Acute exposure to rosiglitazone does not affect glucose transport in intact human skeletal muscle

  • Published on

  • View

  • Download

Embed Size (px)


  • Acute exposure to rosiglitazone does nletb, Asinkich, B

    09; ac

    as anexponond9

    rosiglitazone (1 or 10 mol/L), 5-aminoimidazole-4-carboxamide 1--D-ribonucleoside (1 mmol/L), insulin (120 nmol/L), or a combinationof insulin (120 nmol/L) and rosiglitazone (10 mol/L) in vitro for 1 hour. Glucose transport was analyzed by accumulation of intracellular 3-

    develops [2,3]. Therefore, interventions to improve muscle sensitivity in insulin-resistant subjects such as people with

    to TZDs directly activates adenosine monophosphateactivated protein kinase (AMPK) and increases glucose

    Available online at

    Metabolism Clinical and Experimenta

    Corresponding author. Division of Cardiology, Department of

    Medicine, Helsinki University Central Hospital; Minerva FoundationInstitute for Medical Research, Biomedicum 2U Helsinki, Tukholmankatuglucose metabolism would be beneficial in the preventionand treatment of type 2 diabetes mellitus. Thiazolidinediones(TZDs), such as rosiglitazone or pioglitazone, are widelyused as antidiabetic drugs [4]. Thiazolidinediones improvewhole-body insulin sensitivity by targeting peroxisomeproliferatoractivated receptor (PPAR), a transcriptionfactor highly expressed in adipocytes [5]. Activation of

    obesity, type 2 diabetes mellitus, polycystic ovary disease, orimpaired glucose tolerance (IGT) [7-9].

    The PPAR expression in adipocytes is viewed as theprimary target of TZD action [5]. However, it is possible thatTZDs have direct metabolic effects on other tissues such asskeletal muscle and liver, which may be independent ofPPAR. Transgenic mice with no adipose tissue or with amuscle-specific PPAR deletion improve their insulinsensitivity in response to TZDs [10,11], and acute exposureO-methyl [3H] glucose; phosphorylation of Akt-Ser473 and Akt-Thr308 and phosphorylation of acetyl coenzyme A carboxylase weredetermined using phosphospecific antibodies. 5-Aminoimidazole-4-carboxamide 1--D-ribonucleoside and insulin increased glucosetransport rate 1.5-fold (P b .05) and 1.7-fold (P b .01) in isolated muscle strips, respectively. Exposure to rosiglitazone transiently increasedphosphorylation of acetyl coenzyme A carboxylase , with a maximum effect at 15 minutes and return to baseline at 60 minutes. However,rosiglitazone did not affect basal or insulin-stimulated glucose transport rate, or phosphorylation of Akt-Ser473 or Akt-Thr308 in isolatedmuscle strips. In conclusion, acute exposure to rosiglitazone does not affect glucose transport in human skeletal muscle. 2010 Elsevier Inc. All rights reserved.

    1. Introduction

    Skeletal muscle is the main site of resistance to insulin-stimulated glucose metabolism in type 2 diabetes mellitus[1], and defects at the level of muscle glucose metabolismcan be observed years before overt type 2 diabetes mellitus

    PPAR leads to transcriptional changes in a number of genesregulating adipogenesis and fatty acid uptake and storage.This leads to repartition of lipids from circulation, muscle,and liver into adipose tissue, resulting in improved insulinsensitivity [5,6]. Indeed, a number of clinical studies provideevidence that treatment with TZDs improves insulinhuman skePaulina Skrobuka,b, Heidi Kuoppamaaa,

    aDivision of Cardiology, Department of Medicine, HelbMinerva Foundation Institute for Medical Resear

    Received 27 May 20


    Thiazolidinediones (TZDs) such as rosiglitazone are widely usedmetabolic actions in skeletal muscle. Here, we examined if acuteproximal insulin signaling in isolated skeletal muscle strips fromvastus lateralis from 15 nondiabetic men (50 3 years old, 26.8, 00290 Helsinki, Finland. Tel.: +358 9 191 25 710.E-mail address: (H.A. Koistinen).

    0026-0495/$ see front matter 2010 Elsevier Inc. All rights reserved.doi:10.1016/j.metabol.2009.07.016ot affect glucose transport in intactal musclenne Hiukkaa, Heikki A. Koistinena,b,

    University Central Hospital, 00290 Helsinki, Finlandiomedicum 2U Helsinki, 00290 Helsinki, Finland

    cepted 16 July 2009

    tidiabetic drugs. Animal studies suggest that TZDs may have directsure to rosiglitazone stimulates glucose transport rate and affectsiabetic men. Open muscle biopsies were obtained from musculus1.1 kg/m2). Skeletal muscle strips were isolated and exposed to

    l 59 (2010) 224230www.metabolismjournal.comuptake and palmitate oxidation in intact skeletal muscle fromrodents [12]. Clinically, full therapeutic effect is reached

  • after treatment with TZDs for several weeks to months,demonstrating a delayed onset of action [13-15]. However,LeBrasseur et al [12] demonstrated that an exposure for asshort as 30 minutes to troglitazone increased glucose uptakein isolated rat skeletal muscle. It is not known if TZDs exertacute direct metabolic effects in intact human skeletalmuscle. Therefore, we examined whether rosiglitazoneaffects glucose transport and proximal insulin-signalingevents in isolated skeletal muscle strips from nondiabeticmen. Because we have previously shown that an acuteactivation of AMPK by the nucleoside analog 5-aminoimi-dazole-4-carboxamide 1--D-ribonucleoside (AICAR)increases glucose transport rate in isolated human skeletalmuscle strips [16], we also included AICAR exposure in ourexperiments as a positive control.

    2. Methods

    The study protocol was reviewed and approved by theEthical Committee of the Department of Medicine, HelsinkiUniversity Central Hospital; and written informed consentwas obtained from all subjects before participation. Physio-logic characteristics of the study subjects (15 nondiabetic

    Hip (cm) 100 2

    225P. Skrobuk et al. / Metabolism Clinical and Experimental 59 (2010) 224230Waist-to-hip ratio 0.97 0.02Body fat % 22.8 1.5VO2max (mL kg

    1 min1) 35.8 2.9Systolic blood pressure (mm Hg) 146 5Diastolic blood pressure (mm Hg) 89 3Blood glucose (mmol/L) 5.1 0.1HbA1c (%) 5.2 0.1Serum cholesterol (mmol/L) 4.7 0.2Serum triacylglycerol (mmol/L) 1.6 0.2Serum HDL cholesterol (mmol/L) 1.30 0.08Serum LDL cholesterol (mmol/L) 2.7 0.2Fasting blood glucose at OGTT (mmol/L) 5.3 0.12-h blood glucose at OGTT (mmol/L) 5.3 0.4

    Blood samples were obtained for clinical chemistry analysis from fastedmen. Results are means SE. N = 15 except for body fat percentage (n =11) and VO2max (n = 10). HbA1c indicates hemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; OGTT, oral glucosetolerance are presented in Table 1. Six of the men had no familyhistory of diabetes, whereas 9 men had first- and/or second-degree relatives with type 2 diabetes mellitus. Oral glucosetolerance test (75 g) was performed in all subjects. Nine menhad normal glucose tolerance (NGT), 3 men had impairedfasting glucose (IFG), and 3 men had IGT. None of the menwas taking any regular medications. One man was a smoker.All subjects were instructed to avoid strenuous exercise for

    Table 1Subject characteristics and clinical chemistry

    N 15Age (y) 50 3BMI (kg/m2) 26.9 1.1Waist (cm) 97 372 hours before the muscle biopsy. The subjects wereexamined after an overnight fast.

    2.1. Open muscle biopsy

    Open biopsies were taken from the vastus lateralismuscle under local anesthesia (lidocaine hydrochloride, 5mg/mL), as previously described [16,17]. A 4-cm incisionwas made 15 cm above the proximal border of patella, andthe muscle fascia was exposed. Thereafter, 4 to 5 musclefiber bundles were excised and placed in oxygenated Krebs-Henseleit buffer (KHB), which contained 5 mmol/LHEPES, 5 mmol/L glucose, 15 mmol/L mannitol, and0.1% bovine serum albumin (radioimmunoassay grade;Sigma, St Louis, MO). Smaller skeletal muscle strips weredissected from the muscle biopsy specimen, mounted onPlexiglass clamps, and incubated in vitro in pregassed (95%O2 and 5% CO2) KHB in shaking water bath at 35C. Thegas phase in the vials was maintained during the incubationprocedure. After 30 minutes of incubation in KHB, skeletalmuscle strips were incubated for 30 minutes at 35C without(basal) or with 1 or 10 mol/L rosiglitazone (AlexisBiochemicals, Lausen, Switzerland), 120 nmol/L insulin(Actrapid; Novo Nordisk A/S, Bagsvrd, Denmark), acombination of 10 mol/L rosiglitazone and insulin, or 1mmol/L AICAR (Sigma). These concentrations weremaintained throughout the subsequent incubation proce-dures. As rosiglitazone was dissolved in methanol, equalamounts of methanol were added to the media to maintainconcentration at 0.1% in all conditions.

    2.2. Glucose transport

    Skeletal muscle strips were transferred to fresh KHBcontaining 20 mmol/L mannitol and incubated for 10minutes. Subsequently, muscles were incubated for 20minutes in KHB containing 5 mmol/L 3-O-methyl [3H]glucose (800 Ci/mmol) and 15 mmol/L [14C] mannitol(53 Ci/mmol) (GE Healthcare, Life Sciences, Whitchurch,Cardiff, United Kingdom). Muscle strips were blotted ofexcess fluid, frozen in liquid nitrogen, and stored at 80Cuntil further analysis. Glucose transport was determined bythe accumulation of intracellular 3-O-methyl [3H] glucose,as described [17].

    2.3. Tissue processing

    Muscle strips were homogenized in ice-cold homogeni-zation buffer (90 L/g dry weight muscle) (20 mmol/L Tris[pH 7.8], 137 mmol/L NaCl, 2.7 mmol/L KCl, 1 mmol/LMgCl2,, 1% Triton X-100, 10% [wt/vol] glycerol, 10 mmol/L NaF, 0.5 mmol/L Na3VO4, 1 g/mL leupeptin, 0.2 mmol/L phenylmethyl sulfonyl fluoride, 1 g/mL aprotinin).Homogenates were rotated for 60 minutes at 4C. Subse-quently, homogenates were subjected to centrifugation(12 000g for 10 minutes at 4C); and protein concentrationwas determined using a bicinchoninic acid protein assay(Pierce, Rockford, IL). An aliquot of homogenate was mixed

  • aired Student t test was used in the analysis of paired data.ll statistical analyses were performed with SPSS statisticalackage (version 16.0; SPSS, Chicago, IL). Two-tailed Palue b .05 was considered statistically significant.

    . Results

    .1. Glucose transport

    Isolated muscle strips were incubated for 60 minutes ine absence (basal) or in the presence of rosiglitazone (1 or0 mol/L), AICAR (1 mmol/L), insulin (120 nmol/L), or aombination of rosiglitazone (10 mol/L) and insulin (120mol/L) (Fig. 1). There was no effect of 1 mol/Losiglitazone on glucose transport rate in intact skeletaluscle. Exposure to 10 mol/L rosiglitazone tended tocrease glucose transport rate, but this was not statisticallyignificant (P = .182) (Fig. 1). 5-Aminoimidazole-4-arboxamide 1--D-ribonucleoside and insulin increasedlucose transport rate 1.5-fold (P b .05) and 1.7-fold (P b1), respectively. A combined exposure to both rosiglita-one and insulin did not have any effect on glucose transport

    of 10 mol/L rosiglitazone and insulin (120 nmol/L) for 60 minutes.

    226 P. Skrobuk et al. / Metabolism Clinical and Experimental 59 (2010) 224230with Laemmli buffer containing -mercaptoethanol andheated (95C) for 5 minutes.

    2.4. Western blot analysis

    Total and phosphorylated proteins were determined withcommercially available antibodies: antiphospho-Akt(Ser473) (catalogue #9271), antiphospho-Akt (Thr308)(catalogue #9275), anti-Akt (catalogue #9272), anti-desmin(catalogue #04-585), and antiphospho-ACC (Ser79)(catalogue #07-303) (Upstate Biotechnology, Lake Placid,NY). Proteins were separated by sodium dodecyl sulfatepolyacrylamide gel electrophoresis, transferred to polyviny-lidene difluoride membranes, and blocked with Tris-bufferedsaline with 0.02% Tween containing 2.5% bovine serumalbumin for 3 hours (pAkt-Ser473 and pAkt-Thr308) or 7.5%milk for 2 hours (desmin, total Akt, and pACC-Ser79).Membranes were incubated overnight at 4C with primaryantibodies (1:500 for Akt and desmin, 1:1000 for pACC).Membranes were washed in Tris-buffered saline with 0.02%Tween and incubated with appropriate secondary horse-radish peroxidase-conjugated antibodies (Bio-Rad, Rich-mond, CA). Immunoreactive proteins were visualized byenhanced chemiluminescence (ECL plus; Amersham,Arlington Heights, IL) and quantified by densitometryusing Molecular Analyst Software (Bio-Rad).

    2.5. Clinical chemistry

    Blood glucose concentration was analyzed by glucosedehydrogenase method (Precision-G Blood Testing System;Medisense, Abbott Park, IL). Hemoglobin A1c was deter-mined by an immunological method (DCA 2000+; BayerHealthcare, Elkhart, IN). Serum total and high-densitylipoprotein cholesterol, and triglyceride concentrationswere analyzed enzymatically by an automated analyzer(Konelab 60i; Thermo Electron Oy, Clinical Chemistry andAutomation Systems, Vantaa, Finland). Low-density lipo-protein cholesterol concentration was calculated by using theformula of Friedewald et al [18].

    2.6. Body composition and maximal oxygen uptake

    Waist circumference was measured midway between thelower rib margin and the iliac crest, and hip circumferencewas measured at the level of the trochanters with the use of asoft measuring tape. Body fat percentage was measured bydual-energy x-ray absorptiometry (Hologic Discovery A;Fenno Medical, Vantaa, Finland) (n = 11), and maximaloxygen uptake (VO2max) was determined by using a bicycleergometer (900 ERG Ergometer; Marquette Hellige, Mar-quette Medical Systems, Milwaukee, WI) and a breath-by-breath gas exchange analysis system (Vmax 229, Sensor-medics, Homestead, FL) (n = 10).

    2.7. Statistics

    Data are presented as mean SE. Normal distribution ofthe variables was verified by Kolmogorov-Smirnov test.Glucose transport was analyzed by the accumulation of intracellular 3-O-methyl [3H] glucose. Exposure to insulin and AICAR increased glucosetransport, whereas exposure to rosiglitazone did not modify basal or insulin-stimulated glucose transport rate. *P b .05 vs basal. Data are presented asmeans SE. n = 9 to 15 per condition. Rosi 1 indicates 1 mol/Lrosiglitazone; Rosi 10, 10 mol/L rosiglitazone.rate compared with insulin alone (P = .258) (Fig. 1).To examine if clinical characteristics of the study cohort

    affect glucose transport responses to rosiglitazone, we nextanalyzed glucose transport data according to glucosetolerance of the subjects. Men with NGT (47 4 years old;body mass index [BMI], 25.1 1.3 kg/m2; n = 9) and menwith IFG or IGT (54 4 years old; BMI, 29.5 1.5 kg/m2; n =6) had similar age (P = .28), whereas men with IFG/IGT had ahigher BMI (P b .05). Exposure to 1 or 10 mol/Lrosiglitazone did not affect glucose transport rate in intactskeletal muscle in either men with NGT or men with IFG/IGT(Table 2). We next divided the subjects into 2 groups based

    Fig. 1. Glucose transport. Open muscle biopsies were obtained surgicallyfrom vastus lateralis muscle from 15 nondiabetic men, as described inMethods. Muscle strips were isolated and exposed to 1 or 10 mol/Lrosiglitazone, AICAR (1 mmol/L), insulin (120 nmol/L), or a combinationPApv




  • on the median BMI, body fat percentage, or VO2max of thewhole cohort: (1) subjects with BMI, body fat percentage, or

    ylation reached a peak at 15 minutes (P = .051) and was

    proves insulin sensitivity are still incompletely resolved.

    Table 2Glucose transport rate (in nanomoles per milligram p...