Separate effects of ischemia and reperfusion on vascular permeability in ventilated ferret lungs

Acute lung injury and pulmonary vascular permeability: use of transgenic models

Acute lung injury is a general term that describes injurious conditions that can range from mild interstitial edema to massive inflammatory tissue destruction. This review will cover theoretical considerations and quantitative and semi-quantitative methods for assessing edema formation and increased vascular permeability during lung injury. Pulmonary edema can be quantitated directly using gravimetric methods, or indirectly by descriptive microscopy, quantitative morphometric microscopy, altered lung mechanics, high-resolution computed tomography, magnetic resonance imaging, positron emission tomography, or x-ray films. Lung vascular permeability to fluid can be evaluated by measuring the filtration coefficient (Kf) and permeability to solutes evaluated from their blood to lung clearances. Albumin clearances can then be used to calculate specific permeability-surface area products (PS) and reflection coefficients (σ). These methods as applied to a wide variety of transgenic mice subjected to acute lung injury by hyperoxic exposure, sepsis, ischemia-reperfusion, acid aspiration, oleic acid infusion, repeated lung lavage, and bleomycin are reviewed. These commonly used animal models simulate features of the acute respiratory distress syndrome, and the preparation of genetically modified mice and their use for defining specific pathways in these disease models are outlined. Although the initiating events differ widely, many of the subsequent inflammatory processes causing lung injury and increased vascular permeability are surprisingly similar for many etiologies.

Porcine traumatic lung injury model induced by hilum clamping

OBJECTIVE

To establish a temporary pulmonary hilum clamping model for thoracic damage control surgery, as well as to determine the safety time latitude of this manipulation.

METHODS

Pigs were anaesthetised and instrumented with a thermodilution cardiac output catheter. The left pulmonary hilum was clamped with a urethral catheter after thoracotomy, maintained for three different time periods (n=6 for each group), 90min (C90), 120min (C120) and 150min (C150) and then unclamped. Haemodynamic data were recorded and the serum samples were collected for D-dimer detection and other haematological analysis. A 1-cm(3) pulmonary tissue of the left lower lobe was also obtained for histological study before clamping, at the end of clamping and at 0.5, 1, 1.5, 2 and 4h after unclamping.

RESULTS

Postoperative survival rate in each group of the pigs was as follows: 100% (all six) of C90, 83.3% (five of six) of C120, and 33.3% (two of six) of C150. Blood pressure (BP) and heart rate (HR) increased after clamping and gradually declined after unclamping. The animals of C150 group suffered highest BP and HR, respiratory index, pulmonary dynamic compliance and cardiac output. Platelet count showed no significant changes between the C90 and C120 groups, whereas a decline was noticed in the C150 group. Pulmonary vascular resistance increased significantly after pulmonary hilum clamping; when unclamped, there were minor changes in animals of C90 and C120 groups while there was a persistent elevation in the C150 group. An elevated D-dimer was detected in the C150 group, whereas it was normal in the C90 and C120 groups. There was significantly serious inflammatory cell infiltration, perivascular oedema and haemorrhagic infiltration in the C150 group compared with the C90 and C120 groups.

CONCLUSIONS

We established a pulmonary hilum clamping animal model for investigating pulmonary damage. By studying the haemodynamic and lung function changes of three different unilateral pulmonary hilum clamping time, it was determined that 120min was the longest safety time for hilum clamping without lethal pulmonary injury in porcine models.

Animal models of acute lung injury

Acute lung injury in humans is characterized histopathologically by neutrophilic alveolitis, injury of the alveolar epithelium and endothelium, hyaline membrane formation, and microvascular thrombi. Different animal models of experimental lung injury have been used to investigate mechanisms of lung injury. Most are based on reproducing in animals known risk factors for ARDS, such as sepsis, lipid embolism secondary to bone fracture, acid aspiration, ischemia-reperfusion of pulmonary or distal vascular beds, and other clinical risks. However, none of these models fully reproduces the features of human lung injury. The goal of this review is to summarize the strengths and weaknesses of existing models of lung injury. We review the specific features of human ARDS that should be modeled in experimental lung injury and then discuss specific characteristics of animal species that may affect the pulmonary host response to noxious stimuli. We emphasize those models of lung injury that are based on reproducing risk factors for human ARDS in animals and discuss the advantages and disadvantages of each model and the extent to which each model reproduces human ARDS. The present review will help guide investigators in the design and interpretation of animal studies of acute lung injury.

Effects of thermal preconditioning on the ischemia-reperfusion-induced acute lung injury in minipigs

Lung ischemia-reperfusion (I/R) injury plays an important role in many clinical issues. A series of mechanisms after I/R has been uncovered after numerous related studies. Organ preconditioning (PC) is a process whereby a brief antecedent event, such as transient ischemia, oxidative stress, temperature change, or drug administration, bestows on an organ an early or delayed tolerance to further insults by the same or different stressors. In this study, we want to uncover the optimal thermal PC patterns that cause maximal early or delayed protective effect on the subsequent pulmonary I/R with the use of miniature pig model. Twenty-eight 15- to 20-kg weight Lanyu miniature pigs are used and divided into four groups (seven sham operation control [NC], seven PC only [PC], seven I/R [I/R], and seven PC followed by I/R [PC + I/R]). The PC was performed with the animals being anesthetized and, using an alternative hyperthermic (40 degrees C) and normothermic moist air to ventilate their lungs for 15 min, respectively, for 2 cycles, followed by I/R, which consists of 90 min of blocking the perfusion and ventilation of the left lung followed by 240 min of reperfusion. Control animals had a thoracotomy with hilar dissection only. Indicators of lung injury included hemodynamic parameters, blood gas analysis, histopathological (lung pathology, wet/dry weight ratio, myeloperoxidase assay), and molecular biological profiles (interleukin-1beta [IL-1beta], IL-6, tumor necrosis factor-alpha by enzyme-linked immunosorbent assay analysis). Lung tissue heat shock protein 70 (HSP-70) expression was also detected by Western blotting. This model of lung I/R induced significant lung injury with pulmonary hypertension, increased pulmonary vascular resistance, and pulmonary venous hypoxemia at the ischemia side, increased pulmonary tissue injury score and neutrophil infiltration, increased wet/dry ratio, myeloperoxidase assay, tumor necrosis factor-alpha, IL-1beta, and IL-6 assay. This type of thermal PC would not injure the lung parenchyma or tracheal epithelium. Moreover, it could attenuate the I/R-related lung injury, with some of these parameters improved significantly. Increased expression of HSP-70 was also found in the group of PC plus I/R than the I/R only. Less prominent and transient increase in expression of HSP-70 was found in the PC group. We concluded that the intratracheal thermal PC can effectively attenuate I/R-induced lung injury through various mechanisms, including the decrease of various proinflammatory cytokines. The mechanism of its protective effect might be related to the increased expression of HSP-70.

Organ preconditioning: the past, current status, and related lung studies

Preconditioning (PC) has emerged as a powerful method for experimentally and clinically attenuating various types of organ injuries. In this paper related clinical and basic research issues on organ preconditioning issues were systemically reviewed. Since lung injuries, including ischemia-reperfusion and others, play important roles in many clinical results, including thromboembolism, trauma, thermal injury, hypovolemic and endotoxin shock, reimplantation response after organ transplantation, and many respiratory diseases in critical care. It is of interest to uncover methods, including the PCs, to protect the lung from the above injuries. However, related studies on pulmonary PC are relatively rare and still being developed, so we will review previous literature on experimental and clinical studies on pulmonary PC in the following paragraphs.

Effect of NADPH oxidase inhibition on cardiopulmonary bypass-induced lung injury

Cardiopulmonary bypass (CPB) causes acute lung injury. Reactive oxygen species (ROS) from NADPH oxidase may contribute to this injury. To determine the role of NADPH oxidase, we pretreated pigs with structurally dissimilar NADPH oxidase inhibitors. Low-dose apocynin (4-hydroxy-3-methoxy-acetophenone; 200 mg/kg, n = 6), high-dose apocynin (400 mg/kg, n = 6), or diphenyleneiodonium (DPI; 8 mg/kg) was compared with diluent (n = 8). An additional group was treated with indomethacin (10 mg/kg, n = 3). CPB was performed for 2 h with deflated lungs, complete pulmonary artery occlusion, and bronchial artery ligation to maximize lung injury. Parameters of pulmonary function were evaluated for 25 min following CPB. Blood chemiluminescence indicated neutrophil ROS production. Electron paramagnetic resonance determined the effect of apocynin and DPI on in vitro pulmonary endothelial ROS production following hypoxia-reoxygenation. Both apocynin and DPI attenuated blood chemiluminescence and post-CPB hypoxemia. At 25 min post-CPB with Fi(O(2)) = 1, arterial Po(2) (Pa(o(2))) averaged 52 +/- 5, 162 +/- 54, 335 +/- 88, and 329 +/- 119 mmHg in control, low-dose apocynin, high-dose apocynin, and DPI-treated groups, respectively (P < 0.01). Indomethacin had no effect. Pa(O(2)) correlated with blood chemiluminescence measured after drug administration before CPB (R = -0.60, P < 0.005). Neither apocynin nor DPI prevented the increased tracheal pressure, plasma cytokine concentrations (tumor necrosis factor-alpha and IL-6), extravascular lung water, and pulmonary vascular protein permeability observed in control pigs. NADPH oxidase inhibition, but not xanthine oxidase inhibition, significantly blocked endothelial ROS generation following hypoxia-reoxygenation (P < 0.05). NADPH oxidase-derived ROS contribute to the severe hypoxemia but not to the increased cytokine generation and pulmonary vascular protein permeability, which occur following CPB.

Effect of bronchial artery blood flow on cardiopulmonary bypass-induced lung injury

Cardiovascular surgery requiring cardiopulmonary bypass (CPB) is frequently complicated by postoperative lung injury. Bronchial artery (BA) blood flow has been hypothesized to attenuate this injury. The purpose of the present study was to determine the effect of BA blood flow on CPB-induced lung injury in anesthetized pigs. In eight pigs (BA ligated) the BA was ligated, whereas in six pigs (BA patent) the BA was identified but left intact. Warm (37°C) CPB was then performed in all pigs with complete occlusion of the pulmonary artery and deflated lungs to maximize lung injury. BA ligation significantly exacerbated nearly all aspects of pulmonary function beginning at 5 min post-CPB. At 25 min, BA-ligated pigs had a lower arterial Po2at a fraction of inspired oxygen of 1.0 (52 ± 5 vs. 312 ± 58 mmHg) and greater peak tracheal pressure (39 ± 6 vs. 15 ± 4 mmHg), pulmonary vascular resistance (11 ± 1 vs. 6 ± 1 mmHg·l–1·min), plasma TNF-α (1.2 ± 0.60 vs. 0.59 ± 0.092 ng/ml), extravascular lung water (11.7 ± 1.2 vs. 7.7 ± 0.5 ml/g blood-free dry weight), and pulmonary vascular protein permeability, as assessed by a decreased reflection coefficient for albumin (σalb; 0.53 ± 0.1 vs. 0.82 ± 0.05). There was a negative correlation ( R = 0.95, P < 0.001) between σalband the 25-min plasma TNF-α concentration. These results suggest that a severe decrease in BA blood flow during and after warm CPB causes increased pulmonary vascular permeability, edema formation, cytokine production, and severe arterial hypoxemia secondary to intrapulmonary shunt.

Effect of ischemia and reperfusion without airway occlusion on vascular barrier function in the in vivo mouse lung

Ischemia-reperfusion (I/R) lung injury causes increased vascular permeability and edema. We developed an in vivo murine model of I/R allowing measurement of pulmonary vascular barrier function without airway occlusion. The left pulmonary artery (PA) was occluded with an exteriorized, slipknotted suture in anesthetized C57BL/6J mice. The effect of ischemic time was determined by subjecting mice to 5, 10, or 30 min of left lung ischemia followed by 150 min of reperfusion. The effect of reperfusion time was determined by subjecting mice to 30 min of left lung ischemia followed by 30 or 150 min of reperfusion. Changes in pulmonary vascular barrier function were measured with the Evans blue dye (EBD) technique, dual-isotope radiolabeled albumin (RA), bronchoalveolar lavage (BAL) protein concentration, and wet weight-to-dry weight ratio (WW/DW). Increasing left lung ischemia with constant reperfusion time or increasing left lung reperfusion time after constant ischemic time resulted in significant increases in left lung EBD content at all times compared with both right lung values and sham surgery mice. The effects of left lung ischemia on lung EBD were corroborated by RA but the effects of increasing reperfusion time differed, suggesting binding of EBD to lung tissue. An increase in WW/DW was only detected after 30 min of reperfusion, suggesting edema clearance. BAL protein concentrations were unaffected. We conclude that short periods of I/R, without airway occlusion, increase pulmonary vascular permeability in the in vivo mouse, providing a useful model to study molecular mechanisms of I/R lung injury.

Pulmonary vascular permeability and ischemic injury in gelsolin-deficient mice

Gelsolin is a potent actin filament regulatory protein that controls cytoskeletal assembly and disassembly. Because cellular gelsolin deficiency leads to pronounced actin stress fiber formation and defective chemotaxis, and similar cytoskeletal remodeling results in endothelial barrier dysfunction, we hypothesized that gelsolin deficient mice would exhibit increased vascular permeability. To test this hypothesis, we compared baseline lung lavage (BAL) protein concentration, wet/dry weight ratio, and osmotic reflection coefficient for albumin (sigma alb) in gelsolin-deficient (gsn-/-) and C57BL/6 (wild-type) mice. In addition, we assessed lung permeability in response to ischemia by evaluating BAL protein concentration after 4, 8, or 24 h of left pulmonary arterial (LPA) occlusion, and lung wet/dry weight ratio and histology after 24 h of LPA occlusion, in gsn-/- and wild-type animals, as compared with control and sham-operated mice. Baseline measurements revealed that BAL protein concentration was 18-fold higher in gsn-/- than in wild-type mice, whereas sigma alb averaged 0.62 + 0.15 in wild-type, as compared with 0.31 + 0.05 in gsn-/- animals, indicating that gelsolin deficiency caused increased pulmonary vascular permeability. Ischemia increased lung permeability (BAL protein and lung wet/dry weight) in both wild-type and gsn-/- mice. However, whereas the fold-increase in BAL protein concentration was less in gsn-/- mice (2- to 4-fold) as compared with wild-type (22- to 34-fold), the duration of ischemia-induced permeability changes was prolonged. Lung wet/dry weight and gross histology following ischemia were comparable in wild-type and gsn-/- animals. These data suggest that gelsolin significantly contributes to maintenance of vascular barrier function in the lung.

Differential regulation of diverse physiological responses to VEGF in pulmonary endothelial cells

The mechanisms responsible for the divergent physiological responses of endothelial cells to vascular endothelial growth factor (VEGF) are incompletely understood. We hypothesized that VEGF elicits increased endothelial permeability and cell migration via differential activation of intracellular signal transduction pathways. To test this hypothesis, we established a model of VEGF-induced endothelial barrier dysfunction and chemotaxis with bovine pulmonary endothelial cells. We compared the effects of VEGF on transendothelial electrical resistance (TER), actin cytoskeletal remodeling, and chemotaxis of lung endothelial cells and then evaluated the role of the mitogen-activated protein kinases (MAPKs) p38 and extracellular signal-regulated kinase (ERK)1/2 in VEGF-mediated endothelial responses. The dose response of pulmonary arterial and lung microvascular endothelial cells to VEGF differed when barrier regulation and chemotaxis were evaluated. Inhibition of tyrosine kinase, phosphoinositol 3-kinase, or p38 MAPK significantly attenuated VEGF-mediated TER, F-actin remodeling, and chemotaxis. VEGF-mediated decreased TER was also significantly attenuated by inhibition of ERK1/2 MAPK but not by inhibition of fetal liver kinase-1 (flk-1) or Src kinase. In contrast, VEGF-mediated endothelial migration was not attenuated by ERK1/2 inhibition but was abolished by inhibition of either flk-1 or Src kinase. These data suggest potential mechanisms by which VEGF may differentially mediate physiological responses in vivo.

Short-term "preconditioning" with inhaled nitric oxide protects rabbit lungs against ischemia-reperfusion injury

BACKGROUND

Pulmonary edema, owing to an impairment of microvascular barrier function, is an important feature in lung ischemia/reperfusion (IR) injury. Inhalation of nitric oxide (NO) during the period of reperfusion has previously been shown to reduce this leakage response.

METHODS

We investigated the impact of short-term (30 min) low-dose (10 ppm) pre-ischemic NO inhalation on IR injury in buffer-perfused rabbit lungs, subsequently undergoing 210 min of warm, anoxic-ventilated ischemia.

RESULTS

Far-reaching suppression of the leakage response, reflected by manifold increased capillary filtration coefficients and edema formation, was noted in lungs with pre-ischemic NO administration, corresponding to the beneficial effect of NO inhalation during reperfusion. The effect of NO pre-exposure was not related to vasodilation, because microvascular pressures were unchanged, and was mimicked by pre-ischemic intravascular administration of sodium nitroprusside with subsequent washout of this agent. NO inhalation during reperfusion, but not pre-ischemic, short-term NO administration, provoked a manifold increase in the accumulation of guanosine 3',5'-cyclic monophosphate (cGMP) in the perfusate. The cGMP-analogue, 8-Br-cGMP, mimicked the anti-edematous effect of NO when present during reperfusion, but pre-ischemic, short-term administration of 8-Br-cGMP provided only limited protection. The guanylate cyclase-inhibitor, 1H-[1, 2, 4]-Oxadiazolo-[4,3-a]-quinoxalin-1-one (ODQ), largely antagonized the beneficial effects of NO inhalation during reperfusion but had only minor influence on the effect of NO pre-exposure.

CONCLUSIONS

"Preconditioning" of the lung vasculature with short-term NO administration maintains endothelial integrity in a subsequent ischemia/reperfusion maneuver, with nonvasodilatory and non-cGMP-related mechanisms suggested to be largely responsible. This finding may offer interesting perspectives for donor management in clinical lung transplantation.

Microtubule disassembly increases endothelial cell barrier dysfunction: role of MLC phosphorylation

Endothelial cell (EC) barrier regulation is critically dependent on cytoskeletal components (microfilaments and microtubules). Because several edemagenic agents induce actomyosin-driven EC contraction tightly linked to myosin light chain (MLC) phosphorylation and microfilament reorganization, we examined the role of microtubule components in bovine EC barrier regulation. Nocodazole or vinblastine, inhibitors of microtubule polymerization, significantly decreased transendothelial electrical resistance in a dose-dependent manner, whereas pretreatment with the microtubule stabilizer paclitaxel significantly attenuated this effect. Decreases in transendothelial electrical resistance induced by microtubule disruption correlated with increases in lung permeability in isolated ferret lung preparations as well as with increases in EC stress fiber content and MLC phosphorylation. The increases in MLC phosphorylation were attributed to decreases in myosin-specific phosphatase activity without significant increases in MLC kinase activity and were attenuated by paclitaxel or by several strategies (C3 exotoxin, toxin B, Rho kinase inhibition) to inhibit Rho GTPase. Together, these results suggest that microtubule disruption initiates specific signaling pathways that cross talk with microfilament networks, resulting in Rho-mediated EC contractility and barrier dysfunction.

Effect of changing vascular volume on measurement of protein reflection coefficient in ischemic lungs

In ischemic organs, the protein reflection coefficient (sigma) can be estimated by measuring blood hematocrit (Hct) and protein after increasing static vascular pressure (P(v)). Our original equation for sigma (J Appl Physiol 73: 2616-2622, 1992) assumed a constant vascular volume during convective fluid flux (). In this study, we 1) quantified the rate of vascular volume change (dV/dt) still present in ischemic single ferret lungs after 20 min of P(v) = 30 Torr and 2) developed an equation for sigma that allowed a finite dV/dt. In 25 lungs, we estimated the dV/dt after 20 min at P(v) = 30 Torr by subtracting from the rate of lung weight gain (W(L)). The relationship between (0.15 +/- 0.02 ml/min) and W(L) (0.24 +/- 0.02 g/min) was significant (R = 0.66, P < 0.001), but the slope was <1 (0.41 +/- 0.10, P < 0.05). dV/dt (0.10 +/- 0.02 ml/min) was similar in magnitude to at 20 min. The modified equation for sigma revealed that a finite dV/dt caused the original sigma measurement to underestimate true sigma. A low sigma, high, high baseline Hct, and long filtration time enhanced the error. The error was small, however, and could be minimized by adjusting experimental parameters.

Effect of time and vascular pressure on permeability and cyclic nucleotides in ischemic lungs

We previously found that increased intravascular pressure decreased ischemic lung injury by a nitric oxide (NO)-dependent mechanism (Becker PM, Buchanan W, and Sylvester JT. J Appl Physiol 84: 803-808, 1998). To determine the role of cyclic nucleotides in this response, we measured the reflection coefficient for albumin (sigma(alb)), fluid flux (), cGMP, and cAMP in ferret lungs subjected to either 45 min ("short"; n = 7) or 180 min ("long") of ventilated ischemia. Long ischemic lungs had "low" (1-2 mmHg, n = 8) or "high" (7-8 mmHg, n = 6) vascular pressure. Other long low lungs were treated with the NO donor (Z)-1-[N-(3-ammoniopropyl)-N-(n-propyl)amino]diazen-1-ium -1, 2-diolate (PAPA-NONOate; 5 x 10(-4) M, n = 6) or 8-bromo-cGMP (5 x 10(-4) M, n = 6). Compared with short ischemia, long low ischemia decreased sigma(alb) (0.23 +/- 0.04 vs. 0.73 +/- 0.08; P < 0.05) and increased (1.93 +/- 0.26 vs. 0.58 +/- 0.22 ml. min(-1). 100 g(-1); P < 0.05). High pressure prevented these changes. Lung cGMP decreased by 66% in long compared with short ischemia. Lung cAMP did not change. PAPA-NONOate and 8-bromo-cGMP increased lung cGMP, but only 8-bromo-cGMP decreased permeability. These results suggest that ischemic vascular injury was, in part, mediated by a decrease in cGMP. Increased vascular pressure prevented injury by a cGMP-independent mechanism that could not be mimicked by administration of exogenous NO.

Effect of the NADPH oxidase inhibitor apocynin on ischemia-reperfusion lung injury

Apocynin (4-hydroxy-3-methoxy-acetophenone) inhibits NADPH oxidase in activated polymorphonuclear (PMN) leukocytes, preventing the generation of reactive oxygen species. To determine if apocynin attenuates ischemia-reperfusion lung injury, we examined the effects of apocynin (0.03, 0.3, and 3 mM) in isolated in situ sheep lungs. In diluent-treated lungs, reperfusion with blood (180 min) after 30 min of ischemia (ventilation 28% O(2), 5% CO(2)) caused leukocyte sequestration in the lung and increased vascular permeability [reflection coefficient for albumin (sigma(alb)) 0.47 +/- 0.10, filtration coefficient (K(f)) 0.14 +/- 0.03 g. min(-1). mmHg(-1). 100 g(-1)] compared with nonreperfused lungs (sigma(alb) 0.77 +/- 0. 03, K(f) 0.03 +/- 0.01 g. min(-1). mmHg(-1). 100 g(-1); P < 0.05). Apocynin attenuated the increased protein permeability at 0.3 and 3 mM (sigma(alb) 0.69 +/- 0.05 and 0.91 +/- 0.03, respectively, P < 0. 05); K(f) was decreased by 3 mM apocynin (0.05 +/- 0.01 g. min(-1). mmHg(-1). 100 g(-1), P < 0.05). Diphenyleneiodonium (DPI, 5 microM), a structurally unrelated inhibitor of NADPH oxidase, worsened injury (K(f) 0.32 +/- 0.07 g. min(-1). mmHg(-1). 100 g(-1), P < 0.05). Neither apocynin nor DPI affected leukocyte sequestration. Apocynin and DPI inhibited whole blood chemiluminescence and isolated PMN leukocyte-induced resazurin reduction, confirming NADPH oxidase inhibition. Apocynin inhibited pulmonary artery hypertension and perfusate concentrations of cyclooxygenase metabolites, including thromboxane B(2). The cyclooxygenase inhibitor indomethacin had no effect on the increased vascular permeability, suggesting that cyclooxygenase inhibition was not the explanation for the apocynin results. Apocynin prevented ischemia-reperfusion lung injury, but the mechanism of protection remains unclear.

Oxygen-independent upregulation of vascular endothelial growth factor and vascular barrier dysfunction during ventilated pulmonary ischemia in isolated ferret lungs

Vascular endothelial growth factor (VEGF) is a potent mediator of endothelial barrier dysfunction, and is upregulated during ischemia in many organs. Because ventilated pulmonary ischemia causes a marked increase in pulmonary vascular permeability, we hypothesized that VEGF would increase during ischemic lung injury. To test this hypothesis, we measured VEGF expression by Northern and Western blot analysis in isolated ferret lungs after 45 (n = 12) or 180 (n = 12) min of ventilated (95% or 0% O(2)) ischemia. Pulmonary vascular permeability, assessed by measurement of osmotic reflection coefficient for albumin (sigma(alb)), was evaluated in the same lungs, as was expression of the transcription factor, hypoxia-inducible factor (HIF)-1alpha. Distribution of VEGF as a function of ischemic time and oxygen tension was also evaluated by immunohistochemical staining in separate groups of lungs (n = 3). VEGF messenger RNA (mRNA) increased 3-fold by 180 min of ventilated ischemia, independent of oxygen tension. VEGF protein increased in parallel to mRNA. Immunohistochemical staining demonstrated the appearance of VEGF protein along alveolar septae after 180 min of hyperoxic ischemia, and after 45 or 180 min of hypoxic ischemia. sigma(alb) was not altered by 45 min of hyperoxic ischemia (0.69+/-0.09 versus 0.50+/-0.12, respectively), but decreased significantly after 180 min of hyperoxic ischemia and after 45 and 180 min of hypoxic ischemia (0.20+/-0.03, 0.26+/-0.08, and 0.23+/-0.03, respectively; P<0.05). HIF-1alpha mRNA increased during both hyperoxic and hypoxic ischemia, but HIF-1alpha protein increased only during hypoxic ischemia. These results implicate VEGF as a potential mediator of increased pulmonary vascular permeability in this model of acute lung injury.

Segmental microvascular permeability in ischemia-reperfusion injury in rat lung

Segmental microvascular permeabilities were measured using pre- and postalveolar vessel capillary filtration coefficient (Kfc) values (ml. min-1. cmH2O-1. 100 g-1) in isolated rat lungs subjected to ischemia-reperfusion (I/R). Total Kfc values measured in flowing and nonflowing lungs were highly correlated (r = 0.98, P < 0.0001). Kfc values were then measured in another group of lungs under no-flow conditions when airway pressure was increased to 20 cmH2O and either the arterial or venous pressure was elevated to 7-8 cmH2O to measure the prealveolar and postalveolar Kfc values. Control total and postalveolar Kfc values were 0.0225 +/- 0.001 and 0.0219 +/- 0.001 ml. min-1. cmH2O-1. 100 g-1, respectively, and the prealveolar permeability was extremely small (0.00003 +/- 0.00005 ml. min-1. cmH2O-1. 100 g-1). Kfc values were again made in nonflowing lungs that had been subjected to 45 min of ischemia followed by 30 min of reperfusion. After I/R, the total membrane Kfc increased 10-fold to 0.2597 +/- 0.006 ml. min-1. cmH2O-1. 100 g-1, the prealveolar Kfc increased to 0.0677 +/- 0.003 ml. min-1. cmH2O-1. 100 g-1, and the postalveolar Kfc increased to 0.1354 +/- 0.008 ml. min-1. cmH2O-1. 100 g-1 (P < 0.05 for all I/R values). These data indicate that normal solvent microvascular permeability was predominantly postalveolar, and after I/R damage, the postalveolar (venular) permeability comprised 52% of the total, whereas the prealveolar and alveolar vessels comprised only 27 and 23%, respectively, of the total Kfc.

Effect of ventilation on vascular permeability and cyclic nucleotide concentrations in ischemic sheep lungs

Ventilation during ischemia attenuates ischemia-reperfusion lung injury, but the mechanism is unknown. Increasing tissue cyclic nucleotide levels has been shown to attenuate lung ischemia-reperfusion injury. We hypothesized that ventilation prevented increased pulmonary vascular permeability during ischemia by increasing lung cyclic nucleotide concentrations. To test this hypothesis, we measured vascular permeability and cGMP and cAMP concentrations in ischemic (75 min) sheep lungs that were ventilated (12 ml/kg tidal volume) or statically inflated with the same positive end-expiratory pressure (5 Torr). The reflection coefficient for albumin (sigmaalb) was 0.54 +/- 0.07 and 0.74 +/- 0. 02 (SE) in nonventilated and ventilated lungs, respectively (n = 5, P < 0.05). Filtration coefficients and capillary blood gas tensions were not different. The effect of ventilation was not mediated by cyclic compression of alveolar capillaries, because negative-pressure ventilation (n = 4) also was protective (sigmaalb = 0.78 +/- 0.09). The final cGMP concentration was less in nonventilated than in ventilated lungs (0.02 +/- 0.02 and 0.49 +/- 0. 18 nmol/g blood-free dry wt, respectively, n = 5, P < 0.05). cAMP concentrations were not different between groups or over time. Sodium nitroprusside increased cGMP (1.97 +/- 0.35 nmol/g blood-free dry wt) and sigmaalb (0.81 +/- 0.09) in nonventilated lungs (n = 5, P < 0.05). Isoproterenol increased cAMP in nonventilated lungs (n = 4, P < 0.05) but had no effect on sigmaalb. The nitric oxide synthase inhibitor NG-nitro-L-arginine methyl ester had no effect on lung cGMP (n = 9) or sigmaalb (n = 16) in ventilated lungs but did increase pulmonary vascular resistance threefold (P < 0.05) in perfused sheep lungs (n = 3). These results suggest that ventilation during ischemia prevented an increase in pulmonary vascular protein permeability, possibly through maintenance of lung cGMP by a nitric oxide-independent mechanism.

F2-isoprostane generation in isolated ferret lungs after oxidant injury or ventilated ischemia

Pulmonary edema develops when pulmonary blood flow is interrupted, then restored. Because the lung is not always hypoxic when ischemic, mechanisms of pulmonary ischemia-reperfusion injury are likely to differ from systemic organs, where reactive oxygen species generated during reperfusion mediate organ dysfunction. We previously showed that pulmonary vascular permeability of isolated ferret lungs increased prior to reperfusion, if ventilation was maintained while blood flow was impaired. To determine whether reactive oxygen metabolites generated during ischemia mediated ischemic injury, we measured tissue levels of F2-isoprostanes as an index of lipid peroxidation, 30 min after administration of glucose (5 mM)-glucose oxidase (GOX, 0.1 U/ml), or after short (45 min) or long (180 min) ventilated ischemia, in isolated ferret lungs. Osmotic reflection coefficient for albumin (sigma alb), an estimate of vascular protein permeability, was measured in the same lungs. Tissue F2-isoprostanes increased 375% after exposure to glucose-GOX in association with a 42% decrease in sigma alb, and administration of catalase (CAT, 100,000 U) and superoxide dismutase (SOD, 25,000 U) completely attenuated this lipid peroxidation. In contrast, tissue F2-isoprostanes increased only 60% following 45 min of ischemia, then did not increase additionally. sigma alb was not altered by 45 min of ischemia, but decreased 72% following 180 min of ischemia. CAT+SOD did not alter F2-isoprostane formation during ischemia, but partially attenuated vascular injury. These results suggest that tissue levels of F2-isoprostanes reflect lung lipid peroxidation, but that F2-isoprostane generation does not directly increase vascular permeability following ventilated pulmonary ischemia.

Protective effects of intravascular pressure and nitric oxide in ischemic lung injury

Cessation of blood flow during ischemia will decrease both distending and shear forces exerted on endothelium and may worsen ischemic lung injury by decreasing production of nitric oxide (NO), which influences vascular barrier function. We hypothesized that increased intravascular pressure (Piv) during ventilated ischemia might maintain NO production by increasing endothelial stretch or shear forces, thereby attenuating ischemic lung injury. Injury was assessed by measuring the filtration coefficient (Kf) and the osmotic reflection coefficient for albumin (sigmaalb) after 3 h of ventilated (95% O2-5% CO2; expiratory pressure 3 mmHg) ischemia. Lungs were flushed with physiological salt solution, and then Piv was adjusted to achieve High Piv (mean 6.7 +/- 0.4 mmHg, n = 15) or Low Piv (mean 0.83 +/- 0.4 mmHg, n = 10). NG-nitro-L-arginine methyl ester (L-NAME; 10(-5) M, n = 10), NG-nitro-D-arginine methyl ester (D-NAME; 10(-5) M, n = 11), or L-NAME (10(-5) M)+L-arginine (5 x 10(-4) M, n = 6) was added at the start of ischemia in three additional groups of lungs with High Piv. High Piv attenuated ischemic injury compared with Low Piv (sigmaalb 0.67 +/- 0.04 vs. 0. 35 +/- 0.04, P < 0.05). The protective effect of High Piv was abolished by L-NAME (sigmaalb 0.37 +/- 0.04, P < 0.05) but not by D-NAME (sigmaalb 0.63 +/- 0.07). The effects of L-NAME were overcome by an excess of L-arginine (sigmaalb 0.56 +/- 0.05, P < 0.05). Kf did not differ significantly among groups. These results suggest that Piv modulates ischemia-induced barrier dysfunction in the lung, and these effects may be mediated by NO.

Free radical-mediated vascular injury in lungs preserved at moderate hypothermia

BACKGROUND

Early allograft dysfunction remains a frequently encountered problem in clinical lung transplantation. Lung ischemia-reperfusion injury is associated with increased vascular permeability, which may be due in part to oxygen (O2) free radicals. However, it is not clear whether O2 free radicals are produced during ischemia under storage conditions in clinical lung transplantation.

METHODS

Using an isolated ex vivo rabbit lung model, we studied the effects of preservation temperature on pulmonary capillary filtration coefficient (Kf) and lipid peroxidation in rabbit lungs inflated with 100% O2 after preservation with or without the O2 free radical scavenger dimethylthiourea. New Zealand white rabbits weighing 2.7 to 3.1 kg were intubated and ventilated with room air or 100% O2 (tidal volume = 25 mL). After heparinization and sternotomy, the pulmonary artery was flushed with low-potassium-dextran-1% glucose solution (200 mL). The heart-lung block was excised, submerged, and stored for 24 hours at 1 degree or 10 degrees C. After 24-hour preservation, the heart-lung block was suspended from a strain-gauge force transducer and ventilated with room air. The pulmonary artery cannula was connected to a reservoir of hetastarch solution. The lungs were flushed briefly with the hetastarch solution, and the reservoir was raised sequentially at 8-minute intervals to achieve 1.0 to 1.5 mm Hg increments in pulmonary artery pressure. Lung weight gain, airway pressure, pulmonary artery pressure, and left atrial pressure were measured continuously. The slope of steady-state lung weight gain was used to determine Kf (g.min-1.cm H2O-1 x 100 g-1 wet weight).

RESULTS

Twenty-four-hour lung preservation at both 1 degree and 10 degrees C increased Kf. A similar increase in Kf was observed in lungs stored at 1 degree C while inflated with 100% O2. However, a significant increase in Kf was observed when lungs inflated with 100% O2 were stored at 10 degrees C. This increase in Kf was ameliorated by dimethylthiourea. Thiobarbituric acid-reactive substance levels were increased in lungs stored at 10 degrees C while inflated with 100% O2. This finding was eliminated by dimethylthiourea.

CONCLUSIONS

These results indicate that free radical injury occurs during the ischemic phase when lungs are stored at moderate hypothermia while inflated with 100% O2.

Changes in vascular permeability with ischemic time, temperature, and inspired oxygen fraction in isolated rabbit lungs

The capillary filtration coefficient (Kf) is one of the most accurate measures of change in pulmonary vascular permeability and has been used in various models of acute lung injury. To evaluate the isolated effects of ischemia on Kf, we have developed an ex vivo rabbit lung model in which the influences of reperfusion are eliminated. The current study was designed to validate this model by determining the effect of cold flushing with low-potassium-dextran solution containing 1% glucose (LPDG), ischemic time, temperature, and inspired oxygen fraction on Kf. On completion of the ischemic period, the ventilated lungs, with the heart still attached, were suspended from a strain-gauge force transducer. After the lungs were flushed with 50 mL hetastarch solution (6% hetastarch solution with physiologic saline solution), the left atrial drainage cannula was occluded and the pulmonary artery pressure was incrementally increased by elevation of the reservoir. The Kf was calculated as the slope of the line relating the weight gain rate and pulmonary capillary pressure.(ABSTRACT TRUNCATED AT 250 WORDS)