Acute lung injury (ALI) is an important cause of morbidity and mortality in critically ill patients.1 An underlying cause (e.g., sepsis, trauma) combined with a secondary inflammatory response presents as clinical lung injury.2 Mechanical ventilation is the mainstay of supportive therapy, but it can cause or intensify injury through both mechanical and inflammatory processes.3,4 Thus, in clinical lung injury, mechanical ventilation and lung inflammation are inextricably linked.
Pulmonary inflammation has many different components, and in ALI, the regulation of inflammatory prostanoids may be both important and complex.5 For example, antagonism of thromboxane A2 reduces lung injury resulting from smoke exposure6; however, in contrast, increased expression of prostaglandin (PG)D2 synthase reduces injury7 and PGE2 seems to mediate the development of pulmonary fibrosis.8
Because of observations such as these, interest has developed in modulating cyclooxygenase (COX) enzyme in ALI5 as in other disease states. COX is a membrane-bound enzyme responsible for the oxidation of arachidonic acid to PGG2 and its subsequent reduction to PGH2. Two COX isoforms, COX-1 and COX-2, are expressed in the lung: COX-1 is constitutively expressed whereas COX-2 is inducible. Thus, inhibition of COX may represent a therapeutic target for inflammatory lung injury. Indeed, nonselective inhibition of COX prevented thrombin-induced lung vascular leak in a sheep model9 and attenuates the pulmonary hypertension that accompanies ventilator-induced lung injury (VILI).10 However, nonselective COX inhibition has led to disappointing results in a broad range of at-risk critically ill patients in whom it did not prevent the development of lung injury.11 Therefore, it would be beneficial to better understand which elements of lung injury are attenuated by inhibition of COX. Advances in our understanding of the pathophysiology of VILI include the ubiquitous release of cytokines and eicosanoids, as well as the apparently central role of important signaling pathways such as inhibitor of nuclear factor-κB/nuclear factor-κB (IκB/NF-κB).12,13
The signaling pathways regulating NF-κB activation and its interactions with potential mediators such as cytokines and eicosanoids in VILI are inevitably complex.12 The mechanism of stretch-induced activation of NF-κB is unclear. It is plausible that stretch-induced activation of cytosolic phospholipase A214 with its consequent trigger of COX activity, eicosanoid signaling, and increased intracellular cyclic adenosine monophosphate (cAMP) after activation of prostanoid receptors would upregulate cytokines via cAMP responsive elements (CREs) in their promoters and this might in turn activate NF-κB via cytokine signaling.15 However, NF-κB is an important transcriptional regulator of inducible COX-2,16,17 therefore forming a positive feedback cycle encompassing NF-κB, eicosanoids, and cytokines. Whether interrupting a single component of this cycle, eicosanoid production, could abrogate NF-κB activity and thus protect against its deleterious effects in VILI is unknown.
Ibuprofen has been reported to modulate the activity of NF-κB, although possibly independently of its action as a COX inhibitor.18,19 Nevertheless, activation of NF-κB in the absence of eicosanoid production would indicate that there is an alternative pathway of activation. Because of the failure of clinical trials of COX inhibition in disease states where NF-κB activation is thought to be central (e.g., sepsis),11 we hypothesized that there is such an eicosanoid-independent pathway in VILI.
After institutional ethics approval (conforming to the guidelines of the Canadian Council for Animal Care), male Sprague-Dawley non-SPF rats (300–400 g) obtained from Charles River (Pointe Claire, QC, Canada) were used in all experiments and were anesthetized as previously described.20,21 Rats' lungs were mechanically ventilated (volume controlled) at baseline settings: tidal volume 8 mL · kg−1, positive end-expiratory pressure 1 cm H2O, fraction of inspired oxygen 1.0, and respiratory rate 42 (Harvard Apparatus Rodent Ventilator #683; Harvard Apparatus, St. Laurent, QC, Canada).
Cyclooxygenase Inhibition (Nonselective)
Animals were randomized to pretreatment with high-dose (relative to doses used clinically) IV ibuprofen (100 mg · kg−1, nonselective COX inhibitor; dissolved in saline, pH adjusted to 7.4) or saline in a blinded manner 30 minutes before commencing injurious ventilation.
Injurious ventilation was instituted by reducing positive end-expiratory pressure to zero and increasing tidal volume such that the peak inspiratory pressure was 21 mm Hg (29 cm H2O).21 Arterial blood samples were collected at 10, 30, 60, and 90 minutes for blood gas analysis, and a pressure-volume curve was measured before commencement and after 90 minutes of injurious ventilation.
Measurement of Injury
Animals were then euthanized, the pulmonary arteries flushed with 10 mL normal saline, the left main bronchus was ligated, and the right lung lavaged 3 times using a single 5-mL aliquot of saline. The bronchoalveolar lavage (BAL) fluid was centrifuged at 500 g for 5 minutes and the supernatant stored at −70°C. The left lung was excised; half was snap frozen (−70°C) and the other half weighed for measurement of wet-to-dry lung weight.20 BAL fluid was assayed for protein using the Bradford method (Bio-Rad protein assay kit; Bio-Rad Laboratories, Inc., Hercules, CA).
Concentrations of interleukin (IL)-1β, IL-6, GRO/KC (growth-related oncogene/keratinocyte chemoattractant), and tumor necrosis factor (TNF)α in BAL were measured in duplicate using multiplex immunoassays for Luminex technology from Linco Research, Inc. (St. Charles, MO).22
Eicosanoid concentration in BAL was determined using liquid chromatography–tandem mass spectrometry, and performed using an API4000 triple-quadruple mass spectrometer (MDS SCIEX, Concord, ON, Canada) in the electrospray ionization negative ion mode with TurboIon-Spray as described previously.14
COX-1, COX-2, and IκB-α protein expression was examined using Western analysis of lung tissue. Briefly, frozen lung samples were homogenized in RIPA protein extraction buffer supplemented with phosphatase inhibitors, centrifuged, and the protein in the supernatant was measured. Approximately 60 μg total protein was fractionated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis, 4% to 20% gradient Tris-Glycine gels (Invitrogen, Carlsbad, CA), transferred onto Immobilon-P polyvinylidene difluoride membranes (Millipore, Inc., Billerica, MA), and incubated overnight with blocking buffer. To examine COX-1 and COX-2 expression, membranes were incubated with mouse polyclonal antibodies (Cayman Chemical Co., Ann Arbor, MI). To examine the phosphorylation and expression of IκB-α, membranes were incubated overnight with antibodies (Cell Signaling, Danvers, MA) specific to IκB-α phosphorylated on Ser32 (1:500) or recognizing IκB-α independent of phosphorylation (1:1000). After incubation with horseradish peroxidase–conjugated anti-rabbit immunoglobulin G (1:10,000; BD Biosciences, Mississauga, ON, Canada), immunoreactive proteins were visualized by enhanced chemiluminescence (Amersham Biosciences, Baie d'Urfe, QC, Canada). The signals were normalized by probing for housekeeping proteins GAPDH (glyceraldehyde 3-phosphate dehydrogenase) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or actin (Sigma-Aldrich, Oakville, ON, Canada). Specificities of the COX antibodies were evaluated by preincubation of the primary antibodies with 10-fold excess of the corresponding immunizing peptides for COX-1 and COX-2 before application of primary antibody to the immunoblot. The blocked antibody in each case abolished detection of the band corresponding in molecular weight to COX-1 or COX-2, respectively.
Electrophoretic Mobility Shift Assay
Nuclear protein extracts were prepared from 200 mg frozen lung tissue as described23 and stored in aliquots at −80°C. Lung tissue from a rat pretreated with an IV injection of lipopolysaccharide 1 hour before tissue collection was used as a positive control for NF-κB activity. Protein quantitation was done using the Bio-Rad protein assay (Bio-Rad, Mississauga, ON, Canada). DNA binding activity of NF-κB was assessed using the LightShift® Chemiluminescent EMSA Kit (Pierce, Rockford, IL) according to the manufacturer's instructions. A 30-bp double-stranded oligonucleotide containing the κB site of the rat cytokine-induced neutrophil chemoattractant (CINC) gene was used as probe (5′-CCTGTGCTCCGGGAATTTCCCTGGCCTGGA-3′).24 Ten micrograms of protein was added to a binding reaction containing 20 mM Tris-HCl pH 7.5, 50 mM KCl, 1 mM DTT (dithithreitol), and 0.5 μg poly(dI-dC) (poly(deoxyinosinic-deoxycytidylic) acid). After 15 minutes on ice, 25 fmol biotin-labeled probe was added and the binding reaction was incubated at 20°C for 20 minutes and subjected to electrophoresis on a 4% acrylamide/0.5× TBE gel (TBE 89 mM Tris, 89 mM boric acid, 2 mM EDTA). After electrophoretic transfer to Hybond N+ membrane (GE Healthcare, Baie d'Urfe, QC, Canada), biotin-labeled DNA was detected via conjugated streptavidin– horseradish peroxidase and chemiluminescent substrate. Cold competition was performed by adding 5 pmol of unlabeled probe or of an unlabeled mutant oligonucleotide containing a 3-bp deletion in the κB binding site (5′-CCTGTGCTCCAATTTCCCTGGCCTGGA-3′).
Data are presented as mean ± SD. The Student t test or 1-way analysis of variance followed by Student-Newman-Keuls tests was used unless otherwise specified in figure legends. P < 0.05 was considered significant. Analysis was performed with SigmaStat 3.1 (SPSS, Inc., Chicago, IL).
The baseline compliance was identical in the saline and ibuprofen-pretreated groups (n = 8, each group). The decrement in compliance after 90 minutes of injurious ventilation was significantly greater in the nontreated versus ibuprofen-pretreated group (Fig. 1A), as was the wet-to-dry lung weight ratio (Fig. 1B). The results of blood gas analysis (i.e., pH, PaCO2, PaO2, HCO3−, base excess) did not differ between the groups (Table 1). Ibuprofen was associated with increased survival after injurious ventilation (Fig. 2).
An additional group of nonventilated animals (n = 6) was used for comparisons of BAL fluid cytokine and prostanoid measurements. Injurious ventilation resulted in significantly higher levels of proinflammatory cytokines (TNFα, IL-1β, IL-6, and GRO/KC) in the BAL fluid versus the nonventilated group (Fig. 3), which was not changed by ibuprofen (versus vehicle) pretreatment (n = 7 per pretreated groups).
Injurious ventilation had variable effects on COX-dependent eicosanoids (Fig. 4); for example, most were increased (leukotriene B4 significantly; PGE2, 6-keto-PGF1α, and thromboxane B2 nonsignificantly) and some were decreased (PGF2, PGD2). Ibuprofen obliterated the production of all COX-dependent eicosanoids (Fig. 4), which was not the case with eicosanoids that were not COX dependent (e.g., lipoxygenase-dependent 12-hydroxyeicosatetraenoic acid; data not shown).
COX Protein Expression and NF-κB Activation
Although levels of COX-1 protein demonstrated some variability among individual animals, there was no effect of ventilation or ibuprofen pretreatment on COX-1 (Fig. 5) as expected for this constitutively expressed isoform. In contrast, injurious ventilation was associated with a large increase in expression of the inducible COX-2 isoform, and this increase was attenuated by ibuprofen (Fig. 5).
Because NF-κB is an important regulator of COX-2 induction5,25 and activated by injurious ventilation,12 we examined whether NF-κB activation was modified by COX inhibition. As expected, the phosphorylation of IκB-α protein was increased and its total protein level was attenuated by injurious ventilation (versus no ventilation; Fig. 6A), suggesting activation of NF-κB, but ibuprofen did not abrogate the changes in IκB-α expression (Fig. 6A). Activation of NFκ-B was confirmed by electrophoretic mobility shift assay (Fig. 6B) in lungs after injurious ventilation and this activation also was not blocked by pharmacological inhibition of COX activity.
In the current study, pretreatment with the COX inhibitor ibuprofen attenuated VILI and confirmed that activation of a key injury pathway, IκB/NF-κB, is not eicosanoid dependent.
We report that injurious ventilation caused lung injury (i.e., decrement in compliance, tissue edema) and increased inflammatory cytokines, eicosanoids, and COX-2. Ibuprofen pretreatment, effectively inhibiting COX-2 and reducing key eicosanoids, attenuated the ventilator-induced decrement in respiratory mechanics and lung edema, but did not affect inflammatory cytokines or activation of NF-κB. These data support the role of an eicosanoid-independent stretch-induced pathway of NF-κB in VILI.
Ibuprofen is not a selective COX inhibitor but was chosen because selective inhibition was not possible in vivo. We attempted to selectively inhibit COX-2 using celecoxib to dissect the relative contribution of the inducible COX isoform to VILI. However, we were unable to effectively demonstrate tissue inhibition of COX-2 despite the use of several selective inhibitors and a variety of commercially available COX-2 enzyme assay systems. In addition, the prospect of using COX-2 knockout mice is complicated by the difference in species and the developmental effects of COX-2 gene deletion compared with alternative blockade.26
Stretch-Activated Pathways of Injury
Injurious mechanical ventilation increases the activity of cytosolic phospholipase A2,27 the rate-limiting enzyme for production of arachidonic acid; arachidonic acid is in turn the rate-limiting substrate for the synthesis of eicosanoids by COX. In the current study, injurious ventilation altered the concentrations of COX-dependent eicosanoids and increased the expression of COX-2 (but not COX-1) protein; all of these effects were abolished by ibuprofen. These findings are consistent with our knowledge of COX-1 and COX-2 expression in a variety of biological settings.28,29 In general, COX-1 is most variable during development, and under most conditions, the levels of COX-1 mRNA and protein are stable. In contrast, COX-2 is inducible (in those cells that have the capacity to express it), is not expressed under basal conditions, and when expressed, COX-2 mRNA lasts for hours only.30
In keeping with the difference in responsiveness, the transcriptional regulation of COX-2 is far more complex than that of COX-1; many more factors can affect its gene activation.28 Prominent among the factors known to induce expression of COX-2 are the proinflammatory cytokines IL-1β and TNFα; in addition, NF-κB is perhaps among the best understood of the multiple signaling pathways that mediate the COX-2 expression.28 Indeed, the human COX-2 promoter region has 2 NF-κB binding sites.
The findings in this study are consistent with other reports, i.e., injury associated with minimal change in expression of COX-114 but with a large increase in COX-2 expression.8 However, although NF-κB was activated by injurious ventilation in the current study, it was unaltered by nonselective COX inhibition. Ibuprofen and other nonsteroidal antiinflammatory drugs have been shown to inhibit NF-κB activation in a variety of cell culture models, including Jurkat T cells, prostate cancer cell lines, and human mononuclear cells18,19,31 but not under all stimuli (e.g., see Ref. 31) or consistently in all cell types (e.g., hepatocytes,32 fibroblasts,33 bronchial epithelial cells34). Thus, it is possible that ibuprofen inhibited NF-κB in a key subpopulation of inflammatory cells within the ventilated lung, but that this effect was not detectable in tissue extracts containing a larger background of cells in which NF-κB was not inhibited in vivo.
COX-2 expression is regulated not only by the NF-κB pathway but also by intracellular cAMP signaling pathways that can be upregulated in a positive feedback loop via eicosanoid receptor (e.g., EP) activation.35 Eicosanoids bind to specific G protein–coupled receptors and increase adenylate cyclase activity, thereby increasing the levels of cAMP, which in turn activate transcription factors CREB (cAMP response element binding protein) and AP-1 (activator protein-1). The COX-2 promoter contains a CRE that upregulates transcription of COX-2 via CREB and AP-1.36 Therefore, we speculate that in VILI, the increased level of COX-2 protein requires occupancy of both the NF-κB binding sites and the CRE. This would be consistent with data from endotoxin-treated macrophages in which participation of at least 2 of these cis-acting elements was necessary to achieve maximal induction of COX-2 transcription.37 Thus, the blockade of COX-2 protein expression achieved by nonselective COX inhibition might be explained by reduced eicosanoid levels leading to lessened activation of eicosanoid receptors and decreased production of cAMP.
COX Inhibition and Cytokine Expression
Nonselective COX inhibition in this study did not alter the increased expression of injury-associated proinflammatory cytokines (Fig. 3). The increase in cytokine levels associated with VILI may be mediated by degradation of IκBα and activation of NF-κB,12 and the finding in the current study that the cytokine levels after treatment with ibuprofen were consistently greater than “nonventilation” levels suggests that mechanisms of ibuprofen-associated protection are not likely to be cytokine-related.
Prostanoids and Lung Injury
The role of individual prostanoids is difficult to determine, particularly given their short individual half-lives and the dynamic nature of acute disease. Nonetheless, many roles have been characterized in ALI. For example, thromboxane A2 increases pulmonary artery pressure and microvascular permeability, thereby inducing lung edema.38 In addition, many of the pulmonary vascular and platelet-specific effects are antagonized by PGI2 and PGE25; indeed, both PGI2 and PGE2 are associated with the development of lung injury in several models.39–41 Overall, the “net” effect of global changes in prostanoid levels is difficult to predict, and in the current study, global changes in prostanoid levels (or their stable metabolites) reflected the magnitude of injury, and such changes were generally suppressed by nonselective COX inhibition.
The experiments were short term and performed in a specific animal model. Human critical illness occurs over a far longer interval, and critically ill patients with ALI frequently have disordered coagulation and renal function that would be further complicated with the use of COX inhibition; as such, the current data represent plausibility of effect rather than predictability of human response. In addition, selective inhibition of COX isoforms is not reported. Although the effects of COX inhibition (in the current study) were not mediated by alterations in NF-κB signaling, future studies may examine the role of pharmacological inhibition of elements of this pathway.42
We conclude that nonselective COX inhibition provided partial protection against VILI and demonstrated that NF-κB activation in this context is not exclusively eicosanoid dependent.
1. Zambon M, Vincent JL. Mortality rates for patients with acute lung injury/ARDS have decreased over time. Chest 2008;133:1120–7
2. Puneet P, Moochhala S, Bhatia M. Chemokines in acute respiratory distress syndrome. Am J Physiol Lung Cell Mol Physiol 2005;288:L3–15
3. Dreyfuss D, Saumon G. Ventilator-induced lung injury: lessons from experimental studies. Am J Respir Crit Care Med 1998;157:294–323
4. Vlahakis NE, Hubmayr RD. Cellular stress failure in ventilator-injured lungs. Am J Respir Crit Care Med 2005;171:1328–42
5. Park GY, Christman JW. Involvement of cyclooxygenase-2 and prostaglandins in the molecular pathogenesis of inflammatory lung diseases. Am J Physiol Lung Cell Mol Physiol 2006;290:L797–805
6. Westphal M, Noshima S, Isago T, Fujioka K, Maybauer MO, Maybauer DM, Traber LD, Flynn JT, Westphal-Varghese BB, Traber DL. Selective thromboxane A2 synthase inhibition by OKY-046 prevents cardiopulmonary dysfunction after ovine smoke inhalation injury. Anesthesiology 2005;102:954–61
7. Ando M, Murakami Y, Kojima F, Endo H, Kitasato H, Hashimoto A, Kobayashi H, Majima M, Inoue M, Kondo H, Kawai S, Hayashi I. Retrovirally introduced prostaglandin D2 synthase suppresses lung injury induced by bleomycin. Am J Respir Cell Mol Biol 2003;28:582–91
8. White KE, Ding Q, Moore BB, Peters-Golden M, Ware LB, Matthay MA, Olman MA. Prostaglandin E2 mediates IL-1beta-related fibroblast mitogenic effects in acute lung injury through differential utilization of prostanoid receptors. J Immunol 2008;180:637–46
9. Perlman MB, Johnson A, Malik AB. Ibuprofen prevents thrombin-induced lung vascular injury: mechanism of effect. Am J Physiol 1987;252:H605–14
10. Burger R, Fung D, Bryan AC. Lung injury in a surfactant-deficient lung is modified by indomethacin. J Appl Physiol 1990;69:2067–71
11. Bernard GR, Wheeler AP, Russell JA, Schein R, Summer WR, Steinberg KP, Fulkerson WJ, Wright PE, Christman BW, Dupont WD, Higgins SB, Swindell BB. The effects of ibuprofen on the physiology and survival of patients with sepsis. The Ibuprofen in Sepsis Study Group. N Engl J Med 1997;336:912–8
12. Held HD, Boettcher S, Hamann L, Uhlig S. Ventilation-induced chemokine and cytokine release is associated with activation of nuclear factor-kappaB and is blocked by steroids. Am J Respir Crit Care Med 2001;163:711–6
13. Krakauer T. Nuclear factor-kappaB: fine-tuning a central integrator of diverse biologic stimuli. Int Rev Immunol 2008;27:286–92
14. Copland IB, Reynaud D, Pace-Asciak C, Post M. Mechanotransduction of stretch-induced prostanoid release by fetal lung epithelial cells. Am J Physiol Lung Cell Mol Physiol 2006;291:L487–95
15. Szabo-Fresnais N, Blondeau JP, Pomerance M. Activation of the cAMP pathway synergistically increases IL-1-induced IL-6 gene expression in FRTL-5 thyroid cells: involvement of AP-1 transcription factors. Mol Cell Endocrinol 2008;284:28–37
16. Lee SM, Cheung CY, Nicholls JM, Hui KP, Leung CY, Uiprasertkul M, Tipoe GL, Lau YL, Poon LL, Ip NY, Guan Y, Peiris JS. Hyperinduction of cyclooxygenase-2-mediated proinflammatory cascade: a mechanism for the pathogenesis of avian influenza H5N1 infection. J Infect Dis 2008;198:525–35
17. Xu F, Xu Z, Zhang R, Wu Z, Lim JH, Koga T, Li JD, Shen H. Nontypeable Haemophilus influenzae induces COX-2 and PGE2 expression in lung epithelial cells via activation of p38 MAPK and NF-kappa B. Respir Res 2008;9:16
18. Stuhlmeier KM, Li H, Kao JJ. Ibuprofen: new explanation for an old phenomenon. Biochem Pharmacol 1999;57:313–20
19. Palayoor ST, Youmell MY, Calderwood SK, Coleman CN, Price BD. Constitutive activation of IkappaB kinase alpha and NF-kappaB in prostate cancer cells is inhibited by ibuprofen. Oncogene 1999;18:7389–94
20. Duggan M, McCaul CL, McNamara PJ, Englelberts D, Ackerley C, Kavanagh BP. Atelectasis causes vascular leak and lethal right ventricular failure in uninjured rat lungs. Am J Respir Crit Care Med 2003;167:1633–40
21. Copland IB, Martinez F, Kavanagh BP, Englelberts D, McKerlie C, Belik J, Post M. High tidal volume ventilation causes different inflammatory responses in newborn versus adult lung. Am J Respir Crit Care Med 2004;169:739–48
22. Tsuchida S, Englelberts D, Roth M, McKerlie C, Post M, Kavanagh BP. Continuous positive airway pressure causes lung injury in a model of sepsis. Am J Physiol Lung Cell Mol Physiol 2005;289:L554–64
23. Deryckere F, Gannon F. A one-hour minipreparation technique for extraction of DNA-binding proteins from animal tissues. Biotechniques 1994;16:405
24. Blackwell TS, Blackwell TR, Christman JW. Impaired activation of nuclear factor-kappaB in endotoxin-tolerant rats is associated with down-regulation of chemokine gene expression and inhibition of neutrophilic lung inflammation. J Immunol 1997;158:5934–40
25. D'Acquisto F, Iuvone T, Rombola L, Sautebin L, Di Rosa M, Carnuccio R. Involvement of NF-kappaB in the regulation of cyclooxygenase-2 protein expression in LPS-stimulated J774 macrophages. FEBS Lett 1997;418:175–8
26. Yu Y, Funk CD. A novel genetic model of selective COX-2 inhibition: comparison with COX-2 null mice. Prostaglandins Other Lipid Mediat 2007;82:77–84
27. Miyahara T, Hamanaka K, Weber DS, Anghelescu M, Frost JR, King JA, Parker JC. Cytosolic phospholipase A2 and arachidonic acid metabolites modulate ventilator-induced permeability increases in isolated mouse lungs. J Appl Physiol 2008;104:354–62
28. Kang YJ, Mbonye UR, DeLong CJ, Wada M, Smith WL. Regulation of intracellular cyclooxygenase levels by gene transcription and protein degradation. Prog Lipid Res 2007;46:108–25
29. Smith WL, DeWitt DL, Garavito RM. Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem 2000;69:145–82
30. Perkins DJ, Kniss DA. Rapid and transient induction of cyclo-oxygenase 2 by epidermal growth factor in human amnion-derived WISH cells. Biochem J 1997;321:677–81
31. Scheuren N, Bang H, Munster T, Brune K, Pahl A. Modulation of transcription factor NF-kappaB by enantiomers of the nonsteroidal drug ibuprofen. Br J Pharmacol 1998;123:645–52
32. Callejas NA, Casado M, Bosca L, Martin-Sanz P. Absence of nuclear factor kappaB inhibition by NSAIDs in hepatocytes. Hepatology 2002;35:341–8
33. Gomez PF, Pillinger MH, Attur M, Marjanovic N, Dave M, Park J, Bingham CO III, Al-Mussawir H, Abramson SB. Resolution of inflammation: prostaglandin E2 dissociates nuclear trafficking of individual NF-kappaB subunits (p65, p50) in stimulated rheumatoid synovial fibroblasts. J Immunol 2005;175:6924–30
34. Dauletbaev N, Lam J, Eklove D, Iskandar M, Lands LC. Ibuprofen modulates NF-kB activity but not IL-8 production in cystic fibrosis respiratory epithelial cells. Respiration 2010;79:234–42
35. Steinert D, Kuper C, Bartels H, Beck FX, Neuhofer W. PGE2 potentiates tonicity-induced COX-2 expression in renal medullary cells in a positive feedback loop involving EP2-cAMP-PKA signaling. Am J Physiol Cell Physiol 2009;296:C75–87
36. Freeman AF, Kleiner DE, Nadiminti H, Davis J, Quezado M, Anderson V, Puck JM, Holland SM. Causes of death in hyper-IgE syndrome. J Allergy Clin Immunol 2007; 119:1234–40
37. Mestre JR, Mackrell PJ, Rivadeneira DE, Stapleton PP, Tanabe T, Daly JM. Redundancy in the signaling pathways and promoter elements regulating cyclooxygenase-2 gene expression in endotoxin-treated macrophage/monocytic cells. J Biol Chem 2001;276:3977–82
38. Ishitsuka Y, Moriuchi H, Hatamoto K, Yang C, Takase J, Golbidi S, Irikura M, Irie T. Involvement of thromboxane A2 (TXA2) in the early stages of oleic acid-induced lung injury and the preventive effect of ozagrel, a TXA2 synthase inhibitor, in guinea-pigs. J Pharm Pharmacol 2004;56:513–20
39. Imai Y, Kawano T, Miyasaka K, Takata M, Imai T, Okuyama K. Inflammatory chemical mediators during conventional ventilation and during high frequency oscillatory ventilation. Am J Respir Crit Care Med 1994;150:1550–4
40. Gust R, Kozlowski JK, Stephenson AH, Schuster DP. Role of cyclooxygenase-2 in oleic acid-induced acute lung injury. Am J Respir Crit Care Med 1999;160:1165–70
41. Goggel R, Hoffman S, Nusing R, Narumiya S, Uhlig S. Platelet-activating factor-induced pulmonary edema is partly mediated by prostaglandin E(2), E-prostanoid 3-receptors, and potassium channels. Am J Respir Crit Care Med 2002;166:657–62
42. Sethi G, Tergaonkar V. Potential pharmacological control of the NF-kappaB pathway. Trends Pharmacol Sci 2009;30:313–21
Supported by Canadian Institutes of Health Research (CHIR MOP-15272). Dr. Kavanagh holds the Dr. Geoffrey Barker Chair in Critical Care Medicine and Dr. Post holds a Canada Research Chair in Fetal, Neonatal and Maternal Health.