| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;81:1061-1067
© 2006 The Society of Thoracic Surgeons
Division of Cardiothoracic Surgery, University of Washington Medical Center, Seattle, Washington
Accepted for publication September 15, 2005.
* Address correspondence to Dr Farivar, University of Washington Medical Center, 1959 NE Pacific St, Division of Cardiothoracic Surgery, Box 356310, Seattle, WA 98195 (Email: afarivar{at}u.washington.edu).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: Rats were pretreated with hirudin 30 minutes before warm, in situ left lung ischemia and reperfusion. Multiple in vivo assessments of lung injury were determined, and mechanistic studies assessed transcriptional regulation early in reperfusion and proinflammatory protein secretion late in reperfusion. Immunohistochemistry localized thrombin activation.
RESULTS: Thrombin localized to macrophages and endothelial and epithelial cells early in reperfusion. Hirudin significantly limited lung ischemia–reperfusion injury–induced derangements in vascular permeability and intraalveolar inflammatory cell sequestration, resulting in improved arterial oxygenation after ischemia and 4 hours of reperfusion. The protection was transcriptionally mediated by attenuated activator protein-1 and early growth response-1 transactivation, but not nuclear factor kappa B transactivation. This was associated with reduced chemokine, but not tumor necrosis factor 
, secretion late in reperfusion.
CONCLUSIONS: Thrombin promotes lung ischemia–reperfusion injury, as hirudin protected against experimental acute lung injury. Hirudin conferred protection through a mechanism independent of nuclear factor kappa B and tumor necrosis factor , suggesting that its effects may be mediated by a parallel, synergistic inflammatory pathway through activator protein-1 and early growth response-1.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
There have been a number of publications focusing on the contribution of individual lung cell populations to the development of lung ischemia–reperfusion injury (LIRI). Early after graft reperfusion, alveolar macrophages demonstrate marked transactivation of proinflammatory transcription factors nuclear factor kappa B (NF
B) and activator protein-1 (AP-1), which upregulate inflammatory cytokine, chemokine, and cell-adhesion molecule genes [4]. Furthermore, there is an early burst of tumor necrosis factor (TNF-) and interleukin 1ß from the alveolar macrophage, which primes surrounding epithelial and endothelial cells to amplify their response to LIRI [5, 6]. As such, the NFB and TNF- inflammatory pathway in macrophages makes important autocrine and paracrine contributions to LIRI.
Endothelial cells are also an important proinflammatory cell type regulating myocardial ischemia–reperfusion injury [7], and in the lung are the first line of defense against injurious circulating mediators. Endothelial cells transcriptionally regulate LIRI through NFB and early growth response-1 (EGR-1), promoting the expression of the chemokines cytokine-induced neutrophil chemoattractant (CINC) and monocyte chemoattractant protein-1 (MCP-1) [8]. Although experimental models of LIRI have determined that the early phases of injury are coordinated by the alveolar macrophage and endothelium, late-phase tissue injury is clearly dependent on neutrophil and, to a lesser extent, lymphocyte infiltration [2, 4-6, 9]. As such, interventions targeting early proinflammatory events in these cells (limiting their activation) may be an effective strategy to reduce LIRI, as would characterization of novel inflammatory pathways independent of NFB and TNF-.
Thrombin is the main effector protease of the coagulation cascade, while also augmenting inflammation by activating multiple cell types, as shown in Figure 1. Cytokines and chemokines in turn can activate extravascular coagulation by tissue factor upregulation and thrombin activation, primarily on monocytes and macrophages [10]. Thrombin amplifies inflammation through disruption of endothelial integrity, upregulation of proinflammatory genes, and promotion of neutrophil chemotaxis [11]. In cultured endothelial cells thrombin induces the secretion of chemokines, mediators known to contribute to LIRI, and induces endothelial monolayer derangements [2, 4-7, 12-14]. Lastly, it has long been recognized that acute lung injury, such as seen with acute respiratory distress syndrome or secondary to sepsis, is associated with marked procoagulant activity in the bronchoalveolar lavage (BAL) fluid of affected patients, with unclear physiologic function [15]. As such, thrombin mediates crosstalk between thrombosis and inflammation, and could be a novel target for an intervention aimed at limiting LIRI.
|
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Rodent Model of Lung Ischemia and Reperfusion
Pathogen-free Long-Evans rats (Simonsen Labs, Gilroy, CA), weighing 280 to 320 g, were used for all experiments. Animals were cared for in compliance with the "Principles of Laboratory Animal Care" formulated by the Institute of Laboratory Animal Resources, and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1985).
A warm, in situ ischemia–reperfusion hilar occlusion model was used as described previously [4-6, 8, 16]. In brief, anesthesia was induced with pentobarbital and animals were ventilated (Harvard Apparatus Inc, Holliston, MA) through a tracheostomy with a standardized inspired oxygen content of 60%, a rate of 80 breaths per minute, and 2 cm H2O of positive end-expiratory pressure. Maximal peak pressures were maintained less than 10 cm H2O. After an anterolateral thoracotomy through the fifth intercostal space, the left lung was mobilized atraumatically, and the pulmonary ligament was taken down sharply. The left pulmonary artery, pulmonary vein, and mainstem bronchus were occluded with a noncrushing microvascular clamp with the lung in an inflated state. At the end of ischemia, the clamp was removed and the lung was ventilated and reperfused up to 4 hours.
Three different experimental groups were studied in vivo. Negative controls were sacrificed without undergoing any ischemia or reperfusion, and were surgically unmanipulated. The other two groups underwent 90 minutes of ischemia followed by 4 hours of reperfusion. To assess the effects of direct thrombin inhibition on lung reperfusion injury, rats were treated with recombinant hirudin at a bolus dose of 1 mg/kg intravenously 30 minutes before ischemia. This dose was chosen from dose–titration curves. Our end point sought to determine a dose that would ensure at least a doubling of the activated partial thromboplastin time throughout the experimental protocol (data not shown). Hirudin is a potent, specific, irreversible inhibitor of thrombin activity through competitive inhibition of its catalytic site [21]. No serious bleeding consequences were noted during the experimental protocols in the hirudin-treated animals, and there was no increase in expected mortality. The last group of animals received an equivalent volume of intravenous vehicle (saline) 30 minutes before ischemia. This latter group is referred to as positive controls.
Electromobility shift assays for proinflammatory transcription factor transactivation were done on injured left lung homogenates that had undergone 90 minutes of ischemia and either 15 minutes or 4 hours of reperfusion. The early 15-minute time was chosen as significant transcriptional regulation of proinflammatory genes would need to occur early in these experiments to affect transcription and translation of relevant chemokines and cytokines. Florid lung injury occurs by 4 hours of reperfusion in this model [4-6, 19, 20]. Therefore relevant transcriptional regulation would need to occur early in reperfusion to affect injury known to be full-blown by 4 hours of reperfusion. Furthermore, we have previously published that there is a dramatic increase NFB and AP-1 activity by 15 minutes of reperfusion, but not earlier [17, 18]. Enzyme-linked immunosorbent assays were performed on recovered left lung BAL at 4 hours of reperfusion to quantitate chemokine and cytokine protein content. During the BAL the right hilum was clamped.
Lung Vascular Permeability Index
To quantitate endothelial vascular injury secondary to LIRI, a permeability index was calculated as described previously [16, 17]. Iodine-125–radiolabeled bovine serum albumin (NEN Life Sciences, Boston, MA) was injected intravenously 5 minutes before reperfusion. At 4 hours of reperfusion, the radioactivity counts were quantitated in the left and right lung, as well as in 1 mL of blood. The permeability index was then calculated as follows:
|
|
This ratio corrects for any variation in systemic blood levels of radioactivity and provides a reproducible indicator of lung microvascular permeability secondary to acute oxidative lung injury.
Myeloperoxidase Assay
Tissue myeloperoxidase (myeloperoxidase) content quantitated neutrophil accumulation in lung parenchyma as described previously [4-6, 16]. Assay buffer was composed of 0.0005% H2O2 and 0.167 mol/L O-dianisidine dihydrochloride in a 100 mmol/L potassium phosphate buffer. The change in absorbance at 460 nm over 1 minute was recorded after mixing 50 µL of each sample with 1.45 mL of assay buffer.
Bronchoalveolar Lavage
Left lung BAL assessed inflammatory cell accumulation in injured alveolar spaces of ischemia–reperfused lungs. Left lungs were lavaged selectively through the tracheostomy tube with 3.0 mL of saline after a clamp was placed on the right hilum as described previously [16, 17]. The supernatant was frozen for cytokine and chemokine analysis after the addition of a protease inhibitor. Inflammatory cells were counted using a hemacytometer.
Electromobility Shift Assay
Lungs were snap-frozen and processed as described previously [4-6, 8, 16]. Nuclear protein (10 µg) was incubated with double-stranded phosphate-32 end-labeled oligonucleotide containing the NFµB (Promega, Madison, WI), AP-1 (Promega), or EGR (Santa Cruz Biotechnology, Santa Cruz, CA) consensus binding sequence. Running unlabeled oligonucleotide probe in a cold competition reaction assessed probe specificity. Supershift analysis for the EGR-1 was performed as described by Santa Cruz and us previously [16, 18]. Results were verified in three independent experiments.
Sandwich Enzyme-Linked Immunosorbent Assays for Secreted Cytokine and Chemokine Protein Content
Sandwich enzyme-linked immunosorbent assays for CINC, macrophage inflammatory protein-1, and monocyte chemoattractant protein-1 have been developed in our laboratory, and the methods for quantitating protein were published previously [16–18]. The TNF- enzyme-linked immunosorbent assay was run per manufacturer's protocol (R&D Systems, Minneapolis, MN). The linear sensitivity range of the assays have been determined, and the assays show no crossreactivity. Samples and standards were run in triplicate, and well-to-well variation did not exceed 5%. All enzyme-linked immunosorbent assays were performed three separate times on different samples to ensure consistency of results. Three animals per group were studied with each enzyme-linked immunosorbent assay run.
Arterial Partial Pressure of Oxygen Levels
At the end of the experimental protocol, the right hilum of rats was ligated for 10 minutes through a median sternotomy, and subsequently arterial blood was evaluated by a blood gas analyzer (Chiron Diagnostics, Emeryville, CA). Blood gases were obtained from all animals under equivalent conditions of minute ventilation. Three groups were studied (the vehicle-treated positive controls and the hirudin-treated animals at 4 hours of reperfusion, and the surgically unmanipulated negative control animals), with 4 animals per group.
Immunohistochemistry
Whole-lung tissue specimens were immediately fixed in 10% formalin. Tissue samples were dehydrated through graded alcohol baths and embedded in paraffin. Specimens were cut in 5-µm sections and baked overnight at 50°C, and then rehydrated through graded baths to a final distilled water wash. Specimens were then blocked with 5% serum for 30 minutes at 37°C. The primary antibody was applied at the concentration predetermined by titration experiments. The primary antithrombin antibody (Peprotech, Rocky Hills, NJ; 8 µg/mL) was incubated for 1 hour at 37°C. Manufacturer stock secondary antibody (Vector Laboratories, Burlingame, CA) was then applied at a dilution of 1:250 and incubated for 30 minutes at 37°C. After the secondary antibody incubation, the avidin–biotin–peroxidase complex conjugate (Vector Laboratories) was applied. Staining was performed with diaminobenzidine tetrahydrochloride. Sections were rinsed in running tap water for 10 minutes, dehydrated, cleared, and mounted with permanent mounting media. Stained sections were examined using the image analysis software, Image Pro Plus (Media Cybernetics, Silver Spring, MD). Each run of immunohistochemistry studied more than six sections per representative lung, with four rodent lungs being studied per group. Immunohistochemistry was performed at two separate occasions to confirm the results.
Statistical Analysis
All data are presented as mean value ± the standard error of the mean unless otherwise designated. Comparisons among multiple groups were made using one-way analysis of variance with a post hoc Bonferroni modification for multiple comparisons. Statistical differences among groups were assessed using a two-tailed Student's t test. Statistical significance was defined as p less than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
Myeloperoxidase Content
Myeloperoxidase content increased significantly after ischemia and 4 hours of reperfusion (p = 0.0005). Hirudin did not significantly alter lung myeloperoxidase content in injured lungs relative to positive control lungs at 4 hours of reperfusion (Table 1).
Bronchoalveolar Lavage Cell Counts
When compared with negative control lungs, there was a significant increase in BAL leukocyte counts in positive control lungs at 4 hours of reperfusion (p < 0.0001). Hirudin reduced alveolar inflammatory cell counts by 24% (p < 0.01) relative to vehicle-treated control lungs.
Arterial Partial Pressure of Oxygen Levels
The mean arterial partial pressure of oxygen of positive control animals was 262 ± 19 mm Hg, whereas hirudin increased the partial pressure of oxygen to 387 ± 12 mm Hg at 4 hours of reperfusion. The improvement was statistically significant (p = 0.04) when comparing these two groups. Negative control animals demonstrated a mean arterial partial pressure of oxygen level of 493.3 ± 17 mm Hg.
Electromobility Shift Assay for Nuclear Factor Kappa B, Activator Protein-1, and Early Growth Response
Electromobility shift assay for NFB, AP-1, and EGR transactivation was performed after 90 minutes of ischemia and either 15 minutes or 4 hours of reperfusion. The early reperfusion time was chosen as significant transcriptional regulation of proinflammatory genes would need to occur early to regulate tissue damage known to be fully manifest by 4 hours of reperfusion in this model [16–18]. Representative electromobility shift assay gels are shown for NFB (Fig 2) and AP-1 (Fig 3). There was a dramatic increase in nuclear translocation for EGR early (15 minutes) and late (4 hours) in reperfusion, with the predominant isoform being EGR-1 by supershift analysis. At both reperfusion times, hirudin returned EGR-1 activity to near baseline levels. Although hirudin limited the nuclear translocation of AP-1 and EGR-1 both early and late in reperfusion relative to time-matched control animals, it appeared to have no effect on NFB activity.
|
|
, macrophage inflammatory protein-1, and monocyte chemoattractant protein-1, into the injured left lung alveolar space (Table 2). However, only macrophage inflammatory protein-1 and CINC secretion were reduced significantly by hirudin, as TNF- and monocyte chemoattractant protein-1 secretion were not dramatically altered relative to vehicle-treated control animals.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
B and TNF- for complete development [4-6, 17, 18, 22, 23]. Requirements for other cytokines and chemokines have also been demonstrated in LIRI [2, 6, 16–20]. However, blockade of individual mediators, including TNF- or NFB, provides only partial protection against LIRI, secondary to inflammatory redundancy [6, 19, 20]. This suggests there are other important mediators or pathways independent of NFB/TNF- activated by LIRI that promote tissue injury.
Lipopolysaccharide-induced lung injury requires inflammatory signaling through tissue factor and thrombin in a manner that appears to be independent of TNF- and NFB [24]. This is provocative as it may explain the aforementioned incomplete protection seen with NFB inhibition or cytokine (including TNF-)–chemokine blockade [25]. With respect to systemic lipopolysaccharide in the lung, blockade of extravascular coagulation does appear to independently attenuate the inflammatory response [17], interrupting the self-escalating inflammatory processes induced by sepsis. Blockade of extravascular coagulation also protects against injury that develops after intratracheal lipopolysaccharide instillation [24]. This was associated with reductions in interleukin 1 and interleukin 6 proteins, whereas TNF- production was again not affected. The fact that blockade of coagulation conferred protection against lung injury by a mechanism independent of NFB and TNF- is notable because both are required for injury in numerous models of acute lung injury [24, 26–29]. Interventions aimed at antagonism of multiple inflammatory and coagulation pathways, especially at the transcriptional level, would hypothetically be more effective than strategies aimed at antagonizing a single pathway or mediator. To that end, we are encouraged in that blockade of thrombin significantly limited the transactivation pattern of both AP-1 and EGR-1. With respect to protein elaboration, although TNF- was not modulated, CINC and macrophage inflammatory protein-1 were reduced by hirudin. It has been known for some time that de novo TNF protein translation requires NFB transactivation under conditions of inflammatory stress, such as ischemia–reperfusion injury. The fact that CINC and macrophage inflammatory protein-1 protein elaboration were modulated by hirudin suggests that these chemokines may be more dependent on AP-1 and EGR-1 nuclear translocation than that of NFB under these conditions. These results are exciting as they may provide a molecular target whereby a therapeutic intervention could be developed that would limit multiple synergistic inflammatory pathways at the transcriptional level. This could potentially be achieved if thrombin inhibition were coupled with a known NFB inhibitor. This may partially overcome redundancy inherent to inflammation.
It is not surprising that thrombin localized to macrophages and endothelial cells early in reperfusion. Both cell types are known to be important in promoting proinflammatory events after oxidative stress, as well as in LIRI in particular, and both participate in coagulation. By modulating early inflammatory activation events in these cell types, hirudin likely limited the transcriptional machinery that ultimately resulted in secretion of chemokines into the alveolar space and sequestration of inflammatory cells in the alveolus at 4 hours of reperfusion. Interestingly, myeloperoxidase content of injured left lungs was not altered by hirudin, whereas BAL inflammatory cell counts were reduced. It is not entirely clear why the myeloperoxidase content of injured lungs was the only pathologic indicator not affected by hirudin in this work. When analyzing injured lung sections by histology (data not shown), it was apparent that there were comparable numbers of neutrophils in the interstitial and intravascular spaces in both the hirudin-treated and control sections at 4 hours of reperfusion, whereas there were dramatic reductions in alveolar inflammatory cells in the hirudin-treated animals. This implies that neutrophils are able to traverse the endothelium into the interstitial space, but are then limited in their ability to gain access to the alveolus. Further clarification of these findings is warranted if this trend continues in other preclinical studies using thrombin inhibitors.
There are limitations to this study, which predominantly involve the model used. Although more clinically relevant than ex vivo or buffer-perfused systems, we are limited in that this is a warm, in situ model. Lung transplantation is performed clinically using cold preservation solutions. Therefore, an important limitation of our results may involve the fact that this is a warm ischemia model. Although large-animal, orthotopic transplantation models are more clinically relevant, they can be cost prohibitive. However, the mechanistic insights generated by the molecular studies are generally applicable, and the ability to perform multiple experiments reduces variability of findings. The end points of vascular permeability, myeloperoxidase content, and BAL leukocyte counts are pathologic variables. Physiologic assessments of lung function would add strength to our findings, but we are limited in our ability to measure physiologic variables, a difficult task to perform in rodent models. Lastly, hirudin can be a difficult anticoagulant to monitor clinically. We chose a doubling of the activated partial thromboplastin time throughout the experimental course as a sign of adequate thrombin inhibition for the studies involved in this work, similar to that published previously [30]. Hirudin is a potent, specific thrombin inhibitor and as such is a useful experimental tool to delineate the cellular mechanism involved in protection from acute lung injury. Further preclinical work in large-animal models may use more novel and less tedious thrombin inhibitors that are currently available. Unfortunately, at present, their effects are less specific.
In conclusion, hirudin provides the first evidence in an LIRI model of crosstalk between the inflammatory and coagulation cascades. There may be other molecules not yet defined after ischemia–reperfusion (ie, activated protein C), which also mediate crosstalk between inflammation and thrombosis. The protection afforded by hirudin appears to involve a mechanism dependent on EGR-1 and AP-1 transactivation, and independent of the NFB and TNF- pathway. These are intriguing preliminary findings, as limiting extravascular coagulation (AP-1 and EGR-1) with hirudin, especially when coupled with NFB inhibition, may afford even greater levels of protection against acute LIRI as inflammatory redundancy is further antagonized.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  B.-S. Ding, N. Hong, M. Christofidou-Solomidou, C. Gottstein, S. M. Albelda, D. B. Cines, A. B. Fisher, and V. R. Muzykantov Anchoring Fusion Thrombomodulin to the Endothelial Lumen Protects against Injury-induced Lung Thrombosis and Inflammation Am. J. Respir. Crit. Care Med., August 1, 2009; 180(3): 247 - 256. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Lau, Y. Zhao, J. Kim, I. L. Kron, A. Sharma, Z. Yang, V. E. Laubach, J. Linden, G. Ailawadi, and D. J. Pinsky Enhanced fibrinolysis protects against lung ischemia-reperfusion injury. J. Thorac. Cardiovasc. Surg., May 1, 2009; 137(5): 1241 - 1248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. S. Wolf, H. E. Merry, A. S. Farivar, A. S. McCourtie, and M. S. Mulligan Stress-activated protein kinase inhibition to ameliorate lung ischemia reperfusion injury J. Thorac. Cardiovasc. Surg., March 1, 2008; 135(3): 656 - 665. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
