|
|
||||||||
and IL-6 infusions on insulin sensitivity and expression of IL-18 in humans
1Centre of Inflammation and Metabolism, Department of Infectious Diseases, and Copenhagen Muscle Research Centre, Rigshospitalet University of Copenhagen, Faculty of Health Sciences, Copenhagen, Denmark; and 2Washington University School of Medicine, St. Louis, Missouri
Submitted 28 September 2005 ; accepted in final form 1 February 2006
| ABSTRACT |
|---|
|
|
|---|
and interleukin (IL)-6 may affect glucose uptake. TNF induces insulin resistance, whereas the role of IL-6 is controversial. High plasma levels of IL-18 are associated with insulin resistance in epidemiological studies. We investigated the effects of TNF and IL-6 on IL-18 gene expression in skeletal muscle and adipose tissue. Nine human volunteers underwent three consecutive interventions, receiving an infusion of recombinant human (rh)IL-6, rhTNF, and saline. Insulin sensitivity was assessed by measurement of whole body glucose uptake with the stable isotope tracer method during a euglycemic hyperinsulinemic clamp (20 mU·min1·kg1), which was initiated 1 h after the IL-6-TNF-saline infusion. Cytokine responses were measured in plasma, muscle, and fat biopsies. Plasma concentrations of TNF and IL-6 increased 10- and 38-fold, respectively, during the cytokine infusions. Whole body insulin-mediated glucose uptake was significantly reduced during TNF infusion but remained unchanged during IL-6 infusion. TNF induced IL-18 gene expression in muscle tissue, but not in adipose tissue, whereas IL-6 infusion had no effect on IL-18 gene expression in either tissue. We conclude that TNF-induced insulin resistance of whole body glucose uptake is associated with increased IL-18 gene expression in muscle tissue, indicating that TNF and IL-18 interact, and both may have important regulatory roles in the pathogenesis of insulin resistance.
cytokines; diabetes; inflammation; tumor necrosis factor-
; interleukin-6; interleukin-18
and interleukin (IL)-6 are closely linked to obesity and insulin resistance (5, 74). Plasma concentration of TNF is elevated (33, 44), and TNF is overexpressed in adipose (30, 36) and skeletal muscle (65) tissue in obese and insulin-resistant humans. The role of TNF in insulin resistance has remained controversial. Despite compelling evidence in animal models, attempts to neutralize TNF-
through the administration of a human anti-TNF-
antibody or recombinant TNF-
receptor failed to improve insulin sensitivity in obese type 2 diabetic patients (51, 54). However, recently we and others (60, 62) have shown that TNF infusion impairs glucose uptake in human skeletal muscle by altering insulin signal transduction. Chronically elevated systemic levels of IL-6 are also associated with obesity and insulin resistance (36, 47, 56, 69). IL-6 is overexpressed in adipose tissue in obese humans (21, 26, 45), and IL-6 gene expression is increased in rat muscle tissue (7) and in adipose tissue in humans (37) after insulin stimulation. The role of IL-6 in the development of insulin resistance, however, remains controversial (57). While Di Gregorio et al. (12) found that IL-6 knockout mice do not develop obesity, Wallenius et al. (71) reported that such mice develop impaired glucose tolerance and obesity, which are partially reversed by IL-6 treatment. In addition, we recently demonstrated that even an
100-fold increase in IL-6 levels during postabsorptive conditions does not alter the rate of appearance or disappearance of glucose in humans (68), suggesting that IL-6 per se is unlikely to be involved in the development of insulin resistance in humans.
IL-18, first described in 1989 by Nakamura et al. (48) as an interferon (IFN)-
-inducing factor, is a proinflammatory cytokine belonging to the IL-1 family (16). IL-18 production is induced by a variety of macrophage stimulators, i.e., endotoxin, exotoxins from gram-positive bacteria, and IFN-
(15, 28, 48, 49), and it has recently been shown that TNF stimulates the expression of IL-18 in cardiomyocytes (8). IL-18 is produced in a range of cells, including macrophages, monocytes, epithelial cells, endothelial cells, dendritic cells, keratinocytes, cardiomyocytes, microglial cells, and astrocytes (8, 11, 14, 49, 63) and is also found in different tissues such as rat adrenal cortex (10), human adipose tissue (42), and human skeletal muscle tissue (27). IL-18 is synthesized as a precursor molecule, pro-IL-18, which is cleaved and activated by the enzyme Caspase-1 before it is secreted (13, 22). IL-18 stimulates both Th1 (IFN-
) and Th2 (IL-4, IL-13) responses but can also directly activate nuclear factor-
B (NF-
B) and induce production of different cytokines such as TNF, IL-6, IL-8, and IL-1b (32, 49, 50, 61). IL-18 expression is increased during different autoimmune, inflammatory, and infectious diseases (14, 41, 49, 52, 63), and more recently, IL-18 has also been linked to metabolic diseases such as type 2 diabetes (3, 19, 25, 46), obesity (20), the polycystic ovary syndrome (18), atherosclerosis (17, 43), and the metabolic syndrome (31).
Because high systemic levels of both TNF and IL-18 are found in individuals with insulin resistance, we hypothesized that TNF mediates insulin resistance in humans partially by inducing an increase in IL-18 expression in muscle and fat tissue, whereas IL-6 would induce neither insulin resistance nor expression of IL-18.
To investigate these hypotheses, nine healthy volunteers underwent three consecutive trials, during which a euglycemic low-dose insulin clamp, combined with [6,6-2H2]glucose infusion, was superposed upon a 4-h infusion of recombinant human (rh)IL-6, rhTNF, or saline. We obtained blood samples and muscle and adipose tissue biopsies to measure glucose kinetics and cytokine expression. Gene expression of caspase-1 was measured as an indirect marker for IL-18 protein production within the muscle (13, 22).
| MATERIALS AND METHODS |
|---|
|
|
|---|
The characteristics of the nine volunteers have previously been described (60). In short, nine healthy human males [mean age 26.3 (range 2135) yr; body mass index 23.9 (range 22.525.8) kg/m2] with unremarkable medical pasts were included after oral and written informed consent was obtained. Before the study, all nine subjects underwent a thorough clinical examination and blood samples for evaluation of renal, hepatic and thyroid function, hemoglobin, white blood cell counts, electrolytes, insulin, and plasma glucose. All tests were normal. The study was approved by the Scientific Ethics Committee of Copenhagen and Frederiksberg Municipalities (No. KF-01-006/04), in accordance with the Helsinki Declaration.
Study Design
All subjects underwent three separate trials (
1 wk apart), each consisting of a 4-h infusion of saline (trial A, control), rhTNF (trial B), and rhIL-6 (trial C). To monitor insulin sensitivity in all three trials, [6,6-2H2]glucose was infused as previously described (60). Briefly, a priming dose (22 µmol/kg) of the glucose tracer was given and immediately followed by constant infusions of 0.22 µmol·kg1·min1 for 2 h during basal postabsorptive conditions. During the clamp, the infusion rate was reduced to 0.11 µmol·kg1·min1. Infusions of IL-6, TNF, or saline were initiated at 1 h, and the euglycemic hyperinsulinemic clamp at 2 h, after the start of the stable isotope tracer infusion. The trials were randomized as follows: A, B, C (n = 4); A, C, B (n = 1); C, A, B (n = 2); B, C, A (n = 1); B, A, C (n = 1). Data on insulin resistance have been previously published for trials A and B (60). For the present study, we included data from all three trials.
For each trial, subjects reported to the laboratory at 8:00 AM after an overnight fast. Peripheral catheters were placed in an antecubital vein for blood sampling, in the contralateral antecubital vein for infusion of fluids and medication, and in a dorsal hand vein for blood sampling. The catheterized hand was wrapped in a heating blanket to obtain arterialized venous blood for measurement of glucose and potassium during the clamp. The ECG was continuously monitored; heart rate, noninvasive blood pressure, and tympanic temperature were recorded every hour during the trials. After catheterization, the infusion of the glucose tracer was initiated (time point 120 min). At time point 60 min, infusion of placebo, rhTNF [rate: 1,000 ng·h2·m2 body surface area (BSA); Beromun, Boeringer Ingelheim], or rhIL-6 (rate: 0.47 µg·h2·m2 BSA; Sandoz, Basle, Switzerland) was initiated. RhTNF and rhIL-6 were administered in saline with 20% human albumin; placebo consisted of saline with 20% human albumin. At time point 0 min, the clamp was initiated. Insulin (Actrapid; Novo Nordisk Insulin, Copenhagen, Denmark; 100 IE/ml) was infused continuously at a rate of 20.0 mU·min2·m2 BSA, and the plasma glucose concentration was kept at 5.0 mM by a coinfusion of glucose (200 g/1,000 ml, enriched to 2.5% with the glucose tracer) (24) at a variable rate. Isotonic saline with or without potassium, as appropriate, was infused continuously during the study to maintain plasma potassium concentration within a normal range. Arterialized blood was analyzed for glucose and potassium concentrations at intervals of 5 min during the first hour, and every 10 min during the last 2 h of the clamp these last samples were analyzed for the glucose isotope enrichment as well, as were blood samples taken at time points 20, 10, and 0 min (baseline study period). Venous samples for measurement of cytokines, insulin, and cortisol were drawn at baseline (time points 120 and 60 min), before the start of the clamp (time point 0 min), and at time points 30, 60, 120, and 180 min. Muscle biopsies obtained from musculus quadriceps, vastus lateralis, and abdominal subcutaneous adipose tissue were taken at time points 0, 30, 60, and 180 min. The data obtained from biopsies at time point 0 in the control trial (saline infusion only) are used as baseline data for all trials because the first biopsies during trials B and C were obtained after 60 min of infusion of rhIL-6 or rhTNF.
Measurements
All blood samples were drawn into tubes containing EDTA and immediately centrifuged. Plasma was stored at 80°C until analyzed.
Cytokines. Plasma concentrations of IL-18, IL-6, and TNF were measured by enzyme-linked immunosorbent assay (ELISA: IL-18; MBL, Naka-ku Nagoya, Japan; IL-6 and TNF: high-sensitivity kits; R&D Systems, Minneapolis, MN). Samples were analyzed in duplicate, and mean concentrations were calculated for each sample.
Arterialized blood. Whole blood glucose and potassium concentrations were measured immediately on an EML 105 (Radiometer, Copenhagen, Denmark).
Isotope enrichment. Plasma glucose enrichment was measured using liquid chromatography-mass spectrometry aQa (Finnigan, Thermoquest, Manchester, UK), as previously described (60).
Hormones. Plasma concentrations of insulin and cortisol were analyzed using the ELISA technique (insulin; DAKO, Glostrup, Denmark; and cortisol; DSL, Webster, Texas).
Tissue biopsies. Abdominal subcutaneous adipose tissue and muscle tissue samples were obtained by use of the percutaneous biopsy technique with suction and were immediately frozen in liquid nitrogen and stored at 80°C until further analysis.
RNA extraction. Adipose and muscle tissue RNA extraction was performed with TRIzol (Life Technologies) according to the manufacturer's directions. In brief, 50100 mg of the tissue were dissolved in 1 ml of TRIzol and homogenized with a Brinkman Polytron (version PT 2100; Kinematica, Luzern, Switzerland) rotating at 26,000 rpm. The muscle tissue samples were directly transferred to a fresh tube, 100 µl of chloroform-isoamyl alcohol (24:1) were added, and the tubes were vigorously shaken. The adipose tissue samples were allowed to sit for a few minutes for a triglyceride phase to form. The lower aqueous phase was then transferred to a tube, 100 µl of chloroform-isoamyl alcohol (24:1) were added, and the tubes were shaken. Samples were allowed to sit for 5 min and were subsequently spun at 13,000 rpm for 15 min at 4°C, after which the upper aqueous phase was transferred to a new tube. The aqueous phase was mixed with 0.5 ml of ice-cold isopropanol, and samples were placed at 20°C for 1 h. They were then centrifuged at 13,000 rpm for 15 min at 4°C, and the resulting pellet was washed with 0.5 ml of cold 75% ethanol in diethylpyrocarbonate (DEPC)-treated water (0.05%). After centrifugation at 8,000 rpm for 10 min, pellets were redissolved in 15 µl of DEPC-treated water and allowed to dissolve on ice. RNA was dissolved in DEPC water, and then the concentration of RNA was measured spectrophotometrically at optical density 260.
Reverse transcription. One microgram of total RNA was reverse transcribed using the Applied Biosystems Taqman RT kit, and random hexamers were used as primers.
Real-time PCR. All PCR reagents were obtained from Applied Biosystems.
Gene expression of IL-18, IL-6, TNF, and caspase-1 was analyzed using semiquantitative real-time PCR on an ABI PRISM 7900 HT sequence detector (Applied Biosystems). 18S served as the internal reference gene. We used the predeveloped, primer-limited assay reagents for 18S rRNA, IL-18, TNF, and caspase-1 determination. The IL-6 primers and probe sequences used were obtained from Starkie et al. (66).
An 81-bp fragment was amplified using IL-6 forward primer 5'-GGTACATCCTCGACGGCATCT-3', IL-6 reverse primer 5'-GTGCCTCTTTGCTGCTTTCAC-3', and IL-6 probe 5'-FAM-TGTTACTCTTGTTACATGTCTCCTTTCTCAGGGCT-TAMRA-3'.
A reagent mixture of 75 µl was made up for each sample with 1x MasterMix, 900 nM IL-6 forward primer, 300 nM reverse primer, 100 nM IL-6 probe, 1x 18S, TNF, or IL-18 mix (primers and probe), and 50100 ng of sample and made up into a final volume of 75 µl with water. Each sample was run in triplicates in a reaction volume of 10 µl for 50 cycles using standard real-time PCR cycling conditions. All samples were run in triplicates and normalized to a relative standard curve run at the same plate as the samples, and the transcript quantity of genes was then normalized to 18S rRNA.
Calculations
Endogenous glucose production [rate of appearance (Ra) in plasma] and glucose uptake [rate of disappearance (Rd) from plasma] were calculated using a single-pool non-steady-state model and Steele's equation (24, 67).
Statistics
Data were tested for normal distribution by the Kolmogorov-Smirnov analysis a.m. Lillifors. Plasma levels of insulin, glucose, and cortisol were normally distributed, as were the clinical values and the stable isotope data, on the basis of area under the curve (AUC) of basal (20, 10, and 0 min), and insulin-stimulated (60180 min) rates of whole body glucose appearance and disappearance. Data were analyzed using parametric methods on the absolute data, and reported values are means ± SE. Plasma cytokine (IL-18, TNF, and IL-6) concentrations, as well as gene expression (expressed as an mRNA-to-18S rRNA ratio) of IL-18, TNF, and IL-6 from muscle and adipose tissue and caspase-1 from muscle tissue, were normally distributed when they were log-transformed before analysis. These data were analyzed using parametric methods, and data are expressed as geometric mean (95% confidence interval).
Parametric methods. Within-subject variation over time and variation between trials were analyzed using a repeated-measures (two-way ANOVA, time-by-trial) approach followed by Bonferroni-corrected paired t-tests as appropriate to identify significant differences between trials. If the two-way ANOVA revealed a significant variation over time, variation over time for each trial was then analyzed using one-way ANOVA followed by Bonferroni-corrected paired t-tests to identify significant differences from baseline values.
Differences in AUC of glucose Ra and Rd were analyzed with Bonferroni-corrected paired t-tests.
A P value <0.05 was considered statistically significant. Analysis was performed using a statistical software package (SAS version 9.1).
| RESULTS |
|---|
|
|
|---|
During the hyperinsulinemic euglycemic clamp, insulin levels increased significantly, with no difference between trials, and blood glucose concentrations were kept constant at 5.0 ± 0.09 mM, with no difference between trials (two-way ANOVA P > 0.05). Plasma levels of cortisol revealed a well-described circadian rhythm (64). TNF infusion enhanced cortisol levels (two-way ANOVA P < 0.05; significant difference at 0 and 30 min, paired t-test). There was no difference in plasma cortisol concentration between the control and the IL-6 trial (Table 1).
|
infusion was much smaller than that induced by IL-6 infusion. By contrast, plasma levels of IL-18 did not differ between trials (two-way ANOVA, P > 0.05). Although the one-way ANOVA suggested an increase in plasma levels of IL-18 during the control trial, no significant time points were identified by the post hoc paired t-test. (Fig. 1C).
|
|
|
There was a significant difference between trials in mRNA levels of caspase-1 (two-way ANOVA P = 0.001). This was due to a difference between the IL-6 and the TNF trial (P < 0.001) and with a borderline significant difference between the control and the TNF trial (P = 0.055); there was no variation over time (Fig. 3B).
The mRNA levels of IL-6 did not differ between trials, but the two-way ANOVA revealed a significant variation over time. The subsequent one-way ANOVA identified an increase in IL-6 mRNA during the TNF trial at 180 min and during the IL-6 trial at 60 and 180 min compared with baseline (Fig. 3C).
TNF mRNA levels were not induced by either rhTNF or rhIL-6 infusion (Fig. 3D). In adipose tissue, mRNA levels of IL-18, IL-6, or TNF did not differ between trials (data not shown).
| DISCUSSION |
|---|
|
|
|---|
We recently reported that an eightfold increase in plasma levels of TNF downregulates insulin signaling and whole body glucose uptake in humans because of a decreased insulin-mediated glucose uptake in the skeletal muscle (Rd), and with an unchanged endogenous glucose production (Ra) (60). The present study demonstrates that a 34-fold increase in plasma levels of IL-6, concomitantly with a modest increase in plasma insulin levels, does not affect whole body glucose metabolism. This is in agreement with an earlier study from our group, in which an
100-fold increase in IL-6 levels at a fasting state did not alter rate of appearance or disappearance of glucose in humans (68). Taken together, these data support our hypothesis that TNF and IL-6 influence insulin sensitivity differentially, at least after acute administration. However, it may not be justified to extrapolate our findings to conditions in which cytokine levels are more chronically elevated, like type 2 diabetes and obesity (6, 55, 58, 59).
Given the important role of skeletal muscle and adipose tissue in glucose uptake, a major aim was to explore the interaction between TNF, IL-6, and IL-18 in these tissues. An in vitro study has revealed that TNF induces IL-18 production in cardiomyocytes via NF-
B (8). The present study demonstrates that TNF induces IL-18 gene expression in vivo in human skeletal muscle. Because TNF, and not IL-6, induces IL-18 gene expression, this could be related to TNF-induced NF-
B activation in the present study as well, especially because IL-6 is known not to activate this pathway (53). The increased gene expression of IL-18 is not necessarily related to an increased protein level of this cytokine, but the borderline significant elevation of the gene expression of caspase-1 in the TNF trial vs. the control trial might indicate that the protein has in fact been synthesized. Although the increased IL-18 gene expression did not lead to a detectable increase in systemic levels of IL-18, we suggest that IL-18 may be instrumental in TNF-induced inhibition of glucose uptake in muscle. This suggestion is supported by the fact that IL-18 (1, 70), like TNF (4, 9), activates c-Jun NH2-terminal kinase (JNK), which phosphorylates insulin receptor substrate-1 at serine 307 and thereby inhibits insulin action (2, 40). JNK is, furthermore, suggested to be responsible for obesity-induced insulin resistance in vivo (29). However, definitive proof of the role of IL-18 in TNF-associated insulin resistance would require either infusion of IL-18 or a specific inhibitor of this cytokine.
The finding that IL-6 and TNF enhance IL-6 mRNA levels in skeletal muscle is in accordance with findings in rodents, which demonstrates that endotoxin induces IL-6 expression (39), and previous studies from our group have revealed that muscle-IL-6 is also regulated in an autocrine fashion (34, 35).
In vitro studies have shown that TNF and IL-6 are regulated in an autocrine fashion in adipocytes. Thus TNF stimulates the expression of TNF and IL-6 (23, 72), and IL-6-stimulated adipocytes express IL-6 (23, 38). However, contrary to our hypothesis, cytokine infusion did not affect cytokine gene expression in subcutaneous adipose tissue. Obesity is associated with an increased number of macrophages in adipose tissue (73, 75), and Fain et al. (21) have recently suggested that TNF and, in part, IL-6 are released by the nonfat cells of the adipose tissue. Healthy, lean males were used in the present study, and this could be the explanation for the absent cytokine response in this tissue.
In conclusion, the finding that IL-18 mRNA is expressed in human skeletal muscle and induced by TNF suggests a possible role for IL-18 in the genesis of TNF-induced insulin resistance in skeletal muscle.
| GRANTS |
|---|
|
|
|---|
.
| 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 |
|---|
|
|
|---|
gene expression in human subcutaneous adipose tissue. Am J Physiol Endocrinol Metab 286: E234E238, 2004.
. Am J Physiol Endocrinol Metab 288: E731E740, 2005.This article has been cited by other articles:
![]() |
J. M Bruun, B. Stallknecht, J. W Helge, and B. Richelsen Interleukin-18 in plasma and adipose tissue: effects of obesity, insulin resistance, and weight loss Eur. J. Endocrinol., October 1, 2007; 157(4): 465 - 471. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Musaad and E. N. Haynes Biomarkers of Obesity and Subsequent Cardiovascular Events Epidemiol. Rev., May 10, 2007; (2007) mxm005v1. [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] |
||||
![]() |
E. E. Spangenburg, S. E. Shoelson, C. Weigert, R. Lehmann, E. D. Schleicher, J.-O. Jansson, V. Wallenius, J.-P. Bastard, C. Lagathu, M. Caron, et al. Interleukin-6 does/does not have a beneficial role in insulin sensitivity and glucose homeostasis J Appl Physiol, February 1, 2007; 102(2): 820 - 823. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |