Isolated Cardiac Ryanodine Receptor Function Is Species Specific

Abstract

Concerted robust opening of cardiac ryanodine receptors’ (RyR2) Ca release 1oplasmic reticulum (SR) is fundamental for normal systolic cardiac function. During diastole, infrequent spontaneous RyR2 openings mediate the SR Ca leak that normally constrains SR Ca load. Abnormal large diastolic RyR2-mediated Ca leak events can cause delayed after depolarizations (DADs) and arrhythmias. The RyR2-associated mechanisms underlying these processes are being extensively studied at multiple levels utilizing various model animals. Since there are well-described species-specific differences in cardiac intracellular Ca handing in situ, we tested whether or not single RyR2 function in vitro retains this species specificity. We isolated RyR2-rich heavy SR microsomes from mouse, rat, rabbit, and human ventricular muscle and quantified RyR2 function using identical solutions and methods. The single RyR2 cytosolic Ca sensitivity was similar across these species. However, there were significant species differences in single RyR2 mean open times in both systole and diastole-like solutions. In diastole-like solutions, single rat/mouseRyR2 open probability and frequency of long openings (> 6 ms) were similar, but these values were significantly greater than those of either single rabbit or human RyR2s. We propose these in vitro single RyR2 functional differences across species stem from the species-specific RyR2 regulatory environment present in the source tissue. Our results show the single rabbit RyR2 functional attributes, particularly in diastole-like conditions, replicate those of single human RyR2 best among the species tested.

Graphical Abstract

INTRODUCTION

During the mammalian cardiac action potential (AP), surface membrane Ca influx activates nearby intracellular Ca release channels, type-2 ryanodine receptors (RyR2s), in the sarcoplasmic reticulum (SR). This process is called Ca-induced Ca release (CICR), and the resulting Ca release drives normal systolic contractile function in the heart. Between APs, single RyR2 channels spontaneously open, albeit infrequently, and this produces the diastolic SR Ca leak that normally constrains SR Ca reloading (Bers 2001). Occasionally, when a RyR2 opens during diastole, the released Ca activates neighboring RyR2 channels resulting in a Ca spark (a non-propagating localized bout of inter-RyR2 CICR (Guo et al. 2012). Abnormally large or frequent diastolic sparks can evoke propagating Ca waves and that may drive sufficient electrogenic Na–Ca exchange to generate a delayed after depolarization (DAD).

DADs are abnormal diastolic transient depolarizations that follow a cardiac action potential (Liu et al. 2015). DADS can underlie can arrhythmogenesis in various cardiac diseases (January & Fozzard 1988; Katra & Laurita 2005; Liu et al. 2015) such as heart failure (Pogwizd & Bers 2002), catecholaminergic polymorphic ventricular tachycardia (Kryshtal et al. 2021) and long-QT syndrome (Mohler etal. 2003). If a DAD’s amplitude reaches the AP threshold, then it can trigger a premature AP and dangerous arrhythmogenic activities. Preventing this arrhythmogenic process at its origin requires understanding the spontaneous diastolic RyR2 activity which can trigger the occurrence of DADs.

Intracellular Ca handling in heart is well known to vary between mammals. For example, 90–95% of the Ca underlying the intracellular Ca transient cycles through the SR in rat and mouse myocytes (Bers 2008). The amount cycling through the SR in rabbit and human myocytes is substantially less (~ 70%). In addition, the excitation–contraction coupling gain (the amount of RyR2-mediated SR Ca release divided amount of trigger Ca influx) is between 12 and 20 in rat and mouse while it is far less (2.5–4.0) in other mammals (Bers 2008). This implies that there is substantially differences in Ca between mammalian species. These differences in intracellular Ca handling in the mammalian heart are notable because rat, mouse, and rabbit are commonly used experimental models to address mechanisms underlying human cardiac disease. Given these cross-species differences and RyR2 channel’s central role in cardiac intracellular Ca handling, for a better understanding, we explored the single RyR2 function between these mammals is clear.

Single RyR channel function was first recorded in artificial planar lipid bilayers in 1985 (Smith et al. 1985). Since then, some single RyR2groups have studied chaps-solubilized “purified” RyR2s (Guo et al. 2021; Kryshtal et al. 2021; Loaiza et al. 2013; Mukherjee et al. 2014) that are no longer associated with endogenous regulatory protein partners such as calsequestrin, junctin, triadin, and kinases. Others groups have studied “native” RyR2 s by fusing native SR microsomes into the bilayer (Chen et al. 2013; Guo et al. 2012; Kryshtal et al. 2021; Laver et al. 2004; Porta et al. 2011; Xu & Meissner 1998; Zoghbi et al. 2004). These RyR2s may be associated with their endogenous regulatory partners. The source species of the RyR2 and the recording solutions/conditions also vary between studies as do the recording solutions/conditions, which are often grossly non-physiological. For example, the cytosolic solution may not even include Mg or ATP. These are ever present in cells and both have significant influences on single RyR2 function (Bers 2001; Fill & Copello 2002; Gillespie et al. 2012). For example, ATP binds to RyR2 and will abnormally shift the RyR2’s cytosolic Ca EC50 to sub-micromolar levels if present in the absence of Mg (Fill & Copello 2002; Qin et al. 2009). Magnesium competes with Ca at the RyR2’s cytosolic Ca activation site and, when present with ATP, will make the RyR2’s Ca EC50 about 10 μM (Gillespie et al. 2012). Many single RyR2 studies have been done without cytosolic Mg present. Thus, there is a large body of single RyR2 studies, but it is quite diverse.

This study (we believe) is the first to quantify single “native” RyR2 channel functional variability from healthy hearts across species (mouse, rat, rabbit, and human). We observed that single rat and human RyR2 channel gating (their opening/closing) were consistently and visibly different in identical solutions and after being isolated in the same way. This prompted our quantification of single “native” RyR2channel function across species here and revealed significant differences. These differences have ramifications to the interpretation of single RyR2 channel data and provides a cross-species foundation for future species selection decisions.

METHODS

Wild-type (WT) C57B/6j male mice (2–2.5 months old, n = 30), New Zealand white rabbits (male, 6–7-month-old, n = 10), and adultSprague–Dawley rats were used for this study. All animal studies followed the Guide for the Care and Use of Laboratory Animals (NIH Publication, 8th Edition, 2011) and were approved by the Institutional Animal Care and Use Committee of The Ohio State University(#2021A00000071) and Rush University Medical Center (#12–037 & 16–090). Human left ventricular tissues were obtained from four human organ donors without a history of major cardiovascular diseases like atrial fibrillation and heart failure (see supplemental Table S1). These tissues were not used for heart transplant due to technical reasons and were provided by the Illinois Gift of Hope Organ & Tissue Donor Network (GOH). Consent was obtained by GOH from the donors’ families for use of donor hearts for research. All studies were approved by the Human Study Committees of Rush University Medical Center and The Ohio State University.

Confocal Ca²⁺ Imaging and Action Potential Patch Clamp Measurement in Myocytes

Rabbit and mouse ventricular myocytes were freshly isolated using our well-established protocol as previously described (Ai et al. 2005; Yan et al. 2021; Yan et al. 2018a, b, c; Yan et al. 2018a, b, c). For confocal cellular Ca 2+imaging, myocytes were loaded with a Ca indicator (Fluo4-AM; 5 μM) in 0 Ca Tyrode’s solution for 15 min and washed with Tyrode’s solution containing 1.3 mM Ca for15 min. Calcium imaging (excitation—488 nm; emission—515 nm) was performed while the cells were perfused with the Tyrode’s solution (1.3 mM Ca) at 34.5 °C. SR Ca load was measured using a 10 mM caffeine-induced SR Ca depletion protocol (Yan et al.2021; Yan et al. 2018a, b, c). Caffeine robustly activates RyR2 channels and, thus, depletes the SR of Ca (Porta et al. 2011). SR Caleak was measured using a well-established tetracaine-sensitive protocol as previously described (Shannon et al. 2003; Yan et al. 2018a,b, c). Tetracaine inhibits RyR2 channels and so blocks diastolic RyR2-mediated SR Ca leak (Shannon et al. 2003).

To measure pacing-induced triggered arrhythmic activities (DADs) in freshly isolated myocytes, patch-clamp action potentials (APs)recordings were conducted and analyzed as previously described (Desantiago et al. 2008; Yan et al. 2021). Capillary tubes (Kimax-51,No. 34502; Kimble products) were used to make patch-clamp pipettes (2–4 Megaohms) using a Narishige pipette puller (PP-83). Isolated cardiomyocytes were placed on laminin-coated coverslips then placed into a holding chamber and placed on the stage of an inverted Nikon microscope (eclipse TE300) for perfusion with Tyrode’s solution (1.3 mM Ca). The patch-clamp pipette was filled with internal solution (in mM: L-glutamic acid K salt 130, KCl 10, ATP 4, HEPES-free acid 10, NaCl 10, MgCl 4). Experiments were performed at34.5 °C, and APs were recorded in the whole-cell ruptured patch-clamp configuration with a low-pass filter at 5 kHz and digitized at10 kHz using an Axon Axopatch 200B patch-clamp amplifier, Axon Digidata 1440B low-noise data acquisition interface system, andPClamp-10 Software. APs were evoked by 6 ms stimulation pulses of 6–8 mV (~ 1.5 times activation threshold), steady-state APs were recorded for 40 s, and then stimulation was stopped with continuous recording of resting membrane potential for 20 s. The lowest pacing frequency that induced DAD was determined for each cell and reported as the DAD threshold.

Heavy sarcoplasmic reticulum (SR) microsomes extraction and RyR2 single channel recording

Heavy sarcoplasmic reticulum (SR) microsomes were isolated from ventricular muscle of humans, rabbits, rats, and mice. The ventricular muscle was minced with scissors and then homogenized with a Kinematica Polytron PT 10–35 GT-D in a NaCl 0.9%, 10 mM Tris-Maleate, and pH 7.2. The resulting homogenate was centrifuged at 4 °C (7000 × g; 20 min), and the supernatant collected and recentrifuged (8700 × g; 20 min). The resulting supernatant was collected and centrifuged at 4 °C (43,000 × g; 60 min). The resulting pellet was then resuspended in a small volume of the buffer with sucrose and DTT added. The resulting resuspension, which contained heavy RyR2-rich “native” SR microsomes, was divided into aliquots, flash frozen, and stored at −80 °C until used.

The SR microsomes were fused into artificial planar lipid bilayers in order to record single RyR2 channel function. Briefly, the single-channel recording chamber consists of two solution pools (each 1.2 mL). These two pools are separated by a thin Teflon partition with a100-μM-diameter circular hole at its center. The planar bilayer was formed in this hole. SR microsomes were added to on one side of the bilayer (cis; other side trans). The RyR2’s cytosolic side consistently faced into the cis-chamber after microsome fusion (Tu et al. 1994). Ag/AgCl electrodes in nearby wells were connected to the cis- and trans-solutions by an agar bridge and to the head stage of an Axopatch200B amplifier (Molecular Devices, LLC; San Jose, CA). Solutions on both sides of the bilayer could be independently stirred and perfused. The cytosolic solution (cis) was held at virtual ground and the potential of the other chamber varied (relative to ground) as specified. RyR2-mediated currents were digitized at 50 kHz using an Axon Digidata 1544B AD converter. Single RyR2 recordings were analyzed using pClamp software after being filtered at 1 kHz and events (openings or closings) shorter than 0.5 ms were excluded. Specifically, the durations of individual RyR2 openings and closings were measured applying a 50% current amplitude threshold to define openings and closings. RyR2 open probability was determined per recording by dividing the sum of open times by the total duration of the single-channel recording. Single RyR2 mean open time, mean closed time, and average opening frequency were also determined.

Planar lipid bilayers were formed using 5:4:1 lipid mixture in n-decane (50 mg/ml): bovine brain phosphatidylethanolamine, -serine, and -choline, respectively (Avanti Polar Lipids; Birmingham, AL). This 5:4:1 lipid mixture includes 3 major endoplasmic-sarcoplasmic reticular lipids (van Meer et al. 2008), sufficiently supports the fusion of heavy SR microsome fusion in to the planar lipid bilayers, and is consistent with that used in our previous single RyR2 channel studies. For SR microsome/bilayer fusion, the cis-chamber contained250 mM HEPES, 114 mM Tris, 1 mM CaCl, and pH 7.4. The solution on the other side of the bilayer contained the luminal recording solution (described below). SR microsomes (2 μl) were added to the cis-chamber and vigorously stirred before 50–300 μL of 4 M cesium methanesulfonate (Cs-CHOS) was added to promote microsome fusion. After microsome fusion (marked by the sudden appearance of single-channel activity), the cis-solution was exchanged for the cytosolic recording solution. Recording solutions were prepared following recipes generated by the MaxChelator program. Free Ca was the recording solutions that were verified using a Ca-electrode. The luminal recording solution contained 200 mM Cs-HEPES, 1 mM Ca, and pH 7.4. The cytosolic recording solution contained 114 mMTris, 250 mM HEPES, 0.5 mM EGTA, 5 mM total ATP, 1 mM free Mg, pH 7.2, and various free Ca levels. These solutions minimize/eliminate currents carried by these other non-RyR2 channels that may be present in the heavy SR vesicles. The possible interference of SR Cl channel currents is eliminated by utilizing the impermeable monovalent anion substitute, HEPES. The possible interference of SR K channel currents is minimized by using a very poorly conducting monovalent cation substitute, cesium. No exogenous RyR2 protein partners were applied to single RyR2 channels after they were incorporated into the planar bilayer. Results are presented as means ± standard error (or standard deviation) with the number of determinations specified. These values were statistically compared using an unpaired two-tailed T test. Verification of statistically significant (p < 0.05) outliers was done using the Grubbs test.

RESULTS

Delayed after depolarizations (DADs) are an arrhythmogenic threat that can arise from abnormal diastolic Ca handling in cardiomyocytes (Liu et al. 2015). We first assessed electrical-pacing-induced DADs utilizing patch clamp in freshly isolated myocytes as previously described (Yan et al. 2021). Specifically, we counted DADs in isolated mouse and rabbit myocytes for a period of 2 s starting immediately after a 13 s train of stimulation at the specified frequencies as illustrated in Fig. 1A. Figure 1B shows an example confocal image of a Ca indicator (Fluo4)-loaded mouse myocyte with characteristic morphological features. In this myocyte, intracellular Ca transientswere measured in line-scan mode as shown in Fig. 1C. The minimum (threshold) electrical-pacing frequency after which DADs occurred was determined, and the results are shown in Fig. 1D. The average electrical-pacing frequency threshold evoking DADs in mouse myocytes is significantly lower than that in rabbit myocytes. In other words, DAD inducibility is greater in the mouse cells compared to rabbit cells. It is well-known abnormal diastolic RyR2-opening results in SR Ca leak that may prolong Ca transient decay and consequently trigger Ca waves and DADs by increasing electrogenic Na-Ca exchanger (NCX) activity (Ai et al. 2005; Bers 2000, 2014; Respress et al. 2012). We, thus, measured tetracaine-sensitive diastolic SR Ca leak with NCX blocked (Yan et al. 2021; Yan et al. 2018). Interestingly, mouse myocytes have significantly greater tetracaine-sensitive SR Ca leak than in the rabbit cells (Fig. 1E). The greater tetracaine-sensitive SR Ca leak in mouse is consistent with the greater DAD propensity we observed in the mouse cells. The SR Ca load within these mouse and rabbit myocytes is compared in Fig. 1F. Mouse myocytes have a significantly higher SR Ca 2+loadcompared to that in the rabbit. Since SR Ca leak is higher in mouse, the higher SR Ca load in mouse is not due to reduced SR Ca leak but rather something else (perhaps SR Ca pump up regulation). In addition, we measured the peak amplitude of electrically stimulated cytosolic Ca transients, and those data are shown in Fig. 1G. Calcium transients are significantly larger in the mouse myocytes compared to those in rabbit cells. The higher SR Ca load in mouse explains the larger cytosolic Ca transient in mouse. Overall, these cellular results illustrate existence of some substantial intracellular Ca handling differences in cardiomyocytes across species. As mammals have the similar Ca handling proteins (e.g., RyR2, CaV1.1, SR Ca pump, etc.), this dissimilarity seems likely to stem from cross-species differences in the regulation of intracellular Ca homeostasis.

Fig. 1. Cellular Ca handling differences between mouse and rabbit myocytes.

A. Example mouse myocyte patch-clamp membrane potential recording of action potentials (upward spikes). This myocyte was paced 1 Hz for 13 s and then pacing stimuli was stopped at upward arrow. Post-pacing-induced delayed after depolarizations (DADs; marked) were recorded for a period of 2 s. B. Example confocal Ca 2D image if aFluo4-loaded (green) mouse ventricular myocyte. Note intact morphology of the cell. C. Line scan imaging showing action potential evoked Ca transient at 0.5 Hz. D. Electrical stimulus frequency at which DADs were first observed in mouse (green) and rabbit (blue) myocytes. Small circles are individual determinations. Bars are average ± SEM. Statistical p values of p < 0.005 and < 0.05 are denoted as ** and *, respectively. E. Tetracaine-sensitive diastolic SR Ca leak level in mouse and rabbit myocytes. ΔF and Fo are the change in fluorescence and the baseline fluorescence, respectively. F. SR Ca load level in mouse and rabbit myocytes measured using an intra-SR fluorescent indicator. G. Peak intracellular Ca transient amplitude evoked by the action potential in mouse and rabbit myocytes.

Single-RyR2 Channel Cytosolic Ca Sensitivity

These intracellular regulatory differences may accompany could arise from individual Ca-handling components, like the RyR2 channel complex, when they retained when RyR2s are isolated in heavy SR microsomes and studied in vitro. We measured and compared the function of single RyR2 channels isolated from mouse, rat, rabbit, and human ventricular muscle. In all four cases, we utilized exactly the same SR microsome isolation protocol, single-channel approach, recording solutions, and analysis procedures well known to effectively isolate RyR2 channels (Chen et al. 2013; Fill & Copello 2002; Guo et al. 2013; Porta et al. 2011; Tu et al. 1994; Yan et al. 2018a, b, c).The characteristic modification of channel gating by ryanodol (Ramos-Franco et al. 2010) was used to positively identify the channels after their incorporation into the bilayer (see example in Fig. 1S). Figure 2A compares single RyR2 Po results when plotted as a function of cytosolic-free Ca concentration. The thick lines are Hill fits to the single rat (red) and human (black) RyR2 results. The Ca EC50sare 9.3 and 22.4 for the rat and human results, respectively. These two Hill fits are not statistically different (F-test; significance level0.05). The mouse (green) and rabbit (blue) results were also fit (not shown), and those fits were also not statistically different (compared to the human fit). Figure 2B shows that there was no significant difference in Po of single mouse, rat, rabbit, and human RyR2 channels at 10 μM cytosolic-free Ca2+, which is near their cytosolic Ca EC50. The average Po of single rat, mouse, and rabbit RyR2s are not statistically different compared to that of human RyR2s (0.37 ± 0.06, mean ± SEM). Likewise, there was no significant cross-species difference in RyR2 unitary current amplitude (not shown). Figure 2C shows example recordings of single rat (top), mouse, rabbit, and human (bottom) RyR2 channel activity with 10 μM cytosolic Ca present. However, these single-RyR2 channel recordings appear(visually) different across species. Long RyR2 openings seem to be more frequent in our rat and mouse recordings compared to rabbit and human recordings. In other words, the rabbit and human RyR2s seem to open and close more frequently. Single RyR2-opening frequency s compared across species in supplemental Figure S2A, and indeed, there are significant differences across species. This sparked our interest in further quantifying single RyR2 function across species.

Fig. 2. Single RyR2 cytosolic Ca sensitivity is similar across species.

A. Single RyR2 open probability plotted as a function of cytosolic-free Ca concentration. Rat, mouse, rabbit, and human RyR2 results are shown. Each data point is an average (± SEM) of 5 to 20 determinations. Thick red and black lines are Hill fits of the rat and human data points. Dashed red lines are the upper and lower 95% prediction interval of the rat fit. The two Hill fits were not statistically different (F-test; significance level 0.05). The rat, mouse, rabbit, and human cytosolic Ca EC50s are 5.5, 6.2, 6.6, and 11.1 μM, respectively. The corresponding Hill coefficients are 1.2, 1.4, 1.3, and 1.1. B. Single RyR2 open probability at 10 μM cytosolic-free Ca 2+is compared between species. Small circles are individual determinations. Bars are average ± SEM. The Po was not statistically different across species. C. Sample rat, mouse, rabbit, and human single RyR2 channel recordings at 10 μM cytosolic free (+ 40 mV). Openings are upward deflections from the marked zero current level. Scale bars are 2000 ms and 4 pA.

Single RyR2 Channel Mean Open Time

Figure 3A compares the mean open time (MOT) of single rat (red), mouse (green), rabbit (blue) and human (black) RyR2s at 10 μM cytosolic-free Ca, a systole-like intracellular Ca level. The MOTs at 10 μM Ca are 13.56 ± 1.87, 9.20 ± 1.34, 3.26 ± 0.32, and 2.05 ± 0.21 ms, respectively. The rat and mouse RyR2 MOTs are not statistically different, but these values are statistically (p < 0.005) larger than the MOTs of either rabbit or human channels. The significantly shorter MOTs of rabbit and human RyR2s (compared to rat and mouse) are consistent with the qualitative “flickering” like behavior displayed in the sample recordings shown in Fig. 2C. A similar pattern of MOT differences was also found at 1 μM cytosolic-free Ca 2+(Fig. 3B). At 1 μM Ca, the MOTs are 9.35 ± 1.92, 5.92 ± 1.22,2.14 ± 0.23, and 1.34 ± 0.14 ms, respectively. At the diastole-like intracellular-free Ca level of 100 nM, significant cross-species differences in RyR2 MOT exists as well (Fig. 3C). The MOTs are 2.44 ± 0.42, 2.07 ± 0.49, 1.16 ± 0.08, and 1.32 ± 0.12 ms (respectively)at 100 nM Ca. The associated single RyR2 mean closed times (MCT) at each of these cytosolic Ca levels are shown in supplemental Figure S2B. Figure 3D presents single rat, mouse, rabbit, and human RyR2 MOTs as a function of cytosolic Ca2+. The cytosolic Ca2+dependencies of rabbit and human RyR2s are relatively shallow while those of rat and mouse RyR2s are considerably steeper. Overall, our MOT results reveals that there are clear and significant differences across species. Single RyR2s isolated from rat or mouse on average stay open longer than those isolated from rabbit or human. Single rabbit RyR2 openings were the most human like.

Fig. 3. Single RyR2 Mean Open Time (MOT) Varies Across Species.

Single rat (red), mouse (green), rabbit (blue), and human (gray) RyR2 mean open times were measured with 10 μM (A), 1 μM (B), and 100 nM (C) cytosolic-free Ca present. Small circles are individual determinations, and the bar is their average (± SEM). Statistical p values of < 0.005, < 0.05, and not significant are denoted as **, *, and ns, respectively. D. Mean open times replotted as function of cytosolic free [Ca].

Diastolic Single RyR2 Channel Function

In cardiomyocytes, single RyR2s are largely closed during diastole but can spontaneously open albeit very infrequently. Figure 4A shows example 2-s-long recordings of single rabbit RyR2 function in our diastole-like recording solution (100 nM cytosolic Ca). Four openings are marked (a, b, c & d) and shown on an expanded scale at the bottom. Note events (openings & closings) lasting less than 0.5 ms (like the event marked “a”) were excluded from our analysis (see methods). Opening events are typically infrequent (0.41 ± 0.14 Hz; see Figure S2A) and their MOT brief (1.16 ± 0.08 ms; see Fig. 3C). Figure 4B compares single “diastolic” RyR2 Po across species. The Po of rat and mouse RyR2s (0.0048 ± 0.0012 or 0.0022 ± 0.0004) are not significantly different. The Po of rabbit and humanRyR2s are also similar (0.0011 ± 0.0003 vs 0.0016 ± 0.0003). However, the Po of rabbit RyR2s is significantly different (p < 0.005) compared to either the rat or mouse RyR2 Po’s. Figure 4C compares the open dwell time distributions of RyR2s across species at 100 nM cytosolic Ca. The rat and human open time histograms (normalized) are the red and gray bars, respectively. The lines are single exponential function curve fits. The mouse and rabbit open time histograms are represented by their curve fits (dashed & dotted lines, respectively). The time constant of the fits to the rat, mouse, rabbit, and human histograms are 0.85, 0.51, 0.55, and 0.36 ms (respectively). The inset is an expansion of the same plot to display the open time distributions for openings lasting longer than 6 ms. In a previous study (Porta et al. 2011), we predicted a spontaneous diastolic single RyR2 opening would likely need to last longer than 6 ms in order to trigger a Ca spark. Thus, we quantified the frequency of RyR2 openings longer than 6 ms (FREQ > 6 ms). Figure 4D compares the FREQ > 6 ms of single rat, mouse, rabbit, and human RyR2s in our diastole-like (100 nM free Ca) solution. The FREQ > 6 ms values are 0.042 ± 0.008, 0.034 ± 0.007, 0.004 ± 0.002, and 0.007 ± 0.002 ms, respectively. The rat and mouse RyR2 FREQ > 6 ms are similar (not significantly different). Likewise, the rabbit and human FREQ > 6 ms values are similar. However, the rat/mouse RyR2 FREQ > 6 ms values are significantly higher (p < 0.0005) than either the rabbit or human values.

Fig. 4. In Vitro “Diastolic” Single RyR2 Function Varies Across Species.

Single RyR2 channel activity was collected with 100 nM cytosolic-free Ca present. A. Sample 2 s long recordings (top) from the same single rabbit RyR2 experiment. Openings are upward deflections from the marked zero current level. Labelled openings (a, b, c, d) are shown on an expanded scale below. B. Single rat (red), mouse (green), rabbit(blue), and human (black) RyR2 open probability compared across species. Small circles are individual determinations. Bars are average ± SEM. Statistical p values of < 0.0005, < 0.005, < 0.05, and not significant are denoted as ***, **, *, and ns, respectively. C. Normalized open time rat and human histograms (red and grey bars, respectively). These dwell time histograms were fit by a single exponential function. The rat, mouse, rabbit, and rabbit fits are shown. Mouse and rabbit fits are the dashed and dotted lines, respectively (mouse & rabbit bars not shown). Inset shows the same set of histograms and fits for openings longer than 6 ms. D. Frequency of single RyR2 openings longer than 6ms are compared across species.

DISSCUSSION

We isolated RyR2-rich heavy SR microsomes from mouse, rat, rabbit, and human RyR2s and incorporated them into planar lipid bilayers inexactly the same way then compared their single RyR2 function in identical solutions. The single RyR2 cytosolic Ca sensitivity of Po was similar across species. However, channel gating (opening/closing) was visibly different across these species, and this was associated with significant differences in single RyR2 MOTs in both systole- and diastole-like recording conditions. This is important because these model animals are commonly utilized in cardiovascular research. We propose single rabbit RyR2 channel studies represent a helpful interpretive bridge between single RyR2 experiments that exploit the mechanism defining power of model mice and those seeking to apply that knowledge to RyR2 function in the human heart. We hope the species selection of future single RyR2 channel studies will be improved by our findings here.

Factors that Might Contribute to the Observed Differences

There are many possible factors that might contribute to the in vitro single RyR2 functional differences between species that are quantified here. One possibility is that the human muscle samples were more deteriorated or less healthy compared to the model animal samples. However, our human samples were procured using a protocol that was optimized to maintain viable tissue, which has been experimentally verified (Yan et al. 2021; Yan et al. 2018a,b,c). Moreover, this is unlikely to be a substantial factor because human and rabbit RyR2s had very similar functional attributes (MOT, Po, FREQ > 6 ms) at 100 nM cytosolic Ca which suggests the human RyR2s were not grossly abnormal. Another possibility is RyR2 structural differences between these mammalian species. This is also not likely a substantial factor because mouse, rat, rabbit, and human RyR2 amino acid sequences are 97% identical and even more so (up to 99%) when amino acid attributes are consideration considered. We believe the observed single RyR2 functional differences in vitro across mammalian species stem from the species-specific differences in the compliment of regulatory factors and protein partners which govern RyR2 function in cells. In other words, the RyR2s tested here were likely co-isolated with a subset of those factors and partners that reflects those present in the source tissue. The known compliment of possible RyR2-intracting regulatory factors and protein partners in cells is large. This includes small molecules (like Ca, Mg, ATP, cADPR, adenosine; (Fill & Copello,2002) and various posttranslational modifications (Meissner 2010). It includes kinases (CaMKII, PKA, JNK2; Currie et al. 2004; Marx et al. 2000; Yan et al. 2018a,b,c), phosphatases (PP1, PP1A; Marx et al. 2000), FKBP (Kaftan et al. 1996), sorcin (Lokuta et al. 1997; Meyers et al. 1995), homer (Pouliquin & Dulhunty 2009), calmodulin (Balshaw et al. 2001; Søndergaard et al. 2019), S100A1 (Rebbeck et al. 2016), triadin, and junction (Györke et al. 2004) as well as calseqeustrin (CSQ; Chen et al. 2013; Györke et al. 2004; Qin et al. 2008, 2009). RyR2 function in cells is subject to the actions of all these regulatory factors and protein partners. Some of these protein partners (e.g., CaMKII, PKA, PP1, PP1A) are anchored to RyR2(Camors & Valdivia 2014) while others (e.g., calmodulin) have more labile associations with the RyR2 protein (Meissner et al. 2009). Thus, some will be co-isolated with the RyR2 and others not. Function of isolated RyR2s will logically be influenced by the subset of these regulatory factors and protein partners present. In heart, RyR2-associated intracellular Ca handling is known to vary between species (Bers 2001,2008) so isolation of native RyR2-rich heavy SR microsomes logically may capture elements of this species regulatory specialization. This study confirms the existence of significant cross-species differences in single RyR2 function in vitro in identical solutions and after being isolated in exactly the same way. We believe that this is most likely because the rat, mouse, rabbit, and human RyR2s in the isolated heavy SR microsomes are associated with different subsets of endogenous regulatory factors and protein partners. This implies single rabbit and human RyR2s isolated within heavy SR microsomes retain similar compliments of these regulatory factors and protein partners because their single-channel function in vitro were most similar, particularly in diastolic recording conditions.

Single RyR2 Channel Function in a Diastole-like Solution

This study not only quantifies uniquely single RyR2 function across species but also does so in realistic diastole-like recording solutions. Specifically, diastolic single RyR2 function was quantified with 100 nM free cytosolic-free Ca, 1 mM cytosolic-free Mg, 5 mM cytosolic total ATP, and 1 mM luminal-free Ca present. Single RyR2 Po ranged from 0.001 to 0.005 (i.e., a tenth to a half of one percent) and MOT from 2.5 to 1.2 ms. The FREQ > 6 ms (frequency of long openings > 6 ms) was between 0.05 and 0.004 Hz (i.e., one every 20 to250 s). Importantly, these single RyR2 functional attributes are consistent with normal diastolic RyR2 activity in cells (Bers 2014). These diastolic attributes have rarely been quantified in vitro before because use of such realistic diastole-like recording solutions introduces certain challenges. One of these is simply the rarity of RyR2 openings and very low Po which necessitates making unusually long recordings. Another stems from the multi-ion permeability and lack of Ca/Mg selectivity of the RyR2 pore (Gillespie et al. 2012; Mejía-Alvarez et al. 1999) that makes the RyR2’s unitary net current relatively small in these solutions which reduces the signal-to-noise ratio.

Possible Implications

We quantified diastolic single RyR2 functional metrics that correspond to the two different forms of RyR2-mediated SR Ca leak in cardiomyocytes, spark, and non-spark leak (Bers, 2014; Porta et al. 2011; Zima et al. 2010). Non-spark leak in cells is mediated by all the spontaneous diastolic RyR2 openings that do not evoke sparks (Zima et al. 2010). Thus, diastolic single RyR2 Po likely best correlates to non-spark SR Ca leak in cells, and our diastolic single RyR2 results suggest non-spark leak ought to be 1.4 to 4.4 times larger in rat/mouse compared to rabbit or human cells. This is consistent with our cellular results that show overall tetracaine-sensitive SR Ca leak is about 1.5 times larger mouse compared to rabbit. Abnormally large or frequent diastolic sparks may evoke waves and arrhythmogenic DADs. Our cellular results show that DAD propensity (threshold) is about 1.7-fold greater in mouse compared to rabbit. In a previous study (Porta et al. 2011), we concluded that sparks are likely only trigged by spontaneous diastolic single RyR2 openings lasting > 6 ms. Thus, the diastolic FREQ > 6 ms single RyR2 metric best reflects spark (and thus, DAD) propensity in cells. We show here diastolic FREQ > 6 ms is 4.8- to 10.5-fold greater in rat/mouse compared to rabbit or human cells. Note that one would expect the FREQ >6 ms difference to exceed the DAD propensity difference because not every long (> 6 ms) RyR2 opening triggers a spark and not every spark evokes a DAD.

Single RyR2 regulation is differentially tuned in cells to optimize the species-specific demands of intracellular Ca handling in cardiomyocytes. We propose elements of that species-specific regulatory tuning are retained when RyR2-rich SR microsomes are isolated, explaining the single RyR2 functional variability across species (mouse, rat, rabbit, and human) reported here.