|
|
||||||||
reverses insulin resistance in skeletal muscle but not adipose tissue
1Department of Applied Physiology and Kinesiology, University of Florida, Gainesville 32611; 2Malcom Randall Veterans Affairs Medical Center, Gainesville, Florida 32608; and 3Department of Physiology, Louisiana State University Health Science Center, New Orleans, Louisiana 70112
Submitted 3 February 2004 ; accepted in final form 15 June 2004
| ABSTRACT |
|---|
|
|
|---|
(TNF-
) as a mediator of insulin resistance in maturing male Sprague-Dawley rats. Rats were treated either with goat anti-murine TNF-
IgG (anti-TNF-
) or goat nonimmune IgG (NI) for 7 days. Vascular catheters were implanted, and rats were fasted overnight before hyperinsulinemic euglycemic clamp (HUC) studies were performed. TNF-
neutralization increased the rate of glucose infusion required to maintain euglycemia by 68%. Insulin-stimulated glucose transport into individual tissues was measured after bolus administration of 2-deoxy-[14C]glucose during HUC. Anti-TNF-
administration increased glucose transport in muscles composed predominantly of fast-twitch fibers: white gastrocnemius muscle (68% increase) and tibialis anterior muscle (64% increase). There were nonsignificant trends for increased glucose transport in the slow-twitch soleus muscle and in the mixed-fiber red gastrocnemius muscle. Glucose transport was unchanged in visceral and subcutaneous fat. Anti-TNF treatment did not alter body weight, muscle mass, or fat mass. Anti-TNF-
did not alter the distribution of the 17-kDa and 26-kDa forms of TNF-
in either muscle or fat. However, anti-TNF-
treatment caused an
50% reduction in the secretion of TNF-
bioactivity in vitro by explants of visceral and subcutaneous fat. We conclude that TNF-
neutralization reversed insulin resistance substantially in fast-twitch muscle and may have done so in other muscles, while having little effect in fat. TNF-
neutralization was accompanied by reduced TNF-
bioactivity without tissue depletion of TNF-
protein.
glucose transport; muscle fiber type; soluble tumor necrosis factor-
; membrane tumor necrosis factor-
plays a role in the development of insulin resistance. Infusion of TNF-
results in an insulin-resistant state (13, 24). TNF-
is known to impair insulin signaling in cultured cells by three separate mechanisms (3). First, TNF-
has been shown to cause phosphorylation of Ser307 on insulin receptor substrate (IRS)-1 in murine adipocytes (11), rat hepatoma cells (25), and C2C12 muscle cells (9). This alteration makes IRS-1 resistant to subsequent insulin-stimulated tyrosine phosphorylation, resulting in reduced docking of phosphatidylinositol 3-kinase and in impaired insulin-stimulated glucose transport (11, 20, 26). Second, TNF-
phosphorylates and activates the protein tyrosine phosphatase SH-PTPase, which terminates insulin action by removing tyrosine phosphates from IRS-1 (1). Third, TNF-
phosphorylates and inactivates the protein phosphatase PP-1, preventing insulin activation of glycogen storage (9, 26). Hotamisligil (18) has shown that TNF-
receptor knockout in ob/ob mice improves glucose tolerance. We (4) and others (8, 21) have reported that the neutralization of TNF-
reverses insulin resistance in rodents. We found that administration of anti-TNF-
to S-D rats restores muscle insulin responsiveness, as assessed in isolated strips of soleus muscle (4).
The main purpose of the present study was to determine whether administration of anti-TNF-
could restore insulin responses in skeletal muscle, visceral fat, and subcutaneous fat. Reported overexpression of TNF-
in adipose tissue of insulin-resistant rodents (21) has led to the hypothesis that TNF-
of adipose origin causes systemic insulin resistance. However, we have recently reviewed other evidence suggesting that muscle insulin resistance is caused by locally produced TNF-
(3). A secondary purpose was to determine the effect of antibody administration on TNF-
expression in muscle and fat.
| MATERIALS AND METHODS |
|---|
|
|
|---|
420 g were injected daily, subcutaneously, for 7 days with 8 mg of goat anti-murine TNF IgG (anti-TNF-
) or goat nonimmune IgG (NI). Polyclonal goat anti-murine TNF-
IgGs were prepared, as we have previously reported (4, 10), by using the Ribi adjuvant system containing 0.5 mg each of monophosphoryl lipid A, trehalose dimycolate, and cell wall skeleton in 0.2% Tween 80 (Ribi ImmunoChemistry Research, Hamilton, MT). The serum IgG fraction was obtained by polyethylene glycol precipitation and column chromatography by use of DEAE Bio-Gel A (Bio-Rad Laboratories, Richmond, CA). The neutralizing capacity of the anti-TNF-
IgG fraction in the L929 cytotoxicity assay was 6.5 and 9.0 x 105 50% neutralizing U/mg IgG against recombinant TNF-
and serum TNF-
from LPS-treated rats, respectively. Nonimmune goat IgG was prepared in the same way and had no detectable TNF-
-neutralizing activity. Neither IgG preparation demonstrated effective binding to LPS, IL-1, or interferon-
in an ELISA protocol. Insulin responsiveness. Glucose transport into individual tissues was measured in vivo during hyperinsulinemic euglycemic clamps (HUC), as previously described by Lang and colleagues (22, 23). Briefly, vascular catheters were implanted, and rats were fasted overnight before HUC were performed. A dose of 100 mU of human insulin was administered by bolus injection followed by infusion at a rate of 100 mU/h. We have previously shown that this rate of insulin infusion produces maximal rates of glucose transport into individual skeletal muscles (22, 23) and elevates serum insulin above postprandial concentrations in S-D rats (28) and above the highest concentrations observed during glucose tolerance testing of S-D rats (7). D-Glucose (30%) was infused at an initial rate of 185 µmol·kg1·min1. Blood glucose was measured every 15 min and the rate of glucose infusion adjusted to maintain euglycemia. After HUC had been established for 140 min, an intravenous bolus tracer of 8 µCi 2-deoxy-[14C]glucose was administered. After an additional 40 min, the rat was euthanized and tissues were excised. The integrated specific activity of 2-deoxyglucose is calculated by measuring blood radioactivity periodically over the course of 40 min. The rate of whole body glucose utilization was calculated by dividing the glucose infusion rate by body weight and is expressed in micromoles per kilogram per minute. Glucose transport into individual tissues was estimated by measuring tissue content of phosphorylated 2-deoxy-[14H]glucose (22, 23). Tissue homogenates were precipitated using Somogyi reagents, and phosphorylated 2-deoxy-[14H]glucose was estimated as the difference in radioactivity before and after precipitation. Transport rates were corrected using a lump constant that accounts for competition for transport between glucose and 2-deoxyglucose (22, 23). Serum insulin was assayed by radioimmunoassay (Linco Research, St. Charles, MO).
Secretion of TNF-
protein and TNF-
bioactivity in fat explants.
As previously described (24), 1 g of visceral or subcutaneous fat was minced and suspended in enriched culture medium (DMEM with glutamine, private, 4,500 mg/l glucose, Sigma nonessential amino acids, and 0.5% BSA). After incubation at 37°C for 60 min, medium was removed and centrifuged for 5 min at 13,000 g. TNF-
protein was measured using an ELISA that is specific for rat TNF-
and has a sensitivity of 0.7 pg/ml (BioSource International, Camarillo, CA). TNF-
bioactivity was measured using the WEHI 164 subclone 13 cytotoxicity assay (12). Briefly, 5 x 104 cells in RPMI 1640 containing 10% FCS, 1 mM L-glutamine, and 0.5 mg/ml actinomycin D were added to serially diluted samples in microtiter plates and incubated for 20 h at 37°C and 5% CO2. Cytotoxicity was assessed by measuring MTT-tetrazolium (3,[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) conversion to formazan at 550 nm (16). TNF-
concentrations (pg/ml) were calculated using standard recombinant muTNF-
(Genentech, South San Francisco, CA). Goat anti-muTNF-
IgG reduces plasma TNF-
activity to undetectable levels.
Western analysis of tissue TNF-
.
Tissue extracts were prepared as previously described (5). Tissue samples (0.2 g muscle or 0.5 g fat) were homogenized in buffer containing 0.9% NaCl, 10 mM Na2HPO4, 1 mmol PMSF, 1 mg/l each of pepstatin, aprotinin, and leupeptin, 0.5% Triton X-100, and 0.05% sodium azide, pH = 7.2. Samples were subjected to 1 freeze/thaw cycle, probe sonicated briefly, and centrifuged at 100,000 g for 1 h. Tissue extracts (
5 µg of protein) were subjected to 15% polyacrylamide gel electrophoresis, and proteins were transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA). Membranes were blocked with Starting Block (Pierce, Rockford, IL), incubated first with 1° antibody (1:1,400, rabbit anti-rat, BioSource, Camarillo, CA) and then with 2° antibody [1:2,800, goat F(ab)2 anti-rabbit, BioSource]. Bands were detected with enhanced chemiluminescence plus reagents (Amersham, Piscataway, NJ) and quantitated using phosphor-imaging and ImageQuant software (Molecular Dynamics, Sunnyvale, CA). Rat 17-kDa TNF-
(BioSource) was used as a positive control.
Statistical analysis. Differences among means were tested by ANOVA, with P < 0.05 defined as the threshold of significance. Post hoc analysis was performed using Fisher's exact least significant differences test. Values are reported as means ± SE.
| RESULTS |
|---|
|
|
|---|
IgG was consistently detected in plasma samples taken after 2 and 4 days of antibody treatment. Administration of anti-TNF-
had no effect on food intake or body composition (Table 1). Likewise, weight gain, muscle mass, and fat mass, either visceral or subcutaneous, did not differ between nonimmune IgG and anti-TNF-
-treated animals (Table 1).
|
(128.7 ± 15.9 µmol·kg1·min1 for anti-TNF-
vs. 76.7 ± 5.9 µmol·kg1·min1 for NI; P < 0.05, n = 68; Fig. 1A). The serum insulin concentration during clamps was 19.2 ± 6.9 ng/ml, a concentration that produces maximal glucose transport (22, 23). TNF-
neutralization also increased glucose transport in two muscles composed predominantly of fast-twitch fibers (see Fig. 1B). In white gastrocnemius muscle, treatment with anti-TNF-
caused a 68% increase in glucose transport (74.6 ± 21.7 nmol·g1·min1 for anti-TNF-
vs. 46.6 ± 14.5 nmol·g1·min1 for NI, P < 0.05). In tibialis anterior muscle, anti-TNF-
caused a 64% increase in transport (61.4 ± 9.34 nmol·g1·min1 for anti-TNF-
vs. 37.4 ± 5.7 nmol·g1·min1 for NI, P < 0.05). There were near-significant trends for increased glucose transport in a slow-twitch muscle (soleus: 164 ± 26.6 nmol·g1·min1 for anti-TNF-
vs. 121 ± 29.5 nmol·g1·min1 for NI) and in a mixed fiber muscle (red gastrocnemius: 74.6 ± 21.7 nmol·g1·min1 for anti-TNF-
vs. 46.6 ± 14.5 nmol·g1·min1 for NI). Glucose transport was virtually unchanged in visceral fat (13.8 ± 5.2 nmol·g1·min1 for anti-TNF-
vs. 10.4 ± 1.5 nmol·g1·min1 for NI) and in subcutaneous fat (13.0 ± 2.8 nmol·g1·min1 for anti-TNF-
vs. 15.3 ± 2.4 nmol·g1·min1 for NI).
|
expression.
Western analysis revealed that both the 17-kDa and 26-kDa forms of TNF-
are relatively abundant in fat, whereas in muscle, the 17-kDa form predominates (see Fig. 2). Anti-TNF-
did not alter tissue distribution of 17-kDa and 26-kDa forms of TNF-
in either muscle or fat. After incubation of fat explants in DMEM for 60 min at 37°C, TNF-
protein and TNF-
bioactivity present in conditioned media were measured, as described in MATERIALS AND METHODS (Table 2). Anti-TNF-
treatment had no effect on secretion of TNF-
protein by either visceral or subcutaneous fat (Table 2). However, there was significant overall reduction in secretion of TNF-
bioactivity by fat explants (P < 0.05) and a specific 59% reduction in secretion by visceral fat (P < 0.05) (Fig. 3).
|
|
|
| DISCUSSION |
|---|
|
|
|---|
is known to reverse hepatic insulin responsiveness in Zucker fatty rats. Hotamisligil et al. (21) first demonstrated in 1993 that administration of a TNF-
receptor-IgG chimeric protein construct could increase insulin-stimulated glucose disposal in Zucker fatty rats. Cheung et al. (8) administered a gene construct coding for a similar TNF-
inhibitory peptide to Zucker rats and observed a restoration of insulin suppression of hepatic glucose output without a reduction in adiposity. We have previously reported that administration of anti-TNF-
antibody to maturing S-D rats reverses insulin resistance in isolated strips of soleus muscle. The present study demonstrates that neutralization of TNF-
restores insulin responsiveness in a variety of skeletal muscles, but not in visceral or subcutaneous fat. Glycolytic muscle fibers have a greater dependence on glucose transport. Therefore, insulin resistance may have greater consequences in glycolytic muscles. We chose to study muscles that are relatively homogeneous for each of the major fiber types: soleus for type I slow oxidative fibers, tibialis anterior and white gastrocnemius for type II fast glycolytic fibers, and red gastrocnemius for type II fast oxidative fibers. We found increased insulin responsiveness in postural (soleus) and activity (gastrocnemius) muscles and probably also in muscles composed primarily of slow fibers (soleus), fast fibers (tibialis anterior), and mixed fibers (gastrocnemius). We found that TNF neutralization increased insulin responsiveness in muscle, but not in adipose tissue. The reason for this differential effect is not clear. During maturation, S-D rats develop a reduced response to insulin in muscle (7, 14), but it is not known whether the response in fat is also reduced during the same period. One would predict that TNF neutralization would increase the response to insulin only in those tissues where that response was impaired to begin with. It is also possible that anti-TNF did not increase insulin responsiveness in fat because the antibody did not have sufficient access to fat tissue. However, anti-TNF did reduce secretion of TNF bioactivity in fat explants, which probably indicates that it was present in fat tissue.
Although neutralization of TNF-
reverses insulin resistance, it is not clear that TNF-
is elevated in insulin resistance. Several laboratories have reported that expression of TNF-
mRNA is elevated in adipose tissue of insulin-resistant rodents (17, 21) and humans (2, 15, 19). Saghizadeh et al. (27) found that TNF-
mRNA is elevated sixfold in muscle biopsies obtained from type 2 diabetic subjects. These reports have led to the hypothesis that TNF-
of adipose origin reaches the circulation, causing systemic insulin resistance. However, two observations argue against this hypothesis. First, we have shown that tissue levels of TNF-
protein are several orders of magnitude higher than circulating levels (5), indicating that the role of TNF-
in insulin resistance may be paracrine, rather than endocrine. Second, changes in TNF-
message do not always reflect changes in TNF-
protein, and recent evidence suggests that, whereas TNF-
mRNA is elevated in insulin resistance, TNF-
protein is not. We recently reported that TNF-
protein is decreased in visceral fat, subcutaneous fat, and glycolytic muscle of S-D rats during the onset of age-induced insulin resistance (5). In fat explants, we found that in vitro secretion of TNF-
protein is decreased with age, but secretion of TNF-
bioactivity is unchanged or increased (5). TNF-
bioactivity was measured by the WEHI cytotoxicity assay and reflects the actions both of TNF-
and of its endogenous inhibitor, putatively a soluble form of the TNF-
receptor. Thus it is possible that TNF-
action may be increased in insulin-resistant tissues despite the decrease in tissue content of TNF-
protein. Support for this concept is found in a report by Morin et al. (24), who found that, in fat explants obtained from aged Fischer 344 rats, in vitro secretion of TNF-
protein was decreased, whereas secretion of TNF-
bioactivity was markedly increased (24).
Tissue TNF-
is expressed in two forms, a 17-kDa soluble form (sTNF-
) and a 26-kDa membrane form (mTNF-
) (29). TNF-
is produced in the 26-kDa form and converted to the 17-kDa form by the action of TNF-
-converting enzyme (29). In visceral and subcutaneous fat from S-D rats, we found that both sTNF-
and mTNF-
were relatively abundant. In muscle, sTNF-
was the predominant form. Treatment with anti-TNF-
did not alter TNF-
expression. In Wistar rats, we observed a similar pattern, in which both mTNF-
and sTNF-
are abundant in fat and sTNF-
is predominant in muscle and liver (S. E. Borst, unpublished observation). We also found that high-fat feeding of Wistar rats increased the expression of sTNF-
in visceral fat and muscle (S. E. Borst, unpublished observation).
Our results shown in Fig. 2 indicate an apparent lack of effect by anti-TNF-
on tissue expression of TNF-
protein. In muscle and fat, we detected similar amounts of both mTNF-
and sTNF-
, whether rats were treated with neutralizing antibody or with nonimmune IgG. We interpret this finding to mean that, in all probability, the administered antibody effectively neutralized the biological activity of TNF-
within these tissues but that TNF-
remained present in the tissue and remained detectable by Western analysis. This interpretation is supported by the findings in Table 2 and Fig. 3. Explants of fat obtained from anti-TNF-
-treated and NI-treated rats secreted similar amounts of TNF-
protein in vitro, as assessed by ELISA (Table 2). However, TNF-
bioactivity, as assessed by WEHI cytotoxicity, was 50% lower in the anti-TNF-
-treated group. We interpret the lower bioactivity to reflect the presence of antibody within fat tissue. Therefore, we think that neutralizing antibody was present in both muscle and fat but caused enhanced insulin responsiveness in muscle only.
In the present study, neutralization of TNF-
restored muscle insulin responsiveness, indicating that TNF-
plays a role in the pathophysiology of insulin resistance. The fact that TNF-
is virtually undetectable in the blood of insulin-resistant S-D rats (5) is consistent with TNF-
functioning in a paracrine fashion to decrease insulin responsiveness in skeletal muscle. S-D rats display both muscle (7, 14) and hepatic insulin resistance (8). In this study, whole body glucose disposal was significantly increased, with the increase apparently predominating in skeletal muscle after a 7-day period of reduced TNF-
activity. It is not clear whether increased muscle glucose transport was sufficient in magnitude to account for all of the improvement in glucose disposal.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
in insulin resistance. Endocrine 23: 177182, 2004.
(TNF-
)-induced insulin resistance: evidence for differential regulation of insulin signaling by TNF-
. Endocrinology 139: 49284935, 1998.
and obesity-induced insulin resistance. Science 271: 665668, 1996.
activity is elevated in older rats. J Gerontol 52: B190B195, 1997.
This article has been cited by other articles:
![]() |
Y. Wei, K. Chen, A. T. Whaley-Connell, C. S. Stump, J. A. Ibdah, and J. R. Sowers Skeletal muscle insulin resistance: role of inflammatory cytokines and reactive oxygen species Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2008; 294(3): R673 - R680. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. H St-Pierre, J.-P. Bastard, L. Coderre, M. Brochu, A. D Karelis, M.-E. Lavoie, F. Malita, J. Fontaine, D. Mignault, K. Cianflone, et al. Association of acylated ghrelin profiles with chronic inflammatory markers in overweight and obese postmenopausal women: a MONET study Eur. J. Endocrinol., October 1, 2007; 157(4): 419 - 426. [Abstract] [Full Text] [PDF] |
||||
![]() |
I C van Eijk, M J L Peters, M T Nurmohamed, A W van Deutekom, B A C Dijkmans, and S Simsek Decrease of fructosamine levels during treatment with adalimumab in patients with both diabetes and rheumatoid arthritis Eur. J. Endocrinol., March 1, 2007; 156(3): 291 - 293. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. M Marcora, K. R Chester, G. Mittal, A. B Lemmey, and P. J Maddison Randomized phase 2 trial of anti-tumor necrosis factor therapy for cachexia in patients with early rheumatoid arthritis Am. J. Clinical Nutrition, December 1, 2006; 84(6): 1463 - 1472. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Lambin, R. van Bree,, I. Vergote, and J. Verhaeghe Chronic Tumor Necrosis Factor-{alpha} Infusion in Gravid C57BL6/J Mice Accelerates Adipose Tissue Development in Female Offspring Reproductive Sciences, December 1, 2006; 13(8): 558 - 565. [Abstract] [PDF] |
||||
![]() |
M. I. Lewis, X. Da, H. Li, and M. Fournier Tumor necrosis factor-{alpha} and malnutrition-induced inhibition of diaphragm fiber growth in young rats J Appl Physiol, November 1, 2005; 99(5): 1649 - 1657. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Grohmann, E. Foulstone, G. Welsh, J. Holly, J. Shield, E. Crowne, and C. Stewart Isolation and validation of human prepubertal skeletal muscle cells: maturation and metabolic effects of IGF-I, IGFBP-3 and TNF{alpha} J. Physiol., October 1, 2005; 568(1): 229 - 242. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |