NF-κB activation and polyubiquitin conjugation are required for pulmonary inflammation-induced diaphragm atrophy

Astrid Haegens, Annemie M. Schols, Stefan H. Gorissen, Anon L. van Essen, Frank Snepvangers, Douglas A. Gray, Steven E. Shoelson, Ramon C. Langen

Abstract

Loss of diaphragm muscle strength in inflammatory lung disease contributes to mortality and is associated with diaphragm fiber atrophy. Ubiquitin (Ub) 26S-proteasome system (UPS)-dependent protein breakdown, which mediates muscle atrophy in a number of physiological and pathological conditions, is elevated in diaphragm muscle of patients with chronic obstructive pulmonary disease. Nuclear factor kappa B (NF-κB), an essential regulator of many inflammatory processes, has been implicated in the regulation of poly-Ub conjugation of muscle proteins targeted for proteolysis by the UPS. Here, we test if NF-κB activation in diaphragm muscle and subsequent protein degradation by the UPS are required for pulmonary inflammation-induced diaphragm atrophy. Acute pulmonary inflammation was induced in mice by intratracheal lipopolysaccharide instillation. Fiber cross-sectional area, ex vivo tyrosine release, protein poly-Ub conjugation, and inflammatory signaling were determined in diaphragm muscle. The contribution of NF-κB or the UPS to diaphragm atrophy was assessed in mice with intact or genetically repressed NF-κB signaling or attenuated poly-Ub conjugation, respectively. Acute pulmonary inflammation resulted in diaphragm atrophy measured by reduced muscle fiber cross-sectional area. This was accompanied by diaphragm NF-κB activation, and proteolysis, measured by tyrosine release from the diaphragm. Poly-Ub conjugation was increased in diaphragm, as was the expression of muscle-specific E3 Ub ligases. Genetic suppression of poly-Ub conjugation prevented inflammation-induced diaphragm muscle atrophy, as did muscle-specific inhibition of NF-κB signaling. In conclusion, the present study is the first to demonstrate that diaphragm muscle atrophy, resulting from acute pulmonary inflammation, requires NF-κB activation and UPS-mediated protein degradation.

  • inspiratory muscle
  • inflammation
  • protein degradation

the diaphragm is the most important inspiratory muscle. Diaphragm weakness is a dominant cause of dyspnea in acute and chronic respiratory disease, independent of the severity of airflow obstruction, and it is associated with increased mortality (13). Furthermore, in patients with severe chronic obstructive pulmonary disease (COPD), improvement in inspiratory muscle strength after multimodal intervention was associated with improved survival compared with subjects with constant or decreased baseline inspiratory muscle strength (27). Loss of force-generating capacity occurs, despite continuous activity of the diaphragm muscle, and this is strongly dependent on decreased contractile protein content (23).

In a recent experimental study of endotoxin-induced diaphragm muscle dysfunction, protein loss has been attributed to upregulation of the ubiquitin (Ub) 26S-proteasome system (UPS) (29). The UPS degrades proteins via lysine (K) 48-linkage of multiple Ub molecules to a protein substrate. This poly-Ub conjugation marks proteins for degradation by the 26S proteasome (34). Studied diaphragm biopsies from patients with mild to moderate COPD revealed increased protein levels of subunits, which are part of the 20S proteolytic core of the 26S proteasome (22) and increased protein Ub conjugation (23). Ub conjugation of proteins is catalyzed by E3 Ub ligases, and increased expression of muscle-specific E3 Ub ligases atrogin-1 and muscle RING-finger protein 1 (MuRF1) is required for skeletal muscle atrophy induced by physiological and pathological stimuli (2, 11). Expression of both E3 Ub ligases was demonstrated to be upregulated in atrophied diaphragm muscle of mechanically ventilated subjects (20), and increased atrogin-1 expression has been found in diaphragm muscle biopsies from COPD patients (22).

In addition to COPD, acute respiratory distress syndrome (ARDS) and sepsis are also accompanied by impaired respiration related to diaphragm dysfunction (18). Interestingly, in similarity to COPD, and in particular COPD exacerbation (15), ARDS and sepsis are also accompanied by local and systemic inflammation (8, 21), suggesting that inflammatory signals may contribute to diaphragm weakness in these conditions. NF-κB is an essential regulator of cellular responses to a multitude of inflammatory stimuli (10). A number of inflammatory genes, including macrophage inflammatory protein-2 have been demonstrated to be upregulated in diaphragm muscle during endotoxemia in a NF-κB-dependent manner (5). Besides these inflammatory cytokines, increased murf1 gene expression has also been demonstrated in response to NF-κB activation in skeletal muscle (3). Overall, although these observations clearly suggest participation of UPS-mediated proteolyses and NF-κB signaling in inflammation-associated diaphragm muscle weakness, any causal involvement of these pathways in diaphragm muscle atrophy remains to be addressed.

In the present study, we hypothesized that NF-κB activation in diaphragm muscle and increased protein degradation by the UPS are required for inflammation-induced diaphragm atrophy. This hypothesis was addressed in a well-characterized mouse model of intratracheal (IT) lipopolysaccharide (LPS)-instillation (16, 32, 33), which mimics characteristic lung and systemic inflammatory features of acute COPD exacerbations. Muscle fiber cross-sectional area (CSA), ex vivo tyrosine release, and protein poly-Ub conjugation were determined to assess muscle atrophy, proteolysis, and UPS activity in diaphragm muscle, as well as inflammatory and reporter gene expression as indexes of transcriptional activation of NF-κB were assessed. Subsequently, the contribution of poly-Ub conjugation to diaphragm atrophy was assessed by comparison of this response in transgenic mice expressing human Ub [Ub-wild type (WT), identical in amino acid sequence to mouse Ub], and mice expressing Ub containing a lysine 48 to arginine mutation (Ub-K48R) (30). The dependence of diaphragm muscle atrophy on muscle NF-κB activation was addressed by comparison of WT and transgenic “muscle IκBα super repressor” (MISR) mice, in which NF-κB signaling is repressed in muscle (3).

MATERIALS AND METHODS

Animals.

The Institutional Animal Care Committee of Maastricht University approved the animal studies here described. Twelve-week-old male Ub-K48R (24, 30) mice and control Ub-WT (24), both on a FVB background, were taken into experiment as well as MISR (3) mice and matched background controls (WT C57Bl6) or NF-κB luciferase mice (4). All mice were cared for as described previously (16). Lung inflammation was induced with 1 × 105 endotoxin units of LPS, and mice were killed 24 and 48 h after LPS instillation, as previously described (16). After euthanasia, diaphragm was collected. One hemidiaphragm was snap frozen; the second hemidiaphragm was rolled into a cylinder while embedded in Tissue-Tek OCT (Sakura, Finetek, Zoeterwoude, the Netherlands) for histological analyses.

Diaphragm fiber CSA and fiber-type staining.

Sections were air dried, treated with 0.5% Triton X-100 in PBS, incubated with primary antibodies rabbit anti-laminin (L9393, Sigma, St. Louis, MO) and goat anti-mouse myosin heavy chain I (A4.840, Developmental Studies Hybridoma Bank, Iowa City, IA), and visualized with secondary antibodies Alexa Fluor 680 goat anti-rabbit IgG (A21076, Invitrogen, Paisley, UK) and Alexa Fluor 555 goat anti-mouse IgM (A21426, Invitrogen). Unstained fibers were considered type II fibers. After staining, all images were digitally captured using fluorescence microscopy (Nikon Instruments Europe). Image processing and quantitative analyses were done using the Lucia 4.81 software package. The mean fiber CSA of each diaphragm was calculated 48 h post-IT instillation, by analysis of 300 fibers per diaphragm.

Tyrosine release analysis.

Twenty-four hours following IT instillation of NaCl or LPS, animals were euthanized, and one hemidiaphragm was collected, weighed, and preincubated for 30 min in Krebs-Henseleit buffer [120 mM NaCl, 25 mM NaHCO3, 4.8 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4, 5 mM d-(+)-glucose-monohydrate, 5 mM HEPES, 0.1% BSA, and 0.5 mM cycloheximidine, pH 7.4] of 37°C saturated with 95% O2/5% CO2. After preincubation, the hemidiaphragms were incubated for 60 min in fresh buffer. This buffer was collected and stored at −80°C. Tyrosine levels were determined by HPLC.

NF-κB luciferase reporter activity.

Twenty-four hours post-IT instillation, a hemidiaphragm was collected, mechanically homogenized in 1× Reporter Lysis Buffer (Promega, Madison, WI), and incubated on ice for 10 min. Homogenates were centrifuged (13,000 g, 1 min), and supernatants were analyzed for luciferase activity by luminometry after one freeze-thaw cycle (Promega) and corrected for total protein content (Bio-Rad, Hercules, CA).

RNA isolation, cDNA synthesis, and quantitative PCR analysis.

Forty-eight hours following IT instillation, one hemidiaphragm was collected, and RNA was extracted using the acid guanidium thiocyanate-phenol-chloroform extraction method (Ambion, Ijssel, the Netherlands). cDNA was made with the Transcription First Strand cDNA Synthesis Kit (Roche, Almere, the Netherlands). Gene expression of murf1 (forward: 5′-CTTCCTCTCAAGTGCCAAGCA-3′, reverse: 3′-GTGTTCTAAGTCCAGAGTAAAGTAGTCCAT-5′), atrogin-1 (forward: 5′-CAGCAGCTGAATAGCATCCAGAT-3′, reverse: 3′-TCTGCATGATGTTCAGTTGTAAGC-5′), and cxcl2 (forward: 5′-CCCTGGTTCAGAAAATCATCCAAA-3′, reverse: 5′-TTTGGTTCTTCCGTTGAGGGAC-3′) were determined on the Bio-Rad iCycler apparatus (Bio-Rad, Hercules, CA) using a two-step PCR program with the following cycling conditions: 15′ 95°C, 40 cycles of (15″ 95°C, 45″ 60°C), 30″ 95°C, 30″ 60°C followed by a melting curve (performed by heating from 60°C to 95°C in increments of 0.5°C). Ct values were obtained for the standard curve and each sample. The relative DNA starting quantities of the samples were derived from the standard curve based on the Ct values using the MyiQ analysis software (Bio-Rad) and normalized to the geometric average of at least four out of five reference genes (β-actin, gapdh, tubulin, calnexin, and cyclophilin A) by the geNorm software (31).

Determination of poly-Ub conjugation in diaphragm homogenates.

Approximately 12 mg of diaphragm muscle, collected 48 h post-IT instillation, was homogenized mechanically in 400-μl lysis buffer containing 50 mM Tris (pH 7.4), 150 mM NaCl, 10% glycerol, 0.5% Nonidet P-40, 1 mM EDTA, 1 mM Na-orthovanadate, 5 mM NaF, 1 mM β-glycerophosphate, 1 mM Na-pyrophosphate, 1 mM DTT, 10 μg/ml leupeptin, 1% aprotinin, and 1 mM PMSF, incubated on ice for 30 min, and centrifuged for 30 min at 14,000 rpm at 4°C. For Western blot analysis, 4× Laemmli buffer containing 0.25 M Tris·HCl, pH 6.8; 8% (wt/vol) SDS, 40% (vol/vol) glycerol, 0.4 M DTT, and 0.04% (wt/vol) bromophenol blue was added to the supernatant, followed by boiling samples for 5 min at 100°C and storage at −20°C. Total protein was assessed using the Bio-Rad DC protein assay kit (Bio-Rad). Equal amounts of protein were loaded on a Criterion XT bis-tris gel (4–12%, 18-well). After electrophoresis, transfer, and blocking, blots were incubated with a monoclonal antibody against poly-Ub conjugates (BML-PW8805, Enzo Life Sciences, Lausanne, Switzerland) and a peroxidase-labeled secondary antibody (PI-2000, Vector Laboratories, Burlingame, CA) and visualized with SuperSignal West Pico Chemiluminescent Substrate (Pierce Biotechnology, Rockford, IL) on the ChemiDoc XRS (Bio-Rad). Images were analyzed using Quantity One software (Bio-Rad). All bands were included in the quantification, and the arbitrary units (AU) of each band were added to obtain a value for total poly-Ub conjugates per lane.

Statistical analysis.

Data were reported as means ± SD and analyzed with SPSS 15.0 using a two-sample equal variance Student's t-test and a Mann-Whitney U-test. A P value of <0.05 between groups was considered significant.

RESULTS

Acute pulmonary inflammation results in diaphragm fiber atrophy.

IT-LPS instillation-induced acute pulmonary inflammation, previously extensively characterized by our research group (9, 16, 33) and others (1, 28), was confirmed. LPS instillation resulted in a rapid and significant induction of interleukin-6 gene expression (2.97 ± 0.97 AU, P = 0.00001) in the lung compared with sham treatment (0.006 ± 0.003 AU). This induction (0.47 ± 0.47, P = 0.02) sustained up to 24 h after LPS instillation. Interleukin-6 gene expression was accompanied by an induction of NF-κB regulatory protein IκBε in the lung. Gene expression of IκBϵ increased significantly (2.34 ± 1.10 AU, P = 0.0007) compared with sham instillation (0.32 ± 0.11 AU) 4 h after treatment. IκB gene expression was sustained up to 24 h after LPS instillation (1.06 ± 0.52 AU, P = 0.003). This lung inflammation was accompanied by a decreased (−13%) diaphragm fiber CSA in mice (Fig. 1B, left bars), determined by laminin staining of diaphragm cross sections (Fig. 1A), indicative of diaphragm muscle atrophy. Further immunohistochemical analysis revealed that the decrease in fiber CSA was most pronounced in myosin heavy chain type II muscle fibers of LPS-treated mice (Fig. 1B, right bars). Muscle fiber atrophy was accompanied by increased (+15%) tyrosine release of ex vivo incubated diaphragm (Fig. 1C), reflecting increased protein degradation. Poly-Ub conjugation levels in diaphragm protein homogenates were assessed. A representative blot is shown in Fig. 1D, left. LPS treatment resulted in increased (+20%) levels of poly-Ub conjugates compared with control (Fig. 1D, right), suggesting participation of the UPS in increased proteolysis. To substantiate the potential participation of the UPS, transcript levels of atrogin-1 and murf1, E3 Ub ligases involved in muscle protein poly-Ub conjugation (2, 11), were assessed. Gene expression levels of atrogin-1 (Fig. 1E, ∼1.5-fold) and murf1 (Fig. 1F, ∼3-fold) were increased compared with control. These data demonstrate that acute pulmonary inflammation triggers diaphragm atrophy, which requires increased muscle protein poly-Ub conjugation and degradation.

Fig. 1.

Acute pulmonary inflammation results in diaphragm fiber atrophy. Lung inflammation was induced by intratracheal LPS instillation. Sham treatment consisted of intratracheal NaCl instillation. A: laminin staining was performed on diaphragm cross sections (scale bar 100 μm) of WT C57Bl6 mice. B: the total cross-sectional area (CSA) of diaphragm fibers was determined and specified for myosin heavy chain (MyHC) I and II positive fibers. C: ex vivo tyrosine release from one hemidiaphragm muscle of ubiquitin (Ub)-wild-type (WT) mice was determined by HPLC and expressed as amount of tyrosine released per gram tissue per minute. D: poly-Ub conjugates were determined in diaphragm homogenates of Ub-WT mice by Western blot analysis using an anti-poly-Ub antibody. A representative blot is shown on the left side, with the molecular weight marker (M) in the left lane, a diaphragm homogenate of a sham-treated mouse in the middle lane, and a diaphragm homogenate of an LPS-treated mouse in the right lane. The right panel shows quantification of the poly-Ub conjugates. Gene expression levels of atrogin-1 (E) and murf1 (F) expression in diaphragm muscle of WT C57Bl6 mice were determined by quantitative PCR analysis and normalized using geNorm. Values are means ± SD. *P ≤ 0.05 from sham. AU, arbitrary units.

Inflammation-induced diaphragm fiber atrophy requires K48 poly-Ub conjugation.

To further investigate the contribution of the UPS in diaphragm atrophy induced by IT-LPS instillation, diaphragm muscle of transgenic mice expressing Ub with a mutation at lysine 48 (Ub-K48R) to inhibit poly-Ub conjugation were compared with diaphragms of transgenic mice (over)expressing Ub-WT. Despite similar basal levels of poly-Ub conjugates in diaphragm muscle of sham Ub-WT and Ub-K48R mice, LPS-treated Ub-K48R mice did not demonstrate increased protein poly-Ub conjugation (Fig. 2A, −18%) compared with LPS-treated Ub-WT. This indicates that stress-induced, but not basal, poly-Ub conjugation was suppressed in Ub-K48R mice. Whereas tyrosine release from diaphragm muscle of IT-LPS-treated Ub-WT mice was significantly increased (+15%) compared with sham treatment, this increase in proteolysis after LPS treatment was not present in diaphragm muscle of Ub-K48R mice (Fig. 2B). Figure 2C demonstrates that fiber CSA of diaphragm muscle was decreased in LPS-treated Ub-WT animals (−18%), while the Ub-K48R mice were protected against loss of fiber CSA in diaphragm muscle after LPS treatment. Further analysis of diaphragm muscle atrophy in these mice revealed a redistribution of muscle fiber CSA toward smaller fibers in LPS-treated Ub-WT mice compared with sham-treated Ub-WT mice (Fig. 2D). This shift was absent in LPS-treated Ub-K48R mice compared with sham-treated Ub-K48R mice (Fig. 2E).

Fig. 2.

Inflammation-induced diaphragm fiber atrophy requires K48 poly-Ub conjugation. Acute lung inflammation was induced in mice expressing Ub containing a lysine 48 to arginine mutation (Ub-K48R) and Ub-WT mice by intratracheal LPS instillation. A: poly-Ub conjugates were determined in diaphragm homogenates by Western blot analysis. A representative blot is shown in the left panel: within the left lane the molecular weight marker (M), in the middle Ub-K48R diaphragm homogenates, and on the right Ub-WT homogenates (S = sham, L = LPS) are shown. The right panel shows quantification of the Western blot. B: ex vivo tyrosine release from one hemidiaphragm per Ub-WT and Ub-K48R mice after intratracheal-NaCl or intratracheal-LPS instillation was determined and expressed as amount of tyrosine released per gram tissue per minute. C: the CSA of diaphragm fibers of LPS- and sham-exposed mice was determined on laminin-stained sections. Values are means ± SD. *P ≤ 0.05 from sham. #P ≤ 0.05 from Ub-WT. Diaphragm fiber size distribution was determined in sham- and LPS-treated Ub-WT (D) and Ub-K48R (E) mice.

Altered atrogene expression or food intake do not account for prevention of inflammation-induced diaphragm atrophy in Ub-K48R mice.

No differences were found in atrogin-1 (Fig. 3A) or murf1 (Fig. 3B) expression between Ub-WT and Ub-K48R mice in their response to LPS treatment, consistent with the notion that the protective effect of interference with poly-Ub conjugation on muscle atrophy occurs downstream of the transcriptional regulation of these E3 Ub ligases. Body weight loss was observed in response to pulmonary inflammation, and this was slightly attenuated in LPS-treated Ub-K48R mice compared with Ub-WT mice (Fig. 3C). This effect could not be attributed to a differential response on food intake between Ub-WT and Ub-K48R mice (Fig. 3D).

Fig. 3.

Altered atrogene expression or food intake do not account for prevention of inflammation-induced diaphragm atrophy in Ub-K48R mice. Acute lung inflammation was induced in Ub-K48R and Ub-WT mice by intratracheal LPS instillation. Gene expression levels of atrogin-1 (A) and murf1 (B) in diaphragm muscle were measured by quantitative PCR analysis and normalized using geNorm. Body weight (C) and food intake (D) were determined 24 and 48 h after treatment. Values are means ± SD. *P ≤ 0.05 from sham. #P ≤ 0.05 from Ub-WT.

Inflammation-induced diaphragm atrophy is dependent on NF-κB activation.

To assess if pulmonary (16) and systemic (33) inflammation in response to IT-LPS instillation are accompanied by NF-κB activity in diaphragm muscle, luciferase activity was determined in diaphragm muscle of NF-κB reporter mice. Figure 4A demonstrates increased NF-κB-driven luciferase activity in diaphragm muscle of LPS-treated mice compared with sham-treated animals. Increased levels of endogenously expressed cxcl2 (∼3-fold), a gene described to be NF-κB dependent, confirmed NF-κB activation in diaphragm of WT mice following IT-LPS treatment (Fig. 4B). In diaphragm of MISR mice, however, inflammation-induced cxcl2 expression was strongly attenuated, reflecting suppressed NF-κB activation, resulting from the SR transgene. Body weight loss was slightly attenuated 24 h after LPS instillation in MISR mice compared with WT (Fig. 4C). The reduction in food intake (Fig. 4D) was equal in WT and MISR mice 24 and 48 h after treatment, indicating the differential reduction in body weight could not be attributed to differences in food intake in response to IT-LPS. Inflammation-induced atrogin-1 expression was not affected in MISR compared with WT mice (Fig. 4E). In contrast, the increase in murf1 expression induced by IT-LPS treatment was markedly attenuated in diaphragm of MISR mice (Fig. 4F). Last, the decrease in diaphragm fiber CSA induced by LPS treatment in WT mice (−13%), shown in Fig. 4G, was completely prevented in LPS-treated MISR mice. With these results, we establish the significant contribution of NF-κB activity in the diaphragm to lung inflammation-induced diaphragm muscle atrophy.

Fig. 4.

Inflammation-induced atrophy of the diaphragm is dependent on NF-κB activation. Lung inflammation was induced via intratracheal LPS instillation. A: NF-κB luciferase activity was determined in protein extracts of diaphragm muscle of NF-κB luciferase reporter mice B: gene expression levels of cxcl2 were determined in diaphragm muscle of WT and MISR mice by quantitative PCR analysis and normalized using geNorm. Body weight (C) and food consumption (D) were determined 24 and 48 h after induction of lung inflammation. Atrogin-1 (E) and murf1 expression (F) were determined in diaphragm muscle of WT and MISR mice by quantitative PCR analysis and normalized using geNorm. G: the CSA of diaphragm fibers of LPS- and sham-exposed mice was quantified in laminin-stained sections. Values are expressed as means ± SD. *P ≤ 0.05 from sham. #P ≤ 0.05 from WT. AU, arbitrary units; RLU, relative light units.

DISCUSSION

In the present study, we used a model of acute lung inflammation to investigate whether increased protein degradation by the UPS and NF-κB activation in diaphragm muscle are required for inflammation-induced diaphragm atrophy. We have demonstrated that LPS-induced lung inflammation decreased diaphragm fiber CSA in a NF-κB-dependent manner and requires poly-Ub conjugation. In our study, diaphragm muscle function was not measured, but it is very likely that the decrease in fiber CSA observed in response to pulmonary inflammation has functional consequences, as ample studies demonstrate clearly that reduced muscle fiber CSA correlates with a loss of muscle strength. A recent publication by Jaber et al. (19) describes a reduction in diaphragm CSA of 40% in patients who receive mechanical ventilation. This reduction in CSA was accompanied by an even greater reduction (−50%) in diaphragm function (19). In a mycobacterium ulcerans infection model, reductions of biceps muscle CSA of −29 and −17% resulted in a loss of maximal titanic force of 31 and 18%, respectively (17). Furthermore, a more modest 15% reduction in diaphragm mass resulted in a 13 and 12% reduction of diaphragm muscle strength in two separate studies (6, 26).

The reduced fiber CSA was accompanied by increased ex vivo tyrosine release, indicative of proteolysis-mediated diaphragm atrophy, which was likely mediated by the UPS as inhibition of poly-Ub conjugation prevented increased tyrosine release in Ub-K48R mice. These findings are in keeping with a recently published study by Supinski and colleagues (29), in which it was shown that sepsis, induced by systemic delivery of LPS, resulted in a loss of diaphragm weight and protein content and an increase in diaphragm tyrosine release. UPS-mediated protein degradation is preceded by poly-Ub conjugation of respective protein substrates (34). The increased protein poly-Ub conjugation in diaphragm muscle induced by inflammation as observed here has not been described before. The observations are, however, in line with a human study by Ottenheijm et al. (23), showing overall increased Ub conjugation in diaphragm muscle homogenates from mild to moderate stable COPD patients undergoing thoracotomy for lung cancer compared with cancer patients without COPD.

The present study also demonstrates for the first time a causal relation between diaphragm muscle atrophy and UPS-mediated proteolysis. Our data demonstrating that inflammation-induced tyrosine release from diaphragm is prevented in Ub-K48R transgenic mice are in line with the findings by Supinski et al. (29), who demonstrate that pharmacological inhibition of the UPS could also prevent increased tyrosine release by the diaphragm following systemic administration of endotoxin. It is important to stress that the protection against inflammation-induced diaphragm atrophy in the Ub-K48R mice unlikely resulted from alterations upstream of poly-Ub conjugation of muscle proteins. This can be concluded from the fact that the induction of murf1 and atrogin-1 gene expression following LPS treatment was similar between Ub-K48R and Ub-WT mice. Moreover, endogenous Ub levels are sufficiently abundant in Ub-K48R mice for the dynamic assembly and disassembly of Ub chains, to out-compete K48R-Ub chain-terminating conjugation, which may explain the absence of a phenotype in Ub-K48R mice under basal conditions. In contrast, in response to LPS-induced stress conditions, Ub conjugation increases (Fig. 1D), which may result in depletion of unconjugated endogenous Ub levels to an extent that chain-terminating K48R-Ub conjugation interferes with proteasomal degradation of muscle proteins. Indeed, increased UbB and UbC gene expression, suggestive of depletion of free Ub levels in response to stress conditions, has been demonstrated (7). Our results are also in line with previous findings describing a decreased disease burden or body weight loss in Ub-K48R mutant mice compared with WT control in response to sepsis or cisplatin administration (14). Interestingly, body weight loss in Ub-K48R mice induced by pulmonary inflammation was attenuated compared with that in Ub-WT mice, despite a similar reduction in food intake. This may be indicative of sparing of peripheral muscle due to attenuated proteolysis.

Next, we demonstrated increased NF-κB transcriptional activity, based on elevated NF-κB reporter activity and endogenously NF-κB driven cxcl2 expression in diaphragm muscle of LPS-treated mice. Since NF-κB activation has been implicated in cellular responses to a variety of inflammatory stimuli (10) and sepsis-induced upregulation of NF-κB-dependent proinflammatory genes in the diaphragm in particular (5), we employed a second transgenic approach to investigate the contribution of muscular NF-κB activation to the observed diaphragm muscle atrophy. MISR mice express a mutant form of IκBα specifically in skeletal muscle, which renders NF-κB unresponsive to activating stimuli (3). In these mice, LPS-induced cxcl2 expression in skeletal muscle was significantly attenuated compared with that in WT mice, indicative of decreased NF-κB signaling. Diaphragm fiber CSA was spared in LPS-treated MISR mice, indicating the requirement of NF-κB activation for inflammation-induced diaphragm muscle atrophy. This is in line with findings demonstrating that NF-κB activation is sufficient (3) and required for limb muscle atrophy. In fact, the small but significant attenuation of body weight loss observed in MISR mice after LPS instillation may be the result of peripheral muscle sparing, since NF-κB activation was blocked in all skeletal muscles.

Murf1 expression was significantly reduced in LPS-treated MISR mice compared with LPS-treated WT mice, indicating that murf1 expression in this model is, at least in part, NF-κB dependent and suggests that the increase in murf1 expression is responsible for the diaphragm atrophy observed in response to acute pulmonary inflammation. This finding is in line with increased murf1 expression found following constitutive activation of NF-κB in skeletal muscle (3). In contrast to murf1 expression, LPS-induced atrogin-1 expression was not affected in MISR mice, suggesting that the observed increase in atrogin-1 during active loss of muscle mass is independent of NF-κB transcriptional activity, which is in line with findings by others in vivo (3) and in vitro (25).

Diaphragm dysfunction contributes to respiratory failure in COPD and ARDS and is an unfavorable prognostic factor (13, 35). Our data offer new insight into the potential mechanisms related to diaphragm dysfunction associated with acute pulmonary inflammation as observed in ARDS and during acute exacerbations in COPD (12) and may provide the basis for a therapeutic approach to prevent or reverse diaphragm dysfunction.

In conclusion, our studies show that acute pulmonary inflammation by IT-LPS instillation results in NF-κB activation and increased diaphragm protein breakdown and subsequent diaphragm atrophy. Moreover, diaphragm proteolysis and atrophy require intact K48-Ub conjugation and NF-κB signaling.

GRANTS

This work was supported by the Dutch Top Institute Pharma (project no. T1-201) and the Organization for Scientific Research NWO (VENI 91656112).

DISCLOSURES

S. E. Shoelson is a consultant and is on the board of advisory of Catabasis, has given academic lectures to Merck about research, holds patents based on research conducted in his academic laboratory, which may be licensed to commercial entities by his employer, the Joslin Diabetes Center, and has retirement savings with Fidelity Health Care Mutual Fund.

AUTHOR CONTRIBUTIONS

Author contributions: A.H. and A.M.S. conception and design of research; A.H., S.H.G., A.L.v.E., and F.S. performed experiments; A.H., S.H.G., A.L.v.E., and F.S. analyzed data; A.H. and R.C.L. interpreted results of experiments; A.H. prepared figures; A.H. drafted manuscript; A.M.S. and R.C.L. edited and revised manuscript; D.A.G. and S.E.S. approved final version of manuscript.

ACKNOWLEDGMENTS

The authors thank Menno P. J. de Winther, Molecular Genetics, Maastricht University Medical Centre, The Netherlands, and Harald Carlsen, Institute for Nutrition Research, University of Oslo, Norway, for kindly providing the NF-κB-luciferase mice.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
View Abstract