Online supplement

Methods

The study protocol was approved by the Institutional Animal Care and Use Committee of Indiana University School of Medicine and the Methodist Research Institute, Indianapolis, Indiana, and conformed to the regulations for humane care and treatment of animals established by the NIH. A total of 29 female New Zealand white rabbits (3.5-5.0 kg) were used in this study. Among them, 20 were used for chronically ventricular pacing and 9 normal rabbit hearts that were not paced were used as controls.

Pacing-Induced long-term CM and Optical Mapping

Ventricular pacing was used to induce long-term cardiac memory (CM). Sterile surgery was performed with isoflurane as general anesthesia. Either a left lateral thoracotomy (LV pacing) or a right lateral thoracotomy (RV pacing) was performed and an epicardial pacing lead was sutured to the lateral wall of either the right ventricle (n=9) or the left ventricle (n=11). The pacing leads were then connected to a modified single chamber pacemaker (Medtronic, Inc., Minneapolis, MN). After 1 week of recovery, the ventricles were paced at 270 beats per min (bpm) for 3-5 weeks. Each week, the pacemakers were turned off for 1 hour to assess if there were T-wave changes on Lead II of the ECG. Chronic CM was defined by either a flat or inverted T-wave (Figure 1). Echocardiography was performed both before the pacemaker implantation and at follow up. Measurements of left ventricular diameters and fractional shortening were made from M-mode recordings of the parasternal short-axis view at the tips of the papillary muscles using the leading edge technique in three to four cardiac cycles and averaged. After the echocardiographic recordings were made, optical mapping was performed on both paced hearts and normal hearts of both the RV and LV (Online Supplement Figure 1). The rabbits were intravenously injected with 1,000 units of heparin and anesthetized with sodium pentobarbital (35 mg/kg). A median sternotomy was performed and the hearts were rapidly excised. Excised hearts were Langendorff perfused at 25 to 35 mL/min with Tyrode’s solution (in mmol/L: NaCl 125, KCl 4.5, NaHCO3 24, NaH2PO4 1.8, CaCl2 1.8, MgCl2 0.5, and glucose 5.5) that was bubbled with 95% O2/5% CO2 to maintain a pH of 7.40. The hearts were stained with Rhod-2 AM (1.48 μmol/L) for Cai and RH237 (10 μmol/L) for Vm mapping. Contractility was blocked with blebbistatin (15 – 20 μmol/L). To prevent induction of short term CM the hearts remained in normal sinus rhythm except for specific programmed stimulation during optical recordings.1 The double-stained hearts were illuminated with a laser at 532 nm wavelength. The fluorescence was recorded with 100×100 pixels for a spatial resolution of 0.35×0.35 mm2 per pixel at a 2 ms/frame sampling rate. Optical signals were processed with both spatial (3×3 pixels Gaussian filter) and temporal (3 frames moving average) filtering.

Protocol I – Electrophysiological Effects of Cardiac Memory

The CM model was created in 7 hearts with RV pacing and 3 hearts with LV pacing. Five normal hearts were used as controls. To determine the electrophysiological effects of CM, the ventricular APD80 was determined in each heart by pacing the right atrial appendage at 300 ms. Optical maps were acquired after at least 30 paced beats. To determine the role of the IKAS in this model, apamin (100 nmol/L) was then added to the perfusate and the protocol was repeated after 30 minutes. At that concentration, apamin is a highly specific IKAS blocker.2 Radiofrequency ablation of the atrioventricular (AV) node was then performed to reduce the ventricular rate to < 60 bpm (ventricular escape cycle length >1000 ms). Pacing from the RV apex, a S1/S2/S3 (short/long/short) pacing protocol (S1 30 beats with S1-S1 300 ms, a long S1-S2 of 1000 ms or 2000ms and a S2-S3 starting from 300 ms and gradually shortened to the ventricular effective refractory period (ERP)) were used to simulate the ECG characteristics that initiates early after-depolarizations (EADs) and torsades de pointes (TdP) ventricular tachycardia in the clinical setting.3 Finally, a dynamic ventricular pacing protocol was used to determine arrhythmia inducibility. The PCLs were progressively shortened from 1000 ms (1000, 900, 800, 700, 600, 500, 400, 300, 280, 260, 240, 220, 200, 190, 180, 170, 160, 150,140,130ms) until VF was induced or the loss of 1:1 capture. If VF was not induced by the dynamic pacing protocol, burst pacing (PCL 50 ms, pacing duration 10 s) was used to try to induce VF. Any VF that occurred was allowed to continue for 180 s before a defibrillation attempt was made. Shocks were delivered between stainless steel electrodes placed contiguous to the ventricles connected to a Ventritex HVS-02 programmable defibrillator. Each shock was a truncated biphasic exponential wave form with a pulse width of 3 ms/3 ms at 300 volts. After the return to sinus rhythm, the heart was allowed to rest for 1 min before continuing the pacing protocol. VF induction was attempted 4 times in each heart. Phase singularities (PS) were used to quantify VF characteristics.4

Protocol II – Effects of Cardiac Memory on Arrhythmia Inducibility

In protocol I, all arrhythmia inducibility data was acquired after apamin infusion. To determine the effects of CM on arrhythmia inducibility and the role of IKAS, 3 additional hearts with chronic LV pacing were used to acquire data on ventricular stability at baseline and after apamin infusion. In each heart, AV nodal ablation was performed, and to determine the inducibility of any EADs and VF, ventricular pacing was performed as described in protocol I. Any ventricular arrhythmias that resulted from the pacing were noted and then apamin (100 nmol/L) was added to the perfusate and the protocol was repeated after 30 min.

Western Blotting

Among the 3 different subtypes of SK channels, SK2 is known to be the most sensitive to apamin and contribute the most to IKAS.5 To determine the protein expression of SK2 channels within the myocardial tissue, Western blotting was performed as previously described in 4 normal hearts and 4 chronically paced hearts (2 RV paced and 2 LV paced).6 Briefly, immediately after excising the heart, myocardial blocks from sites distal and proximal to the chronic pacing site were excised and chopped. For comparison, tissue from similar areas were taken in the normal hearts. 100 mg tissues were homogenized by POLY TRON in 1 ml RIPA buffer with protease inhibitor (50 mM Tris pH 8.4, 150 mM NaCl, 1% NP40, 0.5% sodium deoxycholate 1 mM PMSF, 2 μg/ml leupeptin, 1 μg/ml pepstatin A, and 5 μg/ml aprotinin). Homogenates were incubated on ice for 30 min and then centrifuged at 14,000 rpm for 15 min. Supernatants (20 μg) were subjected to electrophoresis using Bio-Rad mini gel system. The separated proteins were transferred to PVDF (Millipore). The membrane was bathed in TBS with 5% milk for one hour, and probed with either anti-KCNN2 antibody (for detecting SK2 channel, Sigma, P0483, 1:200) overnight. After the interaction with primary antibody, the membrane was incubated with HRP-conjugated anti-rabbit or anti-mouse secondary antibodies (sigma, 1:2000) for 30 min. Finally, Luminata Crescendo HRP substrate (Millipore, WBLUR100) was added onto the membrane according to manufacturer’s instructions.

RNA preparation and qPCR
The ventricular tissues of 3 female rabbits with long-term cardiac memory were dissected from the epicardium of proximal and distal of the pacing sites, cut into small pieces (Applied Biosystems/Amibion, Austin, TX) and immersed at 4°C overnight, then stored at -80°C. Total RNA was extracted from the tissues using random primers with RNAqueous 4PCR kit (Applied Biosystems/Amibion, Austin, TX). Complementary DNA (cDNA) was synthesized from 100 ng total RNA for each sample using iScript Select cDNA Synthesis Kit (BioRad, Hercules, CA). qPCR was performed using iQ SYBR Green Supermix with iCycler (BioRad, Hercules, CA). Specific primers for SK2 were sense 5’- GCTCGGTCTCGATGAC -3’ and antisense 5’- GCAAGAAGAAGAACCAGAACA -3’. Specific primers for SK3 were sense 5’-TTGCCCAACTCCAGAGC-3’ and antisense 5’-CAAGCAAGTGGTCATTGAGATTTA-3’.The general protocol for amplification was 95°C for 20 s, 58°C for 20 s, and 72°C for 20 s, for 40 cycles. PCR products were analyzed in agarose gels, and the amplicons were gel-purified with QIAquick Gel Extraction Kit (QIAGEN Sciences, Valencia, CA). The sequentially diluted SK2 or SK3 amplicons were used for standard curves. Data were collected by MyiQ Software (BioRad, Hercules, CA), and analyzed using the threshold cycle (Ct) relative quantification method. The glyceraldehyde 3-phosphate dehydrogenase (GAPDH) reference gene was used for normalizing the data. 2-ΔCt corresponds to the ratio of SK2 or SK3 versus GAPDH. The primers of GAPDH were sense 5′-CCATGGAGAAGGCCGGGG-3′ and antisense 5'-CGAAGTGGTCGTGGATGAC-3'.

Histology

For histology, ventricular tissue from 4 normal hearts and 4 chronically paced hearts was embedded in paraffin, sectioned and stained using Masson's trichrome stain. The collagen content was quantified by identifying and counting the number of blue-staining pixels as a percentage of the total tissue area using digital photomicrographs in Adobe Photoshop CS6 software.

Statistical Analysis

Data are presented as mean and 95% confidence interval (CI). Paired Student t tests were used to compare variables measured at baseline and after apamin infusion. An independent-sample t test was used to compare variables measured between groups. A one way ANOVA was used to compare SK2 protein expression between different groups. Categorical parameters between groups comparing VF vulnerability were compared by either a chi-square test or the Fisher exact test. A 2-sided P value of ≤0.05 was considered statistically significant.

Results

All rabbits survived the chronic pacing protocol. A total of 29 hearts were studied – 20 paced hearts and 9 normal hearts that were not paced hearts were used as controls.

Of the 20 paced hearts, 4 paced hearts were used for protein and histology (2 RV and 2 LV paced), 3 paced hearts were used for qPCR analysis, and 13 paced hearts were used for optical mapping (7 RV and 3 LV paced hearts for protocol I, 3 LV paced hearts for protocol II). Of the 9 normal hearts, 5 normal hearts were used for optical mapping, and 4 normal hearts for protein and histology.

References

1.Chan Y-H, Tsai W-C, Ko J-S, Yin D, Chang P-C, Rubart M, Weiss JN, Everett T, Lin S-F, Chen P-S. Small conductance calcium-activated potassium current is activated during hypokalemia and masks short term cardiac memory induced by ventricular pacing. Circulation 2015;132:1377-1386.

2.Yu CC, Ai T, Weiss JN, Chen PS. Apamin does not inhibit human cardiac Na+ current, L-type Ca2+ current or other major K+ currents. PLoS One 2014;9:e96691.

3.Chang PC, Turker I, Lopshire JC, Masroor S, Nguyen BL, Tao W, Rubart M, Chen PS, Chen Z, Ai T. Heterogeneous upregulation of apamin-sensitive potassium currents in failing human ventricles. J Am Heart Assoc Feb 2013;2:e004713.

4.Valderrabano M, Chen PS, Lin SF. Spatial Distribution of Phase Singularities in Ventricular Fibrillation. Circulation 2003;108:354-359.

5.Weatherall KL, Seutin V, Liegeois JF, Marrion NV. Crucial role of a shared extracellular loop in apamin sensitivity and maintenance of pore shape of small-conductance calcium-activated potassium (SK) channels. Proc Natl Acad Sci U S A Nov 8 2011;108:18494-18499.

6.Hsieh YC, Chang PC, Hsueh CH, Lee YS, Shen C, Weiss JN, Chen Z, Ai T, Lin SF, Chen PS. Apamin-sensitive potassium current modulates action potential duration restitution and arrhythmogenesis of failing rabbit ventricles. Circ Arrhythm Electrophysiol Apr 2013;6:410-418.

Figure 1. Area of the heart that was optically mapped.

1