J Physiol. (http://www.ncbi.nlm.nih.gov/pubmed/16081485) 2005 Oct 1;568(Pt 1):229-42. Epub 2005 Aug 4.
Isolation and validation of human prepubertal skeletal muscle cells: maturation and metabolic effects of IGF-I, IGFBP-3 and TNFalpha.
Grohmann M (http://www.ncbi.nlm.nih.gov/pubmed?term=Grohmann%20M%5BAuthor%5D&cauthor=true&cauthor_uid=16081485), Foulstone E (http://www.ncbi.nlm.nih.gov/pubmed?term=Foulstone%20E%5BAuthor%5D&cauthor=true&cauthor_uid=16081485), Welsh G (http://www.ncbi.nlm.nih.gov/pubmed?term=Welsh%20G%5BAuthor%5D&cauthor=true&cauthor_uid=16081485), Holly J (http://www.ncbi.nlm.nih.gov/pubmed?term=Holly%20J%5BAuthor%5D&cauthor=true&cauthor_uid=16081485), Shield J (http://www.ncbi.nlm.nih.gov/pubmed?term=Shield%20J%5BAuthor%5D&cauthor=true&cauthor_uid=16081485), Crowne E (http://www.ncbi.nlm.nih.gov/pubmed?term=Crowne%20E%5BAuthor%5D&cauthor=true&cauthor_uid=16081485), Stewart C (http://www.ncbi.nlm.nih.gov/pubmed?term=Stewart%20C%5BAuthor%5D&cauthor=true&cauthor_uid=16081485).
The impact of IGF-I, IGFBP-3 and TNFα on skeletal muscle differentiation was also examined. Both IGF-I and its analogue LR3 elicited comparable increases in myotube hypertrophy, protein content, and CK activity over controls, with the greatest difference observed at day 3. The augmented differentiation by IGF-I was concomitant with an increase in myoblast hyperplasia and myoblast incorporation into each myotube fibre; however, the percentage of myotubes formed was unaffected. We suggest that IGF-I is implicated in enhanced myoblast fusion (facilitated by an increase in myoblast number), which may in turn drive a more rapid differentiation process, in part by up-regulating myogenin protein expression (Quinn et al. 1993 (http://jp.physoc.org/content/568/1/229.full#ref-33)). These findings are supported in animal (Mitchell et al. 2002 (http://jp.physoc.org/content/568/1/229.full#ref-30)) and adult human investigations (Foulstone et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-18)).
Isolation and validation of human prepubertal skeletal muscle cells: maturation and metabolic effects of IGF-I, IGFBP-3 and TNFα
Malcolm Grohmann (http://jp.physoc.org/search?author1=Malcolm+Grohmann&sortspec=date&submit=Submit) 1 (http://jp.physoc.org/content/568/1/229.full#target-1),
Emily Foulstone (http://jp.physoc.org/search?author1=Emily+Foulstone&sortspec=date&submit=Submit) 1 (http://jp.physoc.org/content/568/1/229.full#target-1),
Gavin Welsh (http://jp.physoc.org/search?author1=Gavin+Welsh&sortspec=date&submit=Submit) 3 (http://jp.physoc.org/content/568/1/229.full#target-3),
Jeff Holly (http://jp.physoc.org/search?author1=Jeff+Holly&sortspec=date&submit=Submit) 1 (http://jp.physoc.org/content/568/1/229.full#target-1),
Julian Shield (http://jp.physoc.org/search?author1=Julian+Shield&sortspec=date&submit=Submit) 2 (http://jp.physoc.org/content/568/1/229.full#target-2),
Elizabeth Crowne (http://jp.physoc.org/search?author1=Elizabeth+Crowne&sortspec=date&submit=Submit) 2 (http://jp.physoc.org/content/568/1/229.full#target-2)and
Claire Stewart (http://jp.physoc.org/search?author1=Claire+Stewart&sortspec=date&submit=Submit) 4 (http://jp.physoc.org/content/568/1/229.full#target-4)
+ (http://jp.physoc.org/content/568/1/229.full) Author Affiliations
1Department of Surgery, University of Bristol, Bristol Royal Infirmary, Upper Maudlin Street, Bristol, UK 2Department of Paediatric Endocrinology, Institute of Child Health, Royal Hospital for Children, Upper Maudlin Street, Bristol, UK 3Department of Biochemistry, University of Bristol, University Walk, Bristol, UK 4Department of Exercise and Sport Science, Manchester Metropolitan University, Hassall Road, Alsager, UK
Corresponding author
C. Stewart: Department of Exercise and Sport Science, Manchester Metropolitan University, Hassall Road, Alsager, UK. Email:
[email protected] (
[email protected])
Next Section (http://jp.physoc.org/content/568/1/229.full#sec-1)
Abstract
We have developed a primary skeletal muscle cell culture model derived from normal prepubertal children to investigate the effects of insulin-like growth factor-I (IGF-I), insulin-like growth factor binding protein-3 (IGFBP-3) and tumour necrosis factor α (TNFα) on growth, differentiation and metabolism. Cells of myoblast lineage were characterized morphologically by desmin staining and differentiated successfully into multinucleated myotubes. Differentiation was confirmed biochemically by an increase in creatine kinase (CK) activity and IGFBP-3 secretion over time. IGF-I promoted whilst TNFα inhibited myoblast proliferation, differentiation and IGFBP-3 secretion. IGF-I partially rescued the cells from the inhibiting effects of TNFα. Compared to adult myoblast cultures, children's skeletal muscle cells demonstrated higher basal and day 7 CK activities, increased levels of IGFBP-3 secretion, diminished IGF-I/TNFα action and absence of the inhibitory effect of exogenous IGFBP-3 on differentiation. Additional studies demonstrated that TNFα increased basal glucose transport via GLUT1, nitric oxide synthase and p38MAPK-dependent mechanisms. These studies provide baseline data to study the interactivity effects of growth factors and cytokines on differentiation and metabolism in muscle in relation to important metabolic disorders such as obesity, type II diabetes or chronic wasting diseases.
Mechanisms of skeletal muscle maintenance and metabolism have been extensively studied in models of disease states from cachexia (Espat et al. 1994 (http://jp.physoc.org/content/568/1/229.full#ref-12)) to obesity-related insulin resistance (Schmitz-Peiffer, 2000 (http://jp.physoc.org/content/568/1/229.full#ref-40)). Most of the in vitro studies to date have utilized immortalized rat and mouse skeletal muscle cell lines (Roeder et al. 1988 (http://jp.physoc.org/content/568/1/229.full#ref-35); Stewart & Rotwein, 1996a (http://jp.physoc.org/content/568/1/229.full#ref-42)), but more recently primary cell cultures from adult humans have been established, revealing important actions of insulin-like growth factor (IGF) I and II and their binding proteins, as well as cytokines such as tumour necrosis factor α (TNFα) on myoblast survival, proliferation and differentiation (Foulstone et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-18)). The central importance of the IGFs to muscle growth and development has also been demonstrated using ‘knock-out’ studies in mice, which leads to impaired embryonic development, especially in skeletal muscle (Wood, 1995 (http://jp.physoc.org/content/568/1/229.full#ref-49)). Conversely, over-expression of IGF-I has led to elevated bone and muscle growth in transgenic mice (Mathews et al. 1988 (http://jp.physoc.org/content/568/1/229.full#ref-27); Coleman et al. 1995 (http://jp.physoc.org/content/568/1/229.full#ref-6)), as well as increased myogenin mRNA, a transcription factor directly associated with terminal myogenic differentiation (Florini et al. 1991 (http://jp.physoc.org/content/568/1/229.full#ref-16)). Over-expression of IGF-II, on the other hand, has minimal growth promoting effects in vivo (Rogler et al. 1994 (http://jp.physoc.org/content/568/1/229.full#ref-36)), but appears to act as a survival factor in culture by minimizing cell death during the transition from proliferating to differentiating myoblasts (Stewart & Rotwein, 1996b (http://jp.physoc.org/content/568/1/229.full#ref-43)).
The availability of biologically active IGFs is mainly controlled by the IGF binding proteins (IGFBPs), which have high binding affinity for IGF-I and IGF-II and once bound, block IGF receptor activation. They function not only as carrier proteins for the IGFs in the circulation, but also serve to maintain a circulating reservoir of these ligands (Firth & Baxter, 2002 (http://jp.physoc.org/content/568/1/229.full#ref-14)). The inhibitory effects of IGFBPs have been demonstrated using IGF analogues such as des-(1–3)IGF-I or [Arg3]IGF-I (LR3), produced by truncation or mutation of key residues in the amino terminus of IGF, resulting in reduced affinity toward the IGFBPs, while maintaining normal affinity for the IGF-I receptor (IGFIR) (Forbes et al. 1988 (http://jp.physoc.org/content/568/1/229.full#ref-17); Tomas et al. 1993 (http://jp.physoc.org/content/568/1/229.full#ref-46)). In contrast to the cell lines, primary adult skeletal muscle cell cultures express and secrete large quantities of endogenous IGFBP-3 (Crown et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-9)), shown to reduce myoblast differentiation (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)).
In addition to the IGFs and their binding proteins, pro-inflammatory cytokines such as TNFα have also been implicated in aetiology of skeletal muscle growth and degeneration, impacting on muscle wasting (Giordano et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-20)), insulin resistance (Saghizadeh et al. 1996 (http://jp.physoc.org/content/568/1/229.full#ref-38)) and inhibition of differentiation in both murine and adult skeletal muscle cultures (Meadows et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-29); Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)). While muscle is the main determinant of in vivo glucose disposal (DeFronzo et al. 1981 (http://jp.physoc.org/content/568/1/229.full#ref-10)), there is mixed literature concerning the effects of TNFα on glucose homeostasis, showing inhibitory (Lang et al. 1992 (http://jp.physoc.org/content/568/1/229.full#ref-26)), stimulatory (Ciaraldi et al. 1998 (http://jp.physoc.org/content/568/1/229.full#ref-5)) or no impact (Nolte et al. 1998 (http://jp.physoc.org/content/568/1/229.full#ref-32)) on glucose transport in muscle.
The IGF axis and TNFα system play a major role in controlling growth and differentiation of skeletal muscle in adults, but such findings cannot necessarily be extrapolated to children. Since little or no data exist in prepubertal children, we have focused initially on skeletal muscle derived from this population, to avoid the additional variables associated with inherent changes in insulin sensitivity in puberty, due to augmentation of the growth hormone (GH)–IGF axis and the effects of sex steroids. To this end we have developed an in vitro primary skeletal muscle cell culture model derived from prepubertal children to investigate the actions of IGF-I, IGFBP-3 and TNFα on cellular growth, differentiation and metabolism. Using this model we have demonstrated both similarities and differences in the behaviour of skeletal myoblasts derived from children when compared to adult cultures, and add further data to the increasing knowledge of the effects of TNFα on glucose utilization in this clinically relevant model.
Previous Section (http://jp.physoc.org/content/568/1/229.full#abstract-1)Next Section (http://jp.physoc.org/content/568/1/229.full#sec-13)
Methods
Patient data
Skeletal muscle biopsies were taken from the anterior abdominal wall of 14 prepubertal Caucasian children at the onset of routine elective abdominal surgery at the Royal Hospital for Children in Bristol. Patients underwent either pyeloplasty or nephrectomy operations (9 male/5 female), median (range) age was 4.4 (0.9–9) years, median (range) body mass index standard deviation score (BMI SDS) was −0.1 (−2.31 to +1.16). All patients had normal blood pressure and fasting insulin levels (median (range) 1.5 (1–4.0) mU l−1) and displayed normal systemic insulin sensitivity using QUICKI (Quantitative Insulin Sensitivity Check Index) (Uwaifo et al. 2002 (http://jp.physoc.org/content/568/1/229.full#ref-47)) (median (range) 0.47 (0.39–0.54)). Adult biopsies (2 male/2 female), median (range) age 37 (35.5–40.5) years, were taken from the anterior abdominal wall of patients with normal BMI undergoing benign upper gastro-intestinal operations. No patients had sepsis, malignant or endocrine conditions. The study was approved by the United Bristol Healthcare Trust Ethics Committee, and written informed consent was obtained from parent or guardian.
Reagents
Tissue culture plasticware was obtained from Greiner Bio-one (Kremsmunster, Austria). Fetal bovine serum (FBS), Ham's F-10 medium and trypsin–EDTA were obtained from Gibco Invitrogen (Paisley, UK). Streptomycin and penicillin were obtained from the local hospital pharmacy. Human recombinant insulin was obtained from Novo Nordisk (Bagsverd, Denmark) and human recombinant TNFα from Bachem (St Helens, UK). Phosphate buffered saline (PBS) was obtained from Oxoid (Basingstoke, UK). Recombinant human IGF-I and LR3 IGF-I (an analogue of IGF-I with greatly reduced affinity for the IGFBPs) were purchased from Gropep (Adelaide, Australia). Recombinant human non-glycosylated IGFBP-3 was kindly donated by C. Maack (Celtrix, CA, USA). Antibodies to desmin (clone D33) and StrepABComplex/HRP were obtained from Dako (Glostrup, Denmark). Antibodies to GPDH were purchased from Chemicon (CA, USA). In-house antibodies to human IGFBP-3 were raised against rabbit. GLUT4 antibodies (400064) were purchased from Calbiochem (CA, USA) and GLUT1 antibodies were kindly donated by Geoff Holman (University of Bath, UK). 125Iodine, 2-deoxy-d-[3H]glucose, nitrocellulose and enhanced chemiluminescence (ECL) Western blot detection kits were purchased from Amersham Pharmacia Biotech (Buckinghamshire, UK), BCA protein assay and supersignal West-Dura chemiluminescence reagents were purchased from Pierce (Rockford, IL, USA). All other chemicals and reagents unless otherwise stated were purchased from Sigma (Poole, UK).
Isolation and culture of skeletal muscle myoblasts
The isolation of skeletal muscle myoblasts was performed as previously described (Crown et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-9)). Tissue samples (0.2–0.5 g) were collected under sterile conditions at the onset of surgery. The biopsies were washed in PBS, cut into 1 mm3 pieces, and digested in 10 ml 0.05% trypsin−0.02% EDTA in PBS for 15 min with gentle mixing at 37°C. The supernatant was removed and added to 2 ml heat-inactivated fetal bovine serum (hiFBS) before centrifugation at 80 g for 4 min. The pellet was resuspended in 3 ml Ham's F10 growth medium, supplemented with 20% hiFBS, 2 mm l-glutamine, 100 U ml−1 penicillin and 0.1 mg ml−1 streptomycin. This process was repeated a further two times and the cell suspensions pooled and seeded onto a 75 cm2 0.2% gelatin coated tissue culture flask. Cells were maintained at 37°C in a humidified atmosphere of 5% CO2 and 95% air, and medium was changed every 72 h until 80–90% confluency was attained. All experiments were performed on cells at passages 3–5 of their growth kinetics to ensure consistency.
Proliferation assay
Cells were seeded in 24-well plates at 7000 cells cm−2 and cultured in Ham's F10 medium supplemented with 20% FBS. Proliferation was assessed over time at days 1, 3, 6, 8 and 10 post-plating. At each time point cells were washed twice in PBS and either trypsinized and counted by trypan blue exclusion or analysed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; thiazoyl blue (MTT) assay. The MTT assay was initiated by incubation of cells with 500 μl fresh growth medium and 125 μl of MTT stock solution at 2 mg ml−1 in PBS (final concentration 0.4 mg ml−1) for 4 h at 37°C. The MTT medium was aspirated and the accumulated purple formazan dissolved in 150 μl PBS and 750 μl DMSO per well for 5 min. The absorbance of the solution obtained was analysed at 570 nm with background subtraction at 690 nm. Empty wells were similarly treated and used as blanks. All experiments were performed before confluency and hence before contact inhibition occurred.
Differentiation of skeletal muscle myoblasts
Skeletal myoblasts were inoculated into 0.2% gelatin coated six-well plates at a density of 18 000 cells cm−2 and following a 16-h period of cell attachment in growth medium, were washed twice in PBS and induced to differentiate over a 7-day period in low serum medium (LSM) (minimal essential medium (MEM) supplemented with 2% hiFBS, 2 mm l-glutamine, 100 U ml−1 penicillin, and 0.1 mg ml−1 streptomycin). Differentiation was manipulated with a single dose of 50 ng ml−1 IGF-I, 50 ng ml−1 LR3, 150–300 ng ml−1 IGFBP-3 or 20 ng ml−1 TNFα upon transfer to LSM.
Characterization of differentiated skeletal muscle myoblasts
Immunoctyochemical characterization was carried out on cells grown on 35 mm dishes. Myoblasts were identified using antibodies to desmin at a dilution of 1: 100 followed by specific biotinylated secondary antibodies and a StreptAvidin biotinylated horseradish peroxidase (StrepABComplex/HRP) coupled with 3,3′-diamino benzidine (DAB). Nuclei were counterstained with haematoxylin. Cells were imaged using an Olympus BX40 light microscope (Olympus America Inc., NY, USA). Image acquisitions were calibrated with a 1 mm objective micrometer and were controlled and analysed using Image Pro Plus 4.0 software from Media Cybernetics (Silver Spring, MD, USA). Analysis of the average number of cells per field of view, the percentage of positively stained myoblasts/myotubes and the number of nuclei per myotube were calculated from six random fields of view (× 200), for each treatment, from 12 individuals.
Differentiation was characterized biochemically by measuring creatine kinase (CK) activity. Following treatment, cells were washed twice with PBS, scraped and lysed in 60 μl of Tris–Mes–Triton (50 mm Tris-Mes, 1% Triton X-100, pH 7.8) and stored at −80°C. Samples were assayed within 14 days of collection using a commercially available kit (Sigma CK-20) following the manufacturer's instructions. Enzymatic activity was normalized to total protein content as determined by the bicinchonic acid (BCA™) protein assay.
IGFBP-3 production and secretion were determined by western ligand and immunoblotting, described below, and additionally by an in-house radioimmunoassay (RIA) (Yateman et al. 1993 (http://jp.physoc.org/content/568/1/229.full#ref-51)), using an in-house antiserum (Hughes et al. 1995 (http://jp.physoc.org/content/568/1/229.full#ref-25)) at a final dilution of 1: 20 000 (interassay percentage CV5.1). IGF-I and IGF-II secretion were also analysed by RIA as previously described (Bowsher et al. 1991 (http://jp.physoc.org/content/568/1/229.full#ref-4)).
Isolation of cellular proteins and membranes
Following differentiation, conditioned media were removed and stored at −20°C and cell lysates were prepared. After washing cells in ice-cold PBS, total cellular proteins were extracted by scraping into ice-cold lysis buffer (10 mm Tris-Cl pH 7.6, 5 mm EDTA, 50 mm NaCl, 30 mm sodium pyrophosphate, 50 mm sodium fluoride, 100 μm sodium orthovanadate, 1% Triton X-100, and 1 mm phenylmethylsulphonylfluoride (added just prior to use), also supplemented with protease and phosphatase inhibitor cocktails). Lysates were cleared of insoluble material by centrifugation at 2000 g for 5 min at 4°C. Protein concentrations were quantified using the BCA protein assay. Protein extracts were used immediately or aliquoted and stored at −20°C. Total membranes were prepared from cell lysates by homogenization using a Wheaton A (tight fit) glass homogenizer (60 strokes), followed by centrifugation at 2000 g for 15 min at 4°C. The resulting supernatant was centrifuged at 190 000 g for 60 min at 4°C producing a total membrane pellet. The membranes were resuspended in ice-cold lysis buffer and the protein content determined.
Western immuno- and ligand blotting
Conditioned media (80 μl), whole cell extracts and membrane fractions (50 μg sample−1) were separated by 12.5% sodium dodecyl-polyacrylamide gel electrophoresis (SDS-PAGE) under non-reducing conditions (IGFBP analysis) or reducing conditions for GLUT1/4 detection. Following electrophoretic transfer of the proteins to nitrocellulose, non-specific binding sites were blocked with 3% bovine serum albumin (BSA) (IGFBP-3) or 5% non-fat dry milk (GLUT1 and GLUT4) in Tris-buffered saline (TBS; 10 mm Tris-HCl, pH 7.8, 150 mm NaCl), 0.1% Tween 20 (TBS-T) for 1–2 h at room temperature. Ligand blots were probed with a radiolabelled mixture of IGF-I and IGF-II in 50 mm Tris-Cl–3% BSA and visualized by autoradiography. Immunoblots were incubated overnight at 4°C with primary antihuman IGFBP-3 antibodies (1: 5000) or with GLUT1 and GLUT4 antibodies (1: 1000) in the appropriate blocking buffer. Following washing and incubation with horseradish-peroxidase conjugated secondary antibodies (1: 10 000), the proteins were detected by enhanced chemiluminescence after exposure to Kodak X-Omat LS film. Differences in protein abundance were quantified by densitometry using a scanning densitometer and analysed using Molecular Analyst software (Bio-Rad laboratories, Hercules, CA, USA), in addition to re-probing for the housekeeping protein GPDH. Data are expressed in arbitrary optical density units.
Immunocytochemical analysis of GLUT1 expression
Differentiating skeletal myoblasts were exposed to 20 ng ml−1 TNFα for the last 24 h of differentiation. Cells were washed twice in PBS, fixed in 2% paraformaldehyde and blocked in 1.5% goat serum in PBS. Antibodies to GLUT1 were used at a dilution of 1: 100, followed by specific biotinylated secondary antibodies with StrepABComplex/HRP and DAB detection. Nuclei were counterstained with haematoxylin and the cells imaged using an Olympus BX40 light microscope.
Glucose transport in differentiated skeletal muscle
Skeletal myoblasts were differentiated as described and for the last 24 h of differentiation exposed to 20 ng ml−1 TNFα. 2-Deoxy-d-[3H]glucose uptake was determined as a measure of the glucose transport system as previously described (Fletcher et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-15)). Results are expressed as percentage stimulation above basal to correct for interindividual variation in basal glucose transport rates.
Statistics
The data were subjected to Student's t test, and analysis of variance (ANOVA) with Pearson correlation coefficients using SPSS (version 11.5) (SPSS Inc., Chicago, USA). Results were expressed as the mean ± s.e.m. A P-value of less than 0.05 was considered to be significant.
(http://jp.physoc.org/content/568/1/229.full#sec-1) (http://jp.physoc.org/content/568/1/229.full#sec-21)
Results
Characterization of primary skeletal muscle cultures
Myoblasts derived from the small number of satellite cells present in the original muscle biopsies (Mauro, 1961 (http://jp.physoc.org/content/568/1/229.full#ref-28); Moss & Leblond, 1971 (http://jp.physoc.org/content/568/1/229.full#ref-31)) were successfully cultured from all subjects. Since the cultures are not homogeneous populations, the myoblasts were identified using an antibody to the muscle-specific protein desmin. Figure 1A (http://jp.physoc.org/content/568/1/229.full#F1) is a representative view of primary skeletal muscle cells in culture, typically comprising 57 ± 3% myoblasts; the remaining cell population is predominantly composed of muscle-derived fibroblasts. The percentage did not change with subculturing or under normal differentiating conditions (data not shown). Cellular proliferation was assessed over time by manual cell counting and MTT analysis, demonstrating a cell doubling time of 44 ± 4 h and 50 ± 4 h, respectively. All cultures formed multinucleated myotubes when induced to differentiate for 7 days in low serum medium (LSM) (Fig. 1A (http://jp.physoc.org/content/568/1/229.full#F1)). Differentiation was also assessed biochemically by measuring the activity of the muscle-specific enzyme creatine kinase (CK) at day 0, before placing the cells into LSM and at subsequent time points during the differentiation procedure. Maximal differentiation was attained after 7 days in LSM revealing a mean CK activity of 2293 ± 307 mU (mg protein)−1 (Fig. 1B (http://jp.physoc.org/content/568/1/229.full#F1)). ANOVA revealed a highly significant interaction between CK activity and time (P < 0.001). Prolonged differentiation (up to 14 days) did not increase either myotube formation or CK activity above that seen at day 7 (data not shown). Since adult skeletal muscle cultures produce large amounts of intact IGFBP-3 (Crown et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-9)), we wished to establish if our cells were also capable of secreting this protein. Children's skeletal myoblasts do indeed secrete IGFBP-3 into the medium, as shown by radioimmunoassay (RIA) (Fig. 1B (http://jp.physoc.org/content/568/1/229.full#F1)). Change in IGFBP-3 levels was highly significant with time (P < 0.001) and correlated significantly with CK activity (r = 0.83, P < 0.001) under differentiating conditions. The average concentration in day 7 conditioned medium was 290 ± 25 ng ml−1, equating to 346 ± 49 ng (mg protein)−1. Interestingly, when analysed by Western immunoblotting, IGFBP-3 showed evidence of proteolysis with increasing fragmentation over time (Fig. 1C (http://jp.physoc.org/content/568/1/229.full#F1)). Further analysis of day 7 myotube conditioned medium by RIA for IGF-I and IGF-II showed no significant production of endogenous IGF-I. However, low levels of IGF-II were detectable, 10.2 ± 1.4 ng ml−1.
Figure 1.
Differentiation of children's skeletal muscle myoblasts
A, cells were characterized pre- and post-differentiation using antibodies to the muscle-specific protein desmin. Upper panel, myoblasts at day 0; lower panel, myotubes after 7 days of differentiation. Representative micrographs are shown. Scale bar = 100 μm. B, CK activity (filled bars) and IGFBP-3 secretion (open bars) were determined from cell lysates and conditioned medium, respectively, at days 0, 1, 2, 3, 4 and 7 of differentiation. The data represent the mean CK activity and mean IGFBP-3 concentration ± s.e.m. from 14 individuals. C, representative IGFBP-3 Western immunoblot demonstrating IGFBP-3 secretion throughout differentiation.
Effect of IGF-I, LR3, IGFBP-3 and TNFα on myotube formation and CK activity
After characterization of the cell culture model, we next investigated whether differentiation could be manipulated using growth factors and cytokines that have been implicated in skeletal muscle maintenance and metabolism. Both IGF-I and LR3 (50 ng ml−1) enhanced differentiation over day 7 controls, by increasing the size of the myotubes formed, whereas 150 ng ml−1 IGFBP-3 was without effect (Fig. 2A (http://jp.physoc.org/content/568/1/229.full#F2)). Conversely, 20 ng ml−1 TNFα inhibited myotube formation when compared to day 7 LSM control (Fig. 2A (http://jp.physoc.org/content/568/1/229.full#F2)). When differentiation was analysed biochemically the largest impact of growth factors was evident at day 3, for both IGF-I and LR3. Data revealed a 142 ± 50% (P = 0.002) and 157 ± 51% (P = 0.02) increase in CK activity versus day 3 control, respectively (Fig. 2B (http://jp.physoc.org/content/568/1/229.full#F2)). This increase was maintained at day 7 of differentiation although to a lesser degree (Fig. 2C (http://jp.physoc.org/content/568/1/229.full#F2)). When analysed as the percentage increase in CK activity at day 7 versus day 0, differentiation was enhanced comparably with both IGF-I (416 ± 49%, P = 0.006) and LR3 (399 ± 42%, P = 0.003) versus day 7 control 318 ± 32%. There were no significant differences between either IGF-I or LR3 in their capacity to increase CK activity, P = 0.10. TNFα (20 ng ml−1), however, caused a 47 ± 6% decrease in differentiation versus day 7 controls, P < 0.001, while no significant effects were seen with exogenous IGFBP-3 (Fig. 2C (http://jp.physoc.org/content/568/1/229.full#F2)).
Figure 2.
Effect of IGF-I, LR3 IGF-I, IGFBP-3 and TNFα on myotube formation and CK activity
Cells were characterized pre- and post-differentiation using antibodies to the muscle-specific protein desmin. A, differentiation was manipulated with a single dose of 50 ng ml−1 IGF-I, 50 ng ml−1 LR3 IGF-I, 150 ng ml−1 IGFBP-3 and 20 ng ml−1 TNFα. Representative micrographs are shown. Scale bar = 100 μm. B, CK activity at day 3 of differentiation. C, CK activity following 7 days of differentiation. The data represent the mean percentage increase in CK activity versus day 0 controls ± s.e.m. from 14 individuals. Statistical analysis was performed using a Student's paired t test; *P < 0.05, **P < 0.01, ***P < 0.001.
To establish whether the actions of IGF on muscle were via hypertrophy, we calculated the percentage increase in protein versus day 0 controls along with the effects of IGFBP-3 and TNFα. As expected protein levels were significantly increased at day 7 in control cells, by 74 ± 10% (P < 0.001), although cell numbers remained constant. Protein levels were further enhanced by IGF-I (40 ± 8%, P < 0.001), LR3 (48 ± 11%, P < 0.001) and also TNFα (35 ± 7%, P < 0.001). IGFBP-3 was without effect (Fig. 3A (http://jp.physoc.org/content/568/1/229.full#F3)). The increase in protein levels following TNFα exposure was unexpected as TNFα reduced myotube formation and inhibited differentiation (Fig. 2A and C (http://jp.physoc.org/content/568/1/229.full#F2)). To examine if the increased protein levels were due to increased cell numbers, we analysed the number of cells from six random fields of view from 12 individuals. Interestingly, a single dose of 20 ng ml−1 TNFα at day 0 had no impact on myoblast proliferation (Fig. 3B (http://jp.physoc.org/content/568/1/229.full#F3)) but elevated total cell numbers by increasing fibroblast number, 55 ± 9 per field of view, over day 7 controls, 38 ± 6 per field of view, although this did not reach significance. In contrast, both IGF-I and LR3 significantly increased myoblast proliferation, by 56 ± 19% (P = 0.015) and 98 ± 28% (P = 0.002), respectively, versus day 7 control (Fig. 3B (http://jp.physoc.org/content/568/1/229.full#F3)). However, they were ineffective at enhancing fibroblast growth (data not shown). In addition, the average number of nuclei per myotube (a positive indicator of myotube size) was also significantly increased by IGF-I, 10.2 ± 0.6 (P < 0.001), and LR3, 11.1 ± 0.9 (P < 0.001), versus day 7 controls, 5.2 ± 0.3 (Fig. 3B (http://jp.physoc.org/content/568/1/229.full#F3)). Conversely, TNFα significantly reduced the number of nuclei per myotube to 2.6 ± 0.6 (P = 0.009), indicating fewer myoblasts involved in differentiation, while IGFBP-3 was without effect (Fig. 3B (http://jp.physoc.org/content/568/1/229.full#F3)). Having investigated differences in skeletal muscle hyperplasia and hypertrophy, we further analysed the percentage of myotubes formed. The mean percentage myotube formation at day 7 (67 ± 4%) was significantly increased above day 0 controls (1.0 ± 0.4%), P < 0.001, but could not be altered with any of IGF-I, LR3 or IGFBP-3, suggesting that the IGFs play a prominent role in myotube size by increasing the number of cells incorporated into individual myotubes but do not impact on the number of myotubes formed. As expected, the percentage of myoblasts forming myotubes following TNFα exposure was significantly reduced (15 ± 5%, P = 0.004) versus control at day 7.
Figure 3.
Effect of IGF-I, LR3 IGF-I, IGFBP-3 and TNFα on proliferation and differentiation
Skeletal myoblasts were differentiated for 7 days and either lysed for protein determination or stained for desmin and analysed using Image Pro plus as described in Methods. A, protein levels were significantly increased at day 7 differentiation and further enhanced by 50 ng ml−1 IGF-I, 50 ng ml−1 LR3 IGF-I and 20 ng ml−1 TNFα (P < 0.001). IGFBP-3 at 150 ng ml−1 was without effect. B, both IGF-I (P = 0.015) and LR3 IGF-I, (P = 0.002) significantly increased myoblast proliferation (filled bars) and the number of cells forming each myotube above untreated controls (open bars) (P < 0.001). Conversely, TNFα significantly reduced the number of myoblasts involved in myotube formation (P = 0.009). C, the mean percentage change in CK activity versus day 7 controls demonstrates that both IGF-I and LR3 IGF-I rescue the inhibition of differentiation by TNFα. The analytical data ± s.e.m. were determined from 12 individuals (A and B) and 5 individuals (C). Statistical analysis was performed using Student's paired t test; *P < 0.05, **P < 0.01, ***P < 0.001.
Since differentiation was reduced in the presence of TNFα, we next investigated whether IGF-I or LR3 could rescue this inhibition. The mean percentage change in CK activity versus day 7 controls demonstrates that both IGF-I and LR3 can rescue the inhibition of differentiation by TNFα, by 47 ± 12% (P = 0.018) and 50 ± 12% (P = 0.016), respectively, taking differentiation back to control levels (Fig. 3C (http://jp.physoc.org/content/568/1/229.full#F3)). No significant differences were observed between the different analogues of IGF-I (P = 0.50), suggesting that the endogenous production of IGFBP-3 did not effect the response of exogenous IGF-I, thus implicating an IGFBP-independent mode of action. Addition of IGFBP-3 was without effect on rescuing the inhibition of differentiation by TNFα (15 ± 11%, P = 0.25).
Insulin-like growth factors promote and TNFα inhibits IGFBP-3 production and secretion
We have previously reported altered regulation of IGFBP-3 secretion in adults (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)), with evidence of IGF-I increasing and TNFα decreasing this process with parallel changes seen in CK activity. To this end we investigated both production and secretion of this protein in response to IGF and TNFα (alone or in combination) in the children's skeletal muscle cultures. Both IGF-I and LR3 increased IGFBP-3 levels within the cells after 7 days in LSM, by 38 ± 3% (P < 0.001) and 62 ± 6% (P < 0.001), respectively, while TNFα demonstrated an inhibition by 53 ± 3% (P < 0.001) (Fig. 4A (http://jp.physoc.org/content/568/1/229.full#F4)). As for CK analysis, the inhibition of IGFBP-3 production by TNFα was rescued with IGF-I and LR3, by 84 ± 18% (P = 0.009) and 141 ± 33% (P = 0.017), respectively, although there were no significant differences between either IGF-I analogue (P = 0.08). It is interesting to note that while exogenous IGFBP-3 was still detectable after 7 days in culture (indicating either surface binding or internalization) and able to enhance endogenous IGFBP-3 production by 67 ± 6% (P < 0.001), it was unable to rescue the inhibition of IGFBP-3 production by TNFα (Fig. 4A (http://jp.physoc.org/content/568/1/229.full#F4)). Day 7 conditioned media were also analysed to determine IGFBP-3 secretion by Western immunoblotting (Fig. 4B (http://jp.physoc.org/content/568/1/229.full#F4)) and RIA (Fig. 4C (http://jp.physoc.org/content/568/1/229.full#F4)). IGF-I and LR3 increased IGFBP-3 secretion by 36 ± 5% (P < 0.001) and 14 ± 5% (P = 0.016), respectively, although IGF-I was more effective (P = 0.002), possibly in part by increasing the stability of endogenous IGFBP-3 (Fig. 4B and C (http://jp.physoc.org/content/568/1/229.full#F4)). The addition of IGFBP-3 at day 0 could still be detected after 7 days in LSM but showed evidence of proteolytic degradation and was without significant effect on further IGFBP-3 secretion (Fig. 4B and C (http://jp.physoc.org/content/568/1/229.full#F4)) despite the elevated production. TNFα, however, significantly reduced the secretion of intact and fragmented IGFBP-3, by 49 ± 3% (P < 0.001) and it was rescued by a coincubation with either IGF-I or LR3, by 53 ± 13% (P = 0.03) and 38 ± 6% (P = 0.001), respectively.
Figure 4.
Insulin-like growth factors promote and TNFα inhibits IGFBP-3 production and secretion
Differentiation was manipulated with a single dose of 50 ng ml−1 IGF-I, 50 ng ml−1 LR3 IGF-I and 150 ng ml−1 IGFBP-3, or with and without 20 ng ml−1 TNFα. IGFBP-3 levels were determined at day 7 of differentiation. A, representative Western immunoblot of endogenous IGFBP-3 production from 50 μg of whole cell lysates (upper panel) with NHS control and IGFBP-3 spike representing the addition of nonglycosylated protein. The percentage change in IGFBP-3 production versus day 7 controls was calculated by densitometry from 5 individuals (lower panel). B, representative Western immunoblot of day 7 conditioned media, demonstrating IGFBP-3 secretion. C, the secretion of IGFBP-3 was confirmed by RIA. The data represent the mean IGFBP-3 levels ± s.e.m. from 5 individuals. Statistical analysis was performed using Student's paired t test; *P < 0.05, **P < 0.01, ***P < 0.001. NHS, normal human serum.
Child versus adult skeletal muscle myoblast differentiation
To observe if there were any differences in growth factor and cytokine manipulated differentiation between children and adults we differentiated four parallel cultures from each group (children: (2 male/2 female) median (range) age was 4.5 (1.9–8.0) years; adults: (2 male/2 female) median (range) age was 37 (35.5–40.5) years). Children's myoblasts demonstrated a higher basal level of CK activity, 917 ± 21 mU mg−1 ml−1, compared to adult cultures, 152 ± 31 mU mg−1 ml−1 (P < 0.001), and this remained true following 7 days of differentiation, 2270 ± 156 mU mg−1 ml−1 versus 551 ± 82 mU mg−1 ml−1 (P = 0.004) (Fig. 5 (http://jp.physoc.org/content/568/1/229.full#F5)). The IGFs were able to increase differentiation in both groups, but, to our surprise, IGFBP-3 had different effects. While no impact was evident on differentiation in the children's cells, an inhibitory response in the adult cultures was seen when compared to day 7 controls, 366 ± 60 mU mg−1 ml−1 versus 551 ± 82 mU mg−1 ml−1 (P = 0.04) (Fig. 5 (http://jp.physoc.org/content/568/1/229.full#F5)), which has been reported previously (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)). TNFα inhibited differentiation in both groups. When the data were expressed as the percentage change in CK activity at day 7 versus day 0, an interesting pattern emerged, with adult myoblasts demonstrating a greater increase in both control and growth factor-mediated differentiation and a greater decrease following TNFα exposure (Table 1 (http://jp.physoc.org/content/568/1/229.full#T1)). As this may well have arisen from lower levels of CK activity at day 0 (for the IGFs at least), we further analysed the percentage differences from day 7 controls. Interestingly, IGF-I also enhanced differentiation to a greater extent in adult cultures, 157 ± 10%, versus children, 135 ± 6% (P = 0.033), but no differences between the two culture models were observed with LR3, 134 ± 7% versus 127 ± 7%, respectively (P = 0.45). The effect of IGFBP-3 on adult myoblasts was further demonstrated, revealing a 34 ± 7% inhibition of differentiation; no effect was seen in children as previously described. TNFα inhibited differentiation in both cell culture models, but was more effective in adults, −87 ± 4%, versus children, −41 ± 13% (P = 0.040).
Differentiation capacity in child versus adult skeletal muscle
Differentiation was manipulated with a single dose of 50 ng ml−1 IGF-I, 50 ng ml−1 LR3, 150 ng ml−1 IGFBP-3 and 20 ng ml−1 TNFα in both child (filled bars) and adult (open bars) cultures. CK activity was determined at day 0 and 7. The data represent the mean CK activity in triplicate ± s.e.m. from 4 individuals for each group run in parallel. Statistical analysis was performed using Student's paired t test; *P < 0.05, **P < 0.01, ***P < 0.001.
Table 1. Percentage increase in CK activity versus day 0 controls
To elucidate why IGF-I was more effective at enhancing differentiation in adults, we examined the endogenous secretion of IGFBP-3 by RIA, hypothesizing that an increase in IGFBP-3 may sequester exogenous IGF-I and limit its capacity to enhance differentiation. Untreated day 7 controls demonstrated higher IGFBP-3 levels in children by 62 ± 26% (P = 0.013). Interestingly, the effect of IGF-I on increased IGFBP-3 secretion above day 7 controls was comparable in both children, 36 ± 5%, and adults, 42 ± 7%, as was LR3, although to a lesser degree, 14 ± 5% versus 19 ± 6%, suggesting conferred stability of endogenous IGFBP-3 by IGF-I. It is interesting to note that while TNFα was more effective at inhibiting differentiation in adults, its effect on inhibiting IGFBP-3 secretion was comparable in both children, −49 ± 3%, and adults, −48 ± 3%.
The influence of TNFα on glucose transport
Having established differences in the impact of TNFα to inhibit differentiation between children and adults, we next investigated whether differences existed in glucose uptake following exposure to this cytokine, which is known to be implicated in the aetiology of insulin resistance and type 2 diabetes (Hotamisligil & Spiegelman, 1994 (http://jp.physoc.org/content/568/1/229.full#ref-24)), by exerting its effects on insulin signalling (Hotamisligil et al. 1994 (http://jp.physoc.org/content/568/1/229.full#ref-23)), glucose metabolism (Borst et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-3)) or both. Differentiated skeletal muscle myotubes were treated for 24 h with 20 ng ml−1 TNFα, washed, and glucose transport assessed following a 10 min incubation with 2-deoxy-[3H]d-glucose. TNFα stimulated basal glucose transport versus untreated controls in both child, 68 ± 17% (P = 0.003) and adult myotube cultures, 149 ± 30% (P = 0.03) (Fig. 6A (http://jp.physoc.org/content/568/1/229.full#F6)). Adult myotubes were more responsive (P = 0.050).
Effect of TNFα on glucose transport and GLUT1/4 expression
Following differentiation, cells were treated with 20 ng ml−1 TNFα for 24 h and glucose transport determined using 2-deoxy-d-glucose. A, glucose transport in control and TNFα-treated cultures of child and adult skeletal muscle. B, GLUT1 expression in control and TNFα-treated cells. Representative micrographs are shown. Scale bar = 50 μm. C, expression of GLUT1 significantly increases in membrane fractions following TNFα exposure versus untreated controls, while no effect was seen on GLUT4 abundance. The data in A represent the mean 2-deoxy-d-glucose transport in triplicate ± s.e.m. from 4 parallel cultures of child and adult skeletal muscle. The data in C were obtained from densitometric analysis from 3 separate children's biopsies. Statistical analysis was performed using Student's paired t test; *P < 0.05.
GLUT1 and GLUT4 abundance in children's skeletal muscle
To evaluate if the TNFα-induced changes in basal glucose uptake resulted from changes in the expression of GLUT1 (a glucose transporter protein responsible for basal glucose uptake; Ebeling et al. 1998 (http://jp.physoc.org/content/568/1/229.full#ref-11)), we analysed the cultures by immunocytochemistry and by isolating crude membrane fractions from whole cell lysates. Using antibodies to GLUT1, we reveal that exposure to TNFα for 24 h caused an up-regulation in the expression of this protein (Fig. 6B (http://jp.physoc.org/content/568/1/229.full#F6)). This was confirmed by immunoblotting of crude membrane fractions from whole cell lysates as demonstrated in (Fig. 6C (http://jp.physoc.org/content/568/1/229.full#F6)), with a significant increase in membrane GLUT1 abundance (P = 0.019) over untreated controls. In addition to GLUT1 abundance, we examined if changes were present in the insulin-stimulated glucose transporter protein, GLUT4. There was no change in membrane levels in response to TNFα (Fig. 6C (http://jp.physoc.org/content/568/1/229.full#F6)).
Effect of p38MAPK and nitric oxide synthase inhibitors on TNFα induced glucose transport in children's myotubes
The actions of TNFα have been shown to occur through p38MAPK (Ho et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-22)) and also by alterations in the nitric oxide synthase (NOS) pathways (Bedard et al. 1997 (http://jp.physoc.org/content/568/1/229.full#ref-2)). Indeed, we have observed enhanced p38MAPK activation in the presence of TNFα in our model (data not shown), suggesting that it plays an important role in skeletal muscle maintenance and metabolism. We therefore utilized a broad range p38MAPK inhibitor (SB202190) and NOS inhibitor (l-NAME) to investigate if inhibition of these pathways could alter the changes observed in glucose transport following exposure to TNFα. To avoid the inherent problems of inhibitor degradation over time, we performed the experiments acutely over 90 min, with a 30 min preincubation of 1 μm SB202190 or 1 μm l-NAME followed by a 60 min exposure with 20 ng ml−1 TNFα. As for a 24 h exposure, TNFα demonstrated an increase in basal glucose transport by 58 ± 11% over controls (P = 0.002) (Fig. 7 (http://jp.physoc.org/content/568/1/229.full#F7)). The addition of SB202190 and l-NAME reduced the TNFα-induced increases to 37 ± 11% (P = 0.001) and 28 ± 5% (P = 0.05), respectively (Fig. 7 (http://jp.physoc.org/content/568/1/229.full#F7)). These data suggest that TNFα may indeed be acting via both p38MAPK and NOS pathways to increase glucose transport.
p38MAPK and NOS inhibitors reduce the effect of TNFα on glucose transport in children's myotubes
Differentiated cells were treated with 20 ng ml−1 TNFα for 60 min either with or without a predose of 1 μm SB202190 or 1 μml-NAME for 30 min and glucose transport determined. The data represent the mean 2-deoxy-d-glucose transport in triplicate ± s.e.m. from 6 individuals. Statistical analysis was performed using Student's paired t test; *P < 0.05, **P < 0.01.
Previous Section (http://jp.physoc.org/content/568/1/229.full#sec-13)Next Section (http://jp.physoc.org/content/568/1/229.full#sec-22)
Discussion
We have utilized skeletal muscle biopsies from prepubertal children, to establish a unique primary cell culture model. The myoblasts were characterized and the actions of IGF-I and TNFα on differentiation, IGFBP-3 production/secretion and glucose metabolism examined. The model was also compared with primary adult skeletal muscle cultures. Most studies of skeletal muscle cells to date have focused on animal cell lines (Roeder et al. 1988 (http://jp.physoc.org/content/568/1/229.full#ref-35)) or adult myoblast cultures (Crown et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-9)). None have focused on a childhood population where differences may exist, prior to augmentation of the growth hormone–insulin-like growth factor axis and effects of sex steroids during or following puberty.
The isolation, characterization and culture of children's skeletal muscle cells was successful in all biopsies collected. Cultures contained an average of 57 ± 3% myoblasts, higher than we previously reported in adult cultures (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)). This may reflect a greater number of satellite cells in the original biopsies, which are thought to negatively correlate with age (Renault et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-34); Foulstone et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-18)). All cultures successfully differentiated into multinucleated myotubes following 7 days in LSM. Differentiation was associated with an increase in protein content without a change in myoblast or fibroblast number, indicating that myotube formation is accompanied by active protein deposition. Basal and differentiated CK activity were higher in the children's cultures and may have been influenced by the greater number of myoblasts at day 0, as CK activity at day 7 correlates positively with initial myoblast number (Foulstone et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-18)). IGFBP-3, the predominant binding protein produced by human skeletal muscle cultures, accumulated in the medium as the cells differentiated. Interestingly, children's myoblasts secrete higher levels of IGFBP-3 than adult cultures (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)). Since IGFBP-3 has been shown to have IGF-independent effects such as growth inhibition (Valentinis et al. 1995 (http://jp.physoc.org/content/568/1/229.full#ref-48); Yamanaka et al. 1999 (http://jp.physoc.org/content/568/1/229.full#ref-50)), we postulate that local levels may vary throughout life depending on the requirement for cellular growth. Indeed, we previously demonstrated that fetal myoblasts secrete lower levels of IGFBP-3 than adult myoblasts (Crown et al. 2000 (http://jp.physoc.org/content/568/1/229.full#ref-9)), which may influence their proliferation capacity and increase the overall stem cell population. IGFBP-3 then rises before puberty to limit satellite cell proliferation, while avoiding depletion of the myoblast stem cell pool, then lowers during adulthood allowing skeletal muscle maintenance and repair. Analysis of IGF-I and II secretion in our primary myoblast and myotube cultures demonstrated undetectable amounts of IGF-I (similar to adults), but low levels of IGF-II in media from all cultures tested, as has been shown in C2 muscle cells (Tollefsen et al. 1989 (http://jp.physoc.org/content/568/1/229.full#ref-45); Rosen et al. 1993 (http://jp.physoc.org/content/568/1/229.full#ref-37)).
The impact of IGF-I, IGFBP-3 and TNFα on skeletal muscle differentiation was also examined. Both IGF-I and its analogue LR3 elicited comparable increases in myotube hypertrophy, protein content, and CK activity over controls, with the greatest difference observed at day 3. The augmented differentiation by IGF-I was concomitant with an increase in myoblast hyperplasia and myoblast incorporation into each myotube fibre; however, the percentage of myotubes formed was unaffected. We suggest that IGF-I is implicated in enhanced myoblast fusion (facilitated by an increase in myoblast number), which may in turn drive a more rapid differentiation process, in part by up-regulating myogenin protein expression (Quinn et al. 1993 (http://jp.physoc.org/content/568/1/229.full#ref-33)). These findings are supported in animal (Mitchell et al. 2002 (http://jp.physoc.org/content/568/1/229.full#ref-30)) and adult human investigations (Foulstone et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-18)). That IGF-I does not increase the percentage of myotubes formed is highly relevant to the in vivo situation where myofibre number does not change, and muscle size increases through a process of hypertrophy, analogous to what is witnessed here.
The equivalent actions of exogenous IGF-I and LR3 on differentiation suggest that the large amounts of endogenously secreted IGFBP-3 did not affect the responses observed. The IGFs have been shown to increase the levels of IGFBPs in conditioned media of several cell types (as reviewed by Firth & Baxter, 2002 (http://jp.physoc.org/content/568/1/229.full#ref-14)), in part by increasing transcription and stabilization of mRNA and subsequent stabilization of the secreted proteins (Conover, 1991 (http://jp.physoc.org/content/568/1/229.full#ref-7)). IGF-I and LR3 both increased IGFBP-3 levels in a differential manner. LR3 altered IGFBP-3 production and IGF-I impacted on IGFBP-3 secretion. The different affinities of these IGFs for IGFBPs may play a role, allowing LR3 (with insignificant binding affinity toward the IGFBPs) to bind to IGF-IR and initiate its response, while exogenous IGF-I may associate with and stabilize secreted IGFBP-3, which further account for the elevated levels seen in conditioned media. Interestingly, the ability of IGF-I to increase CK activity was more pronounced in adult than children's cultures, while the effect of LR3 was equivalent in both cell models. These data provide us with an interesting model to investigate the different roles that IGFBP-3 may play in skeletal muscle differentiation in childhood and adulthood. IGFBP-3 did not appear to elicit an effect on basal or IGF-induced differentiation in children, but clearly is playing a manipulatory role in adults. Proteolytic status may be important.
As expected, TNFα inhibited morphological and biochemical differentiation along with a reduction in endogenous IGFBP-3 secretion in both culture models, although adult myoblasts were more susceptible. We suggest that differences in TNFα Receptor-1 expression or its proteolytically cleaved soluble form may be involved, since alterations in these can either block or prolong the bioavailability of TNFα (Aderka, 1996 (http://jp.physoc.org/content/568/1/229.full#ref-1)). Alternatively, TNFα may directly alter myoblast fusion, as fewer nuclei were incorporated into myotubes. These marked decreases in differentiation were paralleled with an increase in cellular proliferation of fibroblasts but not myoblasts. The differential effects on growth of these cells remain to be elucidated. That TNFα did not stimulate or inhibit myoblast proliferation suggests that inhibition of myotube formation was not a result of myoblast growth or loss. Interestingly, a co-incubation of IGF-I or LR3 but not IGFBP-3 with TNFα, rescued the inhibition of differentiation, implicating different mechanisms, other than increased production or conferred stability of IGFBP-3 in inhibiting or promoting differentiation in children. In addition, we have shown that children's skeletal muscle not only secretes higher levels of endogenous IGFBP-3 than adult cells, but also respond differently to exogenous levels of this protein, with no evidence of the inhibitory effects, as seen in the differentiation of adult myoblasts (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)). We suggest that differences in proteolytic cleavage of IGFBP-3 may be important for different responses, since previous findings by our group have demonstrated predominantly intact IGFBP-3 present in adults (Foulstone et al. 2003 (http://jp.physoc.org/content/568/1/229.full#ref-19)), while these data from children demonstrate an increase in proteolytically cleaved IGFBP-3. The proteases present are most likely secreted by the cells since cleaved IGFBP-3 was only seen in conditioned media and not in lysates. More importantly, the proteolysis observed at the cellular level may be an important mechanism for regulating IGF–IGFBP interaction and local IGF tissue availability as has been demonstrated in prostate (Conover et al. 1995 (http://jp.physoc.org/content/568/1/229.full#ref-8)) and breast cancer models (Salahifar et al. 1997 (http://jp.physoc.org/content/568/1/229.full#ref-39)).
Further to investigating the effects of TNFα on the parameters of differentiation, we investigated its effects on basal glucose transport, since there are conflicting reports on the effects of TNFα on glucose metabolism in skeletal muscle, with most if not all utilizing cells of either animal or adult human origin. Using parallel cultures of differentiated adult and children's skeletal muscle, we have shown similar findings to those using in vivo models (Evans et al. 1989 (http://jp.physoc.org/content/568/1/229.full#ref-13)), whereby TNFα causes a stimulatory effect on basal glucose transport. TNFα had a greater effect in cells derived from adults, possibly resulting from differences in TNFR1 expression as previously suggested. To elucidate the mechanisms of TNFα-induced glucose transport in children we initially investigated the expression of GLUT1, reported to be up-regulated by TNFα in both adipocytes (Hauner et al. 1995 (http://jp.physoc.org/content/568/1/229.full#ref-21)) and skeletal muscle (Ciaraldi et al. 1998 (http://jp.physoc.org/content/568/1/229.full#ref-5)). Indeed, our findings are similar in this respect. The actions of TNFα on metabolism may also occur through a nitric oxide-dependent mechanism, since the NOS inhibitor l-NAME down-regulated the TNFα-induced glucose transport, which has been previously shown in rat skeletal muscle (Bedard et al. 1997 (http://jp.physoc.org/content/568/1/229.full#ref-2)). Whether this results from a decrease in the expression of the GLUT1 transporter protein in our model remains to be elucidated. We have also shown that TNFα may elicit its responses through p38MAPK, as have others (Ho et al. 2004 (http://jp.physoc.org/content/568/1/229.full#ref-22)). While TNFα appears to be beneficial in the short term, evidence suggests that prolonged exposure can lead to skeletal muscle and whole body insulin resistance (Lang et al. 1992 (http://jp.physoc.org/content/568/1/229.full#ref-26); Storz et al. 1999 (http://jp.physoc.org/content/568/1/229.full#ref-44)), possibly via indirect mechanisms such as elevation of free fatty acids in the circulation (Sethi & Hotamisligil, 1999 (http://jp.physoc.org/content/568/1/229.full#ref-41)).
In summary, we have shown that the IGF system as well as TNFα plays a pivotal role in the regulation of skeletal muscle maintenance and metabolism, with differences observed between children and adult cultures. Alterations in any of these processes could influence the equilibrium of skeletal muscle maintenance and loss associated with many chronic severe illnesses such as cancer cachexia or alternatively in patients with obesity-related type 2 diabetes. Our findings indicate that future studies must use age-specific cell models to appropriately investigate the mechanisms of muscle pathophysiology in both normal and diseased individuals.