Muscle disorders are observed in several pathologies such as sepsis, trauma, acquired immunodeficiency syndrome, cancer, and others (39). In these conditions, a loss of muscle mass is observed, which is associated with higher circulating proinflammatory cytokines, especially tumor necrosis factor alpha (TNF-α) (4). This cytokine disrupts skeletal muscle protein balance by increasing protein degradation and reducing protein synthesis. In C2C12 myotubes, Li et al. reported that TNF-α activates protein degradation through the NF-kB signaling pathway (24). Studies in rats have also showed a TNF-α–induced protein degradation in muscle (9,16). This activation of protein degradation could result from the activation of the ubiquitin-dependent pathway (11,25). Also, TNF-α administration reduces muscle protein synthesis in rats by suppressing protein translation initiation (21). Overall, muscle weight loss has been documented in animals administered with TNF-α (8) and in diseases with elevated endogenous TNF-α (2). Taken together, these data identify TNF-α as a mediator of skeletal muscle loss.
The anti-inflammatory effects of exercise are currently well documented (14,20). In healthy adults and elderly people, physical activity is associated with lower levels of inflammatory markers. Several mechanisms are proposed for this exercise-induced anti-inflammatory response. The interleukin (IL) 6 transiently released by contracting skeletal muscle inhibits the expression of inflammatory cytokines (37). Interleukin 6 also increases the expression of anti-inflammatory cytokines (38). In addition, regular physical activity reduces the expression of Toll-like receptors (TLR) in monocytes, macrophages and skeletal muscle. In this way, regular exercise attenuates TLR downstream responses such as the production of proinflammatory cytokines (15,19). Exercise training also reduces white adipose tissue mass and attenuates the expression of inflammation-related adipokines in obesity (32). All these effects are beneficial for the prevention and treatment of many diseases, turning exercise an important therapy (29).
Although the anti-inflammatory effects of training are well documented, whether they protect skeletal muscle from sudden inflammatory insults is currently unknown. We tested that hypothesis by using an acute TNF-α injection and analyzing the molecular responses in skeletal muscle.
MATERIALS AND METHODS
Animals
Eight-week-old female C57BL/6 J mice (Janvier, France) were housed in a controlled environment (22°C–23°C, 14/10 h light/dark cycle) and fed ad libitum with standard chow. Mice were divided in 4 groups (n = 6 each): untrained, untrained with TNF-α injection, trained and trained with TNF-α injection. Three days after the last training session or an equivalent period in the untrained groups, mice were injected intraperitoneally with recombinant TNF-α (Sigma–Aldrich, St. Louis, MO; 100 ng·g−1 body weight) or with an equal volume of physiological saline solution. Six hours after, the animals were killed by cervical dislocation. The tibialis anterior muscles were dissected and frozen in liquid nitrogen. Samples were then stored at −80°C until processed. All procedures were accepted by the committee for ethical practices in animal experiments of the Université catholique de Louvain. The housing conditions were in accordance with the Belgian Law of May 29, 2013, on the protection of laboratory animals (agreement number: LA-1220548) and with the American College of Sports Medicine standards for animal care.
Exercise training protocol
Mice in the trained groups were familiarized with treadmill exercise by running for 30 min (8 m·min−1) on 2 successive days. Afterward, the maximal velocity was determined by an incremental exercise test. The test started with velocity of 8 m·min−1, and increments of 2 m·min−1 every 2 min were applied until exhaustion. The maximal velocity was defined as the velocity of the last stage completed by the animals. Then, the mice were trained 1 h·d−1 at 60% of maximal velocity, 5 d·wk−1, during 6 wk. Maximal velocity was measured again on the third week, and the velocity was adapted accordingly to ensure the same exercise intensity for the entire training period.
Protein extraction and immunoblotting
The tibialis anterior muscles were placed in ice-cold buffer (20 mM Tris, 270 mM sucrose, 5 mM EGTA, 1 mM EDTA, 1% Triton X-100, 1 mM sodium orthovanadate, 50 mM sodium β-glycerophosphate, 5 mM sodium pyrophosphate, 50 mM sodium fluoride, 1 mM 1,4-dithiothreitol [DTT], and 10% protease inhibitor cocktail 10X [Roche Applied Science, Vilvoorde, Belgium]) and then homogenized using a TissueLyser device (Qiagen). Homogenates were centrifuged at 10,000 g for 10 min at 4°C. Supernatants were collected and stored at −80°C. Protein concentration was determined using the DC protein assay kit (Bio-Rad Laboratories, Hercules, CA) with bovine serum albumin as standard. Fifty micrograms of total proteins were denaturated by mixing with Laemmli buffer. Linearization was achieved by heating the samples for 5 min at 100°C. The samples were loaded with prestained molecular mass markers (Thermo Fisher Scientific). The proteins were then separated on SDS–polyacrylamide gels. The gels were run in an electrophoresis unit at constant voltage (i.e., 100 V). Proteins were then transferred to a PVDF membrane at constant voltage (i.e., 90 V) for 2 h. Membranes were then blocked for 1 h in Tris-buffered saline with 0.1% v/v Tween 20 (TBST) containing 5% of nonfat dried milk. Membranes were then incubated with the following primary antibodies (1:1000 dilution in TBST containing 1% of bovine serum albumin): phospho-Akt (Ser473), Akt, phospho-mTOR (Ser2448), mTOR, phospho-S6 ribosomal protein (Ser235/236), S6 ribosomal protein, phospho-4E-BP1 (Thr37/46), 4E-BP1, phospho-FoxO1 (Thr24)/FoxO3a (Thr32), FoxO3a, TNF-R1, TNF-R2, phospho-IKKα/β (Ser176/180), phospho-IKBα (Ser32), IKBα, and LC3B. All primary antibodies were obtained from Cell Signaling Technology (Leiden, The Netherlands) except for LC3B from Sigma–Aldrich (St. Louis, MO). Membranes were then incubated with a secondary antibody (1:1000 diluted in TBST containing 5% of nonfat dried milk) conjugated to horseradish peroxidase from Cell Signaling Technology. Chemiluminescent detection was carried out using an ECL Western blotting kit (Amersham ECL Plus, GE Healthcare, Diegem, Belgium). Pictures were taken with a charge-coupled device (CCD) camera (Gbox, Syngene, The Netherlands) and signal quantification was determined by GeneTool software (Syngene); α-tubulin was used as loading control, and all results were expressed relative to the untrained vehicle-injected group.
RNA extraction and quantitative real-time PCR
Total RNA was extracted from tibialis anterior using Trizol reagent according to the manufacturer’s protocol (Invitrogen, Vilvoorde, Belgium). The RNA quality and quantity were assessed using Nanodrop spectrophotometry. cDNAs was synthesized from 1 μg of total RNA using iScript cDNA synthesis kit (Bio-Rad Laboratories), then diluted in nuclease-free water and kept at −20°C until use. Real-time PCR experiments were done on a MyIQ2 thermocycler (Bio-Rad Laboratories) using the following conditions: 3 min at 95°C, followed by 35 cycles of 30 s at 95°C, 30 s at 60°C, and 30 s at 72°C. qPCR mixture contained 4.8 μl IQSybrGreen SuperMix (Bio-Rad Laboratories), 0.1 μl of each primer (100 nM final), and 5 μl of cDNA. Primer sequences are in Table 1. The specificity of the amplification was assessed by the melting curve. Each sample was tested in duplicate, and negative controls containing water instead of cDNA were included each time. Data analysis was done using the 2-ΔΔCT method with ribosomal protein L19 (RPL19) as the reference gene. Preliminary data ensured that RPL19 did not change in response to exercise training and TNF-α treatment.
;)
TABLE 1: Primers sequences (5’-3’).
Statistical analysis
Results are presented as means ± SEM. All statistical analyses were performed with GraphPad Prism 6.0. Two-way analysis of variance (ANOVA) was used to test the main effects of TNF-α and training and also the interaction of these 2 factors. In the figures, the results of the ANOVA analyses are presented in a box next to the histograms. Bonferroni post hoc test was used to determine the differences between vehicle and TNF-α-injected mice in the same training condition and differences between untrained and trained mice in the same treatment condition. In the figures, the results of the post hoc tests are presented by asterisks and horizontal lines above the histograms. The threshold of signification was set at P < 0.05.
RESULTS
Training attenuates body weight gain but does not prevent TNF-α–induced cytokine induction in muscle
During the 6 wk of the training program, the trained mice gained less body mass than the untrained mice (P < 0.01, ANOVA, main effect of training; P < 0.05, Bonferroni post hoc test; Fig. 1A). At the end of the program, trained mice presented also a higher muscle mass (P < 0.001, ANOVA, main effect of training; Fig. 1B).
FIGURE 1: Endurance training does not protect skeletal muscle against TNF-α–induced cytokine expression. A, Body weight gain calculated as the difference between the end and the beginning of the experimental period. B, Relative weight of tibialis anterior muscles. C–F, mRNA levels of TNF-α, IL-6, IL-1β, and MCP-1 cytokines. Results are expressed as means ± SEM (n = 6 per group). Vh, vehicle-injected; TNF, TNF-α–injected. Boxes show the main effects and interaction obtained by the 2-way ANOVA analysis. When significant, Bonferroni post hoc analysis was performed and its results added on the bars. ns: nonsignificant, *P < 0.05, **P < 0.01, ***P < 0.001.
Globally, acute TNF-α injection decreased the tibialis anterior mass (P < 0.01, ANOVA; Fig. 1B), but post hoc analysis revealed that the decrease was only significant in untrained mice (−12%, P < 0.05, Bonferroni post hoc).
The mRNA of several cytokines was measured, in tibialis anterior muscle, to assess the inflammatory status of muscle upon TNF-α injection. Independently of the training state, TNF-α-injected mice exhibited higher levels of TNF-α, IL-6, IL-1β, and MCP-1 mRNA as revealed by the ANOVA analysis (Fig. 1C–F).
Endurance training prevents TNF-α–induced NF-κB activation
TNF-α binding to its receptors leads to TRAF2 re-cruitment and subsequently IκBα kinase (IKK) phosphorylation/activation (8). Activated IKK phosphorylates the NF-κB inhibitor alpha (IκBα) inducing its degradation by proteasome, thus allowing NF-κB migration into the nucleus (30). To evaluate the potential anti-inflammatory effect of training, NF-κB activation was estimated through IKKα/β and IκBα phosphorylation states. Whereas TNF-α increased IKKα/β and IκBα phosphorylations in untrained mice, the response was blunted in the tibialis anterior muscle of trained ones (P < 0.001 and P < 0.01 ANOVA, interaction between training and TNF, respectively; Fig. 2A–B). These data suggest that endurance training protects against NF-κB activation induced by TNF-α.
FIGURE 2: Endurance training alters the TNF-α–induced NF-κB activation in skeletal muscle. A–B, Phosphorylation state of IKKα/β and IkBα. α-tubulin was used as loading control. C–D, mRNA levels of TNF-R1 and TNF-R2. E–F, Protein expression of TNF-R1 and TNF-R2; α-tubulin was used as loading control. Results are expressed as means ± SEM (n = 6 per group). Vh, vehicle-injected; TNF, TNF-α–injected. Boxes show the main effects and interaction obtained by the 2-way ANOVA analysis. When significant, Bonferroni post hoc analysis was performed and its results added on the bars. ns: nonsignificant, *P < 0.05, **P < 0.01, ***P < 0.001.
To understand the mechanisms behind the protective effect, we tested the hypothesis that endurance training downregulates the expression of the TNF-α receptors (TNF-R1/TNF-R2). Although the mRNA level of TNF-R1 was lower in trained than untrained mice (P < 0.01 ANOVA, main effect of training; Fig. 2C), the protein levels of TNF-R1 and TNF-R2 were not significantly affected (Fig. 2E–F) by the training program, ruling out our hypothesis.
However, the TNF-α injection increased the mRNA level of both receptors when the trained and untrained groups were considered (P < 0.05 ANOVA, main effect of TNF; Fig. 2C–D). Nevertheless, the post hoc analysis showed that the effect was only significant in the untrained mice (P < 0.05, Bonferroni post hoc; Fig. 2B–C). Again, no significant changes were highlighted at the protein level (Fig. 2E–F).
Endurance training does not prevent the inhibition of the Akt/mTOR signaling induced by TNF-α
The lower NF-κB activation observed in trained muscle suggests that training impairs TNF-α signaling in this tissue. Thus, we analyzed whether training prevents the decrease in protein synthesis induced by acute TNF-α injection (22). For that purpose, different markers of the protein synthesis cascade were analyzed.
Tumor necrosis factor–α reduced Akt phosphorylation in tibialis anterior muscle of untrained mice, but the effect was blunted in the trained ones (P < 0.05, ANOVA, interaction between training and TNF factors; P < 0.01, Bonferroni post hoc in the untrained group; Fig. 3A). We also analyzed Akt downstream target mTORC1, along with 2 mTORC1 substrates: ribosomal protein S6 (rpS6) and 4E-BP1. Tumor necrosis factor–α reduced the phosphorylation of all these proteins independently of the training status (P < 0.05, ANOVA, main effect of TNF; Fig. 3B–D).
FIGURE 3: TNF-α downregulates protein synthesis signaling independently of the training status. A–D, Phosphorylation state of Akt, mTOR, rpS6, and 4E-BP1. Results are expressed as means ± SEM (n = 6 per group). Vh, vehicle-injected; TNF, TNF-α–injected. Boxes show the main effects and interaction obtained by the 2-way ANOVA analysis. When significant, Bonferroni post hoc analysis was performed and its results added on the bars. ns: nonsignificant, *P < 0.05, **P < 0.01.
Endurance training reduces the induction of catabolic markers triggered by TNF-α
Studies in rats (12) and C2C12 myotubes (24) demonstrate that TNF-α induces protein degradation in skeletal muscle. We thus analyzed some catabolic markers after TNF-α administration to test whether training prevents this response. The forkhead box transcription factors (FoxO) stimulate protein degradation by activating 2 systems: ubiquitin proteasome and autophagy (33,40). Globally, TNF-α increased FoxO1 mRNA levels (P < 0.01, ANOVA, main effect of TNF; Fig. 4A), but this increase was only significant in the untrained mice (P < 0.01, Bonferroni post hoc; Fig. 4A). As for Fox3a, neither the mRNA nor the protein levels were modified by any condition (Fig. 4B–C). On the other hand, FoxO3a phosphorylation/inactivation was upregulated by training (P < 0.001, ANOVA, main effect of training; Fig. 4D).
FIGURE 4: Endurance training protects skeletal muscle against the upregulation of ubiquitin-proteasome markers induced by TNF-α. A–B, mRNA levels of FoxO1 and FoxO3a. C, Protein expression of FoxO3a; α-tubulin was used as loading control. D, Phosphorylation state of FoxO3a in relation to total FoxO3a. E–G, mRNA levels of MURF1, MAFbx, and myostatin (Mstn). Results are expressed as means ± SEM. Vh, vehicle-injected; TNF, TNF-α–injected. Boxes show the main effects and interaction obtained by the 2-way ANOVA analysis. When significant, Bonferroni post hoc analysis was performed and its results added on the bars. ns: nonsignificant, *P < 0.05, **P < 0.01, ***P < 0.001.
Regarding the ubiquitin proteasome system, 2 FoxO-regulated ubiquitin-ligases were analyzed, namely, MURF1 and MAFbx (33). When trained and untrained groups were considered together, TNF-α injection increased MURF1 mRNA (P < 0.01, ANOVA, main effect of TNF; Fig. 4E), but the post hoc analysis revealed that this increase was only significant in untrained mice (P < 0.01, Bonferroni post hoc; Fig. 4E). The changes in MAFbx mRNA due to TNF-α were dependent on the training status (P < 0.05, ANOVA, interaction between training and TNF; Fig. 4F). Tumor necrosis factor–α induced a significant increase in MAFbx mRNA only in untrained group (P < 0.01, Bonferroni post hoc; Fig. 4F). We also analyzed myostatin, another regulator of MAFbx activity (31). Whereas TNF-α increased myostatin mRNA in untrained mice, the effect was fully blocked by training (P < 0.01, ANOVA, interaction between training and TNF; P < 0.001, Bonferroni post hoc between the untrained mice; Fig. 4G).
Concerning autophagy, FoxO3a has been reported to increase the mRNA level of several autophagy markers (40). We thus analyzed some of them in the tibialis anterior muscle. The mRNA of BNIP3 tend to increase after TNF-α injection (P = 0.07, ANOVA, main effect of TNF; Fig. 5A). This response tended to be significant only in the untrained mice (P = 0.05, Bonferroni post hoc; Fig. 5A). Another marker, BNIP3L, presented a general downregulation due to training (P < 0.05, ANOVA, main effect of training; Fig. 5B). Additionally, BNIP3L mRNA tended to increase after TNF-α injection, specifically in the untrained mice (P = 0.05, ANOVA, main effect of TNF; P = 0.05, Bonferroni post hoc in the untrained mice; Fig. 5B).
FIGURE 5: Endurance training protects skeletal muscle against the upregulation of autophagy markers induced by TNF-α. A–E, mRNA levels of BNIP3, BNIP3L, Gabarapl1, p62, and LC3b. F, LC3BII/LC3BI protein ratio; α-tubulin was used as loading control. Results are expressed as means ± SEM. Vh, vehicle-injected; TNF, TNF-α–injected. Boxes show the main effects and interaction obtained by the 2-way ANOVA analysis. When significant, Bonferroni post hoc analysis was performed and its results added on the bars. ns: nonsignificant, *P < 0.05, **P < 0.01.
We also measured the autophagy regulated-genes Gabarapl1, p62, and LCB3. Gabarapl1 was induced by TNF-α at the mRNA level (P < 0.05, ANOVA, main effect of TNF; Fig. 5C). Of note, the induction was significant in the untrained mice (P < 0.05, Bonferroni post hoc; Fig. 5C) but not in trained ones. Also, there was a trend to a lower Gabarapl1 mRNA in the trained mice (P = 0.08, ANOVA, main effect of training; Fig. 5C). The mRNA of both p62 and LC3B were downregulated in trained muscles (P < 0.05, ANOVA, main effect of training; Fig. 5D–E). Besides, TNF-α injection tended to increase LC3B mRNA only in the untrained mice (P = 0.06, Bonferroni post hoc; Fig. 5E).
Upon autophagy, LC3BI is converted to LC3BII through a lipidation process. Then, it associates with the autophagic vesicles (17). The LC3BII/LC3BI ratio is thus a marker of autophagic vesicle formation. Tumor necrosis factor–α increased LC3BII/LC3BI ratio in untrained mice, whereas the effect was attenuated in the trained ones (P < 0.05, ANOVA, interaction between training and TNF; P < 0.01, Bonferroni post hoc in the untrained mice; Fig. 5F).
DISCUSSION
Here, we demonstrated that endurance training attenuates some effects of TNF-α in mice skeletal muscle. Trained mice had lower TNF-α–dependent increases in the mRNA of TNF-R1, TNF-R2, FoxO1, MURF1, MAFbx, myostatin, and Gabarapl1. Similarly, trends to lower TNF-α–dependent inductions of BNIP3 and BNIP3L were found with training. Also, the increase in LC3BII/LC3BI ratio induced by TNF-α was almost absent in trained muscles. All this may be the consequence of the blunted NF-κB activation induced by TNF-α in trained muscles.
Tumor necrosis factor–α was used with the goal of upregulating catabolic markers in skeletal muscle. Although an intraperitoneal injection of 100 ng·g−1 TNF-α probably induced a higher plasma concentration than in physiological or even pathological states, we chose this dose on the basis of previous studies showing a large increase of muscle-specific ubiquitin-ligases, MAFbx and MuRF1 (1,23). The mice were killed 6 h after the injection because Li et al., using the same model, reported the peak of MAFbx mRNA at that time (23). Our preliminary experiments revealed that MAFbx and MuRF1 mRNA were more increased in the tibialis anterior muscle than in the soleus and gastrocnemius (data not shown). Consequently, we used the tibialis anterior for further analyses. Previous results of our laboratory show that tibialis anterior adapts to endurance training as evidenced by an increased expression of succinate dehydrogenase (7).
Acute TNF-α exposure induces an inflammatory response in skeletal muscle
Several reports show that TNF-α directly makes protein balance negative (6,13,23,25). In agreement with that, we found a small decrease in muscle mass in TNF-α–injected mice. This was surprising, given the acute nature of the TNF-α stimulation used here. This suggests that chronic and/or repeated exposure to TNF-α may produce dramatic decreases in muscle mass.
As anticipated, TNF-α injection induced an inflammatory response in skeletal muscle of untrained mice. This response was characterized by increases in the mRNA levels of several cytokines. Training had no influence on the cytokine mRNA levels either in the basal (vehicle) or stimulated (TNF-α) state. In the basal state, the lack of change in IL6 mRNA by training agrees with a previous report (3). Nevertheless, a downregulation of TNF-α mRNA was reported in a study wherein TNF-R1/2 knockout mice were used (18). As these transgenic animals had a higher level of TNF-α mRNA in the basal state, their response to acute exercise cannot be extrapolated to adaptations to training of wild-type mice.
On the other hand, the similar cytokine response to TNF-α in untrained and trained mice was surprising. Our data suggest that NF-κB activation induced by TNF-α was blunted in trained mice. A lower cytokine response was thus expected. One possible explanation is that training shortens, but does not blunt, the NF-κB activation induced by TNF-α. Therefore, 6 h after the TNF-α injection, NF-κB would be inactive in the trained group. This may explain our findings. Another possibility is that TNF-α induced cytokine expression in an NF-κB–independent manner. It has been demonstrated that TNF-α activates p38 MAPK in neonatal rat myotubes (6). This MAPK is a known regulator of the expression of several cytokines (34). It is thus plausible that the TNF-α–induced p38 activation increased the cytokine expression in our model. This would explain the normal regulation of the cytokine expression in trained mice in face of an inactive NF-κB. A kinetic of the NF-κB activation upon TNF-α in trained mice would help to clarify these findings.
Altogether, the results demonstrate a normal TNF-α–induced cytokine response in trained muscle. In addition, our data suggest that training alters NF-κB activation in response to TNF-α.
Endurance training attenuates the mRNA induction of TNF-α receptors produced by TNF-α
There are 2 receptors TNF-α can bind, namely, TNF-R1 and TNF-R2 (27). Upon TNF-R1 binding, TNF receptor associate factor-2 (TRAF2) is recruited by TRADD. The formation of the TNF-R1/TRADD/TRAF2 complex leads rapidly to NF-κB activation. Tumor necrosis factor–α binding to TNF-R2 also induces TRAF2-dependent NF-kB activation (27). Notably, different types of TNF-α activate preferentially the receptors. Tumor necrosis factor–R1 is activated by soluble TNF-α, whereas TNF-R2 is mainly activated by membrane-bound TNF-α (27). Another difference between the receptors is the duration of their activity. Whereas TNF-R1 signaling induces a short NF-κB activation (1–3 h), the activation through TNF-R2 is longer (up to 24 h) (27).
In the present study, we showed an upregulation in the mRNA of both TNF-R1 and TNF-R2 by TNF-α. Remarkably, the effect was less pronounced in trained muscles. This suggests that endurance training protects muscles from part of the TNF-α signaling. The mechanisms responsible for the protective effect are unknown. No modification in the expression of TNF-α receptors was found, maybe because of the acute nature of the experiment. Thus, changes in the sensitivity of the receptors or in the intracellular signaling appear as more plausible possibilities.
TNF-α downregulates the protein synthesis pathway independently of the training status
Previous studies show that TNF-α inhibits the insulin signaling (22,35). In the same line, we found that TNF-α decreased Akt phosphorylation in untrained muscles, but trained muscles were protected. However, training had no effect on the repression by TNF-α of the downstream targets of the Akt/mTORC1 pathway, that is, rpS6 and 4E-BP1. This suggests that training may not protect muscle against the inhibition of protein synthesis induced by TNF-α.
Endurance training protects muscle from protein degradation induced by TNF-α
Akt protein not only promotes protein synthesis but also inhibits protein degradation signaling through the FoxO proteins (13). Because we observed a regulation of the activation of Akt by endurance training during the TNF-α treatment, we also investigated the impact of training on signaling pathways regulating protein degradation. As mentioned in the introduction, evidence suggests that TNF-α directly activates protein degradation (9,16). It has been demonstrated that TNF-α activates the non-lysosomal ubiquitin-dependent proteolytic activity during muscle wasting (10,11). In rats, TNF-α increases the mRNA of genes of the ubiquitin proteasome system, together with the levels of ubiquitin (11). Afterward, a direct induction of the ubiquitin proteasome system by TNF-α has been confirmed in rats (25).
Here, we showed that TNF-α increases the mRNA levels of FoxO1, MURF1, MAFbx, and myostatin in untrained mice. Interestingly, endurance training attenuated all those effects. This suggests that endurance training protects skeletal muscle from an acute dose of TNF-α.
Our results fully agree with data from a study in a model of muscle wasting in rats (36). In that study, TNF-α mRNA increased in soleus muscle during heart failure–induced muscle wasting (36). This was associated with upregulations in the expression of NF-κB, FOXO, MAFbx, MURF1, and myostatin (36). Notably, the authors demonstrated that aerobic training prevents the upregulation of all those markers (36). By injecting TNF-α, we found upregulations of the same markers, and training also prevented them.
Previous reports show a link between NF-κB signaling and the ubiquitin pathways (13). Consequently, the attenuated NF-κB response to TNF-α might explain the protection of training against protein degradation. The mechanisms responsible for the deregulation of the NF-κB response in trained mice are unknown. Most of the effects of TNF-α on protein degradation are mediated by TNF-R1 (26). We did not detect any modification in TNF-R1 protein level that could explain the effect of training. Thus, changes in its sensitivity or subsequent signaling may have occurred.
Furthermore, myostatin, a member of TGF-β family, is known to cause muscle atrophy and probably induce the MAFbx and MURF1 expression (31). It is reasonable to envision that myostatin factor linked NF-κB pathway and ubiquitin proteasome system because administration of pentoxyfilline (an inhibitor of TNF-α synthesis) decreased myostatin expression (5). Consistent with this, we observed an upregulation of myostatin induced by TNF-α in untrained mice. This upregulation was prevented with training, probably because of the inhibition of the TNF-α signaling.
Regarding the autophagy-lysosomal degradation system, endurance training also protected muscle from its activation at the single time point of 6 h. This was clear for Gabarapl1 and LC3BII/LC3BI ratio, but there were also trends for BNIP3 and BNIP3L. Again, the alteration on the NF-κB signaling might be behind these effects. This hypothesis is supported by a recent study showing that the ubiquitin-proteasome system and the autophagosome formation were blocked when the TNF-α signaling was altered by inhibiting the adaptor protein TRAF6 in muscle. (28).
In conclusion, endurance training attenuated the TNF-α–induced expression of protein degradation markers in skeletal muscle. The attenuated NF-κB response observed may be the mechanism responsible for such effects. These findings are encouraging for next applications in treatment of muscle disorders because of a high level of inflammation and protein breakdown.
The authors thank Damien Naslain, Robin Rayter, and Cécile Jamart for the technical assistance during the mice training and sacrifice.
Authors’ Contributions: J. R. contributed to the experimental design and the execution of the experiments. J. R., R. F. V., and N. P. contributed to the data analysis and manuscript preparation. F. P. and M. F. supervised and contributed to the overall project and manuscript preparation. All authors revised the final version of the manuscript.
J. R. is a recipient of a postdoctoral fellowship financed by ELEONOR sprl and the Walloon region. R. F. V. is supported by CONICYT Beca CHILE Doctorado en el Extranjero, convocatoria 2012.
The authors have declared no conflict of interest. Results in this study do not constitute endorsement of the American College of Sports Medicine.
REFERENCES
1. Adams V, Mangner N, Gasch A, et al Induction of MuRF1 is essential for TNF-alpha-induced loss of muscle function in mice.
J Mol Biol. 2008; 384( 1): 48–59.
2. Ahmad S, Karlstad MD, Choudhry MA, Sayeed MM. Sepsis-induced myofibrillar protein catabolism in rat skeletal muscle.
Life Sci. 1994; 55( 18): 1383–91.
3. Akerstrom TC, Krogh-Madsen R, Petersen AM, Pedersen BK. Glucose ingestion during endurance training in men attenuates expression of myokine receptor.
Exp Physiol. 2009; 94( 11): 1124–31.
4. Argilés JM, López-Soriano FJ. The role of cytokines in cancer cachexia.
Med Res Rev. 1999; 19( 3): 223–48.
5. Costelli P, Muscaritoli M, Bonetto A, et al Muscle myostatin signalling is enhanced in experimental cancer cachexia.
Eur J Clin Invest. 2008; 38( 7): 531–8.
6. de Alvaro C, Teruel T, Hernandez R, Lorenzo M. Tumor necrosis factor alpha produces insulin resistance in skeletal muscle by activation of inhibitor kappaB kinase in a p38 MAPK-dependent manner.
J Biol Chem. 2004; 279( 17): 17070–8.
7. Deldicque L, Cani PD, Delzenne NM, Baar K, Francaux M. Endurance training in mice increases the unfolded protein response induced by a high-fat diet.
J Physiol Biochem. 2013; 69( 2): 215–25.
8. Dempsey PW, Doyle SE, He JQ, Cheng G. The signaling adaptors and pathways activated by TNF superfamily.
Cytokine Growth Factor Rev. 2003; 14( 3–4): 193–209.
9. Flores EA, Bistrian BR, Pomposelli JJ, Dinarello CA, Blackburn GL, Istfan NW. Infusion of tumor necrosis factor/cachectin promotes muscle catabolism in the rat. A synergistic effect with interleukin 1.
J Clin Invest. 1989; 83( 5): 1614–22.
10. García-Martínez C, Agell N, Llovera M, López-Soriano FJ, Argilés JM. Tumour necrosis factor-alpha increases the ubiquitinization of rat skeletal muscle proteins.
FEBS Lett. 1993; 323( 3): 211–4.
11. García-Martínez C, Llovera M, Agell N, López-Soriano FJ, Argilés JM. Ubiquitin gene expression in skeletal muscle is increased by tumour necrosis factor-alpha.
Biochem Biophys Res Commun. 1994; 201( 2): 682–6.
12. García-Martínez C, López-Soriano FJ, Argilés JM. Acute treatment with tumour necrosis factor-alpha induces changes in protein metabolism in rat skeletal muscle.
Mol Cell Biochem. 1993; 125( 1): 11–8.
13. Glass DJ. Skeletal muscle hypertrophy and atrophy signaling pathways.
Int J Biochem Cell Biol. 2005; 37( 10): 1974–84.
14. Gleeson M, Bishop NC, Stensel DJ, Lindley MR, Mastana SS, Nimmo MA. The anti-inflammatory effects of exercise: mechanisms and implications for the prevention and treatment of disease.
Nat Rev Immunol. 2011; 11( 9): 607–15.
15. Gleeson M, McFarlin B, Flynn M. Exercise and Toll-like receptors.
Exerc Immunol Rev. 2006; 12: 34–53.
16. Goodman MN. Tumor necrosis factor induces skeletal muscle protein breakdown in rats.
Am J Physiol. 1991; 260( 5 Pt 1): E727–30.
17. Hanada T, Noda NN, Satomi Y, et al The Atg12-Atg5 conjugate has a novel E3-like activity for protein lipidation in autophagy.
J Biol Chem. 2007; 282( 52): 37298–302.
18. Keller C, Keller P, Giralt M, Hidalgo J, Pedersen BK. Exercise normalises overexpression of TNF-alpha in knockout mice.
Biochem Biophys Res Commun. 2004; 321( 1): 179–82.
19. Lambert CP, Wright NR, Finck BN, Villareal DT. Exercise but not diet-induced weight loss decreases skeletal muscle inflammatory gene expression in frail obese elderly persons.
J Appl Physiol (1985). 2008; 105( 2): 473–8.
20. Lancaster GI, Febbraio MA. The immunomodulating role of exercise in metabolic disease.
Trends Immunol. 2014; 35( 6): 262–9.
21. Lang CH, Frost RA. Sepsis-induced suppression of skeletal muscle translation initiation mediated by tumor necrosis factor alpha.
Metabolism. 2007; 56( 1): 49–57.
22. Lang CH, Frost RA, Nairn AC, MacLean DA, Vary TC. TNF-alpha impairs heart and skeletal muscle protein synthesis by altering translation initiation.
Am J Physiol Endocrinol Metab. 2002; 282( 2): E336–47.
23. Li YP, Chen Y, John J, et al TNF-alpha acts via p38 MAPK to stimulate expression of the ubiquitin ligase atrogin1/MAFbx in skeletal muscle.
FASEB J. 2005; 19( 3): 362–70.
24. 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( 10): 871–80.
25. Llovera M, García-Martínez C, Agell N, López-Soriano FJ, Argilés JM. TNF can directly induce the expression of ubiquitin-dependent proteolytic system in rat soleus muscles.
Biochem Biophys Res Commun. 1997; 230( 2): 238–41.
26. Llovera M, García-Martínez C, López-Soriano J, et al Role of TNF receptor 1 in protein turnover during cancer cachexia using gene knockout mice.
Mol Cell Endocrinol. 1998; 142( 1–2): 183–9.
27. Naudé PJ, den Boer JA, Luiten PG, Eisel UL. Tumor necrosis factor receptor cross-talk.
FEBS J. 2011; 278( 6): 888–98.
28. Paul PK, Kumar A. TRAF6 coordinates the activation of autophagy and ubiquitin-proteasome systems in atrophying skeletal muscle.
Autophagy. 2011; 7( 5): 555–6.
29. Pedersen BK. The diseasome of physical inactivity—and the role of myokines in muscle—fat cross talk.
J Physiol. 2009; 587( Pt 23): 5559–68.
30. Renard P, Raes M. The proinflammatory transcription factor NFkappaB: a potential target for novel therapeutical strategies.
Cell Biol Toxicol. 1999; 15( 6): 341–4.
31. Rodriguez J, Vernus B, Chelh I, et al Myostatin and the skeletal muscle atrophy and hypertrophy signaling pathways.
Cell Mol Life Sci. 2014; 71( 22): 4361–71.
32. Sakurai T, Ogasawara J, Kizaki T, et al The effects of exercise training on obesity-induced dysregulated expression of adipokines in white adipose tissue.
Int J Endocrinol. 2013; 2013: 801743.
33. Sandri M, Sandri C, Gilbert A, et al Foxo transcription factors induce the atrophy-related ubiquitin ligase atrogin-1 and cause skeletal muscle atrophy.
Cell. 2004; 117( 3): 399–412.
34. Schindler JF, Monahan JB, Smith WG. p38 pathway kinases as anti-inflammatory drug targets.
J Dent Res. 2007; 86( 9): 800–11.
35. Sishi BJ, Engelbrecht AM. Tumor necrosis factor alpha (TNF-α) inactivates the PI3-kinase/PKB pathway and induces atrophy and apoptosis in L6 myotubes.
Cytokine. 2011; 54( 2): 173–84.
36. Souza RW, Piedade WP, Soares LC, et al Aerobic exercise training prevents heart failure-induced skeletal muscle atrophy by anti-catabolic, but not anabolic actions.
PLoS One. 2014; 9( 10): e110020.
37. Starkie R, Ostrowski SR, Jauffred S, Febbraio M, Pedersen BK. Exercise and IL-6 infusion inhibit endotoxin-induced TNF-alpha production in humans.
FASEB J. 2003; 17( 8): 884–6.
38. Steensberg A, Fischer CP, Keller C, Møller K, Pedersen BK. IL-6 enhances plasma IL-1ra, IL-10, and cortisol in humans.
Am J Physiol Endocrinol Metab. 2003; 285( 2): E433–7.
39. van Hall G. Cytokines: muscle protein and amino acid metabolism.
Curr Opin Clin Nutr Metab Care. 2012; 15( 1): 85–91.
40. Zhao J, Brault JJ, Schild A, et al FoxO3 coordinately activates protein degradation by the autophagic/lysosomal and proteasomal pathways in atrophying muscle cells.
Cell Metab. 2007; 6( 6): 472–83.
