The role of G protein beta subunits in the release of ATP from human erythrocytes

Previously, we demonstrated that adenosine triphosphate (ATP) is released from human erythrocytes in response to mechanical deformation and that this release requires activation of a signal-transduction pathway involving adenylyl cyclase and the heterotrimeric G protein, Gs. Here we investigate the role of heterotrimeric G proteins of the Gi subtype in the release of ATP from human erythrocytes. In addition, we determined the profile of heterotrimeric G protein beta subunits present in these erythrocyte membranes. The activity of Gi was stimulated by incubation of erythrocytes (20% hematocrit) with mastoparin (10 microM). ATP release was measured using the luciferin/luciferase assay. Heterotrimeric G protein beta subunits present in erythrocyte membranes were resolved using gel electrophoresis and subunit specific antibodies. Incubation of human erythrocytes with mastoparan (an activator of Gi/o) resulted in a 4.1 +/- 0.6-fold increase in ATP present in the medium (P<0.01). Human erythrocyte membranes stain positively for beta subunit types 1, 2, 3 and 4, all of which been reported to activate of some isoforms of adenylyl cyclase. Activation of the heterotrimeric G protein, Gi, results in ATP release from erythrocytes. This effect is may be related to the activity of beta subunits associated with this G protein in the human erythrocyte.

Participation of cAMP in a signal-transduction pathway relating erythrocyte deformation to ATP release

Previously, we reported that red blood cells (RBCs) of rabbits and humans release ATP in response to mechanical deformation and that this release of ATP requires the activity of the cystic fibrosis transmembrane conductance regulator (CFTR). It was reported that cAMP, acting through a cAMP-dependent protein kinase, PKA, is an activator of CFTR. Here we investigate the hypothesis that cAMP stimulates ATP release from RBCs. Incubation of human and rabbit RBCs with the direct activator of adenylyl cyclase, forskolin (10 or 100 microM), with IBMX (100 microM), resulted in ATP release and increases in intracellular cAMP. In addition, epinephrine (1 microM), a receptor-mediated activator of adenylyl cyclase, stimulated ATP release from rabbit RBCs. Moreover, incubation of human and rabbit RBCs with an active cAMP analog [adenosine 3'5'-cyclic monophosphorothioate Sp-isomer (Sp-cAMP, 100 microM)] resulted in ATP release. In contrast, forskolin and Sp-cAMP were without effect on dog RBCs, cells known not to release ATP in response to deformation. When rabbit RBCs were incubated with the inactive cAMP analog and inhibitor of PKA activity, adenosine 3',5'-cyclic monophosphorothioate Rp-isomer (100 microM), deformation-induced ATP release was attenuated. These results are consistent with the hypothesis that adenylyl cyclase and cAMP are components of a signal-transduction pathway relating RBC deformation to ATP release from human and rabbit RBCs.

Receptor-mediated activation of the heterotrimeric G-protein Gs results in ATP release from erythrocytes

BACKGROUND

Adenosine triphosphate (ATP), released from the erythrocyte in response to mechanical deformation, decreased oxygen tension or reduced pH, has been suggested to be an important determinant of vascular resistance in several vascular beds. Mechanical deformation-induced ATP release from rabbit and human erythrocytes was reported to require the activity of the cystic fibrosis transmembrane conductance regulator (CFTR), suggesting that a signal transduction pathway involving CFTR mediates ATP release from erythrocytes. Here we investigate the hypothesis that the heterotrimeric G-protein Gs is also involved in this signal transduction pathway.

MATERIALS AND METHODS

The heterotrimeric G-protein Gs was identified in rabbit and human erythrocyte membranes, using gel electrophoresis. The concentration of ATP released into a suspension of erythrocytes, incubated with iloprost or epinephrine, was measured using the luciferin/luciferase assay.

RESULTS

The 45 kDa form of the heterotrimeric G-protein Gs was identified in rabbit and human erythrocyte membranes. Incubation of rabbit erythrocytes with iloprost (n=18) or epinephrine (n=6) increased the ATP concentration by 106+/-16% and 156+/-54%, respectively. Epinephrine-induced changes in ATP concentrations were prevented by pretreatment with propranolol.

CONCLUSIONS

The heterotrimeric G-protein Gs is present in erythrocyte membranes. Receptor-mediated activation of Gs results in ATP release. These results are consistent with the hypothesis that Gs is a component of a signal transduction pathway for ATP release from erythrocytes.

Impaired release of ATP from red blood cells of humans with primary pulmonary hypertension

Previously, we reported that in the isolated perfused rabbit lung, red blood cells (RBCs) obtained from either rabbits or healthy humans were a required component of the perfusate to unmask evidence of nitric oxide (NO) participation in regulation of the pulmonary circulation. In addition, we found that mechanical deformation of rabbit and healthy human RBCs released ATP, a known agonist for enhanced NO synthesis. In contrast, RBCs obtained from patients with cystic fibrosis (CF) did not release ATP in response to mechanical deformation. The coexistence of airway disease and alveolar hypoxia in patients with CF precluded the drawing of conclusions relating a defect in RBC ATP release with the pulmonary hypertension associated with CF. Airway disease and alveolar hypoxia are not, however, features of primary pulmonary hypertension (PPH), a human condition of unknown etiology. We postulated that a defect in NO generation might contribute to the increased pulmonary vascular resistance in PPH, and as a first step, we hypothesized that RBCs obtained from patients with PPH would not release ATP. In contrast to RBCs of healthy humans, when RBCs of PPH patients were passed through filters (average pore size 12, 8, or 5 microm), ATP was not released and the RBCs exhibited reduced deformability. Moreover, when incubated with the active cAMP analogue, Sp-cAMP (100 microM), an activator of the CF transmembrane conductance regulator, ATP was not released. These results demonstrate that RBCs obtained from patients with PPH fail to release ATP whether the stimulus is mechanical or pharmacological. Thus, failure of RBCs to release ATP in patients with PPH might be a major pathogenetic factor that accounts for the heretofore unknown etiology of their pulmonary hypertension.

Deformation-induced ATP release from red blood cells requires CFTR activity

Recently, it was reported that rabbit and human red blood cells (RBCs) release ATP in response to mechanical deformation. Here we investigate the hypothesis that the activity of the cystic fibrosis transmembrane conductance regulator (CFTR), a member of the ATP binding cassette, is required for deformation-induced ATP release from RBCs. Incubation of rabbit RBCs with either of two inhibitors of CFTR activity, glibenclamide (10 microM) or niflumic acid (20 microM), resulted in inhibition of deformation-induced ATP release. To demonstrate the contribution of CFTR to deformation-induced ATP release from human RBCs, cells from healthy humans, patients with cystic fibrosis (CF), or patients with chronic obstructive lung disease (COPD) unrelated to CF were studied. RBCs of healthy humans and COPD patients released ATP in response to mechanical deformation. In contrast, deformation of RBCs from patients with CF did not result in ATP release. We conclude that deformation-induced ATP release from rabbit and human RBCs requires CFTR activity, suggesting a previously unrecognized role for CFTR in the regulation of vascular resistance.

5,6-Epoxyeicosatrienoic acid reduces increases in pulmonary vascular resistance in the dog

We recently reported that canine pulmonary microsomes metabolize arachidonic acid to all four regioisomeric epoxyeicosatrienoic acids (EET). 5,6-EET dilates blood vessels in several nonpulmonary vascular beds, often in a cyclooxygenase-dependent manner. The present study was designed to determine whether 5,6-EET can decrease pulmonary vascular resistance (PVR) in the intact pulmonary circulation. In isolated canine lungs perfused with physiological salt solution, a constant infusion of U-46619 (3.28 +/- 0.99 nmol/min) increased PVR 62.1 +/- 4.5%. Administration of 5,6-EET (10(-5) M) into the perfusate reduced the U-46619-mediated increase in PVR by 23.6 +/- 6.1%. These effects of U-46619 and 5,6-EET were limited to changes in resistance solely in the pulmonary venous segment. In contrast, venous as well as arterial segmental resistances were increased in 5-hydroxytryptamine (5-HT)-treated lungs. However, in the latter instance, 5,6-EET reduced arterial but not venous segmental resistance. 5,6-EET increased pulmonary PGI2 synthesis from 70.5 +/- 18.4 to 675.9 +/- 125.4 ng/min. In the presence of indomethacin (10(-4) M), 5,6-EET did not increase PGI2 synthesis nor did it decrease U-46619- or 5-HT-mediated increases in PVR. In canine intrapulmonary vessels, 5,6-EET decreased active tension in veins contracted with U-46619. 5,6-EET decreased active tension in arteries but not veins contracted with 5-HT, consistent with results in the perfused lungs. These results demonstrate that 5, 6-EET is a vasodilator in the intact pulmonary circulation. Its dilator activity depends on the constrictor agent present, the segmental resistance, and cyclooxygenase activity.

Biosynthesis of sulfidopeptide leukotrienes via the transfer of leukotriene A4 from polymorphonuclear cells to bovine retinal pericytes

Administration of exogenous sulfidopeptide leukotrienes (LTs) is associated with enhanced microvascular permeability. In addition, endogenous LTs have been implicated as participants in permeability (nonhydrostatic) edema formation. The source of LTs for interaction with the microvasculature is, however, unknown. We hypothesized that pericytes contribute to vascular LT synthesis. Under basal conditions and after incubation with either the calcium ionophore, A23187 (0-1 microM), or arachidonic acid (20 microM), bovine retinal pericytes (BRPs) did not produce significant amounts of sulfidopeptide LTs. In contrast, in the presence of polymorphonuclear leukocytes (PMNs), which can synthesize LTA4, but not sulfidopeptide leukotrienes, incubation of BRPs with A23187 resulted in dose-dependent increases in LTC4/D4/E4 production (peak: 35.4 +/- 5 pg/microg protein; n = 12). Similarly, BRPs, incubated with exogenous, authentic LTA4 (10 microM), synthesized sulfidopeptide LTs (peak: 18.9 +/- 5 pg/microg protein, n = 3). Preincubation (30 min) of BRPs with PMNs and the lipoxygenase inhibitor, esculetin (1 x 10(-)4 M; n = 12), reduced peak A23187-induced production of LTs by 63.9 +/- 7%. Finally, Northern blot analysis revealed mRNA for 5-lipoxygenase to be present in human and bovine PMNs, but not in BRPs. These results suggest that pericytes produce sulfidopeptide LTs only when provided with LTA4 from an external source such as the PMN. Interactions between pericytes and PMNs may lead to the production of sulfidopeptide LTs, which, in turn, could alter microvascular permeability.

Nitric oxide opposes phorbol ester-induced increases in pulmonary microvascular permeability in dogs

In addition to its effects on vascular tone, nitric oxide (NO) has been suggested to function as a participant in fluid homeostasis affecting interactions between the endothelium and circulating inflammatory cells. The role of NO in the increased microvascular permeability of acute lung injury, however, remains controversial. We investigated the hypothesis that NO opposes increases in pulmonary vascular permeability after phorbol myristate acetate administration, i.e., in a model of neutrophil-dependent acute lung injury. In anesthetized dogs, phorbol myristate acetate (10 microg/kg, i.v.) had no effect on pulmonary arterial pressure (Ppa) or extravascular lung water. After pretreatment with the NO synthesis inhibitor, NG-nitro-L-arginine methyl ester (10 mg/kg, i.v. ; 5 mg/kg/hr), an identical dose of phorbol myristate acetate resulted in a 20 +/- 8 mm Hg (P < .01) increase in pulmonary arterial pressure and a 186 +/- 86% (P < .01) increase in extravascular lung water. To determine if the pulmonary edema was related to increases in microvascular pressure or to changes in the microvascular permeability coefficient, experiments were performed in isolated blood-perfused dog lungs. The addition of phorbol myristate acetate (4.2 x 10(-8) M) to the perfusate was without effect on microvascular pressure or pulmonary capillary filtration coefficient. However, after NG-nitro-L-arginine methyl ester (100 microM), phorbol myristate acetate resulted in increases in both microvascular pressure and permeability coefficient that were prevented by pretreatment with L-arginine (1 mM). These data support the hypothesis that endogenous NO opposes increases in pulmonary vascular permeability as well as microvascular pressure in this neutrophil-dependent model of acute lung injury resulting in preservation of the endothelial barrier to the passage of water and solutes and prevention of the formation of pulmonary edema.

Vascular endothelial growth factor mRNA in pericytes is upregulated by phorbol myristate acetate

Increased microvascular permeability, which occurs in conditions such as the adult respiratory distress syndrome and diabetes mellitus, is related to physicochemical alterations in the microvascular barrier. We postulate that, in part, capillary pericytes affect microvascular permeability via production of a vasoactive cytokine, viz, vascular endothelial growth factor (VEGF), also known as vascular permeability factor. The goal of the present study was to evaluate the effects of phorbol myristate acetate (PMA), a substance known to produce nonhydrostatic pulmonary edema in intact animals, on VEGF gene expression in pericyte cultures. Microvascular pericytes were isolated from bovine retinas using magnetic microspheres coated with 3G5 monoclonal antibody. Pericyte identity was confirmed both morphologically and by immunostaining for alpha-smooth muscle actin and 3G5 ganglioside. The cultured pericytes were stimulated with N(omega)-nitro-L-arginine methyl ester (L-NAME, 1 x 10(-4) mmol/L), angiotensin II (1 x 10(-6) mmol/L), and PMA (5 x 10(-8) mmol/L), selected because of their ability to upregulate VEGF mRNA expressions in other cell types. Northern blot analysis was performed using [32P]dCTP labeled human VEGF cDNA (Genentech). Lane-loading differences were normalized using mouse GAPDH control cDNA probe. VEGF mRNA expression was upregulated by PMA (10(-9) to 10(-6) mol/L) in a dose-dependent manner, whereas neither L-NAME nor angiotensin II affected VEGF mRNA expression in pericytes. These results support the hypothesis that pericytes increase permeability of the endothelial barrier through increased VEGF production.

ATP: the red blood cell link to NO and local control of the pulmonary circulation

Recently, we reported that rabbit red blood cells (RBCs) were required for the expression of nitric oxide (NO) activity on pulmonary vascular resistance (PVR) in rabbit lungs. Here, we investigate the hypothesis that RBCs participate in the regulation of PVR via release of ATP in response to mechanical deformation that, in turn, evokes vascular NO synthesis. We found that rabbit and human RBCs, but not dog RBCs, release ATP in response to mechanical deformation. To determine the contribution of this ATP to NO synthesis and PVR, we compared the effects of human and dog RBCs on pressure-flow relationships in isolated rabbit lungs. In the presence of human RBCs, NG-nitro-L-arginine methyl ester (100 microM) produced a shift in the pressure-flow relationship consistent with a reduction in vascular caliber. NG-nitro-L-arginine methyl ester had no effect in lungs perfused with dog RBCs. These results suggest a unique mechanism for the control of PVR in rabbits and humans whereby release of ATP by RBCs in response to mechanical deformation leads to stimulation of NO synthesis that, in turn, modulates the PVR.

Inhibition of cytochrome P-450 attenuates hypoxemia of acute lung injury in dogs

The intravenous administration of ethchlorvynol (ECV), in dogs, resulted in an acute lung injury (ALI) characterized by a 200 +/- 80% increase in venous admixture and a 142 +/- 30% increase in extravascular lung water (EVLW). Pretreatment with the cytochrome P-450 inhibitor 8-methoxypsoralen prevented the ECV-induced increase in venous admixture but not the increased EVLW. These findings parallel those reported for cyclooxygenase inhibition in ECV-induced ALI and suggest that an arachidonic acid (AA) metabolite of pulmonary cytochrome P-450 activity may mediate the increase in venous admixture of ALI. We demonstrate that canine pulmonary microsomes metabolize [1-(14)C]AA to a variety of products, including the cytochrome P-450 metabolites 5,6-, 8,9-, 11,12-, and 14,15-epoxyeicosatrienoic acid (EET). In prostaglandin F2 alpha-contracted, isolated pulmonary venous rings, 5,6-EET induced relaxation in a concentration-dependent manner. This action of 5,6-EET was prevented by indomethacin (10(-5) M). These results suggest that may serve as the cyclooxygenase-dependent endogenous pulmonary vasodilator responsible for the increase in venous admixture of ECV-induced ALI.

Hypotheses regarding the role of pericytes in regulating movement of fluid, nutrients, and hormones across the microcirculatory endothelial barrier

A decade ago, we initiated studies to define relationship(s) between products of 5-lipoxygenase-mediated arachidonic acid metabolism and altered microvascular permeability. Patients with permeability (nonhydrostatic) pulmonary edema (adult respiratory distress syndrome) and intact animal models of permeability edema, produced with agents that required neutrophils (phorbol myristate acetate) and those that did not (ethchlorvynol), invariably revealed the presence of leukotrienes; in contrast, leukotrienes were not detected in cases of hydrostatic pulmonary edema. In isolated perfused canine lung, we identified increases in microvascular permeability coefficients in response to the injurious agent. Permeability coefficients were not increased when injurious agents were given in the presence of 5-lipoxygenase inhibitors. To define further the relationships between leukotriene generation and edema formation, we postulated that leukotrienes effected contraction of capillary pericytes, thereby increasing pore size of endothelial intercellular junctions and enhancing movement across the microvascular barrier. We isolated pericytes from bovine retinas, identified them morphologically and by staining characteristics, and, in preliminary experiments, found that they do not possess the 5-lipoxygenase enzyme; however, when cocultured with neutrophils, which possess 5-lipoxygenase but cannot synthesize sulfidopeptide leukotrienes because of their lack of glutathione S-transferase, sulfidopeptide leukotriene synthesis ensued. In view of the anatomic position of pericytes, evidence that they participate in endothelial transport, their ability to contract, and evidence of cell-to-cell communication, we propose that pericytes control the movement of fluid, solutes, hormones, and small and large molecules across the microvascular endothelium.

Effect of L-NAME on pressure-flow relationships in isolated rabbit lungs: role of red blood cells

Nitric oxide (NO) is produced by and relaxes pulmonary arteries and veins; however, a role for NO as a participant in the control of pulmonary vascular resistance (PVR) remains to be defined. Here we investigated the hypothesis that for NO to serve as a determinant of PVR in the rabbit requires the presence of blood. In isolated blood-perfused rabbit lungs, NG-nitro-L-arginine methyl ester (L-NAME, 100 microM) increased PVR and the slope of the pressure-flow relationship. These effects of L-NAME were prevented by pretreatment with L-arginine. In contrast, in lungs perfused with a physiological salt solution, L-NAME had no effect on PVR or the pressure-flow relationship. The addition of washed red blood cells (RBCs) to physiological salt solution, but not the addition of plasma and platelets, restored the response to L-NAME. This effect of RBCs was not reproduced by increasing perfusate viscosity with dextran. These results suggest that, in the rabbit lung, NO is a determinant of PVR in the presence of blood. Moreover, that aspect of blood that permits the generation of NO appears to be related to the RBC and not to perfusate viscosity.

Relationship of arachidonic acid release to porcine coronary artery relaxation

In porcine coronary artery endothelium-dependent relaxation to bradykinin is in part attributed to a chemically unidentified factor, termed endothelium-derived hyperpolarizing factor (EDHF). We hypothesize that arachidonic acid, acting through a cyclooxygenase-independent mechanism, is responsible for EDHF production. To define the relationship between EDHF production and arachidonic acid release, we investigated the role of phospholipase C in bradykinin-induced relaxation and prostaglandin I2 production (an index of arachidonic acid release) in porcine coronary artery. The phospholipase C inhibitor U73122 (1 mumol/L) abolished bradykinin-induced, nitric oxide-mediated relaxation but did not inhibit either bradykinin-induced, EDHF-mediated relaxation or prostaglandin I2 production. However, when given at a larger dose (20 mumol/L) U73122 abolished both bradykinin-induced, EDHF-mediated relaxation and prostaglandin I2 production. Similarly, the calcium-ATPase inhibitor thapsigargin, given at a dose (1 mumol/L) that abolished bradykinin-induced increases in intracellular calcium concentration in cultured porcine coronary artery endothelial cells, eliminated both bradykinin-induced. EDHF-mediated relaxation and prostaglandin I2 production. Although thapsigargin abolished bradykinin-induced prostaglandin I2 production, the basal production of prostaglandin I2 was enhanced and contraction of endothelium-intact rings was attenuated. These latter responses are most likely related to enhanced basal arachidonic acid release and associated EDHF production. These observations suggest that phospholipase C activation and increased intracellular calcium concentration are required for both bradykinin-induced arachidonic acid release and EDHF production in porcine coronary artery.(ABSTRACT TRUNCATED AT 250 WORDS)

Inhibition of nitric oxide synthesis improves arterial oxygenation in ethchlorvynol-induced acute lung injury in dogs

In both humans and in experimental animals, acute lung injury (ALI) is characterized by the development of pulmonary edema and arterial hypoxemia. It has been reported that the hypoxemia of ALI is related to the failure of those mechanisms that result in the diversion of blood flow away from hypoxic (edematous) lung units to those that are well oxygenated. One such mechanism is hypoxic pulmonary vasoconstriction (HPV). In the pulmonary circulation, endogenous nitric oxide (NO) has been shown to oppose HPV and, thereby, to support blood flow to hypoxic alveoli. In the present work we investigated the hypothesis that, in ALI, endogenous NO, by virtue of its ability to oppose HPV, supports blood flow to hypoxic lung units resulting in increases in venous admixture (Qva/Qt) and decreases in arterial oxygen tension (PaO2). In anesthetized and mechanically ventilated dogs, the intravenous administration of ethchlorvynol (ECV, 15 mg/kg) resulted in an increase in extravascular lung water (EVLW) of 10 +/- 1 ml/kg body wt (p < 0.001) as well as a 120 +/- 45% increase in Qva/Qt (p < 0.01) and a 23 +/- 5% decrease in PaO2 (p < 0.01) (n = 3). L-NAME (1 mg/kg iv, followed by 5 mg/kg/h, iv), administrated 60 min after ethchlorvynol (ECV), prevented entirely the ECV-induced increase in Qva/Qt and fall in PaO2 with minimal effect on EVLW (n = 3). We conclude that, in this model of ALI, endogenous NO is present in the lung and acts to support blood flow to poorly oxygenated lung units resulting, thereby, in reductions in PaO2.

Endothelium-dependent relaxation to arachidonic acid in porcine coronary artery: is there a fourth pathway?

Endothelium-dependent relaxations to bradykinin (BK) in U46619-contracted, indomethacin (INDO)-treated porcine coronary artery (PCA) rings are modestly attenuated by the nitric oxide (NO) synthase inhibitor, N omega-nitro-l-arginine methyl ester (L-NAME); whereas, when contracted with KCl, L-NAME abolishes BK relaxations. In contrast, endothelium-dependent arachidonic acid (AA) relaxations of U46619-contracted, INDO-treated PCA rings are not affected by L-NAME. AA does not relax KCl-contracted rings. Since BK is known to release AA, we postulated that the non-NO component of BK relaxation of the PCA is mediated by AA or an AA metabolite. Changes in tension of PCA rings to BK and AA were determined in the presence and absence of phospholipase (PLA), cyclooxygenase (CO), lipoxygenase (LO) and cytochrome P-450 (cP450) inhibitors. Responses to BK were attenuated by PLA inhibitors. No other inhibitors, however, eliminated responses to either BK or AA. The results suggest that relaxation to BK in PCA rings requires PLA activity, but relaxation to AA is independent of PLA, CO, LO or cP450 activity. We conclude that relaxation to BK and AA in the PCA is mediated by a product of an unidentified pathway of AA metabolism or by an unknown second messenger system resident within the endothelium and responsive to AA.

Inhibition of nitric oxide synthesis results in a selective increase in arterial resistance in rabbit lungs

Endogenous nitric oxide (NO) opposes the vasoconstriction that occurs when lungs are ventilated with a hypoxic gas mixture. However, the contribution of NO to pulmonary vascular resistance when alveolar gas tension is not reduced remains to be defined. Here, we investigated the hypothesis that endogenous NO is a determinant of pulmonary vascular resistance in isolated perfused rabbit lungs ventilated with a normoxic gas mixture. Moreover, we wished to establish that, as flow rate increases, the contribution of NO to vascular resistance increases. In addition, we examined the contribution of NO to the longitudinal distribution of pulmonary vascular resistance. Pressure-flow curves were generated in isolated blood perfused rabbit lungs by varying flow rate from 50 ml/min to 300 ml/min in the presence and absence of the cyclooxygenase inhibitor, indomethacin (100 microM) and the inhibitor of NO synthesis, NG-nitro-L-arginine methyl ester (L-NAME, 100 microM). Indomethacin did not alter total pulmonary vascular resistance or the longitudinal distribution of resistance. In contrast, L-NAME administration resulted in significant, flow-related increases in total vascular resistance, i.e., after L-NAME, as flow rate increased, the increment in resistance increased. L-NAME-induced increases in total pulmonary vascular resistance were the result of flow-related increases in the arterial component of vascular resistance. These results provide support for the hypothesis that NO is an important determinant of pulmonary vascular resistance in the rabbit and that the major site of NO activity resides in the arterial side of that circulation.

Effect of indomethacin on increases in leukotriene B4 and pulmonary edema in response to phorbol ester administration in dogs

The administration of leukotrienes (LTs) into the pulmonary circulation results in edema formation and increased vascular permeability. We reported previously that the administration of phorbol myristate acetate (PMA, 20 micrograms/kg) to intact anesthetized dogs results in a reduction in circulating white blood cells as well as the development of pulmonary edema concomitant with the appearance of LTs in the lungs. In contrast, when a smaller dose of PMA (10 micrograms/kg) was administered, neither extravascular lung water nor LTs increased, although there was a similar reduction in circulating white blood cells. In the present study, we used a property of indomethacin, namely, its capacity to augment the formation of LTs, to examine further the relationship between LT generation and pulmonary edema formation in response to PMA administration. In intact pentobarbital-anesthetized dogs pretreated with saline (N = 9), the administration of PMA at a dose of 10 micrograms/kg, i.v., did not result in any change in extravascular lung water or in LTB4 present in bronchoalveolar lavage fluid (BALF). In contrast, in six animals pretreated with indomethacin (5 mg/kg), the administration of this dose of PMA resulted in increases in both extravascular lung water (P < 0.05) and LTB4 (P < 0.05) in BALF. These results provide support for the hypothesis that leukotrienes are requisite for PMA-induced increases in extravascular lung water.

Relaxation of porcine coronary artery to bradykinin. Role of arachidonic acid

Bradykinin-induced relaxation of precontracted, porcine coronary artery (PCA) rings is mediated by distinctly different endothelium-derived relaxing factors depending on the contractile agent used. Thus when contracted with KCl, bradykinin-induced relaxation of PCA rings is mediated solely by nitric oxide (NO), whereas when contracted with the thromboxane mimetic U46619, a small component of the relaxation is attributable to NO and a large component is attributable to a non-NO mechanism that is independent of cyclooxygenase activity. We hypothesized that the non-NO component was mediated by arachidonic acid (AA) or by a non-cyclooxygenase product of AA metabolism. Bradykinin-induced relaxations of PCA rings precontracted with U46619 in the presence of indomethacin (10 mumol/L) were moderately attenuated by the NO synthase inhibitor N omega-nitro-L-arginine methyl ester (L-NAME, 100 mumol/L), whereas when precontracted with KCl, L-NAME abolished the relaxations. AA produced endothelium-dependent relaxations of rings precontracted with U46619 that were unaffected by L-NAME, whereas AA did not relax rings precontracted with KCl. In rings precontracted with U46619, in the presence of L-NAME and indomethacin the phospholipase inhibitors quinacrine (50 mumol/L) and 4-bromophenacyl bromide (10 mumol/L) attenuated bradykinin- but not AA-induced relaxations. Inhibitors of both lipoxygenase (BW 755c [100 mumol/L] and nafazatrom [20 mumol/L]) and cytochrome P-450 (proadifen [10 mumol/L] and clotrimazole [10 mumol/L]) pathways did not eliminate bradykinin- or AA-induced relaxations, although clotrimazole partially attenuated AA-induced relaxations. These findings suggest that bradykinin-induced relaxation of PCA rings is mediated by AA through a mechanism that is not dependent on cyclooxygenase, lipoxygenase, or cytochrome P-450 pathways.

OKY-046 prevents increases in LTB4 and pulmonary edema in phorbol ester-induced lung injury in dogs

Thromboxanes (Txs) were implicated as possible participants in the altered microvascular permeability of acute lung injury when the Tx synthase inhibitor, OKY-046, was reported to prevent pulmonary edema induced by phorbol myristate acetate (PMA). Recently, however, we found that OKY-046, at a dose just sufficient to block Tx synthesis in intact dogs, did not prevent PMA-induced pulmonary edema but rather merely reduced it modestly. The present study was designed to explore other mechanisms whereby OKY-046 might prevent PMA-induced pulmonary edema. The finding that 5-lipoxygenase (5-LO) metabolites of arachidonic acid were increased within the lung after PMA administration, coupled with the report that OKY-046 inhibited slow-reacting substance of anaphylaxis formation, permitted formulation of the hypothesis that OKY-046, at a dose in excess of that required to inhibit Tx synthesis, inhibits the formation of a product(s) of 5-LO and, thereby, prevents edema formation. In vehicle-pretreated pentobarbital-anesthetized male mongrel dogs (n = 4), PMA (20 micrograms/kg i.v.) increased pulmonary vascular resistance (PVR) from 4.4 +/- 0.3 to 26.3 +/- 8.8 mmHg.l-1 x min (P < 0.01) and extravascular lung water from 6.7 +/- 0.5 to 19.1 +/- 6.2 ml/kg body wt (P < 0.05). Concomitantly, both TxB2 and leukotriene B4 (LTB4) were significantly increased in the lung. Pretreatment with OKY-046 (100 mg/kg i.v., n = 8) prevented PMA-induced increases in TxB2, LTB4, and pulmonary edema formation but did not prevent the increase in PVR.(ABSTRACT TRUNCATED AT 250 WORDS)

Eicosanoid balance and perfusion redistribution of oleic acid-induced acute lung injury

We have proposed that endogenous prostacyclin opposes the vasoconstriction responsible for redistribution of regional pulmonary blood flow (rPBF) away from areas of increased regional lung water concentration (rLWC) in canine oleic acid- (OA) induced acute lung injury (D. P. Schuster and J. Haller. J. Appl. Physiol. 69: 353-361, 1990). To test this hypothesis, we related regional lung tissue concentrations of 6-ketoprostaglandin (PG) F1 alpha and thromboxane (Tx) B2 in tissue samples obtained 2.5 h after administration of OA (0.08 ml/kg iv) to rPBF and rLWC measured by positron emission tomography. After OA only (n = 16), rLWC increased in dependent lung regions. Some animals responded to increased rLWC by redistribution of rPBF away from the most edematous regions (OA-R, n = 6), whereas others did not (OA-NR, n = 10). In another six animals, meclofenamate was administered after OA (OA-meclo). After OA, tissue concentrations of 6-keto-PGF1 alpha were greater than TxB2 in all groups, but concentrations of 6-keto-PGF1 alpha were not different between OA-R and OA-NR animals. TxB2 was increased in the dependent regions of animals in both OA-R and OA-NR groups compared with controls (no OA, n = 4, P < 0.05). The tissue TxB2/6-keto-PGF1 alpha ratio was smaller in controls and OA-NR in which no perfusion redistribution occurred than in OA-R and OA-meclo in which it did occur. This TxB2/6-keto-PGF1 alpha ratio correlated significantly with the magnitude of perfusion redistribution.(ABSTRACT TRUNCATED AT 250 WORDS)

Enhanced microvascular permeability of PMA-induced acute lung injury is not mediated by cyclooxygenase products

Products of cyclooxygenase activity have been proposed to mediate the pulmonary hypertension and increased microvascular permeability associated with phorbol myristate acetate- (PMA) induced acute lung injury. Previously, we reported that thromboxane (Tx) does not mediate PMA-induced pulmonary hypertension in intact anesthetized dogs. In the present study, PMA was administered to isolated canine lungs perfused with autologous blood at constant flow to investigate a possible role for Tx in the PMA-induced increase in microvascular permeability. Changes in permeability were assessed by determining changes in the capillary filtration coefficient (Kfc). In lobes pretreated with papaverine to prevent PMA-induced increases in pulmonary vascular resistance, Kfc increased from a baseline value of 0.2 +/- 0.03 to 1.5 +/- 0.29 ml.min-1.cmH2O-1.100 g wet lobe wt-1 (P < 0.01) 30 min after PMA (5.8 x 10(-8) M, n = 10). Concomitantly, TxB2, the stable metabolite of TxA2, increased from 138 +/- 44 to 1,498 +/- 505 pg/ml (P < 0.05) in the blood. Both the selective Tx synthase inhibitor, OKY-046 (7 x 10(-4) M, n = 6), and the cyclooxygenase inhibitor, indomethacin (10(-4) M, n = 7), prevented the PMA-induced increase in TxB2, but neither compound attenuated the PMA-induced increase in Kfc. ONO-3708 (10(-6) M), a selective prostaglandin (PG) H2/TxA2 receptor antagonist, prevented the vasoconstriction resulting from administration of U-46619, a stable PGH2/TxA2 receptor agonist, but it did not prevent the PMA-induced increases in Kfc (n = 6).(ABSTRACT TRUNCATED AT 250 WORDS)

Role of endothelium in responses of isolated hepatic vessels to vasoactive agents

The role of the endothelium as a participant in the responses to vasoactive agents was evaluated in isolated canine hepatic artery (HA) and portal vein (PV) rings. Endothelial and smooth muscle integrity was determined by pharmacologic responses as well as by histologic examination. Smooth muscle relaxation was expressed as the percent of decrease of norepinephrine-induced isometric contraction. Acetylcholine (ACh)-induced relaxation of the HA was abolished by removing the endothelium or by the addition of either hemoglobin, methylene blue (MB) or Ng-mono-methyl-L-arginine. In addition, relaxation induced by nitroglycerin, but not that induced by prostaglandin E1, was attenuated by MB. These data suggest endothelium-dependency of the relaxation to ACh and mediation of the response by endothelium-derived relaxing factor through activation of guanylate cyclase. In contrast, ACh produced contraction of the PV which was unaffected by removing the endothelium. The calcium ionophore, A23187, on the other hand, produced relaxation of the PV, which was significantly decreased by removing the endothelium. Relaxation of both HA and PV, produced by 2-chloroadenosine (2-C-Ado) was partially attenuated by removing the endothelium. With the endothelium intact, neither hemoglobin, MB, Ng-monomethyl-L-arginine nor indomethacin affected the responses to 2-C-Ado in the HA and PV, suggesting that the responses were not mediated by endothelium-derived relaxing factor or products of guanylate cyclase or cyclooxygenase activity. Nitroglycerin relaxed both vessels in the presence or absence of endothelium, indicating that removal of the endothelium had not affected smooth muscle function.(ABSTRACT TRUNCATED AT 250 WORDS)

Thromboxane does not mediate pulmonary hypertension in phorbol ester-induced acute lung injury in dogs

Thromboxane (Tx) has been suggested to mediate the pulmonary hypertension of phorbol myristate acetate- (PMA) induced acute lung injury. To test this hypothesis, the relationship between Tx and pulmonary arterial pressure was evaluated in a model of acute lung injury induced with PMA in pentobarbital sodium-anesthetized male mongrel dogs. Sixty minutes after administration of PMA (20 micrograms/kg iv, n = 10), TxB2 increased 10-fold from control in both systemic and pulmonary arterial blood and 8-fold in bronchoalveolar lavage (BAL) fluid. Concomitantly, pulmonary arterial pressure (Ppa) increased from 14.5 +/- 1.0 to 36.2 +/- 3.5 mmHg, and pulmonary vascular resistance (PVR) increased from 5.1 +/- 0.4 to 25.9 +/- 2.9 mmHg.l-1.min. Inhibition of Tx synthase with OKY-046 (10 mg/kg iv, n = 6) prevented the PMA-induced increase in Tx concentrations in blood and BAL fluid but did not prevent or attenuate the increase in Ppa. OKY-046 pretreatment did, however, attenuate but not prevent the increase in PVR 60 min after PMA administration. Pretreatment with the TxA2/prostaglandin H2 receptor antagonist ONO-3708 (10 micrograms.kg-1.min-1 iv, n = 7) prevented the pressor response to bolus injections of 1-10 micrograms U-46619, a Tx receptor agonist, but did not prevent or attenuate the PMA-induced increase in Ppa. ONO-3708 also attenuated but did not prevent the increase in PVR. These results suggest that Tx does not mediate the PMA-induced pulmonary hypertension but may augment the increases in PVR in this model of acute lung injury.

Production of leukotrienes in phorbol ester-induced acute lung injury

Leukotrienes, when administered into the pulmonary circulation of intact animals or isolated perfused lungs, have been associated with the formation of pulmonary edema. In addition, leukotrienes were identified in edema fluid and in bronchoalveolar lavage fluid (BALF) both from patients with the adult respiratory distress syndrome (ARDS) and from dogs with ethchlorvynol-induced acute lung injury (ALI). To determine whether the identification of leukotrienes in BALF was a finding common to ALI, etiology notwithstanding, we produced acute lung injury in dogs with phorbol myristate acetate (PMA). PMA produces a model of ALI thought to differ mechanistically from ethchlorvynol-induced ALI. Leukotriene C4 (LTC4), D4 (LTD4) and B4 (LTB4) were measured in BALF before and after PMA administration in intact pentobarbital-anesthetized dogs. The intravenous administration of 20 or 30 micrograms/kg of PMA produced increases in pulmonary vascular resistance (PVR) and extravascular lung water (EVLW), whereas, 10 or 15 micrograms/kg caused only a modest increase in PVR with no increase in EVLW. LTD4 and LTB4 were increased in BALF solely in those animals that developed increases in EVLW. These results, when viewed together with those reported in humans with ARDS and in dogs with ethchlorvynol-induced ALI, support the hypothesis that leukotriene detection in BALF is a feature common to ALI, etiology notwithstanding.

Proposed role for leukotrienes in the pathophysiology of multiple systems organ failure

The leukotrienes are a group of biologically active products of arachidonic acid metabolism that have been demonstrated to possess the capability to alter vascular reactivity as well as vascular permeability when applied topically to tissues or infused into the vascular bed of various organs. These biologic effects of the exogenous leukotrienes have led to the speculation that these arachidonic acid metabolites could be important mediators in the pathophysiologic events that culminate in the development of acute lung injury and MSOF. Increased production of inflammatory mediators is undoubtedly a contributing factor to the pathophysiologic events that culminate in MSOF. LTB4 has been shown to be a potent stimulator of neutrophil chemotaxis and adhesion. It is possible that LTB4 may be responsible for initiation or amplification of the inflammatory response in this syndrome. The sulfidopeptide leukotrienes have profound effects on cardiac function, which may be mediated through effects on both coronary blood flow and cardiac contractility. These arachidonic acid metabolites are also capable of altering blood flow to several vascular beds and, when synthesized endogenously and released, may be important in the regulation of the peripheral circulation. Endogenous leukotrienes may have physiologic or pathophysiologic effects on cardiac output and its distribution to peripheral vascular beds or on systemic blood pressure. In view of the observations that alterations in cardiac output, blood pressure, and individual organ blood flow have been demonstrated in the clinical setting of ARDs and MSOF, it is attractive to suggest that the leukotrienes may contribute to the hemodynamic alterations observed in these clinical conditions. In addition to their effects on vascular smooth muscle and the myocardium, the leukotrienes have been shown to increase the permeability of blood vessels. In the case of the sulfidopeptide leukotrienes, the enhanced vascular permeability appears to be a direct effect of these lipoxygenase metabolites. Although LTB4 may directly increase vascular permeability, several lines of evidence suggest that the recruitment and activity of neutrophils is required for edema formation to develop following application of this leukotriene. Enhanced capillary permeability, at least in the pulmonary circulation, is a hallmark of ARDS and is also frequently present in MSOF. The leukotrienes, either through direct effects on vessels or through recruitment of inflammatory cells, are likely contributors to the nonhydrostatic edema associated with these syndromes.

Increased concentrations of leukotrienes in bronchoalveolar lavage fluid of patients with ARDS or at risk for ARDS

The sulfidopeptide leukotrienes (LT) C4 and D4 have been reported to promote the formation of pulmonary edema when administered into the pulmonary circulation of laboratory animals. As a first step in the evaluation of the hypothesis that these leukotrienes participate in the edema formation of the adult respiratory distress syndrome (ARDS), we investigated whether LTC4 and LTD4 were present in the bronchoalveolar lavage (BAL) fluid of patients with ARDS compared to nonsmoker control subjects and to patients with acute respiratory failure exhibiting no radiographic evidence of widespread pulmonary infiltrates but having a clinical predisposition for developing ARDS, i.e., the "at risk" group. Bronchoscopic lavage was performed with sterile 0.9% NaCl on 32 control subjects, nine patients with ARDS, and nine patients "at risk" for ARDS. Leukotrienes were measured in BAL fluid by radioimmunoassay after methanol extraction and HPLC purification of a 20-ml aliquot of the BAL sample. LTC4 and LTD4 (mean +/- SE) increased from 1.1 +/- 0.2 and 1.2 +/- 0.5 ng/lavage in the BAL fluid of control subjects to 6.3 +/- 2.3 and 20.1 +/- 5.9 ng/lavage in patients "at risk" for ARDS and to 12.5 +/- 3.0 and 30.5 +/- 7.8 ng/lavage in patients with ARDS, respectively. The sulfidopeptide LTs correlated with BAL fluid protein content. These results suggest that increased amounts of LTs in BAL fluid are a general finding in patients with ARDS and those "at risk" for ARDS.

Relationship of leukotriene C4 and D4 to hypoxic pulmonary vasoconstriction in dogs

Leukotrienes C4 and D4 have been implicated as possible mediators of hypoxic pulmonary vasoconstriction. To test this hypothesis, the relationship between pulmonary leukotriene (LT) synthesis in response to hypoxia and alterations in pulmonary hemodynamics was evaluated in pentobarbital sodium-anesthetized, neuromuscular-blocked, male, mongrel dogs. A reduction in the fraction of inspired O2 (FIO2) in vehicle-treated animals (n = 12) from 0.21 to 0.10 was associated with increases in LTC4 and LTD4 in bronchoalveolar lavage fluid (BALF). After 30 min of continuous hypoxia, LTC4 and LTD4 increased from control values of 59.4 +/- 10.4 and 91.7 +/- 18.1 ng/lavage to 142.7 +/- 31.8 (P less than 0.05) and 156.3 +/- 25.3 (P less than 0.01) ng/lavage, respectively. Concomitantly, mean pulmonary arterial pressure (Ppa) and pulmonary vascular resistance (PVR) were increased over control by 67 +/- 7 (P less than 0.001) and 62 +/- 7% (P less than 0.001), respectively. In contrast, in animals treated with diethylcarbamazine (n = 5), a leukotriene A4 synthase inhibitor, identical reductions in FIO2 were not associated with increases in LTC4 and LTD4 in BALF, although at the same time period, Ppa and PVR were increased over control by 60 +/- 13 (P less than 0.05) and 112 +/- 31% (P less than 0.05), respectively. These results, therefore, do not support the contention that leukotrienes mediate hypoxic pulmonary vasoconstriction in dogs.

Effects of ibuprofen on the hypoxemia of established ethchlorvynol-induced unilateral acute lung injury in anesthetized dogs

Nonsteroidal anti-inflammatory agents, given prior to induction of unilateral acute lung injury with ethchlorvynol (ECV) in anesthetized dogs, prevent the decreases in systemic oxygen tension (PaO2) which are observed when ECV is given alone. We investigated whether ibuprofen, administered after acute lung injury, would result in improvement in arterial oxygenation. In animals not receiving ibuprofen after unilateral acute lung injury with ECV, PaO2 decreased and venous admixture increased significantly from control values at all experimental time-periods. In those animals receiving ibuprofen, significant decreases in venous admixture were noted. The decrease in PaO2 after ECV administration was significantly less than that observed in animals that did not receive ibuprofen after acute lung injury (p less than 0.05). Ibuprofen had no effect on extravascular lung water. These results demonstrate that in an ECV model of acute lung injury the administration of ibuprofen, after the acute lung injury, results in significant decreases in venous admixture.

Increased leukotriene C4 in ethchlorvynol-induced acute lung injury in dogs

We investigated whether ethchlorvynol (ECV)-induced acute lung injury (ALI) is associated with an increase in leukotriene C4 (LTC4) production. In six pentobarbital sodium-anesthetized dogs, ECV (15 mg/kg iv) introduced into the pulmonary circulation resulted in a 164 +/- 31% increase in extravascular lung water 120 min after ECV administration. Concomitantly, the mean (+/- SE) concentration of LTC4 in arterial plasma measured by radioimmunoassay following 80% EtOH precipitation, XAD-7 extraction and high-pressure liquid chromatography purification was 5.0 +/- 1.3 pg/ml, unchanged from control (pre-ECV) values. In contrast, in pulmonary edema fluid 120 min post-ECV, the LTC4 concentration was 35.2 +/- 10.8 pg/ml, sevenfold greater than those values found in the arterial plasma (P less than 0.01). In six additional dogs, 120 min after unilateral ALI had been induced with ECV (9 mg/kg iv), LTC4 in the bronchoalveolar lavage (BAL) of the uninjured lung was 12.1 +/- 1.5 pg/ml, unchanged from pre-ECV values, whereas, LTC4 in the BAL of the injured lung increased from a control value of 10.2 +/- 1.6 to 24.2 +/- 3.5 pg/ml (P less than 0.01) 120 min after ECV administration. These results demonstrate that, in ECV-induced acute lung injury, LTC4 concentrations in pulmonary edema fluid are considerably greater than those found in arterial plasma in the case of bilateral acute lung injury and significantly greater in the BAL of the injured lung compared with the uninjured lung in the case of unilateral acute lung injury. The results are a necessary first step in support of the hypothesis that leukotrienes participate in the altered permeability of ECV-induced acute lung injury.

Effect of cyclooxygenase inhibition on ethchlorvynol-induced acute lung injury in dogs

In anesthetized dogs ethchlorvynol (ECV, 9 mg/kg) was selectively administered into the right pulmonary circulation to produce unilateral acute lung injury (ALI) characterized by nonhydrostatic pulmonary edema and systemic hypoxemia. To investigate the hypothesis that products of cyclooxygenase activity are mediators of the arterial hypoxemia, but not the edema formation in this injury, animals were pretreated with one of two chemically dissimilar cyclooxygenase inhibitors, indomethacin (5 mg/kg), or ibuprofen (12.5 mg/kg), or vehicle (0.1 M sodium carbonate) prior to the administration of ECV. Pretreatment with either inhibitor prevented the ECV-induced systemic hypoxemia observed in animals pretreated with vehicle (P less than 0.01). Despite this protection of systemic oxygenation, there was no redistribution of blood flow to the uninjured lung following unilateral ECV administration. Cyclooxygenase inhibition prior to ALI did not attenuate the accumulation of lung water. In the ibuprofen group, left atrial pressure increased significantly following ECV administration. We conclude that a product(s) of cyclooxygenase-mediated arachidonic acid metabolism is responsible for the altered vascular reactivity and consequent systemic hypoxemia in this model, but that the edema formation following ECV is not related to cyclooxygenase activity. In addition, ibuprofen, administered prior to the induction of ALI, exhibits properties not shared by indomethacin but is not different in its capacity to attenuate hypoxemia or in its failure to limit edema formation.

Prostaglandin I2 supports blood flow to hypoxic alveoli in anesthetized dogs

In an animal model of unilateral alveolar hypoxia, inhibition of cyclooxygenase activity, estimates of immunoreactive 6-ketoprostaglandin F1 alpha (6-keto-PGF1 alpha), and administration of prostaglandin I2 (PGI2) were used to evaluate the hypothesis that endogenous PGI2 opposes hypoxic pulmonary vasoconstriction, thereby producing redistribution of blood flow to hypoxic alveoli and reductions in systemic PO2. In anesthetized dogs, one lung was ventilated with 100% N2 and the other with 100% O2. Thermal dilution coupled with electromagnetic flow measurements permitted estimates of blood flow to each lung. Indomethacin or meclofenamate reduced flow to the N2-ventilated lungs (P less than 0.05) and increased systemic PO2 (P less than 0.05). Simultaneously, aortic concentrations of immunoreactive 6-keto-PGF1 alpha decreased 63 +/- 8% (P less than 0.001). Following cyclooxygenase inhibition, incremental doses of PGI2 (0.01, 0.025, and 0.10 micrograms X kg-1 X min-1) increased flow to the N2-ventilated lungs and reduced systemic PO2 (P less than 0.001) without affecting mixed venous PO2. These results suggest that systemic PO2 was reduced because of increased venous admixture. We conclude that PGI2 attenuates hypoxic vasoconstriction which allows flow to be maintained to hypoxic alveoli, resulting in reduced systemic PO2.

Unilateral acute lung injury induced by ethchlorvynol in anesthetized dogs

Ethchlorvynol (ECV) was used to induce unilateral acute lung injury in anesthetized dogs. Measurements of extravascular thermal volume by double-indicator (thermal-dye) dilution with and without left main pulmonary arterial occlusion permitted sequential estimates of extravascular lung water (EVLW) for each lung. Determinations of EVLW by thermal-dye and gravimetric methods were highly correlated (r = 0.80). ECV (9-15 mg/kg) administered into the right pulmonary circulation produced progressive increases in right lung EVLW, which by 120 min post-ECV was increased 152 +/- 22% (SE) over control (P less than 0.001). Left lung EVLW remained unchanged. Similarly, right, but not left, peak airway pressure was increased. Thermal dilution, coupled with electromagnetic methods, permitted estimates of blood flow to each lung. Despite redistribution of flow to the uninjured lung, systemic PO2 decreased (P less than 0.001) and venous admixture increased (P less than 0.05), suggesting impaired matching of ventilation and perfusion. In summary, introduction of ECV into one lung produced unilateral acute lung injury. EVLW increased solely in the injured lung as did peak airway pressure. Although there was a partial redistribution of blood flow away from the injured lung to the uninjured one, it was apparently inadequate to prevent impaired oxygenation of the blood.

Differential response of the pulmonary circulation to prostaglandins E2 and F2 alpha in the presence of unilateral alveolar hypoxia

Pulmonary hemodynamic responses to several prostaglandins (PGs) are augmented by hypoxia. To test whether responses to vasoconstrictor PGs would be enhanced in hypoxic areas of the lung, resulting in redistribution of blood flow to well-oxygenated areas, the effects of PGE2 and PGF2 alpha were evaluated in anesthetized dogs with divided ventilation, one lung being ventilated with 100% N2 and the other with 100% O2. Thermal dilution techniques coupled with electromagnetic flow measurements permitted estimates of blood flow to each lung. After indomethacin (5 mg/kg i.v.), PGE2 at 0.01, 0.025 and 0.10 micrograms/kg/min i.v. produced significant increases in flow to the O2-ventilated lung at each dose. In addition, PGE2 increased total pulmonary blood flow, but only at the two larger doses. Concomitantly, systemic arterial PO2 increased from 104 +/- 21 to 138 +/- 26 mm Hg (P less than .001). Identical results were obtained when meclofenamate (5 mg/kg i.v.) was used to block PG synthesis. PGF2 alpha increased total pulmonary blood flow (P less than .01) only at the largest dose (0.10 micrograms/kg/min), but did not redistribute flow or increase PO2. Atrial pacing increased total pulmonary blood flow (P less than .001) and flow to both O2 (P less than .001)- and N2 (P less than .05)-ventilated lungs with no change in PO2. We conclude that PGE2 redistributes pulmonary blood flow to well-oxygenated alveoli through mechanisms not related solely to its vasoconstrictor properties or to its capacity to increase cardiac output.

Effect of ventilatory rate on renal venous PGE2 and PGF2 alpha efflux in anesthetized dogs

In anesthetized laparotomized male mongrel dogs with ventilatory rate set at 10 breath.min-1, tidal volume was adjusted so that control arterial pH and PCO2 were within the normal range for unanesthetized dogs. Control renal venous PGE2 and PGF2 alpha concentrations were comparable to those of unanesthetized dogs, namely, 57 +/- 10 and 114 +/- 18 pg.ml-1, respectively. In contrast, control arterial plasma renin activity (PRA), 6.6 +/- 1.2 ng.ml-1.h-1, was considerably greater than in unanesthetized dogs. Stepwise increases in ventilatory rate increased renal venous PGE2 and PGF2 alpha to 109 +/- 18 and 205 +/- 41 pg.ml-1, respectively. Hyperventilation reduced PCO2 and increased pH and PRA but had no effect on renal blood flow, arterial blood pressure, or arterial PGE2 and PGF2 alpha concentrations. When the ventilatory rate was returned to control levels, pH, PCO2, PRA, and renal venous PGE2 and PGF2 alpha concentrations returned to control. Ventilatory rate or some consequence of altering ventilatory rate is, therefore, a determinant of renal venous efflux of PGE2 and PGF2 alpha. Moreover, it may be a more important determinant of "resting" concentrations of prostaglandins in renal venous blood than anesthesia, laparotomy, or PRA.

Alterations in the fraction of oxygen in inspired gas affect rates of renal prostaglandin E2 synthesis in anesthetized dogs

Reductions of oxygen in inspired gas from 20% to 15%, in anesthetized dogs, reduced arterial PO2 and increased the renal efflux of PGE2 but not PGF2a. Renal blood flow, blood pressure, plasma renin activity as well as arterial pH and PCO2 were unaffected PGs may mediate the renal hemodynamic or excretory consequences of alterations in PO2. In addition, minor variations in PO2 might account, in part, for the variable renal venous PGE2 concentrations reported under basal conditions.

Inhibition of prostaglandin synthesis by indomethacin augments the renal vasodilator response to bradykinin in the anesthetized dog

It has been proposed that the increase in renal blood flow (RBF) produced by bradykinin (BK) is mediated or amplified by the intrarenal generation of prostaglandins. The present investigation was designed to explore these relationships further. In anesthetized dogs, the renal arterial infusion of BK (100 ng/kg per min), prior to the intravenous administration of indomethacin, produced a 93 +/- 14% increase in RBF and an increase in the renal venous concentration of a prostaglandin E-like substance ("PGE") from 51 +/- 23 to 235 +/- 73 pg/ml as determined by bioassay. Following indomethacin (5 mg/kg), the same dose of BK produced a 151 +/- 18% increase in RBF (P less than 0.001 compared to the preindomethacin increase) and the concentration of "PGE" remained largely below the threshold of sensitivity of the bioassay system. In three experiments, a highly sensitive and specific radioimmunoassay technique was used to obtain better quantitative estimates of concentrations of E2-like ("PGE2") and F2alpha-like ("PGF2alpha") substances so that determinations of renal efflux could be made. Thus, prior to indomethacin, BK administration increased RBF by 142 +/- 39 ml/min and was associated with a 26-fold increase in renal efflux of "PGE2" and a 12-fold increase in "PGF2alpha." After indomethacin, the effluxes of both "PGE2" and "PGF2alpha" decreased to negligible levels and were not influenced by BK infusion, although RBF increased by 225 +/- 75 ml/min. These results are not compatible with the hypothesis that intrarenal prostaglandins mediate or amplify the renal vasodilator response to BK.

Furosemide exhibits physicochemical properties similar to prostaglandins and interferes with their bioassay

The parallel pharmacological assay system has been of inestimable value in the identification and quantification of prostaglandins. A greater degree of specificity is conferred on the bioassay system by the preparative techniques of extraction and chromatography, for substances which might interfere with the assay of prostaglandins are usually eliminated by these procedures. It would be highly unlikely for a substance to survive the extraction procedure, to exhibit chromatographic properties similar to the prostaglandins and to interfere with their bioassay. We report here such an occurrence. Thus, furosemide not only survived the acidic lipid extraction, but, in addition, it exhibited chromatographic properties identical to those of the prostaglandins of the E series and finally, furosemide, at a dose which in and of itself was without effect on the assay tissues, inhibited the responses to prostaglandins. The quantification of prostaglandins by radioimmunoassay was, on the other hand, not altered by the presence of furosemide. This latter method appears to be the one of choice when interrelations between prostaglandins and furosemide are being examined.

Metabolism of [2-14C]prostaglandin E1 on passage through the pulmonary circulation

A multiple indicator dilution technique was used to study inactivation of [14C]prostaglandin E1 (PGE1) by isolated cat lungs. A bolus containing [2-14C]PGE1, indocyanine green dye and 3HOH was rapidly introduced into the blood entering the pulmonary artery. Sequential samples of the venous effluent were collected and analyzed for 3H, dye, and 14C. The 14C-labelled compounds were separated by thin-layer chromatography, and the quantity of PGE1 and its metabolites contained in each sample was determined (radiochromatogram scanning). Recovery of the dye and 3H was complete within the sampling period. However, only 79% of the injected 14C emerged from the lung within this period, 36% as [14C]PGE1, and 43% as two less polar metabolites. The mean transit times (including connecting tubing) for the unmetabolized [14C]PGE1 and the dye were 9.1 and 8.7 s, respectively. The apparent mean transit times for 14C in the metabolites were considerably longer, 12.4 and 14.3 s. The 3HOH mean transit time was 10.7 s. These data are compatible with an extraluminal site for PGE1 metabolism.

Vascular responses to prostaglandin F 2 alpha in isolated cat lungs

Changes in perfusion pressure in response to graded doses of prostaglandin F 2 alpha (PGF 2 alpha) were measured during both forward and retrograde perfusion of isolated cat lungs perfused at a constant flow rate. During forward perfusion, PGF 2 alpha produced a dose-dependent increase in pulmonary artery pressure and a decrease in lung fluid volume. During retrograde perfusion, PGF 2 alpha also produced a dose-dependent increase in perfusion pressure; however, the dose required was fivefold greater than that needed to produce an identical change in pressure during forward perfusion. In addition, during retrograde perfusion, the lung fluid volume increased in response to PGF 2 alpha. These results suggest that the major site of activity of PGF 2 alpha is on the arterial side of the pulmonary vascular bed and that inactivation of PGF 2 alpha by the lung occurs primarily distal to this arterial site of vasomotion. The changes in perfusion pressure in response to PGF 2 alpha were markedly dependent on pH and oxygen tension (Po-2), being abolished by severe alkalosis and potentiated by both acidosis and hypoxia. In contrast, neither serotonin nor norepinephrine exhibited such a pH or Po-2 dependency. Since the ratio of the forward response to the retrograde response was not decreased by alterations of pH or Po-2, their influence on the responses appears to be through interaction at the site of vascular activity rather than through alteration of the rate of prostaglandin inactivation.

Dependency of renal blood flow on prostaglandin synthesis in the dog

Inhibition of prostaglandin synthesis in chloralose-anesthetized dogs reduced renal blood flow, and this reduction closely correlated (r=0.92, P <0.01) with a decline in the renal efflux of a substance having the properties of PGE 2 . We used solvent extraction and thin-layer chromatography coupled with parallel bioassay to identify and assay the PGE- and PGF-like substances (expressed as PGE 2 and PGF 2α equivalents). Either of two antiinflammatory acids, indomethacin or meclofenamate, that inhibited conversion of 14 C-arachidonic acid to prostaglandins in renal homogenates decreased the basal concentration of a PGE-like substance in renal venous blood to 0.06 ± 0.02 ng/ml from a mean control value of 0.34 ± 0.10 ng/ml ( P <0.01). This change was associated with a mean reduction in renal blood flow of 45% in spite of increased renal perfusion pressure. Femoral blood flow and cardiac output were variably and insignificantly affected. Changes in the renal efflux of a PGF-like substance induced by indomethacin were unrelated to the decline in renal blood flow. Changes in the efflux of a PGE-like substance from the femoral vascular bed were unrelated to the small and variable changes in femoral blood flow. Extrarenal factors, i.e., humoral, nervous, or cardiopulmonary factors, did not account for the decline in renal blood flow produced by the inhibitors of prostaglandin synthesis, since the inhibitors produced identical effects in the isolated blood-perfused canine kidney. We concluded that PGE 2 participates in maintaining renal vascular tone which heretofore has been ascribed to autonomous, intrinsic renal arteriolar activity.

Release of a prostaglandin E-like substance from canine kidney by bradykinin

Renal vasodilation produced by two dissimilar vasodepressor polypeptides, bradykinin and eledoisin, was correlated with changes in renal venous concentrations of substances having the properties of prostaglandins of the E and F series in anesthetized dogs. Samples of renal venous blood were extracted for acidic lipids, and the prostaglandin E and prostaglandin F zones of the chromatographed extracts were eluted and assayed in vitro for prostaglandins of the E and F series by a parallel bioassay system (sensitivity 0.015 ng/ml blood). During the first 2 minutes of infusion, bradykinin increased the concentration of a prostaglandin E-like substance in renal venous blood from a mean control level of 0.16 ng/ml to 1.05 ng/ml ( P <0.01); this increase occurred simultaneously with the greatest increase in renal blood flow to 432 ml/min from a control value of 282 ml/min. After 12 minutes of bradykinin infusion, the concentration of the prostaglandin E-like substance had decreased to 0.30 ng/ml, and renal blood flow had fallen to 398 ml/min. In contrast, eledoisin infused in equidilator doses did not increase the concentration of the prostaglandin E-like substance. The concentration of prostaglandin F-like substances was not affected by either polypeptide. A transient increase in urine flow occurred during the first 2 minutes of bradykinin infusion only. These results suggest that a prostaglandin E-like substance participates in the renal vasodilator and the diuretic responses to bradykinin.

Differential effect of noradrenaline and renal nerve stimulation on vascular resistance in the dog kidney and the release of a prostaglandin E-like substance

1. The concentrations of prostaglandin E(PGE)- and prostaglandin F(PGF)-like substances in renal venous blood were determined by parallel bioassay of extracts of renal venous effluent before and during adrenergic stimulation of the kidney and were related to simultaneous measurements of renal blood flow and urine flow.

2. When noradrenaline was infused continuously into the renal artery, its initial vasoconstrictor and antidiuretic effects diminished on seven of eight occasions in six dogs. Rapid recovery of renal blood flow and urine flow was invariably associated with increasing concentration in renal venous blood of a substance having the physicochemical, chromatographic and biological properties of a prostaglandin of the E series. In the one instance when rapid early recovery of renal blood flow was not observed the concentration of PGE-like substance was not increased.

3. In contrast, during renal nerve stimulation early rapid recovery of renal blood flow and urine flow did not occur and the concentration of a PGE-like substance in renal venous blood did not increase. The concentration of a PGF-like substance in renal venous effluent did not increase in response to either stimulus.

4. Since PGE2, unlike PGF2α, is a potent renal vasodilator and diuretic, the intrarenal release of this substance by noradrenaline in concentrations similar to those determined for a PGE-like substance (&gt;0·50 ng/ml assayed as PGE2 equivalents) would account for the changes in renal blood flow and urine flow in these experiments when the renal actions of noradrenaline were attenuated.

5. These results support the proposal that renal prostaglandins function in an intrarenal negative feedback control system which regulates antidiuretic and vasoconstrictor systems.

Prostaglandin-like substances appearing in canine renal venous blood during renal ischemia. Their patial characterization by pharmacologic and chromatographic procedures

Renal prostaglandins (PC s ) might mediate an antihypertensive function of the kidney. The blood-superfused organ technique possesses the sensitivity (threshold < 0.4 ng/ml blood) and specificity required for identification of PGs in blood. Induction of unilateral renal ischemia in 14 chloraloseanesthetized dogs reduced renal blood flows from a mean value of 257 to 109 ml/min on the ischemic side and from 250 to 209 ml/min on the contralateral side. Concomitantly, PG-like substances were detected by assay organs in the venous blood of ischemic (13 experiments) and contralateral (11 experiments) kidneys. In one experiment, in a spontaneously hypertensive dog, PGs were not detected during renal ischemia.

Renal venous blood and renal medullary tissue were extracted for acidic lipids and assayed for PG-like substances. Extracts of venous blood collected during renal ischemia and extracts of renal medulla yielded substances with biological activity indistinguishable from PG-like substances or PG standards. Chromatographic characterization of PG-like substances suggests that they are predominantly a mixture of PGE 2 and PGF 2α .