adipositas-related gene expression in the atria

International Journal of Cardiology 187 (2015) 604–613

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

Atrial fibrillation and rapid acute pacing regulate adipocyte/ adipositas-related gene expression in the atria R.K. Chilukoti a,1, A. Giese b,1, W. Malenke b,1, G. Homuth a, A. Bukowska c, A. Goette c,d, S.B. Felix e, J. Kanaan f, H.-G. Wollert f, K. Evert g, S. Verheule h, P. Jais i, S.N. Hatem j, U. Lendeckel b,⁎, C. Wolke b a

University Medicine Greifswald, Ernst-Moritz-Arndt-University Greifswald, Interfaculty Institute for Genetics and Functional Genomics, Greifswald, Germany University Medicine Greifswald, Institute of Medical Biochemistry and Molecular Biology, Greifswald, Germany EUTRAF Working Group: Molecular Electrophysiology, University Hospital Magdeburg, Otto-von-Guericke University, Magdeburg, Germany d Cardiology and Intensive Care Medicine, St. Vincenz-Hospital, Paderborn, Germany e University Medicine Greifswald, Department of Cardiology, Greifswald, Germany f Dr. Guth Clinics, Dept. of Cardiovascular Surgery, Karlsburg, Germany g University Medicine Greifswald, Department of Pathology, Greifswald, Germany h Department of Physiology, Cardiovascular Research Institute Maastricht, Maastricht University, Maastricht, The Netherlands i Hôpital Cardiologique du Haut Lévêque, Université Victor-Segalen Bordeaux II, Pessac, France j Sorbonne Universités, UPMC University Paris 06, UMR_S 1166 I, ICAN, Paris, France b c

a r t i c l e

i n f o

Article history: Received 14 December 2014 Received in revised form 19 February 2015 Accepted 3 March 2015 Available online 23 March 2015 Keywords: Atrial fibrillation Atrial adipositas Adipocyte differentiation Structural remodeling

a b s t r a c t Purpose: Atrial fibrillation (AF) has been associated with increased volumes of epicardial fat and atrial adipocyte accumulation. Underlying mechanisms are not well understood. This study aims to identify rapid atrial pacing (RAP)/AF-dependent changes in atrial adipocyte/adipositas-related gene expression (AARE). Methods: Right atrial (RA) and adjacent epicardial adipose tissue (EAT) samples were obtained from 26 patients; 13 with AF, 13 in sinus rhythm (SR). Left atrial (LA) samples were obtained from 9 pigs (5 RAP, 4 sham-operated controls). AARE was analyzed using microarrays and RT-qPCR. The impact of diabetes/obesity on gene expression was additionally determined in RA samples (RAP ex vivo and controls) from 3 vs. 6 months old ZDF rats. Results: RAP in vivo of pigs resulted in substantial changes of AARE, with 66 genes being up- and 53 downregulated on the mRNA level. Differential expression during adipocyte differentiation was confirmed using 3T3-L1 cells. In patients with AF (compared to SR), a comparable change in RA mRNA levels concerned a fraction of genes only (RETN, IGF1, HK2, PYGM, LOX, and NR4A3). RA and EAT were affected by AF to a different extent. In patients, concomitant disease contributes to AARE changes. Conclusions: RAP, and to lesser extent AF, provoke significant changes in atrial AARE. In chronic AF, activation of this gene panel is very likely mediated by AF itself, AF risk factors and concomitant diseases. This may facilitate the development of an AF substrate by increasing atrial ectopic fat and fat infiltration of the atrial myocardium. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Obesity and metabolic syndrome are important risk factors for the development and progression of atrial fibrillation (AF) [1–5]. Previous studies showed an up to 8% increased risk for new-onset AF with each unit increment of the body mass index (BMI) [1,3]. In the general population, obesity leads to a 49% increased risk for AF [2]. Total visceral, epicardial, and intrathoracic fat exert, however, different effects on the cardiovascular system [6,7]. Recent work revealed that the epicardial adipose tissue (EAT) volume is highly associated with paroxysmal and

⁎ Corresponding author at: University Medicine Greifswald, Institute of Medical Biochemistry and Molecular Biology, Ferdinand-Sauerbruch-Strasse, D-17475 Greifswald, Germany. E-mail address: [email protected] (U. Lendeckel). 1 Contributed equally.

http://dx.doi.org/10.1016/j.ijcard.2015.03.072 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.

persistent AF and this is independent of “classical” risk factors such as left atrial diameter (LAD) [8]. Abundance of EAT is independently related to AF recurrence after ablation [9]. Total and LA EAT volumes as well as left periatrial EAT thickness are greater in persistent AF versus paroxysmal or no AF [9–12]. EAT distribution around the LA is uneven with most EAT located within regions superior vena cava, right pulmonary artery, right-sided roof of the LA, aortic root, pulmonary trunk, left atrial appendage, and between left inferior PV and left atrioventricular groove [9]. EAT locations were associated with high dominant frequency (DF) sites, providing a mechanism for EAT's contribution to AF maintenance [11,13]. Recent data suggests that the contribution of EAT to the AF substrate may differ between LA and RA [13]. By releasing e.g. adipokines and pro-inflammatory cytokines, pericardial adipose tissue contributes to the established association of AF with inflammation and obesity [12,14–17]. Elevated levels of CRP, interleukin-6 (IL6), IL-8, tumor necrosis factor 1α (TNF1α), and of the adipokine resistin

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

have been associated with AF [18]. Elevated post-operative serum levels of resistin seem to increase the risk of AF after coronary artery bypass graft surgery [19]. In addition to these humoral effects, epicardial adipocytes can modulate the electrophysiological properties of neighboring cardiomyocytes by direct interaction [20]. Adipocyte accumulation within the myocardium may disturb atrial conduction and favor the development and persistence of re-entry circuits [21]. The origin of myocardial adipose infiltration is not fully understood. Adipocytes may develop from resident or recruited progenitor cells, or invade the myocardium from the adjacent ectopic epicardial fat. Both mechanisms may be supported by AF-dependent changes in atrial (or cardiac) gene expression. The activation of the local renin–angiotensin-system and in particular its classical angiotensin II (AngII) — angiotensin II type I receptor (AT1R) axis represents a hallmark of AF [22–25]. Similarly, increased local and systemic AngII levels have been associated with all four determinants of the metabolic syndrome, namely hypertension, hyperglycemia, obesity, and hyperlipidemia [26–28]. AngII affects the differentiation of progenitor cells and pre-adipocytes [29–32]. Accordingly, AT1R blockers and ACE inhibitors were shown to influence adipogenesis and adipocyte function, including the production and release of adipokines [33–38]. This study aimed (I) to determine the effect of acute rapid atrial pacing (RAP) in vivo on adipocyte-specific atrial gene expression, (II) to assess to what extent similar expression changes could be detected in ex vivo RA and EAT tissue samples from patients with AF or in SR, and (III) to determine the corresponding expression changes during adipocyte differentiation in vitro. 2. Materials and methods 2.1. Patients After written informed consent, right atrial appendages (RA) and epicardial adipose tissue (EAT) were obtained from patients undergoing cardiac bypass surgery or mitral/ aortic valve replacement. EAT was harvested directly from the aortic root and at the lateral

605

wall of the right atrium close to the terminal crest. Tissue samples were removed from 13 consecutive patients with AF and from 13 matched patients with no history of AF suffering from different diseases (SR) (Table 1). The study was approved by the local ethics committee of the University Medicine Greifswald (BB049/13).

2.2. Rapid atrial pacing (RAP) model The tissue samples used in this study were from the same animals as described previously [39]. The animal experiments were approved by the Institutional Animal Care and Use Committee of the University of Magdeburg. Briefly, pigs were subjected to closed chest RAP. In five animals RAP was performed for 7 h at a rate of 600 bpm (twice diastolic threshold, 2-ms pulse duration; RAP-group), and four pigs were instrumented without further intervention (sham). After the 7 h pacing period, chest and pericardial sac were opened and the heart was exposed. The tissue samples were immediately frozen in liquid nitrogen until further use.

2.3. ZDF rats and rapid pacing ex vivo of atrial tissue slices Male Zucker diabetic fatty (ZDF-Leprfa/Crl) were purchased from Charles River Laboratories (France) and housed in a constant temperature room with a 12-h dark/12-h light cycle. ZDF rats received high fat Purina 5008 diet (Charles River France) to accelerate the development of the disease. Animals were studied at 3 (early diabetic stage) and 6 (advanced diabetic stage) months of age. Animal care and treatments were conducted in conformity with the current version of the German Law on the Protection of Animals and according to the guidelines for ethical care of experimental animals of the European Union. RA tissue culture and rapid pacing were performed as described recently [40–42]. RA tissue slices were paced at 4 Hz (RAP) or 0.6 Hz (control) for 24 h (n = 4 per experimental group). Subsequently total RNA was prepared as described below and used for subsequent microarray analysis.

2.4. Cell culture 3T3-L1 cells were grown in high-glucose DMEM supplemented with 10% FCS, 1% penicillin–streptomycin and 2 mM L-glutamine (culture medium) which was replaced twice a week. They were incubated in a humidified atmosphere with 5% CO2 at 37 °C. At confluence (day 0), the medium was supplemented with 0.5 mM 1-methyl-3isobutylxanthine, 0.25 μM dexamethasone and 10 μg/ml insulin (all Sigma Aldrich, Steinheim, Germany ) to initiate adipocyte differentiation. The medium was replaced every 2 days by a culture medium containing 2.5 μg/ml insulin.

Table 1 Patient characteristics. Pat.-Nr.

Rh.

Age

Sex

BMI

CAD

VD

NYHA

LVEF

D/H

ACEi/ATRB

Statin

CRP

2 4 21 33 42 50 59 65 66 68 69 73 77

1 2 2 2 1 1 3 2 1 3 2 1 3 13 SR SR SR SR SR SR SR SR SR SR SR SR SR 13

77 75 74 73 75 66 76 74 73 50 74 68 74 71.5 72 78 73 57 64 74 76 76 76 72 70 73 78 72.2

m m m m f m f m m f f m m 9/4 f f m m m m f f m f m m m 8/5

23 23.2 33.9 32.8 45.2 27.8 23 30.8 35.2 37.1 21.8 25.6 25.5 29.6 28.2 31.2 28.3 34.8 25.9 25.9 21.7 30.5 32.6 23.7 33.5 27.3 23.0 28.2

3 3 3 3 – 3 – – 3 – – 3 3 8 3 3 3 2 2 3 – 3 3 – 2 – 2 10

– – MR AS AS – AS + MR AS – AS TR + MR – – 7 – – – – – – AS – – AS – AR – 5

II–III – III I II II II III II II–III II II II–III

35 38 55 50 55 55 45 55 50 30 55 35 50 55 55 55 60 60 70 55 58 60 55 40 50 55

0/1 1/0 1/0 0/1 0/0 1/0 1/0 1/0 1/0 0/1 1/0 1/0 0/1 8/4 0/0 1/0 1/0 1/0 1/0 1/0 1/0 1/0 1/0 0/1 1/0 0/1 1/0 10/2

1 0 1 1 0 1 0 0 0 1 0 1 1 7 1 0 1 1 1 1 0 1 1 1 1 1 0 10

18.6 6.9 2.5 8.9 8.9 b5 7.2 2.5 7.5 7.1 4.2 4.2 1.2

II II–III III III III II II II III II II II III

H H D+H H H H H H H H D D D+H 2/11 H D H – H D H H H H D+H H – 3/9

5 10 11 17 18 29 34 37 51 53 57 62 72

3.3 2.5 2.5 2.5 2.5 2.5 2.5 10.9 8.3 2.5 10.9 2.5 1

BMI = body mass index (kg/m2), CAD = coronary artery disease, number of diseased coronary arteries; D = diabetes; H = hypertension; LVEF = left ventricular ejection fraction; Rh. = rhythm: SR = sinus rhythm, 1 = paroxysmal, 2 = persistent, 3 = permanent AF; sex: f = female, m = male; VD = valve disease requiring valve replacement: AS = aortic stenosis, AR = aortic regurgitation, MR = mitral regurgitation, MS = mitral stenosis, TR = tricuspid regurgitation.

606

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

2.5. Oil Red O staining

2.8. Microarray expression analysis

Cells were washed twice with DPBS, fixed in 4% paraformaldehyde for 30 min at room temperature and washed again with 60% isopropanol. Dried cells were incubated with an Oil Red O solution (0.33% w/v in 60% isopropanol) for 15 min at room temperature. The cells were then washed immediately with MQ water and 60% isopropanol to remove unbound dye. Stained Oil Red O was eluted with 100% isopropanol for 10 min and quantified by measuring the optical density at 500 nm.

To identify global changes in gene expression, transcriptome analyses were performed as previously described [39].

2.6. RNA isolation and quality control Total RNA was extracted from frozen specimens of RA, LA, and EAT by performing a modified phenol extraction using a TRIzol reagent (Invitrogen, Karlsruhe, Germany) [43] as described previously [39]. Homogenization of snap-frozen tissue was carried out using a bead mill dismembrator (Braun) at 2600 rpm for 2 min. RNA was further purified using the RNA clean-up and concentration micro kit (Norgen, Canada) and concentrations were measured using a ND-1000 spectrophotometer (Thermo Fisher Scientific Inc., Wilmington, USA). RNA integrity was validated by means of the lab-on-chip capillary electrophoresis technology (Bioanalyzer 2100, Agilent Technologies, Santa Clara, CA, USA). Only RNA samples with a RNA integrity number (RIN) N 7.5 [28] and measured absorption quotients of A260 nm / A280 nm ≥ 1.8, A260 nm / A230 nm ≥ 1.9 were used for subsequent experiments. Total RNA from 3T3-L1 cells was extracted using the innuPrep RNA mini kit (Analytik Jena, Jena, Germany) according to the manufacturer's protocol and as previously described [44].

2.7. Reverse transcription and quantitative RT-PCR (RT-qPCR) Reactions were performed as described previously [39]. Primers were obtained from Life Technologies GmbH, Darmstadt, Germany; detailed primer information is given in Table 2.

2.9. In silico-pathway- and functional analysis of microarray data In silico-pathway- and functional analysis of differentially expressed genes was carried out using the commercial systems biology oriented package ingenuity pathways analysis (Ingenuity Systems, Inc. CA, USA) using the annotation details provided by Christopher K. Tuggle [45] with their corresponding gene identifiers and expression values. 2.10. Statistical analysis All data were analyzed statistically using GraphPad Prism 6® (Graph Pad Software Inc., La Jolla, CA, USA). Comparisons between groups were performed by Mann–Whitney post-hoc test considering P b 0.05 significant. Data are shown as the median with 25th and 75th percentiles.

3. Results 3.1. Impact of RAP on left atrial adipocyte/adipositas-related gene expression Previously, microarray-based profiling of RNA prepared from the LA tissue samples from pigs subjected to RAP identified 548 and 453 genes exhibiting increased or decreased mRNA amounts, respectively, under RAP conditions compared to sham-operated animals [39]. Differentially

Table 2 Primers used in the RT-qPCR analyses. Gene

Forward primer (5′→3′)

Reverse primer (5′→3′)

Size (bp)

h-ACACA all tv h-ADIPOQ tv1,2 p-ADIPOQ h-ANGPTL1 all tv h-ANGPTL2 p-ANGPTL2 h-ANGPTL4 h-DDIT3 tv1-4 h-DGAT2 tv1,2 h-FABP4 h-FASN h-HK2 p-HK2 h-IGF1 tv1,2,4 p-IGF1 h-IGF2 p-IGF2 h-IGFBP5 h-IGFBP6 h-LysOX tv1,2 h-NR4A3 tv4 p-NR4A3 h-PLIN2 tv2 h-PLIN5 h-PPARGC1a p-PPARGC1a h-PYGM all tv h-RETN tv1,2 p-RETN h-RPLP0 p-RPLP0 h-ACACA all tv h-ADIPOQ tv1,2 p-ADIPOQ h-ANGPTL1 all tv h-ANGPTL2 p-ANGPTL2 h-ANGPTL4 h-DDIT3 tv1–4 h-DGAT2 tv1,2 h-FABP4

GTT AAG GCG CTG GTT TGT GG TGC TCT GGC TGA GTT GTG TG CTG GCG AGA AGG GTG AGA AA GCA GCC TGA TCT TAC ACG GT CAC GAC ACC AGC TCC ATC TA GCA AGC AAT TCA CCA CCC TG TCT CTG GAG GCT GGT GGT TT AAA CAG GCA TCA GAC CAG CTT GTC ATG GGT GTC TGT GGG TT TGG CAT GGC CAA ACC TAA CA ATG CGG GAC AGA GCA ACT AC GCG TGG ACT ACT CTT CCG AG TGA GGT CTA CTC CGG ATG GG TTG ATG GGG TCT TTC AAG GG CTT CAG TTC GTG TGC GGA GA GTT TAC CTC GCC CCC ACT TG ACA CCC TCC AGT TTG TCT GCG TCC GTT TTC ACC CTT CTC CG GAA CCG CAG AGA CCA ACA GA GGA CTG TGC TGC GTA ATG AG GTG TCT GGC ACC ATG TTA GA TCT GAG ACG TGG TCC ATC CA TGA TTG CAG AGT GTG GTG ACT C GCCT TCT GTT TGG GGC TTT G CTT GCA CTA GCA TGG CCT CT GAT GCA CTG ACA GAT GGA GAT G TCA TCA CCC TGT ACA ACC GC TTA GCT GAG CCC ACC GAG AG CCT CAG GCT TTG CTG TCA CT CAA TGG CAG CAT CTA CAA CC CAG CAG ATC CGC ATG TCT CTC GTT AAG GCG CTG GTT TGT GG TGC TCT GGC TGA GTT GTG TG CTG GCG AGA AGG GTG AGA AA GCA GCC TGA TCT TAC ACG GT CAC GAC ACC AGC TCC ATC TA GCA AGC AAT TCA CCA CCC TG TCT CTG GAG GCT GGT GGT TT AAA CAG GCA TCA GAC CAG CTT GTC ATG GGT GTC TGT GGG TT TGG CAT GGC CAA ACC TAA CA

TGA CTT CTG CTC GCT GAG TG CGT GTT CTG TAA GGA GCC CA GAA CGG TAG ACA TAG GCG CT GGC CCT TTG AAG TAG TGC CA CCT TGC TTG TAC GTC TCC CA AGC CCA GTA CAC TCC ATC CT CAA GTT TCC AGA TGG CCC CT GTC CCG AAG GAG AAA GGC AA AGA AGT GGC TTT CGC CTC TC TGC GAA CTT CAG TCC AGG TC GGT CTG GTT CAG GAA GAG TC GTC GTC ACA GGT GCT CTC AA TGA AGT TAG CCA GGC ACT CG GAG TTG GTG ATG GGG GTC AA CTT CCT TCT GAG CCT TGG GC CAA TGT TCT GAA TGG CCC AC ACG CTT GGC CTC TCT GAC C GCA GAA CAG GTA AGA GGC GT GTT TGA GCC CCT CGG TAG AC ATC CGC CAG TTG ATG ACT GT AGC AGC AGG TCA ATC AAT GC CAC TGA ATG CTC TTG GGG CT TCG TCA CAG CAT CTT TTG CC GCA AAG CAA GCG AGT ATG GC GAT CGT GTT GGG CGA GAG AA GTG CAC TTG TCT CTG CTA CTG ACT TTC TCG GCC AGT GAG AC CTC CAG GCC AAT GCT GCT TA GTG ATG CGC AGA TGC AAA CT ACT CTT CCT TGG CTT CAA CC GGC AGG TAC AGT GAC TTC GCA TGA CTT CTG CTC GCT GAG TG CGT GTT CTG TAA GGA GCC CA GAA CGG TAG ACA TAG GCG CT GGC CCT TTG AAG TAG TGC CA CCT TGC TTG TAC GTC TCC CA AGC CCA GTA CAC TCC ATC CT CAA GTT TCC AGA TGG CCC CT GTC CCG AAG GAG AAA GGC AA AGA AGT GGC TTT CGC CTC TC TGC GAA CTT CAG TCC AGG TC

263 341 154 199 161 170 305 154 304 331 223 157 210 332 223 191 347 185 156 350 346 210 340 304 274 383 220 181 173 329 255 263 341 154 199 161 170 305 154 304 331

h = human; p = porcine; tv = transcript variant.

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

expressed genes could be assigned to the three functional categories Oxidative stress, Tissue remodeling, and Cellular energy depletion [39]. In this study, we specifically analyzed the impact of RAP on the left atrial expression of a newly defined adipocyte-differentiation & adipositas-related gene panel. This panel includes genes that have been assigned to adipocyte differentiation and obesity by ingenuity pathway analysis, adipogenesis and obesity RT2 profiler PCR assays (Qiagen), and review of literature with special emphasis given to results of differential gene expression profiling of in vitro adipogenic differentiation of mesenchymal stem cells [46]. The panel is given in full in Supplementary Table 1. In response to 7 h of RAP in vivo, there were significant changes (≥ 1.5-fold, P b 0.05) of adipocyte/adipositas-related gene expression: RAP led to an increased expression of 66 adipocyte/regulated (Table 3). 3.2. Validation of transcriptome data Microarray-based gene expression data were subsequently confirmed by RT-qPCR for selected targets. As summarized in Table 4, validation experiments fully confirmed the differential gene expressions of adipocyte/adipositas-related genes in response to RAP in vivo in all cases. 3.3. In vitro differentiation of 3T3-L1 pre-adipocytes Next, we confirmed the differential expression of selected members of the adipocyte/adipositas-related gene panel in the course of adipocyte differentiation of 3T3-L1 preadipocytes in vitro. The induction and continuous increase in lipid accumulation during 3T3-L1 cell differentiation is illustrated in Fig. 1A and B. At day 7 of 3T3-L1 cell differentiation, when the cells already exhibit substantial lipid accumulation, there was increased expressions of PLIN1, DGAT2, ACACA, ANGPTL4, and PPRGC1A (compared to day 5 of culture when the cells are still in an early state of lipid accumulation) (Fig. 1C). 3.4. AF-dependent changes in RA adipocyte/adipositas-related gene expression The rapid-pacing experiments revealed that the sole frequency increase provokes differential expression of an adipocyte/adipositasrelated gene panel. We next addressed the question whether similar changes in gene expression could be detected in the right atrial tissue samples (RA) from patients with atrial fibrillation in comparison to SR. In patients with AF (compared to SR), a comparable change in the RA gene expression could be observed only for a fraction of the panel genes. These genes included RETN, IGF1, HK2, PYGM, LOX, and NR4A3 (Fig. 2, Table 4). Although these changes indicated metabolic alterations in the RA tissue during AF, no comparable gene expression changes could be detected for “markers of adipocyte differentiation” such as FASN, ACACA, DDIT3, ANGPTL2, and PLIN2. Moreover, other “markers” (ANGPTL1 and 4, DGAT2, PLIN5, ADIPOQ) were even found to be significantly down-regulated in AF compared to SR (Fig. 2, Table 4). 3.5. AF-dependent changes in EAT adipocyte/adipositas-related gene expression We next assessed if and to what extent AF is capable of provoking changes in the expression of adipocyte/adipositas-related genes within EAT. RA and EAT appeared to be quite differently affected by AF, with the exception of RETN, the expression of which was induced equally strongly in the RA and EAT samples of patients with AF as well as in the LA samples of rapidly paced pigs. There was also a strong tendency toward reduced ADIPOQ expression in all these samples. Whereas NR4A3 and, in particular, HK2 were subject to regulation by AF in the RA myocardium only, AF-dependent expression changes

607

exclusively observed in the EAT included e.g. IGFBP4, 5 and 6, FABP4, and, as a tendency, the induction of ANGPTL4 (Table 4). 3.6. Impact of diabetes/obesity on RA gene expression in ZDF rats RAP-induced changes in adipocyte/adipositas-related gene expression observed in response to rapid pacing in vivo could be only partly confirmed in patients with AF. This led us to hypothesize that in these patients concomitant disease and existing risk factors for AF lead to a strong “pre-activation” of a corresponding gene expression panel. To verify this hypothesis we determined the mRNA levels in the RA tissue samples from 3 month old (early stage of disease) ZDF rats compared to 6 month old (advanced stage of disease) animals. In addition, we assessed the impact of RAP ex vivo on the expression of the same genes of the same animals. AS summarized in Table 5, with increasing age and development/progression of the diabetic phenotype there are substantial changes in gene expression with significant up-regulation of ABHD5, ADIPOR1, ANGPTL4, BMP4, and DGAT2. In addition there was a tendency toward increased mRNA levels of ADFP and IGFBP6. Six month old rats showed reduced expression of ACACA, ACACB, IGF2, and PPARGC1A. When the RA tissue samples of these animals were subjected to RAP, ex vivo gene expression changes were not aggravated to a large extent. In response to RAP, however, the slight and non-significant increase in PPARG mRNA levels from 3 to 6 month old rats (115%, n.s.) became more pronounced and significant (139.5%, P = 0.0286). 4. Discussion Obesity has been associated with a high risk for AF [1,3,47–49]. Perior epicardial adipose tissue (EAT) volume shows an even stronger correlation with AF risk and recurrence after e.g. radiofrequency ablation therapy than traditional BMI measurements [8,10,50–53]. Increased volumes of EAT or other cardiac fat tissue provoke endothelial dysfunction as well as metabolic and pro-inflammatory responses via endocrine and paracrine action of adipokines and may, thereby, contribute to AF development and maintenance [16,53,54]. On the contrary, less is known about the impact of tachyarrhythmia on adipocyte differentiation/proliferation. However, causality has not yet been fully established between AF and EAT volume and the mechanisms underlying the relationship between the amount of EAT and the risk and severity of AF are still largely unknown. A paracrine effect of EAT on the neighboring myocardium has been proposed. EAT produces a number of inflammatory mediators and adipocytokines that can modulate the functional and structural properties of the myocardium. In this line, it has been shown that, the secretome of EAT can induce the fibrosis of the atrial myocardium, an important determinant of the substrate of AF [55]. Another possibility is that adipose tissue infiltrates the atrial myocardium resulting in some degree of functional disorganization as described in the right ventricle of arrhytmogenic right ventricular cardiomyopathy patients [56]. The results of this study demonstrate that in response to acute rapid pacing in vivo, there are significant alterations in atrial gene expression which are consistent with the induction of an adipocyte/adipositasrelated expression profile. From the model used in this study it could be concluded that the sole increase in frequency (600 bpm) over a short period of time (7 h) — in the absence of any AF risk factors or concomitant disease — is capable of inducing this particular expression panel. Substantial changes in atrial gene expression profiles in response to 7 h only of rapid pacing have been described previously [39,57]. Although a simultaneous determination of electrophysiological parameters is not feasible, the effect of short term RAP on electrical remodeling has been determined numerous times in goats (e.g. [58]), dogs (e.g. [59]) and pigs (e.g. [60]). These studies very consistently showed a reduction in effective refractory period (ERP), lack of change in conduction velocity, and an increase in AF vulnerability/stability.

APOC3, CYP7A1

CAMKK2, SMARCA2, DDIT4L WNT5A, CLU, RPS6KA5 ATF4, HES1, BDH1, TSC22D3

Other PPAR targets

Anti-adipogenesis

SLC22A5

IGFBP3

FABP4, FABP6, CEBPD, SLC27A4, LMNA TWIST1, FOXC2, FOXO3, KLF4, CREB1, PPA1, IRAK1, MAPK8

PDK4, AGL, GK, AK2

ACACA, ACSL5, GSK3B, ENO1, PGM2, PPP2CA, DGAT2, FASN, PGLS, ELOVL1, ELOVL5, DEGS1 LysOX, KLF15, ACSL5, HSPA4, CCND3, YWHAH, CDK4, CEBPB, TP53, PLIN5, SGK1, FABP5L7, LEPROTL1, PTX3 DDIT3, CDKN1A Pro-inflammatory cytokines Enzymes (involved in HK2 regulation of FA metabolism) Pro-adipogenesis PLIN2, EIF4E, CRY1, MSX2, TRIB1, RHEB, ROCK2

The level of expression change for each gene is indicated in Supplementary Table 1. The following significance criteria were applied: expression fold-change ≥ 1.5 and P-values refer either to ANOVA or RB P-value . n.s. = non-significant.

MTTP, RB1, FGF7, FN1, NMU, NPY, NCOR1, WNT5B SREBF1, APOE BDNF, NRF1,

IL1A, IL1B ACSM3, CPD, IDH3G, DGAT1, ACSL6, PLA2G6, ACOT6, ACOX3 IGF1, IGF2 DECR1, DECR2, SYK, AK1,FBP1

THRB, RXRG

SST, GH1, CCL2 BMPR2, FGFR2, GRPR, NR4A3, HRH1 ITGB1, SSTR2, TNF IDH2, HMGCL, GK, UGP2, PYGL, PYGM, PGAM2, PLG, PGM1, ACADSB ENO3, PRKAB2 SERPINE1 ANGPTL4 PPARGC1A, TNFRSF1B, NMBR, PPARA, IGF1R, TNFRSF12A TNFRSF1A Adipokines/hormones Receptors

P ≤0.05

RETN, ADIPOQ ADIPOR2, LEPR, TNFRSF5

Decreased expression during RAP

P ≤0.001 P ≤0.001

P ≤0.1 (a considerable n.s. trend toward significance) Induced expression during RAP

Functional gene group

P ≤ 0.05

P ≤ 0.1 (a considerable n.s. trend toward significance) CCK AGTR1 RAMP3

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

Table 3 Lists of probe sets related to adipocyte-differentiation & adipositas related gene panel, indicating functional gene grouping of differentially regulated genes under RAP condition as compared to control (data sheet “Functional Gene Groups”).

608

The expression changes observed here in part reflect metabolic changes and are consistent with AF-dependent flow-limitations and insufficient supply of oxygen and/or nutrients described previously [57,61–64]. In a recent study we demonstrated for the left atria that during AF conduction and oxygen extraction reserve are recruited to compensate for acute supply–demand ischemia [65]. However, the observed increase in lactate production indicated that the oxygen demand during AF still exceeds the actual supply [65]. In this regard, increased HK2 expression observed here could enhance glycolytic flux and, thus, contribute to maintain minimum ATP production under ischemic conditions. The observed decrease of PYGM mRNA levels is in accordance with the glycogen particle accumulation observed in patients with AF by electron microscopy [66]. Fatty acid or triacylglyceride biosynthesis are neither required nor possible under these conditions which is in accordance with decreased atrial mRNA levels of e.g. ACACA, FASN, and DGAT in response to acute RAP. In addition to these changes in “metabolic” gene expression there were also expression changes that point to an induction of genes that are markers of adipocyte-differentiation and adipogenesis such as PLIN5 or ANGPTL4. Therefore, it appears reasonable to conclude that tachyarrhythmia in an otherwise healthy heart (a situation representing lone AF) is associated with transcriptional alterations that favor adipocyte differentiation/infiltration, increase of adipose tissue mass (including EAT), and lipid accumulation. This is in full accordance with the established correlation of AF with EAT volume. RAP-dependent expression changes also regard genes encoding humoral factors that include the adipokines resistin and adiponectin. Both have been associated with important risk factors for AF (for review see [67]). Increased plasma concentrations of resistin have been associated previously with incident AF [67] and are predictive for post-operative AF after coronary artery bypass grafting as well as for paroxysmal and persistent AF [18,19]. High concentrations of adiponectin have been related to persistent AF [68]. The analysis of adipocyte/adipositas-related gene expression within the RA tissue samples from patients with AF revealed the interesting finding that not all the expression changes that had been observed in the RAP in vivo model before could be detected here. Compared to SR, the RA samples from AF patients show expression changes that are in full accordance with metabolic adaption to ischemia/hypoxia [61] and, thus, regard lipid and carbohydrate metabolism. However, there is an apparent lack of differential expression of genes that are indicative of ongoing adipocyte differentiation, increased adipose tissue mass, and lipid synthesis and accumulation. This can be concluded from unchanged or even reduced RA expression levels of IGF1,-2, ANGPTL1, -2, and -4, FABP4, ACACA, FASN, PLIN2, PLIN5, and NR4A3 in AF compared to SR. Perilipins are a multi-protein family with five currently known members (PLIN1–5). They coat lipid droplets and regulate lipid storage and hydrolysis by protecting against neutral lipase action. Perilipin variants have an individual tissue-dependent expression pattern. Plin1 is expressed in white and brown adipose tissue and steroidogenic tissue. Plin2 (previously adipophilin/ADRP) and Plin3 (previously TIP47) are ubiquitously expressed. Plin4 (previously S3–12) is highly expressed in adipocytes, whereas Plin5 (previously OXPAT) is expressed in oxidative tissue including heart, liver, skeletal muscle and brown adipose tissue [69]. DDIT3, also known as CHOP (C/EBP-homologous protein), plays a role in cell survival and differentiation. It binds C/EBPs to form heterodimers, acts as a dominant negative inhibitor and thereby blocks adipogenesis [70]. NR4A3 (also NOR; neuron-derived orphan receptor 1), like other members of the NR4A subfamily of receptors, is induced very early (1–2 h) during adipocyte differentiation [71]. The expression decreased back to very low, nearly undetectable levels within a day after adipogenic induction [72]. However, functional studies in mice suggest that NR4As inhibit adipocyte differentiation [71]. Others observed a

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

609

Table 4 Effects of RAP in vivo (pig) or AF on adipocyte–adipositas-related gene expression. Rapid pacing in vivo pig Microarray (fold change)

Adipositas/adipocyte– Fold change related gene HK2 ANGPTL4 PLIN2 (ADFP)(tv2) PLIN5

Patients with AF versus SR

RT–qPCR validation of microarray

P–value

9.71 3.84

1.62E–7 0.0028

3.42 n.a.

0.0016 n.a.

PPARG1CA

2.96

0.0187

RETN

2.62

0.11*

DDIT3 (tv1–4) DGAT2 FABP4 IGFBP6 ACACA

2.34 2.20 2.09 1.93 1.89

0.0115 0.0638 0.37 0.0311* 0.0306

ADIPOQ

1.46

0.515

↑ n.d. n.d. n.d.

Right atrial tissue

Epicardial adipose tissue

SR (n=12) Median (Q1; Q3)

AF (n=12) Median (Q1; Q3)

P–value (Mann–Whitney)

SR (n=11) Median (Q1; Q3)

100.0 (35.7; 178.7) 100.0 (73.8; 332.2) 100.0 (89.6; 125.3) 100.0 (77.5; 107.8)

329.0 (66.3; 392.1) 66.1 (52.33; 93.8) 98.6 (80.1; 119.3) 47.4 (36.9; 73.6)

0.1031 0.0248 0.5952 0.0127

100.0 (15.0; 181.9) 100.0 (60.2; 297.6) 100.0 (68.2; 146.7) n.d.

↑

100.0 (63,3; 122.7)

96.4 (74.0; 155.1)

↑ n.d. n.d. n.d. n.d. n.d.

100.0 (71.6; 146.4) 100.0 (45.1; 143.9) 100.0 (56.5; 120.5) 100.0 (48.2; 227.1) 100.0 (84.4; 134.5) 100.0 (82.4; 133.6)

229.3 (110.6; 353.8) 26.9 (17.3; 34.5) 9.2 (6.4; 42.4) 102.4 (45.2; 119.5) 89.43 (71.1; 102.7) 112.5 (77.2; 184.5)

↑

100.0 (16.0; 145.3)

7.0 (1.7; 35.9)

AF (n=11) Median (Q1; Q3) 74.0 (31.2; 125.0) 266.3 (145.9; 286.5) 91.4 (70.1; 204.0) n.d.

P–value (Mann–Whitney) 0.866 0.149 0.962

0.4799

100.0 (53.3; 116.9)

61.4 (34.5; 87.9)

0.036

0.0157 0.0017 < 0.0001 0.6407 0.5104 0.5577

100.0 (45.3; 126.6) 100.0 (51.2; 108.7) 100.0 (62.3; 148.9) n.d. n.d. 100.0 (57.1; 149.0)

205.0 (73.6; 326.8) 98.6 (78.6; 155.4) 87.7 (31.9; 112.8) n.d. n.d. 88.9 (66.8; 127.6)

0.027 0.404 0.552

0.743

0.0522

100.0 (70.4; 194.0)

57.6 (50.4; 95.1)

0.101

LOX

1.30

0.691

n.d.

100.0 (82.9; 139.8)

188.6 (110.7; 271.6)

0.0437

100.0 (66.3;124.9)

144.6 (59.6;222.8)

0.360

ANGPTL1

1.10

0.902

n.d.

100.0 (51.2; 160.2)

33.00 (13.1; 63.1)

0.0028

100.0 (82.4; 160.3)

135.5 (115.2; 270.5)

0.113

–22.87

0.016

↓

100.0 (3.1; 114.9)

2.6 (1.2; 8.7)

NR4A3 (tv4) ANGPTL2 PYGM IGFBP5

↓ n.d. n.d.

–3.34 –2.96 –2.53

0.033 6.57E–11 2.46E–7

IGF1 (tv1, 2, 4)

–2.01

0.0223

↓

IGF2 FASN

–1.83 –1.28

0.088 0.945

↓ n.d.

0.0056

100.0 (17.6; 181.1)

165.1 (53.8; 256.4)

0.470

125.1 (104.0; 171.4) 42.0 (33.1; 95.2) 71.2 (55.5; 146.7)

0.3429 0.0086 0.1318

100.0 (54.7; 167.2) n.d. n.d.

124.8 (77.0; 141.5) n.d. n.d.

0.343

100.0 (88.4; 147.0)

58.2 (52.3; 109.9)

0.0471

100.0 (60.1; 157.7)

102.5 (71.5; 145.3)

0.829

100.0 (73.00; 142.0) 100.0 (78.3; 161.5)

144.2 (79.8; 191.7) 85.4 (66.8; 91.3)

0.4316 0.2087

100.0 (51.7; 239.4) 100.0 (79.2; 179.8)

139.7 (68.5; 229.8) 97.1 (57.5; 120.2)

0.786 0.506

100.0 (79.9; 136.4) 100.0 (96.00; 118.9) 100.0 (74.6; 185.6)

tv = transcipt variant; n.d. = not determined; ↑ = significantly up–regulated; ↓ = significantly down–regulated

Fig. 1. Differentiation of 3T3-L1 pre-adipocytes in vitro. 3T3-L1 cells were subjected to adipogenic differentiation as described in Materials and methods section. A: Detection of lipid accumulation in 3T3-L1 cells visualized by Oil red staining of cell cultures at different time points after initiation of differentiation (bar = 200 μm). Lipid accumulation starts at day 5 and continuously increases until day 14. B: Quantitative determination of Oil red staining shown in A. C: Increased expression of adipocyte/adipositas-related genes at day 7 of 3T3-L1 differentiation (compared to day 5). Data are shown as median with 25th and 75th percentiles, n = 4, *P b 0.05.

610

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

Fig. 2. Expression of adipocyte/adipositas-related genes in RA and EAT in patients with AF. AF causes similarly increased expression of RETN and LOX. However, there are tissue-specific differences in the induction of metabolic genes (HK2) and in the genes encoding the differentiation markers ANGPTL4 and NR3A4 which are marginally induced in EAT but drastically reduced in RA. Data are presented as median with 25th and 75th percentiles, n = 13 *P b 0.05 **P b 0.01.

decrease in NR4A expression during the differentiation of human preadipocytes [73]. Angiopoietin-like proteins (ANGPTLs) are a group of eight proteins with similarity to the angiopoietin protein family members. ANGPTL4 regulates the plasma levels of triglyceride, prevents the uptake of dietary lipids into the adipose tissues, and inhibits intravascular lipolysis [74]. Lysyl-oxidase (LOX), a copper-dependent amine oxidase, catalyzes the oxidation of ε-amino groups of lysine or hydroxylysine and is involved in collagen cross-linking. Left atrial expression of LOX is increased during AF and appears to mediate AFassociated fibrosis in an AngII-dependent manner [75]. During AF, LOX may become an important source of H2O2. LOX plays a crucial role in the commitment step of adipocyte formation from pluripotent stem cells during development [76]. These targets appeared to be clearly differentially expressed in the course of adipocyte differentiation of 3T3-L1 pre-adipocytes. It is tempting to speculate, therefore, that in our patients, concomitant disease and the presence of various risk factors for AF had already caused substantial alterations of RA gene expression and, in particular, of adipocyte/

adipositas-related genes. We did not determine the transcriptome of the RA samples from patients with lone AF and have no access to the tissue samples from healthy subjects, which is a clear limitation of our study. However, mounting evidence indicates that e.g. hypertension, diabetes, and obesity do activate adipogenic and inflammatory gene expression and, thereby, exert effects on the heart [16,68,77–81]. EATderived adipokines easily access and act on the adjacent myocardium as no distinct barriers separate both tissues [55]. The close proximity of well-organized adipocytes and myocytes with surrounding fibrosis [55] facilitates direct arrhythmogenic activity of adipocytes in addition to the formation of AF substrate [20,82]. For none of the genes analyzed here a correlation of the expression values with AF duration could be detected. This finding is in full accord with and further supports one major conclusion of our study: in patients with AF, existing concomitant disease substantially contributes to the induction of AARE. A minor role only of AF duration for AARE is also suggested by the fact that even short episodes of RAP/AF strongly induce AARE as has been observed in our RAP in vivo experiments. The number

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

611

Table 5 Impact of the development/progression of diabetes and RAP ex vivo on atrial adipocyte–adipositas-related gene expression in ZDF rats.

ZDF rats

Adipositas/adipocyte– related gene

RA rapid pacing 3 month vs. 6 month

6 month (n=4) Median (Q1; Q3)

P–value (Mann–Whitney)

ADIPOR2 ANGPTL4 BMP4 DDIT3 DGAT2 IGF1 IGF2

100.0 (94.0; 106.8) 100.0 (100.0; 109.0) 100.0 (87.2; 107.5) 100.0 (87.2; 116.5) 100.0 (96.2; 108.5) 100.0 (94.8; 106.0) 100.0 (100.0; 100.8) 100.0 (77.5; 105.3) 100.0 (89.8; 109.0) 100.0 (88.5; 112.3) 100.0 (68.0; 108.8) 100.0 (94.0; 111.3) 100.0 (82.0; 115.8)

120.5 (111.0; 129.3) 88.5 (80.5; 92.8) 47.0 (35.0; 71.0) 121.0 (110.8; 133.5) 100.5 (96.2; 108.5) 113.5 (106.3; 125.3) 93.5 (86.8; 96.5) 138.5 (133.8; 168.0) 130.5 (121.5; 138.8) 108.0 (101.8; 122.5) 141.5 (122.3; 162.3) 116.0 (106.8; 153.8) 66.5 (55.8; 80.2)

0.0571 0.0286 0.0286 0.1143 0.9999 0.0857 0.0286 0.0286 0.0286 0.4857 0.0286 0.1143 0.0571

IGFBP5 IGFBP6 LOX PPARG PPARGC1A

100.0 (76.0; 123.8) 100.0 (91.5; 152.0) 100.0 (94.5; 161.3) 100.0 81.0; 119.0) 100.0 (78.8; 134.8)

68.5 (65.0; 87.8) 182.0 (146.8; 213.5) 108.5 (85.8; 123.8) 115.0 (98.2; 146.0) 59.0 (54.8; 80.5)

0.1143 0.1143 0.8286 0.3429 0.0571

ABHD5 ACACA ACACB PLIN2 (ADFP) ADIPOQ ADIPOR1

3 month (n=4) Median (Q1; Q3)

RA control 3 month vs. 6 month

of patients included in our study is relatively small. Due to this limitation the results of the regression analyses (AF duration versus AARE) should not be over-estimated. Further supporting the view that in our patients (chronic or persistent AF) concomitant disease strongly contributes to a differential expression of adipocyte/adipositas-related genes, we observed substantial changes of the RA gene expression with growing age and progressing metabolic alterations in ZDF rats. The data obtained indicate the profound impact of a diabetic phenotype on the RA gene expression and further demonstrate that on this background tachyarrhythmia is

3 month (n=4) Median (Q1; Q3)

6 month (n=4) Median (Q1; Q3)

P–value (Mann–Whitney)

100.0 (84.25; 103.0) 100.0 (83.5; 120.5) 100.0 (83.2; 120.8) 100.0 (74.0; 104.3) 100.0 (96.0; 109.3) 100.0 (95.2; 113.0) 100.0 (93.2; 106.8) 100.0 (84.8; 105.5) 100.0 (79.5; 116.0) 100.0 (94.8; 103.8) 100.0 (72.5; 114.0) 100.0 (83.2; 106.3) 100.0 (94.2;133.5)

114.5 (111.3; 120.8) 85.5 (77.5; 90.5) 54.5 (46.8; 79.5) 105.0 (100.5; 114.8) 179.0 (112.0; 365.3) 112.5 (108.3; 119.0) 89.5 (82.0; 91.8) 154.0 (139.3; 159.8) 136.0 (130.5; 143.0) 100.0 (96.2; 106.0) 146.0 (124.3; 155.8) 117.0 (106.0; 141.5) 85.8 (82.0; 97.2)

0.0286 0.200 0.0571 0.2286 0.1429 0.200 0.0286 0.0286 0.0286 0.8571 0.0286 0.0857 0.1143

100.0 (73.0; 139.0) 100.0 (82.2; 175.5) 100.0 (85.8; 133.0) 100.0 (74.0; 113.3) 100.0 (78.8; 166.0)

92.0 (84.5; 103.3) 134.5 (116.5; 201.3) 114.0 (75.0; 141.8) 139.5 (132.0; 141.0) 69.0 (56.8; 101.5)

0.6571 0.200 0.8286 0.0286 0.3429

capable of provoking additional changes in adipocyte/adipositasrelated gene expression to some extent. The comparison of AF-dependent expression changes in the RA tissue samples with those in EAT from the same patients revealed both similarly and differentially regulated genes. Notably, metabolic adaption in the cardiomyocytes to AF-dependent flow limitations includes the alteration of glucose and lipid metabolism. Therefore, in particular this subset of genes is equally strongly affected in RA and EAT. In addition, we observed considerably increased expression of RETN and decreased expression of ADIPOQ in both tissues. These

Fig. 3. Induction of adipocyte/adipositas-related gene expression by AF. Rapid pacing or lone AF provokes microvascular flow limitations, increased systemic and tissue levels of AngII, and elevated production of reactive oxygen species (ROS). Resulting adaptive mechanisms include metabolic alterations and the promotion by humoral factors of adipocyte differentiation, adipose tissue expansion, and lipid accumulation. On the contrary, concomitant disease and pre-existing risk factors in patients with chronic AF function as strong inducers of a quite similar atrial (cardiac) expression profile. Adipose tissue-derived factors (adipokines) further contribute to the formation of AF substrate and contribute to AF maintenance and progression.

612

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613

changes may account for adverse (e.g. pro-inflammatory) systemic effects during AF. In summary, the data presented suggest that lone-AF/RAP promotes atrial adipogenesis which in part is regulated by genes specific to metabolic adaptation. These changes favor the development of important risk factors for AF like obesity and T2DM and may contribute to the formation of an AF substrate. In patients with chronic AF, the high prevalence of substantial concomitant disease (hypertension, diabetes, obesity, heart failure) provokes already an induction of metabolic and adipocyte/adipositas-related gene expression to a significant extent and which could not or only marginally further enhanced by tachyarrhythmia (Fig. 3). Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2015.03.072.

Conflict of interest The authors report no relationships that could be construed as a conflict of interest.

Acknowledgments The authors thank Manja Möller and Ines Schultz for excellent technical support. The authors' work is supported by grants from the European Union (FP7 Collaborative project European Network for Translational Research in Atrial Fibrillation, EUTRAF, 261057). We are grateful to Sanofi–Aventis for supporting the in vivo rapid pacing experiments (pigs).

References [1] T.S. Tsang, M.E. Barnes, Y. Miyasaka, S.S. Cha, K.R. Bailey, G.C. Verzosa, et al., Obesity as a risk factor for the progression of paroxysmal to permanent atrial fibrillation: a longitudinal cohort study of 21 years, Eur. Heart J. 29 (2008) 2227–2233. [2] N. Wanahita, F.H. Messerli, S. Bangalore, A.S. Gami, V.K. Somers, J.S. Steinberg, Atrial fibrillation and obesity—results of a meta-analysis, Am. Heart J. 155 (2008) 310–315. [3] T.J. Wang, H. Parise, D. Levy, R.B. D'Agostino Sr., P.A. Wolf, R.S. Vasan, et al., Obesity and the risk of new-onset atrial fibrillation, JAMA 292 (2004) 2471–2477. [4] H. Watanabe, N. Tanabe, T. Watanabe, D. Darbar, D.M. Roden, S. Sasaki, et al., Metabolic syndrome and risk of development of atrial fibrillation: the Niigata preventive medicine study, Circulation 117 (2008) 1255–1260. [5] A. Zacharias, T.A. Schwann, C.J. Riordan, S.J. Durham, A.S. Shah, R.H. Habib, Obesity and risk of new-onset atrial fibrillation after cardiac surgery, Circulation 112 (2005) 3247–3255. [6] A.A. Mahabadi, J.M. Massaro, G.A. Rosito, D. Levy, J.M. Murabito, P.A. Wolf, et al., Association of pericardial fat, intrathoracic fat, and visceral abdominal fat with cardiovascular disease burden: the Framingham Heart Study, Eur. Heart J. 30 (2009) 850–856. [7] G.A. Rosito, J.M. Massaro, U. Hoffmann, F.L. Ruberg, A.A. Mahabadi, R.S. Vasan, et al., Pericardial fat, visceral abdominal fat, cardiovascular disease risk factors, and vascular calcification in a community-based sample: the Framingham Heart Study, Circulation 117 (2008) 605–613. [8] M.O. Al Chekakie, C.C. Welles, R. Metoyer, A. Ibrahim, A.R. Shapira, J. Cytron, et al., Pericardial fat is independently associated with human atrial fibrillation, J. Am. Coll. Cardiol. 56 (2010) 784–788. [9] H.M. Tsao, W.C. Hu, M.H. Wu, C.T. Tai, Y.J. Lin, S.L. Chang, et al., Quantitative analysis of quantity and distribution of epicardial adipose tissue surrounding the left atrium in patients with atrial fibrillation and effect of recurrence after ablation, Am. J. Cardiol. 107 (2011) 1498–1503. [10] O. Batal, P. Schoenhagen, M. Shao, A.E. Ayyad, D.R. Van Wagoner, S.S. Halliburton, et al., Left atrial epicardial adiposity and atrial fibrillation, Circ. Arrhythm. Electrophysiol. 3 (2010) 230–236. [11] K. Nagashima, Y. Okumura, I. Watanabe, T. Nakai, K. Ohkubo, M. Kofune, et al., Does location of epicardial adipose tissue correspond to endocardial high dominant frequency or complex fractionated atrial electrogram sites during atrial fibrillation? Circ. Arrhythm. Electrophysiol. 5 (2012) 676–683. [12] S.Y. Shin, H.S. Yong, H.E. Lim, J.O. Na, C.U. Choi, J.I. Choi, et al., Total and interatrial epicardial adipose tissues are independently associated with left atrial remodeling in patients with atrial fibrillation, J. Cardiovasc. Electrophysiol. 22 (2011) 647–655. [13] Y. Nakatani, K. Kumagai, K. Minami, M. Nakano, H. Inoue, S. Oshima, Location of epicardial adipose tissue affects the efficacy of a combined dominant frequency and complex fractionated atrial electrogram ablation of atrial fibrillation, Heart Rhythm. 12 (2015) 257–265.

[14] R.J. Aviles, D.O. Martin, C. Apperson-Hansen, P.L. Houghtaling, P. Rautaharju, R.A. Kronmal, et al., Inflammation as a risk factor for atrial fibrillation, Circulation 108 (2003) 3006–3010. [15] P. Bruins, H. te Velthuis, A.P. Yazdanbakhsh, P.G. Jansen, F.W. van Hardevelt, E.M. de Beaumont, et al., Activation of the complement system during and after cardiopulmonary bypass surgery: postsurgery activation involves C-reactive protein and is associated with postoperative arrhythmia, Circulation 96 (1997) 3542–3548. [16] T. Mazurek, L. Zhang, A. Zalewski, J.D. Mannion, J.T. Diehl, H. Arafat, et al., Human epicardial adipose tissue is a source of inflammatory mediators, Circulation 108 (2003) 2460–2466. [17] T. Watanabe, Y. Takeishi, O. Hirono, M. Itoh, M. Matsui, K. Nakamura, et al., Creactive protein elevation predicts the occurrence of atrial structural remodeling in patients with paroxysmal atrial fibrillation, Heart Vessels 20 (2005) 45–49. [18] K.S. Ozcan, B. Gungor, S. Altay, D. Osmonov, A. Ekmekci, F. Ozpamuk, et al., Increased level of resistin predicts development of atrial fibrillation, J. Cardiol. 63 (2014) 308–312. [19] H. Gungor, M.F. Ayik, B. Kirilmaz, S. Ertugay, I. Gul, B.S. Yildiz, et al., Serum resistin level: as a predictor of atrial fibrillation after coronary artery bypass graft surgery, Coron. Artery Dis. 22 (2011) 484–490. [20] Y.K. Lin, Y.C. Chen, J.H. Chen, S.A. Chen, Y.J. Chen, Adipocytes modulate the electrophysiology of atrial myocytes: implications in obesity-induced atrial fibrillation, Basic Res. Cardiol. 107 (2012) 293. [21] L. Pantanowitz, Fat infiltration in the heart, Heart 85 (2001) 253. [22] A. Goette, M. Arndt, C. Rocken, A. Spiess, T. Staack, J.C. Geller, et al., Regulation of angiotensin II receptor subtypes during atrial fibrillation in humans, Circulation 101 (2000) 2678–2681. [23] A. Goette, A. Bukowska, U. Lendeckel, M. Erxleben, M. Hammwohner, D. Strugala, et al., Angiotensin II receptor blockade reduces tachycardia-induced atrial adhesion molecule expression, Circulation 117 (2008) 732–742. [24] A. Goette, U. Lendeckel, H.U. Klein, Signal transduction systems and atrial fibrillation, Cardiovasc. Res. 54 (2002) 247–258. [25] A. Goette, T. Staack, C. Rocken, M. Arndt, J.C. Geller, C. Huth, et al., Increased expression of extracellular signal-regulated kinase and angiotensin-converting enzyme in human atria during atrial fibrillation, J. Am. Coll. Cardiol. 35 (2000) 1669–1677. [26] M. Barton, R. Carmona, J. Ortmann, J.E. Krieger, T. Traupe, Obesity-associated activation of angiotensin and endothelin in the cardiovascular system, Int. J. Biochem. Cell Biol. 35 (2003) 826–837. [27] M.E. Frigolet, N. Torres, A.R. Tovar, The renin–angiotensin system in adipose tissue and its metabolic consequences during obesity, J. Nutr. Biochem. 24 (2013) 2003–2015. [28] S. Umemura, N. Nyui, K. Tamura, K. Hibi, S. Yamaguchi, M. Nakamaru, et al., Plasma angiotensinogen concentrations in obese patients, Am. J. Hypertens. 10 (1997) 629–633. [29] P. Fuentes, M.J. Acuna, M. Cifuentes, C.V. Rojas, The anti-adipogenic effect of angiotensin II on human preadipose cells involves ERK1,2 activation and PPARG phosphorylation, J. Endocrinol. 206 (2010) 75–83. [30] F. Jing, M. Mogi, M. Horiuchi, Role of renin–angiotensin–aldosterone system in adipose tissue dysfunction, Mol. Cell. Endocrinol. 378 (2013) 23–28. [31] J. Janke, M. Schupp, S. Engeli, K. Gorzelniak, M. Boschmann, L. Sauma, et al., Angiotensin type 1 receptor antagonists induce human in-vitro adipogenesis through peroxisome proliferator-activated receptor-gamma activation, J. Hypertens. 24 (2006) 1809–1816. [32] G. Ailhaud, Adipose tissue as a secretory organ: from adipogenesis to the metabolic syndrome, C. R. Biol. 329 (2006) 570–577 (discussion 653–5). [33] M. Furuhashi, N. Ura, H. Takizawa, D. Yoshida, N. Moniwa, H. Murakami, et al., Blockade of the renin–angiotensin system decreases adipocyte size with improvement in insulin sensitivity, J. Hypertens. 22 (2004) 1977–1982. [34] F. Tian, R. Luo, Z. Zhao, Y. Wu, D. Ban, Blockade of the RAS increases plasma adiponectin in subjects with metabolic syndrome and enhances differentiation and adiponectin expression of human preadipocytes, Exp. Clin. Endocrinol. Diabetes 118 (2010) 258–265. [35] M. Furuhashi, N. Ura, K. Higashiura, H. Murakami, M. Tanaka, N. Moniwa, et al., Blockade of the renin–angiotensin system increases adiponectin concentrations in patients with essential hypertension, Hypertension 42 (2003) 76–81. [36] M.L. Mathai, S. Naik, A.J. Sinclair, H.S. Weisinger, R.S. Weisinger, Selective reduction in body fat mass and plasma leptin induced by angiotensin-converting enzyme inhibition in rats, Int. J. Obes. (Lond.) 32 (2008) 1576–1584. [37] E.L. Santos, Souza K. de Picoli, E.D. da Silva, E.C. Batista, P.J. Martins, V. D'Almeida, et al., Long term treatment with ACE inhibitor enalapril decreases body weight gain and increases life span in rats, Biochem. Pharmacol. 78 (2009) 951–958. [38] R.S. Weisinger, T.K. Stanley, D.P. Begg, H.S. Weisinger, K.J. Spark, M. Jois, Angiotensin converting enzyme inhibition lowers body weight and improves glucose tolerance in C57BL/6 J mice maintained on a high fat diet, Physiol. Behav. 98 (2009) 192–197. [39] R.K. Chilukoti, J. Mostertz, A. Bukowska, C. Aderkast, S.B. Felix, M. Busch, et al., Effects of irbesartan on gene expression revealed by transcriptome analysis of left atrial tissue in a porcine model of acute rapid pacing in vivo, Int. J. Cardiol. 168 (2013) 2100–2108. [40] A. Bukowska, U. Lendeckel, D. Hirte, C. Wolke, F. Striggow, P. Rohnert, et al., Activation of the calcineurin signaling pathway induces atrial hypertrophy during atrial fibrillation, Cell. Mol. Life Sci. 63 (2006) 333–342. [41] A. Bukowska, L. Schild, G. Keilhoff, D. Hirte, M. Neumann, A. Gardemann, et al., Mitochondrial dysfunction and redox signaling in atrial tachyarrhythmia, Exp. Biol. Med. (Maywood) 233 (2008) 558–574. [42] L. Schild, A. Bukowska, A. Gardemann, P. Polczyk, G. Keilhoff, M. Tager, et al., Rapid pacing of embryoid bodies impairs mitochondrial ATP synthesis by a calciumdependent mechanism—a model of in vitro differentiated cardiomyocytes to study molecular effects of tachycardia, Biochim. Biophys. Acta 1762 (2006) 608–615.

R.K. Chilukoti et al. / International Journal of Cardiology 187 (2015) 604–613 [43] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction, Anal. Biochem. 162 (1987) 156–159. [44] S.A. Ender, A. Dallmer, F. Lassig, U. Lendeckel, C. Wolke, Expression and function of the ACE2/angiotensin(1–7)/Mas axis in osteosarcoma cell lines U-2 OS and MNNG-HOS, Mol. Med. Rep. 10 (2014) 804–810. [45] O. Couture, K. Callenberg, N. Koul, S. Pandit, R. Younes, Z.L. Hu, et al., ANEXdb: an integrated animal ANnotation and microarray EXpression database, Mamm. Genome 20 (2009) 768–777. [46] A. Menssen, T. Haupl, M. Sittinger, B. Delorme, P. Charbord, J. Ringe, Differential gene expression profiling of human bone marrow-derived mesenchymal stem cells during adipogenic development, BMC Genomics 12 (2011) 461. [47] S. Dublin, B. French, N.L. Glazer, K.L. Wiggins, T. Lumley, B.M. Psaty, et al., Risk of new-onset atrial fibrillation in relation to body mass index, Arch. Intern. Med. 166 (2006) 2322–2328. [48] L. Frost, L.J. Hune, P. Vestergaard, Overweight and obesity as risk factors for atrial fibrillation or flutter: the Danish Diet, Cancer, and Health Study, Am. J. Med. 118 (2005) 489–495. [49] U.B. Tedrow, D. Conen, P.M. Ridker, N.R. Cook, B.A. Koplan, J.E. Manson, et al., The long- and short-term impact of elevated body mass index on the risk of new atrial fibrillation the WHS (Women's Health Study), J. Am. Coll. Cardiol. 55 (2010) 2319–2327. [50] G. Thanassoulis, J.M. Massaro, C.J. O'Donnell, U. Hoffmann, D. Levy, P.T. Ellinor, et al., Pericardial fat is associated with prevalent atrial fibrillation: the Framingham Heart Study, Circ. Arrhythm. Electrophysiol. 3 (2010) 345–350. [51] C.X. Wong, H.S. Abed, P. Molaee, A.J. Nelson, A.G. Brooks, G. Sharma, et al., Pericardial fat is associated with atrial fibrillation severity and ablation outcome, J. Am. Coll. Cardiol. 57 (2011) 1745–1751. [52] S.N. Hatem, Is epicardial adipose tissue an epiphenomenon or a new player in the pathophysiology of atrial fibrillation? Arch. Cardiovasc. Dis. 107 (2014) 349–352. [53] S.N. Hatem, P. Sanders, Epicardial adipose tissue and atrial fibrillation, Cardiovasc. Res. 102 (2014) 205–213. [54] N. Girerd, A. Scridon, F. Bessiere, S. Chauveau, A. Geloen, L. Boussel, et al., Periatrial epicardial fat is associated with markers of endothelial dysfunction in patients with atrial fibrillation, PLoS One 8 (2013) e77167. [55] N. Venteclef, V. Guglielmi, E. Balse, B. Gaborit, A. Cotillard, F. Atassi, et al., Human epicardial adipose tissue induces fibrosis of the atrial myocardium through the secretion of adipo-fibrokines, Eur. Heart J. (2013) http://dx.doi.org/10.1093/eurheartj/ eht099. [56] A.P. Burke, A. Farb, G. Tashko, R. Virmani, Arrhythmogenic right ventricular cardiomyopathy and fatty replacement of the right ventricular myocardium: are they different diseases? Circulation 97 (1998) 1571–1580. [57] A. Goette, A. Bukowska, D. Dobrev, J. Pfeiffenberger, H. Morawietz, D. Strugala, et al., Acute atrial tachyarrhythmia induces angiotensin II type 1 receptor-mediated oxidative stress and microvascular flow abnormalities in the ventricles, Eur. Heart J. 30 (2009) 1411–1420. [58] Y. Blaauw, N. Beier, P. van der Voort, A. van Hunnik, U. Schotten, M.A. Allessie, Inhibitors of the Na+/H+ exchanger cannot prevent atrial electrical remodeling in the goat, J. Cardiovasc. Electrophysiol. 15 (2004) 440–446. [59] J.V. Jayachandran, D.P. Zipes, J. Weksler, J.E. Olgin, Role of the Na(+)/H(+) exchanger in short-term atrial electrophysiological remodeling, Circulation 101 (2000) 1861–1866. [60] D. Linz, F. Mahfoud, U. Schotten, C. Ukena, M. Hohl, H.R. Neuberger, et al., Renal sympathetic denervation provides ventricular rate control but does not prevent atrial electrical remodeling during atrial fibrillation, Hypertension 61 (2013) 225–231. [61] A. Bukowska, M. Hammwohner, A. Sixdorf, L. Schild, I. Wiswedel, F.W. Rohl, et al., Dronedarone prevents microcirculatory abnormalities in the left ventricle during atrial tachypacing in pigs, Br. J. Pharmacol. 166 (2012) 964–980. [62] G. Heusch, Heart rate in the pathophysiology of coronary blood flow and myocardial ischaemia: benefit from selective bradycardic agents, Br. J. Pharmacol. 153 (2008) 1589–1601.

613

[63] G. Heusch, D. Baumgart, P. Camici, W. Chilian, L. Gregorini, O. Hess, et al., Alphaadrenergic coronary vasoconstriction and myocardial ischemia in humans, Circulation 101 (2000) 689–694. [64] G.E. Kochiadakis, E.I. Skalidis, M.D. Kalebubas, N.E. Igoumenidis, S.I. Chrysostomakis, E.M. Kanoupakis, et al., Effect of acute atrial fibrillation on phasic coronary blood flow pattern and flow reserve in humans, Eur. Heart J. 23 (2002) 734–741. [65] K.A. van Bragt, H.M. Nasrallah, M. Kuiper, J.J. Luiken, U. Schotten, S. Verheule, Atrial supply–demand balance in healthy adult pigs: coronary blood flow, oxygen extraction, and lactate production during acute atrial fibrillation, Cardiovasc. Res. 101 (2014) 9–19. [66] S. Sharma, G. Sharma, M. Hote, V. Devagourou, V. Kesari, S. Arava, et al., Light and electron microscopic features of surgically excised left atrial appendage in rheumatic heart disease patients with atrial fibrillation and sinus rhythm, Cardiovasc. Pathol. 23 (2014) 319–326. [67] M. Rienstra, J.X. Sun, S.A. Lubitz, D.S. Frankel, R.S. Vasan, D. Levy, et al., Plasma resistin, adiponectin, and risk of incident atrial fibrillation: the Framingham Offspring Study, Am. Heart J. 163 (2012) 119–124. [68] M. Shimano, R. Shibata, Y. Tsuji, H. Kamiya, T. Uchikawa, S. Harata, et al., Circulating adiponectin levels in patients with atrial fibrillation, Circ. J. 72 (2008) 1120–1124. [69] C. Sztalryd, A.R. Kimmel, Perilipins: lipid droplet coat proteins adapted for tissuespecific energy storage and utilization, and lipid cytoprotection, Biochimie 96 (2014) 96–101. [70] R.C. Pereira, A.M. Delany, E. Canalis, CCAAT/enhancer binding protein homologous protein (DDIT3) induces osteoblastic cell differentiation, Endocrinology 145 (2004) 1952–1960. [71] L.C. Chao, S.J. Bensinger, C.J. Villanueva, K. Wroblewski, P. Tontonoz, Inhibition of adipocyte differentiation by Nur77, Nurr1, and Nor1, Mol. Endocrinol. 22 (2008) 2596–2608. [72] M. Fu, T. Sun, A.L. Bookout, M. Downes, R.T. Yu, R.M. Evans, et al., A nuclear receptor atlas: 3T3-L1 adipogenesis, Mol. Endocrinol. 19 (2005) 2437–2450. [73] V.L. Veum, S.N. Dankel, J. Gjerde, H.J. Nielsen, M.H. Solsvik, C. Haugen, et al., The nuclear receptors NUR77, NURR1 and NOR1 in obesity and during fat loss, Int. J. Obes. (Lond.) 36 (2012) 1195–1202. [74] Y. Li, C. Teng, Angiopoietin-like proteins 3, 4 and 8: regulating lipid metabolism and providing new hope for metabolic syndrome, J. Drug Target. 22 (2014) 679–687. [75] O. Adam, K. Theobald, D. Lavall, M. Grube, H.K. Kroemer, S. Ameling, et al., Increased lysyl oxidase expression and collagen cross-linking during atrial fibrillation, J. Mol. Cell. Cardiol. 50 (2011) 678–685. [76] H.Y. Huang, S.Z. Chen, W.T. Zhang, S.S. Wang, Y. Liu, X. Li, et al., Induction of EMT-like response by BMP4 via up-regulation of lysyl oxidase is required for adipocyte lineage commitment, Stem Cell Res. 10 (2013) 278–287. [77] Y.K. Lin, Y.C. Chen, S.L. Chang, Y.J. Lin, J.H. Chen, Y.H. Yeh, et al., Heart failure epicardial fat increases atrial arrhythmogenesis, Int. J. Cardiol. 167 (2013) 1979–1983. [78] J. Butler, A. Kalogeropoulos, V. Georgiopoulou, N. de Rekeneire, N. Rodondi, A.L. Smith, et al., Serum resistin concentrations and risk of new onset heart failure in older persons: the health, aging, and body composition (Health ABC) Study, Arterioscler. Thromb. Vasc. Biol. 29 (2009) 1144–1149. [79] D.S. Frankel, R.S. Vasan, R.B. D'Agostino Sr., E.J. Benjamin, D. Levy, T.J. Wang, et al., Resistin, adiponectin, and risk of heart failure the Framingham Offspring Study, J. Am. Coll. Cardiol. 53 (2009) 754–762. [80] T. Pischon, C.J. Girman, G.S. Hotamisligil, N. Rifai, F.B. Hu, E.B. Rimm, Plasma adiponectin levels and risk of myocardial infarction in men, JAMA 291 (2004) 1730–1737. [81] C. Weikert, S. Westphal, K. Berger, J. Dierkes, M. Mohlig, J. Spranger, et al., Plasma resistin levels and risk of myocardial infarction and ischemic stroke, J. Clin. Endocrinol. Metab. 93 (2008) 2647–2653. [82] Y.K. Lin, Y.J. Chen, S.A. Chen, Potential atrial arrhythmogenicity of adipocytes: implications for the genesis of atrial fibrillation, Med. Hypotheses 74 (2010) 1026–1029.