Animal studies were performed according to the Guide for the Care and Use of Laboratory Animals, published by the US National Institutes of Health (NIH publication No. 85–23, revised 1996) and approved by the Animal Experiments Inspectorate of the Danish Ministry of Justice. Four-week-old male Sprague–Dawley rats (Taconic, Ll. Skensved, Denmark) were randomly stratified into two groups. One group (controls) received unlimited normal chow (13 kcal % fat, Altromin 1319, Brogaarden, Lynge, Denmark) and water, and the other group, fructose-fat fed rats (FFFR), received unlimited high-fat diet (60 kcal % saturated fat, D12492, Research Diets, New Brunswick, USA) and 10 % fructose (F0127, Sigma-Aldrich, Brøndby, Denmark) in the drinking water as previously described . To avoid bacterial growth in the drinking water, citric acid was added to give a pH of 3.6 in both control and fructose water. Rats were kept on this special diet for 6 weeks before entering the experiments described below.
Blood samples, fasting glucose and insulin measurements
Rats fasted for 16 h were gently restrained and their fasting blood glucose was measured on a (~2 μL) tail blood sample using a FreeStyle Lite analyzer (Abbott, Copenhagen, Denmark). Subsequently, a sample of tongue blood (~300 μl) was collected and transferred to heparin containing micro tubes (VWR, Herlev, Denmark). The samples were centrifuged at 1500 g for 15 min at 4 °C. Plasma was collected and stored at −80 °C for later analysis.
Plasma insulin was measured with ELISA (cat no. EIA2048, DRG International, Marburg, Germany) according to the manufactures instructions.
In vivo ECG recordings
Seven needle electrodes were inserted subcutaneously into lightly anesthetized (1.5 % isoflurane in O2) rats. Four electrodes were connected to the limbs, while the remaining electrodes were placed precordially. ECG’s were obtained with an electronic recorder (GE MAC 800, GE Healthcare, Milwaukee, WI). ECG recordings were transferred to a central database (GE Muse Cardiology Information System, GE Healthcare, Milwaukee, WI) and analyzed manually in a blinded fashion using an electronic ruler. Measurements included P-wave duration and amplitude, QRS-complex duration and amplitude as well as QT-interval and corrected QT-interval (QTc) using formulas proposed by Bazett and Kmecova . The QRS width was measured from the beginning of the Q-wave to the negative peak of the S-wave. QT was measured from the beginning of the Q-wave to the point where the downslope of R’ return to the isoelectric line.
Conduction velocity and contractility of ventricular strips
Conduction velocity and contractility was measured in tissue strips from the free wall of the right ventricle as previously described for atrial tissue . In short, hearts were explanted from rats anesthetized with 5 % isoflurane in 35 % O2/N2. Tissue strips were mounted in a 1 ml chamber and perfused with 35 °C, 100 % O2 bubbled Tyrode’s buffer (mM: 136 NaCl, 4 KCl, 0.8 MgCl2, 1.8 CaCl2, 5 HEPES, 5 MES, 10 Glucose, pH 7.3) at a flow rate of 2 ml/min. The tissue was paced with a unipolar stimulation electrode at 1 Hz, impulse duration 0.5 ms and double threshold voltage (Master-8 stimulator, A.M.P.I., Jerusalem, Israel). Conduction velocity was measured using two extracellular microelectrodes (Platinium/Iridium (PI20030.5B10), Micro Probe Inc., Gaithersburg, USA) placed on the longitudinal axis of the tissue strip. Time of local activation under the first and second microelectrode was determined as the time of minimum dU/dt by custom written MatLab script. Conduction velocity was calculated as the inter-electrode distance divided by the inter-electrode delay. Force was recorded continuously using the isometric force-transducer. Muscle length was adjusted to the level where contractions under control conditions were 50 % of maximal. Following a 15 min resting period, conduction velocity was measured over a 20 min period and developed force and passive tension were analyzed and calculated by the custom written MatLab script.
Rats were anaesthetized with 5 % isoflurane in 35 % O2/N2 and the aorta was cannulated as previously described . The hearts were transferred to a perfusion apparatus (Hugo Sachs Elektronik – Harvard Appartus GmbH, March-Hugstetten, Germany) and perfused in the Langendorff mode with modified Krebs-Henseleit solution (mM: 118 NaCl, 4.7 KCl, 1.75 CaCl2, 1.2 KH2PO4, 1.2 Mg2SO4, 24.9 NaHCO3, 11.0 glucose) under continuously bubbling with carbogen (95 % O2/5 % CO2) and a constant perfusion pressure of 80 mmHg. A fluid filled balloon (size 5) (Hugo Sachs Elektronik – Harvard Appartus GmbH, March-Hugstetten, Germany) was inserted through an incision in the left auricle into the left ventricle to allow measurements of left ventricular pressure (LVP). The volume of the balloon was adjusted to give an end-diastolic pressure of approximately 5 mmHg. Subsequently, a ligature was placed around the left anterior descending coronary artery (LAD), enabling induction of ischemia. The experiment progressed as 30 min of normal perfusion (baseline), 30 min of LAD occlusion, and 60 min of re-perfusion. Langendorff studies were recorded and evaluated using iox2 version 22.214.171.124 software (emka TECHNOLOGIES, Paris, France).
Area at risk and infarct size
At the end of Langendorff experiments the LAD ligature was re-clamped and the heart perfused with Evans Blue dye (0.1 %) (SIGMA, E2129) to evaluate the area at risk of infarction. Subsequently, the heart was sliced into ~2 mm thick vertical slices and incubated in 2,3,5-triphenyltetrazoliumchloride (TTC, SIGMA T8877) 10 mg/ml in 0.1 M phosphate buffer (pH 7.4) for 10 min at 37 °C. Subsequently, the tissue slices were washed 3 times in MilliQ water and transferred to 4 % formalin overnight. Finally, the tissue slices were weighed and scanned on both sides at 1200 DPI to analyze the infarct size. Pictures were analyzed in a blinded fashion using ImageJ software.
ECG recordings during Langendorff experiments were obtained using a 6-lead Einthoven ECG recording system (Hugo Sachs Elektronik – Harvard Apparatus GmbH, March-Hugstetten, Germany) and analyzed using ecgAUTO v126.96.36.199 software (emka TECHNOLOGIES, Paris, France). In short, individual ECG libraries were generated for each experiment. QRS and QT duration were analyzed and the number of extra systoles (also known as a ventricular premature complex), as well as the number, morphology and duration of each episode of arrhythmia determined. In general, the arrhythmia analysis complies with the updated Lambeth conventions . Episodes of VT were defined as at least 4 consecutive ventricular complexes, however, we did not distinguish between monomorphic, polymorphic, or Torsades de pointes VTs. VF was defined as a continuous entry for which individual QRS complexes could no longer be distinguished from each other. In regard to QT duration, the end of the T-wave was set as the point where the downslope returned to the isoelectric line.
Isolation of cardiomyocytes
Rats were anesthetized by 5 % isoflurane in 35 % O2/N2 and the heart perfused in a custom made perfusion system at a constant pressure of 60 mmHg. The heart was perfused with Tyrode’s solution (mM: 136 NaCl, 4 KCl, 5 HEPES, 5 MES, 0.8 MgCl2, 1.8 CaCl2, 10 glucose, pH 7.4) for 5 min followed by 2 min perfusion with Ca2+ free Tyrode’s solution and 2 min with potassium-gluconate buffer (mM: 20 NaCl, 120 potassium gluconate, 1 MgCl2, 10 HEPES, 10 glucose, pH 7.4). Subsequently, the heart was perfused with potassium-gluconate buffer including collagenase (140–165 U/ml) (Type 2 from Worthington) until the heart was digested (approximately 20–25 min). The ventricles were then sliced into small pieces, placed in collagenase buffer and bubbled with 100 % O2 till the tissue was dissolved. The solution was filtered and left to settle for 10 min. Cardiomyocytes used for measurements of gap junction conductivity was gradually added Ca2+ to give a final concentration of 0.75 mM.
Gap junction coupling
Intercellular coupling was measured in ventricular cell pairs at room temperature using the dual whole-cell patch-clamp method as previously described . Cells were voltage clamped at −10 mV, and a 1 s −10 mV pulse was applied to one cell at 0.1 Hz. Data was analyzed using Cellworks Reader (NPI Electronic). The average intercellular conductance during a 90 s interval following patch break was calculated as the resulting current deflection in the passive cell (nonpulsed) divided by the transjunctional voltage difference using a custom-made MatLab script.
Whole-cell sodium currents (INa) were recorded at room temperature on ventricular myocytes. The internal solution consisted of 5 mM NaCl, 135 mM CsF, 10 mM EGTA, 5 mM MgATP and 5 mM HEPES (pH 7.2 with CsOH) and the low-sodium external solution consisted of 20 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 0.1 mM CdCl2, 20 mM HEPES, 117.5 mM CsCl and 11 mM glucose (pH 7.4 with CsOH). To characterize the voltage dependence of the peak sodium current, single myocytes were held at −120 mV, and 200 ms voltage steps were applied from −80 to +15 mV in 5 mV increments. Interval between each steps was 3 s. Measurements were made with pClamp10 software and a MultiClamp 700B amplifier sampling at 20 kHz and filtering at 5 kHz (Molecular Devices, Axon Instruments, Sunnyvale, USA). Borosilicate glass pipettes were pulled on a DPZ-Universal puller (Zeitz Instruments, Martinsried, Germany). The pipettes had a resistance of 1.5–2.5 MΩ when filled with intracellular solution. The series resistance recorded in the whole-cell configuration was 2–5 MΩ and was compensated (80 %).
K+-channel voltage-clamp recordings were made using an EPC7 amplifier (HEKA Elektronik, Lambrecht/Pfalz, Germany) and digitized with a Digidata 1440A converter (Axon Instruments, Molecular Devices, Sunnyvale, California). pClamp10 software was used for data acquisition (Axon Instruments, Molecular Devices, Sunnyvale, California). Patch pipettes were fabricated from borosilicate glass capillaries (GC150F, Harvard Apparatus, Edenbridge, UK) using a gravity puller PIP5 puller (HEKA, Lambrecht/Pfalz, Germany) and the pipette resistance ranged from 1 to 3 MΩ. The cardiomyocytes were superfused with a HEPES buffer (mM: 126 NaCl, 5.4 KCl, 1.0 MgCl2, 2.0 CaCl2, 10 HEPES and 11 glucose, pH adjusted to 7.4 with NaOH and 300 μM Cd to block ICaL). The patch pipette solution had the following composition (in mM): 90 K-aspartate, 30 KCl, 5.5 glucose, 1.0 MgCl2, 5 EGTA, 5 MgATP, 5 HEPES, 10 NaCl, pH 7.2 with KOH. The experiments were performed at 37 °C. After a whole-cell patch was established, cell capacitance was measured by applying −5 mV voltage steps. From a holding potential of −80 mV, sodium currents were inactivated by a 10 ms step to −40 mV and potassium currents were activated by a step protocol ranging from −120 to 40 mV for 2 s. A voltage–current relationship of the initial part of the step protocol is shown. Compensation of series resistance to 60–70 % was applied to minimize voltage errors. All analog signals were acquired at 10–50 kHz and filtered at 4–6 kHz. All electrophysiological experiments were blinded during acquisition and data analysis.
Plasma and cardiac lipid measurements
Enzymatic kits were used for measurements of free fatty acids (Wako NEFA C kit, TriChem Aps, Frederikssund, Denmark), cholesterol (CHOD-PAP; Roche Applied Science), and free glycerol and triglycerides (TR0100, serum triglyceride determination kit, Sigma) in plasma samples. Cardiac lipid content was measured by thin-layer chromatography (TLC) as previously described . All samples were analyzed in duplicate on separate TLC plates and quantified by digital image analysis using the ImageJ software.
Fixation of hearts for fibrosis and Cx43 staining
Rats were anaesthetized by 5 % isoflurane in 35 % O2/N2 and the hearts removed, cannulated and connected to a homemade perfusion system as described for isolation of cardiomyocytes. The hearts were perfused at 60 mmHg with modified Krebs-Henseleit solution for 2 min followed by 10 ml of 2 % paraformaldehyde (PFA) dissolved in PBS pH 7.2. The hearts were immersed in 2 % PFA solution and stored at 4 °C overnight. The fixed hearts were subsequently stored in 0.5 % PFA at 4 °C until staining.
Fixated hearts were coated in paraffin and cut in 12 μm thick slices. Slices of right ventricular tissue were deparaffinized, permeabilized and incubated with the primary antibodies anti-Cx43 (C6219, Sigma-Aldrich, St. Louis, MO, 1:1000) and anti- N-cadherin (C3865 Sigma-Aldrich, St. Louis, MO, 1:500) and the appropriate secondary ALEXA conjugated antibodies (Life Technologies, Carlsbad, CA) as previously described . Slices were imaged using a laser confocal microscope (Zeiss LSM 780, Carl Zeiss, Jena, Germany) and a 63x Oil, NA 1.4 objective.
Amounts of total Cx43 and Cx43 not localized within 3 μm of N-Cadherin (regarded as Cx43 not localized in intercalated discs), were analyzed in a blinded fashion.
Sections of fixed ventricular tissue were in coated in paraffin and processed for Masson’s Trichrome staining and histological examination as previously described . The analysis was made in a blinded fashion.
Hearts were explanted from isoflurane anesthetized rats and homogenized as previously described . Ten percent Bis-Tris gels and MOPS running buffer were used and the proteins were blotted onto nitrocellulose membranes (all from Life technologies, Nærum, Denmark). Anti-Cx43 (C6219) 1:50.000, anti-β-tubulin (T0198) 1:2000 (both from Sigma-Aldrich, Brøndby, Denmark), anti-mouse (IRdye680RD) 1:10.000, and anti-rabbit (IRdye800CW) 1:10.000 (LI-COR Biosciences, Cambridge, UK) were used and membranes analyzed using Odyssey CLx Infrared Imaging System (LI-COR Biosciences, Cambridge, UK).
All data are presented as mean ± standard error of the mean (SEM). Statistical analysis was performed using GraphPad Prism 5 (GraphPad Software,Inc. La Jolla, CA) or SAS 9.2 (SAS Institute, Cary, NC). For single parameters, differences between groups were generally analyzed using an unpaired Students t-test. For a subgroup of data sets, variances were significantly different between groups. In these cases, an unpaired t-test with Welch’s correction was performed. The ECG and Langendorff data were analyzed using mixed models with repeated measures in PROC MIXED, followed by Duncan’s new multiple range test. p < 0.05 was considered statistically significant.