Abstract
The role of nitric oxide (NO) in lung injury remains unclear. Both beneficial and detrimental roles have been proposed. In this study, we used mutant mice lacking the inducible nitric oxide synthase (iNOS) to assess the role of this isoform in sepsis-associated lung injury. Wild-type and iNOS knockout mice were injected with either saline or Escherichia coli endotoxin (LPS) 25 mg/kg and killed 6, 12, and 24 h later. Lung injury was evaluated by measuring lactate dehydrogenase activity in the bronchoalveolar lavage, pulmonary wet/dry ratio, and immunostaining for nitrotyrosine formation. In the wild-type mice, LPS injection elicited more than a 3-fold rise in lactate dehydrogenase activity, a significant rise in lung wet/dry ratio and extensive nitrotyrosine staining in large airway and alveolar epithelium, macrophages, and pulmonary vascular cells. This was accompanied by induction of iNOS protein and increased lung nitric oxide synthase activity. By comparison, LPS injection in iNOS knockout mice elicited no iNOS induction and no significant changes in lung NOS activity, lactate dehydrogenase activity, lung wet/dry ratio, or pulmonary nitrotyrosine staining. These results indicate that mice deficient in iNOS gene are more resistant to LPS-induced acute lung injury than are wild-type mice.
Acute respiratory distress syndrome (ARDS) and gram-negative sepsis are major causes of mortality in intensive care units treating more than 500,000 patients per year in the United States (1). ARDS is characterized by hypoxemia, pulmonary infiltrates, and the absence of an elevated pulmonary capillary wedge pressure (1). Several clinical conditions have been associated with the development of ARDS, with severe infection being the most common condition. Pathologic changes in ARDS include neutrophil sequestration in the lungs, intravascular coagulation, disruption of pulmonary capillary integrity leading to pulmonary edema, and increased shunt fraction (1). The majority of these pathologic features of human ARDS have also been observed in experimental animals in response to systemic infusion of live bacteria or endotoxin of gram-negative bacteria.
Numerous mediators have been implicated in ARDS and sepsis-associated lung injury such as inflammatory cytokines, platelet activating factor, prostglandins, thromboxanes, and reactive oxygen species (2). There is more recent evidence that nitric oxide (NO) plays a significant role in the pathogenesis of acute lung injury. NO is a second messenger molecule that is synthesized by a group of hemoproteins known as nitric oxide synthases (NOS) (3). In normal lungs, NO is synthesized by two constitutive NOS isoforms, the neuronal (nNOS) and the endothelial (ecNOS) isoforms, which are localized at airway epithelial cells, pulmonary vascular endothelial cells, and noncholinergic, nonadrenergic nerve fibers (4). The third (inducible) isoform of NOS (iNOS) is known to be induced in many cell types in response to inflammatory cytokines, and LPS exposure and has also been identified in large epithelial cells of normal animals and humans (5, 6).
There are conflicting conclusions regarding the nature of NO contribution to the pathogenesis of acute lung injury. A beneficial role for NO is based on the observation that NO inhibits neutrophil migration and cytokine production and that NOS inhibitors worsen acute lung injury (7-9). There is evidence, on the other hand, that induction of iNOS may play a harmful role by directly inducing tissue damage and through the formation of peroxynitrite (10, 11). In addition, inhibitors of NOS activity have also been reported to attenuate lung injury and reduce the formation of peroxynitrite (12, 13). The controversy regarding the role of iNOS in acute lung injury could be attributed to the lack of isoform-specific NOS inhibitors. Most of the available iNOS inhibitors are likely to inhibit the activity of not only other NOS isoforms but other enzymes such as those mediating prostaglandin synthesis (14). Another deficiency of previous studies on the role of iNOS in acute lung injury is that the contribution of other NOS isoforms to enhanced NO production was not evaluated despite evidence of increased NO release by constitutive NOS isoforms in response to cytokines and LPS (15).
In this study, we assessed the role of iNOS in the pathogenesis of LPS-induced lung injury by using mice in which the iNOS gene has been disrupted (knockout). Unlike the use of NOS inhibitors, the iNOS knockout mouse model provides complete and selective elimination of iNOS activity. Three groups of investigators developed iNOS knockout mice and studied the importance of iNOS in immunologic and cardiovascular response to LPS injection (16-18). Failure to induce iNOS expression in these mice has been shown to be either protective (17, 18) or had no influence (16) on LPS-induced mortality. In addition, MacMicking and colleagues (17) reported that the changes in systemic nNOS and ecNOS mRNA expressions in response to LPS injection are similar between wild-type and iNOS knockout mice. Unlike the vascular responses, little is known about LPS-induced organ failure in iNOS knockout mice. Although hepatotoxicity is similar in wild-type and iNOS knockout mice (17), lymphocyte apoptosis appears to be significantly elevated after LPS injection in iNOS knockout mice compared with wild-type mice.
In this study, we evaluated the differences in the degrees of anatomic and biochemical indices of LPS-induced acute lung injury in wild-type and iNOS knockout mice. Our results indicate that LPS injection in wild type mice elicits a significant degree of acute lung injury that is associated with extensive nitrotyrosine formation and induction of iNOS protein. Mice lacking the iNOS gene showed significantly milder degrees of acute lung injury and less extensive pulmonary nitrotyrosine formation than did wild-type mice. Thus, a genetically caused absence of NO formation by iNOS attenuates acute lung injury in the murine model of septic shock.
Materials for NOS activity, LDH assay, and Escherichia coli endo-toxin were obtained from Sigma Chemical (St. Louis, MO). l-[2,3-3H]arginine was obtained from DuPont Inc. (Boston, MA). Immunoblotting apparatus and reagents were obtained from Novex Inc. (San Diego, CA). Polyclonal anti-iNOS, ecNOS, and nNOS antibodies were obtained from Transduction Laboratories (Lexington, KY). Polyclonal antinitrotyrosine antibody was obtained from Upstate Biotechnology Inc. (Lake Placid, NY). ABC immunostaining kit was obtained from Vector Laboratories (Burlingame, CA). ECL kit was obtained from Amersham Canada Inc. (Oakville, ON).
All experiments were approved by the Animal Care Committee of McGill University. F2 B6/129 hybrid iNOS knockout mice were generated as mentioned previously (16). Wild-type B6/129 hybrid mice were purchased from Jackson Laboratories (Bar Harbor, ME) and bred to obtain an F2 generation to serve as experimental control mice. Twelve groups of animals (n = 6 for each group) were studied. Six groups of iNOS knockout mice were injected intraperitoneally with either 25 mg/kg E. coli lipopolysaccharides (serotype 055:B5, KO+LPS) or an equivalent volume of sodium chloride (KO−LPS) and killed 6, 12, and 24 h later. Similarly, six groups (WT+LPS and WT−LPS) of wild-type B6/129 hybrid mice were injected with either E. coli LPS or sodium chloride and killed after 6 , 12, and 24 h. LPS- and saline-injected mice were matched for age, sex, and weight. Mice ranged from 10 wk to 6 mo of age.
At the end of the experimental period, the animals were anesthetized by asphyxiation (dry ice in a dessicator chamber) and their tracheas were exposed. Bronchoalveolar lavage (BAL) was performed by injecting 3 ml of phosphate buffer solution (PBS) at pH 7.4 through a silastic catheter, and fluid was then collected by gentle suction. BAL fluid was centrifuged for 10 min at 1,800 rpm and the supernatant was collected for lactate dehydrogenase (LDH) assay.
After BAL, 3 ml of PBS were then injected directly into the right ventricle in order to clear the pulmonary circulation of blood cells. For immunoblotting experiments, lungs were removed, cleared of extraneous tissue, blotted dry, snap-frozen in liquid nitrogen, and stored at −80° C. For regular staining and immunostaining, lungs were removed en bloc, placed in 10% paraformaldehyde and perfused via a tracheal catheter with 10% paraformaldehyde for 8 h. After fixation, lungs were stored in 6% sucrose solution with sodium azide at 4° C. In a separate set of experiments, we measured lung wet/dry ratios by removing the lungs after 12 h of LPS or saline injection in wild-type and iNOS knockout mice. Lungs were weighed and then dried in an oven at 50° C for 12 h.
BAL LDH activity was measured using a Sigma LDH determination kit (Sigma Chemical). Two hundred microliters of BAL supernatant were added to 2.5 ml Reagent A for 1 min followed by the addition of 100 μl of Reagent B. Absorbance at 340 nm was recorded every minute for 3 min in order to confirm the linearity of the reaction. LDH activity was expressed in U/L. The assay was linear for increasing LDH concentrations using LDH standard (Sigma Chemical). The assay is also linear over the 3-min time period.
To assess the degree of lung damage, the lungs were obtained after 12 h of LPS injection in wild-type and iNOS knockout mice (three animals in each category). We chose a 12-h time period because of our extensive preliminary experiments, which indicated that the degree of pulmonary histologic abnormalities and the intensity of nitrotyrosine staining peak after 12 h of LPS injection, with a gradual decline to normal values within 48 h after LPS injection. Serial slices from apex to base were acquired, and three random sections were selected from each lung. Sections 8 μm thick were stained with hematoxylin-eosin and submitted to a pathologist who was blinded as to the treatment groups. Ten fields were chosen randomly from each section (a total of 30 fields per animal) and were examined at a magnification ×250. A separate grade from 0 to 3 was calculated for each field depending on the presence or absence of the following: cellular infiltration, air-space cellularity and air-space exudate, and hemorrhage. For each variable, a single score was calculated per animal by summing the field scores. Moreover, a total lung injury score was calculated as the sum of the three components.
For nitrotyrosine immunostaining, 8-μm lung sections (see above) were hydrated with PBS solution, treated with xylene, and blocked with PBS containing 10% normal donkey serum. After washing, the slides were incubated with primary polyclonal antinitrotyrosine antibody (2 μg/ml) overnight at 4° C. After three washes, the slides were probed with biotinylated donkey antirabbit secondary antibody followed by treatment with the avidin-biotin-peroxidase complex. Sites of immunoreactions were visualized by immersing sections in a solution of diaminobenzidine and hydrogen peroxide. Counterstaining was performed with hematoxylin. A similar protocol was used for negative control sections, except primary antibody was replaced with rabbit IgG. Specificity of primary antibody was also assessed by preexposure of the antibody to 10 mM of pure nitrotyrosine.
Frozen lung samples were homogenized in six volumes (wt/vol) of homogenization buffer (pH, 7.4; 10 mM HEPES buffer; 0.1 mM EDTA; 1 mM dithioreitol; 1 mg/ml phenylmethylsulfonylfluoride; 0.32 mM sucrose; 10 μg/ml leuopeptin; 10 μg/ml aprotinin; 10 μg/ml pepstatin A). The crude homogenates were centrifuged at 4° C for 15 min at 10,000 rpm. The supernatant (50 μl) was added to prewarmed (37° C) 10-ml tubes containing 100 μl of reaction buffer of the following composition: 50 mM KH2PO4, 60 mM valine, 1.5 mM NADPH, 10 mM FAD, 1.2 mM MgCl2, 2 mM CaCl2, 1 mg/ml bovine serum albumin, 1 μg/ml calmodulin, 10 μM tetrahydrobiopterin, and 25 μl of 120 μM stock l-[2,3-3H]arginine (150 to 200 cpm/pM). The samples were incubated for 30 min at 37° C and the reaction was terminated by the addition of cold (4° C) stop buffer (pH, 5.5; 100 mM HEPES; 12 mM EDTA). To obtain free l-[3H]citrulline for the determination of enzyme activity, 2 ml of Dowex 50w resin (8% cross-linked, Na+ form) were added to eliminate excess l-[2,3-3H]arginine. The supernatant was assayed for l-[3H]citrulline by using liquid scintillation counting. Enzyme activity was expressed in picomoles of l-citrulline produced/ min/mg total protein. Protein concentration was measured by the Bradford technique with bovine serum albumin as standard. NOS activity was also measured in the presence of 1 mM of NG-nitro-l-arginine methyl ester (NOS inhibitor). Total NOS activity was calculated as the difference between that measured in the presence of Ca2+ and calmodulin and that measured in the presence of NG-nitro-l-arginine methyl ester.
The supernatants (80 μg) of lung homogenates (see above) were heated for 15 min at 90° C and then loaded on gradient (4 to 12%) TRIS- Glycine SDS-polyacrylamide electrophoresis. Proteins were electrophoretically transferred onto PVDF membranes and were blocked overnight (4° C) with 5% nonfat dry milk and subsequently incubated with primary monoclonal anti-iNOS, anti-ecNOS, or anti-nNOS antibodies. Lysate of cytokine-activated murine macrophages, human endothelial cells, and rat pituitary were used as positive control, respectively. Specific proteins were detected using horseradish-peroxidase-conjugated antimouse secondary antibody and chemiluminescence reagents provided with the ECL kit (Amersham Canada Inc.). Predetermined molecular weight standards were used as markers.
Data are shown as means ± standard errors of the means. Statistical differences between mean values were detected by Student's t test. The two-way analysis of variance test was employed in order to calculate statistical heterogeneity for mean lung histology scores, BAL LDH values, NOS activity, and wet/dry ratios both between and within treatment groups at different time points after LPS injection.
Within 3 to 4 h of LPS injection, both iNOS knockout and wild-type mice appeared ill when compared with saline-injected mice. This was characterized by decreased spontaneous movement, decreased reaction to threatening stimuli, and a white periconjunctival exudate. None of the animals died during the 24-h experimental period. On gross examination, there was no apparent difference between lung from knockout or from wild-type mice whether or not they were treated with LPS or with saline.
Hematoxylin-eosin staining in wild-type mice injected with LPS revealed widespread alveolar wall thickening caused by edema, interstitial infiltration with neutrophils and macrophages, as well as significant air-space cellularity and exudation (Figure 1, lower panel). Adhesion of neutrophils to endothelial cells of pulmonary vessels was also found in a few fields. We should emphasize that the extent of these histologic abnormalities peaks after 12 h of LPS injection, with a gradual recovery within 48 h of LPS injection. By comparison, iNOS knockout mice injected with LPS also showed interstitial infiltration and air-space cellularity (Figure 1, middle panel); however, the degree of these changes was significantly less than that of the wild type mice (Table 1). Like the wild-type mice, iNOS knockout mice injected with saline showed normal lung histology (Figure 1, upper panel).
Fig. 1. Hematoxylin-eosin staining of lung tissues obtained from iNOS knockout mice injected with saline (top panel ), injected with E. coli endotoxin (middle panel ), and wild-type mice injected with E. coli endotoxin (bottom panel ). Notice that alveolar exudation and interstitial cellular infiltration were found in lungs obtained from endotoxemic wild-type mice (bottom panel ).
[More] [Minimize]| Cellular Infiltration | Air-Space Cellularity | Air-Space Exudate | Lung Injury Score | |||||
|---|---|---|---|---|---|---|---|---|
| Wild-type+LPS | 66.5 ± 2.5†,‡ | 33.5 ± 1.4†,§ | 37.0 ± 3.0‖,§ | 137.1 ± 1.2‖,‡ | ||||
| iNOS KO+LPS | 52.0 ± 2.3§ | 23.0 ± 3.3 | 21.1 ± 2.5 | 96.0 ± 4.4§ | ||||
| iNOS KO−LPS | 22.5 ± 4.5 | 20.0 ± 2.3 | 19.5 ± 1.2 | 62.0 ± 4.4 |
The changes in BAL LDH activity in various groups are illustrated in Figure 2. No significant changes were observed through the experimental periods in the LDH activity of wild-type and of iNOS knockout mice injected with saline. A significant rise was, however, observed in LDH activity of wild type mice after 12 and 24 h of LPS injection (p < 0.01 compared with wild-type LPS mice). By comparison, LPS injection in iNOS knockout mice elicited no significant rise in BAL LDH activity. Hence, LDH activity after 12 and 24 h of LPS injection was significantly higher than that measured at equivalent time in LPS-injected iNOS knockout mice (p < 0.01).
Fig. 2. Mean changes in bronchoalveolar lavage LDH activity in various animal groups. In wild-type mice, a significant rise in LDH activity was detected after 12 and 24 h of LPS injection. There was no such increase in BAL LDH activity in iNOS knockout mice injected with LPS. **, §§p < 0.01 compared with wild-type −LPS and knockout mice +LPS, respectively.
[More] [Minimize]The changes in lung wet/dry ratio measured after 12 h of saline or LPS injection are shown in Figure 3. LPS injection in wild-type mice elicited a significant rise in wet/dry ratio when compared with saline-injected wild type mice (p < 0.01). No significant differences were observed in lung wet/dry ratio between LPS- and saline-injected iNOS knockout mice (Figure 3), and values of wet/dry ratios in both groups were significantly lower than that of LPS-injected wild type mice (p < 0.01).
Fig. 3. Changes in lung wet/dry ratios after 12 h of saline or LPS injection in various groups of animals. **, §§p < 0.01 compared with wild-type −LPS and knockout mice +LPS, respectively.
[More] [Minimize]LPS injection in wild-type mice elicited a significant rise in lung NOS activity, which peaked at 6 h (p < 0.01 compared with WT−LPS, KO+LPS, and KO−LPS) and then declined thereafter (Figure 4). By comparison, there were no significant changes in lung NOS activity in wild-type mice injected with saline or iNOS knockout mice injected with LPS or saline (Figure 4).
Fig. 4. Changes in lung NOS activity in various groups of mice. Notice that LPS injection in wild-type mice elicited a significant rise in lung NOS activity, whereas no significant alterations in lung NOS activity were detected in iNOS KO mice injected with LPS. Neither wild-type nor iNOS KO mice showed any changes in lung NOS activity in response to saline injection. **, *p < 0.01 and 0.05 compared with wild-type −LPS, respectively. §§, §p < 0.01 and 0.05 compared with iNOS KO+LPS, respectively.
[More] [Minimize]A representative immunoblot of lung samples obtained from various experimental groups is shown in Figure 5. In wild-type mice, no iNOS protein was detected in any of the animals injected with saline (control lane). Injection of LPS elicited a prominent iNOS expression within 6 h, with weak expression evident after 12 h. Expression of iNOS completely disappeared after 24 h of LPS injection. Neither saline-injected (control lane) nor LPS-injected iNOS knockout mice showed any evidence of iNOS protein expression (Figure 5). We detected weak ecNOS and nNOS expressions in lung samples obtained from wild-type and iNOS knockout mice, and LPS injection had no significant effects on the expression of these two isoforms (Figure 6).
Fig. 5. A representative immunoblot of lung samples for iNOS expression in animals injected with LPS. Significant iNOS protein was detected after 6 and 12 h of LPS injection in wild-type mice, whereas no iNOS expression was detected in LPS-injected iNOS KO mice. Similarly, no iNOS protein was detected in any of the animals injected with saline (control lane).
[More] [Minimize]
Fig. 6. Representative immunoblots of lung samples for nNOS (top) and ecNOS (bottom) expressions in animals injected with LPS. Notice that the weak expression of nNOS and ecNOS in the control groups of wild-type and iNOS knockout mice remained unchanged after LPS injection.
[More] [Minimize]Immunostaining of lung samples with antinitrotyrosine antibody is shown in Figure 7. In saline-injected wild type mice, weak nitrotyrosine staining was detected in the epithelial cells of large airways (not shown). However, injection of LPS in the wild type mice elicited widespread nitrotyrosine staining in airway and alveolar epithelial cells ( panel A), alveolar and interstitial macrophages ( panel B), and endothelial and smooth muscle cells of pulmonary vessels ( panel C). Proteinaceous alveolar exudate also stained positively for nitrotyrosine ( panel B). By comparison, only weak expression of nitrotyrosine was detected in airway epithelial cells and blood vessels of LPS-injected iNOS knockout mice (Figure 7, panel D). Nitrotyrosine staining was completely eliminated when primary antibody was incubated with 40 mM nitrotyrosine prior to immunostaining (not shown).
Fig. 7. Immunostaining for nitrotyrosine expression in wild-type and iNOS KO mice injected with LPS. In wild-type mice, strong nitrotyrosine staining was detected in airway and alveolar epithelial cells (A), alveolar exudate and interstitial macrophages (B), and pulmonary vasculature (C ). Only weak nitrotyrosine staining was detected in airway epithelial cells of iNOS KO mice (D).
[More] [Minimize]
The main finding of our study was that injection of LPS in wild-type mice elicited acute lung injury as indicated by the significant rise in BAL LDH activity, increased pulmonary wet/dry ratio, and extensive nitrotyrosine formation in various pulmonary cells. These indices of lung injury were accompanied by expression of iNOS protein. By comparison, the extent of LPS-induced acute lung injury and the intensity of nitrotyrosine staining were significantly milder in mice deficient in iNOS gene when compared with wild-type mice.
Animals with selective genetic alterations in a single NOS isoform provide excellent experimental models without the inherent deficiencies of NOS inhibitors. We used in this study mutant mice lacking iNOS gene to assess whether iNOS plays a significant role in the induction of lung injury (16). Three groups of investigators studied the cardiovascular and immunologic response to LPS injection in iNOS deficient mice (16– 18). Although these investigators agree that deficient iNOS expression impairs host defense against Listeria and Leismania major infections, they disagree with respect to the sensitivity of these mice to LPS-induced death. Although MacMicking and colleagues (17) and Wei and associates (18) concluded that the absence of iNOS induction offers protection against death induced by high doses of LPS, Laubach and colleagues (16) reported that iNOS-deficient mice have a resistance to LPS-induced death similar to that of wild-type mice. Our results indicate that both wild-type and iNOS knockout mice were equally resistant to the lethal effects of endotoxemic shock. These differences can be attributed to various factors such as the use of anaesthesia (as in the study of MacMicking and colleagues), which impairs centrally mediated cardiovascular reflexes and, therefore, enhances LPS-induced cardiovascular failure. Another important factor in determining the mortality to LPS injection is the genetic background of various mice strains (19). Whereas Wei and colleagues (18) compared MF1/ 129-derived iNOS knockout mice of unspecified generation with outbred MF1 mice, we and Laubach and colleagues controlled for both the genetic background and the generation. Finally, the difference in the rate of LPS-induced death between various iNOS knockout mice models can be attributed to differences in the dose, the method of administration, and the strain of E. coli endotoxin.
Interestingly, in contrast to the finding of MacMicking and colleagues (17) of a similar level of LPS-induced hepatotoxicity and similar elevations in serum LDH concentration among wild-type and iNOS knockout mice, we found that iNOS mutant mice generated significantly lower BAL LDH activity, pulmonary wet/dry ratios, and pulmonary nitrotyrosine expression after LPS injection compared with wild-type mice. These results suggest that LPS induces a milder form of acute lung injury in iNOS knockout mice than in wild-type mice.
ARDS is a common complication of diverse clinical problems, including trauma and sepsis, and is characterized by pulmonary capillary-alveolar injury and defective gas exchange (1). The findings that injection of LPS in animals induces acute noncardiogenic pulmonary edema and other features of human ARDS suggest that LPS plays a central role in the pathogenesis of sepsis-associated lung injury (20). A number of mediators contribute to the pathogenesis of lung injury in sepsis. Activation of complement and subsequent neutrophil infiltration are essential events in the induction of organ injury. The rise of adhesion molecule expression and the release of inflammatory cytokines such as tumor necrosis factor (TNF) and interleukin-1 (IL-1) are secondary phenomena that lead to the release of local mediators such as platelet-activating factor, prostaglandins, leukotrienes, oxygen radical species, and NO. These local mediators are involved in cell structural damage as well as of capillary endothelial integrity and tissue edema (1).
The notion that NO is involved in acute lung injury is based on the observations that endotoxemia and sepsis are associated with increased pulmonary NO production (6, 7, 21). However, whether augmented pulmonary NO production serves positive or deleterious roles in normal lung function remains a debatable issue. The findings that NO inhibits neutrophil migration and reduces lung injury suggest that NO plays a beneficial role in attenuating acute lung injury (9). Furthermore, Iuvone and colleagues (8) reported that NO inhibits LPS- induced TNF release in vivo and in vitro. NO is also known to inhibit the activity of NF-κB, the essential nuclear factor for cytokine production (22). The observations that administration of iNOS inhibitor (aminoguanidine) increased lung edema and worsened indices of lung injury in mice injected with LPS (7) confirm a beneficial role for iNOS.
There is a growing belief, on the other hand, that excessive NO production by iNOS plays an important role in the induction of lung injury in patients with ARDS. This belief is based on three major observations. Firstly, iNOS expression in many pulmonary cell types coincides with the development of lung injury in septic humans and animals (6, 23, 24). Secondly, infusion of NOS inhibitors attenuates lung injury and microvascular leakages induced by endotoxemia (12, 25). Thirdly, extensive nitrotyrosine staining, the footprint of peroxynitrite formation, has been detected in lung sections of humans and animals with acute lung injury and coincides with iNOS expression (11).
A major shortcoming of studies dealing with the role of iNOS in lung injury has been the reliance on pharmacologic inhibitors. The diversity of modes of action and the presence of numerous molecular targets for NO renders reliance on NOS inhibitors in delineating the role of iNOS quite difficult. Moreover, most of the available selective iNOS inhibitors also influence ecNOS and nNOS, both of which are expressed in numerous pulmonary cell types and play important roles in the regulation of pulmonary blood flow, airway tone, and epithelial cell function (4). These constitutive isoforms and particularly ecNOS are known to be activated by LPS and may contribute to increased pulmonary NO production in sepsis (15, 24). Finally, there is evidence that NOS inhibitors may influence inflammatory processes through inhibition not only of NOS isoforms but of prostaglandin production (14).
Our results indicate that iNOS-deficient mice develop significantly milder acute lung injury after LPS injection than do wild-type mice. Interestingly, the absence of iNOS induction after LPS injection in iNOS-deficient mice was not associated with changes in pulmonary nNOS or ecNOS expressions (Figure 6). This finding is in agreement with that of MacMicking and colleagues (17) regarding nNOS and ecNOS mRNA expression in various organs after LPS injection. Thus, production of NO by the constitutive NOS isoforms that are normally expressed in pulmonary cells does not appear to increase in compensation for iNOS deficiency.
There exist several mechanisms through which relatively high levels of NO produced by iNOS mediate lung injury. Reaction of NO with superoxide anions produces peroxynitrite, which is a highly oxidative species and is capable of nitrating tyrosine residues of numerous proteins, leading to the formation of nitrotyrosine. High levels of nitrotyrosine formation detected by specific antibody has been shown to develop in human acute lung injury (11, 26) and LPS-injected animals (13). Our results confirm that LPS injection elicited extensive nitrotyrosine formation in alveolar macrophages, endothelial and vascular smooth muscle cells, and epithelial cells of wild-type mice that correlated with indices of cell death and pulmonary edema. Other pathways involved in NO and peroxynitrite-mediated tissue injury includes oxidation of sulfhydryl groups, DNA cleavage, lipid peroxidation, inhibition of mitochondrial respiration, inactivation of α1-proteinase inhibitor, alteration in pulmonary surfactant, and inhibition of ion transport in the epithelial cells (10, 27-30). Another likely mechanism through which iNOS expression may augment the extent of acute lung injury is through potentiation of cytokine production by inflammatory cells (31). Previous experiments have revealed, however, that LPS injection elicits a similar rise in plasma TNF, IL-1, and IL-6 concentrations in iNOS knockout and wild-type mice (17).
It could also be argued that iNOS expression aggravates LPS-induced acute lung injury by influencing the adhesion of inflammatory cells to the pulmonary vascular endothelial cells, leading eventually to worsening of vascular leakage and pulmonary edema (32). This mechanism could explain the differences in pulmonary cellular infiltration and lung edema between iNOS knockout mice and wild-type mice. However, more direct evidence is required to confirm the role of iNOS in determining the degree of adhesion between inflammatory cells and endothelial cells.
An interesting observation in our study was that despite the absence of iNOS expression in iNOS knockout mice, weak nitrotyrosine staining could still be detected in large airway epithelial cells. Similar staining has been reported in normal human lungs (26). These observations suggest that low levels of peroxynitrite exist in normal pulmonary cells that could be attributed to the activity of constitutive NOS isoforms.
In summary, our results indicate that LPS injection induces a milder form of acute lung injury in iNOS-deficient mice than in wild type mice. Moreover, we found that positive nitrotyrosine staining can be detected even in the absence of iNOS expression. Finally, our data suggest that constitutive NOS isoforms normally expressed in pulmonary cells do not compensate for the absence of iNOS induction after LPS injection in iNOS-deficient mice.
The writers are grateful to Ms. J. Longo, Ms. R. Carin, and Dr. S. Magder for their help in editing the manuscript.
Supported by the Medical Research Council of Canada and Quebec Lung Association.
| 1. | Fulkerson W. J., MacIntyre N., Stamler J., Crapo J. D.Pathogenesis and treatment of adult respiratory distress syndrome. Arch. Intern. Med.1519962938 |
| 2. | Brigham K. L., Meyrick B.Endotoxin and lung injury. Am. Rev. Respir. Dis.1331986913927 |
| 3. | Knowles R. G., Moncada S.Nitric oxide synthases in mammals. Biochem. J.2981994249258 |
| 4. | Gaston B., Drazen J. M., Loscalzo J., Stamler J. S.The biology of nitrogen oxides in the airways. Am. J. Respir. Crit. Care Med.1491994538551 |
| 5. | Guo F. H., De Raeve H. R., Rice T. W., Stuehr D. J., Thunnissen F. B., Erzrum S. C.Continuous nitric oxide synthesis by inducible nitric oxide synthase in normal human airway epithelium in vivo. Proc. Natl. Acad. Sci. U.S.A.92199578097813 |
| 6. | Kobzik L., Bredt D. S., Lowenstein C. J., Drazen J., Gaston B., Sugarbaker D., Stamler J. S.Nitric oxide synthase in human and rat lung: immunocytochemical and histochemical localization. Am. J. Respir. Cell Mol. Biol.91993371377 |
| 7. | Pheng L. H., Francoeur C., Denis M.The involvement of nitric oxide in a mouse model of adult respiratory distress syndrome. Inflammation191995599610 |
| 8. | Iuvone T., D'Aquisto F., Carnuccio R., Di Rosa M.Nitric oxide inhibits LPS-induced tumor necrosis factor synthesis in vitro and in vivo. Life Sci.591996207211 |
| 9. | Bloomfield G. L., Holloway S., Ridings P. C., Fisher B. J., Blocher C. R., Sholley M., Bunch T., Sugerman H. J., Fowler A. A.Pretreatment with inhaled nitric oxide inhibits neutrophil migration and oxidative activity resulting in attenuated sepsis-induced acute lung injury. Crit. Care Med.251997584593 |
| 10. | Beckman J. S., Koppenol W. H.Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and the ugly. Am. J. Physiol.2711996c1424c1437 |
| 11. | Haddad I. Y., Pataki G., Galliani C., Beckman J. S., Matalon S.Quantitation of nitrotyrosine levels in lung sections of patients and animals with acute lung injury.J. Clin. Invest.94199424072413 |
| 12. | Arkovitz M. S., Wispe J. R., Garcia V. F., Szabo C.Selective inhibition of the inducible isoform of nitric oxide synthase prevents pulmonary transvascular flux during acute endotoxemia. J. Pediatr. Surg.31199610091015 |
| 13. | Wizemann T. M., Gardiner C. R., Laskin J. D., Quinones S., Durham S. K., Goller N. L., Ohnishi S. T., Laskin D. L.Production of nitric oxide and peroxynitrite in the lung during endotoxemia. J. Leukoc. Biol.561994759768 |
| 14. | Salvemini D., Manning P. T., Zweifel B. S., Seibert K., Connor J., Currie M. G., Needleman P., Masferrer J. L.Dual inhibition of nitric oxide and prostaglandin production contributes to the anti-inflammatory properities of nitric oxide synthase inhibitors. J. Clin. Invest.961995301308 |
| 15. | Salvemini D., Korbut R., Anggard E., Vane J.Immediate release of nitric oxide-like factor from bovine aortic endothelial cells by Escherichia coli lipopolysaccharide. Proc. Natl. Acad. Sci. U.S.A.87199025932597 |
| 16. | Laubach V. E., Shesely E. G., Smithies O., Sherman P. A.Mice lacking inducible nitric oxide synthase are not resistant to lipopolysaccharide-induced death. Proc. Natl. Acad. Sci. U.S.A.9219951068810692 |
| 17. | MacMicking J. D., Nathan C., Hom G., Chartrain N., Fletcher D. S., Tumbauer M., Stevens K., Xie Q. W., Sokol K., Hutchinson N., Chen H., Mudgett J. S.Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide synthase. Cell811995641650 |
| 18. | Wei X. Q., Charles I. G., Smith A., Ure J., Feng G. J., Huang F. P., Xu D., Muller W., Moncada S., Liew F. Y.Altered immune responses in mice lacking inducible nitric oxide synthase. Nature3751995408411 |
| 19. | Heppner G., Weiss D. W.High susceptibility of strain A mice to endotoxin and endotoxin-red blood cell mixtures. J. Bacteriol.901965696703 |
| 20. | Welbourn C. R. B., Young Y.Endotoxin, septic shock and acute lung injury: neutrophils, macrophages and inflammatory mediators. Br. J. Surg.7919929981003 |
| 21. | Stewart T. E., Valenza F., Ribeiro S. P., Wener A. D., Volgyesi G., Mullen J. B., Slutsky A. S.Increased nitric oxide in exhaled gas an early marker of lung inflammation in a model of sepsis. Am. J. Respir. Crit. Care Med.1511995713718 |
| 22. | Peng H. B., Libby P., Liao J. K.Induction and stabilization of IkBα by nitric oxide mediates inhibition of NF-κB. J. Biol. Chem.27019951421414219 |
| 23. | Goode H. F., Howdle P. D., Walker B. E., Webster N. R.Nitric oxide synthase activity is increased in patients with septic syndrome. Clin. Sci.881995131133 |
| 24. | Szabo C.Alterations in nitric oxide production in various forms of circulatory shock. New Horiz.31995232 |
| 25. | Akaike T., Noguchi Y., Ijiri S., Setoguchi K., Suga M., Zheng Y. M., Dietzschold B., Maeda H.Pathogenesis of influenza virus- induced pneumonia: involvement of both nitric oxide and oxygen radicals. Proc. Natl. Acad. Sci. U.S.A.93199624482453 |
| 26. | Kooy N. W., Royall J. A., Ye Y. Z., Kelly D. R., Beckman J. S.Evidence for in vitro peroxynitrite production in human acute lung injury. Am. J. Respir. Crit. Care Med.151199512501254 |
| 27. | Moreno J. J., Pryor W. A.Inactivation of α1-proteinase inhibitor by peroxynitrite. Chem. Res. Toxicol.51991425431 |
| 28. | Cassina A., Radi R.Differential inhibitory action of nitric oxide and peroxynitrite on mitochondrial electron transport. Arch. Biochem. Biophys.3281996309316 |
| 29. | Rubbo H., Radi R., Trujillo M., Telleri R., Kalyanaraman B., Barnes S., Kirk M., Freeman B. A.Nitric oxide regulation of superoxide and peroxynitrite-mediated lipid peroxidation: formation of novel nitrogen-containing oxidized lipid derivatives. J. Biol. Chem.26919942606626075 |
| 30. | Haddad I. Y., Zhu S., Crow J., Barefield E., Gadilhe T., Matalon S.Inhibition of alveolar type II cell ATP and surfactant synthesis by nitric oxide. Am. J. Physiol.2701996L898L906 |
| 31. | Van Dervort A. L., Yan L., Madara P. J., Cobb J. P., Wesley R. A., Corriveau C. C., Tropea M. M., Danner R. L.Nitric oxide regulates endotoxin-induced TNF-α production by human neutrophils. J. Immunol.152199441024109 |
| 32. | Fukatsu K., Saito H., Han I., Furukawa S., Hashiguchi Y., Lin M. T., Matsuda T., Inaba T., Inoue T., Ikeda S., Yasuhara H., Muto T.Nitric oxide inhibition decreases neutrophil adhesion at the inflammatory site, while increasing adhesion in remote organs in peritonitis. J. Surg. Res.6819977986 |
