Sevoflurane Protects Ventricular Myocytes against Oxidative Stress-induced Cellular Ca²⁺ Overload and Hypercontracture

IT is clear that, if the ischemic myocardium is ultimately to survive, coronary blood flow must be restored promptly. However, reperfusion of the ischemic myocardium is not entirely benign and can initiate additional cellular damage beyond the injury inflicted by the ischemia itself. This phenomenon is known as myocardial reperfusion injury.¹  Although the pathogenesis of myocardial reperfusion injury is multifactorial, there is good evidence that reactive oxygen species (ROS), which are produced during early reperfusion period,^2–4  play a major role in the development of reperfusion injury, characterized by myocardial Ca²⁺ overload and hypercontracture.^1,4–6 

ROS can modify a number of cardiac ion channels and transporters either through direct oxidation, or through the activation of signaling molecules, such as Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), and disrupt the cellular Ca²⁺ homeostasis to cause excess Ca²⁺ loading in cardiac myocytes.^7–9  For example, ROS including H₂O₂ stimulate the cardiac ryanodine receptor Ca²⁺-release channel (RyR2) in sarcoplasmic reticulum (SR).^8,10,11  In addition, ROS enhance the cardiac Na⁺/Ca²⁺ exchange (NCX) activity in sarcolemma, which would alter cellular Ca²⁺ regulation and electrical activity in cardiac myocytes.^8,12  Whereas ROS were previously shown to have diverse effects on the L-type Ca²⁺ current (I_Ca,L) in ventricular myocytes,^8,13,14  a more recent study found that H₂O₂ substantially potentiates I_Ca,L, primarily through the activation of CaMKII.¹⁵  These functional alterations of the RyR2, NCX and/or I_Ca,L by oxidative stress likely contribute to the impaired cellular Ca²⁺ homeostasis associated with myocardial reperfusion injury.^6,8 

Volatile anesthetics, including isoflurane, sevoflurane, and desflurane, provide cardioprotection against reperfusion injury, when administered before ischemia (anesthetic preconditioning) or even during reperfusion (anesthetic postconditioning).^16–20  Although these beneficial actions of volatile anesthetics appear to be predominantly mediated through attenuation of cellular Ca²⁺ overload,^18,20  its precise ionic basis remains to be fully elucidated. Our recent study found that sevoflurane protects ventricular myocytes against Ca²⁺ paradox-mediated Ca²⁺ overload, an important experimental cellular model for reperfusion injury,^6,21,22  by preventing excess Ca²⁺ loading through inhibition of both Ca²⁺ influx via transient receptor potential canonical channels in sarcolemma, and Ca²⁺ release from the SR through the RyR2.²³  In addition, sevoflurane potently inhibits the store-operated Ca²⁺ entry that has been suggested to mediate Ca²⁺ overload during ischemia/reperfusion in the heart.^24,25  These observations suggest the possibility that Ca²⁺ channels and transporters could be the targets through which volatile anesthetics exert their preventive action against myocardial injury. To date, few studies have investigated the effects of volatile anesthetics on Ca²⁺ transport mechanisms during oxidative stress. However, as judged from their cardioprotective effects against myocardial reperfusion injury,^18,20  volatile anesthetics should reasonably be expected to correct the possible functional alteration of Ca²⁺ channels and transporters and to restore the impaired Ca²⁺ homeostasis during oxidative stress.

In this study, we exogenously applied H₂O₂ to mouse ventricular myocytes, which has been used to mimic ROS exposure in reperfused myocardium in vivo,^5,7,12,14  and found that clinically used concentrations of sevoflurane almost completely protected ventricular myocytes against H₂O₂-induced cellular Ca²⁺ overload and hypercontracture, by correcting electrophysiological abnormalities associated with cellular Ca²⁺ handling.

All animal care and experimental procedures complied with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996) and were approved by the institutional Animal Care and Use Committee of Shiga University of Medical Science (Otsu, Shiga, Japan; No. 2011-2-2). Single ventricular myocytes were enzymatically isolated from the hearts of 7- to 12-week-old male C57BL/6J mice (body weight, 20–25 g) and stored at 37°C in normal Tyrode solution, containing NaCl, 140 mm; KCl, 5.4 mm; CaCl₂, 1.8 mm; MgCl₂, 0.5 mm; NaH₂PO₄, 0.33 mm; glucose, 5.5 mm; and HEPES, 5 mm (pH adjusted to 7.4 with NaOH), supplemented with 2% bovine serum albumin and antibiotics (penicillin/streptomycin). The full details of the cell isolation procedure have been described previously.^22,23,26  The present isolation procedure resulted in a high yield (70–80%) of rod-shaped quiescent ventricular myocytes,^22,23,26  with normal electrophysiological and contractile properties.²⁶  The myocytes were used within 8 h of isolation.

Fluorescence imaging of intracellular Ca²⁺ was performed using a confocal laser scanning microscope, as described previously.^22–24  An aliquot of fluo-3 acetoxymethyl ester (Dojin Chemicals, Kumamoto, Japan)-loaded ventricular myocytes was transferred to a recording chamber (0.5 ml in volume) mounted on the stage of an Eclipse TE2000-E inverted microscope, which was equipped with a C1si spectral imaging confocal laser scanning system (Nikon, Tokyo, Japan). The chamber was continuously superfused at a constant rate of 2 ml/min with normal Tyrode solution at room temperature (23°–25°C). The myocytes were excited with an argon laser beam (wavelength 488 nm) at 30-s intervals, and data were collected for the emission intensity at a wavelength of 515 nm through a ×10 objective lens. In most experiments, 15–20 rod-shaped viable myocytes showing fluo-3 fluorescence were observed within a single field of view. Fluo-3 fluorescence images were analyzed frame by frame using the Nikon EZ-C1 software to calculate the average intensity in each cell, which was used to estimate the cytosolic Ca²⁺ levels. The fluo-3 fluorescence intensities were expressed as arbitrary units. The length/width ratio was also measured in each myocyte image, and a decrease in the length/width ratio of less than 2 was defined to indicate a hypercontracted or dead cell.²⁷  Because ROS have been detected in micromolar concentrations in both in vivo and in vitro ischemia-reperfused myocardium,^3,4,7  we examined the effects of H₂O₂ at concentrations in the range of 1–100 μm on fluo-3 fluorescence intensity and cell morphology for 15 min, in each experiment. The percentage of hypercontracted myocytes, accompanied by the elevation of fluo-3 fluorescence intensity, was 16.3 ± 5.4% (the number of experiments [n] = 5, the number of cell isolations [animals, N] = 3), 26.5 ± 10.6% (n = 5, N = 3), and 56.8 ± 24.5% (n = 11, N = 4) after 15 min exposure to 1, 10, and 100 μm H₂O₂, respectively. In the present experiments, we investigated the protective effects of volatile anesthetics and other agents against cellular Ca²⁺ overload during relatively potent oxidative stress imposed by 100 μm H₂O₂.²⁸  In the confocal Ca²⁺ imaging experiments, mouse ventricular myocytes were not electrically stimulated, because myocytes failed to respond to electrical stimulation during superfusion with H₂O₂.⁷ 

Perforated and conventional (ruptured) whole-cell patch-clamp techniques were used to record the membrane potentials and I_Ca,L in the current- and voltage clamp modes, respectively.^29,30  An EPC-8 patch-clamp amplifier (HEKA, Lambrecht, Germany) was used for these recordings. Fire-polished pipettes pulled from borosilicate glass capillaries (Narishige Scientific Instrument Lab., Tokyo, Japan) had a resistance of 2.0–3.5 MΩ, when filled with the pipette solution. An aliquot of isolated ventricular myocytes was transferred to a recording chamber (0.5 ml in volume) mounted on the stage of an inverted microscope (ECLIPSE TE-2000U; Nikon) and was continuously superfused at a constant rate of 2 ml/min with normal Tyrode solution at 35°–37°C.

Membrane potentials (including action potentials) were recorded using the amphotericin B-perforated patch-clamp method with a pipette solution containing potassium aspartate, 70 mm; KCl, 50 mm; KH₂PO₄, 10 mm; MgSO₄, 1 mm; and HEPES, 5 mm (pH adjusted to 7.2 with KOH). Amphotericin B (100 μg/ml; Wako Pure Chemical Industries, Osaka, Japan) was added to the pipette solution just before use, and the measurements were started 10–20 min after giga-seal formation. Action potentials were evoked every 5 s by applying suprathreshold current pulses of 2–5 ms duration via the patch pipette. The action potential duration was measured at 90% repolarization level (APD₉₀). In some experiments, ventricular myocytes were not electrically stimulated.

I_Ca,L was recorded by conventional (ruptured) patch-clamp method with a Cs⁺-rich pipette solution containing cesium aspartate, 90 mm; CsCl, 30 mm; tetraethylammonium chloride, 20 mm; MgCl₂, 2 mm; adenosine 5′-triphosphate, 5 mm (Mg salt; Sigma Chemical Company, St. Louis, MO); phosphocreatine, 5 mm (disodium salt; Sigma); guanosine 5′-triphosphate, 0.1 mm (dilithium salt; Roche Diagnostics GmbH, Mannheim, Germany); EGTA, 0.1 mm; and HEPES, 5 mm (pH adjusted to 7.2 with CsOH). In some experiments, a highly specific CaMKII inhibitor autocamtide-2-related inhibitory peptide (AIP; Calbiochem, San Diego, CA) was added to the pipette solution to dialyze the cell inside through a ruptured patch. Approximately 10–15 min were allowed to elapse to ensure an adequate inhibition of CaMKII by AIP before conducting the measurements of I_Ca,L.³¹  A concentrated stock solution of AIP (1 mm) was prepared in distilled water. The bath solution was a Cs⁺-Tyrode solution containing NaCl, 140 mm; CsCl, 5.4 mm; CaCl₂, 1.8 mm; MgCl₂, 0.5 mm; NaH₂PO₄, 0.33 mm; glucose, 5.5 mm; and HEPES, 5 mm (pH adjusted to 7.4 with NaOH). I_Ca,L was activated by 200-ms depolarizing steps applied from a holding potential of −40 mV to test potentials of −30 to +60 mV, and the amplitude of I_Ca,L was measured at the peak inward current level. The conductance for I_Ca,L (g_Ca,L) at each test potential was obtained by dividing the peak current amplitude by the driving force for Ca²⁺ and was fitted with a Boltzmann equation: g_Ca,L = g_Ca,L,max/(1 + exp[(V_h − V_t)/k]), where g_Ca,L,max is the fitted maximal conductance, V_h is the voltage at half-maximal activation, V_t is the test potential, and k is the slope factor. In some analyses, g_Ca,L at each test potential was normalized with reference to its maximal value (g_Ca,L,max). The late current level of I_Ca,L was also measured near the end of 200-ms depolarizing test steps. The voltage-clamp protocols and data acquisition were controlled with PATCHMASTER software (HEKA), and current records were low-pass filtered at 1 kHz, digitized at 2 kHz through an LIH-1600 interface (HEKA), and stored on a Macintosh computer (Apple Inc., Cupertino, CA). Current amplitude was presented as current density (in pA/pF), obtained by normalizing with reference to cell membrane capacitance. The zero-current and zero-potential levels are denoted by dashed lines.

Various test agents were added to the normal Tyrode solution for fluorescence Ca²⁺ imaging and/or patch-clamp experiments. These were H₂O₂ (Wako Pure Chemical Industries), I_Ca,L blocker nifedipine (Sigma), RyR2 blocker tetracaine (Sigma), NCX blocker SEA0400 (Taisho Pharmaceutical Co., Ltd., Saitama, Japan), adenosine A₁ receptor antagonist 8-cyclopentyl-1,3-dipropylxanthine (Sigma), α-adrenergic receptor antagonist prazosin (Sigma), β-adrenergic receptor antagonist propranolol (Sigma), protein kinase C inhibitor chelerythrine (Sigma), phosphatidylinositol-3-kinase inhibitor wortmannin (Sigma), mitogen-activated protein kinase-extracellular signal-regulated kinase-1 inhibitor PD098059 (Sigma), endothelial nitric oxide synthase inhibitor N-nitro-l-arginine methyl ester (Sigma), mitochondrial adenosine triphosphate-sensitive K⁺ channel blocker 5-hydroxydecanoic acid (Sigma), and large-conductance Ca²⁺-activated K⁺ channel blocker paxilline (Sigma). Sevoflurane (Abbott Laboratories, North Chicago, IL), isoflurane (Abbott Laboratories), or desflurane (Baxter, Deerfield, IL) was equilibrated in bathing solutions in a reservoir by passing air (flow rate, 0.5 l/min) through the respective vaporizers for at least 15 min before entering a recording chamber for both fluorescence Ca²⁺ imaging and patch clamp experiments.

Ventricular myocytes were incubated at 37°C for 10 min in normal Tyrode solution without or with the addition of H₂O₂ (100 μm) and/or sevoflurane (3%). Ventricular myocytes were then harvested and lysed in Tris buffer (pH = 7.5) containing Tris-HCl, 50 mm; NaCl, 150 mm; EDTA, 5 mm; and 1% Triton X-100 supplemented with protease and phosphatase inhibitors (Complete Mini; Roche). Protein concentration was measured by the Lowry method (DC protein assay; Bio-Rad, Richmond, CA). The protein samples were dissolved in NuPAGE LDS sample buffer (Invitrogen, Carlsbad, CA) supplemented with 2% 2-mercaptoethanol. These reduced proteins were separated on 4–12% sodium dodecylsulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes (Bio-Rad). The polyvinylidene difluoride membrane was blocked with 5% bovine serum albumin in Tris-buffered saline (Tris-HCl, 10 mm and NaCl, 100 mm; pH = 7.5) supplemented with 0.1% Tween-20, and then incubated with primary antibodies against CaMKII (polyclonal rabbit, 1:700; Abcam, Cambridge, United Kingdom), phosphorylated CaMKII at Thr 287 (polyclonal rabbit, 1:500; Cell Signaling Technology, Beverly, MA), and glyceraldehyde-3-phosphate dehydrogenase (polyclonal rabbit, 1:5,000; Cell Signaling Technology), respectively. The immunosignals were detected using the ECL Plus Western blotting detection system (GE Healthcare, Buckinghamshire, United Kingdom) and were analyzed with Lumino-image analyzer (LAS-4000; Fujifilm, Tokyo, Japan). The band intensity for Thr287-phosphorylated CaMKII was normalized to that of CaMKII in the same sample, and this relative band intensity (Thr287-phosphorylated CaMKII/CaMKII) was used to estimate the CaMKII activity.³² 

The effects of experimental protocols and animals on the data were initially examined by using the two-way layout ANOVA (GLM procedure; SAS 9.1.3; SAS Institute, Inc., Cary, NC). This statistical analysis demonstrates that the mean square for animals was much smaller than the mean square for experimental protocol, which indicates that the effect of animals on the data is negligibly small in the present experiments. Results are expressed as the means ± SD, whereas n and N represent the number of experiments and the number of cell isolations (animals), respectively. One to three experiments (n) were conducted from one cell isolation (animal, N) for a given experimental protocol. The error bars in the figures indicate SD with n given in parentheses. The statistical power calculation was carried out for each experimental protocol, using the StatMate version 2.0 (GraphPad Software, La Jolla, CA), with a statistical power of 0.8 at a significance level (α) of 0.05 (two-tailed). This showed that a group size of five would allow detection of a difference of 50% between group means in the percentage fraction of hypercontracted myocytes in fluo-3 fluorescence imaging experiments. However, a group size of four would allow detection of a difference of 25% between group means in the amplitude of I_Ca,L in patch-clamp experiments. In Western blot analysis, a group size of five was required to detect a difference of 30% between group means in the intensity ratio of Thr287-phosphorylated CaMKII/CaMKII. Statistical comparisons were performed using either one-way ANOVA or repeated measure ANOVA, as appropriate, which was followed by Dunnett or Tukey test. Fisher exact test was used to compare the fraction of myocytes that showed abnormal action potentials during exposure to H₂O₂ without and with SEA0400 or sevoflurane. All statistical analyses were conducted, using either the SAS 9.1.3 or GraphPad Prism 5 (GraphPad Software). We used two-tailed hypothesis testing for all tests. A P value less than 0.05 was considered statistically significant.

We first examined the effects of exposure to H₂O₂ on cytosolic Ca²⁺ levels and cell morphology without and with the addition of sevoflurane, by assessing confocal images of fluo-3 fluorescence in mouse ventricular myocytes (fig. 1). In control experiment, there were 16 rod-shaped viable myocytes detected before exposure to 100 μm H₂O₂ (fig. 1A, Baseline), and nine out of the 16 myocytes (56.3%) eventually underwent Ca²⁺ overload accompanied by hypercontracture after a 15-min exposure to H₂O₂ (fig. 1A; 15 min), as evidenced by a marked elevation of fluo-3 fluorescence intensity (fig. 1Ca) along with a reduction of length/width ratio to less than 2 during exposure to H₂O₂ (fig. 1Cb). In time-matched control experiments, there were no significant changes in fluo-3 fluorescence intensity and cell morphology in confocal images of fluo-3-loaded ventricular myocytes without exposure to H₂O₂ (data not shown). In contrast, cellular hypercontracture associated with Ca²⁺ overload was not evoked in any of the 16 viable myocytes during the 15-min exposure to H₂O₂ by the concomitant administration of 3% sevoflurane (fig. 1, B and D).

As summarized in figure 2A, the percentage of myocytes that developed Ca²⁺ overload-mediated hypercontracture was progressively increased during exposure to H₂O₂ (100 μm) and reached 56.8 ± 24.5% (n = 11, N = 4) after 15-min exposure (black empty circles) in control condition. Sevoflurane at concentrations of 2% or more significantly reduced the percentage of hypercontracted myocytes at 10 and 15 min of H₂O₂ exposure. Figure 2B depicts the percentage of myocytes that hypercontracted during 15-min exposure to H₂O₂ as a function of sevoflurane concentrations, which was well described by a Hill equation with an IC₅₀ of 0.194 mm (which corresponds to 1.35%). Sevoflurane was thus, found to markedly protect ventricular myocytes against the H₂O₂-induced cellular Ca²⁺ overload accompanied by hypercontracture, in the range of clinically relevant concentrations.

We also examined the effects of isoflurane and desflurane on H₂O₂-induced Ca²⁺ overload and hypercontracture, using nearly equipotent concentrations with 3% sevoflurane in the mouse.^33,34  Two percent isoflurane and 10% desflurane reduced the occurrence of H₂O₂-induced Ca²⁺ overload accompanied by hypercontracture to 6.2 ± 9.4% (n = 6, N = 3) and 10.9 ± 5.0% (n = 6, N = 3), respectively, which are similar to the effects of 3% sevoflurane (4.2 ± 7.9%; n = 8, N = 3; see Supplemental Digital Content 1, fig. 1, https://links.lww.com/ALN/A926).

To elucidate the ionic mechanisms underlying the H₂O₂-induced cellular Ca²⁺ overload accompanied by hypercontracture, we investigated the effect of blockers for Ca²⁺ channels and transporters; namely, the I_Ca,L blocker nifedipine, the RyR2 blocker tetracaine,³⁵  and the NCX blocker SEA0400.^36,37  As demonstrated in figure 3, the percentage of myocytes that showed cytosolic Ca²⁺ elevation accompanied by hypercontracture after a 15-min exposure to H₂O₂ (100 μm) was partly but significantly decreased by nifedipine (10 μm), tetracaine (100 μm), and SEA0400 (1 μm), suggesting that I_Ca,L, RyR2, and NCX are all involved in mediating the H₂O₂-induced cellular Ca²⁺ overload.

Previous patch-clamp studies have demonstrated that H₂O₂ evokes both early and delayed afterdepolarizations in isolated ventricular myocytes.^15,38  Early afterdepolarizations (EADs) are secondary depolarizations that originate during the plateau or final repolarization of action potential, whereas delayed afterdepolarizations appear as small-amplitude voltage fluctuations at the resting potential after completion of action potential repolarization. These electrophysiological abnormalities seem to be linked to the H₂O₂-induced alterations of cellular Ca²⁺ homeostasis.

In the next series of experiments, we examined the effect of H₂O₂ (100 μm) on membrane potentials of mouse ventricular myocytes, using the amphotericin B-perforated patch-clamp technique (fig. 4), which minimizes the disruption of normal intracellular Ca²⁺ buffering mechanisms.³¹ Figure 4A shows a representative response of a quiescent (not electrically stimulated) ventricular myocyte to H₂O₂ (100 μm). The resting membrane potential averaged −77.9 ± 6.3 mV (n = 23, N = 12), and no appreciable voltage fluctuations were observed in the resting membrane potential under control conditions (fig. 4Aa). However, in most of the ventricular myocytes examined (10 out of 11 myocytes, 90.9%), a series of small-amplitude voltage fluctuations appeared in the resting membrane potential approximately 5–10 min after exposure to H₂O₂, and it is notable that irregular and repetitive action potentials were triggered by these voltage fluctuations in the resting potential (fig. 4, Ab, Ac, and D). Because ventricular myocytes subsequently exhibited repetitive abnormal action potentials, and eventually evolved to hypercontracture (data not shown), these electrophysiological abnormalities seem to be responsible for the H₂O₂-induced cellular Ca²⁺ overload accompanied by hypercontracture.

It is well documented that membrane voltage fluctuations typically result from the activation of NCX inward currents, triggered by spontaneous Ca²⁺ release from the SR.^39–41  We, therefore, examined the effect of H₂O₂ on membrane potentials in the presence of the NCX inhibitor SEA0400.^36,37  In these experiments, ventricular myocytes were pretreated with SEA0400 (1 μm) for approximately 8 min before exposure to H₂O₂ (100 μm). As expected, spontaneous voltage fluctuations and abnormal action potentials were abolished in six out of seven myocytes by the presence of SEA0400 (fig. 4, B and D), thus, supporting the involvement of NCX in membrane voltage fluctuations during H₂O₂ exposure. Sevoflurane (3%) also stabilized the resting potentials by preventing the occurrence of spontaneous voltage fluctuations in seven out of eight myocytes examined (fig. 4, C and D). It should be noted that membrane voltage fluctuations leading to abnormal automaticities appeared after washout of sevoflurane during the presence of H₂O₂ (fig. 4C). These observations suggest that sevoflurane, possibly by decreasing the amount of overloaded Ca²⁺ that must be eliminated from the cell, reversibly decreases the net depolarizing NCX current that underlies voltage fluctuations in resting potential leading to abnormal automaticity.

We also investigated the effect of H₂O₂ (100 μm) on action potentials evoked by electrical stimulation of 5-s intervals. As demonstrated in figure 5, action potential duration was abruptly prolonged after an approximately 8-min exposure to H₂O₂ (fig. 5B), which was accompanied by the occurrence of EADs (fig. 5Cb). With further exposure to H₂O₂, small-amplitude voltage fluctuations were induced in the resting membrane potential, which frequently triggered abnormal repetitive action potentials (fig. 5, Cc and Cd). Those myocytes that showed abnormal automaticities also eventually underwent hypercontracture during exposure to H₂O₂ for approximately 10 min or more (data not shown).

Previous investigators have demonstrated that I_Ca,L reactivation plays a predominant role in the genesis of EADs in ventricular myocytes during oxidative stress.^15,38  We then examined the effects of nifedipine and sevoflurane on the occurrence of H₂O₂-induced EADs (fig. 6). Figure 6A illustrates superimposed action potentials recorded before and during exposure to H₂O₂ (100 μm) and after further addition of nifedipine (10 μm). H₂O₂-induced EADs were almost completely abolished by nifedipine, supporting that I_Ca,L reactivation is involved in the EAD upstroke during exposure to H₂O₂. Sevoflurane (3%) also suppressed the occurrence of EADs during exposure to H₂O₂ (fig. 6C). Figure 6, B and D, demonstrates that H₂O₂ (100 μm)-induced abnormal action potentials were also prevented by subsequent addition of nifedipine (10 μm) and sevoflurane (3%), respectively. Thus, the inhibitory action of nifedipine and sevoflurane on EADs may contribute to their preventive actions against the abnormal automaticities during exposure to H₂O₂ (fig. 6, B and D).

We next examined the effects of H₂O₂ on I_Ca,L in the absence and presence of sevoflurane, using the conventional (ruptured) patch-clamp method. As demonstrated in figure 7A, H₂O₂ (100 μm) gradually increased the peak and late current amplitudes of I_Ca,L activated by 200-ms depolarizing steps to 0 mV, which were markedly reduced by the subsequent application of sevoflurane (3%). Figure 7B illustrates superimposed traces of I_Ca,L recorded at test potentials of −30 to +50 mV in control (left panel), during exposure to H₂O₂ (middle), and after further addition of sevoflurane (right). The peak amplitudes of I_Ca,L were significantly increased at test potentials between −20 to +30 mV by H₂O₂, and were significantly reduced to below the control levels at potentials of −10 to +40 mV by sevoflurane, even in the presence of H₂O₂ (fig. 7C). The effects of H₂O₂ (100 μm) and sevoflurane (3%) on I_Ca,L were analyzed by constructing the conductance (g_Ca,L)–voltage relationship fitted with Boltzmann equation (fig. 7D). H₂O₂ significantly increased the maximal g_Ca,L (g_Ca,L,max) from control value of 219.0 ± 20.0 to 268.0 ± 27.44 pS/pF, and subsequent administration of sevoflurane markedly decreased it to 141.2 ± 15.6 pS/pF (n = 5, N = 3). We also investigated the effects of isoflurane and desflurane on I_Ca,L in the presence of H₂O₂. Both 2% isoflurane and 10% desflurane decreased g_Ca,L,max potentiated by H₂O₂ (100 μm) to 57.9 ± 7.0% (n = 5, N = 3) and 56.2 ± 5.7% (n = 5, N = 3), respectively, which were similar to g_Ca,L,max reduction by 3% sevoflurane (51.0 ± 6.3%, n = 5, N = 3; see Supplemental Digital Content 1, fig. 2, https://links.lww.com/ALN/A926). H₂O₂ also slightly but significantly shifted V_h for I_Ca,L in a negative direction from −12.0 ± 4.3 to −17.0 ± 5.9 mV, and subsequent administration of sevoflurane (3%) largely reversed it to −13.0 ± 5.1 mV (n = 5, N = 3; fig. 7D, inset).

It has been demonstrated that H₂O₂-induced enhancement of I_Ca,L is mediated through the activation of CaMKII in ventricular myocytes.^9,15  To examine the role of CaMKII in the modulation of I_Ca,L by H₂O₂ and sevoflurane, we applied the specific CaMKII inhibitor AIP to the cell through a recording pipette.³¹ Figure 8A illustrates I–V relationships for peak I_Ca,L in myocytes dialyzed with AIP (1 μm), recorded before and during exposure to H₂O₂ (100 μm), and after further addition of sevoflurane (3%). In the presence of AIP, peak amplitudes of I_Ca,L were not appreciably affected by H₂O₂ but were significantly reduced at test potentials of −10 to +40 mV by subsequent addition of sevoflurane. As summarized in figure 8, B, C, and D, AIP abolished the H₂O₂-induced changes in g_Ca,L,max, late current amplitude and V_h of I_Ca,L, suggesting that CaMKII activation is involved in these biophysical changes in I_Ca,L during exposure to H₂O₂.

However, the effects of sevoflurane on late current amplitude and shift in V_h during exposure to H₂O₂ were largely abolished when CaMKII activity was inhibited by AIP (fig. 8, C and D). These observations suggest that sevoflurane reversed these H₂O₂-induced changes in I_Ca,L by inhibiting CaMKII activity. It should be noted, however, that sevoflurane significantly decreased the g_Ca,L,max during exposure to H₂O₂, despite the presence of AIP, suggesting that the inhibitory effect of sevoflurane on peak I_Ca,L also involves a CaMKII-independent mechanism (fig. 8B). Western blot analysis showed that the band intensity of Thr287-phosphorylated CaMKII (p-CaMKII) was increased by exposure to H₂O₂ (100 μm), which was largely and significantly reversed by concomitant administration of sevoflurane (fig. 8E).

It has been demonstrated that volatile anesthetics produce their protective action against myocardial ischemia/reperfusion injury by affecting several G-protein coupled cell surface receptors, downstream signaling molecules, and targets.^42,43  These include adenosine A₁ receptor, α-adrenergic receptor, β-adrenergic receptor, protein kinase C, phosphatidylinositol-3-kinase, mitogen-activated protein kinase-extracellular signal-regulated kinase-1, endothelial nitric oxide synthase, mitochondrial adenosine triphosphate-sensitive K⁺ channel, and large-conductance Ca²⁺-activated K⁺ channel. We then examined the possible involvement of these signaling components in the protective effect of sevoflurane against H₂O₂-induced cellular Ca²⁺ overload and hypercontracture, using their respective blockers. None of these blockers were able to affect the sevoflurane protection against H₂O₂-induced Ca²⁺ overload and hypercontracture (see Supplemental Digital Content 1, fig. 3, https://links.lww.com/ALN/A926). Therefore, the protective effects of sevoflurane against at least H₂O₂-induced Ca²⁺ overload and hypercontracture appears to occur independent of these receptors, signaling molecules, and targets.

The present experiments using confocal imaging of fluo-3 fluorescence in mouse ventricular myocytes demonstrate that exposure to H₂O₂ evoked severe hypercontracture due to cellular Ca²⁺ overload (fig. 1). There are a number of evidences that hypercontracture occurs during reperfusion of ischemic myocardium through a multifactorial mechanism, including ROS-induced cellular Ca²⁺ overload and that hypercontracture and its associated mechanical stress act as prominent causes of cardiomyocyte death during reperfusion.^5,6  It is, therefore, reasonable to assume that the H₂O₂-induced ventricular myocyte hypercontracture (fig. 1) represents an important pathomechanism of reperfusion-induced cardiomyocyte death. This study further found that clinically used concentrations of sevoflurane protected ventricular myocytes against the H₂O₂-induced hypercontracture by preventing cellular Ca²⁺ overload, in a concentration-dependent manner (IC₅₀ = 1.35%; fig. 2). Isoflurane and desflurane also protected ventricular myocytes against H₂O₂-induced cellular injuries to a degree similar to that produced by sevoflurane at nearly equipotent anesthetic concentrations (see Supplemental Digital Content 1, fig. 1, https://links.lww.com/ALN/A926). This preventive effect of volatile anesthetics against H₂O₂-induced hypercontracture seems to be involved in their cardioprotective action during reperfusion of ischemic myocardium, mediated primarily through attenuation of the cellular Ca²⁺ overload.^17,18,20 

ROS cause Ca²⁺ dyshomeostasis, leading to Ca²⁺ overload, by affecting the functions of a variety of ion channels and transporters involved in the regulation of intracellular Ca²⁺ concentrations.⁸  The present current-clamp experiments demonstrate that H₂O₂-induced voltage fluctuations in resting membrane potential (figs. 4A and 5C) act as a key to electrophysiological abnormality, responsible for subsequent development of abnormal automaticities, cellular Ca²⁺ overload, and hypercontracture. It is well established that spontaneous Ca²⁺ release from the SR through RyR2 activates the forward mode of NCX, that is three Na⁺ enter for one Ca²⁺ exit, causing a net depolarizing current. This transient inward current results in an isolated or series of membrane fluctuations that could trigger abnormal action potentials in cardiac myocytes.^39–41  Importantly, spontaneous Ca²⁺ release from the SR is enhanced by either excess Ca²⁺ loading to the SR, RyR2 dysfunction, or a combination of both.^{8,10,11,39–41}  For example, ROS including H₂O₂ enhance RyR2 activity, presumably by direct RyR2 oxidation, and thereby facilitate Ca²⁺ release from the SR.^8,10,11  Alternatively, recent studies indicated that RyR2 activity is enhanced by CaMKII-dependent phosphorylation under some pathological conditions, such as heart failure and atrial fibrillation.^41,44,45  It is, thus, probable that Ca²⁺ release through RyR2 during exposure to H₂O₂ can be mediated through multiple mechanisms, including direct oxidation and CaMKII-dependent phosphorylation. In addition, NCX itself can be potentiated by H₂O₂ in ventricular myocytes.^8,12  Taken together, it seems likely that exposure to H₂O₂ favors the sequential activation of RyR2 and NCX for the transport of Ca²⁺, which generates the electrical NCX current causing membrane voltage fluctuations that could lead to triggered action potentials (fig. 4A). This view may be supported by the prevention of H₂O₂-induced membrane voltage fluctuations by the NCX blocker SEA0400 (fig. 4, B and D).

Sevoflurane has been demonstrated not only to inhibit the spontaneous Ca²⁺ release from the SR through RyR2,^23,24,46  but also to block the NCX current in ventricular myocytes.⁴⁷  These dual inhibitory actions of sevoflurane appear to represent an electrophysiological basis for the effective prevention of membrane voltage fluctuations that lead to abnormal automaticities and hypercontracture during H₂O₂ exposure (fig. 4, C and D).

The present voltage-clamp experiments revealed that H₂O₂ increased peak and late amplitudes of I_Ca,L and shifted the voltage-dependent activation toward more negative potentials through the activation of CaMKII (figs. 7 and 8). H₂O₂ activates CaMKII by direct oxidation of Met281/282 at the regulatory domain, accompanied by the autophosphorylation of CaMKII at Thr287.^9,32  The Western blot analysis also detected significant elevation of Thr287-phosphorylated CaMKII by incubation with H₂O₂ (fig. 8E), supporting that CaMKII is activated by H₂O₂. It is important to note that potentiation of I_Ca,L by CaMKII-mediated phosphorylation during repetitive depolarizations (Ca²⁺-dependent I_Ca,L facilitation) is associated with slower inactivation kinetics,⁴⁸  which appears to be qualitatively similar to the H₂O₂-induced enhancement of I_Ca,L with increased late current amplitude (fig. 8C). Thus, the voltage-clamp results are consistent with CaMKII activation, resulting in phosphorylation of and increased current through I_Ca,L channels (fig. 8, A–D). In addition, because there is Ca²⁺-dependent inactivation of I_Ca,L,⁴⁹  it is possible that H₂O₂ interferes with that process, leading to a larger I_Ca,L.

The present experiments indicate that H₂O₂-induced EADs are dependent on the reactivation of I_Ca,L (fig. 6A), consistent with previous observations in rabbit ventricular myocytes.^15,38  H₂O₂ increased amplitude of inward I_Ca,L and shifted the voltage-dependence of current activation toward more negative potentials (fig. 7), and these biophysical changes in I_Ca,L appear to be favorable not only for the increase in total Ca²⁺ influx through I_Ca,L, but also for the induction of EADs,⁵⁰  accompanied by further Ca²⁺ loading.

Sevoflurane, isoflurane, and desflurane markedly and similarly inhibited the H₂O₂-stimulated I_Ca,L to below control levels (fig. 7 and see Supplemental Digital Content 1, fig. 2, https://links.lww.com/ALN/A926). There are a number of reports showing that these volatile anesthetics substantially inhibit basal I_Ca,L in ventricular myocytes.^51–53  This inhibitory action seems to contribute to a reduction of I_Ca,L even during H₂O₂ exposure. The reversing effects of sevoflurane on H₂O₂-induced changes in late current and V_h appear to be mediated through the inhibition of CaMKII activation, which is consistent with the reduction in the ratio of Thr287-phosphorylated CaMKII/CaMKII to near the control levels by addition of sevoflurane in the presence of H₂O₂ (fig. 8E). These inhibitory actions of sevoflurane on H₂O₂-stimulated I_Ca,L may also contribute to the prevention of excess Ca²⁺ influx and EAD induction.

Because CaMKII is selectively enriched within sarcolemmal-SR junction where CaMKII colocalizes with I_Ca,L channels and RyR2,⁵⁴  CaMKII activity appears most likely to be influenced by Ca²⁺ entering the cell through I_Ca,L⁴⁸ and Ca²⁺ released from the SR through RyR2.³²  At present, the precise mechanism by which sevoflurane reduces the CaMKII activity has yet to be fully elucidated. However, there are some plausible explanations for the ability of sevoflurane to alter CaMKII activity: (1) sevoflurane reduces CaMKII activity by decreasing Ca²⁺ entry through I_Ca,L;⁴⁸  (2) sevoflurane alters the Ca²⁺ binding affinity of calmodulin, as has been demonstrated for other volatile anesthetics, halothane and isoflurane;⁵⁵  (3) alternatively, sevoflurane exerts a direct effect on CaMKII. It is interesting to note that a recent study demonstrated that Thr287-phosphorylated CaMKII is increased during reperfusion of ischemic rat heart, which is significantly reduced by administration of sevoflurane.⁵⁶  Because CaMKII activation has been implicated in the induction of cellular apoptosis and necrosis associated with irreversible ischemia/reperfusion injury,⁵⁷  an inhibitory action of sevoflurane on CaMKII activity is expected to play a role in its cardioprotective action against ischemia or reperfusion injury.

Although the present experiments provide evidence that CaMKII activation contributes at least partly to the H₂O₂-induced cellular Ca²⁺ overload and hypercontracture by potentiating I_Ca,L (fig. 8), future study should quantitatively examine the precise role of CaMKII activation in causing Ca²⁺ overload during H₂O₂ exposure mediated through other mechanisms, including RyR2-mediated Ca²⁺ release.

Whereas myocardial reperfusion injury has a multifactorial genesis,^1,6  the current study focused on one dominant factor of ROS responsible for reperfusion injury. In addition, this study examined isolated ventricular myocytes, which lack the cell-to-cell interactions and thereby cannot transmit the mechanical forces of hypercontracture and electrical disturbances that occur during ischemia/reperfusion in the whole heart.⁵⁸  The ischemia/reperfusion injury is also mediated through inflammatory cell activation and vascular endothelial cell dysfunction, which can be ameliorated by volatile anesthetics to protect the heart.^59–61  The current study thus, does not necessarily represent all aspects of cardiac reperfusion injury and its prevention by sevoflurane. Nevertheless, because many studies have demonstrated that typical features of reperfusion injury in the whole heart, such as Ca²⁺ overload and hypercontracture, can be readily reproduced in isolated cardiac myocytes by exogenous application of ROS,^5–7  the present findings, using ventricular myocyte model, should provide important information relevant to the reperfusion injury and its prevention by sevoflurane in the whole heart.

Sevoflurane at clinically used concentrations adequately prevented the H₂O₂-induced EADs, membrane voltage fluctuations in resting potential, and triggered activities in ventricular myocytes (figs. 4 and 6). These electrophysiological abnormalities are responsible not only for cellular Ca²⁺ overload and hypercontracture, but also for ventricular arrhythmias during reperfusion of ischemic heart.^13,62  Our findings, therefore, have potentially important clinical implications that sevoflurane can exert an antiarrhythmic action, in addition to a protective action against severe irreversible tissue damages during reperfusion. These beneficial effects of sevoflurane were afforded by the simultaneous administration with H₂O₂ (figs. 1 and 4C), suggesting that sevoflurane could protect the heart against these reperfusion injuries associated with oxidative stress, when applied during reperfusion (anesthetic postconditioning).

In conclusion, sevoflurane protects ventricular myocytes against H₂O₂-induced cellular injuries, such as electrophysiological abnormalities, Ca²⁺ overload, and hypercontracture, by affecting multiple Ca²⁺ channels and transporters. This multichannel- or multitransporter-blocking property of sevoflurane appears to play a critical role in exerting the cardioprotective action against reperfusion injury and will provide potential new avenues to develop strategies for protecting the heart during ischemia/reperfusion.

The authors thank Teisuke Takahashi, Ph.D. (Taisho Pharmaceutical Co., Ltd., Saitama, Japan), for the kind help with the use of SEA0400, Tadanori Sugimoto, Ph.D. (Dainippon Sumitomo Pharma Co., Ltd., Suita, Osaka, Japan), for valuable help with statistical analyses, and Mr. Takefumi Yamamoto and Mr. Yasuhiro Mori (Technical Instructors, Central Research Laboratory, Shiga University of Medical Science, Otsu, Shiga, Japan) for helpful advice with the use of the confocal laser scanning microscope.