Electrical stimulation modulates IGF binding protein transcript levels in C2C12 myotubes

Electrical stimulation (ES) of skeletal muscle can produce changes in metabolic enzyme and contractile protein gene expression resulting in fast‐to‐slow phenotypic changes. The molecular mechanism by which ES induces changes in phenotype is not entirely understood but recent reports have demonstrated that the calcineurin/NF‐AT signalling pathway is involved. IGF‐1 is also capable of inducing changes in phenotype through the same calcineurin/NF‐AT pathway but little is known of the direct effect of ES on the IGF system. In this study, we examined the effects of ES on the expression of igf‐1, igf‐2 and the six igfbp genes in the C2C12 muscle cell line. Results showed that ES induced a change in phenotype that was accompanied by downregulation of igf‐2 and upregulation of igfbp‐4 mRNA levels. However, ES did not significantly alter the transcription of igf‐1, igfbp‐2, igfbp‐5 and igfbp‐6 genes. This study demonstrates that ES of muscle cells in vitro not only directly modulates the gene expression of contractile proteins but also modulates proteins that are part of the IGF regulatory system, in particular IGFBP‐4. Copyright © 2004 John Wiley & Sons, Ltd.


INTRODUCTION
Numerous in vivo studies have demonstrated that electrical stimulation (ES) of skeletal muscle can produce dramatic changes in metabolic enzyme and contractile protein gene expression resulting in a fast-to-slow phenotypic change. 1 Subsequently, several studies have shown that in vitro ES can produce similar phenotypic changes in primary cell cultures 2 and in the C2C12 cell line. 3 More recently, transfection studies carried out in vitro have shown that over-expression of IGF-1 in the C2C12 cell line induced both a switch in phenotype and myotube hypertrophy, and that the IGF-1 actions were mediated by the calcineurin/NF-ATc1 pathway. 4 Furthermore, the fast-to-slow phenotypic transformation in response to ES was also shown to be mediated through the same calcineurin/NFATc1 signalling pathway. 5 These two recent studies therefore suggest that muscle phenotypic changes induced by ES could involve the IGF system as an upstream regulator of the NF-AT/calcineurin pathway.
It is well documented that IGF-1 actions are modulated at the protein level by at least six binding proteins (IGFBP). Evidence of a possible role for the IGFBPs in regulating muscle phenotype came from an in vivo study which showed that IGFBP-5 mRNA was differentially expressed in slow and fast muscles. The levels of this transcript were lower in the slow soleus muscle than in the fast EDL, gastrocnemius, plantaris and tibialis anterior muscles. 6 Furthermore, the mRNA levels for IGFBP-4 and IGFBP-5 were increased in the gastrocnemius muscle in response to denervation, which was also shown to induce phenotypic changes and muscle fibre atrophy. 7 These observations suggest that both igfbp-4 and igfbp-5 genes may be influenced by electrical activity and the gene products may play a role in fibre type switch. However, in vivo situations are complex and denervation can trigger other changes in which the IGF system could play a role. In this study, we therefore examined the transcriptional response of the IGF system, including IGF-1, IGF-2 and six IGFBPs, to electrical stimulation settings which were capable of inducing phenotypic changes in the C2C12 cell line.

Cell culture and electrical stimulation
Cells from a mouse skeletal muscle cell line C2C12 were grown to confluence on 0.1% gelatin-coated six-well multi dishes in Dulbecco's MEM-glutamax-1 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/steptomyocin/fungizone (PSF; all from Life Technologies). Myotube differentiation was induced by reduction of FBS to 2%. ES of the myotubes was initiated after 7 days of culture in the differentiation medium (DM) and was maintained for either 1, 2 or 3 days. During the period of ES, the culture media from both the control and stimulated cells were replaced daily. ES was carried out following a model previously described and optimized by Thelen et al. 3 Platinum electrodes (0.5 mm) were integrated into the lid of a six-well culture dish. In three of the six wells, the electrodes were disposed in two half circles to obtain a maximal area of stimulation, the three other wells did not contain electrodes and served as controls. The lid and electrodes were sterilized by immersion in 75% ethanol for at least 24 h and the ethanol was left to fully evaporate in a sterile cell culture hood before use. The platinum electrodes were attached to an electrical stimulator (GRASS, SD9 stimulator) and the C2C12 cells were stimulated at 3 V cm À2 at a frequency of 2 Hz.

RNA extraction and Northern blot analyses
Total RNA was extracted from the cell culture samples using RNAzol-B (BiotecX). RNA (10 mg) was denatured with de-ionized glyoxal/DMSO at 50 C for 1 h and separated by 1% agarose gel electrophoresis in 10 mM sodium phosphate buffer, pH 7.0. RNA samples were transferred onto nylon membranes (MSI, Osmonics, Inc), exposed to UV light for 2 min and heated at 80 C for 2 h. Pre-hybridization and hybridization of the nylon membrane was carried out at 48 C in 2.5Â Denhardt's reagent, 50% formamide, 0.3% SDS and 250 g cm À3 salmon sperm DNA. cDNA probes for IGF-1 and IGF-2 (ATCC), IGFBPs 1-6 (kindly supplied by S. Shimasaki), and embryonic and type IIB MyHC (kindly supplied by M. Buckingham) were isolated from plasmid vectors after digestion with the appropriate restriction enzymes. An 18 S rRNA probe was used as a loading control. The cDNA fragments were radioactively labelled using the random priming method with [ 32 P-dCTP and a multiprime DNA labelling kit (Amersham Int). The membranes were finally washed twice in 2 Â SSC, 0.1% SDS at room temperature followed by a wash with 0.2 Â SSC, 0.1% SDS at 60 C and exposed to X-Omat XAR-5-F6613 film (Kodak).

Data analysis
The intensity of the signal on the developed X-ray was assessed by image analyses (Molecular Analyst Software, Bio-Rad). Results were expressed as a ratio of 18 S rRNA and given as mean AE SEM. Data was analysed by two-tailed Student's t-test and considered significant when P < 0.05.

Morphological analyses
In the electrically-stimulated cultures, approximately 60% of the myotubes contracted synchronously whereas no contractions were observed in the nonstimulated cultures. Contractile activity in the stimulated C2C12 cells indicated that the differentiated myotubes had functional contractile apparatus (sarcomeres).

Effects of ES on Myosin Heavy Chain (MyHC) gene expression
Myotube maturation following ES did not affect the transcription of the embryonic MyHC gene, as levels of its mRNA transcript remained unchanged even after 3 days of ES. In contrast the levels of the adult fast type IIB MyHC transcript were significantly reduced after 2 days and maintained after 3 days showing that ES at 3 V cm À2 and at a frequency of 2 Hz was sufficient to induce phenotypic changes.

Effect of ES on the IGF system
The cDNA probe for IGF-1 detected three mRNA transcripts of 0.8, 1.5 and 7.0 kb in the C2C12 myotubes (data not shown), but these transcripts were only detectable at extremely low levels by the Northern blotting technique. ES did not increase the endogenous igf-1 gene expression in the C2C12 myotubes. The levels of IGF-2 mRNA were significantly reduced after 3 days of ES (Figures 1 and 2).
IGFBP-1 and IGFBP-3 mRNA transcripts were undetectable in both the control and stimulated C2C12 myotubes. IGFBP-2 and IGFBP-6 mRNAs were detected at very low levels in differentiated C2C12 myotubes and ES had no effects on the transcription of these genes. In contrast IGFBP-4 and IGFBP-5 were expressed at high levels in both control and stimulated myotubes (Figure 1). The IGFBP-4 mRNA levels significantly increased following 2 and  (Figures 1 and 2).

DISCUSSION
Numerous studies have demonstrated that ES of skeletal muscle in vivo can produce marked changes in muscle phenotype through changes in contractile protein and metabolic enzyme gene expression. 1 It has also been demonstrated that ES of 'aneural' myotubes in vitro can produce alterations in phenotype. 2 In the present study we examined the levels of two MyHC transcripts as markers of the efficacy of the ES regime employed. Myotubes in culture, without the trophic, neural, endocrine, mechanical and other environmental cues received by muscles in vivo, remain relatively immature during their viable period. Early developmental isoforms are continually expressed in cultures during this entire period. We have shown that the expression of one such isoform, embryonic MyHC, is unaffected by ES alone. In contrast, the adult fast type IIB MyHC isoform was significantly downregulated on ES of the C2C12 myotubes. This result is consistent with previous studies, where ES induced phenotypic changes in cultured myotubes, and demonstrates the efficacy of the stimulation protocol employed. 3 It is not clear whether the IGF system is involved in the phenotypic changes in muscles subjected to altered electrical activity. It has recently been demonstrated that IGF-1 can produce changes in muscle phenotype through a calcineurin/NF-AT pathway and that ES of myotubes in culture activates this pathway. 4,5 However, in vivo studies in the rabbit EDL muscle, have shown that IGF-1 mRNA levels are not affected by ES alone, whereas the fast type 2X MyHC isoform is markedly reduced. 8 Results from our in vitro study showing that IGF-1 mRNA levels were not affected by ES but those of the type IIB MyHC isoform were reduced, is consistent with these in vivo observations.
Our results for igf-2 gene expression in vitro also correlate with in vivo studies. Ishii showed that high IGF-2 mRNA levels in fetal rat muscle were reduced when the muscle was innervated. 9 IGF-2 mRNA levels were elevated in adult denervated muscle, and subsequently reduced on re-innervation of that muscle. In the present study we found that the expression of the igf-2 gene was directly influenced by increased electrical activity. The high levels of IGF-2 mRNA seen in the control unstimulated samples were decreased in cells that had been subjected to ES. These observations support the suggestion of Ishii that IGF-2 may be a key regulator of muscle innervation.
The actions of both IGF-1 and IGF-2 are regulated at the protein level by the IGFBPs. The IGFBPs bind IGF-1 and IGF-2 with different affinities and are capable of either increasing or inhibiting the actions of the IGF in the tissues in which they are expressed. Thus, even if the paracrine and endocrine levels of IGF-1 remain the same, changes in the levels of the IGFBPs may alter its action. Our study showed that IGFBP-1 and IGFBP-3 were undetectable in the C2C12 cells and, thus, they appear to have little function in this cell line. IGFBP-2 and IGFBP-6 mRNA were only detectable at very low levels and were not affected by ES suggesting that they have little influence on the mechanism of muscle adaptation to ES. IGFBP-5 and IGFBP-4 mRNAs were detected at relatively high levels in control cultures which is consistent with previous observations that these binding proteins may have a role in muscle differentiation. 6 Denervation studies showed that these two IGFBPs were both upregulated following 2 days of denervation in the adult rat muscle suggesting the expression of the two IGFBPs may be regulated by neural activity. However denervation is complex and produces a response not simply explained by the lack of electrical activation. In this in vitro study using C2C12 cultures, we observed no changes in IGFBP-5 mRNA levels following ES, whereas the levels of IGFBP-4 mRNA were significantly increased compared with control cultures following 2 and 3 days of ES. This suggests, that of the six binding proteins, IGFBP-4 may have the most important role in muscle adaptation to ES.
In conclusion, our in vitro study, which directly examined the role of electrical activity in isolation from other factors, suggests that the role of IGFBP-4 in mediating the effects of electrical activity needs further investigation.