|
|
||||||||
, but not IL-6, stimulates plasminogen activator inhibitor-1 expression in human subcutaneous adipose tissue
The Copenhagen Muscle Research Centre, and The Department of Infectious Diseases, Rigshospitalet, Denmark
Submitted 29 October 2004 ; accepted in final form 25 January 2005
| ABSTRACT |
|---|
|
|
|---|
and IL-6 on PAI-1 gene induction in human adipose tissue. Twenty healthy men underwent a 3-h infusion of either recombinant human TNF-
(n = 8), recombinant human IL-6 (n = 6), or vehicle (n = 6). Biopsies were obtained from the subcutaneous abdominal adipose tissue at preinfusion, at 1, 2, and 3 h during the infusion, and at 2 h after the infusion. The mRNA expression of PAI-1 in the adipose tissue was measured using real-time PCR. The plasma levels of TNF-
and IL-6 reached 18 and 99 pg/ml, respectively, during the infusions. During the TNF-
infusion, adipose PAI-1 mRNA expression increased 2.5-fold at 1 h, 6-fold at 2 h, 9-fold at 3 h, and declined to 2-fold 2 h after the infusion stopped but did not change during IL-6 infusion and vehicle. These data demonstrate that TNF-
rather than IL-6 stimulates an increase in PAI-1 mRNA in the subcutaneous adipose tissue, suggesting that TNF-
may be involved in the pathogenesis of related metabolic disorders. human; subcutaneous adipose tissue; in vivo; gene expression; plasminogen activator inhibitor type I
The cytokines interleukin (IL)-6 and tumor necrosis factor (TNF)-
are produced by various cells, including the adipose tissue (6, 10, 13, 28) and macrophages (3, 30), and exert major regulatory effects on the acute-phase protein C-reactive protein (36). Overexpression of TNF-
occurs in the adipose tissue of both obese animals with insulin resistance (19) and obese humans (20). Adipose tissue secretes IL-6 (28), which is a primary regulator of the acute-phase response (36). TNF-
may stimulate PAI-1 secretion in human adipose tissue fragments (18), and circulating plasma IL-6 is associated with PAI-1 plasma levels (42). Increased levels of both TNF-
and IL-6 are observed in obese individuals (20), in Type 2 diabetes (29), and in patients with atherosclerosis (1). In two population-based studies, plasma concentrations of IL-6 predict all-cause mortality as well as cardiovascular mortality (17, 49). Furthermore, plasma concentrations of IL-6 and TNF-
predict the risk of myocardial infarction in several studies (2, 15, 37). Recently, it was shown that the C-reactive protein level (which is induced by TNF-
and IL-6) is a stronger predictor of cardiovascular events than the LDL cholesterol level and that C-reactive protein adds prognostic information to that conveyed by the Framingham risk score (38).
The facts that TNF-
, IL-6, and PAI-1 are all associated with the metabolic syndrome (22, 52), that TNF-
induces PAI-1 expression in vitro, and that IL-6 is associated with elevated PAI-1 plasma levels (32) stimulated us to test the hypothesis that elevations of TNF-
and IL-6 were involved in the in vivo regulation of PAI-1 in human adipose tissue.
| MATERIALS AND METHODS |
|---|
|
|
|---|
(n = 8), rhIL-6 (n = 6), or vehicle (n = 6). The groups did not differ with regard to age or body mass index. The study was approved by the Ethical Committee of the Copenhagen and Frederiksberg Communities, Denmark, and performed according to the Declaration of Helsinki. Subjects were informed about possible risks and discomfort before they gave their informed, written consent to participate. Subjects had no medical history, and physical examination revealed no abnormalities. The volunteers did not use any medication and did not have any febrile illness in the fortnight preceding the study. Furthermore, subjects abstained from heavy exercise 2 days before the experiments. Protocol. On the day of experiment, subjects arrived at 0800 after an overnight fast.
Eight subjects were infused with rhTNF-
(Beromun) 700 ng·m2·h1, (Boeringer-Ingelheim, Biberach an der Riss, Germany) for 3 h; six subjects were infused with rhIL-6 for 3 h with a constant infusion rate at 5 µg/h (Sandoz, Basle, Switzerland). The rhTNF-
and rhIL-6 were administered in 20% human albumin (Statens Serum Institut, Copenhagen, Denmark). The controls were infused with vehicle (20% human albumin) for 3 h.
Biopsies were obtained from abdominal subcutaneous adipose tissue before infusion, after 1, 2, and 3 h of infusion, and 2 h postinfusion. Biopsies were obtained using the Bergstrøm cannula; the skin was anesthetized using Lidokain (20 mg/ml; Sygehus Apotekerne Danmark, Copenhagen, Denmark). The anesthetic did not contain epinephrine to avoid unintended effects on gene expression. A 5-mm incision was made, and the cannula was introduced into the subcutaneous adipose tissue. Suction was applied, and 510 cuts were made. The biopsy was rinsed using physiological saline to remove blood contamination. The cleaned biopsy was quickly frozen in liquid nitrogen and stored at 80°C. Blood samples were collected before infusion and at 0.5, 1, 2, 3, 4, and 5 h after infusion. The blood was spun, and plasma was stored at 80°C. Subjects were permitted to consume only water during the experiment.
RNA extraction.
Total RNA was extracted from
100 mg of adipose tissue following the procedure for the TRIzol Reagent kit from In Vitrogen. The adipose tissue was homogenized using a Brinkman Polytron (version PT 2100), and a layer of triglycerides collected on the surface was removed. Chloroform was added, and tubes were spun to separate the phenol and aqueous phase. The aqueous phase was transferred to a new tube, and RNA was precipitated by adding isopropyl alcohol and centrifugation. The RNA pellet was washed in 75% ethanol and redissolved in 15 µl of diethyl pyrocarbonate-treated water. The amount of RNA was determined spectrophotometrically at 260 nm.
Reverse transcription. One microgram of RNA was reverse transcribed to single-stranded cDNA using TaqMan reverse-transcription reagents from Applied Biosystems. Random hexamers were used as primers, and the conditions were in agreement with the directions from the manufacturer.
Semi-quantitative real-time PCR. The expression of mRNA was measured using an ABI PRISM 7900HT Sequence Detection System from Applied Biosystem. 18S rRNA was used as a housekeeping reference gene. To determine the 18S rRNA and PAI-1 mRNA levels, a predeveloped primer limited assay from Applied Biosystems was used. The reagents used for the PCR reaction were all obtained from Applied Biosystems, and each sample was run in triplicate in a volume of 10 µl for 50 cycles using standard, real-time cycling conditions. The cycle threshold values were normalized to a relative standard curve run at the same plate as the samples. The standard curves confirmed an acceptable efficiency, and the ratio of amount of PAI-1 mRNA to amount of 18S rRNA was calculated (see Fig. 2).
|
and IL-6 were measured using high sensitivity kits from R & D Systems (Minneapolis, MN). PAI-1 was measured using the Biopool kit from UMEÅ, and cortisol was measured using an enhanced immunoassay kit from Diagnostic Systems Laboratories. All samples were run in duplicates.
Statistics.
All data were logarithmically transformed to obtain a normal distribution. The means are geometric means ± SE. Due to the nonlinear shape of the curves of plasma TNF-
, IL-6, and PAI-1 mRNA, the area under the curve was calculated, and an ANOVA was performed to evaluate the difference between the three groups. The analysis was performed using Excel 2000 from Microsoft. A repeated-measurement ANOVA was used to evaluate the difference over time. A paired t-test with Bonferroni correction was used as a post hoc analysis to compare individual time points with the 0-h sample. Because the plasma PAI-1 curves are linear, a repeated-measurement two-way ANOVA was applied to evaluate the effect of time and difference between groups. Systat version 8.0 for Windows (SPSS) was used for this analysis.
| RESULTS |
|---|
|
|
|---|
and IL-6 levels in response to rhTNF-
and rhIL-6 infusions appear in Fig. 1. The concentration of plasma TNF-
increased in response to rhTNF-
infusion, only reaching a level of
20 pg/ml after 30 min (Fig. 1A). This level was stable until the cessation of the infusion, after which a steep decline was seen. Likewise, plasma concentrations of IL-6 increased only during the rhIL-6 infusion and reached a level of
100 pg/ml (Fig. 1B). TNF-
and IL-6 concentrations did not change in the control trial. The subjects had unchanged temperature, blood pressure, and heart rate during the cytokine infusions (data not shown).
|
infusion, reaching a maximal level of approximately ninefold (P = 0.004) at the end of the infusion (Fig. 2). Two hours after cessation of rhTNF-
infusion, the level of PAI-1 mRNA had almost returned to baseline. Neither rhIL-6 infusion nor vehicle infusion had any effect on the level of PAI-1 mRNA.
PAI-1 plasma levels showed high interindividual variation. PAI-1 plasma levels declined with time. This decline was abolished by TNF-
infusion (2-way, repeated-measures, time x group ANOVA; P = 0.001) but not by IL-6 (2-way, repeated-measures, time x group ANOVA; P = not significant) (Fig. 3). Plasma cortisol levels increased during rhIL-6 infusion compared with placebo (P = 0.000) but did not changes during TNF-
infusion (P = 0.24) (data not shown).
|
| DISCUSSION |
|---|
|
|
|---|
induce PAI-1 mRNA expression in human adipose tissue in vivo and further induce a relative increase in the systemic levels of PAI-1. Thus the decline in plasma PAI-1, owing to circadian changes (8, 43), was counteracted by TNF-
. The finding of a regulatory role of TNF-
on PAI-1 is in accordance with in vitro studies (18, 31, 40). Circulating cortisol levels increased during the IL-6 infusion, as previously demonstrated (44), but did not change in response to TNF-
administration. Therefore, cortisol does not seem to mediate the effect of TNF-
on PAI-1 expression. Previous clinical studies find a correlation between plasma cytokines and PAI-1 (45). However, because cytokines appear in cascades, it has not previously been possible to define the independent role of each cytokine.
PAI-1 is considered to be a risk marker in the metabolic syndrome (22, 25, 34), but indications exist that PAI-1 is also directly involved in the pathogenesis of this disorder. Studies using genetic models of PAI-1 deletion (41) strongly indicate that PAI-1 is involved in adipose tissue accumulation. PAI-1 knock-out mice were protected against high-fat diet, induced obesity, and insulin resistance compared with wild-type mice (23). Recent observations indicate that the p75 TNF receptor plays a role in attenuating TNF-
-induced PAI-1 mRNA expression in acute inflammatory conditions (31). The latter study is compatible with the fact that TNF-
is directly involved in the regulation of PAI-1 and is in accordance with the findings in the present study.
Both TNF-
and IL-6 are elevated in individuals with the metabolic syndrome (22, 52) and in nonacute inflammatory conditions. The major source of these cytokines may be adipose tissue (28). TNF-
is a strong inducer of IL-6 release from adipocytes (39), and circulating levels of IL-6 are associated with markers of the metabolic syndrome (12, 35). Given that TNF-
mainly works locally, TNF-
transcription may or may not be reflected in enhanced systemic levels of TNF-
. However, TNF-
may stimulate IL-6 production and consequently other inflammatory markers. Therefore, chronically elevated levels of IL-6 are likely to reflect local ongoing TNF-
production (33).
The metabolic effect of IL-6 suggests, however, that IL-6 may not be the underlying cause of metabolic disturbances. Thus IL-6 knock-out mice developed late-onset obesity and impaired glucose tolerance (50), and rhIL-6 infusion to healthy people induced lipolysis and fat oxidation (47). In contrast, TNF-
has been shown to impair insulin-stimulated rates of glucose storage in cultured human muscle cells (4) and to impair insulin-mediated glucose uptake in rats (51). Also, obese mice with a gene knock-out of TNF-
are protected from insulin resistance (46). In vitro studies suggest that TNF-
inhibits insulin signaling through inhibition of serine phosphorylation of insulin receptor substrate-1 (21), suppressor of cytokine signaling 3 (7), and STAT5b (48). Data in the literature pointing to TNF-
as being a direct player in insulin resistance are therefore accumulating, whereas IL-6 may have the opposite effect (9, 50). The conclusion from the present data is limited to the acute effect of TNF-
in lean healthy humans. However, the fact that TNF-
, and not IL-6, induces PAI-1 expression may support the theory that TNF-
, rather than IL-6, is the actual "driver" behind the metabolic syndrome.
| 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 |
|---|
|
|
|---|
is associated with dementia in centenarians. J Gerontol A Biol Sci Med Sci 54: M357M364, 1999.[Abstract]
(TNF-
) and atherosclerosis. Clin Exp Immunol 121: 255260, 2000.[CrossRef][Web of Science][Medline]
on glucose metabolism in cultured human muscle cells from nondiabetic and type 2 diabetic subjects. Endocrinology 139: 47934800, 1998.
in the adipose tissue of obese mice. J Biol Chem 276: 4794447949, 2001.
(TNF-
) polymorphisms 857C/A and 863C/A are associated with TNF-
secretion from human adipose tissue. Diabetologia 45: 149150, 2002.[Medline]
and atherosclerosis risk indicators in a healthy family population: the STANISLAS cohort. Atherosclerosis 170: 277283, 2003.[CrossRef][Web of Science][Medline]
and its receptors during drug and dietary modulation of insulin resistance. Endocrinology 134: 264270, 1994.
in human obesity and insulin resistance. J Clin Invest 95: 24092415, 1995.[Web of Science][Medline]
- and obesity-induced insulin resistance. Science 271: 665668, 1996.[Abstract]
, in vivo. J Clin Endocrinol Metab 82: 41964200, 1997.
in the pathogenesis of insulin resistance and type 2 diabetes. Trends Endocrinol Metab 11: 212217, 2000.[CrossRef][Web of Science][Medline]
receptors in the induction of plasminogen activator inhibitor-1. Am J Pathol 162: 933941, 2003.
, overexpressed in human fat cells from insulin-resistant subjects. J Biol Chem 278: 4577745784, 2003.
is a key component in the obesity-linked elevation of plasminogen activator inhibitor 1. Proc Natl Acad Sci USA 96: 69026907, 1999.
function. Nature 389: 610614, 1997.[CrossRef][Medline]
. Diabetes 49: 19041909, 2000.[Abstract]
This article has been cited by other articles:
![]() |
R. C Parish and J. D Evans Inflammation in Chronic Heart Failure Ann. Pharmacother., July 1, 2008; 42(7): 1002 - 1016. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. L. Holland and S. A. Summers Sphingolipids, Insulin Resistance, and Metabolic Disease: New Insights from in Vivo Manipulation of Sphingolipid Metabolism Endocr. Rev., June 1, 2008; 29(4): 381 - 402. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Plomgaard, C. P. Fischer, T. Ibfelt, B. K. Pedersen, and G. van Hall Tumor Necrosis Factor-{alpha} Modulates Human in Vivo Lipolysis J. Clin. Endocrinol. Metab., February 1, 2008; 93(2): 543 - 549. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Bruunsgaard Physical activity and modulation of systemic low-level inflammation J. Leukoc. Biol., October 1, 2005; 78(4): 819 - 835. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Plomgaard, K. Bouzakri, R. Krogh-Madsen, B. Mittendorfer, J. R. Zierath, and B. K. Pedersen Tumor Necrosis Factor-{alpha} Induces Skeletal Muscle Insulin Resistance in Healthy Human Subjects via Inhibition of Akt Substrate 160 Phosphorylation Diabetes, October 1, 2005; 54(10): 2939 - 2945. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |