Ciliary Neurotrophic Factor Prevents Acute Lipid- Induced Insulin ...

0013-7227/06/$15.00/0 Printed in U.S.A.

Endocrinology 147(5):2077–2085 Copyright © 2006 by The Endocrine Society doi: 10.1210/en.2005-1074

Ciliary Neurotrophic Factor Prevents Acute LipidInduced Insulin Resistance by Attenuating Ceramide Accumulation and Phosphorylation of c-Jun N-Terminal Kinase in Peripheral Tissues Matthew J. Watt, Andrea Hevener, Graeme I. Lancaster, and Mark A. Febbraio Cellular and Molecular Metabolism Laboratory (M.J.W., G.I.L., M.A.F.), School of Medical Sciences, RMIT University, Bundoora, Victoria 3083, Australia; and Department of Medicine (A.H.), University of California, San Diego, La Jolla, California 92093 Ciliary neurotrophic factor (CNTF) is a member of the gp130 receptor cytokine family recently identified as an antiobesity agent in rodents and humans by mechanisms that remain unclear. We investigated the impact of acute CNTF treatment on insulin action in the presence of lipid oversupply. To avoid confounding effects of long-term high-fat feeding or genetic manipulation on whole-body insulin sensitivity, we performed a 2-h Intralipid infusion (20% heparinized Intralipid) with or without recombinant CNTF pretreatment (Axokine 0.3 mg/kg), followed by a 2-h hyperinsulinemic-euglycemic clamp (12 mU/kg䡠min) in fasted, male Wistar rats. Acute Intralipid infusion increased plasma free fatty acid levels from 1.0 ⴞ 0.1 to 2.5 ⴞ 0.3 mM, which subsequently caused reductions in skeletal muscle (insulin-stimulated glucose disposal rate)

S

KELETAL MUSCLE insulin resistance is a key factor in the pathogenesis of type 2 diabetes and is a metabolic disorder that precedes overt diabetes by several decades. Elevated plasma free fatty acid (FFA) levels are often observed in human and rodent models of type 2 diabetes or insulin resistance, and it has been shown that acute elevation of FFA levels, achieved via heparinized infusion of a triglyceride emulsion, leads to skeletal muscle and hepatic insulin resistance (1–3). An important consequence of chronically elevated circulating FFA levels is the accumulation of fatty acid metabolites within insulin-responsive tissues. Accumulation of these lipid intermediates, including ceramides and diacylglycerol, also correlate with insulin resistance (4, 5). The production of fatty acid metabolites within insulin-responsive tissues is known to activate a host of serine kinases including protein kinase C␪, IKK-nuclear factor ␬B (NF␬B), and c-Jun N-terminal kinase (JNK), which have been shown to impair insulin signal transduction and glucose disposal in vitro (3, 6). First Published Online January 5, 2006 Abbreviations: CNTF, Ciliary neurotrophic factor; FFA, free fatty acid; HF, high fat; HGP, hepatic glucose production; IRS, insulin receptor substrate; IS-GDR, insulin-stimulated glucose disposal rate; JNK, c-Jun N-terminal kinase; MLK3, mixed linage kinase 3; NF␬B, nuclear factor ␬B; PI3, phosphatidylinositol 3. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

and liver (hepatic glucose production) insulin sensitivity by 30 and 45%, respectively. CNTF pretreatment completely prevented the lipid-mediated reduction in insulin-stimulated glucose disposal rate and the blunted suppression of hepatic glucose production by insulin. Although lipid infusion increased triacylglycerol and ceramide accumulation and phosphorylation of mixed linage kinase 3 and c-Jun N-terminal kinase 1 in skeletal muscle, CNTF pretreatment prevented these lipid-induced effects. Alterations in hepatic and muscle insulin signal transduction as well as phosphorylation of cJun N-terminal kinase 1/2 paralleled alterations in insulin sensitivity. These data support the use of CNTF as a potential therapeutic means to combat lipid-induced insulin resistance. (Endocrinology 147: 2077–2085, 2006)

Given that insulin sensitization is a key therapeutic strategy to combat type 2 diabetes, our laboratories have been actively investigating mechanisms to ameliorate insulin resistance by promoting fatty acid use in skeletal muscle. To this end, in the current investigation, we studied the impact of ciliary neurotrophic factor (CNTF) administration on acute lipid-induced insulin resistance. CNTF is a member of the gp130 cytokine family and was first identified as a trophic factor for motor neurons in the ciliary ganglion (7), and later as an antiobesity agent in rodents and humans (3, 6, 8 –10). Because obesity is highly correlated with type 2 diabetes and because obese CNTF-treated mice exhibit an improved metabolic profile, (i.e. reduced basal levels of blood glucose, insulin, FFA, and triglycerides and improved glucose tolerance) (8, 9, 11, 12), we examined the mechanism(s) by which CNTF improves metabolism. Although other studies have described the effects of CNTF administration on improving insulin signaling in 3T3-L1 adipocytes (13) and steatotic hepatocytes isolated from diabetic mice (11), we investigated the impact of CNTF administration on skeletal muscle insulin action and fatty acid metabolism because skeletal muscle accounts for the majority (75–95%) of whole-body insulinstimulated glucose disposal (14) and 25% of whole-body FFA uptake (15). The CNTF receptor ␣ heterodimerizes with the transmembrane gp130 receptor upon activation, and this heterodimer is highly expressed in skeletal muscle (16). Moreover, the gp130 receptor shares significant sequence homology with

2077

2078

Endocrinology, May 2006, 147(5):2077–2085

the long form leptin receptor in that it contains a Src homology 2 domain capable of activating the Jak/Stat signaling pathway after ligand binding. It is well known that leptin increases fat oxidation and prevents insulin resistance (17); therefore, we hypothesized that activation of the CNTFR␣ would induce similar metabolic actions to leptin, namely increased fatty acid oxidation (18) and reduced fatty acid deposition in skeletal muscle (19). Therefore, the aim of the present study was to elucidate the role of CNTF on insulin action in vivo and correlate these measures with analyses of skeletal muscle fat metabolite accumulation. To avoid confounding variables of long-term high-fat feeding or genetic manipulation, we imposed a lipid challenge via acute heparinized Intralipid infusions with or without exogenous CNTF treatment. After the lipid challenge, hyperinsulinemic-euglycemic clamp studies were performed to assess whole-body insulin sensitivity, and tissues subsequently were excised for analysis of insulin signal transduction and fatty acid metabolite accumulation. In this study, we report that although lipid overload caused marked accumulation of lipid intermediates, decrements in insulin signaling, and increased proinflammatory signaling in muscle and liver, CNTF pretreatment completely prevented these deleterious effects of lipid oversupply. Materials and Methods Animals Male Wistar rats (Monash Animal Services, Clayton, Australia) aged 13–14 wk were housed two per cage at 22 ⫾ 1 C under a 12-h light, 12-h dark cycle. Rats were fed a standard chow diet (Purina Rodent Chow, 3% calories from fat, 77% from carbohydrate and 20% from protein; Purina, St. Louis, MO) and water was made available ad libitum. Rats were housed for at least 1 wk before experimental procedures and were acclimated daily to human handling to reduce stress. All experimental procedures were approved by the RMIT Animal Ethics Committee and were in accordance with the National Health and Medical Research Council Guidelines on Animal Experimentation.

Surgical procedures Three days before the glucose clamp, all animals were chronically cannulated under single-dose anesthesia (100 mg/ml ketamine HCl, 20 mg/ml xylazine, 10 mg/ml acepromazine maleate; 1.5:3:1; 0.1 cc/100 g body weight administered im) in the jugular vein for infusion (glucose, [3H]glucose tracer, and insulin) [dual cannula; SILASTIC brand (Dow Corning Corp., Midland, MI), inside diameter ⫽ 0.03 cm] and in the carotid artery (Clay Adams; PE-50) for arterial blood sampling. Cannulae were tunneled sc, exteriorized at the back of the neck, and encased in SILASTIC brand tubing (0.2 cm inside diameter) sutured to the skin. Animals were allowed 3 d of postoperative recovery and were fasted 12 h before the clamp experiment. To determine whether CNTF was effective at overcoming acute lipid-induced insulin resistance, rodents were randomly divided into one of three study groups: 1) control (Con) saline no CNTF, 2) Intralipid infusion no CNTF, and 3) CNTF treatment (0.3 mg/kg) ⫹ Intralipid infusion.

Hyperinsulinemic-euglycemic clamp experiments Animals were weighed and placed in a modified metabolic chamber at 140 min before the clamp. After 20 min, a basal sample was drawn from the carotid line and plasma was stored at ⫺20 C for subsequent analyses. A primed (5 ␮Ci) constant infusion (0.16 ␮Ci/min) of d-[33 H]glucose (Amersham Biosciences, Buckinghamshire, UK) was then administered. After a 60-min tracer equilibration period and basal sampling at 15 min before and immediately before lipid infusion (to determine basal glucose turnover), either saline ⫹ tracer (0.16 ␮Ci/min) or

Watt et al. • CNTF Prevents Lipid-Induced Insulin Resistance

heparinized (40 U/ml) 20% lipid emulsion ⫹ tracer (Intralipid; Monash Medical Centre, Clayton, Australia) was infused at a rate of 0.0167 ml/min for 120 min. The total volume of Intralipid ⫹ heparin infused was 2.0 ml. Although others have typically infused lipid emulsions and heparin for 3–5 h to induce insulin resistance (3, 20 –22), preliminary work from our laboratory demonstrated insulin resistance after 2 h of infusion and more reliable steady-state glucose infusion rates during the subsequent hyperinsulinemic-euglycemic clamp. Where appropriate, the CNTF analog Axokine, a gift from Regeneron Pharmaceuticals, was administered slowly as a bolus injection (0.3 mg/kg) into the carotid line and flushed into the animal with saline just before the constant infusions of Intralipid. Axokine is not identical to human CNTF as Gln63 is replaced by Arg63, which appears to make Axokine more specific for the CNTF receptor (23). Axokine at this dose does not induce cachexia, stress, inflammation, or nausea in rodents (8). During the saline or Intralipid infusion period preclamp, samples were drawn at 60 and 120 min for the measurement of FFA concentration. After this, the lipid or saline infusion was terminated and the euglycemic-hyperinsulinemic clamp was initiated. This infusion protocol was designed to increase intracellular lipid content in peripheral tissues, yet ensure low circulating fatty acid levels during the clamp procedure. Unlike previous studies that have implemented a triglyceride-heparin infusion throughout the hyperinsulinemic-euglycemic clamp, stopping the triglycerideheparin infusion before the clamp allowed us to discriminate between events mediated within the muscle from events mediated by pharmacologically high circulating FFAs. Moreover, because glucose provides almost all of the substrate during hyperinsulinemic-euglycemic clamp, we did not expect a reduction in tissue lipids. After the saline or lipid infusion period, animals were each exposed to the same general glucose clamp protocol. Glucose (variable infusion; 50% dextrose; Abbott Labs, Chicago, IL) and tracer ⫹ insulin (16.7 ␮l/min) infusions were initiated simultaneously. Small blood samples (5 ␮l) were drawn at 10-min intervals and immediately analyzed for glucose (HemoCue, Angelholm, Sweden) to maintain the integrity of the glucose clamp throughout the duration of the experiment. Animals were clamped at a target glycemia 6 mm, and there were no differences in clamp glycemia between the groups (P ⫽ 0.72). Steady-state was achieved in most animals by 60 min and maintained until termination of the clamp at 120 min. Arterial samples (100 ␮l) for measurement of glucose turnover were drawn at 100 and 120 min. Specific activity of radiolabeled tracer in the blood was not different between the two time points (P ⫽ 0.35). Given that arterial blood glucose did not fluctuate more than 5% and that glucose infusion rate was held constant for a minimum of 30 min, we are confident that stringent steady-state conditions were achieved. After the blood sample for determination of glucose turnover was drawn, an additional 500-␮l sample of blood was taken from the carotid line for the subsequent analysis of FFA and steady-state insulin concentrations. Animals were then euthanized with a lethal dose of sodium pentobarbital (Nembutal, 100 mg/kg) administered iv. Tissue samples were rapidly excised and snap frozen for subsequent metabolic analyses.

Plasma metabolite and hormone analysis Blood was analyzed for fasting concentrations of glucose, insulin, and FFA. Plasma was obtained by centrifugation of whole blood and stored at ⫺80 C until analysis. Blood glucose levels were monitored with the MediSense2a¨ Blood Glucose Testing System (MediSense Australia Pty. Ltd., Victoria, Australia). For determination of plasma metabolites and hormones, blood was withdrawn, immediately spun at 4 C (5 min at 5000 ⫻ g), and rapidly frozen to limit lipoprotein lipase activation and, thereby, prevent triglyceride degradation, which would result in overestimation of plasma FFA levels. Plasma insulin levels were determined using enzyme RIA kits (Linco, St. Charles, MO), and plasma FFA were measured using an enzymatic colorimetric assay kit (Roche, Basel, Switzerland).

Western blot analysis Soleus muscle and liver were homogenized in ice-cold buffer consisting of 20 mm HEPES, 1 mm dithiothreitol, 1 mm Na4P2O7, 2 mm EDTA, 1% Triton X-100, 10% glycerol (vol/vol), 3 mm benzamidine, 1 mm phenylmethylsulfonyl fluoride, 5 ␮l/ml phosphatase inhibitor cock-


tail 2 (Sigma, St. Louis, MO), 5 ␮l/ml protease inhibitor cocktail (Sigma) and rotated for 40 min at 4 C. Homogenates were centrifuged at 16,000 ⫻ g for 25 min and the supernatant was removed and rapidly frozen in liquid nitrogen. Protein concentration of the lysates was determined according to the bicinchoninic acid method (Pierce Kit; Progen Industries, Darra, Queensland, Australia). Tissue lysates (1 mg) were immunoprecipitated with anti-insulin receptor substrate (IRS)-1 or anti-IRS-2 antibody (Upstate, Lake Placid, NY) for 16 h and solubilized in Laemmli sample buffer and boiled for 5 min, resolved by SDS-PAGE on polyacrylamide gels, transferred to a nitrocellulose membrane, blocked with 5% milk, and immunoblotted with anti-IRS-1 tyrosine (Upstate) or antip85 (Cell Signaling Technology, Beverly, MA) primary antibody overnight. After incubation with horseradish peroxidase-conjugated secondary antibody (Amersham Biosciences, Castle Hill, New South Wales, Australia), the immunoreactive proteins were detected with enhanced chemiluminescence (Amersham Biosciences, Castle Hill, New South Wales, Australia) and quantified by densitometry (ChemiDoc XRS; BioRad Laboratories, Regents Park, New South Wales, Australia). Membranes were stripped, washed, and reprobed for total protein content where appropriate. Primary antibodies for phospho-mixed linage kinase 3 (MLK3) (Thr277/Ser281), JNK, phospho-JNK (Thr183/Tyr185), NF-␬B p65, phospho-NF-␬B p65 (Ser536), phospho-I␬B␣ (Ser32), and total I␬B␣ were obtained from Cell Signaling Technology, and MLK3 total was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Protein contents were determined as described after solubilizing 80 ␮g protein for SDS-PAGE.

Lipid analysis Triacylglycerol content was analyzed as previously described (24). Tissues were dissected free of visible connective tissue and blood. Lipid was extracted by a Folch extraction, the triacylglycerol was saponified in an ethanol/KOH solution at 60 C, and glycerol content was determined fluorometrically. Diacylglycerol and ceramide were extracted and quantified according to the methods of Preiss et al. (25). Lipids were extracted from freeze-dried, powdered soleus muscle using chloroformmethanol-PBS ⫹ 0.2% SDS (1:2:0.8). Diacylglycerol kinase and [32P]ATP (15 mCi/mmol cold ATP) were added to lysates preincubated with cardiolipin/octylglucoside, and the reaction was stopped after 2 h by the addition of chloroform-methanol (2:1). Samples were spotted onto thinlayer chromatography plates and developed. 32P-labeled phosphatidic acid and ceramide-1-phosphate were identified, dried, scraped from the thin layer chromatography plate, and counted in a liquid scintillation analyzer (Tri-Carb 2500TR; Packard, Canberra, Australia).


2079

Results Plasma metabolites and hyperinsulinemic-euglycemic clamp

In the basal fasted state, circulating levels of insulin, glucose, and FFA as well as glucose turnover rate (10.34 ⫾ 0.42 mg/kg䡠min) were identical between the groups. After basal measurements were taken, CNTF or saline was acutely administered via the carotid line. Animals were then infused with a heparinized lipid emulsion or saline, and at the end of this 2-h period, FFA levels were elevated 3-fold above basal (P ⬍ 0.05; Fig. 1) for animals receiving the lipid emulsion, whereas FFA levels were not different from basal for animals infused with saline. After 2 h, these constant infusions were ceased and the subsequent effect of insulin on whole-body glucose disposal rate and suppression of hepatic glucose production (HGP) was measured during hyperinsulinemic-euglycemic clamps (12 mU/kg䡠min). Clamp insulin and glucose concentrations at 12 mU/kg䡠min averaged 840 ⫾ 61 pm and 6.1 ⫾ 0.2 mm, respectively, and by design, no differences for either parameter were detected between the groups (Table 1). Insulin infusion caused a decrease in plasma FFA levels (P ⬍ 0.05) from basal, and despite the 2-h lipid preinfusion, no differences in fatty acid concentration were detected at 60 or 120 min during the clamp procedure between the groups (Fig. 1). Despite the insulin-stimulated suppression of circulating FFA for animals preinfused with lipid, the glucose infusion rate required to maintain euglycemia was decreased by approximately 30% (P ⫽ 0.03) compared with Con saline-infused rats (Fig. 2A). Similar findings were observed during preliminary submaximal clamp studies (1.2 mU/kg䡠min) where only glucose infusion rate was assessed (data not shown). In animals receiving CNTF before lipid, the steady-state glucose infusion rate was identical to Con (saline infused) values. Assessment of tracer kinetics allowed us to distinguish whether these differences could be attributed to alterations in metabolism within liver, skeletal muscle, or both. Insulin caused a 93% suppression of HGP in

Cell culture In later experiments, we determined the effects of CNTF administration on fat oxidation in skeletal muscle cell culture. L6 myoblasts were maintained at 37 C (95% O2/5% CO2) in ␣-modified Eagle’s medium containing 10% fetal bovine serum, 1% penicillin/streptomycin, and 5 mm glucose. Differentiation was induced by switching to medium containing 2% horse serum when the myoblasts were approximately 90% confluent. To examine the effects of CNTF on fatty acid oxidation, cells were serum starved for 4 h and, after washing twice with PBS, fresh media (2 ml) consisting of ␣-modified Eagle’s medium, 2% BSA, and 0.5 mm palmitate, 1 ␮Ci/ml [9,10-3H]palmitate (Amersham Biosciences, Buckinghamshire, UK), was added to the cells. Cells were treated with 10 ng/ml CNTF, 2 mm 5-aminoimidazole-4-carboxamide 1-␤-d-ribofuranoside (AICAR), or 100 nm epinephrine. The reaction was stopped after 2 h. Fatty acid oxidation was determined by separation of 3H2O from the substrate by adding a 1-ml media to chloroform-methanol (2:1), subsequent addition of 2 m KCl-2 m HCl, and removal of the upper phase for liquid scintillation counting.

Statistics Data are expressed as the mean ⫾ sem. Statistical analysis was performed by one-way ANOVA with repeated measures, and specific differences were detected using a Student-Newman-Keuls post hoc test. Statistical significance was set at P ⬍ 0.05.

FIG. 1. Plasma FFA concentrations measured before and after 20% Intralipid and heparin or saline infusion, and during and immediately after hyperinsulinemic euglycemic clamp. Values are expressed as means ⫾ SEM, n ⫽ 6 for the three experimental groups: vehicle ⫹ saline infusion (Con), vehicle ⫹ Intralipid (HF Sal), and 0.3 mg/kg CNTF ⫹ Intralipid (HF CNTF). *, Significant difference vs. Con, P ⬍ 0.05.

2080



TABLE 1. Metabolic parameters during hyperinsulinemiceuglycemic clamp in male Wistar rats with (CNTF) or without (SAL) CNTF pretreatment and/or prior triglyceride and heparin infusion

Body mass (g) Insulin (pM) Glucose (mM) FFA (mM) Liver TG (␮mol/g)

CON

HF SAL

HF CNTF

367 ⫾ 15 835 ⫾ 70 6.0 ⫾ 0.3 0.30 ⫾ 0.12 12.8 ⫾ 1.8

401 ⫾ 14 860 ⫾ 26 6.0 ⫾ 0.4 0.16 ⫾ 0.03 8.4 ⫾ 2.3

382 ⫾ 13 828 ⫾ 88 6.2 ⫾ 0.2 0.22 ⫾ 0.04 10.9 ⫾ 0.8

Values are mean ⫾ SEM, n ⫽ 6. TG, Triacylglycerol; CON, control.

The insulin-stimulated glucose disposal rate (IS-GDR), reflecting predominantly skeletal muscle insulin sensitivity, was decreased by 25% (P ⫽ 0.04) after acute elevation of tissue lipid content (HF—see below). CNTF pretreatment caused a normalization (P ⫽ 0.005 vs. HF) of IS-GDR back to Con levels in HF animals. These data indicate that CNTF counteracts the deleterious effects of elevated circulating FFAs as it restored the ability of insulin to stimulate peripheral glucose disposal as well as suppress HGP despite prior lipid oversupply. Skeletal muscle lipids

untreated rats during clamps, whereas acute elevation in FFA concentration before the clamp caused a significant blunting (P ⫽ 0.01; Fig. 2B) in HGP suppression for high fat (HF). Despite HF, acute CNTF treatment restored the ability of insulin to suppress HGP to rates observed in Con (Fig. 2B).

FIG. 2. A, Glucose infusion rate during a 120-min hyperinsulinemiceuglycemic clamp. Animals were pretreated with a 120-min saline (Con) or 20% Intralipid and heparin (HF) infusion. HF animals were also treated with 0.3 mg/kg (CNTF) or an equal volume of 0.9% NaCl (saline). Insulin infusion rates were 12 mU/kg䡠min. HGP expressed as percent suppression from basal values (B) and IS-GDR (C) during a 120 min hyperinsulinemic-euglycemic clamp (B). Values are means ⫾ SE, n ⫽ 6. *, Different from Con; †, different from HF CNTF, P ⬍ 0.05.

The intracellular deposition of triacylglycerols and fatty acid metabolites have been implicated in lipid-induced skeletal muscle insulin resistance via alterations in insulin signaling (3, 26). Triacylglycerol concentration in skeletal muscle averaged 4.26 ⫾ 0.29 ␮mol/g wet mass and was increased by 155% (P ⫽ 0.04; Fig. 3A) after heparinized Intralipid infusion. CNTF pretreatment completely prevented this increase in intramyocellular triglyceride accumulation. Skeletal muscle ceramide content was elevated (P ⫽ 0.06) after acute heparinized Intralipid infusion, and CNTF pretreatment lowered these levels by 45% (P ⫽ 0.03; Fig. 3C). Unlike

FIG. 3. Intracellular fat metabolites in soleus muscle after hyperinsulinemic-euglycemic clamp preceded by saline or triglyceride and heparin infusion. Triacylglycerol (A), diacylglycerol (B), and ceramide (C) contents. Values are means ⫾ SE, n ⫽ 6. *, Different from Con; †, different from HF CNTF, P ⬍ 0.05.


the potent effects of CNTF on prevention of skeletal muscle triacylglycerol and ceramide accumulation, CNTF had no effect on the accumulation of diacylglycerol induced by lipid infusion (Fig. 3B). To determine whether the reduced skeletal muscle fatty acid accumulation seen in CNTF-treated animals is mediated by increased fat oxidation, we examined CNTF effects in skeletal muscle L6 myotubes. CNTF increased palmitate oxidation by 88%, which was comparable to other known activators of fatty acid oxidation including AICAR (2 mm, 39% increase) and epinephrine (100 nm, 55% increase). Skeletal muscle insulin signaling

The increase in skeletal muscle triacylglycerol and ceramide in HF animals was associated with impaired insulin signaling. As previously shown, IRS serine phosphorylation leads to reduced tyrosine phosphorylation and reduced glucose transport (3, 27). Consistent with this, IRS-1 Ser307 phosphorylation tended to increase (P ⫽ 0.06) in HF animals; however, CNTF treatment prevented this lipid-induced signaling effect (P ⫽ 0.02; Fig. 4A). Although lipid preinfusion tended to cause a modest decrease (P ⫽ 0.1, not significant) in insulin-stimulated IRS-1 tyrosine phosphorylation (Fig. 4B) for HF vs. Con, similar to our findings for IS-GDR, CNTF completely prevented this effect. Insulin-stimulated association between IRS-1 and the p85 subunit of phosphatidyl-

FIG. 4. Insulin signaling in soleus muscle after hyperinsulinemic-euglycemic clamp preceded by saline or triglyceride and heparin infusion. IRS-1 Ser307 phosphorylation (A), IRS-1 tyrosine phosphorylation (B), IRS-1 PI3 kinase association (C), Akt Ser473 phosphorylation (D), representative blots after immunoprecipitation with IRS-1 and immunoblotting with anti-PI3 kinase p85 or anti-IRS-1 tyrosine or serine antibodies (E), and representative immunoblots after probing with anti-Akt Ser473 and anti-Akt antibodies. Values are means ⫾ SE, n ⫽ 6. *, Different from Con; †, different from HF CNTF, P ⬍ 0.05.


2081

inositol 3 (PI3) kinase was also impaired (P ⫽ 0.04) by prior lipid oversupply (HF) vs. Con, but was prevented by CNTF pretreatment (Fig. 4C). Insulin-stimulated phosphorylation of Akt at Ser473 was diminished by prior acute lipid infusion and CNTF pretreatment in HF animals completely restored insulin signal transduction from IRS-1 through Akt (Fig. 4D). MLK3 is a member of the MAPK kinase kinase family of kinases and is activated in mammalian cells by ceramide (28). MLK3 phosphorylation was elevated (P ⫽ 0.03) in muscle after lipid infusion, and this finding is consistent with the increased ceramide levels observed in this group. CNTF treatment prevented the lipid-induced increase in ceramide content, and hence unlike HF alone, no subsequent elevation in MLK3 phosphorylation was observed for CNTF ⫹ HF (Fig. 5A). Additionally, because MLK3 activates JNK in vitro (29), our findings are consistent with these observations in that JNK phosphorylation was elevated in HF rats alone (P ⫽ 0.04, Fig. 5B), but not after CNTF treatment. Hepatic lipids, insulin signaling, and serine kinase activation

Lipid infusion did not result in liver steatosis (Table 1), but activated proinflammatory pathways and inhibited insulin signaling. Insulin-stimulated IRS-1 association with the p85 subunit of PI3 kinase p85 was decreased (P ⫽ 0.02) in HF Sal compared with Con, and although not significant (P ⫽ 0.14),

2082



Discussion

FIG. 5. Serine kinase phosphorylation in skeletal muscle after hyperinsulinemic-euglycemic clamp preceded by saline or triglyceride and heparin infusion. Phospho-MLK3 (Thr277/Ser281)/total MLK3 (A), phospho-JNK (Thr183/Tyr185)/total JNK (B), and phospho-NF␬B p65 (Ser536)/total NF␬B p65 (C). Values are means ⫾ SE, n ⫽ 6. *, Different from Con; †, different from HF CNTF, P ⬍ 0.05.

there was a tendency for this marker of insulin signaling to be attenuated in HF CNTF (Fig. 6A). IRS-2 activation as assessed by IRS-2 association with p85 subunit of PI3 kinase was decreased (P ⫽ 0.05) in HF Sal compared with Con (Fig. 6B). CNTF restored PI3 kinase p85 association with IRS-2 after lipid infusion. Akt Ser473 phosphorylation was decreased in both HF groups compared with Con (P ⬍ 0.05, Fig. 6C). The phosphorylation of Ser536 on the p65 subunit of NF-␬B, a critical event in the regulation of NF-␬B signaling, was elevated in HF Sal compared with Con (P ⫽ 0.01) (Fig. 6D). HF ⫹ CNTF partially blocked the phosphorylation of Ser536. We observed no increase in Ser32 phosphorylation of I␬B␣, or I␬B␣ degradation, both critical events in the activation of the NF-␬B signaling complex. However, these events occur before the phosphorylation of Ser536 on the p65 subunit of NF-␬B, and are tightly regulated so as to avoid the generation of highly proinflammatory conditions. Given that we collected tissue 2 h after the cessation of lipid infusion, it is likely that we missed these early activation markers of the NF-␬B signaling complex. The phosphorylation of both p46 JNK1/2 and p54 JNK1/2 were elevated after HF Sal compared with Con (P ⬍ 0.05; Fig. 6E). HF CNTF completely prevented JNK1/2 phosphorylation.

The beneficial effects of CNTF on insulin action have been previously attributed to occur secondary to effects on lipogenic gene expression and decreased lipid accumulation in nonadipose tissues (11). Our findings show that acute CNTF treatment prevents decrements in whole-body insulin sensitivity induced by acute lipid infusion. This was accompanied by increased insulin-stimulated tyrosine phosphorylation of IRS-1, p85 association with IRS-1, and activation of Akt in skeletal muscle samples obtained from animals undergoing a hyperinsulinemic euglycemic clamp. CNTF treatment also prevented the lipid-induced increase in skeletal muscle triacylglycerol and ceramide content, and subsequent lipid-induced MLK3 and JNK activation. We show that CNTF protects insulin action in the face of a lipid challenge via mechanisms other than altered lipogenic gene expression. These mechanisms are likely associated with the prevention of cellular accumulation of ceramides and phosphorylation of JNK in peripheral tissues, thereby preventing impaired insulin signal transduction. A strong correlation between skeletal muscle lipid accumulation and the severity of insulin resistance suggests that the accumulation of lipid-derived metabolites may antagonize directly insulin signal transduction (4, 5). In the present study, lipid infusion increased skeletal muscle triacylglycerol, diacylglycerol, and ceramide content in conjunction with reduced insulin sensitivity. CNTF treatment completely prevented the increase in IMTG and ceramide, but not diacylglycerol, suggesting CNTF may mediate its insulin-sensitizing effects by reducing select skeletal muscle fatty acid metabolite levels, perhaps via increased fatty acid oxidation. Ceramide is a fatty acyl CoA metabolite that is elevated in the skeletal muscle from insulin-resistant humans (30) and rodents (31, 32), and is correlated negatively to insulin sensitivity as assessed by the hyperinsulinemic-euglycemic clamp technique (33). Studies using short-chain ceramide analogs or saturated fatty acids to increase ceramide levels in skeletal muscle cells show that ceramide inhibits Akt phosphorylation/activation (26, 34) by dual mechanisms: first by preventing translocation of Akt and second via dephosphorylation of Akt by protein phosphatase 2A (35, 36). The reduction in ceramide content and increased insulin-stimulated Akt Ser473 phosphorylation after CNTF administration in lipid-treated rats is consistent with these in vitro observations and may explain, at least in part, the insulin sensitizing effects of CNTF. The decrease in ceramide content may also explain the improved insulin-stimulated IRS-1 tyrosine phosphorylation seen in CNTF-treated rats after the lipid challenge (37), although this remains controversial (26). Fatty acids promote serine kinase activation resulting in increased serine phosphorylation of IRS-1. Phosphorylation of IRS-1 on certain Ser residues impairs insulin-stimulated tyrosine phosphorylation and downstream PI3 kinase activation, and thus insulinstimulated glucose disposal is reduced (38). Although many kinases phosphorylate IRS-1, evidence suggests JNK, a member of the MAPK family, is an important negative mediator of fatty acid-induced insulin resistance. In vitro, JNK activation causes phosphorylation at IRS-1 on residue Ser307 and



2083

FIG. 6. Insulin and serine kinase signaling in the liver after hyperinsulinemic-euglycemic clamp preceded by saline or triglyceride and heparin infusion. IRS-1 PI3 kinase association (A), IRS-2 PI3 kinase association (B), Akt Ser473 phosphorylation (C), NF␬␤ p65 phosphorylation (D), and JNK 1/2 phosphorylation (E). Values are means ⫾ SE, n ⫽ 6. *, Different from Con; †, different from HF CNTF, P ⬍ 0.05.

reduces tyrosine phosphorylation, effects which are reversed by inhibition of JNK in cultured cells (27, 39), whereas similarly in vivo, ob/ob and high fat-fed mice have elevated JNK activity in peripheral tissues (14). Additional support for JNK1 in the pathogenesis of lipid-induced insulin resistance is provided by the JNK1 null mice which are protected from the deleterious effects of high fat feeding on whole-body insulin sensitivity (40). Although the molecular pathway linking JNK to reduced insulin action is well supported, the in vivo effects of ceramide in this pathway remain unknown. Previous studies using dsRNAi-mediated gene silencing and cell-permeable agonists show that ceramide activates MLK3 and thus JNK (28). In the present in vivo experiments, elevated ceramide content (1.5-fold) was associated with increased MLK3, JNK, and IRS-1 phosphorylation on residue Ser307 in skeletal muscle, and decreased whole-body insulinstimulated glucose disposal. Moreover, these signaling and biological effects were prevented with CNTF pretreatment, which also coincided with prevention of ceramide accumulation. It is possible that other serine kinases phosphorylate IRS-1 at Ser307 during lipid infusion. IKK␤ can directly bind to IRS-1 Ser307 (6); however, we measured no change in skeletal muscle NF␬B signaling, which is consistent with human studies using acute lipid infusion (1). Fatty acids may also

cause insulin resistance through diacylglycerol activation of a serine kinase cascade involving protein kinase C␪ (2, 3). Although diacylglycerol and IRS Ser307 were increased after lipid infusion, CNTF had no effect to diminish diacylglycerol after Intralipid infusion despite improved skeletal muscle insulin signaling and whole-body glucose disposal. The disconnect between skeletal muscle diacylglycerol concentration and peripheral insulin action, particularly CNTF-induced insulin sensitization in this case, indicate that diacylglycerol and protein kinase C activation may not be essential to the etiology of lipid-induced insulin resistance (1, 3). Lipid-induced hepatic insulin resistance in the present study was associated with reduced IRS-1 and IRS-2 association with the p85 subunit of PI3 kinase, which are important components of hepatic insulin signal transduction. The finding that CNTF administration partially reversed these deleterious events in lipid-infused rats is consistent with findings from Sleeman et al. (11), who demonstrated that 10 d of Axokine treatment improved insulin-stimulated signaling in db/db mice exhibiting severe hepatic steatosis. However, we show that CNTF can improve insulin action acutely in liver, and insulin sensitization in our model does not occur due to reductions in liver steatosis and/or alterations in the expression of key hepatic genes. There is now an abundance of

2084


evidence linking inflammation, serine/threonine kinase activation, and insulin resistance (41), and accordingly, we demonstrated increased NF␬B and JNK activation and hepatic insulin resistance after lipid infusion. A significant finding of the present study was the ability of CNTF to completely suppress lipid-induced JNK activation and partially prevent NF␬B signaling. Thus, our data ascribes an acute direct role for CNTF on hepatic insulin action. In summary, these data provide new insights into the mechanisms of CNTF-induced insulin sensitization in skeletal muscle and liver. In this study, we provide compelling evidence that CNTF prevents the accumulation of triacylglycerol and ceramides in skeletal muscle and this yields protection against acute lipid-induced insulin resistance. The maintenance of normal insulin signaling despite lipid oversupply is associated with 1) elimination of ceramide-induced Akt inhibition and or 2) enhanced IRS-1 tyrosine phosphorylation secondary to attenuated MLK3/JNK activation and IRS-1 Ser307 phosphorylation. Acute CNTF administration also enhances hepatic insulin action acutely by inhibition of inflammatory serine kinases. These data confirm the efficacy of CNTF receptor ligands as potential therapeutic targets for lipid-induced insulin resistance.


9.

10.

11.

12.

13.

14.

15.

16.

17.

Acknowledgments

18.

We thank Anna Holmes for excellent technical assistance. 19.

Received August 23, 2005. Accepted December 23, 2005. Address all correspondence and requests for reprints to: Matthew J. Watt, Ph.D., St. Vincent’s Institute of Medical Research, 9 Princes Street, Fitzroy, Victoria 3065, Australia. E-mail: [email protected]. This research was supported by a research grant (251558), a Peter Doherty postdoctoral fellowship (to M.J.W), and a senior research fellowship (to M.A.F) from the National Health and Medical Research Council of Australia. A.H. is supported by the National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases (DK60484). M.J.W., A.H., G.I.L., and M.A.F. have no disclosures to declare.

20.

21. 22.

23.

References 1. Itani SI, Ruderman NB, Schmieder F, Boden G 2002 Lipid-induced insulin resistance in human muscle is associated with changes in diacylglycerol, protein kinase C, and I␬B-␣. Diabetes 51:2005–2011 2. Griffin ME, Marcucci MJ, Cline GW, Bell K, Barucci N, Lee D, Goodyear LJ, Kraegen EW, White MF, Shulman GI 1999 Free fatty acid-induced insulin resistance is associated with activation of protein kinase C␪ and alterations in the insulin signaling cascade. Diabetes 48:1270 –1274 3. Yu C, Chen Y, Cline GW, Zhang D, Zong H, Wang Y, Bergeron R, Kim JK, Cushman SW, Cooney GJ, Atcheson B, White MF, Kraegen EW, Shulman GI 2002 Mechanism by which fatty acids inhibit insulin activation of insulin receptor substrate-1 (IRS-1)-associated phosphatidylinositol 3-kinase activity in muscle. J Biol Chem 277:50230 –50236 4. Krssak M, Falk Petersen K, Dresner A, DiPietro L, Vogel SM, Rothman DL, Roden M, Shulman GI 1999 Intramyocellular lipid concentrations are correlated with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabetologia 42:113–116 5. Pan DA, Lillioja S, Kriketos AD, Milner MR, Baur LA, Bogardus C, Jenkins AB, Storlien LH 1997 Skeletal muscle triglyceride levels are inversely related to insulin action. Diabetes 46:983–988 6. Gao Z, Zuberi A, Quon MJ, Dong Z, Ye J 2003 Aspirin inhibits serine phosphorylation of insulin receptor substrate 1 in tumor necrosis factor-treated cells through targeting multiple serine kinases. J Biol Chem 278:24944 –24950 7. Varon S, Manthorpe M, Adler R 1979 Cholinergic neuronotrophic factors: I. Survival, neurite outgrowth and choline acetyltransferase activity in monolayer cultures from chick embryo ciliary ganglia. Brain Res 173:29 – 45 8. Lambert PD, Anderson KD, Sleeman MW, Wong V, Tan J, Hijarunguru A, Corcoran TL, Murray JD, Thabet KE, Yancopoulos GD, Wiegand SJ 2001 Ciliary neurotrophic factor activates leptin-like pathways and reduces body

24.

25.

26.

27.

28.

29.

30.

31.

32. 33.

fat, without cachexia or rebound weight gain, even in leptin-resistant obesity. Proc Natl Acad Sci USA 98:4652– 4657 Gloaguen I, Costa P, Demartis A, Lazzaro D, Di Marco A, Graziani R, Paonessa G, Chen F, Rosenblum CI, Van der Ploeg LH, Cortese R, Ciliberto G, Laufer R 1997 Ciliary neurotrophic factor corrects obesity and diabetes associated with leptin deficiency and resistance. Proc Natl Acad Sci USA 94:6456 – 6461 Ettinger MP, Littlejohn TW, Schwartz SL, Weiss SR, McIlwain HH, Heymsfield SB, Bray GA, Roberts WG, Heyman ER, Stambler N, Heshka S, Vicary C, Guler HP 2003 Recombinant variant of ciliary neurotrophic factor for weight loss in obese adults: a randomized, dose-ranging study. JAMA 289:1826 –1832 Sleeman MW, Garcia K, Liu R, Murray JD, Malinova L, Moncrieffe M, Yancopoulos GD, Wiegand SJ 2003 Ciliary neurotrophic factor improves diabetic parameters and hepatic steatosis and increases basal metabolic rate in db/db mice. Proc Natl Acad Sci USA 100:14297–14302 Bluher S, Moschos S, Bullen Jr J, Kokkotou E, Maratos-Flier E, Wiegand SJ, Sleeman MW, Mantzoros CS 2004 Ciliary neurotrophic factor Ax15 alters energy homeostasis, decreases body weight, and improves metabolic control in diet-induced obese and UCP1-DTA mice. Diabetes 53:2787–2796 Zvonic S, Cornelius P, Stewart WC, Mynatt RL, Stephens JM 2003 The regulation and activation of ciliary neurotrophic factor signaling proteins in adipocytes. J Biol Chem 278:2228 –2235 Baron AD, Brechtel G, Wallace P, Edelman SV 1988 Rates and tissue sites of non-insulin- and insulin-mediated glucose uptake in humans. Am J Physiol 255:E769 –E774 van Hall G, Sacchetti M, Radegran G, Saltin B 2002 Human skeletal muscle fatty acid and glycerol metabolism during rest, exercise and recovery. J Physiol 543:1047–1058 Davis S, Aldrich TH, Valenzuela DM, Wong VV, Furth ME, Squinto SP, Yancopoulos GD 1991 The receptor for ciliary neurotrophic factor. Science 253:59 – 63 Arner P 2005 Insulin resistance in type 2 diabetes-role of the adipokines. Curr Mol Med 5:333–339 Minokoshi Y, Kim YB, Peroni OD, Fryer LG, Muller C, Carling D, Kahn BB 2002 Leptin stimulates fatty-acid oxidation by activating AMP-activated protein kinase. Nature 415:339 –343 Steinberg GR, Bonen A, Dyck DJ 2002 Fatty acid oxidation and triacylglycerol hydrolysis are enhanced after chronic leptin treatment in rats. Am J Physiol Endocrinol Metab 282:E593–E600 Boden G, Lebed B, Schatz M, Homko C, Lemieux S 2001 Effects of acute changes of plasma free fatty acids on intramyocellular fat content and insulin resistance in healthy subjects. Diabetes 50:1612–1617 Boden G, Chen X 1995 Effects of fat on glucose uptake and utilization in patients with non-insulin-dependent diabetes. J Clin Invest 96:1261–1268 Bachmann OP, Dahl DB, Brechtel K, Machann J, Haap M, Maier T, Loviscach M, Stumvoll M, Claussen CD, Schick F, Haring HU, Jacob S 2001 Effects of intravenous and dietary lipid challenge on intramyocellular lipid content and the relation with insulin sensitivity in humans. Diabetes 50:2579 –2584 Schuster B, Kovaleva M, Sun Y, Regenhard P, Matthews V, Grotzinger J, Rose-John S, Kallen KJ 2003 Signaling of human ciliary neurotrophic factor (CNTF) revisited. The interleukin-6 receptor can serve as an ␣-receptor for CTNF. J Biol Chem 278:9528 –9535 Frayn KN, Maycock PF 1980 Skeletal muscle triacylglycerol in the rat: methods for sampling and measurement, and studies of biological variability. J Lipid Res 21:139 –144 Preiss J, Loomis CR, Bishop WR, Stein R, Niedel JE, Bell RM 1986 Quantitative measurement of sn-1,2-diacylglycerols present in platelets, hepatocytes, and ras- and sis-transformed normal rat kidney cells. J Biol Chem 261:8597– 8600 Schmitz-Peiffer C, Craig DL, Biden TJ 1999 Ceramide generation is sufficient to account for the inhibition of the insulin-stimulated PKB pathway in C2C12 skeletal muscle cells pretreated with palmitate. J Biol Chem 274:24202–24210 Lee YH, Giraud J, Davis RJ, White MF 2003 c-Jun N-terminal kinase (JNK) mediates feedback inhibition of the insulin signaling cascade. J Biol Chem 278:2896 –2902 Sathyanarayana P, Barthwal MK, Kundu CN, Lane ME, Bergmann A, Tzivion G, Rana A 2002 Activation of the Drosophila MLK by ceramide reveals TNF-␣ and ceramide as agonists of mammalian MLK3. Mol Cell 10:1527–1533 Rana A, Gallo K, Godowski P, Hirai S, Ohno S, Zon L, Kyriakis JM, Avruch J 1996 The mixed lineage kinase SPRK phosphorylates and activates the stressactivated protein kinase activator, SEK-1. J Biol Chem 271:19025–19028 Adams 2nd JM, Pratipanawatr T, Berria R, Wang E, DeFronzo RA, Sullards MC, Mandarino LJ 2004 Ceramide content is increased in skeletal muscle from obese insulin-resistant humans. Diabetes 53:25–31 Gorska M, Dobrzyn A, Zendzian-Piotrowska M, Gorski J 2004 Effect of streptozotocin-diabetes on the functioning of the sphingomyelin-signalling pathway in skeletal muscles of the rat. Horm Metab Res 36:14 –21 Turinsky J, O’Sullivan DM, Bayly BP 1990 1,2-Diacylglycerol and ceramide levels in insulin-resistant tissues of the rat in vivo. J Biol Chem 265:16880 –16885 Straczkowski M, Kowalska I, Nikolajuk A, Dzienis-Straczkowska S, Kinal-


34. 35. 36.

37.


ska I, Baranowski M, Zendzian-Piotrowska M, Brzezinska Z, Gorski J 2004 Relationship between insulin sensitivity and sphingomyelin signaling pathway in human skeletal muscle. Diabetes 53:1215–1221 Chavez JA, Summers SA 2003 Characterizing the effects of saturated fatty acids on insulin signaling and ceramide and diacylglycerol accumulation in 3T3–L1 adipocytes and C2C12 myotubes. Arch Biochem Biophys 419:101–109 Stratford S, Hoehn KL, Liu F, Summers SA 2004 Regulation of insulin action by ceramide: dual mechanisms linking ceramide accumulation to the inhibition of Akt/protein kinase B. J Biol Chem 279:36608 –36615 Cazzolli R, Carpenter L, Biden TJ, Schmitz-Peiffer C 2001 A role for protein phosphatase 2A-like activity, but not atypical protein kinase C␨, in the inhibition of protein kinase B/Akt and glycogen synthesis by palmitate. Diabetes 50:2210 –2218 Kanety H, Hemi R, Papa MZ, Karasik A 1996 Sphingomyelinase and ceramide

38. 39.

40. 41.

2085

suppress insulin-induced tyrosine phosphorylation of the insulin receptor substrate-1. J Biol Chem 271:9895–9897 Shulman GI 2000 Cellular mechanisms of insulin resistance. J Clin Invest 106:171–176 Jiang G, Dallas-Yang Q, Liu F, Moller DE, Zhang BB 2003 Salicylic acid reverses phorbol 12-myristate-13-acetate (PMA)- and tumor necrosis factor ␣ (TNF␣)-induced insulin receptor substrate 1 (IRS1) serine 307 phosphorylation and insulin resistance in human embryonic kidney 293 (HEK293) cells. J Biol Chem 278:180 –186 Hirosumi J, Tuncman G, Chang L, Gorgun CZ, Uysal KT, Maeda K, Karin M, Hotamisligil GS 2002 A central role for JNK in obesity and insulin resistance. Nature 420:333–336 Wellen KE, Hotamisligil GS 2005 Inflammation, stress, and diabetes. J Clin Invest 115:1111–1119

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.

Erratum The article “Evaluation of an Estrogen Receptor-␤ Agonist in Animal Models of Human Disease” by Heather A. Harris, Leo M. Albert, Yelena Leathurby, Michael S. Malamas, Richard E. Mewshaw, Chris P. Miller, Yogendra P. Kharode, James Marzolf, Barry S. Komm, Richard C. Winneker, Donald E. Frail, Ruth A. Henderson, Yuan Zhu, and James C. Keith, Jr. (Endocrinology 144:4241– 4249, 2003) described a rat mammotrophic model in which ERB-041 was reported to regulate gene expression. The authors have determined that the gene marker measured in this model was, in fact, not casein kinase II and at present have not identified the gene product measured. The authors regret the error.