| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;79:1010-1016
© 2005 The Society of Thoracic Surgeons
University of Washington Medical Center, Department of Surgery, Division of Cardiothoracic Surgery, Seattle, Washington
Accepted for publication August 30, 2004.
Bnuclear factor kappa B; PAECpulmonary artery endothelial cells; SEMstandard error of mean; TACtacrolimus; TNFtumor necrosis factor
* Address reprint requests to Dr Mulligan, University of Washington Medical Center, Division of Cardiothoracic Surgery, 1959 NE Pacific St., Box 356310, Seattle, WA 98195 (E-mail: msmmd{at}u.washington.edu).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: Isolated rat PAEC were subjected in vitro to 2 hours hypoxia followed by up to 4 hours reoxygenation. Cells were pretreated with CSA or a cremaphor vehicle. Differences in activation of signaling kinases and transcription factors were assessed, as was cytokine-chemokine protein secretion.
RESULTS: There was significant signaling kinase (extracellular signal regulated kinase [ERK 1/2]) activation by 15 minutes reoxygenation, which was temporally associated with marked activation of the transcription factors nuclear factor kappa B (NFB) and early growth response one (EGR-1). At 4 hours reoxygenation there were significant increases in chemokine protein secretion. The CSA decreased ERK 1/2 phosphorylation and significantly attenuated transcription factor transactivation at 15 minutes reoxygenation. The CSA was found to be selective in reducing cytokine-chemokine elaboration at 4 hours reoxygenation.
CONCLUSIONS: Hypoxia-reoxygenation induces ERK 1/2 phosphorylation, as well as transactivation of the transcription factors NFB and EGR-1 in PAEC. Cyclosporine selectively reduces proinflammatory mediator secretion, likely by transcriptional regulation through NFB and EGR-1. This is the first demonstration of ERK 1/2 inhibition afforded by CSA.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Recently there have been multiple publications characterizing the contributions of individual cell types to the development of LIRI [3, 4]. Lung ischemia-reperfusion injury in rats is known to be biphasic in nature, with an early neutrophil-independent phase, and a later neutrophil-dependent phase [5]. Alveolar macrophages (AM) appear to orchestrate early activation signaling events upon reperfusion. They are the predominant early source of tumor necrosis factor alpha (TNF-
) and interleukin one beta (IL-1ß), and oxygen and nitrogen-derived free radicals [6, 7]. Oxidative stress appears to prime many of the constituent cell populations [8], with macrophage secretory products amplifying the inflammatory responses of epithelial and endothelial cells (unpublished data). Furthermore, there is likely a positive feedback loop to the AM, which also serves to augment AM humoral response (monocyte chemoattractant protein [MCP-1], macrophage inflammatory proteins [MIP-1, MIP-2], cytokine induced neutrophil chemoattractant [CINC]) later in reperfusion. These intercellular signaling patterns implicate the macrophage as the early orchestrator of inflammatory tissue injury after reperfusion. However, depletion of AM using intratracheal instillation of liposomes containing clodronate, or inactivation of AM using gadolinium chloride only confers partial protection against the development of tissue injury, suggesting a role for other proinflammatory effector cells [8].
Both neutrophils and lymphocytes also contribute to the development of tissue injury after ischemia and reperfusion (IR), and have been implicated in the later phase of lung reperfusion injury [5, 9]. However, recent strategies attempting to deplete or inactivate neutrophils (ie, leukocyte-depleting filters or pentoxyfilline), also confer only moderate protection against LIRI [10]. de Perrot and colleagues [9] studied the effects of lymphocyte depletion utilizing nude donor and recipient rats in an isograft model of LIRI, and found that while recipient T lymphocytes do contribute to late phase reperfusion injury, their depletion only partially prevented injury.
Endothelial cell activation is centrally important in promoting inflammatory responses to ischemia-reperfusion injury in other vascular beds, most notably that of the heart and kidney [11, 12]. In those organ systems, the endothelium responds to oxidative stress by activating inflammatory pathways that transduce messages from the cell surface to the nucleus. These events result in transactivation of proinflammatory transcription factors such as nuclear factor kappa B (NFB) and early growth response one (EGR-1), with eventual transcription and translation of proinflammatory genes [13]. The end result is secretion of an array of vasoactive, procoagulant, and proinflammatory mediators that can induce such clinical problems as acute vasospasm, IR injury, coagulopathy, and systemic inflammatory response syndrome after coronary artery bypass grafting [14]. Because endothelial cells respond continuously to extracellular stimuli, such as circulating chemokines and cytokines, and are known to upregulate adhesion molecule genes after lung IR, it would be logical to study their individual contribution to LIRI. The actual contribution of pulmonary artery endothelial cells to LIRI is still undefined. Characterization of the proximal signaling pathways that eventually result in transcription factor transactivation and proinflammatory mediator secretion are subjects of ongoing studies.
Mitogen activated protein kinases (MAPK), including extracellular signal related kinase (ERK 1/2), c-Jun N-terminal kinase (JNK), and p38 kinase, are signaling proteins that mediate cellular responses to extracellular stimuli [15], and are known to govern such protean cellular functions as proliferation and differentiation, apoptosis, and responses to oxidative stress [16]. The NFB and EGR-1 directly upregulate expression of adhesion molecule and cytokine genes in cerebral and myocardial endothelial cells in response to oxidative stress [17], and have been shown in studies to be activated by MAPK in response to hypoxia and reoxygenation (H&R) [18, 19]. It may be, therefore, that MAPK activation is a proximate and fundamental signaling event that promotes transcriptional activation and the amplification of inflammatory pathways in pulmonary artery endothelium. If so, MAPK activation in pulmonary artery endothelial cells (PAEC) could contribute significantly to the ultimate development of lung injury after reperfusion.
Calcineurin inhibitors, such as cyclosporine (CSA), have been found to have antiinflammatory properties in addition to their immunomodulatory effects [20]. Interestingly, we have demonstrated that this antiinflammatory effect occurs at a dose one-tenth of that needed for immunomodulation in lymphocytes, and involves modulation of a signaling pathway dependent upon NFB and EGR-1 transactivation [9, 21]. This diverges from the well-published immunomodulatory pathway afforded by calcineurin inhibition that acts by attenuation of nuclear factor of activated T cells (NFAT) activation. At doses used in this and other studies from our laboratory with calcineurin inhibition, NFAT nuclear translocation does not appear to be affected by our dosing regimen after the stimulus of ischemia-reperfusion to the lung [22, 23].
Given the difficulty in determining the contribution of PAEC to tissue injury in vivo, we chose to isolate primary cultures of rat PAEC to determine not only their response to H&R at the pretranscriptional and protein levels, but also whether calcineurin inhibition with CSA modulated that response. Since it has been previously demonstrated that the transcription factors NFB and EGR-1 upregulate proinflammatory genes in LIRI and other models of IR injury, that MAPK are activated by oxidative stress, and that MAPK directly phosphorylate these transcription factors, we hypothesized that MAPK activation would induce transcription factor nuclear translocation in PAEC subjected to H&R. These events would lead to eventual inflammatory chemokine production. Furthermore, we hypothesized that CSA would be protective against this response, reducing eventual proinflammatory protein secretion by attenuation of MAPK and transcription factor activation.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Primary Rat Pulmonary Artery Endothelial Cell Culture
Pathogen free 21-day-old Long-Evans rats (Simonsen Labs, Gilroy, CA) were used for all experiments. Animals were euthanized with 120 mg/kg of intraperitoneal pentobarbital. A 14-gauge angiocatheter was inserted into the trachea through a midline neck incision and secured with a 4-0 braided silk suture. A median sternotomy was performed and the heart-lung block was rapidly excised. Intratracheal lung lavage was performed 15 times with 6 to 9 mL volumes of phosphate buffered saline containing 0.25 mmol/L ethylenediaminetetraacetic acid to deplete alveolar macrophages [22]. Two millimeter strips of peripheral lung were removed from all lung lobes. Tissue was minced, rinsed in Roswell Park Memorial Institute (RPMI) media, transferred to a dispase (10 mg/mL) solution, and then incubated for 60 minutes at 37°C. The cell suspension was homogenized and incubated for an additional 5 minutes at 37°C. Ten milliliters of complete media containing 10% fetal bovine serum (FBS) was added to terminate the reaction, and the cellular suspension was then filtered through a 100µm mesh. The filtrate was spun at 800g for 8 minutes, resuspended in supplemented RPMI media, and plated on gelatin-coated culture dishes.
The media was changed every 48 hours during the incubation period until confluent [23]. Cells were labeled for 8 hours with 4 µg/mL of acetylated low density lipoprotein, which binds selectively to endothelial cells. Cells were then separated using flow cytometry (FAC STAR Plus; Becton Dickinson, Franklin Lakes, NJ), which facilitated maintenance of a pure culture of endothelial cells. All cells used in these experiments were from passages 4 to 8. Cells were plated in 12-well plates at a concentration of 500,000 cells/well, and were considered optimal for experiments just before they reached confluence.
One hour before performing the H&R experiments, cells were pretreated in a 37°C incubator with either cyclosporine (500 ng/mL) or cremaphor (6,500 ng/mL), the vehicle in which CSA is solubilized. These agents were added at appropriate concentrations directly to the cell media. Since it has previously been shown that cremaphor induces a capillary leak both experimentally [24] and clinically [25], wells pretreated with cremaphor vehicle served as controls at all time points studied.
Hypoxia and Reoxygenation
Plated cells were placed in a humidified hypoxic chamber (Coy Lab Products, Ann Arbor, MI) at a partial percentage of oxygen of 0.5% for 2 hours. Hypoxic RPMI media supplemented with 5% FBS was used in all experiments (media was allowed to equilibrate overnight in the chamber). Reoxygenation was achieved by removing the plate from the hypoxic chamber and placing it into a normoxic, normothermic, humidified incubator for up to 4 hours. Media reaches atmospheric Po2 within 5 minutes of removal from the hypoxic chamber [21]. For all experiments, cell viability, as determined by trypan blue exclusion, was greater than or equal to 95%. Negative controls refer to cells that are unstimulated by conditions of hypoxia and reoxygenation, while positive controls are cremaphor-treated cells that have undergone hypoxia and reoxygenation for the specified time.
Electromobility Shift Assay (EMSA)
Ten microgram aliquots of nuclear protein were incubated in a binding reaction with a double stranded 32P end-labeled oligonucleotide containing the consensus NFB binding sequence 5'-GCCATTGGGGATTTC-CTCTTTACTGG-3' (Promega, Madison, WI) or the consensus EGR binding sequence 5'-GGATCCAGCGGGGAGCGGGGGCGA-3' (Santa Cruz Biotechnology, Santa Cruz, CA) [26]. The binding reaction was carried out as published previously [4, 21]. Supershift analysis for the EGR-1 isoform was performed as described by the Santa Cruz protocol and as published by us previously [4]. Multiple samples for each condition were analyzed and results were verified in at least three independent experiments. Densitometry was performed with Image J software (Version 1.2, Silver Spring, MD) to assess relative signal intensity.
Western Blotting
Total cellular protein was used for all MAPK immunoblotting studies, and protein was extracted as described previously [4]. Protein concentration was determined using the bicinchoninic acid assay (Pierce, Rockford, IL). Twenty micrograms of protein were electrophoresed, transferred, and blocked as previously described [4]. Membranes were probed with a phosphospecific ERK 1/2 (Thr202/Tyr204) antibody, a phosphospecific p38 (Thr180/Tyr182) antibody (New England Biolabs, Beverly, MA), or an antiactive JNK (Thr183/Tyr 185) polyclonal antibody (Promega Corp, Madison, WI) overnight at 4°C. These westerns were run alongside others that were probed with the corresponding antibody against total levels of p38, JNK, and ERK 1/2 in order to confirm their actual presence in the PAEC and verify equal protein loading in each lane. The membranes were developed as we have published previously [4]. Multiple samples for each condition were analyzed, and results were verified in at least three independent experiments.
Enzyme Linked Immunosorbent Assay (ELISA)
At the conclusion of the H&R protocol, the experimental media was removed and spun at 1,000g for 2 minutes. The resultant supernatant was snap frozen after the addition of a protease inhibitor cocktail (described previously [4, 21]) until used for ELISA studies. The ELISAs for IL-1ß were performed per the manufacturer’s (R&D Systems, Minneapolis, MN) suggested protocol. Sandwich ELISAs for CINC and MCP-1 were developed as published by our group previously [4, 21]. The linear sensitivity range of the assays for these chemokines was determined and the assays showed no cross reactivity with one another. Samples and standards were run in triplicate, and well-to-well variation did not exceed 5%. The results are the mean of three independent experiments, with each time point being represented by at least three separate samples.
Statistical Analysis
All data are presented as mean values ± the standard error of the mean unless otherwise designated. Comparisons between groups were made using one-way analysis of variance, with a Bonferroni modification performed for multiple comparisons. Differences between individual groups were determined with a two-tailed Student’s t test. Statistical significance was defined for all tests as a p less than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
EMSA for NFB and EGR-1 Activation
Since we have previously demonstrated in vivo that NFB activation promotes the expression of multiple proinflammatory mediators, we chose to study this transcription factor in this set of experiments [8, 20]. Furthermore, EGR-1 transactivation has previously been shown to be induced by ERK 1/2 activation in a murine model of LIRI, and in that study EGR-1 was localized to the lung vasculature using immunohistochemical techniques [28]. Therefore, the patterns of in vitro activation of both NFB and EGR-1, including their modulation by CSA, were studied by EMSA in PAEC subjected to H&R.
As shown in Figure 2, there was significant nuclear translocation of NFB in the vehicle treated cells that underwent 2 hours of hypoxia and 15 minutes of reoxygenation. The cold competition and assay positive control lanes verified the band as NFB. The CSA pretreatment dramatically reduced NFB nuclear translocation relative to the vehicle controls at this time. Densitometry confirmed significant reductions in NFB activation after CSA treatment. (Fig 2).
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
B and EGR nuclear translocation in inflammatory models [28]. However, before these studies, the specific contribution of PAECs to LIRI was poorly understood. Additionally, there has been no work to date describing PAEC specific MAPK activation patterns in response to H&R.
In rat PAEC subjected to H&R, there does not appear to be any significant activation of p38 or JNK. However, there was strong activation of ERK 1/2. This is analogous to what is seen in type II pneumocytes subjected to similar conditions of H&R [4]. The ERK 1/2 activation is classically believed to be most functional in promoting cellular differentiation and survival [29], while JNK and p38 activation generally were thought to promote inflammatory responses [28, 30]. However, MAPK activation patterns appear to be very tissue and stimulus specific in nature, and recent evidence suggests that ERK 1/2 may also participate in inflammatory signaling [28]. The peak in ERK 1/2 activation 15 minutes after reoxygenation correlates temporally with early and significant NFB and EGR-1 transactivation seen in PAEC. The fact that CSA reduces the phosphorylation of ERK 1/2, the downstream transactivation of both NFB and EGR-1, and the eventual secretion of CINC protein, indirectly implies that CSA in part limits inflammatory activation by ERK 1/2 inhibition. Work with specific ERK 1/2 inhibitors, such as U0126, will help more directly assess whether activation of ERK 1/2 is directly associated with the nuclear translocation of NFB and EGR-1 in this model, as well as downstream secretion of CINC protein.
Transcriptional regulation through NFB controls a wide range of vital functions, including cellular proliferation, apoptosis, regulation of immune function, and inflammatory mediator production [31]. Therefore, it is not surprising that NFB plays a central role in pulmonary endothelium activation. The activation of this transcription factor promotes transcription of a number of proinflammatory cytokine, chemokine, and cell-adhesion molecule genes [32]. If this transcription factor is consistently important in the activation of other lung cell populations after oxidative stress, it may be a potential therapeutic target in the setting of LIRI. Calcineurin inhibition appears to exert its antiinflammatory effects partly through a reduction of NFB activity. The use of CSA, a readily available clinical compound, in this context has significant translational implications.
Early growth response-one is known to be critical in the production of hypoxia-induced acute pulmonary injury [33]. Yan and colleagues [28], in a murine lung reperfusion model, found that EGR-1 expression was increased in the pulmonary vasculature after IR, and that EGR-1 null mice demonstrated decreased expression of intercellular adhesion molecule-1 as well as reduced secretion of an alpha chemokine, macrophage inflammatory protein-2. Furthermore, a link between ERK 1/2 activation and EGR-1 transactivation has been published previously. An inhibitor of ERK (PD98059) markedly reduced EGR-1 translocation to the nucleus in RAW 264.7 monocytes stimulated with lipopolysaccharide. The ERK 1/2 inhibition ultimately resulted in an abrogation of TNF- messenger ribonucleic acid expression in that study [34]. Therefore, it is likely that ERK 1/2 activation mediates EGR-1 dependent transcriptional activation of the alpha chemokine gene CINC in PAEC subjected to H&R. Conversely, attenuation of ERK 1/2 activation by CSA may in turn limit the transactivation of EGR-1, and the eventual secretion of CINC.
Calcineurin inhibition has been shown to be protective against experimental ischemia-reperfusion injury of the brain, heart, and liver [35]. The proposed mechanism for these effects fall into three broad categories, involving mitochondrial stabilization, modulation of endothelin and nitric oxide activity, and finally transcriptional regulation [6, 8, 12, 36]. Cyclosporine has been shown to block mitochondrial permeability subsequent to increases in intracellular calcium, yet most of this work has been performed in tissues rich in mitochondria (such as the liver and heart), while the lung is known to have relatively lower mitochondrial concentrations [37]. Second, attempts have been made to identify relationships between changes in tissue microcirculation and alterations in endothelin-1 and nitric oxide production [14]. However, conclusions drawn from the published studies have been inconsistent and while it is possible that calcineurin inhibition is protective against LIRI by a mechanism partly involving these mediators, this is unlikely the central, protective mechanism of its action in the lung.
Based on previous work from multiple laboratories [7, 8, 21], it is likely that protection afforded by calcineurin inhibition centrally involves transcriptional regulation of proinflammatory genes. We have previously demonstrated that calcineurin inhibition protects against the development of in vivo LIRI, and this was mediated, in part, by an attenuation of NFB-dependent transcriptional activation of chemokine and cytokine genes [20]. Alveolar macrophages, subjected to in vitro H&R, also responded with TNF-, IL-1ß, MCP-1, MIP-2, and CINC production [21]. This response is NFB-dependent, and all but IL-1ß are reduced by CSA pretreatment. In PAEC, CSA reduces proinflammatory protein secretion after H&R and studies are presently underway to validate these findings using tacrolimus, the other potent calcineurin inhibitor commonly used in tranplantation. Therefore, the administration of CSA in the future to lung donors may have therapeutic benefit in protecting against LIRI by reducing cytokine and chemokine protein secretion from alveolar macrophages and pulmonary artery endothelium upon reperfusion of the allograft, as both cell types appear to be rendered transcriptionally quiescent by this intervention. The effect of CSA on cells in the alveolar and vascular compartment of the lung has translational implications regarding nebulized and/or parenteral treatment of lung donors.
Due to its insolubility in water, CSA is dissolved in a vehicle called cremaphor EL. Cremaphor consists of a polyethoxylated castor oil at a concentration of 650 ng/mL with 33% ethanol. Cremaphor has been implicated clinically in anaphylactic reactions and the release of histamine [38]. Experimentally, CSA has been shown to be directly toxic to endothelial cells at concentrations greater than 10 µg/mL [39]. We chose to use a relative dosage far beneath what is utilized clinically with CSA for immunomodulatory purposes, and determined appropriate dosage based on drug levels in vivo after intravenous CSA infusion. The CSA dosage in vivo is adjusted so that trough levels are close to 0.1 µg/mL, while levels as high as 5 µg/mL are reached for 1 to 2 hours after intravenous bolus infusion of 5 mg/kg [40]. Because we pretreated our PAEC with 500 ng/mL of CSA, a dose that is one tenth of what is achieved in the serum of humans clinically shortly after bolus IV infusion, it is apparent that smaller doses are able to provide potent antiinflammatory effects that are distinct from its immunomodulatory properties. In other words, while CSA provides immunomodulation by an NFAT-IL-2 pathway, there also is an antiinflammatory pathway in PAEC that involves attenuated NFB and EGR-1 activation, which in turn likely limits CINC secretion.
In conclusion, rat PAEC contribute to tissue injury in vivo by dramatic increases in CINC secretion, as well as MCP-1. The ability of CSA to reduce PAEC activation through MAPK inhibition, combined with its known effects on alveolar macrophages, make it an intriguing agent to potentially give to lung donors in an attempt to reduce IR associated lung injury after transplantation. Further work with ERK 1/2 inhibitors will expand our knowledge of multilevel inflammatory pathways in PAEC.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
release from the pulmonary macrophage in lung ischemia-reperfusion injury J Thorac Cardiovasc Surg 2004;127:1502-1508. from resident lung cells is a key initiating factor in pulmonary ischemia-reperfusion injury J Thorac Cardiovasc Surg 2004;127:541-547.B, activator protein-1, and mitogen-activated protein kinases in lung parenchyma FASEB J 2003;17:1800-1811.B and expression of ICAM-1 in ischemic-reperfused canine myocardium J Mol Cell Cardiol 2001;33:109-119.[Medline]
B Adv Exp Med Biol 2002;507:503-508.[Medline]
B in the cardioadaptive response to ischemia Ann Thorac Surg 1999;68:377-382.B activation: a reassessment of the evidence in the light of recent discoveries Biochem Pharmacol 2000;59:13-23.[Medline]
B—a pivotal transcription factor in inflammatory diseases N Engl J Med 1997;366:1066-1071.
This article has been cited by other articles:
|
|
 |
|
  A. S. McCourtie, A. S. Farivar, S. M. Woolley, H. E. Merry, P. S. Wolf, B. Mackinnon-Patterson, J. C. Keech, E. FitzSullivan, and M. S. Mulligan Alveolar Macrophage Secretory Products Effect Type 2 Pneumocytes Undergoing Hypoxia-Reoxygenation. Ann. Thorac. Surg., December 1, 2008; 86(6): 1774 - 1779. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Wang, Y. Song, D.-B. Chen, and J. Zheng Protein Phosphatase 3 Differentially Modulates Vascular Endothelial Growth Factor- and Fibroblast Growth Factor 2-Stimulated Cell Proliferation and Signaling in Ovine Fetoplacental Artery Endothelial Cells Biol Reprod, October 1, 2008; 79(4): 704 - 710. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Bhattacharyya, S.-J. Chen, M. Wu, M. Warner-Blankenship, H. Ning, G. Lakos, Y. Mori, E. Chang, C. Nihijima, K. Takehara, et al. Smad-Independent Transforming Growth Factor-{beta} Regulation of Early Growth Response-1 and Sustained Expression in Fibrosis: Implications for Scleroderma Am. J. Pathol., October 1, 2008; 173(4): 1085 - 1099. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Ngiam, M. Post, and B. P. Kavanagh Early growth response factor-1 in acute lung injury Am J Physiol Lung Cell Mol Physiol, November 1, 2007; 293(5): L1089 - L1091. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Khomenko, S. Szabo, X. Deng, M. R. Jadus, H. Ishikawa, K. Osapay, Z. Sandor, and L. Chen Suppression of early growth response factor-1 with egr-1 antisense oligodeoxynucleotide aggravates experimental duodenal ulcers Am J Physiol Gastrointest Liver Physiol, June 1, 2006; 290(6): G1211 - G1218. [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 |
