This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

Author's personal copy

Journal of Steroid Biochemistry & Molecular Biology 141 (2014) 71–76

Contents lists available at ScienceDirect

Journal of Steroid Biochemistry and Molecular Biology

j ourna l h omepage: www.elsev ier .com/ locate / j sbmb

Dihydrotestosterone deteriorates cardiac insulin signaling andglucose transport in the rat model of polycystic ovary syndrome

Snezana Tepavcevic a, Danijela Vojnovic Milutinovic b, Djuro Macutc, Zorica Zakulaa,Marina Nikolic b, Ivana Bozic-Antic c, Snjezana Romic a, Jelica Bjekic-Macutd,Gordana Matic b, Goran Koricanaca,∗

a Laboratory for Molecular Biology and Endocrinology, Vinca Institute of Nuclear Sciences, University of Belgrade, Belgrade, Serbiab Institute for Biological Research “Sinisa Stankovic”, University of Belgrade, Belgrade, Serbiac Clinic for Endocrinology, Diabetes and Metabolic Diseases, Clinical Center of Serbia and Faculty of Medicine, University of Belgrade, Belgrade, Serbiad CHC Bezanijska kosa, Belgrade, Serbia

a r t i c l e i n f o

Article history:Received 11 October 2013Received in revised form24 December 2013Accepted 16 January 2014Available online 26 January 2014

Keywords:Polycystic ovary syndromeDihydrotestosteroneHeartInsulin signaling pathwayGlucose transporters

a b s t r a c t

It is supposed that women with polycystic ovary syndrome (PCOS) are prone to develop cardiovasculardisease as a consequence of multiple risk factors that are mostly related to the state of insulin resistanceand consequent hyperinsulinemia. In the present study, we evaluated insulin signaling and glucose trans-porters (GLUT) in cardiac cells of dihydrotestosterone (DHT) treated female rats as an animal model ofPCOS. Expression of proteins involved in cardiac insulin signaling pathways and glucose transporters,as well as their phosphorylation or intracellular localization were studied by Western blot analysis inDHT-treated and control rats. Treatment with DHT resulted in increased body mass, absolute mass ofthe heart, elevated plasma insulin concentration, dyslipidemia and insulin resistance. At the molecularlevel, DHT treatment did not change protein expression of cardiac insulin receptor and insulin receptorsubstrate 1, while phosphorylation of the substrate at serine 307 was increased. Unexpectedly, althoughexpression of downstream Akt kinase and its phosphorylation at threonine 308 were not altered, phos-phorylation of Akt at serine 473 was increased in the heart of DHT-treated rats. In contrast, expressionand phosphorylation of extracellular signal regulated kinases 1/2 were decreased. Plasma membranecontents of GLUT1 and GLUT4 were decreased, as well as the expression of GLUT4 in cardiac cells at theend of androgen treatment. The obtained results provide evidence for alterations in expression and espe-cially in functional characteristics of insulin signaling molecules and glucose transporters in the heart ofDHT-treated rats with PCOS, indicating impaired cardiac insulin action.

© 2014 Elsevier Ltd. All rights reserved.

Abbreviations: PCOS, polycystic ovary syndrome; DHT, dihydrotestosterone;GLUT, glucose transporter; IR, insulin receptor; IRS, insulin receptor substrate;MAPK, mitogen-activated protein kinase; Akt, protein kinase B; ERK1/2, extracellu-lar signal regulated kinases 1/2; BCA, bicinchoninic acid; HOMA, homeostasis modelassessment; PM, plasma membrane; LDM, low density microsomes; TG, triglyc-erides; NEFA, nonesterified fatty acids.

∗ Corresponding author at: Laboratory for Molecular Biology and Endocrinology,Vinca Institute of Nuclear Sciences, PO Box 522, 11001 Belgrade, Serbia.Tel.: +381 11 6442532; fax: +381 11 6455561.

E-mail addresses: sradivojsa@vinca.rs (S. Tepavcevic), dvojnovic@ibiss.bg.ac.rs(D. Vojnovic Milutinovic), djmacut@gmail.com (D. Macut), zzakula@vinca.rs(Z. Zakula), mnikolic@ibiss.bg.ac.rs (M. Nikolic), dr.ivana.bozic@gmail.com(I. Bozic-Antic), snjezana.romic@vinca.rs (S. Romic), jbjekic@yahoo.com(J. Bjekic-Macut), gormatic@ibiss.bg.ac.rs (G. Matic), gogi@vinca.rs (G. Koricanac).

1. Introduction

Polycystic ovary syndrome (PCOS) is a common repro-ductive disorder frequently associated with various metabolicabnormalities, and substantially leading to the increased risk fortype 2 diabetes [1]. According to the Androgen Excess and PCOS(AE-PCOS) Society, PCOS is defined by the obligatory presence ofhyperandrogenism, combined with ovarian dysfunction indicatedby oligo-ovulation and/or polycystic ovaries, after exclusion ofother related disorders [2].

According to the literature data PCOS may have its onsetbefore or during puberty and is associated with excessive andro-gen production during early puberty [3–5]. Since PCOS is primaryhyperandrogenic state, several androgens have been used toinduce PCOS conditions in rats. In this study, we used 5�-dihydrotestosterone (DHT) to induce a state of androgen excessin prepubertal age of rats. DHT is non-aromatizable androgen with

0960-0760/$ – see front matter © 2014 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.jsbmb.2014.01.006

Author's personal copy

72 S. Tepavcevic et al. / Journal of Steroid Biochemistry & Molecular Biology 141 (2014) 71–76

high affinity for the androgen receptor. The chosen dose of DHT wasused to induce a hyperandrogenic state mimicking that is seen inwomen with PCOS, whose plasma DHT levels are 1.7 times higherthan those of healthy controls [6,7]. As Manneras et al. [8] havedemonstrated, treatment of juvenile rats with DHT induces bothreproductive and metabolic abnormalities similar to those seen inwomen with PCOS. They have shown that DHT treated rats hadirregular estrous cycles and other ovarian features similar to humanPCOS, including increased numbers of large atretic follicles andfollicular cysts with diminished granulosa layer.

Insulin resistance and hyperinsulinemia appear to be majorcontributors to the pathophysiology of PCOS. There is a generalagreement that obese women with PCOS are insulin resistant,while some groups of lean affected women may have normalinsulin sensitivity [9].

Insulin is a hormone involved in regulation of cardiacmetabolism, contractility, protein synthesis, growth, apoptosis andsurvival of cardiomyocytes, and myocardial blood perfusion [10].When insulin binds to insulin receptor (IR), it induces the phos-phorylation of IR substrates (IRSs) at tyrosine residues, therebyactivating a complex signal transduction network. The phos-phatidylinositol 3-kinase (PI3K) and the mitogen-activated proteinkinase (MAPK) pathways are two main pathways of that network.PI3K is considered to be the main player of the metabolic actionof insulin, whereas the MAPK pathway is principally involved incell growth and differentiation. A key effector of the PI3K pathwayis the protein kinase termed protein kinase B (Akt). Akt phospho-rylates a variety of intracellular substrates, regulating cell growth,metabolism, and survival [11]. Insulin stimulates glucose uptakeby increasing the translocation of the insulin-responsive glucosetransporters (GLUT). The most abundant GLUT in the heart is GLUT4,which is insulin-regulated via PI3K/Akt pathway, while GLUT1 isprimarily responsible for basal glucose transport [12].

Post-binding defect in IR signaling has been detected in PCOS,likely due to increased receptor and IRS-1 serine phosphorylationthat selectively affects metabolic, but not mitogenic pathway inclassic insulin target tissues and in the ovary [9]. On the other hand,constitutive activation of serine kinases in the MAPK-extracellularsignal regulated kinases 1/2 (ERK1/2) pathway may contribute toresistance to insulin metabolic actions in skeletal muscle. Thesefindings have been directly translated into PCOS therapy withinsulin-sensitizing drugs [9,13]. In addition, a growing body ofevidence supports a notion that women with PCOS display anincreased prevalence of cardiovascular disease risk factors, mostof which are etiologically correlated with insulin resistance and aconsequent hyperinsulinemia [14].

Although insulin resistance was already studied in skeletalmuscles [15–20], adipose tissue [20–23], fibroblasts [24,25] orendometria [26,27] of women with PCOS and in skeletal musclesof experimental animals with PCOS [28], there are no data regard-ing insulin signaling and action in the heart of women with PCOSor animal models of the syndrome. To shed new light on the issueof possible PCOS-accompanied cardiac insulin resistance, femaleWistar rats were exposed to a 90-days long treatment with subcuta-neous DHT pellets to obtain an animal model of PCOS [8]. At the endof the treatment, the protein level of insulin signaling molecules (IR,IRS-1, Akt and ERK1/2) and glucose transporters GLUT1 and GLUT4,as well as their activating/inhibiting phosphorylation or subcellularlocalization were analyzed in the whole heart.

2. Materials and methods

2.1. Materials

Anti-phospho-IRS-1 (Ser307) antibody (07–247) was a productof Merck Millipore (Billerica, MA, USA). Anti-Akt 1/2/3 antibody

(#9272), anti-phospho-ERK1/2 (Thr202/Tyr204) antibody (#9101)and anti-ERK1/2 antibody (#9102) were products of Cell Sig-naling Technology, Inc. (Danvers, MA, USA). Anti-IR� (sc-711),anti-IRS-1 (sc-8038), anti-phospho Akt 1/2/3 (Ser473) (sc-7985-R),anti-phospho Akt 1/2/3 (Thr308) (sc-16646-R), anti-GLUT4 antibody(sc-7938), anti-GLUT1 (sc-7903), and anti-actin antibody (sc-1616-R) were products of Santa Cruz Biotechnology, Inc. (Santa Cruz, CA,USA). RIA kit for insulin was a product of INEP (Zemun, Serbia).Reagents for the bicinchoninic acid (BCA) assay were purchasedfrom Pierce (Rockford, IL, USA). DHT and placebo pellets werepurchased from Innovative Research of America (Sarasota, FL).Electrophoretic reagents were purchased from Sigma–Aldrich Cor-poration (St. Louis, MO, USA).

2.2. Animals and treatment

At 21-days of age female Wistar rats were separated from lactat-ing dams and randomly divided into two groups (n = 12 per group).At that age rats were implanted subcutaneously on the back ofthe neck with 90-days continuous-release pellets containing 7.5 mgDHT (daily dose, 83 �g, DHT group) or 7.5 mg placebo pellets (con-trol group).

The animals were housed three per cage and kept in atemperature-controlled room (22 ± 2 ◦C) with a 12 h light/darkcycle (lights on at 07:00 h) and constant humidity. Rats were fedwith commercial chow and drinking water available ad libitum. Atthe end of the experimental period rats were killed by decapitationin the diestrus phase of estrous cycle. All animal procedures were incompliance with the EEC Directive (86/609/EEC) on the protectionof animals used for experimental and other scientific purposes, andwere approved by the Ethical Committee for the Use of LaboratoryAnimals of the Institute for Biological Research “Sinisa Stankovic”,University of Belgrade (No. 2-20/10).

2.3. Glucose and insulin concentration determination and HOMAindex calculation

For measurement of glucose and insulin concentration, animalswere fasted overnight before collection of blood samples to avoidchanges of insulin and glucose induced by food intake. The glu-cose concentration was measured in whole blood using Accutrendglucometer (Roche Diagnostics GmbH, Mannheim, Germany). Forthe determination of plasma insulin concentration blood was col-lected at decapitation in EDTA-pretreated tubes and centrifugedat 3000 rpm for 10 min to obtain plasma samples. The plasmainsulin level was determined by the RIA method, using ratinsulin standards. The reference level of rat fasting insulin was12.06–48.26 mIU/l. Assay sensitivity was 0.6 mIU/l and an intra-assay coefficient of variation was 5.24%.

Homeostasis model assessment (HOMA) index, as an indicatorof insulin resistance, was calculated from fasted plasma insulin andglucose concentration using the formula described by Matthewset al. [29]: insulin (mU/l) × [glucose (mmol/l)/22.5].

2.4. Measurement of circulating triacylglycerols andnon-esterified fatty acids

The TG level was measured using a Multicare analyzer (Bio-chemical Systems International, Arezzo, Italy) and the plasmanon-esterified fatty acid (NEFA) level was determined using a col-orimetric method [30].

2.5. Preparation of cardiac cell lysate

Cardiac tissue samples were homogenized at 4 ◦C with an Ultra-Turrax homogenizer (3× 30 s) in four volumes (m:V) of modified

Author's personal copy

S. Tepavcevic et al. / Journal of Steroid Biochemistry & Molecular Biology 141 (2014) 71–76 73

RIPA buffer (50 mmol/l Tris–HCl, pH 7.4, 150 mmol/l NaCl, 1%Triton X-100, 0.2% Na-deoxycholate, 0.2% SDS, 1 mM EDTA, pro-tease inhibitors, phosphatase inhibitors). The homogenate wascentrifuged at 4 ◦C, 15,000 × g, 30 min. Obtained supernatant wasused as a cardiac cell lysate. Protein concentration was determinedby BCA method according to the manufacturer’s instruction and allsamples were prepared for Western blot in Laemmli sample buffer[31].

2.6. Preparation of cardiac cell plasma membranes andlow-density microsomes

Hearts were removed immediately after decapitation, washedby cold saline, dried, excised in small pieces and incubated for30 min in a cold high-salt solution (2 mol/l NaCl, 20 mmol/l HEPESpH 7.4, and 5 mmol/l NaN3). Thereafter, the samples were cen-trifuged for 5 min at 1000 × g and the pellet homogenized inTES buffer (20 mmol/l Tris, 250 mmol/l sucrose, 1 mmol/l EDTA,pH 7.4) supplemented with protease inhibitors (10 �g/ml leu-peptin, 10 �g/ml aprotinin, 2 mM phenylmethylsulfonyl fluoride),using an Ultra-Turrax homogenizer. The resulting homogenate wascentrifuged for 5 min at 1000 × g, after which the pellet was reho-mogenized in TES buffer and then recombined with the 1000 × gsupernatant. The sample was submitted to further sequential cen-trifugations:

1. Homogenate – 10 min at 100 × g – pellet P12. Supernatant – 10 min at 5000 × g – pellet P23. Supernatant – 20 min at 20,000 × g – pellet P34. Supernatant – 30 min at 48,000 × g – pellet P45. Supernatant – 80 min at 200,000 × g – pellet P5

All pellets were resuspended in TES buffer. According to sug-gestion of the authors of original protocol [32], we decided to referto P2 as plasma membrane (PM) fraction and to P5 as low densitymicrosomes (LDM) fraction. Protein concentration was determinedby the BCA assay, using bovine serum albumin as a standard.

2.7. SDS polyacrylamide electrophoresis and Western blot

Cardiac lysate, PM or LDM proteins were separated on 7.5% or10% SDS polyacrylamide gels and transferred to PVDF membranes.The membranes were blocked with 5% bovine serum albumin andblotted with an antibody against IR, IRS-1, Akt, ERK1/2, GLUT1 orGLUT4. After extensive washing, membranes were incubated withsecondary HRP-conjugated antibody and used for detection withECL reagents. To ensure that protein loading was equal in all lysatesamples, blots were striped and reprobed with the actin antibody.Films were scanned and analyzed using ImageJ software (NIH, USA).

2.8. Statistics

Values are expressed as the means ± SD. The experiments wereperformed with twelve animals per group, according to standardsof statistics and ethics. The significance of differences betweentwo groups (DHT group vs. control group) was estimated by theStudent’s t-test. A value of p < 0.05 was considered statistically sig-nificant.

3. Results

As presented in Table 1, blood glucose concentration was notaltered, while plasma insulin concentration and HOMA index weresignificantly higher in DHT-treated rats in comparison to con-trol rats (p < 0.05). Concentration of circulating TG and NEFA weresignificantly increased in DHT-treated rats (p < 0.001 and 0.05,

Table 1Physical and biochemical characteristics of experimental animals. Glucose andtriglycerides were measures in the whole blood, while insulin and nonesterifiedfatty acid concentration were determined in the plasma. HOMA index was calcu-lated from data for fasting glucose and insulin concentration. Body mass and massof the heart were determined at the end of the experiment. Data are presented asmean ± SD from the experiment with 12 animals per group.

Control rats DHT-treated rats

Body mass (g) 255.10 ± 13.50 295.80 ± 40.27**

Mass of the heart (g) 0.71 ± 0.07 0.81 ± 0.10**

Mass of the heart/body mass 0.28 ± 0.018 0.27 ± 0.02Glucose (mmol/l) 3.48 ± 0.49 3.20 ± 0.43Insulin (mIU/l) 9.61 ± 5.15 18.83 ± 12.17*

HOMA index 1.47 ± 0.72 2.62 ± 1.75*

TG (mmol/l) 1.09 ± 0.17 1.59 ± 0.34***

NEFA (mmol/l) 0.46 ± 0.10 0.59 ± 0.14*

Abbreviations: DHT, dihydrotestosterone; HOMA, homeostasis model assessment;TG, triglycerides; NEFA, nonesterified fatty acids.

* p < 0.05.** p < 0.01.

*** p < 0.001.

respectively). Body mass and mass of the heart of androgen-treatedrats were significantly higher (p < 0.01 for both parameters), butrelative heart-to-body mass was not altered by the hormone treat-ment (Table 1).

The expression of IR and IRS-1 protein in cardiac cells of DHT-treated rats was not changed compared to control ones. However,phosphorylation of IRS-1 at the serine 307 (expressed as phos-pho/total IRS-1), which is inhibitory and connected with insulinresistance, was significantly increased in DHT-treated rats (Fig. 1,p < 0.05).

Akt kinase protein content in cardiac cells was not alteredby androgen treatment, while stimulatory phosphorylation of thekinase at serine 473 was increased in DHT-treated rats (p < 0.05).

Fig. 1. Cardiac IR and IRS-1 protein level and phosphorylation of IRS-1 at ser-ine 307. Protein level of insulin receptor (IR), insulin receptor substrate 1 (IRS-1)and phospho-IRS-1 Ser307 in the hearts of dihydrotestosterone-treated (DHT) andcontrol (C) rats were determined by Western blot using specific antibodies. Rep-resentative blots are placed above histograms. Results are expressed as mean ± SDand presented as percentage of control value. The content of phospho-IRS-1 wasadditionally calculated as a ratio of phosphorylated vs. total protein content. Com-parisons between DHT-treated and control rats were made by unpaired Student’st-test. Asterisks indicate significant differences. *p < 0.05.

Author's personal copy

74 S. Tepavcevic et al. / Journal of Steroid Biochemistry & Molecular Biology 141 (2014) 71–76

Fig. 2. Protein expression and phosphorylation of Akt kinase at serine 473 and thre-onine 308 in the cardiac cells. Protein content of Akt and its phosphorylations weredetermined in the hearts of dihydrotestosterone-treated (DHT) and control (C) ratsby Western blot using specific antibodies against the Akt 1/2/3. Representative blotsare placed above histograms. Results are expressed as mean ± SD and presented aspercentage of control value. The content of phospho-Akt was additionally calculatedas a ratio of phosphorylated vs. total protein content. Comparisons between DHT-treated and control rats were made by unpaired Student’s t-test. Asterisks indicatesignificant differences. *p < 0.05.

Phosphorylation of Akt at threonine 308 residue, expressed per totalAkt content, did not differ in androgen-treated vs. control animals(Fig. 2).

In contrast to Akt, the expression of mitogen insulin signalingpathway members, ERK1/2 kinases, was lower in DHT-treated rats(p < 0.05), as well as their phosphorylation (phospho ERK1/2 pertotal ERK1/2) at threonine 202/tyrosine 204 (Fig. 3, p < 0.05).

The expression of main cardiac glucose transporter, GLUT4, wasdecreased in DHT-treated rats (p < 0.05). In concert with this, PMcontent of GLUT4 was lower than in control rats (p < 0.05), whileGLUT4 content in intracellular pool did not change (Fig. 4).

Similar to GLUT4, content of GLUT1 protein, which is responsiblefor basal glucose transport in cardiac cells, was decreased in PMfraction of DHT-treated rats (p < 0.05). Total expression and LDMcontent of GLUT1 was not altered by DHT treatment (Fig. 5).

4. Discussion

Insulin resistance is a recognized feature of PCOS with a preva-lence rate of 44–70% [9]. However, the molecular mechanisms ofpathophysiology of the reduced insulin sensitivity is not completelyclarified. Some of the existing data have shown that a postbindingdefect in insulin signaling could be an underlying mechanism ofinsulin resistance in women with PCOS [13,16]. To the best of ourknowledge, this is the first study showing insulin resistance in theheart of experimental animals or humans with PCOS.

In our study, chronic treatment with DHT induced continuousanestrous and decreased weight of ovaries and uteri, while ren-dering the concentration of plasma estradiol unaltered (data notshown). Accordingly, the animals treated with DHT were anovu-latory and hyperandrogenic, which fulfilled AE-PCOS criteria fordiagnosis of PCOS.

In the present study increased insulin plasma concentration andthe consequent increase of HOMA index, accompanied with dys-lipidemia were observed in DHT-treated rats, indicating impaired

Fig. 3. Protein expression and phosphorylation of ERK1/2 kinase at threonine202/tyrosine 204 in the heart. Protein content and phosphorylations at threonine202/tyrosine 204 of extracellular signal regulated kinases 1/2 (ERK1/2) were deter-mined by Western blot using specific antibodies. The representative blots are placedabove histograms. Results are expressed as mean ± SD and presented as percent-age of control value. The content of phospho-ERK1/2 was additionally calculatedas a ratio of phosphorylated vs. total protein content. Comparisons between DHT-treated (DHT) and control (C) rats were made by unpaired Student’s t-test. Asterisksindicate significant differences. *p < 0.05.

insulin sensitivity. Alteration of cardiac IR� level related to DHTtreatment was not detected, which is in accordance with findingsin other tissues from women with PCOS [15,23,27]. However, mod-ification at the level of IR phosphorylation was described, in termsof increased serine phosphorylation [33] and decreased tyrosinephosphorylation in fibroblasts from women with PCOS [34].

Protein expression of IRS-1 in the heart was also not changedin DHT-treated rats in the present study, while IRS-1 phospho-rylation at serine 307, as a mark of tissue insulin resistance, was

Fig. 4. Protein expression and subcellular localization of GLUT4 in the heart. Pro-tein content of GLUT4 was determined by Western blot in lysate, plasma membranes(PM) and low density microsomes (LDM) using specific antibody. The representativeblots are placed above histograms. Results are expressed as mean ± SD and pre-sented as percentage of control value. Comparisons between DHT-treated (DHT) andcontrol (C) rats were made by unpaired Student’s t-test. Asterisks indicate significantdifferences. *p < 0.05.

Author's personal copy

S. Tepavcevic et al. / Journal of Steroid Biochemistry & Molecular Biology 141 (2014) 71–76 75

Fig. 5. Protein expression and subcellular localization of cardiac GLUT1. Protein con-tent of GLUT1 was determined by Western blot in lysate, plasma membranes (PM)and low density microsomes (LDM) using specific antibody. The representative blotsare placed above histograms. Results are expressed as mean ± SD and presented aspercentage of control value. Comparisons between DHT-treated (DHT) and con-trol (C) rats were made by unpaired Student’s t-test. Asterisk indicates significantdifference. *p < 0.05.

increased. No significant differences in the abundance of IRS-1in muscles and preadipocytes from women with PCOS were alsoobserved [16,19,23], while IRS-1 protein level in cultured myotubesfrom women with PCOS was shown to be increased [15]. Increasedinhibitory phosphorylation of IRS-1 could be related to increasedinsulin concentration or increased circulating NEFA in DHT-treatedrats [35]. In line with this, decreased IRS-1 tyrosine phospho-rylation in adipose tissue and endometria from PCOS patientswas shown to be related to insulin resistance [21]. Consequently,changes of IRS-1/p85 association accompanied with PCOS were alsodescribed [16].

Previous experiments showed no changes in the expressionor maximal insulin stimulation of Akt phosphorylation in skele-tal muscles, myotubes, adipocytes or fibroblasts from women withPCOS [20,25] or in primary cultures of rat myoblasts pre-exposedto testosterone [28]. In contrast, reduced insulin-stimulated Aktphosphorylations at serine 473 and threonine 308 were observed inmuscles of PCOS patients [17]. Although there were no alterationsin Akt protein expression and Akt phosphorylation at threonine 308in the heart of DHT-treated rats in the present study, phosphory-lation of Akt at serine 473 was increased. This unexpected result,which is opposite to the alterations of upstream and downstreammolecules, could be related to the fact that the used antibodydetects phosphorylation of all three Akt isoforms having differ-ent functions in the heart [36,37]. For instance, possible increase inphosphorylation of Akt 1, regarding its function in cardiac growth[36], could be connected with the observed gain of the absolutemass of the heart. To analyze the effect of DHT treatment on Akt 2phosphorylation, which is responsible for metabolic effects of Aktsignaling [37], further experiments with specific antibody againstthis isoform should be performed.

In contrast to Akt, expression and activating phosphorylationof ERK1/2, were significantly lower in the heart of DHT-treatedrats, suggesting downregulation of insulin mitogenic pathway.Rajkhowa et al. [19] showed an attenuation of insulin stimulationof the ERK pathway and trend toward higher basal phosphoryla-tion of ERK1/2 in the muscles from women with PCOS. However,ERK1/2 activation was shown to be increased in skeletal musclesand in cultured myotubes from women with PCOS, both basally andin response to insulin [18]. On the other hand, ERK1/2 activation didnot alter in preadipocytes from women with PCOS [23].

Our finding of decreased content of GLUT4 in lysate and PMfraction of cardiac cells of DHT-treated rats, is in accordance withthe data showing decreased GLUT4 content in adipocytes fromwomen with PCOS [22]. In primary cultures of adipocytes fromPCOS patients, GLUT4 gene expression and protein level werereduced. GLUT4 gene expression was lower in PCOS with insulinresistance than in PCOS without it [38]. In addition, hyperinsuline-mic PCOS women exhibited lower levels of endometrial GLUT4 thaneuinsulinemic ones [27]. However, there were no differences in theexpression of GLUT4 in skeletal muscles, myotubes, or adipocytesfrom PCOS subjects in the study of Ciaraldi et al. [20]. Decrease ofGLUT4 in PM could be explained by a possible decrease in Akt 2 con-tent/phosphorylation in DHT-treated rats, or by downregulation ofother insulin signaling pathways that regulate GLUT4 transloca-tion, such as PKC �/� or cbl/CAP pathway [39], in the heart of ratswith PCOS. This would be in concert with other indication of car-diac insulin resistance in the present study (increased IRS-1 Ser 307phosphorylation, decreased ERK1/2 expression/phosphorylation,decreased plasma membrane GLUT1). On the other hand, increasedNEFA and TG, together with downregulation of PM GLUT4 andGLUT1 can also indicate enhanced uptake of fatty acid and redi-rection of the heart toward fatty acid metabolism in the insulinresistance conditions [40].

Contrary to the results in the present study, showing that PMcontent of GLUT1 protein is decreased in PCOS heart, which canbe also attributed to insulin signaling defects, because translo-cation of cardiac GLUT1 is insulin-regulated [12], basal andinsulin-stimulated glucose transport and GLUT1 abundance weresignificantly increased in cultured myotubes from women withPCOS [15].

It is questionable whether the effects of androgen treatment thatlead to impairment of cardiac insulin sensitivity are direct or indi-rect. The literature data suggest positive effects of androgen on Aktand ERK1/2 phosphorylation [41] and on GLUT4-dependent glucosetransport [42] in cardiomyocytes. We suggest that the observedchanges in cardiac insulin signaling and glucose transport in ouranimal model of PCOS are mediated by metabolic disturbancesand general impairment of insulin sensitivity, indicated by theobserved increase of insulin, HOMA index, circulating TG and NEFA.In addition, studies showing that treatment of PCOS with insulinsensitizers class of drugs, such as metformin or thiazolidinediones,which establish normal insulin sensitivity and androgen level, sug-gest that insulin resistance is crucial event in the genesis of PCOS[43].

There are no published data about cardiac androgen receptorin PCOS women or experimental animals. Cardiac cells of malesand females express the androgen receptor [44], and some andro-gen effects in the heart have been described [45], including theeffects on insulin signaling molecules [41] and glucose transport[42]. However, in order to make distinction between direct andindirect effects of androgen treatment, in vitro treatment of car-diomyocytes with androgen receptor antagonists would be anappropriate approach.

The results presented herein indicate that hyperinsulinemiaand insulin resistance in female rats treated with DHT are accom-panied by a decrease in expression of some molecules involved incardiac insulin signaling (ERK1/2) or insulin-regulated molecules(GLUT4), and particularly by an impairment of their functionalcharacteristics (increased phosphorylation of IRS-1 at serine 307,decreased phosphorylation of ERK1/2, decreased PM abundanceof GLUT4 and GLUT1). In conclusion, findings of the present studypoint out to impaired insulin action in the heart of DHT-treatedrats, which is in compliance with the notorious link betweenPCOS and insulin resistance. The precise molecular mechanismof impaired cardiac insulin signaling and glucose transport inDHT-treated rats remains to be elucidated. The contribution of

Author's personal copy

76 S. Tepavcevic et al. / Journal of Steroid Biochemistry & Molecular Biology 141 (2014) 71–76

direct vs. indirect effects of androgen treatment in the observedphenomenon is particularly intriguing issue. Eventually, it shouldbe very careful with extrapolation of insulin signaling deteriorationobserved in the animal model to the unfavorable cardiovascularoutcomes of PCOS in women.

Conflict of interest

No potential conflicts of interest were disclosed by the authors.

Acknowledgements

This study was supported by the Project grant No. III41009, fromthe Ministry of Education, Science and Technological Development,Republic of Serbia.

References

[1] G.W. Bates, R.S. Legro, Longterm management of polycystic ovarian syndrome(PCOS), Mol. Cell. Endocrinol. 373 (2013) 91–97.

[2] R. Azziz, E. Carmina, D. Dewailly, et al., The androgen excess and PCOS soci-ety criteria for the polycystic ovary syndrome: the complete task force report,Fertil. Steril. 91 (2009) 456–488.

[3] D. Apter, How possible is the prevention of polycystic ovary syndrome devel-opment in adolescent patients with early onset of hyperandrogenism, J.Endocrinol. Invest. 21 (1998) 613–617.

[4] S. Franks, Adult polycystic ovary syndrome begins in childhood, Best Pract. Res.Clin. Endocrinol. Metab. 16 (2002) 263–272.

[5] J. Bronstein, S. Tawdekar, Y. Liu, et al., Age of onset of polycystic ovarian syn-drome in girls may be earlier than previously thought, J. Pediatr. Adolesc.Gynecol. 24 (2011) 15–20.

[6] M. Fassnacht, N. Schlenz, S.B. Schneider, et al., Beyond adrenal and ovar-ian androgen generation: increased peripheral 5 alpha-reductase activity inwomen with polycystic ovary syndrome, J. Clin. Endocrinol. Metab. 88 (2003)2760–2766.

[7] M.E. Silfen, M.R. Denburg, A.M. Manibo, et al., Early endocrine, metabolic, andsonographic characteristics of polycystic ovary syndrome (PCOS): comparisonbetween nonobese and obese adolescents, J. Clin. Endocrinol. Metab. 88 (2003)4682–4688.

[8] L. Manneras, S. Cajander, A. Holmang, et al., A new rat model exhibitingboth ovarian and metabolic characteristics of polycystic ovary syndrome,Endocrinology 148 (2007) 3781–3791.

[9] E. Diamanti-Kandarakis, A. Dunaif, Insulin resistance and the polycystic ovarysyndrome revisited: an update on mechanisms and implications, Endocr. Rev.33 (2012) 981–1030.

[10] F. Iliadis, N. Kadoglou, T. Didangelos, Insulin and the heart, Diabetes Res. Clin.Pract. 93 (Suppl. 1) (2011) S86–S91.

[11] B.J. DeBosch, A.J. Muslin, Insulin signaling pathways and cardiac growth, J. Mol.Cell. Cardiol. 44 (2008) 855–864.

[12] J.J. Luiken, S.L. Coort, D.P. Koonen, et al., Regulation of cardiac long-chain fattyacid and glucose uptake by translocation of substrate transporters, PflugersArch. 448 (2004) 1–15.

[13] E. Diamanti-Kandarakis, A.G. Papavassiliou, Molecular mechanisms of insulinresistance in polycystic ovary syndrome, Trends Mol. Med. 12 (2006) 324–332.

[14] E. Kassi, E. Diamanti-Kandarakis, The effects of insulin sensitizers on the cardio-vascular risk factors in women with polycystic ovary syndrome, J. Endocrinol.Invest. 31 (2008) 1124–1131.

[15] A. Corbould, Y.B. Kim, J.F. Youngren, et al., Insulin resistance in the skeletalmuscle of women with PCOS involves intrinsic and acquired defects in insulinsignaling, Am. J. Physiol. Endocrinol. Metab. 288 (2005) E1047–E1054.

[16] A. Dunaif, X. Wu, A. Lee, et al., Defects in insulin receptor signaling in vivo inthe polycystic ovary syndrome (PCOS), Am. J. Physiol. Endocrinol. Metab. 281(2001) E392–E399.

[17] K. Hojlund, D. Glintborg, N.R. Andersen, et al., Impaired insulin-stimulatedphosphorylation of Akt and AS160 in skeletal muscle of women with polycys-tic ovary syndrome is reversed by pioglitazone treatment, Diabetes 57 (2008)357–366.

[18] A. Corbould, H. Zhao, S. Mirzoeva, et al., Enhanced mitogenic signaling in skele-tal muscle of women with polycystic ovary syndrome, Diabetes 55 (2006)751–759.

[19] M. Rajkhowa, S. Brett, D.J. Cuthbertson, et al., Insulin resistance in polycysticovary syndrome is associated with defective regulation of ERK1/2 by insulin inskeletal muscle in vivo, Biochem. J. 418 (2009) 665–671.

[20] T.P. Ciaraldi, V. Aroda, S. Mudaliar, et al., Polycystic ovary syndrome is asso-ciated with tissue-specific differences in insulin resistance, J. Clin. Endocrinol.Metab. 94 (2009) 157–163.

[21] Y.L. Chu, Y.Y. Sun, H.Y. Qiu, et al., [Tyrosine phosphorylation and proteinexpression of insulin receptor substrate-1 in the patients with polycystic ovarysyndrome], Zhonghua Fu Chan Ke Za Zhi 39 (2004) 176–179.

[22] D. Rosenbaum, R.S. Haber, A. Dunaif, Insulin resistance in polycystic ovary syn-drome: decreased expression of GLUT-4 glucose transporters in adipocytes,Am. J. Physiol. 264 (1993) E197–E202.

[23] A. Corbould, A. Dunaif, The adipose cell lineage is not intrinsically insulin resis-tant in polycystic ovary syndrome, Metabolism 56 (2007) 716–722.

[24] C.B. Book, A. Dunaif, Selective insulin resistance in the polycystic ovary syn-drome, J. Clin. Endocrinol. Metab. 84 (1999) 3110–3116.

[25] A.M. Venkatesan, A. Dunaif, A. Corbould, Insulin resistance in polycysticovary syndrome: progress and paradoxes, Recent Prog. Horm. Res. 56 (2001)295–308.

[26] C. Rosas, F. Gabler, D. Vantman, et al., Levels of Rabs and WAVE family proteinsassociated with translocation of GLUT4 to the cell surface in endometria fromhyperinsulinemic PCOS women, Hum. Reprod. 25 (2010) 2870–2877.

[27] R. Fornes, P. Ormazabal, C. Rosas, et al., Changes in the expression of insulinsignaling pathway molecules in endometria from polycystic ovary syndromewomen with or without hyperinsulinemia, Mol. Med. 16 (2010) 129–136.

[28] M.C. Allemand, B.A. Irving, Y.W. Asmann, et al., Effect of testosterone on insulinstimulated IRS1 Ser phosphorylation in primary rat myotubes—a potentialmodel for PCOS-related insulin resistance, PLoS One 4 (2009) e4274.

[29] D.R. Matthews, J.P. Hosker, A.S. Rudenski, et al., Homeostasis model assess-ment: insulin resistance and beta-cell function from fasting plasma glucoseand insulin concentrations in man, Diabetologia 28 (1985) 412–419.

[30] W.G. Duncombe, The colorimetric micro-determination of non-esterified fattyacids in plasma, Clin. Chim. Acta 9 (1964) 122–125.

[31] U.K. Laemmli, Cleavage of structural proteins during the assembly of the headof bacteriophage T4, Nature 227 (1970) 680–685.

[32] J.J. Luiken, D.P. Koonen, J. Willems, et al., Insulin stimulates long-chain fatty acidutilization by rat cardiac myocytes through cellular redistribution of FAT/CD36,Diabetes 51 (2002) 3113–3119.

[33] A. Dunaif, J. Xia, C.B. Book, et al., Excessive insulin receptor serine phosphory-lation in cultured fibroblasts and in skeletal muscle. A potential mechanism forinsulin resistance in the polycystic ovary syndrome, J. Clin. Invest. 96 (1995)801–810.

[34] M. Li, J.F. Youngren, A. Dunaif, et al., Decreased insulin receptor (IR)autophosphorylation in fibroblasts from patients with PCOS: effects of ser-ine kinase inhibitors and IR activators, J. Clin. Endocrinol. Metab. 87 (2002)4088–4093.

[35] P. Gual, Y. Le Marchand-Brustel, J.F. Tanti, Positive and negative regulation ofinsulin signaling through IRS-1 phosphorylation, Biochimie 87 (2005) 99–109.

[36] B. DeBosch, I. Treskov, T.S. Lupu, et al., Akt1 is required for physiological cardiacgrowth, Circulation 113 (2006) 2097–2104.

[37] A.J. Muslin, Akt2: a critical regulator of cardiomyocyte survival and metabolism,Pediatr. Cardiol. 32 (2011) 317–322.

[38] Y.H. Chen, S. Heneidi, J.M. Lee, et al., miRNA-93 inhibits GLUT4 and is overex-pressed in adipose tissue of polycystic ovary syndrome patients and womenwith insulin resistance, Diabetes 62 (2013) 2278–2286.

[39] C. Montessuit, R. Lerch, Regulation and dysregulation of glucose transport incardiomyocytes, Biochim. Biophys. Acta 1833 (2013) 848–856.

[40] S.L. Coort, A. Bonen, G.J. van der Vusse, et al., Cardiac substrate uptake andmetabolism in obesity and type-2 diabetes: role of sarcolemmal substratetransporters, Mol. Cell. Biochem. 299 (2007) 5–18.

[41] F. Altamirano, C. Oyarce, P. Silva, et al., Testosterone induces cardiomyocytehypertrophy through mammalian target of rapamycin complex 1 pathway, J.Endocrinol. 202 (2009) 299–307.

[42] C. Wilson, A. Contreras-Ferrat, N. Venegas, et al., Testosterone increases GLUT4-dependent glucose uptake in cardiomyocytes, J. Cell. Physiol. 228 (2013)2399–2407.

[43] M. Jensterle, A. Janez, B. Mlinar, et al., Impact of metformin and rosiglitazonetreatment on glucose transporter 4 mRNA expression in women with polycysticovary syndrome, Eur. J. Endocrinol. 158 (2008) 793–801.

[44] E. Lizotte, S.A. Grandy, A. Tremblay, et al., Expression, distribution and regula-tion of sex steroid hormone receptors in mouse heart, Cell. Physiol. Biochem.23 (2009) 75–86.

[45] K.L. Golden, J.D. Marsh, Y. Jiang, Testosterone regulates mRNA levels of cal-cium regulatory proteins in cardiac myocytes, Horm. Metab. Res. 36 (2004)197–202.