Intestinal Ischemia and Reperfusion Impairs Vasomotor Functions of Pulmonary Vascular Bed

Abstract

Objective

To investigate the effects of intestinal ischemia and reperfusion (I/R) on the pulmonary vascular endothelium and smooth muscle.

Summary Background Data

Respiratory failure is an important cause of death and complications after intestinal I/R. Although the mechanism of respiratory failure in this setting is complex and poorly understood, recent studies of lung injury suggest that endothelial dysfunction may play a significant role.

Methods

A rat model of acute lung injury was studied after 60 minutes of superior mesenteric arterial occlusion followed by either 120 or 240 minutes of reperfusion. The pulmonary vasomotor function was examined in isolated lungs perfused at a constant flow rate.

Results

Sixty minutes of intestinal ischemia followed by 120 or 240 minutes of reperfusion led to a significant reduction in the ability of the pulmonary vasculature to respond to angiotensin II, acetylcholine, and calcium ionophore but not to nitroglycerin. The vasoconstriction response to N^G-nitro-L-arginine methyl ester, which is a measure of basal nitric oxide release, was diminished in the 240-minute reperfusion group. Intestinal I/R was also associated with pulmonary leukosequestration and increased pulmonary microvascular leakage.

Conclusions

Basal and agonist-stimulated release of nitric oxide from the pulmonary vascular endothelium and the ability of pulmonary smooth muscle to contract in response to angiotensin II were impaired by intestinal I/R. Such functional impairment in both pulmonary vascular endothelium and smooth muscle may contribute to the alveolocapillary dysfunction and pulmonary hypertension found in acute lung injury after intestinal I/R.

Respiratory failure is a common cause of death and complications after intestinal ischemia and reperfusion (I/R). ¹ Various cytotoxic inflammatory agents, including cytokines, activated complement, arachidonic acid metabolites, adhesion molecules, and free radicals, have been implicated in the development of respiratory failure. ^2–4 All of these agents are powerful chemoattractants and chemoactivators for neutrophils. They can induce activation, sequestration, and adherence of phagocytic cells at particular target sites, such as the alveolocapillary bed, promoting endothelial cell injury, increased microvascular permeability, and pulmonary hypertension. ^5–8

In addition to its function as an anatomic barrier, the endothelium produces various vasodilators and vasoconstrictors that regulate vascular tone. One of the most important vasodilators is nitric oxide (NO), which also inhibits platelet aggregation and neutrophil sequestration. ⁹ Endothelial cells release NO continuously and help regulate vascular smooth muscle tone and prevent neutrophil adhesion to endothelial cells. ^9,10 Basal NO release contributes to low pulmonary vascular smooth muscle tone by stimulating guanylate cyclase and increasing cyclic guanosine monophosphate production, which induces pulmonary smooth muscle relaxation. ¹⁰

In acute lung injury, increased levels of systemic or local vasoconstricting agonists can increase pulmonary vascular tone. ¹¹ However, impaired pulmonary vasodilatation mechanisms may also contribute to increased pulmonary vascular tone. ⁸ Acute lung injury after intestinal I/R is associated with increased microvascular permeability, diminished pulmonary endothelial cell ATP levels, and an upregulation of inducible NO synthase activity. 6,8,12,13 Although these findings cannot be extrapolated to the entire pulmonary vasculature, it is clear that there is an imbalance of pulmonary artery vasodilatation and vasoconstriction in pulmonary artery rings after intestinal I/R. ⁸ In this study, we hypothesized that intestinal I/R causes endothelial and smooth muscle cell dysfunction in the pulmonary vascular bed. To test this hypothesis, we examined the whole pulmonary vascular bed for vasomotor functions in an isolated perfused lung model after intestinal I/R.

METHODS

All animals were given humane care according to Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health, as well as institutional guidelines.

Animal Model

Male Wistar albino rats weighing 220 to 300 g were kept at 25°C with 12-hour light/dark cycles. They were fed rat chow and given water ad libitum. The rats were weighed and then anesthetized with pentobarbital (30 mg/kg intraperitoneally). Just before laparatomy, 10 mL/kg saline was given intravenously via the penile vein to the animals. Via a midline laparatomy, the superior mesenteric artery (SMA) was carefully isolated at its origin from the abdominal aorta and occluded with an atraumatic microvascular clamp. The collateral branches originating from the celiac axis and the inferior mesenteric artery were also clamped. After 60 minutes of SMA occlusion, all microvascular clamps were removed, and reperfusion was initiated. The existence of intestinal I/R in this model was confirmed in our laboratory by fluorescein angiography in preliminary experiments. Briefly, 10 minutes before declamping, 0.1 mL/kg 10% fluorescein dye was injected intravenously and the intestinal circulation was examined under a Wood’s lamp.

The study consisted of three groups: time-matched, sham-operated animals that underwent laparatomy and dissection of the proximal SMA without occlusion (controls, n = 17); animals that had 60 minutes of ischemia and 120 minutes of reperfusion (I/R 60/120, n = 28); and animals that had 60 minutes of ischemia and 240 minutes of reperfusion (I/R 60/240, n = 26). All analytic procedures were done in living animals only at the end of the I/R period (Fig. 1).

Figure 1. Experimental protocol used to examine the effects of intestinal ischemia and reperfusion on pulmonary vasomotor functions and lung neutrophil accumulation, as well as lung microvascular permeability. The study consisted of a control group, a group subjected to 60 minutes of ischemia and 120 minutes of reperfusion, and a group subjected to 60 minutes of ischemia and 240 minutes of reperfusion.

Isolated Perfused Lung Preparation

At the end of the in vivo study protocol, heparin sulfate (500 USP) was injected into the inferior vena cava and the chest was opened. After exsanguination, tracheotomy and cannulation of the main pulmonary artery trunk via the right ventricle were performed. The pulmonary vasculature was flushed with Krebs-Hanseleit buffer and the lungs were excised en bloc and then suspended in an isolated perfusion system. ¹⁴ The lungs were ventilated with room air at a tidal volume of 10 mL/kg, 2 to 3 cm H₂O positive end-expiratory pressure, and 60 cycles per minute. The lungs were perfused with a constant flow peristaltic pump (Harvard Apparatus Corp., South Natick, MA) at 0.02 mL/g per minute. Lungs were perfused for an average of 70 minutes before the experiment to eliminate circulating blood elements from the vasculature and to allow the lung to reach a stable perfusion pressure. Lung perfusion was carried out with Krebs-Hanseleit solution containing (mmol/L) 138.2 Na⁺, 5 K⁺, 2.5 Ca²⁺, 0.5 Mg²⁺, 123 Cl⁻, 25 HCO₃⁻, 1.2 H₂PO₄⁻, and 11.5 dextrose, with a mixture of 95% O₂ and 5% CO₂, providing a pH of 7.4 at 37°C. The perfusate was not recirculated. Pulmonary perfusion pressure was measured continuously by a pressure transducer (Statham P23 BB, Statham Instruments, Inc., Los Angeles, CA) and recorded on a polygraph (Grass Polygraph, Model 79D, Quincy, MA). The absolute change from baseline was measured when each of the agents was added to the perfusion solution. Baseline values for perfusion pressure were obtained after the equilibration period.

Responsiveness of pulmonary smooth muscle to contractile and relaxant agonists was assessed as a measure of endothelial and smooth muscle function. We tested the smooth muscle contraction mechanism by using angiotensin II, an agent commonly used to evaluate the receptor-dependent smooth muscle contraction. ¹⁴ A partial dose-response curve to angiotensin II was generated to evaluate the receptor-dependent smooth muscle contraction mechanism. Angiotensin II (50 ng/0.1 mL) was injected into the circuit (as a bolus) just proximal to the pulmonary artery catheter. Peak perfusion pressure was measured after injection, and a 10-minute recovery period was allowed. This procedure was repeated for each subsequent challenge of 100 and 200 ng of angiotensin II, with baseline values for perfusion pressure measured before each dose. After recovery from the third angiotensin II challenge, submaximal vasoconstriction (approximately 75% of maximum) was elicited by continuous perfusion of Krebs-Hanseleit solution containing phenylephrine (Phe 2 × 10⁻⁷ mol/L). The vasodilator effects of other drugs were tested in submaximally contracted pulmonary vasculature.

Different levels of pulmonary vasodilatation—via endothelial receptors, endothelium, and smooth muscle—were evaluated by specific vasodilators. The endothelial receptor-dependent vasodilator acetylcholine chloride (Ach), the receptor-independent but endothelial-dependent vasodilator calcium ionophore A23187, and the endothelial-independent, smooth muscle relaxant nitroglycerin were used. Ach and A23187 cause vasodilatation by inducing the release of NO from the endothelium. Therefore, we used this method to test agonist-stimulated NO release from vascular endothelium. As in the angiotensin II protocol, baseline values were obtained in submaximally contracted pulmonary vasculature after the equilibrium period. A partial dose-response curve to different concentrations of Ach (50 ng/0.1 mL, 100 ng/0.2 mL, 200 ng/0.4 mL) was generated to evaluate the endothelial receptor-dependent vasodilatation mechanism. A similar procedure was repeated for each subsequent challenge of 50 ng/0.1 mL and 100 ng/0.2 mL A23187. After a 10-minute recovery period, the new baseline value was measured, followed by the nitroglycerine challenge (50 ng/0.1 mL and 100 ng/0.2 mL). After each drug challenge with subsequent recovery, baseline values for perfusion pressure were obtained. All injections were made in a random sequence, and sufficient time was permitted between agonist injections for pressure to return to baseline values.

To evaluate basal NO release, N^G-nitro-L-arginine methyl ester (L-NAME), a competitive inhibitor of NO synthesis from L-arginine, was added to the perfusion solution. When the basal NO release is prevented by L-NAME, the balance between vasorelaxation and vasoconstriction changes toward vasoconstriction. Therefore, this method is an indirect indicator of basal NO release. ¹⁰ In this study, endothelial-basal NO release was assessed indirectly by measuring L-NAME–induced vasoconstriction. Lungs were perfused with a solution containing L-NAME (1 mmol/L) for 30 minutes. L-NAME elicited an incremental increase in the pulmonary perfusion pressure. When the perfusion pressure reached its steady state, we recorded the value of L-NAME–induced vasoconstriction. At that time, the complete inhibition of NO release by L-NAME was also confirmed by the absence of the relaxation response to acetylcholine (100 ng) and by the presence of the relaxation response to nitroglycerine (50 ng). The changes in perfusion pressure were measured on the recorder and expressed as ΔmmHg. The agents (Sigma, St. Louis, MO) were used with solutions prepared daily in saline from their frozen (−20°C) stock solutions.

Preparation of the Lungs for Myeloperoxidase Assay and Evans Blue Dye Concentration

In another part of the experiment, intestinal I/R-induced pulmonary neutrophil sequestration and microvascular permeability were quantified by measuring myeloperoxidase and the concentration of Evans blue dye in the lung, respectively (see Fig. 1). Just before laparatomy, Evans blue dye (30 mg/kg) was infused intravenously via the penile vein. At the end of the sham operation or intestinal I/R experiment, living rats were heparinized (500 USP). After median sternotomy, the lungs were perfused via the pulmonary artery for 2 minutes with saline at 37°C (0.04 mL/g per minute) using an infusion pump (Becton Dickinson, Brezins, France) to eliminate circulating blood elements from the pulmonary vasculature. Finally, the lungs were excised, externally rinsed with saline, blotted dry, and weighed to evaluate assay results per tissue weight. Tissues were then processed for measurement of myeloperoxidase activity and Evans blue dye concentration.

The right lung, which was used for the myeloperoxidase assay, was frozen at −80°C until analysis. Myeloperoxidase is a hemoprotein that is released extracellularly by activated neutrophils. The lung myeloperoxidase assay is a reliable technique widely used to detect experimentally induced pulmonary sequestration of leukocytes. We used a modification of this assay described by Suzuki et al. ¹⁵ Briefly, tissue samples were homogenized in 1:10 potassium phosphate buffer (50 mM, pH 7.4) using a dounce homogenizer. The homogenate was centrifuged at 15,000 g and the pellet was resuspended in an equal volume of a detergent-containing buffer (50 mM potassium phosphate, pH 6, 0.5% hexadecyltrimethylammonium bromide, 10 mmol/L EDTA). A standard reaction mixture contained 1.6 mmol/L tetramethyl benzidine. The reaction was started by the addition of H₂O₂ to a final concentration of 0.88 mmol/L (0.003%). The rate of the myeloperoxidase-catalyzed oxidation of tetramethyl benzidine was followed by recording the absorbance increase at 655 nm. Considering the initial linear phase of the reaction, we calculated the absorbency change per minute, and the enzyme activity was expressed as the amount of enzyme producing one absorbency change per minute under assay conditions.

Intestinal I/R-induced pulmonary microvascular permeability was quantitated in vivo by measuring the concentration of Evans blue dye in the left lung. Evans blue dye binds strongly to albumin and has been used as a marker of protein extravasation in models of acute lung injury. ¹² This technique compares favorably with methodology using radiolabeled albumin and is more sensitive to microvascular dysfunction than lung weight changes. ^16,17 We used a method that is a modification of the assay described by Turnage et al. ¹² The lung was weighed, placed in an oven, and heated at 90°C for 16 hours. The dried tissues were incubated in 2 mL formamide at 37°C for 24 hours. The dye concentration of the elute was then measured by spectrophotometry at 620 nm. The concentration of Evans blue dye extracted from the lungs was expressed as micrograms of dye per gram wet lung weight.

Statistical Analysis

The data were analyzed for significant differences between groups using one-way analysis of variance and post hoc Scheffe’s test. P < .05 was considered significant. Values given in the text and figures are means ± standard error of the mean (SEM).

RESULTS

Although all control animals survived during the experiment, intestinal I/R led to a significantly high death rate. Death rates for the 60/120 and 60/240 groups were 28% (8/28) and 46% (12/26), respectively (P < .01).

The initial pulmonary perfusion pressure was not different among groups in the isolated lung perfusion (Table 1). However, after 3 hours of pulmonary perfusion, the perfusion pressure was higher in the I/R groups than in controls (P < .01). Intestinal I/R led to a significant reduction in the ability of pulmonary vasculature to respond to angiotensin II. With equal ischemic periods, the longer the duration of reperfusion, the greater the reduction in the ability of pulmonary vasculature to respond to angiotensin II. Submaximal vasoconstriction elicited by phenylephrine was also significantly higher in the I/R groups than controls.

Table 1. PULMONARY PERFUSION PRESSURE DURING THE EXPERIMENT AND VASOCONSTRICTIVE RESPONSE TO PHENYLEPHRINE AND ANGIOTENSIN II

Data are mean ± SE.

*P < .01 vs. controls.

†P < .01 vs. I/R 60/120.

Lungs harvested from both groups of rats subjected to I/R showed prominent endothelial dysfunction. As shown in Figure 2, the vasorelaxation responses to both acetylcholine and A23187 were significantly diminished in both I/R groups. However, the vasorelaxation response to the endothelial-independent dilator nitroglycerin was not significantly altered by I/R, indicating that the vascular smooth muscle response remained normal throughout the reperfusion period.

Figure 2. Changes in pulmonary artery perfusion pressure as a response to different concentrations of various vasodilators. Circle = sham-operated controls (n = 9); square = 60 minutes of ischemia and 120 minutes of reperfusion (n = 12); triangle = 60 minutes of ischemia and 240 minutes of reperfusion (n = 7). Mean ± SEM. (A) Both ischemia and reperfusion groups had significantly decreased endothelium- and receptor-dependent vasodilatation by different concentration of acetylcholine, as reflected by a reduced change in perfusion pressure. *P < .05 vs. controls, †P < .01 vs. controls. (B) Both ischemia and reperfusion groups had significantly decreased endothelium-dependent but receptor-independent vasodilatation by bolus injection of 100 ng calcium ionophore A23187, as reflected by a reduced change in perfusion pressure. ‡P < .01 vs. controls. (C) The relaxation response to nitroglycerin in isolated perfused lungs was not significantly altered by ischemia and reperfusion.

In the 60/120 group, the vasoconstriction response to L-NAME, which is a measure of basal NO release, was not different from that in controls (Fig. 3). However, L-NAME–induced vasoconstriction was significantly diminished in the 60/240 group.

Figure 3. Responses to N^G-nitro-L-arginine methyl ester (L-NAME) (1 mmol) infusion of isolated perfused lungs harvested from the control group (n = 9), the group that underwent 60 minutes of ischemia and 120 minutes of reperfusion (n = 12), and the group that underwent 60 minutes of ischemia and 240 minutes of reperfusion (n = 7). The L-NAME–induced increase in the pulmonary perfusion pressure was significantly diminished in the I/R 60/240 group. Values are mean ± SEM. *P < .01 vs. controls.

Intestinal I/R caused a significant increase in pulmonary myeloperoxidase activity compared with controls (Table 2). Further, myeloperoxidase activity in the 60/240 group was significantly higher than in the 60/120 group. A significant increase was also observed in pulmonary microvascular permeability, as measured by the concentration of Evans blue dye, in the lungs of both I/R groups.

Table 2. LUNG MYELOPEROXIDASE VALUE AND EVANS BLUE DYE CONCENTRATION

Data are mean ± SE.

*P < .05 vs. controls.

†P < .01 vs. controls.

‡P < .05 vs. I/R 60/120.

DISCUSSION

Acute lung injury is a common cause of organ failure accompanying intestinal I/R. ¹ Previous studies report increased pulmonary microvascular permeability, leukosequestration, pulmonary hypertension, histologic evidence of alveolar capillary endothelial cell injury, diminished pulmonary endothelial cell ATP level, upregulated pulmonary inducible NO synthase activity, and pulmonary artery vasomotor dysfunction after intestinal I/R. 6,8,12,13 The results of the present study indicate that intestinal I/R affects the functional activity of the lung vascular wall. The reduced receptor and the endothelial-dependent vasodilatation, the receptor-independent but endothelial-dependent vasodilatation, the diminished basal NO release, and the vasoconstrictive response to angiotensin II all support this theory. These functional changes in pulmonary vascular endothelium and smooth muscle after intestinal I/R are associated with increased pulmonary neutrophil accumulation and microvascular permeability. Our results also demonstrate that with equal ischemic periods, the longer the duration of reperfusion, the greater the impairment of vasomotor functions in the pulmonary vascular bed.

The mechanism of endothelium-dependent dilation in response to Ach and A23187 appears to be mediated by the well-described release of endothelium-derived relaxing factor, now known to be NO. ⁹ However, the response to endothelial-independent vascular smooth muscle dilator nitroglycerin was not altered. This implies that there is marked endothelial dysfunction characterized by diminished agonist- stimulated NO release from the entire pulmonary vascular bed after intestinal I/R. Our results suggest that endothelial injury may play a role in the development of pulmonary hypertension after intestinal I/R. Several studies have found endothelial dysfunction in vessel segments exposed to I/R, including pulmonary, ¹⁸ femoral, ¹⁹ and coronary arteries. ²⁰ In their study assessing the effects of intestinal I/R on pulmonary artery vascular tone, Fullerton et al ²¹ found an impairment of endothelial-dependent vascular smooth muscle relaxation. After reperfusion, the release of toxic mediators, including endothelin peptides, cytokines, free oxygen radicals, and thromboxane A₂, and leukocyte adherence to endothelium may decrease the number of Ach receptors, lead to endothelial cell membrane injury, or alter the biosynthetic pathway and release of NO. ^9,18,22

Relatively little attention has been given to basal NO release from endothelium after reperfusion injury. When the basal NO release is prevented by L-NAME, the balance between vasorelaxation and vasoconstriction changes toward vasoconstriction. Therefore, this method is an indirect indicator of basal NO release. ¹⁰ In this study, we found diminished vasoconstrictor response to L-NAME infusion after intestinal I/R, which indicates that intestinal I/R leads to significant impairment of basal NO release. These findings on the basal release of NO in pulmonary circulation are consistent with the result of Davenpeck et al ¹⁸ in another lung I/R study. These investigators found diminished basal NO release in pulmonary artery rings after I/R of the rabbit lung. In the present study, administration of L-NAME abolished endothelium-dependent vasodilatation to Ach but did not reduce endothelium-independent vasodilatation to nitroglycerin, a finding consistent with previous studies. ¹⁶ The diminished release of basal NO release may be more important than the impairment of agonist-stimulated NO release. Research postulates that basal NO release prevents neutrophil and platelet aggregation and adherence to endothelium. ^9,18 This impairment in basal NO release may contribute to further endothelial injury, increased microvascular permeability, and pulmonary vascular tone, resulting in pulmonary hypertension as a prominent feature of I/R-induced lung injury.

Our results also indicate that intestinal I/R significantly affects pulmonary vascular smooth muscle contraction. Although we did not obtain a dose-response curve, submaximal vasoconstriction elicited by phenylephrine, an alpha-adrenergic receptor agonist, was significantly higher in lungs subjected to intestinal I/R. In studies of canine femoral arteries, the dose-response curves to norepinephrine and methoxamine (the selective agonist of α-adrenoceptor) demonstrated an increased sensitivity of α-adrenoceptors in ischemic vessels. ²³ However, impaired pulmonary vasodilatation may also contribute to the increased sensitivity to circulating vasoconstrictors.

We also found that intestinal I/R significantly reduces the ability of pulmonary vascular smooth muscle to contract in response to angiotensin II. Vasoconstrictive responses to angiotensin II in the pulmonary vascular bed are mediated in part by activation of protein kinase C through angiotensin receptor signaling in the rat. ²⁴ Diminished vasoconstrictive response to angiotensin II has been demonstrated in other lung injury models. Li et al ²⁵ reported that isolated lungs from septic rats exhibited significantly diminished pulmonary vascular contractility to angiotensin II. Hypoxia and reoxygenation reduce the vasoconstrictor response to angiotensin II in isolated perfused lung experiments. ^14,26 In those experiments, free radical scavengers attenuated hypoxia/reoxygenation-induced vascular injury, as determined by diminished angiotensin II response. ¹⁴ In addition to oxygen free radicals, platelet activating factor, released during reperfusion injury, significantly reduces angiotensin II-induced pulmonary vasoconstriction. ^26,27 Chang ²⁷ reported that lungs isolated from rats after lipopolysaccharide administration have reduced vasoconstriction ability in response to angiotensin II and increased vasoconstriction response to platelet activating factor. This finding is similar to our results demonstrating an increased sensitivity to phenylephrine and a diminished response to angiotensin II. These observations suggest that oxygen free radicals and cytokines contribute in varying degrees to smooth muscle dysfunction.

We observed that perfusion pressure increased in the later phases of isolated perfusion, where vasoconstrictor response to angiotensin II was diminished. We believe that the initially undetectable changes in pulmonary vascular tone became apparent as lung perfusion progressed. Although we did not thoroughly evaluate the hemodynamics in the pulmonary vasculature, evidence for increased vascular tone at the beginning of lung perfusion has been cited by studies in other acute lung injury models. Li et al ²⁵ reported that a normal baseline perfusion pressure accompanies increased pulmonary vascular resistance in isolated perfused lung harvested from experimentally septic rats. In that study, they also found decreased pulmonary vascular contractility to both angiotensin II and KCl, despite increased pulmonary vascular resistance. Therefore, baseline perfusion pressure may not reflect the changes in the pulmonary vascular bed, despite impaired endothelial and smooth muscle functions.

Intestinal I/R is associated with a generalized inflammatory response culminating in acute lung edema. Evans blue dye binds strongly to albumin and has been used as a marker of protein extravasation in models of acute lung injury. ^16,17 At the beginning of reperfusion, a high Evans blue dye concentration in the lung reflects intraparenchymal extravasation rather than liquid within the pulmonary air space. ²⁸ In the present study, the lungs of animals sustaining intestinal I/R had two- to fourfold higher Evans blue dye concentration than controls. Any loss of vascular surface area due to pulmonary edema may affect some functions of the pulmonary vasculature. However, our findings regarding pulmonary endothelial and smooth muscle functions cannot be explained by edema alone. If the edema had caused the loss of vascular surface area, we would have observed an altered response to nitroglycerin. The normal response to nitroglycerin, despite the diminished responsiveness to Ach and A23187, shows the presence of functioning vascular surface area. However, research suggests that the pulmonary edema resulting from reperfusion injury occurs because of changes in pulmonary microvascular membrane permeability to both fluids and proteins and also because of an increase in microvascular hydrostatic pressure. Iglesias et al ²⁸ evaluated the forces affecting pulmonary edema during intestinal I/R in an isolated perfused lung model using zone III conditions (i.e., P arterial > P venous > P airway), in which the microvascular membrane surface area is kept constant despite changes in vascular resistance. They reported that pulmonary arterial and capillary pressures within the lungs sustaining intestinal I/R were nearly twice those of controls. In that study, the Evans blue dye content in lungs of animals subjected to intestinal I/R was 48% greater than that of controls. They concluded that pulmonary vasoconstriction is likely to be an important contributing factor for the development of pulmonary edema. Therefore, endothelial and smooth muscle dysfunction, culminating in increased pulmonary artery pressure and a damaged microvascular barrier, contributes to the pulmonary edema after intestinal I/R.

Despite the impaired endothelial functions, the initial pulmonary perfusion pressure was normal and pulmonary vascular vasomotor functions still existed after reperfusion injury. Although the interaction between endothelium and neutrophils ranges from increased permeability to endothelial cell death, hypoxia and reoxygenation may activate neutrophils to detach from the endothelial cell monolayer in human endothelial vein cell cultures. ²⁹ Therefore, endothelial detachment is one of the mechanisms that may explain increased endothelial leak or dysfunction without irreversible cell damage. The results mentioned above confirm that the pulmonary vascular endothelium can undergo a loss of integrity and yet sustain normal metabolic function.

In conclusion, the results of this study suggest that intestinal I/R impairs the vasomotor functions of both pulmonary vascular endothelium and smooth muscle. This functional impairment is characterized by diminished basal and agonist-stimulated NO release from the endothelium and reduced ability of the pulmonary smooth muscle to contract in response to angiotensin II. Such data also suggest an imbalance of pulmonary vasodilatation and vasoconstriction, which may culminate in a net increase in pulmonary vascular smooth muscle tone. The current findings suggest that endothelial dysfunction of the pulmonary vascular bed plays an active role in the development of pulmonary hypertension and the increased microvascular permeability found in acute lung injury after intestinal I/R. A thorough understanding of the molecular mechanisms of reperfusion injury will be helpful in the development of therapeutic options to manage this life-threatening problem.