Tumor Necrosis Factor- Decreases Insulin-Like Growth Factor-I ...

47 downloads 103 Views 433KB Size Report
skeletal muscle was independent of changes in two house- keeping genes [L32 and ..... Colson A, Le Cam A, Maiter D, Edery M, Thissen JP 2000 Potentiation of.
0013-7227/03/$15.00/0 Printed in U.S.A.

Endocrinology 144(5):1770 –1779 Copyright © 2003 by The Endocrine Society doi: 10.1210/en.2002-220808

Tumor Necrosis Factor-␣ Decreases Insulin-Like Growth Factor-I Messenger Ribonucleic Acid Expression in C2C12 Myoblasts via a Jun N-Terminal Kinase Pathway ROBERT A. FROST, GERALD J. NYSTROM,

AND

CHARLES H. LANG

Department of Cellular and Molecular Physiology, Pennsylvania State University, College of Medicine, Hershey, Pennsylvania 17033 IGF-I is a major anabolic hormone for skeletal muscle in vivo. Yet the mechanisms by which GH and cytokines regulate IGF-I expression remain obscure. Lipopolysaccharide (LPS) dramatically alters the circulating concentration of both TNF␣ and IGF-I, and TNF␣ in part mediates the cachectic activity of LPS. Little is known about the local synthesis of IGF-I and TNF␣ in skeletal muscle per se. The purpose of the present study was to determine whether LPS alters the expression of TNF␣ and IGF-I in mouse skeletal muscle and whether TNF␣ directly inhibits IGF-I mRNA expression in C2C12 myoblasts. Intraperitoneal injection of LPS in C3H/SnJ mice increased the expression of TNF␣ protein in plasma (16fold) and TNF␣ mRNA in skeletal muscle (8-fold). LPS also decreased the plasma concentration of IGF-I (30%) and IGF-I mRNA in skeletal muscle (50%, between 6 and 18 h after LPS administration). Addition of LPS or TNF␣ directly to C2C12 myoblasts decreased IGF-I mRNA by 50 – 80%. The TNF␣induced decrease in IGF-I mRNA was both dose and time dependent and occurred in both myoblasts and differentiated myotubes. TNF␣ selectively decreased IGF-I but not IGF-II

I

NFECTION, TRAUMA, cancer cachexia, AIDS, and chronic alcohol abuse are often associated with severe muscle wasting that cannot be solely explained by a reduction in caloric intake (1). Potential mediators of the wasting syndromes include inflammatory cytokines such as TNF␣, IL-1␤, and IL-6 (2) as well as negative regulators of muscle mass such as myostatin (3, 4) and glucocorticoids (5). Administration of lipopolysaccharide (LPS) or proinflammatory cytokines to rats mimics the protein hypercatabolic aspects of cachexia. Conversely, prophylactic administration of TNF␣-binding protein (TNFBP) (6) and IL-1␤ receptor antagonist (7) partially reverses some of the effects of endogenous cytokines released in response to a bacterial infection. Cytokines such as TNF␣ may have a direct effect on skeletal muscle protein metabolism and/or alter the expression and biological activity of anabolic hormones such as GH and IGF-I (8, 9). LPS and TNF␣ stimulate the hepatic synthesis of suppressors of cytokine signaling, and this is thought to be Abbreviations: 1400W, N-3-Aminomethyl benzylacetamidine; DRB, 5,6-dichloro-␤-d-ribofuranosyl-benzimidazole; GAPDH, glyceraldehyde3-phosphate dehydrogenase; IGFBP, IGF-binding protein; JNK, Jun N-terminal kinase; L-NMMA, NG-monomethyl l-arginine; LPS, lipopolysaccharide; MEK, MAPK kinase; NF␬B, nuclear factor ␬B; NO, nitric oxide; NOS, NO synthase; NOS2, inducible form of NOS; PDTC, pyrrolidinedithiocarbamate; RPA, ribonuclease protection assay; SNP, sodium nitroprusside; STAT, signal transducer and activator of transcription; TNFBP, TNF␣-binding protein.

mRNA levels, and the effect of TNF␣ was blocked by a specific TNF-binding protein. TNF␣ did not alter IGF-I mRNA levels in the presence of the protein synthesis inhibitor cycloheximide. TNF␣ did not change the half-life of IGF-I mRNA. TNF␣ completely prevented GH-inducible IGF-I mRNA expression, but this GH resistance was not attributable to impairment in signal transducer and activator of transcription-3 or -5 phosphorylation. TNF␣ increased both nitric oxide synthase-II mRNA and protein, and the nitric oxide donor sodium nitroprusside decreased IGF-I mRNA levels in C2C12 cells. Yet inhibitor studies indicate that nitric oxide did not mediate the effect of TNF␣ on IGF-I mRNA expression. TNF␣ stimulated the phosphorylation of c-Jun and specific inhibition of the Jun N-terminal kinase pathway, but not other MAPK pathways, completely prevented the TNF␣-induced drop in IGF-I mRNA. These data suggest that LPS stimulates TNF␣ expression in mouse skeletal muscle and autocrine-derived cytokines may contribute to the reduced expression of IGF-I in this tissue. (Endocrinology 144: 1770 –1779, 2003)

important for the development of GH resistance at the level of the liver (10, 11). As a consequence, the ability of GH to stimulate hepatic IGF-I synthesis and increase blood-borne IGF-I is greatly attenuated (12). Infusion of TNF␣ in rats lowers both the plasma concentration of IGF-I and expression of IGF-I mRNA in the gastrocnemius muscle (13). Because IGF-I is a major anabolic hormone for skeletal muscle, a local deficit in IGF-I may also be partially responsible for the muscle wasting that occurs in sepsis, trauma, and other catabolic conditions (14, 15). It is not known whether TNF␣ generates GH resistance in skeletal muscle or acts directly on skeletal muscle to alter the local expression of IGF-I. TNF␣ may alter the endocrine expression of other cytokines or signaling molecules that in turn suppress the expression of IGF-I in muscle. Investigators have had varied success in observing skeletal muscle responses to TNF␣ in intact animals, epitrochlears incubated in vitro, and muscle cell lines. Yet a constant infusion of TNF␣ decreases basal muscle protein synthesis in rats (16). TNF␣ also decreases protein content in C2C12 myotubes (17), and we have shown that TNF␣ decreases protein synthesis in human skeletal muscle cells (18). In the present study, we examined whether TNF␣ is expressed in mouse skeletal muscle in response to an LPS challenge and whether this increase in TNF␣ is associated with a subsequent drop in IGF-I mRNA expression. Addi-

1770

Frost et al. • Regulation of Muscle IGF-I by TNF␣

Endocrinology, May 2003, 144(5):1770 –1779 1771

tional experiments were performed in C2C12 myoblasts to examine potential mechanisms by which TNF␣ decreases IGF-I mRNA expression in muscle. This includes the effect of TNF␣ on IGF-I mRNA half-life and the ability of GH to stimulate IGF-I mRNA expression. Finally, potential signal transduction pathways by which TNF␣ may inhibit IGF-I mRNA expression were investigated including interference with the phosphorylation of signal transducer and activator of transcription (STAT) factors, generation of nitric oxide (NO), and activation of MAPK. Materials and Methods Experimental protocol for C3H/HeSnJ mice C3H/HeSnJ mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All mice were housed in a controlled environment and provided water and rodent chow ad libitum for 3 wk before their use. At the time of the study, mice were 8 –9 wk old and weighed 21.4 ⫾ 0.3 g. In the experiment depicted in Fig. 1, C3H/HeSnJ mice were injected ip with LPS derived from Escherichia coli 026:B6 (25 ␮g per mouse, Difco Laboratories, Detroit, MI) or an equal volume of saline (250 ␮l/mouse). This dose was based on a preliminary dose-response study and is similar to that used by other investigators (19). After 2 h (for TNF␣ mRNA) or 2, 6, and 18 h (for IGF-I mRNA), mice were anesthetized with a mixture of ketamine (Fort Dodge Animal Health, Fort Dodge, IA) and xylazine (Bayer Corp., Shawnee Mission, KS) at 90 and 9 mg/kg, respectively. Blood was collected from the inferior vena cava in heparinized syringes. Hind limb skeletal muscle from both legs was dissected from each animal, wrapped in aluminum foil, and flash-frozen in liquid nitrogen. Mice were killed by cardiac excision and subsequent exsanguination. Tissues were later powdered under liquid nitrogen using a mortar and pestle and stored at ⫺70 C. All experiments were approved by the Institutional Animal Care and Use Committee at the Pennsylvania State University College of Medicine and adhere to the National Institutes of Health Guidelines for the Use of Experimental Animals.

Cell culture The C2C12 mouse myoblast cell line was purchased from the American Type Culture Collection (Manassas, VA) and used for all studies. Cells were grown in 100-mm Petri dishes (Becton Dickinson and Co., Franklin Lakes, NJ) and cultured in MEM containing 10% bovine calf serum, penicillin (100 U/ml), streptomycin (100 ␮g/ml), and Amphotericin B (25 ␮g/ml; all from Sigma, St. Louis, MO). Cells were grown to confluence and switched to fresh serum-containing media before addition of LPS, cytokines, or other agents. Cells for STAT-3 and -5 phosphorylation studies were grown in serum-free media as previously described (20). Most studies used C2C12 cells at the myoblast stage, but selected experiments were repeated using differentiated myotubes. Experiments were performed with LPS B derived from E. coli 026:B6 (Difco). A variety of compounds were used to characterize the response to TNF including TNFBP (Amgen, Inc., Boulder, CO), cycloheximide, 5,6-dichloro-␤-d-ribofuranosyl-benzimidazole (DRB), sodium nitroprusside (SNP), glyco-SNAP (acetyl-S-nitroso-d,l-penicillaminamide), NG-monomethyl l-arginine (L-NMMA), N-3-aminomethyl benzylacetamidine (1400W), pyrrolidinedithiocarbamate (PDTC), PD98059, SB202190, and SP600125 (all from Calbiochem, La Jolla, CA). Additional experiments used the recombinant cytokines IL-1␤ and TNF␣ (Peprotech, Rocky Hill, NJ). Cell extracts were prepared in SDS-PAGE sample buffer. Urea was added to a final concentration of 3 m to improve the solubility of contractile proteins when myoblast and myotube cell extracts were compared.

RNA isolation and ribonuclease protection assay (RPA) Total RNA, DNA, and protein were extracted from C2C12 cells or tissues in a mixture of phenol and guanidine thiocyanate (TRI reagent, Molecular Research Center, Inc., Cincinnati, OH) using the manufacturer’s protocol. RNA was separated from protein and DNA by the addition of bromocholoropropane and precipitation in isopropanol. Af-

FIG. 1. Effect of LPS on circulating TNF␣ and TNF␣ mRNA abundance in mouse skeletal muscle. Wild-type (C3H/HeSnJ) mice weighing 21.4 ⫾ 0.3 g were injected ip with either saline (Sal) or a nonlethal dose of E. coli LPS (25 ␮g/mouse). Blood and tissue samples were collected after 2 h. Muscle was flash-frozen in liquid nitrogen, powdered, and analyzed at the peak of cytokine expression (2 h). RNA was isolated and hybridized to a cytokine mRNA RPA template as described in Materials and Methods and run on a 5% acrylamide gel. The dried gel was exposed to a PhosphorImager screen and quantified with ImageQuant software. All data are normalized to L32 mRNA as described in Materials and Methods and expressed as a fold increase relative to animals injected with saline alone. We measured plasma TNF␣ by ELISA (A). An image of TNF␣, L32, and GAPDH mRNA in skeletal muscle is shown (B). A PhosphorImager analysis of TNF␣ mRNA in the dried gel was quantified and is plotted (B). Values are means ⫾ SEM. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05). ter a 75% ethanol wash and resuspension in formamide, RNA samples were quantified by spectrophotometry. Ten micrograms of RNA were used for each assay. Riboprobes were synthesized from a custom multiprobe mouse template set containing probes for both TNF␣ and IGF-I mRNA detection (PharMingen, San Diego, CA). The labeled riboprobe was hybridized with RNA overnight using an RPA kit and the manufacturer’s protocol (PharMingen). Protected RNAs were separated using a 5% acrylamide gel (19:1 acrylamide/bisacrylamide). Gels were transferred to blotting paper and dried under vacuum on a gel dryer. Dried gels were exposed to a PhosphorImager screen (Molecular Dynamics, Inc., Sunnyvale, CA), and the resulting data were quantified using ImageQuant software and normalized to the mouse ribosomal protein L32 mRNA signal in each lane. Some RNA samples were also electrophoresed under denaturing conditions on a 1.1% agarose gel containing 6% formaldehyde. RNA was transferred to a Nytran Supercharge membrane (Schleicher & Schuell,

1772

Endocrinology, May 2003, 144(5):1770 –1779

Inc., Keene, NH) using the Turbo Blotting System. After baking, blots were hybridized at 42 C in ULTRAhyb (Ambion, Inc., Austin, TX). The membrane was probed with a 325-bp rat IGF-I cDNA (Peter Rotwein, Portland, OR) or IGF-II cDNA (21) that was labeled with [␣-32P]deoxyATP (Amersham, Arlington Heights, IL) using a random primed DNA labeling kit (Roche Diagnostics Corp., Indianapolis, IN). For normalization of RNA loading, an oligonucleotide directed against rat 18S RNA was radioactively labeled with [␥-32P]ATP (Amersham) using terminal deoxynucleotidyl transferase. All membranes were washed twice in 2⫻ saline sodium citrate/0.1% sodium dodecyl sulfate at 42 C for 15 min followed by an additional wash at 62 C. Membranes were exposed to a PhosphorImager screen (Molecular Dynamics, Inc.), and the resultant data were quantified using ImageQuant software.

Western blot analysis Cell extracts were electrophoresed on denaturing polyacrylamide gels and electrophoretically transferred to nitrocellulose with a semidry blotter (Bio-Rad Laboratories, Inc., Melville, NY). The resulting blots were blocked with 5% nonfat dry milk for 1.5 h and incubated with antibodies against either total or phosphorylated STAT3 or STAT5 as previously described (20). Additional blots were probed with antibodies to NO synthase (NOS)-II, phosphorylated ERK-1 and -2, p38, Jun N-terminal kinases (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), and sarcomeric myosin (MF-20; provided by D. A. Fishman, National Institute of Child Health and Human Development, Hybridoma Bank, Iowa City, IA). Unbound primary antibody was removed by washing with Tris-buffered saline containing 0.05% Tween-20, and blots were incubated with antirabbit or antimouse immunoglobulin conjugated with horseradish peroxidase. Blots were briefly incubated with the components of an enhanced chemiluminescent detection system (Amersham, Buckinghamshire, UK). Dried blots were used to expose x-ray film for 1–3 min.

Frost et al. • Regulation of Muscle IGF-I by TNF␣

2 illustrates that LPS decreased the plasma concentration of IGF-I by 30% (Fig. 2A) and IGF-I mRNA expression in skeletal muscle 6 h after LPS injection by 50% (Fig. 2B). The drop in skeletal muscle IGF-I mRNA was sustained through 18 h. The change in TNF␣ and IGF-I mRNA in skeletal muscle was independent of changes in two housekeeping genes [L32 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH)] that were used to normalize the steady-state levels of the RNAs. Addition of LPS to C2C12 myoblasts also decreased IGF-I mRNA. A significant decrease was observed after 4 h, and this was sustained through 18 h, similar to that observed in mouse skeletal muscle in vivo (Figs. 2 and 3). Whether RNA was analyzed by RPA or Northern blotting similar results were obtained (Fig. 3, A and B). LPS also transiently increased TNF␣ mRNA in C2C12 cells (Fig. 3C).

IGF-I RIA and TNF␣ ELISA Serum samples were acid ethanol extracted with an additional cryoprecipitation step to remove IGF-binding proteins (IGFBPs). This procedure quantitatively removes IGFBPs from serum. IGF-I was assayed by RIA using IGF-I antibody (National Institute of Diabetes and Digestive and Kidney Diseases) and [125I]IGF-I (Amersham Biosciences, Arlington Heights, IL) as previously described (22). TNF␣ was measured in mouse plasma using mouse-specific anti-TNF␣ antibody pairs (PharMingen).

Statistics Values are means ⫾ sem. Unless otherwise noted, each experimental condition was tested in triplicate, and each experiment was repeated two times. Data were analyzed by ANOVA followed by Student-NewmanKeuls test. Statistical significance was set at P ⬍ 0.05. For animal studies, the number of mice per group was four (control) and six (LPS). IGF-I mRNA half-life was calculated from the slope of the regression line using the formula t1/2 ⫽ 0.5/m, where m is the slope of the line in arbitrary units per hour. Half-lives were compared by t test where t ⫽ [(m1 ⫺ m2)/(公(se12 ⫹ se22)]. Statistical significance was set at P ⬍ 0.05.

Results LPS increases TNF␣ mRNA and decreases IGF-I mRNA in skeletal muscle

Mice were injected with a nonlethal dose of LPS and hindlimb skeletal muscle obtained near the peak of cytokine expression (2 h) and after a period of time that we have previously demonstrated a fall in the plasma concentration of IGF-I (6 h). LPS increased the concentration of TNF␣ in plasma from undetectable levels (⬍0.015 ng/ ml) to 16 ng/ml (Fig. 1A). LPS also increased TNF␣ mRNA in skeletal muscle at this time point (9-fold; Fig. 1B). Figure

FIG. 2. Effect of LPS on IGF-I in plasma and IGF-I mRNA abundance in skeletal muscle. Blood and skeletal muscle (gastrocnemius/plantaris) from saline (Sal) and LPS-injected mice were collected 2, 6, and 18 h after LPS as described in Fig. 1. We measured plasma IGF-I 6 h after LPS injection by RIA (A). IGF-I mRNA in skeletal muscle at the same time point was determined by RPA as described in Materials and Methods. A phosphorimage of the dried gel for IGF-I, L32, and GAPDH mRNA is shown (B). IGF-I mRNA content in skeletal muscle 2, 6, and 18 h after LPS was also quantified (B). All mRNA data were normalized to L32 mRNA as described in Materials and Methods and expressed as a percentage of animals injected with saline. Values are means ⫾ SEM. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05).

Frost et al. • Regulation of Muscle IGF-I by TNF␣

FIG. 3. LPS time dependently decreases IGF-I mRNA abundance in C2C12 myoblasts. C2C12 cells were grown as described in Fig. 2 and treated with 1 ␮g/ml E. coli LPS for 0.33, 0.66, 1, 2, 4, 8, or 18 h. RNA was isolated and hybridized to an IGF-I mRNA RPA template as described in Materials and Methods and run on a 5% acrylamide gel. The dried gel was exposed to a PhosphorImager screen. A representative image of the dried gel is shown (A). The same samples were also run on an agarose gel under denaturing conditions, transferred to a Nytran membrane, and probed with a radioactively labeled rat IGF-I cDNA. A representative phosphorimage of the membrane is shown (B). When normalized to 18S mRNA, the 7.5-kb IGF-I signal was 97, 96, 105, 82, 69, 40, and 2% of control for each of the time points listed above, respectively. Additional cells were treated with LPS for 1 h, RNA isolated and hybridized to an RPA template containing a TNF␣ probe, and analyzed as described above. A representative image of RNA from the saline (Sal)- and LPS-treated cells is shown for TNF␣, L32, and GAPDH (C).

TNF␣ dose and time dependently decreases IGF-I mRNA expression in C2C12 myoblasts

Because TNF␣ mediates many of the effects of LPS, we examined whether this cytokine could directly inhibit IGF-I mRNA expression in C2C12 myoblasts. TNF␣ decreased IGF-I mRNA dose and time dependently. TNF␣ decreased IGF-I mRNA with an ED50 of approximately 10 –15 ng/ml over an 18-h period (Fig. 4A), and the response to TNF␣ was sustained for at least 48 h (Fig. 4B). TNF␣ decreased IGF-I mRNA abundance by 50% in myocytes after incubation for 16 h (Fig. 5A). The TNF␣-induced drop in IGF-I mRNA was completely blocked by a TNFBP, suggesting that the change in IGF-I mRNA is due to TNF receptor binding and not to a potential contaminant in the TNF␣ preparation. The negative effect of TNF␣ on IGF-I mRNA expression was also relatively selective. In contrast to IGF-I mRNA, TNF␣ increased the

Endocrinology, May 2003, 144(5):1770 –1779 1773

FIG. 4. TNF␣ dose and time dependently decreases IGF-I mRNA in C2C12 cells. C2C12 myoblasts were grown as described in Fig. 3 and treated with TNF␣ (0 –100 ng/ml) for 18 h. IGF-I mRNA was isolated and quantified as described above. The dose-response curve for TNF␣ is presented (A). C2C12 cells were also treated with 40 ng/ml TNF␣ and RNA isolated 12, 24, and 48 h later. The quantification of IGF-I mRNA in control and TNF␣-treated cells at these time points is presented for cells treated with saline f or LPS 䡺 (B). All data were normalized for L32 mRNA as described in Materials and Methods. Values are means ⫾ SEM.

level of another mRNA (Mac25, Fig.5B) and had no affect on the level of IGF-II mRNA (Fig. 5C). Response of C2C12 cells to TNF␣ requires ongoing transcription and translation

The ability of C2C12 cells to respond to TNF␣ was examined in the presence of the protein synthesis inhibitor cycloheximide or the transcriptional inhibitor DRB. Cycloheximide completely blocked the ability of TNF␣ to decrease IGF-I mRNA abundance (Fig. 6A). When the transcriptional inhibitor DRB was given to C2C12 cells at the same time as TNF␣, the half-life of IGF-I mRNA was unchanged (Fig. 6B). TNF␣ blunts the ability of GH to stimulate IGF-I mRNA expression but not STAT3 or STAT5 phosphorylation

Previously we have shown that GH stimulates IGF-I mRNA expression in C2C12 myoblasts, and this response is associated with the phosphorylation and nuclear accumulation of both STAT3 and STAT5. It was therefore of interest to determine whether TNF␣ could inhibit GH-stimulated IGF-I mRNA expression. Pretreatment of C2C12 cells with TNF␣ for 30 min inhibited the subsequent ability of GH to stimulate IGF-I mRNA expression (Fig. 7A). In contrast,

1774

Endocrinology, May 2003, 144(5):1770 –1779

Frost et al. • Regulation of Muscle IGF-I by TNF␣

FIG. 6. The TNF␣-induced decrease in IGF-I mRNA requires ongoing protein synthesis and transcription. C2C12 myoblasts were grown as described in Fig. 3 and treated with either TNF␣ alone (40 ng/ml) or TNF␣ and cycloheximide (10 ␮M). RNA was isolated 8 h later. IGF-I mRNA content in these cultures was quantified and is presented (A). Additional cells were treated with either saline or TNF␣ and DRB to examine the half-life of IGF-I mRNA. RNA was isolated 3, 6, 12, and 18 h after the addition of DRB. The mRNA half-life curves for control and TNF␣-treated cells are presented (B). All data are normalized for L32 mRNA as described in Materials and Methods. Values are means ⫾ SEM of triplicate dishes. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05).

FIG. 5. TNF␣ selectively decreases IGF-I mRNA in C2C12 myoblasts. C2C12 myoblasts were grown as described in Materials and Methods and treated with TNF␣ (40 ng/ml) for 16 h. RNA was isolated in Tri-reagent and hybridized to a cytokine mRNA RPA template as described in Materials and Methods and run on a 5% acrylamide gel. A and B, Cells were treated with either TNF␣ alone or TNF and a molar excess of a TNFBP. The level of IGF-I and Mac-25 RNA are presented. C, The level of IGF-II RNA in saline (Sal)-treated cells was compared with cells treated with TNF␣ by Northern blot analysis. IGF-I mRNA was normalized to L32 mRNA as described in Materials and Methods and expressed as a percentage of cells treated with saline alone. Northern blots for Mac-25 and IGF-II were normalized to 18S RNA. Values are means ⫾ SEM. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05).

TNF␣ pretreatment did not alter the GH-stimulated phosphorylation of either STAT3 or STAT5 (Fig. 7B). NO donors inhibit IGF-I mRNA expression, but the TNF␣induced drop in IGF-I is not mediated through an NOSdependent mechanism

TNF␣, IL-1␤, and LPS all are capable of stimulating the expression of the inducible form of NOS and generating NO

in immune cells. In addition, NO and cytokines have been shown to have a negative effect on anabolic processes such as bone matrix formation. NO also decreases protein synthesis by regulating the kinase that phosphorylates the eukaryotic initiation factor eIF2␣. Therefore, we examined the role of NO in regulating IGF-I mRNA expression. The NO donor SNP inhibited IGF-I mRNA expression dose dependently with a 50% inhibition of IGF-I mRNA at 50 ␮m SNP (Fig. 8A). Similar results were obtained with the NO donor glyco-SNAP (data not shown). To determine whether TNF␣ could stimulate NO via up-regulation of the inducible form of NOS (NOS2), we compared the ability of TNF␣ and IL-1␤ to induce NOS2 mRNA. IL-1␤ and TNF␣ maximally stimulated NOS2 mRNA expression in myocytes 4 h after exposure to either cytokine (Fig. 8B). LPS, IL-1␤, TNF␣, and IL-6 all increased NOS2 protein as detected by Western blotting (Fig. 8C). Despite the fact that cytokines can induce NOS2 expression and NO donors can decrease IGF-I mRNA levels, neither the generalized NOS inhibitor L-NMMA nor the NOS2 specific inhibitor 1400W altered the basal level of IGF-I mRNA in C2C12 cells. In addition, neither inhibitor blocked the ability of TNF␣ to inhibit IGF-I mRNA expression. We

Frost et al. • Regulation of Muscle IGF-I by TNF␣

Endocrinology, May 2003, 144(5):1770 –1779 1775

itor] slightly stimulated IGF-I mRNA expression (Fig. 9B). Exposure of myocytes to the p38 kinase inhibitor SB202190 decreased IGF-I mRNA expression, whereas the JNK inhibitor SP600125 had no effect on the basal level of IGF-I mRNA. Of the MAPK inhibitors examined, only the JNK inhibitor (SP600125) blocked the TNF␣-induced decrease in IGF-I mRNA expression (Fig. 9D). The efficacy of the specific doses of each inhibitor has previously been determined to be effective in myocytes and other cell types. TNF␣ inhibits IGF-I mRNA expression in myoblasts and myotubes

Although myoblasts and myotubes share many regulatory pathways, myotubes also express a large number of musclespecific genes. To examine whether TNF␣ regulates IGF-I gene expression comparably in myoblasts and myotubes, we differentiated C2C12 cells to a point at which they formed multinucleated myotubes. The cells exhibited occasional spontaneous contractions and expressed contractile proteins such as sarcomeric myosin (Fig. 10A). TNF␣ suppressed IGF-I mRNA expression by 50 – 80% in both myoblasts and myotubes at 24 h, and this was sustained for at least 48 h (Fig. 10, B and C). Discussion

FIG. 7. TNF␣ inhibits GH-stimulated IGF-I mRNA expression but not STAT phosphorylation. C2C12 myoblasts were grown as described in Fig. 3 and pretreated with TNF␣ for 30 min. GH was added to the cultures and RNA isolated 18 h later. IGF-I mRNA in the cultures was quantified and is presented (A). Additional cells were pretreated with either saline or TNF␣ for 30 min followed by GH for 10 min. Cell extracts from duplicate dishes were isolated in sample buffer and run on quadruplicate 7.5% acrylamide gels. The gels were transferred to nitrocellulose and probed for phosphorylated (pSTAT) and total STAT (tSTAT)-3 and -5, respectively. The proteins were detected with a chemiluminescent substrate and captured on x-ray film. A representative set of images is presented (B).

were also unable to block the TNF␣-induced decrease in IGF-I mRNA with two inhibitors of nuclear factor ␬B (NF␬B) activation, PDTC and MG132 (data not shown). TNF␣ stimulates the phosphorylation of Jun N-terminal kinase (JNK) and a JNK inhibitor blocks TNF␣-mediated down-regulation of IGF-I mRNA

TNF␣ and other proinflammatory cytokines are known to activate multiple MAPK via their phosphorylation on serine and tyrosine residues. We examined whether the MAPK ERK-1 and -2 and the JNK substrate c-Jun become phosphorylated after exposure of C2C12 myocytes to TNF␣. TNF␣ stimulated the phosphorylation of c-Jun but had no effect on ERK-1 and -2 (Fig. 9A). TNF␣ decreased the basal expression of IGF-I mRNA by approximately 50 – 80% (Fig. 9, B–D). MAPK inhibitors had differing effects on the basal expression of IGF-I mRNA. As observed in a previous study by our laboratory, PD98059 [a MAPK kinase (MEK)-1 inhib-

The molecular mechanisms regulating IGF-I mRNA expression in skeletal muscle have remained largely obscure because of the lack of a good model system for examining positive and negative regulators of IGF-I expression. Recently our laboratory and others have shown that C2C12 myoblasts and myotubes express IGF-I mRNA and that its abundance can be positively regulated by GH (20, 23). It is likely that a number of cytokines and serum hormones can also influence IGF-I mRNA abundance. These factors may have direct effects on muscle protein content and/or indirectly influence muscle function by down-regulating IGF-I expression. In this study, we demonstrated that LPS is a potent stimulus for both the systemic expression of TNF␣ in blood as well as the expression of TNF␣ mRNA in mouse skeletal muscle. The temporal expression of TNF␣ at 1–2 h followed by a fall in skeletal muscle IGF-I mRNA by 6 h suggests that TNF␣ may contribute to the diminution in muscle IGF-I mRNA. This finding is consistent with our previous studies in rats in which infusion of TNF␣ decreased IGF-I mRNA content in multiple muscles including the gastrocnemius, soleus, and heart (21). In addition, we demonstrated that an anti-TNF␣ antibody can completely prevent the LPSinduced reduction in IGF-I mRNA in rat skeletal muscle (13). A limitation of in vivo studies is that skeletal muscle is composed of multiple cell types and as a result TNF␣ and IGF-I mRNA may be derived from not only muscle cells but also blood-borne immune cells or endothelial cells that are components of the muscle vasculature. To examine the regulation of muscle IGF-I mRNA in more detail, we determined whether LPS and TNF␣ could down-regulate IGF-I mRNA in C2C12 myoblasts. These cells embody a muscle precursor phenotype much like satellite cells that are resident in mature muscle. In addition, we have found that the C2C12 cell line

1776

Endocrinology, May 2003, 144(5):1770 –1779

Frost et al. • Regulation of Muscle IGF-I by TNF␣

FIG. 8. Influence of TNF␣ and NO on IGF-I mRNA. C2C12 myoblasts were grown as described in Fig. 3 and treated with 0.06, 0.125, 0.25, 0.5, or 1 mM SNP. RNA was isolated 8 h later. IGF-I mRNA in the cultures was quantified and is presented (A). Additional cells were treated with TNF␣ or IL-␤ (40 ng/ml) and RNA isolated 0.33, 0.66, 1, 2, 4, or 8 h later. NOS2 mRNA was determined by RPA and is presented (B). NOS2 protein was also determined in cell extracts from cells treated with LPS (1 ␮g/ml), IL-1␤, TNF␣, and IL-6 (40 ng/ml) by Western blotting (C). Some cells were pretreated with the NOS inhibitors L-NMMA and 1400W for 30 min followed by TNF␣. IGF-I mRNA was determined and is presented (D). All data were normalized for L32 mRNA as described in Materials and Methods. Values are means ⫾ SEM of triplicate dishes. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05).

is responsive to multiple proinflammatory molecules at the myoblast stage including LPS from Gram-negative bacteria, peptidoglycan from Gram-positive bacteria, IL-1␤, and TNF␣ (24). C2C12 myoblasts are therefore a good model cell type to assess the interaction between the immune response and autocrine/paracrine growth factors and cytokines. Although TNF␣ can suppress the differentiation of C2C12 cells and this response is associated with many changes in gene expression, our cells were grown as myoblasts in 10% serum and did not exhibit significant differentiation into myotubes under basal or stimulated conditions. Use of myoblasts has also allowed us to more easily compare our observations here with previous observations we have made on the regulation of IGF-I mRNA by GH in C2C12 cells (20). C2C12 myoblasts responded to LPS and TNF␣ with a 50 – 80% decrease in IGF-I mRNA, and the time course of this drop was similar to that observed in vivo in skeletal muscle. The TNF␣-induced decrease in IGF-I mRNA expression in myocytes was not due to a generalized decrease in the abundance of mRNA. RNA transcribed from a number of genes including GAPDH, L32, and 18S rRNA remained stable. IGF-II mRNA content, another IGF-system component, was also unchanged by TNF␣. In addition, TNF␣ increased the mRNA abundance of another mRNA Mac-25. The changes in magnitude and specificity of IGF-system components in

C2C12 myoblasts was comparable to that found in both mouse and rat skeletal muscle (21). In addition, the concentration of TNF␣ that was necessary to observe a 50% decrease in IGF-I mRNA in C2C12 cells was comparable to that found in the circulation of mice injected with LPS (compare Figs. 1 and 4). The C2C12 myoblasts exhibited a sustained decrease in IGF-I mRNA relative to control cultures and mice that were injected with LPS demonstrated a 50% decrease in IGF-I mRNA that also persisted for at least 18 h. C2C12 myoblasts therefore express growth factors and cytokines that are regulated in a manner comparable to that observed in skeletal muscle in vivo (24). We have previously found that GH stimulates IGF-I mRNA expression in C2C12 myoblasts and that the response requires both ongoing transcription and translation. GHinducible IGF-I mRNA expression is blocked by either cycloheximide or DRB (20). Ongoing protein synthesis is also required for the TNF␣-induced fall in IGF-I mRNA because this response was blocked by cycloheximide. TNF␣ has the potential to decrease IGF-I mRNA abundance by inhibiting transcription of the IGF-I gene or accelerating the decay of the IGF-I message. In C2C12 cells we found that IGF-I mRNA has a half-life of 10 h and that this measurement was not altered by TNF␣. The half-life of IGF-I mRNA in C2C12 cells was similar to that described previously for both C6 glioma cells

Frost et al. • Regulation of Muscle IGF-I by TNF␣

Endocrinology, May 2003, 144(5):1770 –1779 1777

FIG. 9. TNF␣ stimulates c-Jun phosphorylation and a JNK inhibitor blocks the TNF␣-induced fall in IGF-I mRNA. C2C12 myoblasts were grown as described in Fig. 3 and treated with TNF␣ for 15 min. Cell extracts were isolated in sample buffer and run on 10% SDS-PAGE gels. After transfer to nitrocellulose, the blots were probed for phosphorylated ERK-1 and -2 and c-Jun. A representative blot is shown (A). Additional cells were pretreated with an MEK-1 inhibitor (PD98059), p38 inhibitor (SB202190), or JNK inhibitor (SP600125) for 30 min followed by TNF␣ for 8 h. Phosphorimages of the dried gels were quantified and are plotted (B–D). All data were normalized for L32 mRNA as described in Materials and Methods. Values are means ⫾ SEM of triplicate dishes. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05).

(25) and osteoblasts (26). Additional studies that examine IGF-I promoter activity in C2C12 cells are necessary to determine whether GH and TNF␣ activate and inhibit IGF-I gene transcription, respectively. GH-induced IGF-I mRNA expression has been shown to be STAT5b dependent in the liver (27). Likewise, we have previously demonstrated that GH stimulates the phosphorylation of STAT3 and STAT5 in C2C12 myoblasts (20). It was therefore of interest to determine whether TNF␣ could alter GH responsiveness in C2C12 cells. TNF␣ decreased both basal and GH-stimulated IGF-I mRNA expression, suggesting that the TNF␣ effect is dominant. Although GH-induced IGF-I mRNA expression is associated with STAT3 and STAT5 phosphorylation, TNF␣ did not alter GH-induced phosphorylation of these two transcription factors. This suggests that TNF␣ induces GH resistance via signaling components that act either after or independently of STAT phosphorylation. This may include the induction of suppressors of cytokine signaling-1, -2, and -3 and the cytokine-inducible Src homology 2-containing protein CIS. Alternatively, TNF␣ may induce other secondary mediators that in turn blunt IGF-I mRNA expression. This would be consistent with the need for ongoing protein synthesis to occur to observe the TNF␣-induced decrease in IGF-I mRNA expression.

The well-known ability of proinflammatory cytokines, such as TNF␣, to induce NO synthesis by up-regulating NOS2 in immune cells led us to examine whether NO donors would decrease IGF-I mRNA in C2C12 cells. SNP and glycoSNAP, two NO donors, both decreased IGF-I mRNA abundance. TNF␣ also induced NOS2 mRNA and protein expression in the C2C12 cells, albeit to a lesser extent than that found in response to IL-1␤ or LPS. Despite these differences, neither a generalized inhibition of NOS with L-NMMA nor a specific inhibition of NOS2 with 1400W blocked the TNF␣induced fall in IGF-I mRNA. These data suggest that TNF␣ uses an NO-independent pathway to regulate IGF-I mRNA in myocytes. Although many cytokine-mediated events occur via the activation of NF␬B. We were also unable to prevent the TNF␣-induced decrease in IGF-I mRNA with PDTC, which is both an antioxidant and inhibitor of NF␬B activation. In addition, we were unsuccessful at blocking the decrease in IGF-I mRNA with the proteasomal inhibitor MG-132 (data not shown). This compound prevents the proteolysis of inhibitor of NF␬B and thereby blocks NF␬B activation. These data suggest that activation of NF␬B does not mediate TNF␣induced down-regulation of IGF-I mRNA abundance in C2C12 myoblasts.

1778

Endocrinology, May 2003, 144(5):1770 –1779

FIG. 10. TNF␣ inhibits IGF-I mRNA expression in differentiated myotubes. C2C12 myoblasts were grown as described in Fig. 3. Additional cells were allowed to differentiate in low serum containing media until they formed multinucleated myotubes that exhibited spontaneous contractions. Two dishes of myoblasts and myotubes were isolated in gel sample buffer containing 3 M urea and run on an SDS-PAGE gel. The resulting blot was probed for sarcomeric myosin and is shown (A). Myoblasts (B) and myotubes (C) were also treated with TNF␣ for either 24 or 48 h and RNA isolated for quantification of IGF-I mRNA by RPA. All data were normalized to L32 mRNA as described in Materials and Methods and expressed as a percentage of cells treated with saline alone. Values are means ⫾ SEM. Bars with different lowercase letters are significantly different from each other (P ⬍ 0.05).

Because TNF␣ and other cytokines strongly activate MAPK and downstream substrates, we examined whether TNF␣ induced the phosphorylation of ERK-1 and -2 and c-Jun in C2C12 cells. Phospho-specific antibodies to these substrates demonstrated c-Jun phosphorylation in C2C12 cells in response to TNF␣. Only JNK activation was associated with the ability of TNF␣ to down-regulate IGF-I mRNA. A specific JNK inhibitor (SP600125) completely blocked the TNF␣-induced decrease in IGF-I mRNA, whereas MEK and p38 inhibitors were without effect. JNK phosphorylates a number of factors including the activator protein-1 transcription factor c-Jun. An activator protein-1 enhancer has been shown to be necessary for the response of the chicken IGF-I gene to 12-O-tetradecanoyl-

Frost et al. • Regulation of Muscle IGF-I by TNF␣

phorbol 13-acetate in HepG2 cells (28). It is not known whether this mechanism is pertinent in the tissue-specific context of muscle cells. Transcription factor c-Jun is part of a much larger family of transcription factors including Jun B and Jun D and is influenced by its potential heterodimerization partners c-Fos, Fos B, and Fra-1 and -2 as well as coactivators such as Jun-activation domain-binding protein 1. Additional studies examining whether these transcription factors can activate and/or inhibit the IGF-I promoter in C2C12 cells is necessary to determine whether they play a role in IGF-I gene transcription in muscle. In addition, although SP600125 shows good selectivity to JNK, we cannot completely exclude the possibility that this inhibitor affects other kinases or signaling pathways. We were unable to block the effect of TNF␣ with epigallocatechin-3-gallate, an active polyphenol and a proposed JNK inhibitor, present in green tea (29). It is likely that this compound is much less specific than SP600125. Low serum IGF-I levels are seen in a variety of catabolic states including sepsis, AIDS, thermal injury, cancer cachexia, and chronic alcohol abuse. Skeletal muscle from experimental animals and patients also has reduced levels of IGF-I peptide and mRNA. Because muscle wasting is due primarily to an increase in skeletal muscle protein degradation (30) and a decrease in protein synthesis (14), it has been hypothesized that a deficit of IGF-I in skeletal muscle is causally related to muscle wasting. GH has therefore been used for the restoration of muscle mass in patients with AIDS (31) and thermal injury (32). An IGF-I and IGFBP-3 complex also is effective at restoring muscle protein synthesis in septic rats (33). In addition, IGF-I delivered locally into muscle by adenovirus assisted gene transfer increases proliferation of satellite cells and induces skeletal muscle hypertrophy (34). Future studies are needed to examine the feasibility of selectively blocking the drop in endogenous IGF-I in skeletal muscle in vivo to determine whether such an intervention prevents the loss of muscle mass after catabolic insults. Acknowledgments We thank the National Hormone and Pituitary Program for providing antiserum to IGF-I. The MF-20 monoclonal antibody developed by D. A. Fishman was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the National Institute of Child Health and Human Development and maintained by the University of Iowa Department of Biological Sciences (Iowa City, IA). Received August 5, 2002. Accepted January 14, 2003. Address all correspondence and requests for reprints to: Robert A. Frost, Ph.D., Department of Cellular and Molecular Physiology, Pennsylvania State University College of Medicine, Hershey Medical Center: H166, Hershey, Pennsylvania 17033. E-mail: [email protected]. This work was supported in part by NIH Grants GM-38032 and AA-11290.

References 1. Tisdale MJ 1997 Biology of cachexia. J Natl Cancer Inst 89:1763–1773 2. Espat NJ, Copeland EM, Moldawer LL 1994 Tumor necrosis factor and cachexia: a current perspective. Surg Oncol 3:255–262 3. Lee SJ, McPherron AC 1999 Myostatin and the control of skeletal muscle mass. Curr Opin Genet Dev 9:604 – 607 4. Thomas M, Langley B, Berry C, Sharma M, Kirk S, Bass J, Kambadur R 2000 Myostatin, a negative regulator of muscle growth, functions by inhibiting myoblast proliferation. J Biol Chem 275:40235– 40243

Frost et al. • Regulation of Muscle IGF-I by TNF␣ 5. Hasselgren PO 2000 Catabolic response to stress and injury: implications for regulation. World J Surg 24:1452–1459 6. Cooney R, Kimball SR, Eckman R, Maish 3rd G, Shumate M, Vary TC 1999 TNF-binding protein ameliorates inhibition of skeletal muscle protein synthesis during sepsis. Am J Physiol 276:E611–E619 7. Lang CH, Fan J, Cooney R, Vary TC 1996 IL-1 receptor antagonist attenuates sepsis-induced alterations in the IGF system and protein synthesis. Am J Physiol 270:E430 –E437 8. Thissen JP, Verniers J 1997 Inhibition by interleukin-1 ␤ and tumor necrosis factor-␣ of the insulin-like growth factor I messenger ribonucleic acid response to growth hormone in rat hepatocyte primary culture. Endocrinology 138: 1078 –1084 9. Wolf M, Bohm S, Brand M, Kreymann G 1996 Proinflammatory cytokines interleukin 1␤ and tumor necrosis factor ␣ inhibit growth hormone stimulation of insulin-like growth factor I synthesis and growth hormone receptor mRNA levels in cultured rat liver cells. Eur J Endocrinol 135:729 –737 10. Wang P, Li N, Li JS, Li WQ 2002 The role of endotoxin, TNF-␣, and IL-6 in inducing the state of growth hormone insensitivity. World J Gastroenterol 8:531–536 11. Colson A, Le Cam A, Maiter D, Edery M, Thissen JP 2000 Potentiation of growth hormone-induced liver suppressors of cytokine signaling messenger ribonucleic acid by cytokines. Endocrinology 141:3687–3695 12. Yumet G, Shumate ML, Bryant P, Lin CM, Lang CH, Cooney RN 2002 Tumor necrosis factor mediates hepatic growth hormone resistance during sepsis. Am J Physiol Endocrinol Metab 283:E472–E481 13. Fan J, Char D, Bagby GJ, Gelato MC, Lang CH 1995 Regulation of insulin-like growth factor-I (IGF-I) and IGF-binding proteins by tumor necrosis factor. Am J Physiol 269:R1204 –R1212 14. Lang CH, Frost RA 2002 Role of growth hormone, insulin-like growth factor-I, and insulin-like growth factor binding proteins in the catabolic response to injury and infection. Curr Opin Clin Nutr Metab Care 5:271–279 15. Frost RA, Lang CH 1998 Growth factors in critical illness: regulation and therapeutic aspects. Curr Opin Clin Nutr Metab Care 1:195–204 16. Lang CH, Frost RA, Nairn AC, MacLean DA, Vary TC 2002 TNF-␣ impairs heart and skeletal muscle protein synthesis by altering translation initiation. Am J Physiol Endocrinol Metab 282:E336 –E347 17. Li YP, Schwartz RJ, Waddell ID, Holloway BR, Reid MB 1998 Skeletal muscle myocytes undergo protein loss and reactive oxygen-mediated NF-␬B activation in response to tumor necrosis factor ␣. FASEB J 12:871– 880 18. Frost RA, Lang CH, Gelato MC 1997 Transient exposure of human myoblasts to tumor necrosis factor-␣ inhibits serum and insulin-like growth factor-I stimulated protein synthesis. Endocrinology 138:4153– 4159 19. Baumgarten G, Knuefermann P, Nozaki N, Sivasubramanian N, Mann DL, Vallejo JG 2001 In vivo expression of proinflammatory mediators in the adult heart after endotoxin administration: the role of Toll-like receptor-4. J Infect Dis 183:1617–1624 20. Frost RA, Nystrom GJ, Lang CH 2002 Regulation of IGF-I mRNA and signal

Endocrinology, May 2003, 144(5):1770 –1779 1779

21. 22. 23. 24. 25.

26. 27. 28.

29.

30. 31.

32. 33.

34.

transducers and activators of transcription-3 and -5 (Stat-3 and -5) by GH in C2C12 myoblasts. Endocrinology 143:492–503 Lang CH, Nystrom GJ, Frost RA 2001 Tissue-specific regulation of IGF-I and IGF-binding proteins in response to TNF␣. Growth Horm IGF Res 11:250 –260 Frost RA, Nachman SA, Lang CH, Gelato MC 1996 Proteolysis of insulin-like growth factor-binding protein-3 in human immunodeficiency virus-positive children who fail to thrive. J Clin Endocrinol Metab 81:2957–2962 Sadowski CL, Wheeler TT, Wang LH, Sadowski HB 2001 GH regulation of IGF-I and suppressor of cytokine signaling gene expression in C2C12 skeletal muscle cells. Endocrinology 142:3890 –3900 Frost RA, Nystrom GJ, Lang CH 2002 Lipopolysaccharide regulates proinflammatory cytokine expression in mouse myoblasts and skeletal muscle. Am J Physiol Regul Integr Comp Physiol 283:R698 –R709 Wang L, Adamo ML 2001 Cyclic adenosine 3⬘, 5⬘-monophosphate inhibits insulin-like growth factor I gene expression in rat glioma cell lines: evidence for regulation of transcription and messenger ribonucleic acid stability. Endocrinology 142:3041–3050 Bichell DP, Rotwein P, McCarthy TL 1993 Prostaglandin E2 rapidly stimulates insulin-like growth factor-I gene expression in primary rat osteoblast cultures: evidence for transcriptional control. Endocrinology 133:1020 –1028 Davey HW, Xie T, McLachlan MJ, Wilkins RJ, Waxman DJ, Grattan DR 2001 STAT5b is required for GH-induced liver IGF-I gene expression. Endocrinology 142:3836 –3841 Kajimoto Y, Kawamori R, Umayahara Y, Iwama N, Imano E, Morishima T, Yamasaki Y, Kamada T 1993 An AP-1 enhancer mediates TPA-induced transcriptional activation of the chicken insulin-like growth factor I gene. Biochem Biophys Res Commun 190:767–773 Hwang KC, Lee KH, Jang Y, Yun YP, Chung KH 2002 Epigallocatechin-3gallate inhibits basic fibroblast growth factor-induced intracellular signaling transduction pathway in rat aortic smooth muscle cells. J Cardiovasc Pharmacol 39:271–277 Fang CH, Li BG, Wray CJ, Hasselgren PO 2002 Insulin-like growth factor-I inhibits lysosomal and proteasome-dependent proteolysis in skeletal muscle after burn injury. J Burn Care Rehabil 23:318 –325 Schambelan M, Mulligan K, Grunfeld C, Daar ES, LaMarca A, Kotler DP, Wang J, Bozzette SA, Breitmeyer JB 1996 Recombinant human growth hormone in patients with HIV-associated wasting. A randomized, placebocontrolled trial. Serostim Study Group. Ann Intern Med 125:873– 882 Jeschke MG, Barrow RE, Herndon DN 2000 Recombinant human growth hormone treatment in pediatric burn patients and its role during the hepatic acute phase response. Crit Care Med 28:1578 –1584 Svanberg E, Frost RA, Lang CH, Isgaard J, Jefferson LS, Kimball SR, Vary TC 2000 IGF-I/IGFBP-3 binary complex modulates sepsis-induced inhibition of protein synthesis in skeletal muscle. Am J Physiol Endocrinol Metab 279: E1145–E1158 Barton-Davis ER, Shoturma DI, Sweeney HL 1999 Contribution of satellite cells to IGF-I induced hypertrophy of skeletal muscle. Acta Physiol Scand 167:301–305