Memantine targets glutamate receptors in atrial cardiomyocytes to prevent and treat atrial fibrillation

Atrial ﬁ brillation (AF), the most common sustained cardiac arrhythmia, can lead to severe consequences such as heart failure and stroke 1 . Recently, we have demonstrated that atrial cardiomyocytes express ionotropic glutamate receptors (iGluRs) in high abundance, inhibi-tion of which can signi ﬁ cantly reduce the incidence of AF and effectively block AF progression in an experimental AF model using isolated rat hearts 2 . However, to date, none of medicines targeting iGluRs has been prescribed in clinical AF treatment. Memantine is a commonly used drug in clinic for the treatment of Alzheimer ’ s disease by antagonizing iGluRs in neurons 3 . We therefore hypothe-sized that memantine may also exert an effect against AF by blocking iGluRs in atrial cardiomyocytes. First, we assessed the effect of memantine on the prevention and termination of AF. To examine the preventive effect of memantine on AF, we constructed three in vitro rat AF models, covering three common clinical types of AF, i.e., stretch-induced AF, cholinergic AF and ischemia-induced AF 4 . As shown in Fig. 1a, memantine perfusion effectively reduced the incidence of AF in a concentration-dependent

CaCl2, 1.2 KH2PO4, 5 HEPES, 1 MgCl2, and 5.5 glucose (pH adjusted to 7.4 with NaOH). After washing out blood from the heart chambers, the hearts were connected to a Langendorff setup and perfused with 37°C oxygenated Tyrode's solution. For ECG recording of perfused hearts, electrocardiographic electrodes were attached to the free wall of the outside of the atria and screwed into the ventricular apex. The cardiac electrical signals were amplified and acquired with BioAmp and PowerLab (AD Instruments, USA). The data were analyzed manually using LabChart7 (AD Instruments, USA). All hearts were kept at 37 °C during the whole course of recording.
AF was defined as the occurrence of rapid irregular atrial rhythms with irregular R-R intervals. The duration of AF was measured from the end of burst pacing to the first Pwave detected after the rapid irregular atrial rhythm.
A stretch-induced AF model was established as previously described with modifications 1 . Briefly, acute atrial dilation was conducted by ligating the inferior vena cava and then adjusting the perfusion speed to increase the atrial volume to twice the initial volume. After 40 min of atrial stretching, AF was induced by 20 Hz burst atrial pacing (pulse width 2 ms) for 40 seconds followed by 10 cycles of 40 seconds atrial pacing at 1-minute intervals.
A cholinergic AF (acetylcholine-induced AF) model was induced as previously described with modifications 1,2 . Acetylcholine (2 μM) was administered to generate a rat model of acetylcholine-induced AF. Rat hearts were infused in vitro with acetylcholine for approximately 10 min, and then AF was induced by 20 Hz burst atrial pacing (pulse width 2 ms) for 40 seconds followed by 10 cycles of 40 seconds atrial pacing at 1-minute intervals.
An ischemia-induced AF model was established as previously described with modifications 1 . In brief, atrial ischemia was induced by perfusing hearts with one-fifth of normal perfusion pressure throughout. At the same time, right atrial pressure was maintained to prevent collapse during ischemia. After 30 min of ischemia, AF was induced by 20 Hz burst atrial pacing (pulse width 2 ms) for 40 seconds followed by 10 cycles of 40 seconds atrial pacing at 1-minute intervals.

In vivo AF induction
An asphyxia-induced rat AF model was induced by rapid atrial pacing during brief episodes of asphyxia as previously described with minor modifications 3 . The rats were anesthetized using 1.5% isoflurane, artificially respirated, and kept at 37°C. A pacing catheter (1.4F, Millar Instruments, USA) was inserted via the right jugular vein into the right atria. The pacing threshold was determined by incrementally increasing the voltage until atrial capture occurred. Pacing was then performed at 2× pacing threshold.
The rats were subjected to asphyxia, which was induced by clamping the tracheal tube with a pair of hemostatic forceps at the end of the inspiratory cycle, for 30 seconds. Ten seconds after the beginning of asphyxia stimulation, 20 Hz burst pacing was performed at 2× threshold voltage for 30 seconds to trigger AF. AF was defined as rapid and fragmented atrial electrograms with an irregular ventricular rhythm lasting for at least 2 min immediately following burst pacing.
A cholinergic rat AF model was established by injection of acetylcholine and rapid atrial pacing. Briefly, a pacing catheter (1.4F, Millar Instruments, USA) was inserted into the right atria of each anesthetized rat. Then, the rats received a single intravenous bolus injection of acetylcholine (1 mg/kg, 0.1 mL) within 5 seconds. Next, 20 Hz burst pacing was performed at 2× threshold voltage for 30 seconds after acetylcholine administration to trigger AF.
A transverse aortic constriction (TAC)-induced rat AF model was established as previously described with modifications 4,5 . A pacing catheter (1.4F, Millar Instruments, USA) was inserted into the right atrium of the rat, and 20 Hz burst pacing was performed at 2× threshold voltage for 30 seconds to trigger AF.

Isolation of adult rat atrial cardiomyocytes
Atrial cardiomyocytes were isolated from the Langendorff-perfused hearts of adult male Sprague Dawley rats as previously described with minor modifications 6 . Briefly, the heart was removed, mounted on a Langendorff apparatus and perfused with Tyrode's solution to wash out the blood. Then, the heart was perfused with Ca 2+ -free Tyrode's solution buffer containing 1 mg/mL collagenase type II (Worthington, USA) and 1 mg/mL BSA for digestion at 37°C for 15 min. When the heart became flaccid, the atria were dissected out from the heart, cut into small pieces, and resuspended in

Patch-clamp recording
Whole-cell patch-clamp recording of atrial cardiomyocytes was performed as previously described 7

Optical mapping
Optical mapping experiments were performed essentially as described previously 8 . In brief, adult male Sprague-Dawley rats were anesthetized by intraperitoneal injection of pentobarbital (25 mg/kg) containing 120 IU heparin, the chest was opened, and the heart was harvested. After rapid excision and Langendorff perfusion at 37 °C with oxygenated Tyrode's solution, the heart was allowed to recover for 10 min and then loaded with the voltage-sensitive dye RH237 (10 μM, AAT bioquest, USA) for 5 min.
Blebbistatin (10 μM, MCE, USA) was added to the perfusate to prevent motion artifacts.
The dye was excited using LED light sources centered at 550 nm. Images were captured with a high-speed camera (MiCAM ULTIMA, USA). Activation time was determined as the time point of maximum change in fluorescence over time (dF/dt) for each fluorescent signal in the array. Reentry cycles were defined as repetitive returns of an impulse into the same atrial area with a constant rotation pattern over at least 2 cycles.
All the data were collected and analyzed using MiCAM ULTIMA acquisition software and BV_Ana software (SciMedia, USA).

Multielectrode array (MEA) recording
Human induced pluripotent stem cell-derived atrial cardiomyocytes (iPSC-ACMs) were obtained from the Institute of Biophysics of the Chinese Academy of Sciences (Beijing, China). After recovery, the cells were cultured in a T25 flask precoated with vitronectin (0.01 μg/μL, Cauliscell, China) in cardiomyocyte maintenance medium (Cauliscell, China) at 37°C and 5% CO2. The medium was refreshed every 2 days.
At day 12, the cells were dissociated and resuspended in cardiac recovery medium at a density of 3 x 10 6 cells/mL. Aliquots of the cell suspensions (10 μL) were plated on CytoView MEA 24-well plates (Axion BioSystems, USA). The MEA device automatically adjusted and controlled the environment (37°C and 5% CO2) to maintain the temperature and pH of the medium. The data were acquired using Maestro MEA System (Axion BioSystems, USA). The induction of arrhythmic events on MEA recording was carried out as previously reported with some modification 9,10 , human iPSC-ACMs were treated with 50 μM sotalol and cultured in a nitrogen-supplied hypoxia incubator subjected to a cycle of hypoxia (1% oxygen) for 3 h and overnight normoxia (19% oxygen).

Western blotting
The protein was extracted by RIPA lysis buffer (Beyotime Biotechnology, China) supplemented with a cocktail of protease inhibitors and phosphatase inhibitors (Roche, USA) at 4°C. Protein concentrations were determined with BCA kit (Beyotime Biotechnology, China), and equal amount of protein was separated by 10% SDS-PAGE (Thermo Fisher Scientific, USA), then transferred onto PVDF membranes (Millipore, USA). Next, the membranes were blocked with 5% BSA in TBS + 0.1% Tween (TBST) for 1 h, then incubated with primary antibody against RyR2 (Abcam, USA), RyR2 (phospho-S2814) (Badrilla, UK) or GAPDH (Abcam, USA) overnight at 4℃. The next day, the membranes were washed with TBST and incubated with the HRP-conjugated secondary antibody (Invitrogen, USA) for 1 h, and the bands were visualized with ChemiDoc Touch Gel Imaging System (Bio-Rad, USA).

Statistics
Data normality was tested using the Kolmogorov-Smirnov test. For comparisons between two groups, two-tailed Student's t-test or the nonparametric Mann-Whitney test was used. Statistics on percentages were performed with a Chi-squared test.
Statistical significance was defined as p ≤ 0.05. Each animal sample and drug used in the electrophysiological experiments was assigned a code, and the data were unmasked upon completion of the study. All experiments were repeated at least 3 times. All the data were presented as the mean ± SEM. All statistical analyses were performed with GraphPad Prism 8.