High dexamethasone concentration prevents stimulatory effects of TNF-α and LPS on IL-6 secretion from the precursors of human muscle regeneration

Oja Prelovsek, Tomaz Mars, Marko Jevsek, Matej Podbregar, Zoran Grubic


A frequent finding in patients surviving critical illness myopathy is chronic muscle dysfunction. Its pathogenesis is mostly unknown; one explanation could be that muscle regeneration, which normally follows myopathy, is insufficient in these patients because of a high glucocorticoid level in their blood. Glucocorticoids can prevent stimulatory effects of proinflammatory factors on the interleukin (IL)-6 secretion, diminishing in this way the autocrine and paracrine IL-6 actions known to stimulate proliferation at the earliest, myoblast stage of muscle formation. To test this hypothesis, we compared the effects of major proinflammatory agents [tumor necrosis factor (TNF)-α and endotoxin lipopolysaccharide (LPS)] on the IL-6 secretion from the muscle precursors and then studied the influence of dexamethasone (Dex) on these effects. Mononuclear myoblasts, which still proliferate, were compared with myotubes in which this capacity is already lost. For correct interpretation of results, cultures were examined for putative apoptosis and necrosis. We found that constitutive secretion of IL-6 did not differ significantly between myoblasts and myotubes; however, the TNF-α- and LPS-stimulated IL-6 release was more pronounced (P < 0.001) in myoblasts. Dex, applied at the 0.1–100 nM concentration range, prevented constitutive and TNF-α- and LPS-stimulated IL-6 release at both developmental stages but only at high concentration (P < 0.01). Although there are still missing links to it, our results support the concept that high concentrations of glucocorticoids, met in critically ill patients, prevent TNF-α- and LPS-stimulated IL-6 secretion. This results in reduced IL-6-mediated myoblast proliferation, leading to the reduced final mass of the regenerated muscle.

  • sepsis-induced myopathy
  • critical illness myopathy
  • myoblasts
  • myotubes
  • cytokines
  • glucocorticoids

critically ill patients treated in intensive care units frequently develop skeletal muscle dysfunction (9, 13), which often persists after hospital discharge. Complete functional recovery of patients regaining the ability to both breathe spontaneously and to walk independently was reported in only two-thirds of cases (13, 24). Pathological processes responsible for poor recovery of muscle strength in these patients have not been precisely identified and are, to our present understanding, multifactorial (9). Catabolism of muscle proteins, because of a high concentration of glucocorticoids (GCs) in the blood of such patients because of increased endogenous release (33) and/or high therapeutic doses administered to such patients, could explain temporary but not permanent myopathic effects (23). A more likely explanation for the prolonged myopathic effects might be reduced capacity of muscle regeneration in these patients. Normally, skeletal muscle has a fairly good ability to regenerate. In response to myopathic changes, dormant mononuclear satellite cells become activated and turn into myoblasts. These then proliferate, differentiate, and eventually fuse into myotubes that mature into myofibers (22). In critically ill patients, myoblast proliferation may be impaired so that muscle mass, which under such conditions develops from the reduced number of precursors, is also substantially reduced.

It is now well documented that myoblast proliferation, an essential step in muscle regeneration, is controlled by cytokine signaling. Interleukin (IL)-6, which is the major cytokine, released from the skeletal muscle under various stress conditions (17) is a potent stimulator of myoblast proliferation (2–4, 11, 13) and hence muscle regeneration. It is possible that IL-6 release from the early muscle precursors is reduced in critically ill patients so that its auto- and paracrine proliferative effects on these precursors are insufficient for normal muscle regeneration.

To test this hypothesis, we studied how proinflammatory factors, present in the blood of critically ill patients, influence IL-6 release from the cultured precursors of the human muscle and also whether this effect persists if cultures are exposed to various concentrations of GCs, which are well-known inhibitors of interleukin release (31). The influence of two factors, considered to have the pivotal stimulatory effect on the IL-6 release from the muscle under such conditions, was tested. Endotoxin lipopolysaccharide (LPS), which is derived from the cell wall of gram-negative bacteria under septic conditions, is frequently elevated in critically ill patients. LPS binds to specific Toll-like receptors and induces expression and release of IL-6 from the muscle (19, 20). The second factor studied here was the endogenous proinflammatory cytokine tumor necrosis factor (TNF)-α, which is synthesized mainly by monocytes in response to various stressful stimuli (7) and stimulates IL-6 release from the human muscle (6). GCs, including dexamethasone (Dex) used in this study, are on the other hand, known to downregulate secretion of a great number of proinflammatory cytokines, including IL-6 (37). The effects of these factors were studied differentially at the myoblast and myotube levels, since the above hypothesis implies that the role of IL-6 secretion in muscle regeneration is more important and therefore more sensitive to the influences of proinflammatory factors in proliferation-competent myoblasts than in myotubes that can no longer divide, therefore not contributing substantially to the final mass of regenerated muscle.


Preparation of Human Muscle Cultures

Experiments were conducted on the primary cultures of human muscles at the myoblast and myotube developmental stages. All studies reported here were approved by the Ethical Commission of the Ministry of Health of the Republic of Slovenia (permit no. 63/01/99) and in accordance with the Declaration of Helsinki.

Myoblast Cultures

Myoblast cultures were prepared as described previously (1, 21, 26). Briefly, satellite cells were prepared from muscle tissue routinely discarded at orthopedic operations on patients without muscular disease. The muscle tissue was cleaned of adhering connective tissue, cut into small pieces, and trypsinized to release muscle satellite cells. Cells were grown at clonal density in 100-mm petri dishes in advanced minimum essential medium (Advanced MEM; GIBCO, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS; GIBCO) at saturated humidity in a mixture of 5% CO2 and air at 37°C. Confluent myoblast cultures were trypsinized before myoblast fusion. Cells were plated on glass cover slips coated with a 1:2 mixture of 1.5% gelatin (Sigma, St. Louis, MO) and human plasma in six-well dishes and grown for 3 days in growth medium as described above.

Myotube Cultures

To induce myogenic differentiation, subconfluent cultures were shifted from growth medium to differentiation medium (advanced MEM supplemented with 2% FBS). Myotube cultures were obtained after fusion of the myoblasts and were grown in differentiation medium in six-well dishes for 3 wk.

Treatment with TNF-α, LPS, and Dex

Myoblast and myotube cultures were exposed to TNF-α or LPS (Sigma-Aldrich) for 24 h at three different concentrations (1, 10, and 100 ng/ml) or to synthetic GC Dex at 0.1, 1, and 100 nM for 6 h.

To examine the effect of Dex on the IL-6 release induced by proinflammatory factors, we exposed myoblast and myotube cultures to either TNF-α (10 ng/ml) or LPS (100 ng/ml) in combination with 0.1, 1, and 100 nM Dex. The selected Dex concentrations spanned the range of GC concentrations in the human blood under normal and under stress conditions (32, 33). The TNF-α (10 ng/ml) and LPS (100 ng/ml) concentration ratio was also selected so that it approximately reflected the ratio determined for these two cytokines in the blood of septic patients (10). After 6 h of incubation, culture supernatants were collected and frozen at −80°C for further analyses. The cells were either collected in 175 μl lysis buffer (SV Total RNA Isolation kit; Promega, Madison, WI) and frozen at −80°C for further RNA analyses or fixed on the cover slips with 4% paraformaldehyde for microscopic examination and counting of Hoechst 33342-stained nuclei. In control experiments, samples were incubated for the same time in the same medium without the addition of the TNF-α or LPS (control for TNF-α or LPS effects on IL-6 secretion) or stimulated with either TNF-α (10 ng/ml) or LPS (100 ng/ml) without the addition of Dex (control for Dex effects).

Determination of IL-6

Concentrations of the secreted IL-6 were determined with an ELISA kit (Endogen, Rockford, IL) according to the manufacturer's instructions. Standards and samples were diluted in the same culture medium. The concentration range in the standard curve was 0–400 pg/ml, and our samples were diluted 1:50 and 1:100 before the determination of IL-6. The comparability of cell material in each condition is given by the number of nuclei belonging to either myoblasts or myotubes. This was determined by counting 10 vision fields/well (100 vision fields/each experiment). The amount of IL-6 was calculated per 100,000 nuclei to be able to compare results acquired for myoblasts and myotubes and for the different experimental conditions.

Determination of Expression of Receptor Subunits

Total RNA was isolated from myoblasts and myotubes using the SV Total RNA Isolation kit (Promega). RNA was eluted with 100 μl of diethyl pyrocarbonate-treated water, and total RNA was determined spectrophotometrically. With the use of a first-strand cDNA synthesis kit (Reverse Transcription System; Promega), mRNAs were reverse transcribed by Avian Mieloblastosis Virus-RT. Reverse transcription was primed with oligo(dT) in a volume of 40–80 μl.

The final volume (25 μl) of PCR reaction mixture contained 0.1–3 μl of cDNA, 0.5 units of HotStarTaq DNA Polymerase, 40 μmol dNTPs, 1× Q-Solution in PCR buffer (all from Qiagen, Valencia, CA), and 100 pmol of Invitrogen primers. After 15 min of DNA polymerase activation at 95°C (MJ Research thermocycler), 35–40 cycles of amplification were performed with a final 10-min extension at 72°C. Each PCR cycle consisted of denaturation at 94°C for 1 min, annealing at 46 or 54°C for 1 min, and extension at 72°C for 1 min. After electrophoresis in a 1% agarose gel, PCR products were examined for their sizes using a 100-bp ladder PCR marker (Promega). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a standard. Primers for receptors were as follows: GAPDH forward 5′-GTCAGTGGTGGACCTGACCT-3′ and GAPDH reverse 5′-TGCTGTAGCCAAATTCGTTG-3′. Primers used for the determinations of the receptor subunits were as described before for TNF receptor (TNFR) 1 and TNFR2 (15) and for Toll-like receptor (TLR) 2 and TLR4 (38).

Testing Muscle Cultures for Apoptosis and Necrosis


Because TNF-α and Dex cause apoptosis in some cell types (18, 35) and this outcome would significantly influence our IL-6 measurements and comparison between the treated and control cultures, we tested our cultures for this effect. Cultures were tested for apoptosis on the basis of DNA fragmentation and nuclear chromatin changes.


Cells were incubated with 100 nM Dex for 24 h. After harvesting, total DNA was isolated using the Wizard Genomic DNA Purification Kit (Promega). Electrophoresis was performed on ethidium bromide-stained 1.8% agarose gel and visualized with a transilluminator. To obtain a positive control of the DNA ladder, cells were treated with 10% DMSO for 3 days.


Cells grown on sterile collagen-coated cover slips or floating dead cells were fixed in PBS containing 4% paraformaldehyde, stained with 1 mM Hoechst 33258 in PBS, and examined under a fluorescence microscope. Cells were scored as apoptotic if they exhibited unequivocal nuclear chromatin condensation and/or fragmentation as described previously (16).


To test myoblasts and myotubes for the integrity of their membranes after treatment, we determined the concentration of lactate dehydrogenase (LDH, EC in supernatant. LDH was determined in clear supernatants in one batch at the end of the study with the analyzer Vitros 950 (Ortho-Clinical Diagnostics, Neckerdgemünd, Germany). Samples were stored at −70°C for no longer than 3 mo. Imprecision of triplicate determination was <2.1%.


Results were expressed as means ± SE. Univariate two-way or three-way ANOVA was used to test the difference between myoblasts and myotubes in IL-6 secretion under different experimental conditions. Group differences were tested post hoc using Multiple Student's t-test with Bonferroni correction. Statistical significance was set at P < 0.05. Data were compared using SPSS 13.0 for Windows (SPSS, Chicago, IL) and Microsoft Excel (Microsoft Office Excel 2003).


Individual Effects of TNF-α, LPS, and Dex on the Secretion of IL-6 from the Myoblasts and Myotubes

Both mononuclear myoblasts and myotubes constitutively secrete IL-6 in the cultured human muscle. Among muscle donors, secretion varied from 1,000 to 7,000 pg of IL-6 per 100,000 nuclei/24 h. To prevent the interference of this variability with our determinations of the effects of the pro- and anti-inflammatory factors on the IL-6 secretion, we used only cultures from one donor in each series of experiments.

In most cases, the constitutive secretion rate of IL-6 in our untreated, control cultures was ∼300,000 molecules of IL-6 per nucleus per 24 h. No significant difference was observed in this respect between mononuclear myoblasts and myotubes. Secretion of IL-6 was strongly stimulated by both LPS and TNF-α, the latter being more potent in this regard: a five- to sixfold increase of IL-6 secretion in TNF-α-treated and a two- to fourfold increase in LPS-treated myoblasts was observed. Myotubes were less sensitive to both TNF-α and LPS than myoblasts [interaction concentration by cell type: F(3, 32) = 8.1, P < 0.001 for TNF-α treatments; interaction concentration by cell type: F(3,38) = 40.5, P < 0.001 for LPS treatments; Fig. 1, A and B]. The effects of TNF-α and LPS on the IL-6 secretion were concentration dependent [main effect concentration: F(3,32) = 100, P < 0.001 for TNF-α treatments; main effect concentration: F(3,38) = 77.5, P < 0.001 for LPS treatments]. Within the same concentration range, we found this concentration dependency in myoblasts close to linear, whereas in myotubes it appeared more bell shaped with maximal IL-6 secretion at the concentration of 10 ng/ml for both TNF-α and LPS (Fig. 1, A and B).

Fig. 1.

Effects of tumor necrosis factor (TNF)-α (A), lipopolysaccharide (LPS; B), and dexamethasone (Dex; C) on the secretion of interleukin (IL)-6 in myoblasts (open bars) and myotubes (filled bars). Columns represent means ± SE of 3–6 determinations of IL-6 secretion (n in parentheses). Data correspond to IL-6 secretion in the untreated cultures (control) and in the cultures treated with three different concentrations of TNF-α, LPS, and Dex. IL-6 secretion is presented in arbitrary units so that 1 in A and B and 100 in C represent the amount of IL-6 secreted in 24 h (A and B) or 6 h (C) per 100,000 nuclei. *Significant differences (P < 0.05) between treated and control samples. †Significant differences between 1 ng/ml and other concentrations of TNF-α or LPS. ‡Significant difference between 10 ng/ml LPS and other LPS concentrations. Gauges above pairs of columns indicate significant difference between myoblasts and myotubes under the same experimental condition (P < 0.001).

Dex treatment inhibited constitutive IL-6 secretion equally at both developmental stages [interaction concentration by cell type: F(3,16) = 1.0, P = 0.417; main effect concentration: F(3,16) = 6.7, P = 0.004]; however, this effect was statistically significant only at the highest Dex concentration tested (P < 0.05 for both myoblasts and myotubes). At the Dex concentrations up to 1 nM, we did not observe statistically significant effects on the constitutive IL-6 secretion (Fig. 1C).

Effects of Dex on TNF-α- and LPS-stimulated Secretion of IL-6 from Human Myoblasts and Myotubes

In one series of experiments, we exposed myoblasts and myotubes simultaneously to Dex and either TNF-α (10 ng/ml) or LPS (100 ng/ml). The secreted IL-6 was measured after 6 h of combined treatment. Dex treatment decreased TNF-α- and LPS-induced IL-6 secretion [main effect concentration: F(3,48) = 14.13, P < 0.001]. We observed no difference between myoblasts and myotubes with regard to the Dex effects [main effect cell type: F(3,48) = 0.404, P < 0.751]. Again, only at the highest (100 nM) concentration did Dex statistically significantly decrease TNF-α (P < 0.05)- and LPS (P < 0.05)-induced IL-6 secretion in both myoblasts and myotubes (Fig. 2).

Fig. 2.

Effects of Dex on TNF-α- and LPS-induced IL-6 secretion from human myoblasts and myotubes. IL-6 secretion is expressed in percentages: 100% represent the amount of IL-6 secreted in 6 h/100,000 nuclei in control, untreated myoblasts or myotubes. MEM, minimum essential medium. Inhibition of IL-6 secretion was statistically significant (P < 0.05) only at the 100 nM concentration of Dex (*) in both myoblast and myotubes. Bars represent means ± SE. F and P in top right refer to the Dex effects from 3-way ANOVA.

Expression of Receptors for TNF-α and LPS in Cultured Human Myoblasts and Myotubes and Analyses for Apoptosis and/or Necrosis

The observed differences in the responses of myoblasts and myotubes to the TNF-α and LPS could be the result of differential expression of specific receptors i.e., TLRs (TLR2 and TLR4) and TNFRs (TNFR1 and TNFR2) in the myoblasts and myotubes. Testing the expression of these receptors, we found that, of all the receptors tested, only TLR4 was not expressed. mRNAs of TNFR1, TNFR2 and TLR2 were expressed at both myoblast and myotube levels (Fig. 3).

Fig. 3.

Expression of transcripts encoding receptors for TNF-α and LPS. Representative electrophoretic identification of mRNAs encoding TNF receptor (TNFR) 1, TNFR2, and Toll-like receptor (TLR) 2 subtypes in myoblast cultures. TLR4 could not be detected under our experimental conditions. No difference could be observed between myoblasts and myotubes regarding the expression of these receptors. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

TNF-α, LPS, and Dex did not cause apoptosis or necrosis under our experimental conditions. Apoptosis was tested by analyzing our cultures for apoptotic DNA fragmentation and apoptotic nuclei exhibiting chromatin condensation. Such condensation was found only in the DMSO-treated cultures, serving as a positive control, whereas TNF-α-, LPS- and Dex-treated cultures remained intact (Fig. 4). Release of LDH in culture media as a marker for muscle lysis was also negative. LDH concentrations in the media were in the range of 10−4 catalytic units and in many samples under the detection limit. There were no significant differences in LDH release between the treated and control cultures.

Fig. 4.

A: analysis for apoptosis in human muscle cultures. This representative plate shows TNF-α-treated mixed cultures in which both myoblasts (arrow) and myotubes (arrowhead) are present. Using fluorescence nuclear staining with Hoechst 33342, we could not find apoptotic nuclei. The results (not shown) were the same after LPS or Dex treatment. B: apoptotic nuclei were, however, found in the myoblast cultures after 24 h treatment with 10% DMSO, which served as a positive control. Bar = 10 μm.


This is the first report in which the individual effects and the cross talk of major pro- and anti-inflammatory factors on IL-6 release from the human skeletal muscle have been studied differentially at the myoblast and myotube stages. Such an approach is necessary to gain more detailed understanding of the effects of these factors on muscle regeneration. Namely, regeneration is a repeat of muscle ontogenesis and proceeds through the same stages as met during embryonic development. In this development, mononuclear myoblasts differ substantially from the myotubes with regard to the expressed set of proteins, especially muscle-specific proteins, and consequently in response to various extrinsic factors (8, 22).

Constitutive secretion of IL-6 did not differ significantly between myoblasts and myotubes. Our estimation of ∼300,000 molecules of IL-6 per nucleus per 24 h resembles constitutive secretion in other cell types, for which the data are available, for example, microglia (25). The physiological meaning of the constitutive secretion of IL-6 from the muscle precursors is not known. In other tissues, it participates in the modulation of the immune response. Contractions were demonstrated to induce the release of IL-6 in the skeletal muscle; it has been suggested that the released IL-6 contributes to the metabolic adaptations of the muscle fiber and of the organism in general to the physical activity (17). This role of IL-6 is, however, unlikely in our system, since cultured human myoblast and myotubes do not contract unless innervated (1, 26).

TNF-α and LPS stimulated IL-6 secretion from both myoblasts and myotubes. However, unlike constitutive IL-6 secretion, which was practically the same in myoblasts and myotubes, TNF-α- and LPS-stimulated IL-6 release was differential and was significantly more pronounced in the myoblasts than in the myotubes. It is important to note that our results are underestimating this differentiation stage specificity, since myotube cultures were always contaminated with myoblasts that, for technical reasons, could not be avoided. This observation is in accordance with the view that proinflammatory factors stimulate muscle regeneration by potentiating the release of IL-6, which has strong stimulatory effects on myoblast proliferation (2–4, 10, 13). Given the same number of satellite cells, it is the proliferation rate of myoblasts that most importantly determines the mass of differentiated muscle developed in the further stages and therefore the final extent of regeneration. Autocrine and paracrine proliferatory effects of IL-6, released from myoblasts as a result of stimulatory effects of proinflammatory factors, are therefore of outmost importance in this process.

Dex decreased not only constitutive but also TNF-α- and LPS-stimulated IL-6 secretion, which is in accordance with the findings of Carballo-Jane et al. (12). However, in our experiments, this effect was mild and statistically insignificant in the concentration range corresponding to the normal daily concentrations of endogenous cortisol [14–55 nM, which is approximately equivalent to 0.5–2 nM Dex (32)]. Even at the two to three times higher endogenous cortisol concentration (corresponding to 1.5–6 nM Dex), which is met under various stress conditions (27), IL-6 secretion is probably not seriously affected. However, 500 nM cortisol concentration (equivalent to ∼20 nM Dex), determined in survivors with severe sepsis (33), might already be within the range when the suppressive effects of Dex on the IL-6 secretion become significant. Because there are also substantial individual variations in constitutive IL-6 secretion among donors (up to 7-fold differences were observed in our experiments) and because high doses of GCs are still sometimes used in the treatment of critically ill patients (5, 28), individuals with constitutively low capacity of IL-6 secretion might, after such treatment, have IL-6 secretion reduced to the level where its proliferative effects on myoblasts is greatly diminished. Under such conditions, IL-6-stimulated muscle regeneration might be substantially impaired, supporting the proposed concept that insufficient muscle regeneration is an important contributor to the severe chronic muscular disability manifested in almost one-third of all surviving critically ill patients.

Differentiation stage dependency of the proinflammatory factors TNF-α and LPS and differentiation stage independency of the anti-inflammatory Dex effect on the IL-6 secretion suggest that, although coupled, the mechanisms of these factors are under separate control. According to the generally accepted scheme, TNF-α and LPS activate the nuclear factor (NF)-κB signaling pathway, which results in the increased transcription of cytokines, including IL-6 (29). On the other hand, GCs, after activating their receptors, which then bind to the glucocorticoid-response elements stimulate the transcription of inhibitory factor IκBα, which binds and inactivates NF-κB and cytokine expression through the NF-κB pathway. Because we could not find differences in the patterns of expression of receptor subtypes between the two developmental stages, the reason for the more pronounced response of myoblasts to proinflammatory stimuli might be in the differential expression of other signal-transducing molecules or because of some other more general differences between the two differentiation stages.

Although TNFR1 is known to mediate apoptosis (36), we found no differences in this regard between control and TNF-α-treated cultures. This observation suggests that early myogenic precursors destined for proliferation and muscle regeneration are somehow protected against apoptosis. As for the effects of LPS, we demonstrated expression of only TLR2 but not TLR4. According to some authors, TLR4 binds gram-negative bacterial LPS, and TLR2 is the receptor for gram-positive peptidoglycan and lipoproteins (30, 34). Our results are more in accordance with the study of Yang et al. (38), who reported that TLR-2 also mediates LPS-induced cellular signaling, given the presence of LPS-binding protein in plasma, which is analogous to serum in the culture medium.

In conclusion, constitutive secretion of IL-6 did not differ significantly between myoblasts and myotubes; however, TNF-α- and LPS-stimulated IL-6 release was significantly more pronounced in myoblasts than in myotubes. This is in accordance with the view that proinflammatory factors stimulate muscle regeneration by potentiating the release of IL-6. Its autocrine and paracrine actions, with IL-6 released from the myotubes participating in the latter, strongly stimulate myoblast proliferation. Inhibition of IL-6 release at the high doses of GCs might therefore seriously hamper muscle regeneration in constitutively more susceptible critically ill patients.


This work was supported by the Slovenian Research Agency.


We gratefully acknowledge help from Drs. Branka Wraber and Darko Cerne. Bojana Ziberna and Zvonka Frelih are acknowledged for technical assistance. We are also grateful to Gaj Vidmar from the Institute of Biomedical Informatics, Medical School Ljubljana, for statistical advice. We are especially grateful to Dr. Matthew Watt and Dr. Graeme Lancaster for critical reading of the manuscript.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract