INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue damage may result from a temporary interruption of the blood supply to an organ, which may be augmented by the reintroduction of oxygenated blood; this is the phenomenon of ischemia-reperfusion (IR) injury (1). Clinically, the lung is thought to suffer IR injury in a variety of circumstances, including following the removal of massive pulmonary embolus by surgery or thrombolysis, following surgery involving cardiopulmonary bypass and following lung transplant (2, 3).

Pulmonary IR injury results in endothelial damage and disruption, leading in extreme cases to the development of high permeability pulmonary edema, or the acute respiratory distress syndrome (ARDS). ARDS is associated with increased pulmonary vascular resistance (4) and refractory hypoxemia, though to be attributable to the loss of the reflex of hypoxic pulmonary vasoconstriction (HPV) (5). HPV is thought to be modulated at least in part through endothelial-derived vasoactive substances, including nitric oxide (NO).

In acute lung injury, dysfunction of endothelial cells may be associated with alteration in the production of NO, which is synthesized by NO synthase (NOS) through the conversion of L-arginine to L-citrulline. To date, three different NOS isoforms have been identified: two constitutive, Ca²⁺/calmodulin-dependent isoforms (c-NOS) and one Ca²⁺-independent isoform (i-NOS) regulated at the transcriptional level by cytokines and bacterial endotoxin. Recently, investigators have suggested that i-NOS inhibition may prevent endothelial dysfunction, suggesting that NO may be cytotoxic in certain circumstances (6). Further, NO has also been implicated in mediating certain forms of lung injury, and may be implicated in modulating IR through its ability to form the toxic reactive oxygen species peroxynitrite in the presence of superoxide (7). By contrast, other studies indicate the therapeutic potential of administering exogenous NO or NO donors in IR injury (8). NO may protect against IR injury via the inhibition of neutrophil adhesion to the endothelium, and through its scavenging capacity for reactive oxygen species (ROS) (9, 10). The mechanisms by which NO may be protective and/or injurious in circumstances leading to pulmonary IR injury is not clear. It is possible that overproduction of NO or induction of i-NOS modulates vascular damage, but to our knowledge this possibility has not been explored in experimental IR injury.

The aims of the present study were therefore threefold. First, to investigate the time course of i-NOS mRNA expression and the reciprocal changes of NOS enzyme activity in the isolated, blood-perfused rat lung subjected to IR. Second, in the same model, to study the role of NO in IR injury after short (30 min) and long periods (180 min) of reperfusion. Third, to study the role of i-NOS in IR lung injury by intraperitoneal injection of Salmonella lipopolysaccharide (LPS) to induce i-NOS, followed by 30 min ischemia and 30 min reperfusion.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolated, Blood-perfused Rat Lung Preparation

All experimental protocols were carried out in accordance with the Animals (Scientific Procedures) Act 1986 (UK). The isolated, blood-perfused, and ventilated in situ preparation of rat lungs originally described by Hague and modified by Emery and coworkers (11) was used throughout. Male Wistar rats (275-299 g; Charles River, Kent, UK) were anesthetized with Hypnovel (midazolam 0.6 mg/kg intraperitoneal) and Hypnorm (fentanyl 0.315 mg/ml and fluanisone 10 mg/ ml intramuscular). A tracheostomy was performed and the rat was ventilated mechanically (Harvard Small Animal Ventilator; Harvard, Dartford, Kent, UK) using gas of composition 21% O₂, 5% CO₂, balance N₂ at a rate of 50 breaths/min, tidal volume 2.5 ml and 2.0 cm H₂O positive end expiratory pressure (PEEP). Animals were exsanguinated from the abdominal aorta after heparinization (30 mg/kg) and the blood kept in a reservoir maintained at 40° C by a surrounding water bath. Blood was also used to fill a perfusion circuit consisting sequentially of a bubble trap, side-arm pressure transducer (SensoNor 840; Lectromed Ltd, Herts, UK) and a pulmonary artery cannula. Blood was propelled by a roller pump (MHRE 200; Watson Marlow, Falmouth, UK) at a rate of 15 ml/min and returned to the reservoir from a free-draining left atrial cannula. Three pressure transducers were employed, connected to the side-arms of the pulmonary artery cannula, left arterial cannula and tracheal cannula, respectively. Pressure changes were plotted synchronously using a multi-channel recorder (Multitrace 4P; Lectromed Ltd).

After exsanguination, the anterior thoracic cage of each rat was removed by bilateral thoracotomy. A cannula was placed and secured in the left atrium to drain the blood freely into the reservoir, thereby ensuring a left atrial pressure (P_LA) of zero. A second cannula was inserted into the right main pulmonary artery through a ventriculotomy and tied in place, the ascending aorta being tied off simultaneously. The lungs were perfused with whole blood. Samples of blood were taken anaerobically from the left atrial cannula during the stabilization period and gas tensions and pH measured. Base excess was maintained between 2 and 2 mmol/L by adding small volumes of NaHCO₃. In all cases, after institution of the perfusion circuit, a 30-min equilibrium period was permitted. Ischemic conditions were created by switching off the perfusion pump, and reperfusion established by switching it on. The lungs were kept ventilated throughout the experiment.

Evaluaton of Lung Injury

Extravascular albumin accumulation in lung tissue was used to assess lung injury after ischemia-reperfusion. This technique has been previously used by us and others to evaluate lung injury both experimentally and clinically (12, 13). Briefly, extravascular albumin volume was determined by adding ¹²⁵I-human serum albumin (¹²⁵I-HSA;  500 nCi/lung) after 15 min during an initial 30-min equilibrium period (see protocol). To measure intravascular volume, ¹³¹I-HSA ( 250 nCi/lung) was added to the reservoir at the end of each experiment, and lung and plasma samples were harvested after a further 5 min. ¹³¹I-HSA was prepared according to standard techniques (14). At the end of each experiment, 2 ml of blood perfusate was centrifuged and three plasma samples, each of 100 µl, were collected for the estimation of radioactivity expressed as counts per minute (cpm) with appropriate correction for spilldown (Cobra Autogamma; Canberra Packard, Berks, UK). The lungs were divided into eight approximately equal pieces and the radioactivity contained therein was also determined. Whole lung radioactivity was derived by summing the cpm of all tissue samples. The albumin volume (µl plasma equivalents) for each isotope was determined by the ratio of whole lung cpm to the cpm/µl plasma. The volume of extravascular albumin accumulated in lung tissue was then calculated as follows:

Measurement of Pulmonary Artery Pressure (Ppa)

The pressure transducers were calibrated using a sphygmomanometer (Baumanometer; W. A. Baum Co. Inc., New York, NY) and the pressure was expressed as mm Hg. Ppa was monitored and recorded throughout each experiment and the difference (Ppa) between the lowest and peak Ppa was calculated.

Measurement of NOS Activity

Both Ca²⁺-dependent and Ca²⁺-independent NOS activity were examined in homogenized tissue samples after the different experimental protocols. Intravascular blood was flushed out with Kreb's solution through the pulmonary artery and the lung was blotted dry after each experiment. Lung samples were then weighted, frozen in liquid nitrogen and stored at 80° C until NOS activity was measured by the ability of lung homogenates to covert [³H]L-arginine to [³H]L-citrulline as described in full elsewhere (15). Briefly, lungs were homogenized on ice in Tris (50 mM, pH 7.4), containing the protease inhibitor phenylmethylsulfonyl fluoride (1 mM), in a ratio of 1:5 (wt/vol). Lung homogenates were incubated at room temperature with NADPH (1 mM), calmodulin (300 U/ml), tetrahydrobiopterin (5 µM), L-valine (50 µM), L-arginine (10 mM) and [³H]L-arginine (0.03 µM). Total NOS activity was measured in the presence of Ca²⁺ (2 mM). Ca²⁺-independent NOS activity was measured with the addition of EGTA (1 mM) instead of Ca²⁺. Any NOS-independent conversion of L-arginine to L-citrulline was measured by adding N^G-nitro-L-arginine methyl ester (L-NAME; 1 mM); this conversion was subtracted from the measurement of both Ca²⁺-dependent and -independent NOS activity (pmol of citrulline/mg tissue).

Northern Blot Analysis

Rat i-NOS and GAPDH cDNA probes were generated using RT-PCR as previously described (16). Frozen lung samples were granulated and homogenized. Total RNA was extracted following the method described by Chomczynski and Sacchi (17). mRNA was isolated using a Poly(A) mRNA isolation kit (Promega, Madison, WI). Approximately 5 µg mRNA from each sample was separated in a 1.2% denaturing agarose gel and transferred onto Hybond-N nylon filter (Amersham, Arlington Heights, IL). The filter was incubated at 42° C for at least 4 h in a prehybridization buffer consisting of 50% formamide, 5× concentrated Denhardt's solution, 5× SSC (0.75 M sodium chloride, 75 mM sodium citrate), 5 mM sodium phosphate, 200 µg/ml sonicated denatured salmon sperm DNA and 0.1% sodium dodecyl sulfate (SDS). The filter was hybridized with [³²P]-labeled probes at 0.5 to 2 × 10⁶ cpm/ml at 42° C for 14 to 16 h. The blots were washed twice in 2× SSC/0.1% SDS at 42° C for 15 min, once in 0.5× SSC/0.1% SDS at 50° C for 15 min, in 0.2× SSC/0.1% SDS at 55° C for 15 min, and in 0.1× SSC/0.1% SDS for 20 min. Blots were exposed to Kodak X-OMAT S (Sigma, Poole, UK) film in the presence of an intensifying screen for 2 to 5 d. The expression bands were quantified using laser densitometry (Howtek, Hudson, NH) linked to a computer analysis system (PDI, Huntington Station, NY). Filters were stripped and rehybridized with the [³²P]GAPDH probe. GAPDH was used as an internal standard against which i-NOS mRNA was expressed.

Myeloperoxidase (MPO) Assay

The extent of neutrophil sequestration was quantified by measuring MPO activity (18) in whole lung tissue. The assay was performed in homogenized tissue samples. At the end of each protocol, the great vessels were tied to avoid leakage of intravascular blood. The lung was removed en bloc, blotted dry, frozen in liquid nitrogen and stored at 80° C until assayed. The homogenized sample was centrifuged at 260 × g for 10 min and supernatant ultracentrifuged at 100,000 × g for 1 h. The pellet was resuspended in hexadecyltrimethyl-ammonium bromide (0.5%, pH 5.4) and incubated with equal volumes of 3,3,5,5-tetramethylbenzidine (16 mM) and sodium acetate (0.02 M, pH 5.4) for 5 min. The reaction was initiated by the addition of MPO substrate (H₂O₂) for 5 min and stopped by catalase. MPO activity in the final mixture was assayed by measuring the change in spectrophotometric absorbance (optical density, O.D.) at 690 nm and the result expressed as O.D./g of tissue.

Experimental Protocols

Effect of IR on i-NOS mRNA expression and NOS activity. Animals were subjected to 30 min ischemia before reperfusion. Lungs were harvested after varying periods of reperfusion as follows: reperfusion groups: R0, R30, R60, R90, R120, R180 (after 30 min ischemia and 0, 30, 60, 90, 120, and 180 min reperfusion, respectively). In control experiments, lungs were taken immediately after anesthesia (fresh lung group) or immediately after setting up the isolated lung (C0) and after 240 min continuous perfusion (C180). Lung samples were first frozen in liquid nitrogen immediately after harvesting and stored at 80° C. NOS activity (n = 4) and i-NOS mRNA expression (n = 3) were measured in all groups.

Effect of endogenous nitric oxide on IR lung injury. The effect of endogenous NO on IR was studied using the dual i-NOS and c-NOS inhibitor L-NAME and the selective i-NOS inhibitor aminoguanidine. After 30 min ischemia, animals were subjected to either 30 min (I₃₀R₃₀) or 180 min (I₃₀R₁₈₀) reperfusion. L-NAME (10⁴ M) or aminoguanidine (10⁴ M) were added to the reservoir 5 min before ischemia. D-NAME (10⁴ M) and H₂O were used as controls for L-NAME and aminoguanidine, respectively, throughout the study. Lung injury was measured as extravascular albumin accumulation (n = 5 for each group).

Effect of i-NOS on IR lung injury. The effect of i-NOS on ischemia-reperfusion injury was studied by purposely inducing i-NOS activity through intraperitoneal injection of Salmonella enteritidis LPS (5 mg/ kg). Lung samples were assessed for NOS activity 4 h after LPS (n = 4). Another four animals were subjected to I₃₀R₃₀ 4 h after LPS, and NOS enzyme activity was measured. The effect of i-NOS inhibition on IR lung injury was also studied in these LPS-treated animals. Thus, either L-NAME (10⁴ M, n = 5) or aminoguanidine (10⁴ M, n = 5) was added to the reservoir 5 min prior to ischemia and extravascular albumin accumulation was measured after I₃₀R₃₀. D-NAME and H₂O were used as the controls.

Effects of L-NAME on neutrophil sequestration in IR lung injury. Four groups of animals (each n = 3) were studied: two I₃₀R₁₈₀ groups in which either L-NAME (10⁴ M) or D-NAME (10⁴ M) were added to the reservoir 5 min before ischemia, one C180 group (continuous perfusion for 240 min) and one fresh lung group. Circulating neutrophils were counted in 20-µl samples aspirated from the reservoir prior to starting perfusion and at 60-min intervals during the experiment. Total leukocytes were quantified in a Coulter counter (Model ZM) and differential counts performed on stained (Hema-Gurr, BDH) blood smears. At the end of the experiment, lung MPO activity was assessed.

Statistics

All results are presented as mean ± SEM. Data were analyzed by one-way ANOVA and followed by Newman-Keuls post-test or one sample test where appropriate. p Values less than 0.05 were regarded as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of IR on i-NOS mRNA Expression and NOS Activity

The ratio of i-NOS mRNA/GAPDH in lungs subjected to ischemia and reperfusion increased progressively with time of reperfusion, reaching significance (compared to that in the fresh lung group) at 180 min of reperfusion (one-way ANOVA, p < 0.05, Figure 1A and B). Moreover, the level of i-NOS mRNA in lungs subjected to 30 min ischemia and 180 min reperfusion (R180) was significantly greater than in sham lungs (C180, unpaired t test, p < 0.05). Lung homogenate from the fresh lung group contained detectable levels of NOS activity (pmol of citrulline/mg wet weight) which was completely calcium-dependent (Figure 2A, Table 1). R0, R30, R60, R90, R120, R180, and C180 similarly contained detectable levels of NOS activity that was not significantly different (one-way ANOVA) from that found in lungs from the fresh lung group (Table 1). However, in contrast to other groups, calcium-independent NOS enzyme activity increased significantly in R180 compared to the fresh lung group (one-way ANOVA, p < 0.05, Figure 2B). In addition, the amount of calcium-independent NOS activity was significantly greater than that in sham lungs (C180, unpaired t test, p < 0.05, Figure 2B).

View larger version (33K): [in this window] [in a new window]   Figure 1.   (A) Northern blot autoradiograph of inducible nitric oxide synthase (i-NOS) mRNA expression in rat lungs subjected to different periods of ischemia-reperfusion. Glyceraldehyde phosphate dehydrogenase (GAPDH) mRNA served as internal control. Numeric numbers represent mRNA from 1. fresh lung, 2. C0, 3. R0, 4. R60, 5. R90, 6. R120, 7. R180, 8. C180 (see text for group name description). (B) Graph showing the i-NOS mRNA expression in rat lungs subjected to ischemia-reperfusion at different time points, quantified by optical densitometry and expressed as an optical density ratio of i-NOS/GAPDH. The expression ratio of R180 is significantly increased compared to that of fresh lung group (p < 0.05, ANOVA). Values are mean ± SEM of three lungs per group.

View larger version (23K): [in this window] [in a new window]   Figure 2.   (A) Total NOS enzyme activity in fresh lung, R180 and C180 groups. No significant difference was found between these three groups. (B) The i-NOS enzyme activity of R180 was increased significantly compared to the fresh lung group and C180. Values are mean ± SEM of three lungs per group.

Effect of L-NAME and Aminoguanidine on Pulmonary Artery Pressure and Extravascular Albumin Accumulation in Rat Lungs Subjected to IR

Lungs subjected to 30 min ischemia followed by reperfusion resulted in a transient increase in Ppa. Inhibition of NOS by L-NAME and aminoguanidine did not alter this reperfusion-induced peak pressure change in either 30 min or 180 min reperfusion groups compared with vehicles (D-NAME and H₂O, Table 2). In addition, Ppa remained similar in all groups when compared at hourly intervals over the 180 min of reperfusion (data not shown). Extravascular albumin accumulation in lungs subjected to 30 min ischemia followed by either 30 min reperfusion (106.6 ± 11.6 µl) or 180 min (259.8 ± 40.9 µl) reperfusion was significantly increased compared to the respective controls C30 (continuous perfusion for 120 min, 8.6 ± 2.1 µl) and C180 (continuous perfusion for 240 min, 103.0 ± 19.4 µl). In lungs subjected to I₃₀R₃₀, neither L-NAME nor aminoguanidine significantly influenced extravascular albumin accumulation compared to their respective controls (D-NAME and H₂O, Table 3). In lungs subjected to I₃₀R₁₈₀, treatment with aminoguanidine did not alter the extravascular albumin accumulation compared to the vehicle (H₂O, Table 3). However, treatment with L-NAME significantly increased the extravascular albumin accumulation in I₃₀R₁₈₀ compared with the vehicle (D-NAME, Table 3).

Effect of L-NAME and Aminoguanidine on Pulmonary Artery Pressure and Extravascular Albumin Accumulation in LPS-treated Rat Lungs Subjected to IR

When rats were pre-treated with LPS (5 mg/kg) for 4 h, there was a large increase in the level of NOS activity in the lungs, which was mostly calcium-independent (activity without calcium 107.6 ± 15.23 pmol/mg, and activity with calcium 116.4 ± 10.63 pmol/mg). Thus these animals were used to assess the effects of NO derived from i-NOS on IR lung injury. Similarly, to the observations in lungs from untreated rats, reperfusion of the ischemic lungs from LPS-treated animals resulted in a transient increase in Ppa which was not significantly different from that of untreated animals (data not shown). However, in lungs from LPS-treated animals, L-NAME significantly increased the reperfusion-induced rise in Ppa (Ppa, 26.2 ± 3.8 mm Hg, n = 5) compared to that with D-NAME (Ppa, 7.6 ± 1.4 mm Hg, n = 5) (Figure 3). Not only was Ppa increased with L-NAME but the elevated pressure was sustained for a longer period than with the treatment of D-NAME (Figure 3). Aminoguanidine did not alter the reperfusion-induced pressure increase compared to that seen with H₂O (data not shown). In addition, L-NAME and aminoguanidine significantly increased extravascular albumin accumulation compared to their respective vehicles (D-NAME and H₂O, one sample t test, Table 3).

View larger version (123K): [in this window] [in a new window]   Figure 3.   Typical pulmonary artery pressure of traces in LPS-treated animals subjected to 30-min ischemia followed by 30-min reperfusion with the treatment of D-NAME or L-NAME. Not only was the peak pressure in the L-NAME group higher than in the D-NAME group but the elevated pressure was sustained for a longer period.

Effects of L-NAME on Neutrophil Sequestration in IR Lung Injury

More than 80% of neutrophils circulating at the onset of perfusion subsequently sequestered in the lung during the experiment (Figure 4A). There was no significant difference in the circulating neutrophil numbers between L-NAME, D-NAME and control groups during the 4 h experiment, suggesting that the relatively low flow rate (15 ml/min) determined the number of neutrophils sequestered in the lung. The loss of neutrophils from the circulating blood was reflected in an increase in MPO activity in lung tissue (Figure 4B). Thus, compared to fresh lung group, there was a twofold increase in MPO activity in lungs continuously perfused for 240 min (C180 group). However, this did not increase further in lungs subjected to I₃₀R₁₈₀ in the presence of D-NAME or L-NAME.

View larger version (22K): [in this window] [in a new window]   Figure 4.   Circulating neutrophil numbers and MPO activity in lungs subjected to I30R180. (A) Neutrophils in the reservoir were counted before (T0) and during the experiments. Numbers decreased immediately after starting perfusion, but did not vary between treatment (L-NAME and D-NAME) and no treatment (C180: continuous perfusion for 240 min without ischemia) groups. (B) Lung MPO activity, assessed at the end of the experiments, was significantly increased in the L-NAME , D-NAME, and C180 groups compared to the fresh lung group (*p < 0.05). By contrast, no significant difference was found between L-NAME, D-NAME, and C180 groups. Values are mean ± SEM of three lungs per group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

NO has been reported as being both injurious and protective to tissue. It is cytotoxic by virtue of its ability to impair mitochondrial function, inhibiting DNA synthesis and releasing iron from tissue storage sites (19). Overproduction of NO has also been implicated in many disease states including neurodegenerative disorders, rheumatoid arthritis and septic shock (20, 21). NO is also a neurotransmitter and vasodilator and is cytotoxic to bacteria. It is currently believed that tissue injury from NO overproduction takes place particularly in an oxidant-rich environment due to the formation of toxic peroxynitrite (22). It is, however, not clear whether i-NOS is induced and therefore overproduces NO and contributes to injury in lungs subjected to IR.

Our results show that i-NOS mRNA expression increased progressively over a 180-min reperfusion. The major stimuli inducing i-NOS include bacterial endotoxin, interleukin-1 (IL-1), tumor necrosis factor- (TNF-) and interferon- (IFN-). It is therefore surprising that IR alone may also induce the i-NOS mRNA expression. However, there are several possible mechanisms to explain our results. Two major 5'-flanking regulatory regions have been cloned in the i-NOS gene, one LPS-sensitive and the other IFN--sensitive (23). The LPS-sensitive region contains a binding site for transcription factor NF-B which has been shown to be upregulated during reoxygenation after a period of hypoxia (24). NF-B may also be activated in the presence of ROS which can be rapidly generated during ischemia-reperfusion (25). In addition, NF-B, IL-1 and TNF- can also be generated by ischemia-reperfusion (26). i-NOS mRNA may therefore be indirectly induced by ischemia-reperfusion. Recently, a hypoxia-responsive element (i-NOS-HRE) was described as a novel regulatory element of i-NOS promoter activity and provided the evidence that i-NOS is also inducible by hypoxia (27). This may not be fully applicable to our ventilated lung during ischemia, although measured Pa_O2 was the lowest at the start of reperfusion suggesting possible tissue hypoxia during ischemia (data not shown). It is, however, possible that in similar clinical circumstances following lung transplantation, i-NOS mRNA may be induced by i-NOS-HRE in the donor lung during preservation.

Our data showed an increase in i-NOS mRNA expression associated with an increase in i-NOS enzyme activity over a 180-min reperfusion. A trend of decreased constitutive NOS enzyme activity was also observed. The major mechanism controlling c-NOS is a Ca²⁺/calmodulin system (28). Increased c-NOS enzyme activity may be triggered by the bulky influx of intracellular Ca²⁺ ([Ca²⁺]i) during initial reperfusion, a phenomenon reflected in our experiments (Table 1). The down regulation of c-NOS enzyme activity may be regulated at post-translational or transcriptional levels. Phosphorylation of c-NOS by Ca²⁺/calmodulin-dependent protein kinase C (PKC) has been suggested as a feedback mechanism in regulating NO production (29, 30). Elevated [Ca²⁺]i may thus activate PKC thereby inhibiting c-NOS enzyme activity. TNF- has also been shown to decrease e-NOS mRNA levels by increasing the rate of mRNA degradation (31). In addition, increases in ROS production may selectively injure a component in the receptor/ G-protein complex linking receptor-stimulus coupling to the activation of NOS, thus inhibiting the production of NO (32).

The effect of endogenous NO on I₃₀R₃₀ and I₃₀R₁₈₀ was evaluated by blocking NOS with L-NAME and aminoguanidine. Neither agent altered the early (I₃₀R₃₀) reperfusion injury which has been shown in other studies to be exaggerated in the presence of NO (7, 33). The discrepancy may come from the measurement used for defining lung injury and perhaps the different models tested. In the present study, extravascular albumin accumulation was measured throughout the experiment thereby including the injury attributed to ischemia. We also used autologous blood, which might provide a larger acid-buffering capacity and binding capacity for NO from hemoglobin. The importance of perfusate acid-base status has been demonstrated previously in an isolated, buffer-perfused rat lung model, in that both L-NAME and D-NAME at a concentration of 5 mM significantly decreased IR lung injury (34). However, both agents also significantly decreased perfusate pH to which the protective effect was attributable (35). By contrast, in lungs subjected to I₃₀R₁₈₀, lung injury increased significantly when NOS was inhibited by L-NAME indicating a beneficial effect of endogenous NO to lungs subjected to longer periods of reperfusion. The selective i-NOS inhibitor, aminoguanidine, had no effect in the same preparation, suggesting that up to this time i-NOS enzyme contributed a lesser role to NO production.

Transient post-reperfusion pulmonary hypertension is a common feature of experimental or clinical ischemia-reperfusion lung injury. Previously, we have shown that this pressure change (Ppa) significantly contributed to lung injury (36). However, there was no difference in Ppa after reperfusion in either the L-NAME- or aminoguanidine-treated lungs compared with groups treated with their respective vehicles. This suggests that mechanisms other than hemodynamic dysfunction alone modulate the increased injury seen with NOS blockade. Others have shown that NO attenuates lung injury by inhibiting neutrophil and platelet function. In orthotopically transplanted rat lungs, nitroglycerin (NO donor) significantly increased survival rate and prevented platelet and neutrophil sequestration in lungs, but vasodilators alone did not achieve similar effects (37). We have found that neutrophil depletion attenuated IR lung injury after 180 min (Lu and colleagues, unpublished data) but not following 30-min reperfusion (36), indicating a possible role of neutrophils in modulating this injury at the later stages of reperfusion. We obtained a similar result in the present study, in that blocking NOS by L-NAME significantly increased lung injury in animals with 180-min reperfusion, but not in animals with 30-min reperfusion. This implies that endogenous NO prevents IR lung injury perhaps by inhibiting neutrophil activity. However, the number of sequestered neutrophils and MPO activity were not different between treatment and control groups, suggesting that other mechanisms are operating to induce lung injury.

Because the total NOS enzyme activity did not increase over a 180-min reperfusion in the current study, it would be reasonable to argue that NO was not overproduced, and therefore would be expected to be protective against ischemia-reperfusion lung injury. We studied this by inducing i-NOS enzyme activity using LPS. The i-NOS enzyme activity generated by LPS was about 20-fold higher than in lungs subjected to I₃₀R₁₈₀. In this study, blocking NOS by L-NAME significantly increased extravascular albumin accumulation volume nearly threefold compared with vehicle-treated controls (Table 2) and also significantly increased Ppa. The exact mechanism through which NO prevented IR lung injury in the LPS-treated rats is not clear. However, LPS enhances endothelial production of endothelin, a strong vasoconstrictor, whose vasoactivity might be counterbalanced by NO which was, in this study, mainly produced by i-NOS. LPS is more potent in inducing an inflammatory response than IR alone (e.g., IL-1, IL-8, colony-stimulating factors, tissue factors, etc.), the production of vasoactive substances (e.g., endothelin, PGI₂, PGE₂, PGF_2a, etc.) and most importantly capable of inducing endothelial injury before IR. LPS is also known to increase oxidant activity, which in the presence of NO, generates the more reactive species peroxynitrite. Histologically, there is a marked and diffuse neutrophil infiltration in the pulmonary microcirculation with evidence of interstitial edema and disrupted endothelium in LPS-treated rat lungs (data not shown). It is possible that the presence of endogenous NO produced by i-NOS inhibits neutrophil function, thus protecting the lung from further injury, although there are potential pathophysiological differences between LPS-treated and normal rats undergoing IR.

In conclusion, we have shown that IR alone can upregulate i-NOS mRNA and i-NOS enzyme activity, but that constitutive NOS enzyme activity decreased over 180-min reperfusion. Endogenously produced NO is protective in IR lung injury in both normal and in septic rats. NO is also pivotal in maintaining pulmonary vascular homeostasis in septic rat lungs undergoing ischemia-reperfusion.