2.1 In vivo experiments
The rats were anesthetized with zoletil (Virbac S.A., Carros, France, 15 mg/kg, i.m.) and xylazine (Interchemie, Castenray, Netherlands, 0.1 mg/kg, i.m.) and ventilated mechanically. The chest was opened via a midsternal thoracotomy and the heart was exposed. Blood [K+] was decreased by hypokalemiс diet contained only cereal and furosemide administration (2 mg/kg dose i.m. daily, 3 days before experiment). To increase blood [K+], i.v. infusion of 10% KCl solution was performed during the acute experiment.
Acute local ischemia was induced by a ligation of a proximal third of the left anterior descending coronary artery (LAD) by a coated braided polyester ligature (№ 5-0, Ti-Cron, Cardiopoint, Covidien, Dominican Republic). After 5-min of coronary occlusion ligature was loosened to provide reperfusion. This technique of short ischemia followed by reperfusion was used in a number of our previous studies and was verified to produce a sufficient number of ventricular arrhythmias27, 30. The measurements of blood [K+] were done immediately before the onset of ischemia (EasyStat, Medica Corp, USA) in blood samples withdrawn from the femoral vein. Ventricular tachycardia and fibrillation were assessed at first min of reperfusion. Extrasystolic burden (ESB) was calculated as a sum of ESs during the first min of reperfusion divided by 60 sec excluding periods of VF or VT, if any.
Epicardial mapping was performed from the epicardium of the left ventricle (LV) using a square multielectrode array (8×8 leads, 0.5 mm interelectrode spacing) by a custom-designed 128-channel recording system (16 bits; bandwidth 0.05 to 1000 Hz; sampling rate of 4000 Hz). The mapping array was located on the anterior surface of LV in a way to cover equally (at approximately 1:1 ratio spatially) the normal and affected areas. The latter was recognized by cyanotic coloration (Fig. 6 A, B) at the time of electrode application. At subsequent analysis, the affected zone was defined as lead sites demonstrating ST-segment elevation during the period of ischemia. In each lead, activation time (AT) and repolarization time (RT) were determined as the instants of dV/dt min during QRS complex and dV/dt max during T-wave respectively. ARI, a surrogate for action potential duration (APD), was calculated as ARI = RT – AT (Fig.6 C).
Conduction velocity (CV) was measured using isochrone activation mapping under electrical stimulation (400 bpm, 2 mA, 2 ms) in the middle of the LV. Pacing stimuli were delivered to the unaffected area (Fig.1D). In order to calculate local CVs, local gradients of activation (AG) were computed using finite difference method 31, 32 well suited for the regularly spaced electrode array (Fig. 6E). The horizontal and vertical components of AG were computed for the sites of the electrode array using standard first-order finite-difference stencil (Fig. 6F) as AGx=(ti+1,j – ti–1,j)/2d, AGy=(ti,j+1 – ti,j–1)/2d.
Along with AG, which is an averaged parameter, we also considered the minimal activation velocity between the given point and its eight nearest neighbors located top, bottom, right, left and diagonally (Fig. 6F). The magnitudes of the minimal activation velocity were used to detect the boundary between the normal and ischemic zones and the direction of activation wavefront, which was not known a priori. The boundary electrodes located on the edge of the multielectrode array (Fig. 6E) were excluded from the analysis, since they did not have enough neighboring electrodes for accurate determination of activation velocity. In some neighboring electrodes, activation times could coincide; such points were also excluded from the analysis.