help button home button
AJRCCM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published ahead of print on June 15, 2006, doi:10.1164/rccm.200509-1511OC
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Online Supplement
Right arrow All Versions of this Article:
200509-1511OCv1
174/6/646    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Demoule, A.
Right arrow Articles by Petrof, B. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Demoule, A.
Right arrow Articles by Petrof, B. J.
American Journal of Respiratory and Critical Care Medicine Vol 174. pp. 646-653, (2006)
© 2006 American Thoracic Society
doi: 10.1164/rccm.200509-1511OC

Original Article

Endotoxin Triggers Nuclear Factor-{kappa}B–dependent Up-regulation of Multiple Proinflammatory Genes in the Diaphragm

Alexandre Demoule, Maziar Divangahi, Linda Yahiaoui, Gawiyou Danialou, Dusanka Gvozdic, Katherine Labbe, Weisheng Bao and Basil J. Petrof

Meakins-Christie Laboratories, McGill University; and Respiratory Division, McGill University Health Center, Montreal, Quebec, Canada

Correspondence and requests for reprints should be addressed to Basil J. Petrof, M.D., Respiratory Division, Room L411, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, PQ, Canada H3A 1A1. E-mail: basil.petrof{at}mcgill.ca


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Sepsis-induced diaphragmatic force loss and failure are associated with an increased exposure of the muscle to proinflammatory mediators.

Objectives: Our objectives were to test the hypothesis that force-inhibiting mediators may arise in large part from the diaphragm itself and to evaluate the roles of mechanical stress, free radicals, and the nuclear factor (NF)-{kappa}B transcription factor pathway in endotoxin (LPS)-induced proinflammatory responses of the diaphragm.

Methods: Murine diaphragm and limb muscle cells were exposed to LPS in vitro and in vivo. Proinflammatory gene expression was measured using RNase protection assays (tumor necrosis factor [TNF]-{alpha}, TNF-{alpha} receptor p55, interleukin [IL]-1{alpha}, IL-1beta, IL-6, macrophage inflammatory peptide-2, intercellular adhesion molecule-1, Fas ligand, and inducible nitric oxide synthase) and ELISAs (TNF-{alpha}, IL-6, and macrophage inflammatory peptide-2). Cyclical muscle cell stretch and free-radical scavengers (N-acetylcysteine and catalase) were used to alter mechanical and oxidative stress levels, respectively. Pharmacologic (pyrrolidinedithiocarbamate) and dominant-negative transfection strategies were used to inhibit the NF-{kappa}B pathway.

Results: In primary diaphragm muscle cell cultures, modulation of mechanical stress levels or free-radical exposure did not alter responses to LPS stimulation. However, pharmacologic blockade of the NF-{kappa}B pathway and dominant-negative molecular inhibition of IKB kinase-beta strongly suppressed LPS-induced proinflammatory gene expression. In vivo, acute endotoxemia induced significantly greater mRNA and protein levels for proinflammatory mediators in the diaphragm as compared with limb muscle. Basal expression levels of proinflammatory genes were significantly higher in the diaphragm.

Conclusions: Constitutive and LPS-induced proinflammatory gene expression are exaggerated in the diaphragm compared with limb muscles and are critically dependent on the NF-{kappa}B pathway. We suggest the diaphragm may be relatively predisposed to proinflammatory responses.

Key Words: cytokines • diaphragm • inflammation • LPS • nuclear factor-{kappa}B

Acute respiratory failure occurs in a large proportion of patients with severe sepsis from various causes (1, 2), and diaphragmatic dysfunction has been shown to be a potential cause of ventilatory failure in this setting (3). The diaphragm is the main inspiratory muscle, and major diaphragmatic weakness has been demonstrated in several experimental models of sepsis, such as after intraperitoneal (4) or intravascular administration of LPS (5), cecal ligation/perforation-induced bacterial peritonitis (6), and Pseudomonas aeruginosa lung infection (7). Several mechanisms have been implicated in the pathogenesis of sepsis-induced respiratory muscle pump failure. Chief among these is an increased exposure of respiratory muscles to proinflammatory mediators such as reactive oxygen species (8, 9), nitric oxide (4, 6), and cytokines (e.g., tumor necrosis factor [TNF]-{alpha}) (10, 11). A number of proinflammatory mediators are able to cause direct inhibition of contractile mechanisms or myofiber protein loss with attendant muscle wasting (12–14). Although the cell signaling pathways that are involved in producing such effects in the septic diaphragm have not been clearly elucidated, the nuclear factor (NF)-{kappa}B transcription factor pathway is one plausible candidate for playing an important role because it acts as a central coordinator of the inflammatory response in other organs and cell types (for reviews, see References 15 and 16).

There is abundant experimental evidence linking the actions of proinflammatory mediators to skeletal muscle weakness and wasting in vitro and in vivo (10, 11, 17–20). In the setting of severe sepsis, cytokines and other proinflammatory mediators released from distant organs have the potential to enter the circulation and act upon the diaphragm in an endocrine fashion. Another potentially important source of cytokines is the diaphragm itself. Hence, the proinflammatory mediators, which suppress diaphragmatic function during sepsis, could arise in large part from the diaphragm per se, with consequent autocrine/paracrine actions on the constituent muscle fibers. This latter hypothesis is supported by studies pointing to an active participation of skeletal muscle cells in the response to different inflammatory stimuli. For example, human skeletal cell cultures constitutively express low levels of proinflammatory cytokines, such as interleukin (IL)-1 and IL-6 (21–23), and exposure to exogenous proinflammatory cytokines further up-regulates this expression and induces the expression of additional cytokines, such as TNF-{alpha} (23). In addition, it has been reported that TNF-{alpha} expression is up-regulated within rat diaphragm muscle fibers after LPS administration in vivo (11). Taken together, these findings suggest that the diaphragm could play an active role in the overproduction of proinflammatory mediators that contribute to its own mechanical failure during sepsis.

In the present investigation, our primary goal was to test the hypothesis that the diaphragm can act as a major participant in the overproduction of cytokines and other proinflammatory mediators during endotoxemia. In addition, we hypothesized that the magnitude of proinflammatory gene expression under these conditions would be greater in the diaphragm than in limb muscles because several studies have suggested that the diaphragm is more vulnerable to the adverse effects of sepsis in terms of depressed contractile function and myofiber injury (4, 6, 7, 24, 25). Finally, identification of the cellular mechanisms and signaling pathways responsible for inducing the expression of these mediators in the diaphragm could provide new insights into potential therapies for respiratory muscle dysfunction during sepsis. Therefore, we also sought to evaluate the influence of mechanical stress and reactive oxygen species and the role of the NF-{kappa}B transcription factor pathway in the pathogenesis of cytokine overproduction by diaphragm muscle cells during sepsis. Some of these data have been previously reported in the form of an abstract (26).


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
See the online supplement for additional details.

Animal Procedures
Male C57BL/6 mice aged 8–10 wk (23–26 g) were studied. LPS was administered at a dose previously shown to produce diaphragmatic dysfunction in mice (25 mg/kg intraperitoneally) (27, 28), and overall mortality was less than 15% before mice were killed. A sham-treated control group underwent the same procedure with an equivalent volume of saline. The mice were killed at 6, 12, or 24 h after injection. The diaphragm and limb muscle (tibialis anterior [TA]) were removed and stored at –80°C.

Muscle Cell Cultures
Primary diaphragm and TA muscle cell cultures were established as previously described using single living muscle fibers to isolate adult myoblasts (29). All experiments were performed on the fifth day after the induction of differentiation into myotubes. For LPS stimulation, the cells were washed with Dulbecco's modified Eagle's medium and stimulated for 4 h with Escherichia coli LPS (1 µg/ml) (30), used without serum alone or in combination with TNF-{alpha} (1 ng/ml), IL-1{alpha} (5 U/ml), and IFN-{gamma} (20 U/ml). To study the participation of oxidative stress, myotubes were treated with catalase (2,000 U/ml) or N-acetylcysteine (NAC) (10 mM) initiated 1 h before LPS exposure. The same experimental paradigm was used to treat LPS-exposed diaphragmatic myotubes with pyrrolidinedithiocarbamate (PDTC; at 100 µM), a compound that inhibits the ubiquitin E3 ligase function required for NF-{kappa}B activation (31). To apply mechanical stress, diaphragm muscle cells were grown on plates containing a flexible silicone base coated with type I collagen. By applying a vacuum to the base of the wells (Flexercell Strain Unit; Flexcell Corp., McKeesport, PA), myotubes were subjected to cyclical biaxial strain (10% elongation at a frequency of 1.5 Hz for 4 h) as previously described (5). A murine myoblast cell line (C2C12) was used for plasmid transfection experiments using lipofectamine. Dominant-negative mutant forms of the IkB kinases, IKK{alpha} and IKKbeta, were used to inhibit NF-{kappa}B activation as previously described (32).

Assays of Proinflammatory Mediator mRNA and Protein Expression
Total RNA was extracted, and reverse transcriptase–polymerase chain reaction (RT-PCR) was performed to detect Toll-like receptor (TLR)-4 expression using standard procedures. For RNase protection assays, 32P-labeled riboprobes were synthesized from commercial template panels (BD Bioscience, Pharmingen, San Diego, CA) and were directed against transcripts for the following molecules involved in the immune response: TNF-{alpha}, TNF-{alpha} receptor (p55), IL-1{alpha}, IL-1beta, IL-6, macrophage inflammatory peptide (MIP)-2, intercellular adhesion molecule (ICAM)-1, Fas ligand (Fas-L), and inducible nitric oxide synthase (iNOS). The autoradiographic signals were normalized to the L32 housekeeping gene to control for loading in each lane and expressed in arbitrary units. To quantify selected cytokines/chemokines at the protein level, commercially available ELISA kits were used according to the manufacturer's instructions (R&D Systems, Minneapolis, MN).

Detection of Inflammatory Cells in the Diaphragm
Myeloperoxidase activity in diaphragm tissue was assayed to evaluate the presence of activated neutrophils (7). Macrophages in the diaphragm were detected by immunohistochemistry and quantified by stereology as previously described (29).

Statistical Analysis
For in vivo and in vitro studies, six independent replicate analyses were performed for each experimental condition and time point unless otherwise noted. Groups were compared using the Wilcoxon or Mann-Whitney tests as appropriate, with statistical significance set at p < 0.05.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Constitutive Expression of ProInflammatory Genes Is Higher in the Diaphragm than in the Limb Muscle In Vivo
To compare the baseline constitutive levels of proinflammatory gene expression within the diaphragm and limb muscle in vivo, total RNA was extracted and analyzed by RNase protection assay. Figure 1 shows group mean quantification of mRNA signal intensities for TNF-{alpha} and its receptor, IL-1{alpha}, IL-1beta, IL-6, and MIP-2. In all cases, there was higher cytokine expression in the diaphragm than in the TA limb muscle under basal conditions (similar results were also obtained for two other hindlimb muscles, the extensor digitorum longus and soleus; data not shown). We also examined basal expression levels of ICAM, Fas-L, and iNOS. As with the cytokines, ICAM and Fas-L exhibited higher constitutive expression in the diaphragm, whereas iNOS expression did not differ between the diaphragm and TA under baseline conditions.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
  		Figure 1. Baseline mRNA expression levels for multiple proinflammatory genes are higher in the diaphragm than in limb muscle in vivo. All data are group means ± SE (n = 6 mice/group). Data are normalized to the L32 housekeeping gene and expressed as a fraction of the mean value (arbitrarily defined as = 1.0) obtained in the diaphragm. *p < 0.05 for comparisons between diaphragm and tibialis anterior muscles. Fas-L = Fas ligand; ICAM = intercellular adhesion molecule; IL = interleukin; iNOS = inducible isoform of nitric oxide synthase; MIP = macrophage inflammatory protein; TNF = tumor necrosis factor; TNF-{alpha}R = p55 TNF-{alpha} receptor.

 
Differential Proinflammatory Mediator Responses of Diaphragm and Limb Muscle to Acute Endotoxemia In Vivo
Quantification of mRNA signal intensities in the diaphragm and TA for the three different time points after LPS administration are shown in Figure 2 (see also online supplement). Some differences in the magnitude and timing of the individual proinflammatory gene responses were found, but the maximal induction of gene expression was generally observed at 6–12 h after LPS administration. Abnormally elevated expression levels were frequently present at 24 h after LPS delivery (Figure 2B). Furthermore, although LPS administration induced significant increases in proinflammatory gene expression within the diaphragm and the TA, the absolute magnitude of proinflammatory gene expression under these conditions was almost always greater in the diaphragm than in the limb muscle. In addition, the differences observed at the mRNA level were closely paralleled by results obtained at the protein level by ELISA for representative proinflammatory mediators (TNF-{alpha}, IL-6, and MIP-2; Figure 3). Finally, the increases in proinflammatory mediator expression by the diaphragm after LPS administration were not associated with an increased presence within the muscle of activated neutrophils or macrophages (see online supplement).


Figure 2
Figure 2
View larger version (76K):
[in this window]
[in a new window]
  		Figure 2. Multiple proinflammatory genes are up-regulated in the diaphragm and limb muscle during acute endotoxemia in vivo. (A) Proinflammatory gene expression levels in the diaphragm (filled bars) and tibialis anterior (open bars) were measured in sham-treated (S) and LPS-treated (L) mice at 6, 12, and 24 h after intraperitoneal injection. Data are normalized to the L32 housekeeping gene and expressed as a fraction of the mean value (arbitrarily defined as = 1.0) obtained in the sham-treated diaphragm. All data are group means ± SE (n = 6 mice/group). *p < 0.05 for comparisons between diaphragm and tibialis within the same condition (LPS or saline); {dagger}p < 0.05 for comparisons between sham- and LPS-treated mice within the same type of muscle (diaphragm or tibialis). ND = signal not detectable. (B) Representative autoradiograph of ribonuclease protection assay showing the responses of different proinflammatory genes within the diaphragm and tibialis anterior at 24 h after LPS administration. In many cases, the basal levels of proinflammatory gene expression observed in the diaphragms of sham-treated mice (i.e., Saline group) were similar to the tibialis muscles of LPS-treated mice.

 

Figure 3
View larger version (5K):
[in this window]
[in a new window]
  		Figure 3. Quantification of cytokine/chemokine protein levels in the diaphragm and limb muscle after LPS administration in vivo. Protein levels of selected proinflammatory mediators (TNF-{alpha}, IL-6, and MIP-2) were measured in the diaphragm (filled bars) and tibialis anterior (open bars) at 6 h after sham (S) or LPS (L) treatment. All data are group means ± SE (n = 6 mice/group). *p < 0.05 for comparisons between diaphragm and tibialis within the same condition (LPS or saline); {dagger}p < 0.05 for comparisons between sham- and LPS-treated mice within the same type of muscle (diaphragm or tibialis).

 
Responses of Diaphragm and Limb Muscle Cells to LPS Exposure In Vitro
RT-PCR showed similar signal intensities in diaphragm and TA for the LPS receptor, TLR-4, from in vivo tissues and from cultured myotubes in vitro (Figure 4). To determine whether diaphragm muscle cells are intrinsically more sensitive to LPS-induced proinflammatory gene up-regulation, we compared the responses of diaphragm and TA muscle primary cell cultures to direct LPS exposure in vitro. RNase protection assays were performed on primary cultures at the differentiated myotube stage after 4 h of LPS exposure. Although IL-6 and MIP-2 could be detected at very low levels in unstimulated cultures, there were no differences in constitutive expression between diaphragm and TA myotubes. After LPS exposure in vitro, several proinflammatory genes that demonstrated increased expression after LPS administration in vivo were up-regulated (TNF-{alpha}, IL-1{alpha}, IL-6, MIP-2, and iNOS). Moreover, although the diaphragm and limb muscle cultures responded in a similar fashion, the magnitude of responses seemed to be greater in the limb muscle cell cultures (Figure 5A).


Figure 4
View larger version (40K):
[in this window]
[in a new window]
  		Figure 4. The LPS receptor (TLR-4) is expressed by diaphragm and limb muscles in vivo and in vitro. Reverse transcriptase–polymerase chain reaction revealed the presence of Toll-like receptor 4 (TLR-4) transcripts in intact muscles (in vivo) and in cultured myotubes (in vitro) derived from the diaphragm and tibialis anterior muscles.

 

Figure 5
View larger version (13K):
[in this window]
[in a new window]
  		Figure 5. Quantification of proinflammatory gene mRNA levels after direct stimulation of diaphragm and limb muscle cells with LPS in vitro. (A) Proinflammatory gene expression levels in differentiated myotubes derived from the diaphragm (Di) or tibialis anterior (TA) after 4 h of exposure to 1 µg/ml endotoxin (LPS) or saline vehicle (CTL). *p < 0.05 for comparisons between diaphragm and TA myotubes within the same condition (LPS or saline); {dagger}p < 0.05 for comparisons between LPS and CTL within the same type of myotubes (diaphragm or TA). All data are normalized to the L32 housekeeping gene. ND = signal not detectable. (B) Proinflammatory gene expression levels in diaphragmatic myotubes after exposure to LPS alone (1 µg/ml), LPS + Stretch (10% elongation at 1.5 Hz for 4 h), or LPS + Cytokines (TNF-{alpha}: 1 ng/ml; IL-1{alpha}: 5 U/ml; IFN-{gamma}: 20 U/ml). All data are normalized to the L32 housekeeping gene and expressed as a fraction of the mean value (arbitrarily defined as = 1.0) obtained in the group exposed to LPS alone. *p < 0.05 for comparisons to the control (CTL) condition, which consisted of exposure to saline vehicle; {dagger}p < 0.05 for comparisons to the LPS group. ND = signal not detectable.

 
We determined whether additional factors present within the in vivo context, such as mechanical loading of muscle cells or the presence of other proinflammatory mediators classically associated with sepsis, could influence proinflammatory gene expression by diaphragm muscle cells. Diaphragmatic myotubes were subjected to cyclical stretch for 4 h during simultaneous LPS exposure. This had no demonstrable effect on the proinflammatory gene expression response to LPS (Figure 5B). On the other hand, after the cultures were exposed to a combination of LPS and prototypical sepsis-associated cytokines (TNF-{alpha}, IL-1{alpha}, and IFN-{gamma}), there was significantly greater up-regulation than with LPS alone for several of the proinflammatory genes examined (TNF-{alpha}, IL-1{alpha}, MIP-2, and iNOS).

Role of NF-{kappa}B and Reactive Oxygen Species in the Regulation of ProInflammatory Gene Expression by Diaphragm Muscle Cells Exposed to LPS
Because the transcription factor NF-{kappa}B and free radicals could be implicated in cytokine production by the diaphragm during sepsis, we examined the effects of a NF-{kappa}B inhibitor (PDTC) and classical free radical scavengers (NAC and catalase) on proinflammatory gene expression by LPS-stimulated diaphragmatic myotubes. PDTC had a dramatic effect in down-regulating the expression levels of several proinflammatory genes (Figure 6). These findings suggest that NF-{kappa}B plays an important role in LPS-induced cytokine production by diaphragmatic myotubes. On the other hand, NAC and catalase exerted no measurable effects, suggesting that LPS-induced cytokine production by diaphragm muscle cells was not mediated in a major way by reactive oxygen species.


Figure 6
View larger version (46K):
[in this window]
[in a new window]
  		Figure 6. Effects of pharmacologic inhibition of nuclear factor (NF)-{kappa}B and free radical scavengers on LPS-induced proinflammatory gene expression by diaphragm muscle cells in vitro. Ribonuclease protection assay showing the effects of the NF-{kappa}B inhibitor pyrrolidinedithiocarbamate (PDTC) and the free radical scavengers, N-acetylcysteine (NAC) and catalase, on proinflammatory gene expression by diaphragmatic myotubes exposed to LPS. The primary myotube cultures were pretreated with these compounds for 1 h before initiating LPS (1 µg/ml) stimulation for 4 h.

 
To further confirm the role of NF-{kappa}B in LPS-induced cytokine production by muscle cells, C2C12 cells were transfected with plasmids expressing dominant-negative mutant forms of IKK{alpha} and IKKbeta, which are involved in the "alternative" and "classical" activation pathways for NF-{kappa}B, respectively. Although the IKK{alpha} dominant-negative transfection had no significant effect, inhibition of IKKbeta led to a substantial down-regulation of proinflammatory gene expression by LPS-stimulated myotubes (Figure 7). These findings point to involvement of the classical activation pathway for NF-{kappa}B in LPS-induced cytokine production by skeletal muscle cells.


Figure 7
Figure 7
View larger version (50K):
[in this window]
[in a new window]
  		Figure 7. Effects of dominant-negative inhibition of the NF-{kappa}B pathway on LPS-induced proinflammatory gene expression by muscle cells in vitro. (A) Ribonuclease protection assay showing the effects of transfecting expression plasmids encoding dominant-negative constructs for IKK{alpha} and IKKbeta on proinflammatory gene expression by murine C2C12 myotubes exposed to LPS. Cells were transfected in an identical fashion with empty control plasmid (CTL). All myotube cultures were stimulated for 4 h with LPS (1 µg/ml) and immediately harvested for total RNA extraction. (B) Quantification of proinflammatory gene mRNA levels in LPS-stimulated C2C12 myotubes previously transfected with IKK{alpha} or IKKbeta dominant-negative mutant constructs. All data are normalized to the L32 housekeeping gene and expressed as a fraction of the mean value (arbitrarily defined as = 1.0) obtained in muscle cells transfected with the control plasmid (CTL). *p < 0.05 for comparisons to the CTL group.

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Contractile dysfunction and muscle fiber injury in the diaphragm are well described in sepsis, and proinflammatory mediators likely play an important role in the pathogenesis of these phenomena (4, 6, 8, 10, 11, 17). In the present investigation, we used a combination of in vivo and in vitro approaches to gain a better understanding of the pattern and regulation of proinflammatory gene expression by the diaphragm during endotoxemia. The specific objectives of our study were (1) to measure in vivo diaphragmatic expression levels of proinflammatory cytokines previously implicated in skeletal muscle weakness or wasting and other immunologically active factors that could play a role in the diaphragmatic dysfunction associated with severe sepsis; (2) to evaluate whether there are differential proinflammatory gene responses in the diaphragm and limb muscles within the intact endotoxemic animal and in primary myogenic cell cultures exposed to LPS; (3) to determine whether LPS-induced cytokine up-regulation by diaphragm muscle cells in vitro is modified by the level of exposure to mechanical stress or free radicals, both of which are associated with increased muscle activation in vivo and hence are expected to be increased in the septic diaphragm; and (4) to ascertain whether interference with the NF-{kappa}B transcription factor pathway, through pharmacologic or dominant-negative molecular inhibition, leads to a reduction in LPS-induced cytokine up-regulation by diaphragm muscle cells.

In vivo, our major novel findings were that (1) the diaphragm normally expresses a wide repertoire of proinflammatory genes even under nonseptic conditions, (2) the constitutive baseline expression levels for these genes is significantly higher in the diaphragm than in limb muscles, and (3) acute endotoxemia leads to significantly greater upregulation of proinflammatory genes in the diaphragm than in limb muscles. To determine whether the greater constitutive and LPS-induced cytokine expression by the diaphragm observed in vivo reflects intrinsic properties of myogenic cells derived from this muscle, primary cell cultures from diaphragm and limb muscle were compared. We found that (1) diaphragm and limb muscle cells expressed the receptor for LPS (TLR-4), with no greater intrinsic sensitivity of diaphragm muscle cell cultures to LPS-induced cytokine up-regulation; (2) in vitro modulation of mechanical stress and free radical exposure did not significantly modify the response to LPS stimulation; and (3) in contrast, modulation of the NF-{kappa}B transcription factor pathway had dramatic effects on cytokine production by LPS-stimulated muscle cells. Our data suggest a critically important role for the NF-{kappa}B transcription factor and its activation via the classical IKKbeta pathway in diaphragm muscle cells exposed to LPS and thus point to a potential therapeutic target for inhibiting cytokine up-regulation within the septic diaphragm.

Differential Responses of Diaphragm and Limb Muscles to LPS
Although diaphragm and limb muscle exhibited expression of multiple proinflammatory genes in response to LPS, constitutive and inducible expression levels were almost uniformly higher in the diaphragm in vivo. Moreover, stimulation of primary diaphragmatic muscle cell cultures by LPS led to up-regulation of IL-1{alpha}, IL-6, TNF-{alpha}, and MIP-2 in vitro, confirming the hypothesis that diaphragm muscle cells per se are able to produce cytokines and chemokines in response to endotoxin. The mRNA expression levels of these proinflammatory mediators were not altered by the addition of serum to the culture medium during LPS stimulation (data not shown). This suggests that accessory factors, such as exogenous LPS-binding protein or soluble CD14 (33), are not required, which is consistent with a previous study of IL-6 regulation in C2C12 myoblasts exposed to LPS (30).

To our knowledge, this is the first study to examine the proinflammatory gene response to LPS in primary cultures of diaphragm and limb muscle. Cultured muscle cells derived from different types of skeletal muscle have been reported to maintain certain distinct patterns of gene expression in vitro, suggesting inherent differences in their genetic programming (34). To determine whether this was the case for cytokine gene expression, diaphragm- and limb-derived myogenic cell characteristics were compared. In contrast to the in vivo situation, the levels of constitutive or LPS-induced cytokine expression in vitro were not greater in the diaphragm than in limb muscle cell cultures. Therefore, caution must by exercised in extrapolating between LPS exposure in vitro and endotoxemia in vivo. For example, a higher level of muscle activity by the diaphragm in vivo could be responsible for differential responses of the two muscle types in the endotoxemic animal. Mechanical stress and free radicals, two factors strongly associated with increased muscle activity, do not seem to play a significant role in the cytokine response of muscle cells to LPS stimulation. Other investigators have also reported a dissociation between LPS-induced and reactive oxygen species–induced TNF-{alpha} up-regulation in certain cell types (35–37). However, an effect by other components of increased muscle activation in vivo, such as increased neural stimulation or blood flow, remains possible. It has recently been reported that increasing the workload of the diaphragm under nonseptic conditions leads to up-regulation of multiple cytokines within the muscle (38). In addition, it is possible that the more vigorous cytokine response of the diaphragm to LPS in vivo is linked to differences in nonmuscle cell types present within the muscle. There is considerable evidence for cross-talk between myogenic cells and other cell types, and muscle fibers are able to express major histocompatibility complex molecules and act as antigen-presenting cells in certain inflammatory pathologies (22). Finally, the higher constitutive expression of cytokines found within the diaphragm in vivo could, in and of itself, help to explain the more vigorous expression of proinflammatory genes during sepsis through enhancement of a positive feedback mechanism. This is suggested by the fact that stimulation of diaphragm muscle cells with a combination of LPS and proinflammatory cytokines led to a greater induction of these same cytokines (TNF-{alpha} and IL-1-{alpha}) as well as other proinflammatory genes (MIP-2 and iNOS) when compared with stimulation by LPS alone.

Regulation of Proinflammatory Gene Expression by NF-{kappa}B and IKKbeta
The transcription factor NF-{kappa}B activates a broad spectrum of proinflammatory genes within multiple cell types (15). Therefore, we sought to determine whether NF-{kappa}B exerts a similar function in diaphragm muscle cells. In its inactive state, NF-{kappa}B is held in the cytoplasm, where it is complexed to the inhibitory protein IkB. NF-{kappa}B activation depends upon sequential modification of IkB by two protein complexes, the IKK complex and the E3 ubiquitin ligase complex (15). In the present study, these protein complexes were separately inhibited in differentiated muscle cells, and in both cases this greatly reduced the levels of cytokine/chemokine expression triggered by LPS stimulation.

Due to the very low transfectability of primary cells (39), a pharmacologic approach (PDTC) was used to inhibit NF-{kappa}B activation in cultured primary diaphragmatic myotubes. We observed a dramatic suppression of LPS effects on cytokine/chemokine (IL-1{alpha}, IL-6, TNF-{alpha}, and MIP-2) expression in diaphragmatic myotubes treated with PDTC, a compound that interferes with the E3 ubiquitin ligase required for NF-{kappa}B activation (31). PDTC also has antioxidant properties (40) that could theoretically contribute to its inhibitory effects on proinflammatory gene expression after LPS stimulation. However, this is unlikely in our study because two potent antioxidants (NAC and catalase), with documented abilities to inhibit reactive oxygen species–mediated cytokine up-regulation in muscle cells at the doses given in this study (41), had no significant effects on LPS-induced cytokine gene expression.

To further elucidate the role of the NF-{kappa}B pathway in LPS-induced cytokine up-regulation by skeletal muscle cells, we used a dominant-negative molecular approach to inhibit the IKK{alpha} or IKKbeta subunits of the IKK complex in a murine myogenic cell line (C2C12 cells). Phosphorylation of IkB by IKK{alpha} has been implicated in various homeostatic functions, whereas IKKbeta is known to play a central role in mediating NF-{kappa}B signaling triggered by proinflammatory stimuli such as TNF-{alpha} (15, 16). In addition, muscle-specific overexpression of IKKbeta in transgenic mice was recently found to cause severe muscle wasting in limb muscles (20). In the present study, transfection with an IKKbeta dominant-negative construct decreased the expression of proinflammatory genes in differentiated C2C12 myotubes exposed to LPS, whereas an IKK{alpha} dominant-negative construct had no apparent effect. This is consistent with the fact that other cell types lacking IKK{alpha} generally show normal induction of NF-{kappa}B activity in response to proinflammatory stimuli (42), although IKK{alpha} can play a role in NF-{kappa}B activation under certain inflammatory conditions (43, 44). Our findings suggest that the classical, IKKbeta-mediated pathway for NF-{kappa}B activation plays the predominant role in the LPS-induced inflammatory response of muscle fibers.

Functional and Clinical Implications
To our knowledge, the present study is the first to evaluate the expression of such a large variety of inflammatory mediators in any skeletal muscle during endotoxemia. Previous investigations have indicated that many of the proinflammatory cytokines up-regulated within the diaphragm by LPS in our study can cause muscle wasting or contractile dysfunction. For example, IL-1 inhibits muscle protein synthesis (19), and IL-6 up-regulates the cathepsin and ubiquitin pathways of muscle proteolysis (18). TNF-{alpha} has the ability to inhibit force production (14), promote protein degradation (45), and destabilize myogenic transcription factors such as MyoD or myogenin (12, 46). In our study, we show that the TNF-{alpha} receptor (p55) was also up-regulated by LPS, suggesting an additional mechanism for increasing TNF-{alpha} effects on the diaphragm during endotoxemia. In addition, mRNA levels for MIP-2 (a CXC chemokine) and ICAM-1 (a cell adhesion molecule), two molecules that are important for inflammatory cell migration and activation (47), were also found to be greatly increased within the diaphragm after LPS administration.

In several different experimental models of sepsis, a greater vulnerability of the diaphragm to sepsis-induced diaphragmatic dysfunction has been reported (4, 6, 7, 24, 25). This is likely due to the fact that the diaphragm is a functionally and anatomically unique skeletal muscle. For example, it is the only skeletal muscle that performs cyclical contraction on a continuous 24-h basis. The diaphragm also acts as a thin, flexible, anatomic barrier separating the abdominal and thoracic cavities. Recent gene expression profiling studies have revealed a pattern that indicates a greater presence of antigen-presenting cells in the diaphragm than in limb muscles (48). Taken together with our own findings of greater constitutive and inducible levels of proinflammatory gene expression within the diaphragm, this suggests that the diaphragm may be preferentially "primed" to respond in a vigorous fashion to infectious or inflammatory insults. We speculate that this could be related to a protective barrier function performed by the muscle because the diaphragm is well situated anatomically to help prevent the spread of infectious processes between the peritoneal and pleural cavities. Such a function would be enhanced by the ability to produce cytokines, chemokines, and other proinflammatory mediators. However, for the reasons discussed previously, this latter property may also be responsible for a greater vulnerability of the muscle to sepsis-induced contractile dysfunction.


   Acknowledgments
 
The authors thank Dr. Phil Barker for kindly providing the plasmid constructs containing the IKK{alpha} and IKKbeta dominant-negative mutants.


   FOOTNOTES
 
Supported by the Canadian Institutes of Health Research, Canadian Lung Association, Fonds de la Recherche en Santé du Quebec, Prix Marianne Josso, and Fondation pour la Recherche Medicale.

This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org

Originally Published in Press as DOI: 10.1164/rccm.200509-1511OC on June 15, 2006

Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Received in original form September 27, 2005; accepted in final form June 14, 2006


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Brun-Buisson C, Meshaka P, Pinton P, Vallet B. EPISEPSIS: a reappraisal of the epidemiology and outcome of severe sepsis in French intensive care units. Intensive Care Med 2004;30:580–588.[CrossRef][Medline]
  2. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003;348:1546–1554.[Abstract/Free Full Text]
  3. Hussain SN, Simkus G, Roussos C. Respiratory muscle fatigue: a cause of ventilatory failure in septic shock. J Appl Physiol 1985;58:2033–2040.[Abstract/Free Full Text]
  4. Boczkowski J, Lanone S, Ungureanu-Longrois D, Danialou G, Fournier T, Aubier M. Induction of diaphragmatic nitric oxide synthase after endotoxin administration in rats: role on diaphragmatic contractile dysfunction. J Clin Invest 1996;98:1550–1559.[Medline]
  5. Ebihara S, Hussain SN, Danialou G, Cho WK, Gottfried SB, Petrof BJ. Mechanical ventilation protects against diaphragm injury in sepsis: interaction of oxidative and mechanical stresses. Am J Respir Crit Care Med 2002;165:221–228.[Abstract/Free Full Text]
  6. Lin MC, Ebihara S, El Dwairi Q, Hussain SN, Yang L, Gottfried SB, Comtois A, Petrof BJ. Diaphragm sarcolemmal injury is induced by sepsis and alleviated by nitric oxide synthase inhibition. Am J Respir Crit Care Med 1998;158:1656–1663.[Abstract/Free Full Text]
  7. Divangahi M, Matecki S, Dudley RW, Tuck SA, Bao W, Radzioch D, Comtois AS, Petrof BJ. Preferential diaphragmatic weakness during sustained Pseudomonas aeruginosa lung infection. Am J Respir Crit Care Med 2004;169:679–686.[Abstract/Free Full Text]
  8. Nethery D, DiMarco A, Stofan D, Supinski G. Sepsis increases contraction-related generation of reactive oxygen species in the diaphragm. J Appl Physiol 1999;87:1279–1286.[Abstract/Free Full Text]
  9. Van Surell C, Boczkowski J, Pasquier C, Du Y, Franzini E, Aubier M. Effects of N-acetylcysteine on diaphragmatic function and malondialdehyde content in Escherichia coli endotoxemic rats. Am Rev Respir Dis 1992;146:730–734.[Medline]
  10. Wilcox PG, Wakai Y, Walley KR, Cooper DJ, Road J. Tumor necrosis factor alpha decreases in vivo diaphragm contractility in dogs. Am J Respir Crit Care Med 1994;150:1368–1373.[Abstract]
  11. Shindoh C, Hida W, Ohkawara Y, Yamauchi K, Ohno I, Takishima T, Shirato K. TNF-alpha mRNA expression in diaphragm muscle after endotoxin administration. Am J Respir Crit Care Med 1995;152:1690–1696.[Abstract]
  12. Guttridge DC, Mayo MW, Madrid LV, Wang CY, Baldwin AS Jr. NF-kappaB-induced loss of MyoD messenger RNA: possible role in muscle decay and cachexia. Science 2000;289:2363–2366.[Abstract/Free Full Text]
  13. Li YP, Schwartz RJ, Waddell ID, Holloway BR, Reid MB. Skeletal muscle myocytes undergo protein loss and reactive oxygen-mediated NF-kappaB activation in response to tumor necrosis factor alpha. FASEB J 1998;12:871–880.[Abstract/Free Full Text]
  14. Reid MB, Lannergren J, Westerblad H. Respiratory and limb muscle weakness induced by tumor necrosis factor-alpha: involvement of muscle myofilaments. Am J Respir Crit Care Med 2002;166:479–484.[Abstract/Free Full Text]
  15. Hayden MS, Ghosh S. Signaling to NF-kappaB. Genes Dev 2004;18:2195–2224.[Abstract/Free Full Text]
  16. Karin M, Yamamoto Y, Wang QM. The IKK NF-kappa B system: a treasure trove for drug development. Nat Rev Drug Discov 2004;3:17–26.[CrossRef][Medline]
  17. Wilcox P, Osborne S, Bressler B. Monocyte inflammatory mediators impair in vitro hamster diaphragm contractility. Am Rev Respir Dis 1992;146:462–466.[Medline]
  18. Tsujinaka T, Fujita J, Ebisui C, Yano M, Kominami E, Suzuki K, Tanaka K, Katsume A, Ohsugi Y, Shiozaki H, et al. Interleukin 6 receptor antibody inhibits muscle atrophy and modulates proteolytic systems in interleukin 6 transgenic mice. J Clin Invest 1996;97:244–249.[Medline]
  19. Cooney RN, Maish GO III, Gilpin T, Shumate ML, Lang CH, Vary TC. Mechanism of IL-1 induced inhibition of protein synthesis in skeletal muscle. Shock 1999;11:235–241.[Medline]
  20. Cai D, Frantz JD, Tawa NE Jr, Melendez PA, Oh BC, Lidov HG, Hasselgren PO, Frontera WR, Lee J, Glass DJ, et al. IKKbeta/NF-kappaB activation causes severe muscle wasting in mice. Cell 2004;119:285–298.[CrossRef][Medline]
  21. Bartoccioni E, Michaelis D, Hohlfeld R. Constitutive and cytokine-induced production of interleukin-6 by human myoblasts. Immunol Lett 1994;42:135–138.[CrossRef][Medline]
  22. Nagaraju K, Raben N, Merritt G, Loeffler L, Kirk K, Plotz P. A variety of cytokines and immunologically relevant surface molecules are expressed by normal human skeletal muscle cells under proinflammatory stimuli. Clin Exp Immunol 1998;113:407–414.[CrossRef][Medline]
  23. De Rossi M, Bernasconi P, Baggi F, de Waal MR, Mantegazza R. Cytokines and chemokines are both expressed by human myoblasts: possible relevance for the immune pathogenesis of muscle inflammation. Int Immunol 2000;12:1329–1335.[Abstract/Free Full Text]
  24. Hussain SN, Giaid A, El Dawiri Q, Sakkal D, Hattori R, Guo Y. Expression of nitric oxide synthases and GTP cyclohydrolase I in the ventilatory and limb muscles during endotoxemia. Am J Respir Cell Mol Biol 1997;17:173–180.[Abstract/Free Full Text]
  25. Hussain SN, El Dwairi Q, Abdul-Hussain MN, Sakkal D. Expression of nitric oxide synthase isoforms in normal ventilatory and limb muscles. J Appl Physiol 1997;83:348–353.[Abstract/Free Full Text]
  26. Divangahi M, Demoule A, Danialou G, Bao W, Petrof BJ. Expression of pro-inflammatory mediators is highly upregulated in the mouse diaphragm in comparison to limb muscles after endotoxin (LPS) administration [abstract]. Am J Respir Crit Care Med 2006;169:A244.
  27. Comtois AS, Barreiro E, Huang PL, Marette A, Perrault M, Hussain SN. Lipopolysaccharide-induced diaphragmatic contractile dysfunction and sarcolemmal injury in mice lacking the neuronal nitric oxide synthase. Am J Respir Crit Care Med 2001;163:977–982.[Abstract/Free Full Text]
  28. Comtois AS, El Dwairi Q, Laubach VE, Hussain SN. Lipopolysaccharide-induced diaphragmatic contractile dysfunction in mice lacking the inducible nitric oxide synthase. Am J Respir Crit Care Med 1999;159:1975–1980.[Abstract/Free Full Text]
  29. Demoule A, Divangahi M, Danialou G, Gvozdic D, Larkin G, Bao W, Petrof BJ. Expression and regulation of CC class chemokines in the dystrophic (mdx) diaphragm. Am J Respir Cell Mol Biol 2005;33:178–185.[Abstract/Free Full Text]
  30. Frost RA, Nystrom GJ, Lang CH. Lipopolysaccharide regulates proinflammatory cytokine expression in mouse myoblasts and skeletal muscle. Am J Physiol Regul Integr Comp Physiol 2002;283:R698–R709.[Abstract/Free Full Text]
  31. Hayakawa M, Miyashita H, Sakamoto I, Kitagawa M, Tanaka H, Yasuda H, Karin M, Kikugawa K. Evidence that reactive oxygen species do not mediate NF-kappaB activation. EMBO J 2003;22:3356–3366.[CrossRef][Medline]
  32. Bhakar AL, Tannis LL, Zeindler C, Russo MP, Jobin C, Park DS, MacPherson S, Barker PA. Constitutive nuclear factor-kappa B activity is required for central neuron survival. J Neurosci 2002;22:8466–8475.[Abstract/Free Full Text]
  33. Cowan DB, Poutias DN, Del Nido PJ, McGowan FX Jr. CD14-independent activation of cardiomyocyte signal transduction by bacterial endotoxin. Am J Physiol Heart Circ Physiol 2000;279:H619–H629.[Abstract/Free Full Text]
  34. Feldman JL, Stockdale FE. Skeletal muscle satellite cell diversity: satellite cells form fibers of different types in cell culture. Dev Biol 1991;143:320–334.[CrossRef][Medline]
  35. Strassheim D, Asehnoune K, Park JS, Kim JY, He Q, Richter D, Mitra S, Arcaroli J, Kuhn K, Abraham E. Modulation of bone marrow-derived neutrophil signaling by H2O2: disparate effects on kinases, NF-kappaB, and cytokine expression. Am J Physiol Cell Physiol 2004;286:C683–C692.[Abstract/Free Full Text]
  36. White JE, Tsan MF. Differential induction of TNF-alpha and MnSOD by endotoxin: role of reactive oxygen species and NADPH oxidase. Am J Respir Cell Mol Biol 2001;24:164–169.[Abstract/Free Full Text]
  37. Jennings GR, Castresana MR, Newman WH. Regulation of tumor necrosis factor-alpha production in the isolated rat heart stimulated by bacterial lipopolysaccharide or reactive oxygen. Am Surg 2004;70:797–800.[Medline]
  38. Vassilakopoulos T, Divangahi M, Rallis G, Kishta O, Petrof B, Comtois A, Hussain SN. Differential cytokine gene expression in the diaphragm in response to strenuous resistive breathing. Am J Respir Crit Care Med 2004;170:154–161.[Abstract/Free Full Text]
  39. Pampinella F, Lechardeur D, Zanetti E, MacLachlan I, Benharouga M, Lukacs GL, Vitiello L. Analysis of differential lipofection efficiency in primary and established myoblasts. Mol Ther 2002;5:161–169.[CrossRef][Medline]
  40. Schreck R, Meier B, Mannel DN, Droge W, Baeuerle PA. Dithiocarbamates as potent inhibitors of nuclear factor kappa B activation in intact cells. J Exp Med 1992;175:1181–1194.[Abstract/Free Full Text]
  41. Kosmidou I, Vassilakopoulos T, Xagorari A, Zakynthinos S, Papapetropoulos A, Roussos C. Production of interleukin-6 by skeletal myotubes: role of reactive oxygen species. Am J Respir Cell Mol Biol 2002;26:587–593.[Abstract/Free Full Text]
  42. Hu Y, Baud V, Delhase M, Zhang P, Deerinck TEM, Johnson R, Karin M. Abnormal morphogenesis but intact IKK activation in mice lacking the IKKalpha submit of IkappaB kinase. Science 1999;284:316–320.[Abstract/Free Full Text]
  43. Yamamoto Y, Verma UN, Prajapati S, Kwak YT, Gaynor RB. Histone H3 phosphorylation by IKK-alpha is critical for cytokine-induced gene expression. Nature 2003;423:655–659.[CrossRef][Medline]
  44. Li X, Massa PE, Hanidu A, Peet GW, Aro P, Savitt A, Mische S, Li J, Marcu KB. IKKalpha, IKKbeta, and NEMO/IKKgamma are each required for the NF-kappa B-mediated inflammatory response program. J Biol Chem 2002;277:45129–45140.[Abstract/Free Full Text]
  45. Li YP, Chen Y, John J, Moylan J, Jin B, Mann DL, Reid MB. TNF-alpha acts via p38 MAPK to stimulate expression of the ubiquitin ligase atrogin1/MAFbx in skeletal muscle. FASEB J 2005;19:362–370.[Abstract/Free Full Text]
  46. Langen RC, Schols AM, Kelders MC, Wouters EF, Janssen-Heininger YM. Inflammatory cytokines inhibit myogenic differentiation through activation of nuclear factor-kappaB. FASEB J 2001;15:1169–1180.[Abstract/Free Full Text]
  47. Albelda SM, Smith CW, Ward PA. Adhesion molecules and inflammatory injury. FASEB J 1994;8:504–512.[Abstract]
  48. Porter JD, Merriam AP, Leahy P, Gong B, Feuerman J, Cheng G, Khanna S. Temporal gene expression profiling of dystrophin-deficient (mdx) mouse diaphragm identifies conserved and muscle group-specific mechanisms in the pathogenesis of muscular dystrophy. Hum Mol Genet 2004;13:257–269.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Eur Respir JHome page
R. C. J. Langen and A. M. W. J. Schols
Inflammation: friend or foe of muscle remodelling in COPD?
Eur. Respir. J., October 1, 2007; 30(4): 605 - 607.
[Full Text] [PDF]

Home page
 J. Immunol.Home page
M. Divangahi, T. Yang, K. Kugathasan, S. McCormick, S. Takenaka, G. Gaschler, A. Ashkar, M. Stampfli, J. Gauldie, J. Bramson, et al.
Critical Negative Regulation of Type 1 T Cell Immunity and Immunopathology by Signaling Adaptor DAP12 during Intracellular Infection
J. Immunol., September 15, 2007; 179(6): 4015 - 4026.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Respir. Crit. Care Med.Home page
E. B. Milbrandt, A. Ishizaka, and D. C. Angus
Update in Critical Care 2006
Am. J. Respir. Crit. Care Med., April 1, 2007; 175(7): 638 - 648.
[Full Text] [PDF]

Home page
 Am. J. Respir. Cell Mol. Bio.Home page
M. Divangahi, A. Demoule, G. Danialou, L. Yahiaoui, W. Bao, Z. Xing, and B. J. Petrof
Impact of IL-10 on Diaphragmatic Cytokine Expression and Contractility during Pseudomonas Infection
Am. J. Respir. Cell Mol. Biol., April 1, 2007; 36(4): 504 - 512.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Online Supplement
Right arrow All Versions of this Article:
200509-1511OCv1
174/6/646    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Demoule, A.
Right arrow Articles by Petrof, B. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Demoule, A.
Right arrow Articles by Petrof, B. J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Proc. Am. Thorac. Soc. Am. J. Respir. Cell Mol. Biol.
Copyright © 2006 American Thoracic Society