain was markedly increased compared to experiments performed with application of 5 μmol/liter AA at the same inhibitor concentration.
Application of angiotensin II provoked a PAP elevation similar to that elicited by AA. Pretreatment with the Cox-2 inhibitor NS-398 did not inhibit the pressor response to angiotensin II (table2). The thromboxane analogue U46619 was additionally used to evoke responses comparable to those after AA administration and to demonstrate intact vasoreactivity in the presence of selective Cox-2 inhibition (table 2). After NS-398 pretreatment, the PAP elicited by U46619 was not significantly reduced as compared to the reference experiments in the absence of Cox-2 inhibitor.
Experiments probing the solvent DMSO alone showed no effect (data not given in detail). None of the cyclooxygenase inhibitors investigated (ASA, NS-398, DUP697, SC-236) caused any change in base-line pulmonary artery pressure in the absence of AA and vasoconstrictor challenge.
Perfusate analysis.
After termination of the steady state period (0 min) only minute amounts of TxB2 and 6-keto PGF1alpha were detected in the recirculating perfusate (figs. 3 and4). After arachidonic acid injection, a marked prostanoid release occurred in control lungs (without inhibitor), in parallel with the pressor response.
The levels of TxB2 and 6-keto PGF1alpha did not return to baseline but remained on a higher level before the second and third AA application. The subsequent challenges with this fatty acid then provoked further increments of TxB2 and 6-keto PGF1alpha liberation.
TxB2 and 6-keto PGF1alpha levels after the first AA challenge were dose-dependently inhibited by NS-398 (figs. 3and 4). Virtually complete blockage of prostanoid generation was noted in the presence of 25 and 50 μmol/liter NS-398. The increments in prostanoid release to the subsequent second and third AA challenge were correspondingly inhibited (data not given in detail).
Discussion
In our study in perfused rat lungs, pulmonary artery pressure responses were provoked by intravascular administration of free arachidonic acid, and the release of prostanoids (TxA2, PGI2) into the recirculating buffer fluid was monitored. Three different Cox-2 inhibitors, NS-398, DUP697 and SC-236, suppressed both pressor response and prostanoid release. In contrast, control vasoconstrictor responses to angiotensin II and the stable thromboxane analogue U46619 were not affected by Cox-2 inhibition. These data suggest a physiological role of Cox-2 rather than Cox-1 in prostanoid related pulmonary vasoregulation.
The currently used model of perfused lungs has been extensively investigated in preceding studies (for review see Seeger et al., 1994). It is characterized by constant base-line pulmonary artery pressure in the physiological range, absence of significant lung edema formation over several hours and excellent physiological ventilation-perfusion matching. Accordingly, normal histological appearance was noted even after more than 2 hr of extracorporeal perfusion. Altogether these data support the view that the perfused lung may be considered to be in a noninflamed state.
It has since long been known that free AA, offered to the intact lung vasculature, is metabolized to a variety of prostanoids, with PGI2 and TxA2 representing the predominant ones (Spannhake et al., 1978; Seeger et al., 1982,1986; Malik et al., 1985; Kramer et al., 1993). Despite the parallel appearance of both vasoconstrictive and vasodilatory prostanoids, a net pressor response ensues, mostly due to the strong vasoconstrictive potency of TxA2, which surpasses the vasodilatory capacities of PGI2 and PGE2. As reproduced in our study, sequential and well reproducible AA challenges may be undertaken in this model, with intermediate return of pulmonary artery pressure to base-line levels. Using such protocol, accumulation of the stable scission products of TxA2 and PGI2, TxB2 and 6-keto-PGF1alpha, occurs in the recirculating buffer fluid. This is more obvious for 6-keto-PGF1alpha, neither taken up nor metabolized in the lung vasculature, than for TxB2which undergoes some further metabolism in the pulmonary circulation (Schulz and Seeger, 1986; Shore et al., 1989; Benedettoet al., 1987; Granström et al., 1987); these secondary metabolites are not detected by the antibody used in the currently used ELISA.
All Cox-2 inhibitors presently engaged, NS-398, DUP697 and SC-236, belong to a group of new nonsteroidal antiinflammatory drugs, which elicit their therapeutic effect by inhibition of prostaglandin synthesis via the cyclooxygenase-2 isoenzyme (Copeland et al., 1994; Masferrer et al., 1994; Reitz et al., 1994; Futaki et al., 1994). Concerning NS-398, the currently used concentration range between 1 and 50 μmol/liter has been shown to be highly selective for Cox-2-inhibition, with no effect on Cox-1 activity (Copeland et al., 1994; Masferrer et al., 1994; Futaki et al., 1994; Endo et al., 1995). This agent caused dose-dependent inhibition of both pulmonary artery pressure elevation and prostanoid release in response to the intravascular AA application. When calculated for the first AA challenge, 25 μmol/liter NS-398 sufficed to suppress all presently assessed events, PAP increase and TxA2 and PGI2formation, to less than 10% of the control response. The IC50 concentration for both inhibition of pressor response and prostaglandin synthesis was between the used experimental dosages 1 and 5 μmol/liter. In sharp contrast, the vasoconstrictor responses to angiotensin II and U46619, not demanding preceding prostanoid synthesis, were not at all affected. Similar to NS-398, the Cox-2 inhibitor DUP697 blocked the pressor response to the first AA challenge to less than 10% of control at a concentration of 25 μmol/liter. This finding fits well basic pharmacological data of corresponding affinity to Cox-2 and mode of inhibition of this enzyme by NS-398 and DUP697 (Copeland et al., 1994). Moreover, AA evoked pressor responses were also suppressed by the third specific Cox-2 inhibitor employed, SC-236, at the low concentration of 10 μmol/liter. In addition application of high concentration of AA did not overcome Cox-2 inhibition and did not result in pressor responses due to possible Cox-1 activity. Collectively these finding leave no doubt that the AA related pressor response in perfused rat lungs proceeds predominantly if not exclusively via a cyclooxygenase-2 activity.
Interestingly, the pressure response to intravascular AA bolus administration somewhat “recovered” during the second and third as compared to the first AA challenge, which was particularly evident for the high NS-398 doses of 25 and 50 μmol/liter. This observation may not be explained by some increase in circulating AA levels due to the mode of sequential application of this fatty acid, as the pressor responses to AA in the absence of inhibitor were perfectly stable. As NS-398 is known to irreversibly inactivate Cox-2, with no recovery of the enzyme activity upon removal of the drug (Copeland et al., 1994), such finding might signal some de novosynthesis of Cox-2 in the intervals of 30 and 60 min elapsing between the first and the second and third AA challenge. Although Cox-2 inhibitor was present in the perfusate, the inhibitor was given only once and might not be able to inactivate all of the newly generated Cox-2. By immunohistochemical detection we could recently demonstrate that large amounts of Cox-2 can be found in rat lungs, especially in partially muscular vessels (Ermert et al., 1998). Corresponding evidence for a rapid de novo synthesis of Cox-2 was recently obtained in cultured human bronchial smooth muscle cells in vitro (Viganò et al., 1997).
Besides provoking vasoconstriction, AA is known to increase lung vascular permeability independent of the level of microvascular filtration pressure in different species, and this event is assumed to proceed largely via lipoxygenase products of AA (Townsley et al., 1985; Seeger et al., 1986; Westcott et al., 1988; Grimminger et al., 1991; Walmrath et al., 1991). Such effect might well underly the AA-induced increase in lung weight currently observed. Alternatively, the AA-induced vasoconstrictor response might engage postcapillary vessels, thereby increasing the capillary filtration pressure and promoting lung fluid filtration. Data from different species do, indeed, suggest that the site of action of thromboxane in the pulmonary vasculature are the smooth muscle cells in small pre- and postcapillary vessels (Greenberget al., 1981; Townsley et al., 1985; Yoshimuraet al., 1989). Such role of Tx-induced postcapillary vasoconstriction in lung edema formation might well explain the current observation that the AA-elicited weight gain was also dose-dependently suppressed by Cox-2 inhibition. Moreover, this interpretation is well in line with the preceding immunohistochemical studies in the perfused rat lungs, demonstrating the presence of Cox-2 in small partially muscular vessels of pre- and postcapillary localization (Ermertet al., 1998). In contrast, Cox-1 was predominantly found in the airways, where bronchial epithelial cells displayed high staining intensity, and in smooth muscle cells of the large hilum veins, but not in smaller vessels.
In previous studies in different vascular sites, evidence was presented that inflammatory stimuli may induce smooth muscle cell Cox-2 expression (Feng et al., 1993; Rimarachin et al., 1994; Coroneos et al., 1995). In companion with the immunohistochemical demonstration of Cox-2 in small vessel smooth muscles in non-inflamed lungs (Ermert et al., 1998), this study strongly supports the notion that the cyclooxygenase-2 is constitutively expressed in the lung vasculature. Moreover, the potent inhibitory capacities of the different Cox-2 inhibitors suggest that prostanoid related pulmonary vasomotor events, known to contribute to vasoregulation under physiological and pathological conditions, proceed virtually exclusively via this type of cyclooxygenase pathway in the rat lungs. To the best of our knowledge, such fundamental role of Cox-2 in physiological vasoregulation has thus far not been described. The findings are, however, reminiscent of the observations that Cox-2 is constitutively expressed in the rat stomach (Iseki, 1995) and kidney (Harris et al., 1994), implicated in physiological gastric mucosa cytoprotection and kidney function.
In addition to broadening our knowledge on physiological functions of Cox-2, these findings may be crucial for further drug development. The Cox-2 selective nonsteroidal antiinflammatory drugs were designed to block inflammatory without affecting physiological prostanoid generation, to avoid side effects inherent in the use of nonselective antiinflammatory drugs (Klein et al., 1994; Masferreret al., 1994; Futaki et al., 1994). At least for the rat lung vasculature, such expectation does evidently not hold true, but interference with basic vasoregulatory effects is to be awaited by inhibition of Cox-2 activity.
Footnotes
-
Send reprint requests to: Dr. Leander Ermert, Institut fuer Anatomie und Zellbiologie, Aulweg 123, 35385 Giessen, Germany.
-
↵1 This work was supported by the Deutsche Forschungsgemeinschaft (SFB 547 “Kardiopulmonales Gefäßsystem”). This publication includes parts of the thesis of A.A. in partial fulfillment for the degree of M.D.
- Abbreviations:
- AA
- arachidonic acid
- Cox
- cyclooxygenase
- ASA
- acetylsalicyclic acid
- PGI2
- prostacyclin
- 6-keto PGF1alpha
- 6-keto prostaglandin F1alpha
- TxA2
- thromboxane A2
- TxB2
- thromboxane B2
- PGE2
- prostaglandin E2
- DMSO
- dimethyl sulfoxide
- ANOVA
- analysis of variance
- PAP
- pulmonary artery pressure
- Δ W
- lung weight gain
- ELISA
- enzyme linked immunosorbent assay
- Received January 6, 1998.
- Accepted April 27, 1998.
- The American Society for Pharmacology and Experimental Therapeutics