Bupivacaine Suppresses [Ca²⁺]_i Oscillations in Neonatal Rat Cardiomyocytes with Increased Extracellular K⁺ and Is Reversed with Increased Extracellular Mg²⁺

Abstract

Lidocaine is used to treat cardiac arrhythmias, whereas bupivacaine is noted for its cardiotoxicity. A precise mechanism for these differences is unclear, and there is no well defined antidote for local anesthetic cardiotoxicity. Our study compares the effect of lidocaine and bupivacaine on oscillations of intracellular Ca²⁺ coupled with contractions in neonatal rat cardiomyocytes by using digital imaging. In medium containing 5.6 mM K⁺, both 42 μM lidocaine and 5.5 μM bupivacaine significantly reduced the oscillation rate. The oscillatory patterns were highly irregular, and the rates were increased in the presence of bupivacaine in 7.6 mM K⁺ medium, eventually degenerating into a loss of oscillations after several minutes of bupivacaine exposure. Irregular oscillations did not occur with lidocaine until the K⁺ concentration was increased to 10 mM. Increasing the Mg²⁺ and Ca²⁺ concentrations by 2 mM each recovered oscillation that had been suppressed by bupivacaine in high K⁺ buffer. Evaluation of intracellular Ca²⁺ oscillations in neonatal rat suggests that increased extracellular K⁺ may be an important component of bupivacaine cardiotoxicity.

Implications 

Evaluation of intracellular Ca²⁺ oscillations in neonatal rat myocytes suggests that increased extracellular K⁺ may be an important component of bupivacaine cardiotoxicity.

Accidental IV injection of bupivacaine in humans (¹) or its experimental injection in laboratory animals (^2–5) results in a variety of cardiac tachyarrhythmias, including the common occurrence of ventricular fibrillation. Bupivacaine has greater toxicity than lidocaine (¹), but even lidocaine can be proarrhythmic (⁶). The mechanisms by which local anesthetics produce cardiac rhythm disturbances are incompletely characterized, but have been assumed to result from excessive blockade of Na⁺ channels (^5,7). However, local anesthetics interact with dihydropyridine-sensitive Ca²⁺ channels (^3,8), K⁺ channels (^9,10), the ryanodine receptor (RyR) (¹¹), and many other ion channels and enzymes. Any of these actions could trigger arrhythmias and account for the greater cardiotoxicity of bupivacaine over lidocaine. With excessive local anesthetic concentrations in blood, convulsions usually precede cardiovascular collapse and the resulting sympathetic (⁴), acid-base, and electrolyte disturbances (^1,2,12,13) may contribute to the cardiotoxicity of these drugs in the clinical setting.

Arrhythmias ultimately arise from alterations in the electrical activity of myocytes, which is determined by the flow of ionic currents across both the sarcolemmal membrane (SLM) and the sarcoplasmic reticulum (SR) (¹⁴). Models have only recently been able to link the flow of intracellular Ca²⁺ ([Ca²⁺]_i) across the SR with the movement of ions across the SLM, and, as a result, much less is known about the role of [Ca²⁺]_i on the generation of arrhythmias (¹⁵). Aberrations in [Ca²⁺]_i are nevertheless thought to be important in generation of arrhythmias associated with ischemia and reperfusion, especially ventricular tachycardia, fibrillation, and torsades de pointes associated with myocardial infarction (^16–18). Possible mechanisms of these disturbances include Ca²⁺ overload leading to early and delayed after depolarization. These studies were therefore conducted to determine whether abnormal [Ca²⁺]_i oscillations were produced by bupivacaine, but not by lidocaine, which might explain the greater toxicity of bupivacaine in the clinical setting. It is our hypothesis that bupivacaine disrupts [Ca²⁺]_i oscillations, and that this disruption can be used experimentally to better define the pharmacological differences between bupivacaine and lidocaine.

Methods

Cardiomyocyte Culture

This study had institutional review board approval. Myocytes were from rat pups 3–5 days old. Hearts were sterilely dissected and placed in 25 mL of Hanks balanced salt solution with 10% trypsin. Hearts were cut into 1–2 mm sections with a razor and enzymatically degraded at room temperature for 4 h. Every 10 min, cells were further dissociated with trituration through a 10 mL plastic pipette. After this enzymatic-trituration process, undigested chunks were removed; cells were centrifuged for 15 min at 2000 g, the supernatant discarded, and the pellet resuspended in 20 mL of Dulbecco’s Modified Eagle’s medium (containing 5.6 mM K⁺, 1.2 mM Mg²⁺, and 1.5 mM Ca²⁺) lacking bicarbonate (pH 7.4 with NaOH). The pellet was triturated through a 10 mL pipette multiple times until cells were suspended, and 1 mL of the suspension was place on glass cover slips coated with poly-l-lysine. Cells were allowed to adhere to the surface for the next 20 min, after which, the suspension was aspirated and antibiotic-free Dulbecco’s Modified Eagle’s medium containing 6% fetal bovine serum (JRH Biosciences, Lenex, KS), 6% defined/supplemented bovine calf serum (Hyclone, Logan, UT) was added. Cells were grown for 3–7 days in vitro and periodically examined for spontaneous beating that was easily seen in >90% of the cultures under phase-contrast microscopy. Cardiomyocytes were plated sparsely so that they were not confluent.

Measurement of [Ca²⁺]_i

[Ca²⁺]_i levels were determined with digital imaging of fluorescent dyes with the Ca²⁺-indicator dye, fluo-3. Cells were incubated at 37°C for 1 h in growth medium with fluo-3/acetoxymethyl ester (2 μM; Molecular Probe, Eugene, OR). Cells were then washed and placed in a microincubator perfusion well (Peltier) mounted on a Nikon Diaphot (Monmouth Junction, NJ) microscope and maintained at 37°C. Intracellular fluo-3 was excited at 488 nm using a 75 W xenon lamp with a diffraction spectrophotometer (Photon Technology International). Dye-emitted light was directed through a 530-nm filter to a photo intensifier and charged-coupled device camera (Hamamatsu, Model C 2400-68). Images were recorded on an industrial videocassette recorder for later analysis. Five to 7.5 points/s were obtained, and changes in [Ca²⁺]_i were recorded as changes in the intensity of fluorescence in arbitrary units. The cardiomyocyte could be seen contracting during the experiment. The cell flashes brightly during each contraction, so it could be visually seen that excitation-contraction (E-C) coupling was linked with [Ca²⁺]_i oscillations. Low ambient light from the fluorescent dye allows visualization of the cell even when it is not contracting. Tracings were made by measuring from the smallest area allowed by the software and were taken from a single cardiomyocyte. This area represented from 2% to 10% the total area of the cell (size of cardiomyocyte varied markedly). [Ca²⁺]_i oscillations as seen from the 530 nm emission were bright enough to visualize the contractions occurring in the myocytes, and increases in [Ca²⁺]_i correlated with each contraction. In preliminary experiments, a concentration response of spontaneously contracting cardiomyocytes to local anesthetics was examined by incrementally increasing the local anesthetic in the perfusion medium. Otherwise, cells were treated with a single dose of local anesthetic, and repeated measures were not taken unless explicitly stated in the washout experiments. A 4–60 min section of spontaneous contractions was recorded in control cells before adding a local anesthetic. Longer control times were taken to determine that changes did not occur as a result of increased length of time when cells were contracting under 488 nm light exposure. Thus, each tracing shown in the figures represents the [Ca²⁺]_i oscillations from a single cardiomyocyte.

Significance was determined in an experiment before and after treatment with a local anesthetic by the paired t-test. In comparing between groups, a repeated-measures analysis of variance was performed.

Results

Initial studies were performed on 10 coverslips to determine the concentration response to bupivacaine. There was wide variability with 127 μM bupivacaine required to prevent [Ca²⁺]_i oscillations and the associated E-C coupling in 3 of 10 slides; 55 μM in 5 of 10 slides; and 17 μM in 2 of 10 slides. Because of the wide variability in our model, we focused on smaller concentrations of local anesthetics that are clinically associated with toxicity.

Response in 5.

6 mM K⁺ Medium

Typical responses in growth medium containing 5.6 mM K⁺ to 5.5 μM bupivacaine and 42 μM lidocaine are shown in Fig. 1, and the changes in contraction rates are listed in Table 1. Bupivacaine resulted in a statistically significant rate decrease from 95 ± 8 to 75 ± 9 bpm, and lidocaine decreased the rate from 92 ± 14 to 80 ± 16 bpm (P < 0.004 and 0.01, respectively). Neither bupivacaine nor lidocaine suppressed the amplitude of maximal fluorescence of the Ca²⁺ transients from pretreatment amplitudes or resulted in a disorganized oscillation pattern at these concentrations, and the decrease in rate induced by lidocaine was not significant from that induced by bupivacaine (P = 0.27). [Ca²⁺]_i levels are not quantitated using fluo-3. Figure 2 shows the response to a three times larger concentration of bupivacaine (17 μM). Oscillations were not inhibited, but peak Ca²⁺ transients were initially suppressed with partial recovery over the next several minutes.

Figure 1: Rate slowing induced by bupivacaine and lidocaine in 5.6 mM K⁺. Bupivacaine and lidocaine concentrations were in the range that have been associated with seizures in experimental animals and patients. Time series are shown for intracellular Ca²⁺ oscillations in cardiomyocytes before and after treatment with (A) 5.5 μM bupivacaine or (B) 42 μM lidocaine. A decrease in rate is seen soon after the application of drug. Thirty-second tracings of pre- and posttreatment are shown below the longer tracings revealing no changes in the overall pattern of Ca²⁺ oscillations.

Table 1: Contraction Rate Before and After Treatment with a Local Anesthetic in Medium Containing Normal and Large Concentrations of K⁺

Figure 2: Partial suppression of Ca²⁺ transients by 16.5 μM bupivacaine in 5.6 mM K⁺. Large concentrations of bupivacaine suppress Ca²⁺ transients. Recovery of the suppression occurs several minutes after bupivacaine application. Bottom tracings were taken from control and experimental sections and show 30-second segments with no changes in the overall pattern other than a rate decrease.

Response in 7.

6 mM K⁺ Medium

When the K⁺ in the perfusion medium was increased to 7.6 mM, concentrations of bupivacaine that decreased the rate previously and had no effect on the oscillation pattern, now suppressed the Ca²⁺ transients, increased the oscillation rate from 125 ± 21 bpm to 215 ± 34 bpm (P < 0.003, n = 8), and after a brief period of rapid beating activity, completely suppressed all [Ca²⁺]_i oscillations and E-C coupling (Fig. 3). Suppression occurred whether the K⁺ was increased before bupivacaine exposure (Fig. 3, top tracing) or when the K⁺ was increased after bupivacaine exposure (bottom tracing). The contraction rate is presented in Table 1 and shows that there was an initial increase in rate before a complete suppression of contractile activity in 5 of 8 experiments. One of the 3 of 8 experiments where bupivacaine failed to suppress contractions is shown in Fig. 4. In these failed cases, bupivacaine caused highly irregular oscillatory activity, suppressing Ca²⁺ transients in two of the three experiments; but, in the final one, bupivacaine caused very large Ca²⁺ transients with a rate of three per minute before irregular activity (Fig. 4). Cardiomyocytes treated with lidocaine are less sensitive to the effect of the K⁺ than those treated with bupivacaine and had a reduced rate in 7.6 mM K⁺ medium (significantly different from bupivacaine, P < 0.0003). However, a further increase in the K⁺ to 10 mM resulted in highly irregular oscillations (Fig. 5) and a suppression of activity in 3 of 8 cardiomyocytes. There was also a significant increase in the contraction rate from 128 ± 18 to 220 ± 34 (P < 0.02, Table 1).

Figure 3: Suppressive effect of 5.5 μM bupivacaine in 7.6 mM K⁺. When bupivacaine is applied to cardiomyocytes in medium containing 7.6 mM K⁺, an irregular beating pattern occurs with an increase in the rate followed by a complete suppression of activity. Suppression occurs whether the K⁺ is increased before (top tracing) or after (bottom tracing) the application of bupivacaine.

Figure 4: Irregular intracellular Ca²⁺ oscillations in a cardiomyocyte whose contractions were not suppressed by 5.5 μM bupivacaine and 7.6 mM K⁺. There is a complete suppression of activity in about 60% of the experiments. In the remaining 40%, there are highly irregular Ca²⁺ oscillations as seen in this tracing. After several minutes of increased rate and irregular activity, the Ca²⁺ transients are more than tripled in size and the rate is slowed to only three per minute. Then, the cardiomyocyte begins disorganized activity with Ca²⁺ oscillations occurring in a myocyte that visually appears to be fibrillating. Thirty-second tracings were taken from the regions as labeled in the three lower tracings.

Figure 5: Lidocaine-induced irregular intracellular Ca²⁺ oscillations in a cardiomyocyte continuously exposed to 10 mM K⁺. In contrast to the suppression of oscillations in cells treated with bupivacaine, those treated with lidocaine (42 μM) are more resistant to changes, even when the extracellular K⁺ is increased to 7.6 mM. However, with further increases in K⁺, irregular activity is often seen as illustrated in this tracing. Thirty-second tracings are shown under each respective section before, during, and after washout.

Reversal by Extracellular Ca²⁺ and Mg²⁺ Addition

Both Mg²⁺ or Ca²⁺ added to high K⁺ medium caused [Ca²⁺]_i oscillations and contractions to reappear in cardiomyocytes suppressed by bupivacaine. Mg²⁺ was added in incremental boluses of 1 mM to achieve an increase from 1.2 mM in the medium to 4.2 mM final concentration (reversed 4 of 10 slides). Similarly, when Ca²⁺ was added in 1 mM boluses from 1.5 mM in the medium to 3.5 mM final concentration, there was reversal of suppression in 5 of 10 culture slides. Finally, a single large bolus of 2 mM Mg²⁺ + 2 mM Ca²⁺ reversed the suppression in 9 of 10 culture slides. The reversal rate for 7.6 mM K⁺ with 5.5 μM bupivacaine was 4 of 10 and 5 of 10, respectively, with 4.2 mM Mg²⁺ and 3.5 mM Mg²⁺. Similarly, with 10 mM K⁺, reversal rates with 4.2 mM Mg²⁺ and 3.5 mM Mg²⁺ were 3 of 10 and 3 of 10, respectively. This is shown for the case of the addition of Mg²⁺ in Fig. 6. Cells were more likely to be reversed if treated with combined Mg²⁺ and Ca²⁺ (90% and 70% reversal in 7.6 and 10 mM K⁺, respectively).

Figure 6: Suppressive effect of bupivacaine and reversal by Mg²⁺ in a cardiomyocyte continuously exposed to 7.6 mM K⁺. A tracing is shown in a cardiomyocyte before and after treatment with 5.5 μM bupivacaine and its reversal by Mg²⁺ (2 mM). Note that the pretreatment conditions are different from those in Fig. 1 only by increased extracellular K⁺.

Discussion

The results of this study show that bupivacaine (5.5 μM) and lidocaine (42 μM) produce similar reductions in the oscillatory rate of spontaneously contracting cardiomyocytes. These concentrations were used because they represent the lower limits for drug (central nervous system [CNS]) toxicity (^2–5). The findings are consistent with a slower conduction velocity in anesthetized animals that precedes ventricular arrhythmias, as myocytes specialized in conduction would be expected to respond similarly (^2,19,20). It has been suggested that bupivacaine-induced ventricular fibrillation results from reentrant arrhythmias that arise after bupivacaine slows conduction (²¹). Our data suggest another explanation: bupivacaine toxicity may arise from irregular [Ca²⁺]_i oscillations that coexist with elevated extracellular K⁺ that may be expected to result from seizure activity induced by toxic concentrations of bupivacaine.

There are several reasons that elevated K⁺ might combine with bupivacaine to induce arrhythmias. Arrhythmias arise from alterations in the electrical activity of myocytes (¹⁴), and [Ca²⁺]_i oscillations are linked to the membrane potential via the close juxtaposition of SLM L-type Ca²⁺ channels and the RyR on the SR (²²). Local anesthetics are known to block the K⁺ channels (^9,10) in addition to Na⁺ channels. It is possible that the kinetics of the RyR (adaptation) (²³) and its link to membrane potential are altered with increased extracellular K⁺. Certainly, heart rate increases, alone, result in higher [Ca²⁺]_i (^16–18) that may predispose the heart to abnormal activity, such as delayed and early after-depolarizations leading to ventricular tachycardia and fibrillation.

A possible interaction between local anesthetics and extracellular Mg²⁺ and Ca²⁺ reversal is through a combined interaction at the RyR. All three agents—local anesthetics, Ca²⁺, and Mg²⁺—can regulate [Ca²⁺]_i at the RyR (^12,16,22), and this receptor is crucial for [Ca²⁺]_i oscillations. Ca²⁺ regulates its own release at the RyR, a term called Ca²⁺-induced-Ca²⁺-release (CICR). However, influx of Ca²⁺ through SLM L-type channels is not necessary for [Ca²⁺]_i oscillations in small mammals such as rats (²⁴), unlike the case for large mammals such as humans (²²). In rats, the sarcolemma Na⁺ and Ca²⁺ channels can be blocked with continued contractions; but, when the RyR is blocked, contractions no longer occur (²⁴). Even in humans and other large mammals, blocking CICR from the RyR produces a decrease in the heart rate (^25,26). Thus, the key processes involved with E-C coupling in the rat and small mammals with faster heart rates are those involving CIRC from the RyR and Ca²⁺ reuptake into the SR by Ca²⁺-ATPase (²⁷). Ca²⁺ released from the SR is controlled by kinetics of the RyR (adaptation), and the kinetics may change depending on the [Ca²⁺]_i (^16,25). A combination of adaptation at the RyR (^16–18), K⁺-induced membrane potential abnormalities that feedback to the RyR, and [Ca²⁺]_i modulation of SLM ionic currents (¹⁸) may contribute to bupivacaine toxicity.

The effects of increased extracellular K⁺ highlight differences between lidocaine and bupivacaine on [Ca²⁺]_i oscillations in myocytes. Concentrations of bupivacaine (5.5 μM) that result in a rate decrease produced irregular activity when the extracellular K⁺ was increased to 7.6 mM. However, lidocaine (42 μM) spawned similar irregular oscillations only when the K⁺ was increased to 10 mM, and even at that lethal concentration, the chaotic activity was characterized by a smaller decrease in the Ca²⁺ transients. Thus, bupivacaine-induced chaotic activity is more sensitive to the effects of elevated K⁺ levels, an effect that may be clinically relevant as both anesthetics produce toxic CNS events before cardiac events, and even modest hyperkalemia resulting from muscular contractions associated with convulsions could dramatically potentiate bupivacaine’s toxic effects. Moreover, cardiac ischemia results in a local increase in K⁺ that may not be reflected in serum K⁺ (²⁸), and this increase may contribute to bupivacaine’s cardiotoxicity. The increase in [Ca²⁺]_i oscillation rate (and contractions) under conditions of increased K⁺ induced by bupivacaine and lidocaine contrasts with the in vivo effect of procainamide under hyperkalemia (7.0 mM). Treatment of dogs with procainamide slows through the left ventricle in the presence of hyperkalemia, probably from prolongation of the rested phase of the action potential during which time procainamide binds with the Na⁺ channel (²¹). Although both lidocaine and bupivacaine slowed the oscillation rate at concentrations that approximate CNS toxicity, the consequence of a decrease in beating frequency on the pattern of [Ca²⁺]_i oscillations is not clear. We saw no change in diastolic [Ca²⁺]_i with the concomitant rate slowing. However, rate changes that result from epinephrine or pacing are accompanied by changes in diastolic [Ca²⁺]_i, which would explain the rate-dependent increase in contractility. However, rate increases that occur with temperature changes (either increases or decreases), like local anesthetics, are not accompanied by a corresponding change in diastolic [Ca²⁺]_i.

Mg²⁺ and Ca²⁺ (2 mM each to the extracellular medium) were able to restore spontaneous beating and [Ca²⁺]_i oscillations to myocytes whose oscillations had been completely suppressed by bupivacaine in 7.6 mM K⁺ buffer. Multiple in vivo experimental studies (^2–5,23,29), as well as case reports (¹), have described bupivacaine-induced ventricular arrhythmias that are resistant to treatment, and this finding might offer a potential treatment for local anesthetic-induced toxicity. In such a scenario, these results strongly suggest that any condition that would lead to increased serum K⁺ levels should be treated aggressively and prevented from occurring. Magnesium has a well defined role in the treatment of some ventricular arrhythmias (e.g., torsade de pointes), but the mechanism of reversal by Mg²⁺ and Ca²⁺ is unclear. Some studies suggest that bupivacaine’s cardiotoxicity derives from a blockade of dihydropyridine-sensitive Ca²⁺ channels (^3,8), especially at larger bupivacaine plasma concentrations. It is possible that this explains the findings herein, as Mg²⁺ is known to compete with Ca²⁺ binding at these sites, but both Mg²⁺ and local anesthetics also bind to the RyR. Certainly, the reversal of bupivacaine suppression by Mg²⁺ should be experimentally verified in vivo before clinical use, as it is possible that cardiosuppressive effects of Mg²⁺ will potentiate the suppression induced by bupivacaine and could also potentiate the Ca²⁺ channel blockade by local anesthetics. However, nimodipine has been reported to reduce the toxicity of bupivacaine in rats (³), and the suppression induced by Mg²⁺ should not be greater than a Ca²⁺ channel blocker. Finally, it should be noted that these experiments were performed in cells containing 10% bovine serum, and no corrections were made for protein binding.

In summary, measurements of spontaneous oscillations of cardiac myocytes provide a good in vitro model of toxic local anesthetic effects. Smaller concentrations of both lidocaine and bupivacaine that are associated with convulsions reduce the contraction rate in cardiomyocytes. The major difference between toxic concentrations of lidocaine and bupivacaine is the production of irregular oscillations with subsequent suppression of oscillations seen only with bupivacaine when the extracellular K⁺ was 7.6 mM. Lidocaine produces similar activity only when the extracellular K⁺ was increased to 10 mM. This irregular activity seen with increased extracellular K⁺ and bupivacaine can be partially reversed by Ca²⁺ and Mg²⁺.