Muscle contractions induce interleukin-6 mRNA production in rat skeletal muscles
I. H. Jonsdottir,
The Copenhagen Muscle Research Centre, Rigshospitalet, Denmark
Corresponding author's present address I. H. Jonsdottir: Institute of Physiology and Pharmacology, Department of Physiology, Göteborg University, Box 432, 405 30 Göteborg, Sweden. Email: email@example.com
1The present study explored the hypothesis that interleukin-6 (IL-6) might be locally produced in response to skeletal muscle contractions and whether the production might reflect the type of muscle contraction performed. Rats were anaesthetized and the calf muscles of one limb were stimulated electrically for concentric or eccentric contractions (4 × 10 contractions with 1 min of rest between the 4 series, 100 Hz). The contralateral muscles served as unstimulated controls. The mRNA levels for IL-6, the glucose transport protein GLUT-4 and β-actin in the rat muscles (white and red gastrocnemius and soleus) were quantified by quantitative competitive RT-PCR.
2The IL-6 mRNA level, measured 30 min after the stimulation, increased after both eccentric and concentric contractions and there were no significant differences in IL-6 mRNA levels between the different muscle fibre types. No significant increase in IL-6 mRNA level was seen in the unstimulated contralateral muscle fibres.
3No increase in GLUT-4 mRNA level was detected, indicating that the increase in IL-6 mRNA level was not due to general changes in transcription.
4We conclude that IL-6 is locally produced after muscle contraction, with no significant differences between different muscle fibre types. This local production of IL-6 is not due to general changes in transcription, since no changes in the level of GLUT-4 mRNA were found. The fact that increased IL-6 mRNA levels were seen after both concentric and eccentric contractions indicates that the production of IL-6 is not solely due to muscle damage, seen primarily after eccentric exercise.
It is well documented that unaccustomed exercise, particularly involving eccentric muscle contractions, results in disruption of contractile tissue and cytoskeletal components (Smith, 1991; Friden & Lieber, 1992). Furthermore, several investigators agree that some aspects of an inflammatory response are seen following eccentric exercise, involving production of cytokines that are released at the site of inflammation (Smith, 1991). Increased plasma levels of inflammatory cytokines, primarily IL-6, have been found in response to exercise (Pedersen et al. 1997). IL-6 has been shown to be a potent inducer of acute phase protein synthesis in both rats and humans (Heinrich et al. 1990) and it has been suggested that IL-6 might be involved in the generation of acute phase responses seen after exercise (Northoff & Berg, 1991).
Increased plasma levels of IL-6 have primarily been described in relation to eccentric exercise but IL-6 has also been shown to be elevated after concentric exercise. Ullum et al. (1994) investigated the effect of concentric exercise (cycling) on cytokine plasma levels and found a 2-fold increase in IL-6 in plasma whereas IL-1α, IL-1β and TNF-α remained below detection limits.
Bruunsgaard et al. (1997) showed that high-intensity eccentric exercise caused a more pronounced increase in the plasma level of IL-6, than concentric exercise, suggesting that the elevation of IL-6 could be related to muscle damage seen predominantly after eccentric exercise. We have recently shown, using a qualitative PCR technique, that mRNA for IL-6 was detectable in human skeletal muscle after intense prolonged exercise. IL-6 mRNA was not concomitantly detectable in circulating blood mononuclear cells (BMNCs), suggesting that IL-6 could be locally produced in the skeletal muscle (Ostrowski et al. 1998).
The aim of the present study was to explore further the hypothesis that IL-6 could be locally produced in skeletal muscle following muscle contractions. By using the quantitative competitive reverse transcription polymerase chain reaction (QCRT-PCR) technique IL-6 mRNA levels were measured in muscle after concentric or eccentric contractions and in the contralateral non-contracting muscle. The study was also designed to examine if there were any differences in IL-6 levels between different types of muscle fibres.
Male Wistar rats weighing 200–250 g were used in this study. All animals were kept on a 12–12 h light-dark cycle and fed on an ad libitum standard chow diet.
The animals were divided randomly into the following groups: (a) non-stimulated controls (n= 5); (b) concentric stimulation group (n= 6); (c) eccentric stimulation group (n= 6). In the eccentric stimulation group, both the stimulated muscles and the contralateral unstimulated muscles were included in the study.
All experiments were approved by the Danish Animal Inspectorate.
Eccentric/concentric contraction model
The calf muscles on one side were stimulated for concentric or eccentric contraction. The contralateral muscles of the rats stimulated with eccentric contractions served as unstimulated controls.
The rats were anaesthetized by an intraperitoneal injection of Dormicum (midazolam, 0.5 mg (kg body wt)−1, Roche, Switzerland) and Hypnorm (fentanyl, 20 mg (kg body wt)−1 and fluoanison, 1 mg (kg body wt)−1, Janssen, High Wycombe, Bucks, UK), laid on their backs, and both hindlimbs were fixed over a plastic block by the use of Velcro strips. Supplementary doses (Dormicum, 0.5 mg kg−1; Hypnorm, 20 mg kg−1) of anaesthetic were administered as indicated by the withdrawal reflex response to toe and/or tail pinch. The limbs were placed so the hip joint had an approximately 60 deg angle, the knee joint an approximately 60 deg angle and the ankle joint had a 90 deg angle. Hooks were placed under the Achilles' tendon and connected to a strain gauge and an air pressure system. One lower limb was stimulated directly and externally for contraction through a needle placed in the most distal part of the quadriceps muscle and the ipsilateral hook under the Achilles' tendon. The stimulus consisted of four sessions of ten trains (duration of 1000 ms, pulse duration 1 ms, 100 Hz). In order to gain approximately the same force development during the sessions the stimulus amplitude was increased gradually from 10 V during the first session to 15 V during the second session, 20 V during the third session and 25 V during the final session. The trains in each session were separated by 4 s, and the sessions were separated by a couple of minutes. The electrical stimulus per se made the calf contract concentrically, and this procedure was used in the concentric group. In the eccentric group the hindlimb was first stimulated for concentric contraction as described above and with a 400 ms delay the active calf was stretched using the air pressure system. A spring was inserted in series to smooth the muscle stretch, and a strain gauge was also inserted to measure the power output and input. All exercise sessions were finished before 12 am.
Bilateral calf muscles were obtained 30 min after the stimulation. The superficial part of the gastrocnemius muscle, which consists mainly of fast-twitch white fibres was cut out and clamped. The soleus muscle, which consists mainly of slow-twitch fibres was reflected and clamped. Finally, a portion of the deep part of the medial head of the gastrocnemius, consisting mainly of fast-twitch red fibres, was cut out and clamped. The animals were killed with an overdose of barbiturate. All samples were frozen immediately in liquid nitrogen and stored at −80°C.
We have previously shown, using this model, that no muscle damage was inflicted by concentric contractions as suggested by lack of histological changes and normal glycogen resynthesis. A prior study revealed no histological signs of muscle damage and the muscle glycogen concentration returned to the pre-stimulation level less than 24 h after the concentric stimulation. In contrast, eccentric contractions caused histological signs of muscle damage and the muscle glycogen remained subnormal more than 48 h after the stimulation (Asp et al. 1995).
Total RNA was isolated essentially as described by Chomczynski & Sacchi (1987). First a 100–300 mg muscle tissue sample was homogenized in 3 ml GIT solution (4 M guanidine thiocyanate, 25 mM sodium citrate, 100 mM β-mercaptoethanol, pH 7) using a polytron (Ultra-Turrax T8, Ika Labortechnik, Staufen, Germany). Then 50 μl Sarkosyl, 300 μl sodium acetate (pH 4.0) and 3 ml phenol were added and the sample was mixed thoroughly. After this, 600 μl chloroform-isoamyl alcohol (49:1) was added and the sample mixed thoroughly, followed by 15 min incubation on ice. After centrifugation at 5000 g at 4°C for 20 min, the aqueous phase was precipitated on ice with one volume of isopropanol for 1 h. After a further centrifugation at 5000 g at 4°C for 30 min, the pellet was resuspended in 600 μl GIT solution and precipitated on ice with 600 μl isopropanol for 1 h. Finally, after centrifugation at 13 000 g at 4°C for 10 min, the pellet was washed once with 1 ml 80 % ethanol. The dry pellet was resuspended in 100 μl RNase-free water and stored at −80°C for later use.
Competitors for QCRT-PCR were constructed using PCR to introduce a small deletion in the wild-type PCR product (Celi et al. 1993). The Pfu polymerase (Stratagene, La Jolla, USA) was used to amplify both the competitor and the wild-type PCR products from rat muscle cDNA using the primers listed in Table 1. The PCR products were cloned into the Sma I site of the pBlueScript II SK(+) vector (Alting-Mees & Short, 1989), with orientation parallel to the lac Z gene, resulting in the plasmids listed in Table 1. The plasmids were cut with Pvu II to release the insert before further use in PCR reactions.
Table 1. Primers used for PCR
Plasmid name †
*Using the respective sense primer in PCR. †PCR product cloned into pBlueScript II SK(+).
Total RNA was reverse transcribed into cDNA using the MMLV Reverse Transcriptase (SuperScript II, Life Technologies, Grand Island, New York, USA). A 1–3 μg total RNA sample was mixed with 0.5 ng poly-dT12–18 in a total of 12 μl and heated to 70°C for 10 min before transfer to ice. The sample was mixed with 4 μl 5 × First Strand Buffer (250 mM Tris-HCl, 375 mM KCl, 15 mM MgCl2, pH 8.3), 2 μl 100 mM dithiothreitol, and 1 μl 10 mM dNTP (10 mM each). After a short incubation at 42°C, 1 μl 200 U μl−1 SuperScript II was added and the sample was incubated at 42°C for 50 min followed by 15 min at 70°C to inactivate the transcriptase.
Quantitative competitive RT-PCR
The mRNAs for β-actin, IL-6 and GLUT-4 were quantified using a quantitative competitive RT-PCR assay (Sun et al. 1996). The competitive PCR reactions for each cDNA were performed by preparing a ‘master mix’ containing everything necessary but the primers and competitors. This master mix was allocated to three tubes containing the respective primers and then further subdivided into PCR tubes containing the respective ‘competitor mix’. Each PCR tube contained a final volume of 25 μl, consisting of 0.1 μl cDNA, 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 2 mM MgCl2, 200 μM dNTP (200 μM each), 5 μl competitor mix, and 1.25 U Platinium Taq polymerase (Life Technologies). The competitor mix consisted of equal amounts of a specific number of the three cut competitor plasmids diluted in TE solution (10 mM Tris-HCl, 1 mM EDTA, pH 8) containing 1 ng ml−1 Salmon Sperm DNA (Sigma, St Louis, MO, USA). The PCR reactions were performed in a PTC-200 PCR machine (MJ Research, Watertown, MA, USA) as follows: 1 cycle of 94°C for 1 min, followed by 45 cycles of 94°C for 5 s, 56°C for 30 s, 72°C for 30 s, followed by 1 cycle of 4°C for 15 s. After the PCR, the two PCR products were separated on a 2 % agarose gel (NuSieve 3:1, FMC BioProducts, Rockland, ME, USA) and stained with ethidium bromide. The bands were quantified using a video camera (Electrophoresis documentation and Analysis System 120, Eastman Kodak Company, Rochester, New York, USA). The validity of the quantitative competitive RT-PCR assays for β-actin, IL-6 and GLUT-4 mRNA was tested in two ways: (1) by quantifying different but known amounts (10–107 molecules) of the wild-type plasmids; (2) by quantifying different dilutions of a real cDNA sample and checking that the values were consistent with the dilution. Both tests gave the expected results, confirming the validity of the assay (data not shown).
Single-stranded DNA probes
Templates for probes were made by PCR using biotinylated primers. The biotinylated strand was retained using streptavidin-coated DynaBeads (Dynal, Oslo, Norway) and the complementary strand was resynthesized in the presence of α-32P-dATP (3000 mCi mmol−1) using the non-biotinylated primer and T7 DNA polymerase. The radioactive strand was isolated and used as probe. The GLUT-4 template was made from PCR on pCM43 using the GLUT-4 sense (biotinylated) and antisense primers (see Table 1). The β-actin template was made from cDNA using two primers specific for the 3′-UTR to avoid hybridization to other actin isoforms. The two primers were 3′-β-actin sense (biotin-AAAACTGGAACGGTGAAGGC) and 3′-β-actin antisense (GCCTCAACACCTCAAACCAC).
Northern blot analysis
Northern blot analysis was performed using the NorthernMax kit from Ambion (Austin, USA). Briefly, the RNA was separated on a denaturing formaldehyde agarose gel and blotted onto a positively charged nylon membrane (Positive membrane, Appligene Oncor, Illkirch, France). The membrane was then hybridized to the β-actin and GLUT-4 probes at 55°C overnight, followed by a wash at the same temperature. The signals were detected and quantified on a PhosphorImager (Storm, Molecular Dynamics, Sunnyvale, USA).
The data were tested for a normal distribution. The β-actin and GLUT-4 data were normally distributed, whereas the IL-6 data needed to be log-transformed to show a normal distribution. The data were tested by analysis of variance (ANOVA) for dependence on contractions or muscles. Student's paired t test was used to compare measurements with their respective control muscles. Indicated P values are two-tailed.
The mRNA levels for IL-6, GLUT-4 and β-actin in the rat muscles (white and red gastrocnemius and soleus) were quantified by using quantitative competitive RT-PCR. Figure 1 shows the amount of β-actin mRNA related to the amount of total RNA. No significant dependence on contractions or muscles could be detected (ANOVA, P > 0.05), confirming that the β-actin mRNA level is sufficiently constitutive to be used as an internal control.
Figure 2 shows that the IL-6 mRNA level did increase in the stimulated muscles following eccentric contractions (5- to 23-fold). However, a similar increase was seen in the stimulated muscles following concentric contractions (8- to 17-fold). Furthermore, there was no significant increase in the IL-6 mRNA level in the resting leg of the rats performing eccentric contractions, indicating that the IL-6 production is not due to a systemic signal. Also, there was no significant difference in IL-6 mRNA levels between the different muscles (ANOVA, P > 0.05).
To confirm that the IL-6 mRNA increase was not due to general changes in transcription, the level of GLUT-4 was measured (Fig. 3). As expected, over the short time span of the experiment, no significant dependence on contractions could be detected (ANOVA, P > 0.05). Furthermore, no significant differences between the different muscles could be detected.
To support the results obtained using the PCR method, the β-actin and GLUT-4 mRNA levels were also measured by Northern blot analysis (the IL-6 mRNA level is too low to be measured using this method). To facilitate comparison between the PCR results and the Northern blot results the values were normalized to the level of β-actin in white gastrocnemius within each leg (Fig. 4A). The results showed that the GLUT-4 mRNA level was similar in the three different muscles when measured both by PCR and by Northern blot and was related to the β-actin mRNA level. Figure 4B shows the GLUT-4 mRNA level normalized to the 28S ribosomal RNA level.
The major finding of this study is that IL-6 is locally produced in the skeletal muscle after muscle contractions. Thus, an increased level of IL-6 mRNA was seen only in the stimulated muscle, but not in the contralateral unstimulated muscle. There were no significant differences between the different muscle fibre types, e.g. similar levels of IL-6 mRNA were found in white and red fibres from gastrocnemius, as well as in soleus slow-twitch red fibres. Furthermore, we found no increase in the mRNA level for GLUT-4, indicating that the increase in IL-6 mRNA is not due to general changes in transcription within the time period studied. Surprisingly, similar increases in IL-6 mRNA levels were found after eccentric and concentric contractions, indicating that the production of IL-6 is not solely due to muscle damage seen primarily after eccentric exercise. A prior study, using the same model, revealed that after the concentric stimulation there were no histological signs of muscle damage and the muscle glycogen concentration returned to the pre-stimulation level in less than 24 h. In contrast eccentric contractions caused histological signs of muscle damage and the muscle glycogen remained subnormal more than 48 h after the stimulation (Asp et al. 1995)
The measurements of GLUT-4 mRNA show that the choice of internal control is important. However, the perfect control for muscles has yet to be identified. Since only mRNA was converted to cDNA using this RT-PCR method ribosomal RNA cannot be used as an internal control. Furthermore, ribosomal RNA may not be a good control for contraction-induced changes since it has been shown to increase 5-fold in rabbit muscles subjected to chronic stimulation when related to the muscle weight (Neufer et al. 1996). On the other hand, β-actin mRNA has been shown to increase with endurance training in rats (Ploug et al. 1990). In the present study, the β-actin mRNA level was normalized to the ribosomal RNA and the change could therefore reflect a decrease in the ribosomal RNA level. Actually, this explanation results in a calculated percentage increase in mRNA that is similar to the increase in GLUT-4. Nevertheless, the difference between 28S RNA and β-actin mRNA levels was small compared to the change in the IL-6 mRNA level and therefore should not compromise the conclusion of our study. Our results for GLUT-4 show that the mRNA level is similar in the three different muscles when measured both by PCR and by Northern blot analysis and is related to the β-actin mRNA level. However, several authors have previously shown that the levels of GLUT-4 mRNA and protein are 50 to 100 % higher in soleus than in white gastrocnemius, with an intermediate level in red gastrocnemius (Megeney et al. 1995; Han et al. 1995; Ploug et al. 1997; Kristiansen et al. 1997). This inconsistency seems to be due to the choice of reference mRNA. In previous studies, the GLUT-4 mRNA level was normalized to the ribosomal RNA or total RNA levels, whereas in the present study the level is normalized to the β-actin mRNA level. If the GLUT-4 mRNA level is normalized to the 28S ribosomal RNA level instead of the β-actin mRNA level, the previously reported differences between the muscles can be seen. However, a similar difference can be seen for the β-actin mRNA level when it is normalized to 28S ribosomal RNA levels. Therefore, the difference between the muscles vanishes when the GLUT-4 mRNA level is normalized to the β-actin mRNA level.
Ostrowski and coworkers (1998) have previously shown in humans that mRNA for IL-6 is detectable in skeletal muscle but not in circulating BMNCs and concluded that the presence of mRNA for IL-6 is not due to contamination of the muscle by blood. The present study supports the hypothesis that IL-6 is locally produced in skeletal muscle after muscle contractions, since the levels of IL-6 mRNA in the contralateral unstimulated muscle were not significantly increased.
Ostrowski et al. (1998) discussed likely sources of this locally produced IL-6, including the possibility that during adherence or infiltration of neutrophils or macrophages due to inflammation in the damaged tissue the cells are activated and thus become the local source of IL-6 mRNA. Asp and coworkers (1995) have previously shown, in the same eccentric exercise model as used in the present study, that accumulation of inflammatory cells is seen only after eccentric exercise and only in the white and red fibres of gastrocnemius but not in soleus. Thus, the present data do not support the idea that the source of IL-6 is infiltrating inflammatory cells.
The physiological role of IL-6 production in relation to muscle contractions is not known. However, overexpression of IL-6 in transgenic mice activates protein degradation and causes atrophy in the muscles (Tsujinaka et al. 1995). These changes can be prevented by administration of anti-IL-6 receptor antibody to the mice (Tsujinaka et al. 1996). Furthermore, injection of IL-6 in mice decreases the level of circulating IGF-1 (De Benedetti et al. 1997). These data suggest that one role of IL-6 production might be the regulation of muscle protein turnover. Therefore, the contraction-induced increase in IL-6 production may play a role in the regulation of muscle protein turnover in relation to muscle contractions.
We conclude that IL-6 is locally produced in the skeletal muscle after muscle contractions, with no significant differences between the various muscle fibre types. This local production of IL-6 is not due to general changes in transcription, since no changes in the levels of GLUT-4 mRNA were found. Furthermore, similar levels of local IL-6 production were found in response to both eccentric and concentric muscle contractions, indicating that IL-6 production is not closely related to muscle damage.
This study was supported by The Danish National Research Foundation (nr 504–14). Dr Jonsdottir is supported by The Swedish Medical Research Council, The Swedish Association For The Promotion Of Sports and the Swedish National Centre for Research in Sports. The technical assistance of Ruth Rousing and Hanne Villumsen is acknowledged.