Left Ventricular Hypertrophy Increases Susceptibility to Bupivacaine-induced Cardiotoxicity through Overexpression of Transient Receptor Potential Canonical Channels in Rats

All experiments were approved by the ethics committee on animal experiments of Osaka City University and performed according to the Guiding Principles for the Care and Use of Animals recommended by the Physiological Society of Japan. Male Sprague–Dawley rats were purchased from Kiwa Laboratory Animals Co. (Japan) and housed in cages (two rats per cage) at the animal center of Osaka City University in a standard 12-h reverse day/night cycle at an ambient temperature of 22°C and 55 to 60% humidity. The rats were randomly assigned to the left ventricular hypertrophy model or the sham model. The rats underwent surgery to generate the aortic constriction rats at the age of 3 weeks to induce pressure overload conditions. The rats were anesthetized with 2 to 4% sevoflurane (Nikko Pharmaceutical Co., Japan) and underwent tracheostomy to place a shortened 18-gauge cannula. Subsequently, the rats were positioned on their right side, and the distal aortic arches were exposed through a left thoracotomy in the third intercostal space. The aortic arches were constricted using 3/0 silk sutures against a paralleled blunted 22-gauge needle just distal to the left subclavian artery. The needle was immediately removed after constriction. The left thoracotomy incision and tracheotomy incision were closed with 4/0 and 6/0 sutures, respectively. After the procedure, cefazolin sodium (200 mg/kg) was subcutaneously administered. Sham-operated rats underwent an identical procedure without aortic ligation at the same age. Experiments were performed 4 weeks after surgery. The investigators performing the experiments were not blinded to the type of rat models due to the differences in appearance of rat models or their hearts, however data analyses were performed by another investigator who was blinded to the group assignment.

Twelve sham rats and 12 aortic constriction model rats were studied to investigate differences in the cardiovascular toxicity of bupivacaine; one aortic constriction model rat was excluded due to inappropriate procedures. There were no missing data for the remaining rats. The rats were anesthetized with 20% urethane (1.2 to 1.5 g/kg, intraperitoneal; Sigma-Aldrich, USA). The caudal vein (for bupivacaine infusion) and right carotid artery (for recording arterial pressure and blood sampling) were cannulated using polyethylene catheters (O.D. 0.965 mm; Becton Drive, USA). The trachea was cannulated via tracheostomy and mechanically ventilated with room air to maintain Paco₂ at 35 to 40 mmHg (SN480-7-10CC; Shinano Seisakusho, Japan). Needle electrodes were placed to record lead II of the electrocardiogram (ECG). Next, rocuronium bromide (30 mg/kg; MSD Co. Ltd., Japan) was administered via the caudal vein for immobilization. Arterial blood was sampled via the right carotid artery, and blood gas was analyzed. The rats were warmed with heating blankets during the experiments, and rectal temperature was maintained at 36.5° to 37.5°C. All experiments were started from 9:00 am and performed during daytime hours.

Hemodynamic parameters were obtained with a transducer (P23XL; Nihon-Koden, Japan) and recorded with a data acquisition system (Powerlab 4/25; AD Instruments, Australia). After stable vital signs were confirmed, 0.5% bupivacaine (Sigma-Aldrich) was infused continuously at a dose of 3 mg · kg^-1 · min^-1 using an electronic syringe pump (NE 1000; World Precision Instruments, USA) until cardiac arrest was determined by ECG for 1 min. Mild and severe cardiovascular toxicities were defined as the time points of doubled duration of QRS complex and cardiac arrest on ECG, respectively. We assessed the time to mild and severe cardiovascular toxicities and the cumulative dose of bupivacaine until cardiac arrest. After cardiac arrest, the plasma concentration of bupivacaine was measured in blood sampled from the right carotid artery or directly from the heart. Serum was separated and frozen at −40°C. Methanol (1 ml) and internal standard (10 μl; propranolol hydrochloride 100 μg/ml in methanol) were added to plasma samples (50 μl) for deproteinization. The mixture was centrifuged at 13,000 × g for 5 min at 4°C, and the supernatants (50 μl) combined with 200 μl of solvents (methanol:water = 1:1, v/v) were analyzed. If the concentration of the plasma sample exceeded the upper limit of quantification, the sample was diluted with blank rat plasma to 100- or 1,000-fold. The samples were analyzed by high-performance liquid chromatography-electrospray ionization-tandem mass spectrometry using the API5000 mass spectrometer system (Applied Biosystems; MDS SCIEX, USA) coupled with a high-performance liquid chromatography system (Acquity UPLC system; Waters Corporation, USA). An Ascentis Express column (50 mm × 2.1 mm I.D., 2.7 μm particle size; Sigma-Aldrich) was used and operated at 40°C. The mobile phase used was 10 mM ammonium formate/formic acid (1,000:2, v/v) (phase A) and acetonitrile (phase B), and the gradient was applied. The injection volume was 1 μl, and the flow rate was at 300 μl/min. A multiple reaction monitoring was used. The precursor-to-product ion transitions used for quantification, and collision energy values were as follows: bupivacaine, 289.4 to 139.9 m/z, 45 eV; and propranolol, 259.8 to 183.1 m/z, 25 eV. The low-, medium-, and high-quality control concentrations were 20, 100, and 800 ng/ml, with coefficient of variation of 6.0%, 5.5%, and 4.1%, respectively. The limit of quantification was 10.0 ng/ml.

Male rats (7 to 8 weeks of age; 200 to 300 g) were weighed and anesthetized with sodium pentobarbital (120 mg/kg; Kyoritsu Seiyaku, Japan) according to body weight. The heart of each rat was excised after heparinization (100 units), immediately arrested in cold iso-osmotic saline containing 20 mM KCl, and mounted on a Langendorff apparatus. Heart weights were measured before ventricular myocyte isolation. The hearts were perfused with Tyrode solution for 2 min followed by calcium ion–free Tyrode solution for 5 min. Then, hearts were perfused with 50 μM Ca⁺ Tyrode solution containing 0.3 mg/ml collagenase (Fujifilm Wako Pure Chemical Corporation, Japan) at 50 cm H₂O for 7 to 9 min for sham rats or 0.4 mg/ml collagenase at 80 cm H₂O for 20 to 30 min for aortic constriction rats. The atrium and right ventricle were removed, and the left ventricle was fragmented and dissociated in modified Kraft–Bruhe solution (70 mM KOH, 30 mM KCl, 70 mM L-glutamic acid, 1 mM MgCl₂, 20 mM taurine, 10 mM KH₂PO₄, 10 mM HEPES, 0.3 mM ethylene glycol tetraacetic acid, and 10 mM glucose; the pH was adjusted to 7.47 with KOH at room temperature). The isolated cardiomyocytes were stored in calicum-free Tyrode solution. The calcium ion concentration in the cell-containing Tyrode solutions was then consecutively increased to 50, 100, 500, and 1,800 μM every 15 min at room temperature.

In response to peer review, additional experiments using transfected HEK-293T (ATCC) cells were conducted to confirm the effects of TRPC3 channels on the bupivacaine-induced cardiotoxicity without other pathologic factors which are related to left ventricular hypertrophy. HEK-293T cells were cultured in Dulbecco Modified Eagle Medium (Nissui Pharmaceutical Co., LTD, Japan), supplemented with 10% heat-inactivated fetal calf serum, 2 mM L-glutamine, and 1% penicillin/streptomycin (Sigma-Aldrich) at 37°C in a humidified atmosphere with 5% CO₂. The expression vectors carrying genes of the cardiac sodium ion channel (SCN5A) (NM_198056, gift from Dr. Hisao Yamamura, Nagoya City University, Japan),²¹  with and without human TrpC3-NMyc (Addgene #21085) were transiently transfected in HEK-293T cells using FuGENE HD (Promega Co., USA) in accordance with the manufacturer’s instructions. A plasmid encoding the green fluorescent protein (OriGene Technologies, USA) was cotransfected to identify transfected cells for patch clamp recordings. The cells were used at 24 to 48 h after transfection.

Sodium ion channel currents were recorded in ventricular myocytes (n = 35 and 33 from sham and aortic constriction rats, respectively). Patch pipettes were pulled from borosilicate glass capillaries (World Precision Instruments, USA) and were fire-polished to a resistance of 1.5 to 2.5 MΩ. Whole cell voltage-clamp recordings were performed at room temperature using a patch clamp amplifier (AxoPatch 200B; Axon Instruments, USA). Data were filtered at 5 kHz, digitized at 10 to 20 kHz with an analog–digital converter (Digidata 1440A; Axon Instruments), and were analyzed using pCLAMP 10 software (Axon Instruments). The external solution for isolated cardiomyocytes contained 10 mM NaCl, 20 mM HEPES, 120 mM CsCl, 3 mM CoCl₂, 10 mM glucose, 1 mM CaCl₂, and 2 mM MgCl₂ (pH 7.35 with CsOH at room temperature). The pipette solution for isolated cardiomyocytes contained 10 mM NaF, 20 mM HEPES, 115 mM CsF, 2 mM ethylene glycol tetraacetic acid, 20 mM CsCl, 1 mM MgCl₂·6H₂O, and 5 mM Mg₂ATP (pH 7.25 with CsOH at room temperature). Cardiomyocytes were dialyzed with the pipette solution for 5 min before recording currents. The holding potential was maintained at −80 mV, which is nearly equal to the physiologic resting potential of ventricular cardiomyocytes.²²  The capacitance and leak currents were canceled by the P/4 subtraction protocol. Only cells with a seal resistance of 1 GΩ or greater were used. Series resistance (4.3 ± 2.1 MΩ; n = 68 cells) was compensated by 70% to reduce voltage errors. Cell membrane capacitance was calculated using the automatic procedure of the amplifier (sham, 145 ± 24 pF; aortic constriction, 171 ± 35 pF). In the same manner, whole cell voltage-clamp recordings were performed at room temperature for HEK-293T cells using the following solutions: the external solution contained 140 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 10 mM HEPES, 10 mM Glucose (pH 7.35 with NaOH at room temperature), and the internal solution contained 130 mM CsCl, 15mM NaCl, 2 mM MgCl₂, 10 mM EGTA, 10 mM HEPES (pH 7.25 with CsOH at room temperature). A concentrated stock solution of 1-oleoyl-2-acetyl-sn-glycerol or pyrazole compound 3 (Pyr3) was prepared in dimethyl sulfoxide. The final dimethyl sulfoxide concentration was less than 0.2%, which affected neither sodium ion currents nor the cell shapes (data not shown).

Immunostaining of TRPC3 channels was performed with anti-TRPC3 antibody in rat cardiomyocytes. Left ventricular cardiomyocytes from both sham and aortic constriction rats were fixed with 4% paraformaldehyde in phosphate-buffered saline, washed with phosphate-buffered saline, permeabilized with 0.2% Triton-X in phosphate-buffered saline for 15 min, and then blocked for 30 min in phosphate-buffered saline containing 3% bovine serum albumin. Then, the cardiomyocytes were incubated with rabbit anti-TRPC3 antibody (1:100 Sigma T5067; Sigma-Aldrich) in phosphate-buffered saline for 1 h at room temperature and stained with Alexa Fluor 488-conjugated goat anti-rabbit secondary antibody (1:100; ab150077; Abcam, United Kingdom). Immunofluorescent images were obtained using the scanning laser confocal microscope (Zeiss LSM700) equipped with an argon laser and imaging software (Zen; Zeiss, Germany). Fluorescence intensities were expressed in arbitrary pixel units in the range of 0 to 255.

Data were presented as mean ± SD. Student t test was used to determine intergroup differences between the two rat groups both in vivo and in vitro experiments, and paired Student t tests were used for intragroup comparisons in the same rat groups in the patch clamp experiments after confirming a normal distribution using the Shapiro–Wilk test. If the data distribution failed the normality test, a Mann–Whitney U test was used (data presented as median with 25th and 75th percentiles). The sample size for each group was calculated based from preliminary experiments. The cumulative bupivacaine dose for severe cardiovascular toxicity was 66.9 ± 9.6 mg/kg (n = 3) in preliminary in vivo experiments using naive rats. Assuming the expected SD of both models are 9.6, 10 rats in each group were necessary to detect a 20% difference with a power of 0.80 and a type 1 error protection of 0.05 (two-tailed). Meanwhile, in patch clamp experiments, a group size of four cells each was necessary to detect a 30% difference (SD ± 8%) in intergroup means of peak sodium ion currents based on our preliminary experiments. Dose–response studies were conducted by means of the four-parameter logistic nonlinear regression model using the Hill equation:

where Y is the expected response at dosage X, a is the minimum asymptote, b is the maximum asymptote, IC₅₀ is the half-maximal inhibitory concentration, and H is the Hill slope. The Hill coefficient was calculated automatically with SigmaPlot 13.0 (Systat Software Inc, USA).

Statistical analysis was performed using SigmaPlot 13.0. Outliers, if any, were included in the analyses. Differences with a P value less than 0.05 (two-tailed) were considered statistically significant.

The macroscopic appearance and hematoxylin/eosin staining of a short-axis cross-section showed obviously enlarged hearts in aortic constriction rats (fig. 1, A and B). Body weights were not different between the sham and aortic constriction rats (table 1). The heart weights of aortic constriction rats were greater than those of sham rats (1.35 ± 0.17 g vs. 1.98 ± 0.33 g; P < 0.0001, fig. 1C). The heart weight to body weight ratio of the aortic constriction rats was significantly greater than that of the sham rats (0.53 ± 0.07% vs. 0.78 ± 0.02%; P < 0.001; fig. 1D).

Baseline mean arterial pressure (MAP) was significantly higher and QRS complex was significantly prolonged in aortic constriction rats compared to sham rats (table 1). Heart rates (HRs) and blood gas analysis results did not differ significantly (table 1). Continuous infusion of bupivacaine decreased HR (fig. 2A) and MAP (fig. 2B) in both rat models. The data at mild cardiovascular toxicity (indicated by “a” and “b”) and severe cardiovascular toxicity (HR = 0, i.e., cardiac arrest) are shown in figure 2. The time to mild and severe cardiovascular toxicities in aortic constriction rats was 147 ± 35 s and 1,034 ± 211 s (n = 11), respectively, which was significantly shorter than that in sham rats (197 ± 55 s and 1,302 ± 324 s; n = 12; P = 0.017 and P = 0.030, respectively). The cumulative bupivacaine dose until cardiac arrest was significantly lower in aortic constriction rats (50.6 ± 11.0 mg/kg; n = 11) than in sham rats (68.9 ± 15.4 mg/kg; n = 12; P = 0.004; fig. 2C). The plasma bupivacaine concentration at each cardiac asystole was also significantly lower in aortic constriction rats (41.4 ± 12.8 μg/ml; n = 11) than in sham rats (58.1 ± 23.3 μg/ml; n = 12; P = 0.048) (fig. 2D). Therefore, the concentration of bupivacaine that induced cardiac arrest was lower in aortic constriction rats.

Voltage-gated sodium ion channels were activated by a series of 50-ms depolarizing pulses from a holding potential of −80 mV to different membrane potentials (−60 to 20 mV) in 10 mV increments at 0.2 Hz (fig. 3, A and B). The sodium ion current density at −30 mV in the control was 19.7 ± 8.1 pA/pF (n = 35 cells from 33 rats) and 14.0 ± 7.7 pA/pF (n = 33 cells from 25 rats) in sham and aortic constriction rats, respectively (P = 0.009).

Sodium currents were decreased by application of 3 μM bupivacaine alone and was further decreased by adding 100 μM 1-oleoyl-2-acetyl-sn-glycerol, a TRPC3 channel activator, in both sham (fig. 3A) and aortic constriction rats (fig. 3B). Current–voltage relations normalized by the maximum current at −30 mV in the two rat models were almost overlapped (fig. 3C). The peak sodium ion currents at −30 mV were used for subsequent analyses. Semilogarithmic dose–response curves for peak sodium ion currents were fitted using the Hill equation (fig. 3D). The half-maximal inhibitory concentration for bupivacaine alone was 4.5 and 4.3 μM in the sham and aortic constriction rats, respectively. Coapplication of bupivacaine and 100 μM 1-oleoyl-2-acetyl-sn-glycerol decreased the half-maximal inhibitory concentration to 3.9 and 2.6 μM in sham and aortic constriction rats, respectively. Thus, 1-oleoyl-2-acetyl-sn-glycerol increased the sensitivity to bupivacaine in both groups, but this increase was stronger in the aortic constriction rats. 1-oleoyl-2-acetyl-sn-glycerol alone did not affect sodium ion currents under the protocol and solutions used in the present experiments (data not shown).

As the HR at bupivacaine-induced mild cardiotoxicity was ~5 Hz, the use-dependent block of sodium ion channels was examined. We applied a series of 50 depolarizing pulses (−30 mV) at 4 Hz. In the presence of bupivacaine, 1-oleoyl-2-acetyl-sn-glycerol slightly accelerated the decline of sodium ion currents and decreased their amplitudes during the train of pulses in both sham (fig. 4, A and C) and aortic constriction (fig. 4, B and D) rats. There were no statistically significant differences in the relative amplitudes of the 50th sodium ion currents between the two groups regardless of 1-oleoyl-2-acetyl-sn-glycerol co-application (n = 4 cells per group); bupivacaine alone caused the peak sodium ion currents to reduce to 63 ± 6% and 60 ± 10% of the control value in sham and aortic constriction rats, respectively (P = 0.634); bupivacaine and 1-oleoyl-2-acetyl-sn-glycerol coadministration caused the peak sodium ion currents to reduce to 56 ± 7% and 51 ± 9% of the first pulse current in sham and aortic constriction rats, respectively (P = 0.385) (fig. 4, C and D).

To confirm the contribution of TRPC3 channels to increased sensitivity to bupivacaine, the effects of QX-314 (a membrane impermeable, positively charged lidocaine derivative) on sodium ion currents were examined. Individually, 50 μM QX-314 did not decrease the peak sodium ion currents (sham rats, 102 ± 13% of control, n = 5; aortic constriction rats, 91 ± 25%, n = 5; P = 0.413) (fig. 5A), whereas coapplication of 50 μM QX-314 and 100 μM 1-oleoyl-2-acetyl-sn-glycerol significantly decreased sodium ion currents compared to QX-314 alone in both groups (sham rats, 79 ± 10% of control; n = 5; P = 0.004; aortic constriction rats, 47 ± 27%; n = 5; P = 0.020), and the intermodel difference was significant (P = 0.040) (fig. 5A). At a higher QX-314 concentration (100 μM), sodium ion currents decreased slightly (88 ± 9% of control in sham rats; P = 0.039; 87 ± 20% in aortic constriction rats; P = 0.158; n = 5 each group) (fig. 5B, black column). The inhibitory effects of 1-oleoyl-2-acetyl-sn-glycerol coapplication were potentiated, as sodium ion currents decreased further to 48 ± 9% of control in sham rats (P = 0.003) and 33 ± 10% in aortic constriction rats (P = 0.027; fig. 5B, grey column). However, the intermodel difference with the coapplication of 100 μM QX-314 and 100 μM 1-oleoyl-2-acetyl-sn-glycerol was not significant (P = 0.076; fig. 5B). Furthermore, the inhibitory effect by QX-314 and 1-oleoyl-2-acetyl-sn-glycerol was cancelled by pre- and coapplication of pyrazole compound 3 (Pyr3), a specific inhibitor for TRPC3 channels, in other cells (sham rats, 72 ± 17% of control; P = 0.028; aortic constriction rats, 89 ± 20% of control; P = 0.007; fig. 5B).

Voltage-gated sodium ion channels were activated by a series of 50-ms depolarizing pulses from a holding potential of −110 mV to different membrane potentials (−70 to 20 mV) in 10 mV increments at 0.2 Hz (fig. 6, A and B). Sodium ion currents were reduced by 3 μM bupivacaine, while there was no further reduction caused by 1-oleoyl-2-acetyl-sn-glycerol in HEK-293T cells transfected with Nav1.5 alone (n = 4 cells; 90 ± 3% and 89 ± 5% of control, respectively; P = 0.617). On the other hand, there were further reductions caused by 1-oleoyl-2-acetyl-sn-glycerol in HEK-293T cells cotransfected with Nav1.5 and TRPC3 (n = 4 cells; 86 ± 7% and 72 ± 7% of control, respectively; P = 0.048). There was a significant difference between Nav1.5 alone group and Nav1.5 with TRPC3 group in sodium ion current reductions caused by bupivacaine and 1-oleoyl-2-acetyl-sn-glycerol (n = 4 cells per group; P = 0.006) (fig. 6C).

Immunostaining with the anti-TRPC3 antibody revealed that TRPC3 channels were predominantly expressed on cardiomyocyte plasma membranes in both sham and aortic constriction rats (fig. 7, A and B). The plasmalemmal expression tended to be more prominent in aortic constriction rats. Fluorescence intensity profiles on a traverse line showed the distribution of TRPC3 protein in these cells (fig. 7, C and D). In order to quantify the localization, the ratio of the mean intensity at the two peripheries to the intracellular minimal intensity was calculated (fig. 7E).²³  The ratio of mean intensity at the two peripheries to intracellular minimal intensity levels was greater in aortic constriction rats than in sham rats (median [interquartile range]; sham rats: n = 15; 3.0 [2.6 to 3.8]; aortic constriction rats: n = 19; 4.9 [4.1 to 6.6]; P = 0.002) (fig. 7F).

In the current study, we examined the cardiac toxicity of bupivacaine on a left ventricular hypertrophy model with in vivo and in vitro experiments. The results indicated that the impact of bupivacaine infusion on hemodynamics was greater in rats undergoing modified transverse aortic constriction than in sham rats in vivo, and bupivacaine-induced inhibition of sodium ion channels in cardiomyocytes was greater in aortic constriction rats than in sham rats. Uptake of protonated bupivacaine through upregulated TRPC3 channels (“pore phenomenon” effect) was suggested to underlie the higher sensitivity to local anesthetics in left ventricular hypertrophy.

Prolonged QRS complex is a feature of ECG changes due to local anesthetic toxicity and is used to assess the potency of cardiovascular suppression.^24,25  We defined a doubled duration of QRS complex as mild cardiovascular toxicity and cardiac arrest as severe cardiovascular toxicity. The times to mild and severe cardiovascular toxicities were both shortened in aortic constriction rats despite lower cumulative doses, suggesting that vulnerability to bupivacaine was higher in the left ventricular hypertrophy model. This was unlikely due to the increase in plasma bupivacaine concentration resulting from heart failure, considering that the concentration at cardiac arrest was lower in aortic constriction rats than sham rats. The cumulative doses needed to cause cardiac arrest in this study were relatively greater than those in a previous report.²⁶  One reason for this disparity may be the difference in anesthetics used in the experiments (urethane in the current study and isoflurane in the report). The baseline QRS complex duration in aortic constriction rats was prolonged, as was reported in rodent left ventricular hypertrophy models²⁷  and in humans with left ventricular hypertrophy^3,4  (table 1). This effect is one of the electrophysiologic properties of left ventricular hypertrophy and is attributed to delayed conduction and morphological changes.

Several reports have been conducted on differences in cardiomyocyte electrophysiology in left ventricular hypertrophy models compared to non–left ventricular hypertrophy models.^6,28–32  Electrophysiologic changes accompanied with left ventricular hypertrophy or heart failure in animal models include prolonged action potential durations,^32–34  altered sodium ion currents,²⁹  increased sodium ion/calcium ion exchanger currents, and decreased calcium ion currents.^6,28  Changes in sodium ion current density were inconsistent among the previous reports.^28,29  In the present study, sodium ion current density was significantly lower in aortic constriction rats. Consistent with a previous report,³¹  cell membrane capacitance, which monitors cell size, was also increased. In addition, electric remodeling induced by aortic constriction (late sodium ion current) has been reported, in which sodium ion channel gating properties were altered, i.e., a shifted steady-state inactivation curve, increased intermediate inactivation, and slowed recovery from inactivation.^30,35  Because the inhibition of cardiac sodium ion channels by local anesthetics occurs in a state-dependent manner, it is possible that higher sensitivity to bupivacaine in aortic constriction rats is caused by altered gating properties. However, the current-voltage relationships and half-maximal inhibitory concentration of bupivacaine alone were comparable between sham-operated rats and aortic constriction rats in this study, suggesting that the changes in channel properties caused by left ventricular hypertrophy were not crucial for tonic blockage of sodium ion channels. Coapplication of bupivacaine and 1-oleoyl-2-acetyl-sn-glycerol decreased sodium ion currents to a greater degree in aortic constriction rats than in sham rats (fig. 3D). Use-dependent block by bupivacaine was not significantly different between the two models and was greater when 1-oleoyl-2-acetyl-sn-glycerol was co-applied with bupivacaine in both models. In vivo, regarding the time of mild cardiovascular collapse between sham and aortic constriction rats, where high heart rate was retained (~5 Hz) and use-dependent block could be promoted, the differences were statistically significant.

Ferreira et al. reported that TRPA1, transient receptor potential cation channel subfamily M member 8 (TRPM8), and TRPV1-4 subtypes could promote uptake of large molecular weight solutes up to 900 Da.¹³  However, TRPC channels have not been reported to pass large molecular weight solutes. We demonstrated that sodium ion currents were decreased by the combination of 1-oleoyl-2-acetyl-sn-glycerol and QX-314, indicating that QX-314 passed through TRPC3 channels activated by 1-oleoyl-2-acetyl-sn-glycerol. When TRPC3 channels are sufficiently activated with adequate doses of 1-oleoyl-2-acetyl-sn-glycerol, the amount of passed QX-314 is dependent on the number of TRPC3 channels expressed on the plasma membrane. This implied that the expression of TRPC3 channels in aortic constriction rats was higher than in sham rats. Previous reports suggested that QX-314 inhibits sodium ion channels in cardiomyocytes from outside of the membrane.^36,37  However, in those studies, relatively high QX-314 concentrations (0.5 mM) were used for sodium ion channels expressed in Xenopus oocytes or single-channel recordings.^38,39  We observed that, at 100 mM, QX-314 alone decreased sodium ion currents slightly, and its effects on the currents with 1-oleoyl-2-acetyl-sn-glycerol or pyrazole compound 3 (Pyr3) were not significantly different between sham and aortic constriction rats. There is a possibility that high QX-314 concentrations decrease sodium ion currents partly through unidentified mechanisms other than the TRPC3-mediated pathway.

Since 1-oleoyl-2-acetyl-sn-glycerol can activate not only TRPC3, but also TRPC6 and TRPC7^40,40,41  pyrazole compound 3 (Pyr3), a selective and direct inhibitor of TRPC3^4242–44 was used to identify the subtype. Under the condition of preperfusion with pyrazole compound 3 (Pyr3), sodium ion currents were not decreased by the coapplication of QX-314 and 1-oleoyl-2-acetyl-sn-glycerol. These results unveiled that TRPC3 rather than TRPC6 is mainly involved in this mechanism. Additionally, TRPC6 protein expression decreases immediately after birth,⁴⁵  whereas TRPC3 channels are expressed in cardiomyocytes dominantly in the plasma membrane and are stored in submembrane pools, consistent with our immunocytochemical images. Expression of plasmalemmal TRPC3 channels was enhanced in aortic constriction rats with left ventricular hypertrophy. In the experiments using transiently transfected HEK-293T cells where other pathologic factors with left ventricular hypertrophy were not present, we obtained additional evidence that sodium ion current reduction caused by bupivacaine was exacerbated only by activated TRPC3 channels.

1-oleoyl-2-acetyl-sn-glycerol is an analog of an essential second messenger, diacylglycerol. In in vivo experiments, bupivacaine infusion alone led to greater cardiovascular suppression in aortic constriction rats. This could be explained by the constitutive activity of TRPC3 channels, i.e., TRPC3 channels are expressed at relatively high levels on the plasma membrane and constitutively activated by diacylglycerol, which is physiologically produced by phospholipase C^45–47  and is increased in cardiomyocytes by transverse aortic constriction surgery.⁴⁸  Additionally, TRPC3 channels can be activated by various stimuli, such as mechanical stress and neurohumoral factors. For instance, a heteromeric channel with TRPC3 and TRPC1 is activated by stromal interaction molecule 1 (STIM1), a sensor for store-operated calcium ion release.⁴⁶  In that sense, factors other than diacylglycerol or its analog, 1-oleoyl-2-acetyl-sn-glycerol, could also stimulate TRPC3 channels in vivo.⁴⁷  In the present in vivo study, there was a possibility that constitutively activated TRPC3 channels promoted bupivacaine transfer into the cell, which led to exacerbated bupivacaine-induced cardiac toxicity in aortic constriction rats. This assumption was supported by the greater decrease in half-maximal inhibitory concentration for bupivacaine-induced inhibition of sodium ion currents, probably due to the “pore phenomenon” of TRPC3 channels in aortic constriction rats. This “pore phenomenon” effect might explain, at least partly, the previous report that administration of QX-314 alone led to cardiac toxicity in the whole animal model.⁴⁹ 

There are several limitations in this study. First, TRPC6 proteins can assemble heterotetramers with TRPC3 proteins. It is possible that TRPC6 is reupregulated in cardiac hypertrophy, although ventricular TRPC6 is mainly expressed in fetuses and has low basal activity. However, in the case of heterotetramers, the inhibition of TRPC3 proteins alone affected the total channel activity.⁵⁰  Hence, the presence of heterotetramers seemed to be uninfluential on the results. Second, we employed one 1-oleoyl-2-acetyl-sn-glycerol concentration, leaving the possibility that differences in agonist concentration alter the effect on QX-314 permeability. However, there is no evidence that higher agonist concentrations lead to a higher conductance state.¹³  Third, we utilized only male rats to exclude female hormonal changes during the menstrual cycle which might affect the drug distribution and metabolism. Since the current study has demonstrated the similarity of the results in vivo and in vitro, it is possible to some extent to investigate the effect of left ventricular hypertrophy on the cardiotoxicity induced by local anesthetics using cardiomyocytes.

We concluded that left ventricular hypertrophy rats are more vulnerable to bupivacaine-induced cardiotoxicity, which may be attributed to the increased uptake of protonated bupivacaine through TRPC3 channels that are upregulated in left ventricular hypertrophy.

The authors thank Masayuki Shiota, Ph.D. and Junko Kawawaki, M.S. (Research Support Platform of Osaka City University Graduate School of Medicine, Osaka, Japan) for technical support, and Hisao Yamamura, Ph.D. (Department of Molecular & Cellular Pharmacology, Graduate School of Pharmaceutical Sciences, Nagoya City University, Nagoya, Japan) for kindly providing the expression vectors of Nav1.5 channels.

This study was supported by a grant-in-aid for scientific research (18K16491) from the Japan Society for the Promotion of Science (Tokyo, Japan).

The authors declare no competing interests.