In Vitro Negative Inotropic Effect of Low Concentrations of Bupivacaine Relates to Diminished Ca²⁺ Sensitivity but Not to Ca²⁺ Handling or β-Adrenoceptor Signaling

Although technical advances, such as the use of ultrasound-guided injection, have led to a reduced incidence of local anesthetic systemic toxicity,¹  such critical incidents are still estimated at 0.04 to 1.8 per 1,000 peripheral nerve blocks.^2,3  Systemic toxicity of local anesthetics is predominantly complicated by their myocardial toxicity.

At a cellular level, local anesthetics do not only block sodium,⁴  potassium,⁵  and Ca²⁺ channels,⁶  possibly inducing arrhythmia, but also reduce myocardial contractility.⁷  This negative inotropic effect of local anesthetics has been attributed to inhibition of intracellular Ca²⁺ release,⁸  decreased sensitivity of myofilaments to Ca²⁺,⁹  inhibition of 3′,5′-cyclic adenosine monophosphate production,¹⁰  and interference with mitochondrial bioenergetics.¹¹  Yet these concentration-dependent in vitro effects of local anesthetics like bupivacaine have been described at concentrations far higher than local anesthetic concentrations described in clinical toxicity.¹² 

Given this discrepancy in concentrations measured in vitro and in clinical studies, we hypothesized that some of the previously described mechanisms of toxicity may not play a role in negative inotropy at clinically relevant concentrations. Therefore, we evaluated the negative inotropic effect of bupivacaine by simultaneously measuring action potentials and force of contraction in isolated intact heart muscle preparations. Additionally, we examined the same concentration effect of bupivacaine on L-type Ca²⁺ channels, β-adrenoceptor signaling, sarcoplasmic reticulum function, and myofilament Ca²⁺ sensitivity to pinpoint the negative inotropic effect of bupivacaine in skinned muscle fibers and isolated cardiac myocytes.

The investigation conformed to the guide for the care and use of laboratory animals published by the National Institutes of Health (Bethesda, Maryland; Publication No. 85–23, revised 2011, published by National Research Council, Washington, D.C.). The experimental procedures were in accordance with the German Law for the Protection of Animals and with the guidelines of the European Community. Euthanization of animals and harvesting of organs for experiments were performed following approved protocols accepted by the Regierungspräsident Dresden, Dresden, Germany (permit no. 24D-9168.24-1/2007–17) and accepted by the Ministry of Science and Public Health of the City State of Hamburg, Hamburg, Germany (Organ extraction permit No. 653). The experiments were performed in Dresden and Hamburg; therefore, approval was obtained from the respective local authorities. Contractility analysis of papillary muscles, action potential, and L-type Ca²⁺ current recordings were performed with guinea pig preparations, as guinea pigs have a cardiac physiology that is relatively close to human physiology. Single cell contractility and Ca²⁺ transients, 3′,5′-cyclic adenosine monophosphate concentration, and myofilament Ca²⁺ sensitivity were analyzed in mouse cells, because Fura-based measurements of intracellular Ca²⁺ changes gave more reliable results in mouse than in guinea pig cells with our established protocols. The Förster Resonance Energy Transfer–based 3′,5′-cyclic adenosine monophosphate measurements were done with cells isolated from mice transgenically expressing a 3′,5′-cyclic adenosine monophosphate sensor, which gives better signals than using transfected sensors. In total, 34 guinea pigs and 20 mice were euthanized for this study.

Right ventricular papillary muscles from guinea pig hearts were mounted in an organ bath and continuously perfused with oxygenated Tyrode’s solution of the following composition: 126.7 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl₂, 1.05 mM MgCl₂, 22 mM NaHCO₃, 0.42 mM NaHPO₄, and 5 mM glucose. (Unless mentioned otherwise, all materials were purchased from Sigma-Aldrich, Germany.) After equilibration with carbogen (95% oxygen and 5% carbon dioxide), the pH of the solution was 7.4 (36°C). One end of the muscle was pinned to the floor of the chamber; the free end was connected to a force transducer (AE 801, SensoNor, Germany) with a loop of silk thread. The preparations were stimulated electrically via silver/silver chloride electrodes at a regular frequency of 1 Hz. The muscles were allowed to equilibrate for at least 90 min before intracellular action potentials were recorded with conventional glass pipettes filled with 2.5 M KCl solution (tip resistances of 30 to 40 MΩ). Drugs were added to the superfusion solution from concentrated stock solutions to yield the final concentration. Action potentials were stored on a personal computer, and the following parameters were analyzed offline: resting membrane potential, force of contraction, action potential duration at 20%, 50%, and 90% of repolarization, and maximum upstroke velocity. All data acquisition and analysis were carried out with the ISO 2 system (MFK, Germany). Right ventricular papillary muscles were chosen because they more reliably allow parallel recordings of both contractility and action potentials than left ventricular preparations.

After an equilibration period, bupivacaine was added to the superfusion solution from concentrated stock solutions to yield the final concentrations (100 nM to 100 µM). Action potentials from stable preparations were recorded 30 min after drug addition before the individual drug concentrations were increased. At the end of each experiment, a washout period of 60 min was performed.

A free wall of right ventricles was prepared from guinea pig hearts (male Dunkin Hartley Crl. HA guinea pigs of 280 to 350 g body weight; Charles River, Germany) and mounted in a 50-ml organ bath, with one end of the muscle clamped down between a pair of platinum electrodes and the free end tied to a strain gauge. Right ventricles were stimulated at 2 Hz, 10% above stimulation threshold. Resting tension was adjusted to yield approximately half-maximum active force development. The organ bath contained the above mentioned oxygenated Tyrode’s solution at 37°C.

First, preparations underwent an equilibration period of 60 min. An incubation time of 60 min with the α-adrenoceptor blocker phenoxybenzamine (6 µM) followed. Phenoxybenzamine is an unselective α-adrenoceptor antagonist and increases release of noradrenaline.¹³  Trabeculae were washed from released catecholamines and left to stabilize during an additional period of 30 min. In a first set of experiments, the negative inotropic effect of bupivacaine was analyzed by increasing concentrations in a cumulative manner. In a second protocol, right ventricles were pretreated with the estimated effective concentration producing a 50% inhibitory effect (5 µM) of bupivacaine for 30 min, and subsequently a cumulative concentration response curve (0.1 nM to 1 µM) for the β-adrenoceptor agonist isoprenaline was measured. At the end of each experiment, the adenylyl cyclase stimulator forskolin (1 µM) and calcium chloride (8 mM) were added.

Ventricular myocytes were isolated from guinea pig hearts as described before.¹⁴  In brief, tissue was cut into small pieces and transferred to Tyrode’s solution containing 0.54 mg/mL collagenase type I (Worthington, USA) and 4.6 mg/mL protease type XXIV (Sigma-Aldrich) for digestion and gently stirred for 45 min under light perturbation. The myocytes were stored in storage solution containing 20 mM KCl, 10 mM KH₂PO₄, 10 mM glucose, 70 mM K-glutamate, 10 mM β-hydroxybutyrate, 10 mM taurine, 10 mM EGTA, and 1 mM albumin, pH 7.4, at 37°C until final use. Only striated, rod-shaped myocytes were used. The cells were studied in a small chamber continuously perfused at a rate of 1.6 ml/min and a temperature of 37°C. A system for rapid solution changes allowed application of drugs in the close vicinity of the cells (Cell Micro Controls, USA; ALA Scientific Instruments, USA). Currents were measured in the whole cell configuration of the single electrode voltage clamp technique as described elsewhere¹³  using the EPC-7 patch-clamp amplifier (List Electronics, Germany) and the ISO 2 software (MFK, Germany) for data acquisition and analysis. Borosilicate glass pipettes had a tip resistance of 2 to 4 MΩ. Membrane capacitance was measured with a small depolarizing ramp pulse from 0 to –2 mV (duration, 5 ms). We allowed 5 min after establishing the whole cell configuration for equilibration before starting to record. L-type Ca²⁺ current was activated with a 200-ms test pulse to +10 mV from a holding potential of –80 mV every 2 s. The current amplitude was determined as the difference between peak inward current and current at the end of the depolarizing step; contaminating potassium currents were blocked by replacing potassium with cesium. The composition of the superfusion solution was 120 mM tetraethylammonium chloride, 10 mM CsCl, 10 mM HEPES, 2 mM CaCl₂ 2, 1 mM MgCl₂, and 20 mM glucose, pH 7.4 (adjusted with cesiumhydroxide). The pipette solution (pH 7.2) included 90 mM cesium methanesulfonate, 20 mM CsCl, 10 mM HEPES, 4 mM Mg-ATP, 0.4 mM Tris-GTP, 10 mM EGTA, and 3 mM CaCl₂, with a calculated free Ca²⁺ concentration of ~60 nM (computer program EQCAL, Biosoft, United Kingdom).

Cardiomyocytes were isolated from transgenic mice ubiquitously expressing the 3′,5′-cyclic adenosine monophosphate Förster Resonance Energy Transfer sensor Epac1-camps (produced by V.N.) via Langendorff perfusion and enzyme digestion as described previously.¹⁵  After stepwise recalcification to 1 mM of Ca²⁺ cells were plated onto laminin-coated glass coverslips (Invitrogen, USA) and allowed to settle down for 1 h before commencement of Förster Resonance Energy Transfer measurements as follows: cells with and without 3 min pretreatment with 5 µM bupivacaine were stimulated with 100 nM isoprenaline followed by simultaneous stimulation with 100 µM of the unspecific phosphodiesterase inhibitor 3-isobutyl-1-methylanxthine and 10 µM of the adenylyl cyclase activator forskolin. Additionally, cells were treated with 5 µM bupivacaine only followed by 100 µM 3-isobutyl-1-methylanxthine and 10 µM forskolin. A detailed description of the Förster Resonance Energy Transfer setup and analysis was published previously.¹⁵ 

Ventricular cardiomyocytes were isolated from wild-type Black Swiss mouse hearts as described before.¹⁶  In brief: After mice had been euthanized by cervical dislocation under carbon dioxide anesthesia, hearts were excised and then retrogradely perfused through the aorta with a Langendorff perfusion system at 37°C. Perfusion buffer contained 113 mM NaCl, 4.7 mM KCl, 0.6 mM KH₂PO₄, 0.6 mM Na₂HPO₄, 10 mM KHCO₃, 1.2 mM MgSO₄, 12 mM NaHCO₃, 30 mM taurine, 5.55 mM glucose, 10 mM 2,3-butanedione monoxime, and 10 mM HEPES, pH 7.46. Hearts were first perfused with this buffer for 6.5 min, and then 0.075 mg/mL Liberase TM (Roche Diagnostics, Germany) and 12.5 µM calcium chloride were added for 6 to 7 min. Then the ventricles were separated from the atria, and digestion was stopped by addition of fetal calf serum. Ventricles were dissected into small pieces, and single cardiomyocytes were obtained by pipetting steps. After stepwise reintroduction of calcium chloride to 1 mM, cells were transferred to modified Tyrode’s solution (135 mM NaCl, 4.7 mM KCl, 0.6 mM KH₂PO₄, 0.6 mM Na₂HPO₄, 1.2 mM MgSO₄, 1.5 mM CaCl₂, 20 mM glucose, and 10 mM HEPES, pH 7.46). Cells were then incubated for 15 min with 0.6 µM Fura-2 am (Invitrogen), followed by a 20-min washing period, which also allowed deesterification of Fura-2.

In baseline conditions, cardiomyocytes were perfused with modified Tyrode’s solution at 37°C and electrically stimulated at 1 Hz (4-ms pulses, 10 V). Sarcomere shortening was detected with a video-based system, and Ca²⁺ transients were detected by alternate excitation at 340 and 380 nm while emitted light was recorded at 510 nm with a photomultiplier tube. Data were recorded and analyzed with IonWizard software. Equipment and software were purchased from IonOptix (USA).

For evaluation of the bupivacaine effect, cells were recorded for 1 min at baseline conditions. Then 5 µM bupivacaine was added to the perfusion, and cells were recorded further for up to 5 min. For evaluation of the influence of bupivacaine on the effect of isoprenaline, cells were preincubated with 5 µM bupivacaine for 5 min, then electrically stimulated at 1 Hz while being perfused with buffer containing 5 µM bupivacaine. After 1 min, 100 nM isoprenaline was added to the perfusion, and cells were recorded until the full inotropic effect of isoprenaline was reached.

Force measurements on skinned ventricular mouse muscle strips were done as previously published.¹⁶  In brief, muscle strips were skinned overnight at 4°C with 1% Triton X-100 in relaxing solution (–log[M] of Ca²⁺ that contained 5.89 mM Na₂ATP, 14.5 mM creatine phosphate, 6.48 mM MgCl₂, 40.76 mM potassium propionate, 100 mM N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid, and 7 mM EGTA, pH 7.1). Skinned strips were mounted between a length controller and a force transducer (Test System 1400A, Aurora Scientific, Canada) and stretched until they started to develop force in activating solution (EGTA buffer with pCa 4.5). Ca²⁺ force relationships were measured from pCa 9 to pCa 4.5. The effect of 5 µM bupivacaine was analyzed in paired measurements (with and without bupivacaine) on every strip. In every second experiment, bupivacaine buffer was tested before control condition was measured to account for time-dependent loss of force.

All drugs were purchased from Sigma-Aldrich. We used racemic bupivacaine in all experiments.

All data are expressed as means ± SD. Curve fitting and statistical analyses were performed using the GraphPad 5.0 Prism Software (GraphPad Software, USA). Values for half-maximum effect or inhibition were estimated from fitting a Hill function with variable slopes to concentration–response curves and are given as EC₅₀ or 50% inhibitory effect values (M), respectively. Two-way ANOVA with Bonferroni posttest for multiple comparisons was used to test for statistically significant differences between groups. Parameters of concentration–response curves (top, bottom, 50% inhibitory effect /EC₅₀/–log[M] of Ca²⁺) for bupivacaine, isoprenaline, and Ca²⁺-evoked tension were compared by the extra sum-of-squares F test after nonlinear regression fit for sigmoidal curves.

Samples were randomly assigned to bupivacaine treatment or control groups, except for skinned muscle experiments, in which all samples were measured in both conditions. Experimenters were not blinded. Sample sizes were chosen by previous experience of sufficient amount of data. The methods described above were well established in the respective laboratory facilities, and all experiments were planned with the reported parameters as primary endpoints.

Any reduction in sodium influx should lead to a reduced Ca²⁺ load via the sodium/Ca²⁺ exchanger. Accordingly, the simplest explanation of bupivacaine-induced negative inotropy would be a reduction in sodium current. Potency values for bupivacaine to block cardiac sodium channels vary widely and are hard to interpret since drug binding depends on resting membrane potential, stimulation rate, and extracellular sodium concentration.^17,18  Since there is an almost linear relation between sodium currents and upstroke velocity,¹⁹  we simultaneously measured action potential and force in intact muscle preparations of guinea pig ventricles exposed to a wide range of bupivacaine concentrations.

In time-matched controls, all parameters remained stable more than 4 h. In bupivacaine-treated muscles, there was a clear reduction of force at 5 µM bupivacaine (time-matched controls 174 ± 44 µN, bupivacaine 98 ± 25 µN; 50% inhibitory effect, 5.1 µM), a concentration reported as clinically relevant (fig. 1, A and D).^12,20  This reduction of force was reproducible in mouse ventricular muscle strips, where 5 µM bupivacaine reduced force by about 40% (from 0.98 ± 0.7 to 0.61 ± 0.6 mN, n = 10) while time-matched controls did not show reduction of force (data not shown). However, the negative inotropic effect of 5 µM bupivacaine in guinea pig muscles was not accompanied by a decline in maximum upstroke velocity (fig. 1, B and D). Similarly, action potential duration was not changed by 5 µM bupivacaine (fig. 1, C and D). At higher concentrations of up to 300 µM, bupivacaine further diminished contractile force (from 173 ± 119 µN at baseline to 28 ± 13 µN), reduced maximum upstroke velocity (from 181 ± 34 V/s to 116 ± 21 V/s; 50% inhibitory effect, 72 µM) and shortened action potential duration at 50% repolarization (from 97 ± 21 ms to 65 ± 16 ms; 50% inhibitory effect, 263 µM) and action potential duration at 90% repolarization (from 169 ± 19 ms to 135 ± 26 ms, n = 7; 50% inhibitory effect, 230 µM). While effects on action potential duration and maximum upstroke velocity were reversible, force remained low even after 60 min of washout.

To estimate if bupivacaine directly affects Ca²⁺ influx, we measured L-type Ca²⁺ currents in isolated guinea pig cardiomyocytes exposed to bupivacaine. Baseline L-type Ca²⁺ current density was 12.6 ± 6.3 pA/pF (n = 20) in ventricular cardiomyocytes. L-type Ca²⁺ currents showed the typical rundown over time (fig. 2). When compared to TMC, bupivacaine had no effect on L-type Ca²⁺ currents even up to a concentration of 100 µM (fig. 2), a concentration that induces maximum negative inotropy (fig. 1A).

Clinically, catecholamines are frequently employed to overcome negative inotropy by bupivacaine. Therefore, we evaluated whether bupivacaine has an impact on β-adrenoceptor–evoked inotropy. We measured concentration–response curves for the positive inotropic effect of isoprenaline in guinea pig right ventricular muscle strips under control conditions and in the presence of 5 µM bupivacaine.

As seen before in right ventricular papillary muscles (fig. 1), 5 µM bupivacaine reduced force significantly by about 50% (time-matched controls, 0.62 ± 0.29 mN, bupivacaine, 0.24 ± 0.12 mN; fig. 3A). Subsequent isoprenaline challenge could not completely overcome bupivacaine-induced negative inotropy. As a result, maximum force development was reduced significantly by about 30% in the presence of 5 µM bupivacaine (fig. 3B). EC₅₀ of isoprenaline tended to be slightly but nonsignificantly lower in the presence of 5 µM bupivacaine (130 ± 194 nM in time-matched controls vs. 301 ± 373 nM with 5 µM bupivacaine, n = 17 vs. 18).

To test if bupivacaine interferes with other parts of the β-adrenoceptor pathway, we measured whether 5 µM bupivacaine reduces the isoprenaline response of L-type Ca²⁺ currents (fig. 2C). In time-matched controls, 1 µM isoprenaline increased L-type Ca²⁺ currents by 83 ± 42% (n = 12; fig. 2C). Bupivacaine (5 µM) did not affect basal current or isoprenaline-induced increase of L-type Ca²⁺ currents (88 ± 72%, n = 16).

The finding that bupivacaine does not reduce maximum effects of isoprenaline on L-type Ca²⁺ currents does not rule out general effects of bupivacaine on 3′,5′-cyclic adenosine monophosphate generation. Concentrations of 3′,5′-cyclic adenosine monophosphate are regulated spatially,¹⁵  and L-type Ca²⁺ currents increase may reflect elevation of 3′,5′-cyclic adenosine monophosphate only in the subsarcolemmal space. In order to investigate whether bupivacaine has a direct effect on global 3′,5′-cyclic adenosine monophosphate concentration, Förster Resonance Energy Transfer measurements were performed with isolated cardiomyocytes from transgenic mice expressing the cytosolic 3′,5′-cyclic adenosine monophosphate sensor Epac1-camps (fig. 4).

Bupivacaine (5 µM) alone did not affect basal 3′,5′-cyclic adenosine monophosphate concentration (n = 7), nor did bupivacaine significantly reduce effects of subsequent exposure to isoprenaline to stimulate β₁-adrenoceptors (12 ± 21%, n = 14; vs. 9 ± 16%, n = 12) or 3-isobutyl-1-methylanxthine plus forskolin to provoke maximum 3′,5′-cyclic adenosine monophosphate concentration (25 ± 36%, n = 14; vs. 25 ± 20%, n = 12; fig. 4, B–D).

To investigate whether bupivacaine mediates its negative inotropic effect by interaction with intracellular targets involved in electromechanical coupling, we analyzed the speed of relaxation in guinea pig heart muscle strips as an indirect measure of Ca²⁺ handling. To minimize time-dependent effects, we exposed muscles to one concentration only. Time to 50% relaxation as a surrogate parameter for phospholamban and/or troponin I function was not affected by 5 µM bupivacaine. Furthermore, isoprenaline led to a similar acceleration of relaxation in TMC- and bupivacaine-treated preparations (fig. 5). Next we addressed more directly whether bupivacaine affects intracellular Ca²⁺ handling. Therefore, we measured Ca²⁺ transients in isolated mouse ventricular cardiomyocytes. Bupivacaine (5 µM) did not change Ca²⁺ transients (fig. 6, A, B, and F) but again showed a negative inotropic effect (sarcomere shortening before bupivacaine, 8.4 ± 3%, after bupivacaine, 6.4 ± 3.3%; fig. 6E). Also, isoprenaline-induced increases in Ca²⁺ transients were not reduced by 5 µM bupivacaine (fig. 6, C, D, and H). In contrast to intact muscles, maximum increase in force generation induced by isoprenaline was not influenced by the presence of bupivacaine (fig. 6G).

As intracellular Ca²⁺ transients were not affected by 5 µM bupivacaine, negative inotropy might be the result of direct effects on myofilaments. Therefore, we measured Ca²⁺-evoked force in skinned mouse ventricular muscle strips. Bupivacaine (5 µM) treatment shifted the Ca²⁺-tension force curve to the right, leading to a significant change in pCa₅₀ (negative logarithm of the half-maximum effective Ca²⁺ concentration in M) from 5.66 ± 0.05 in control to 5.56 ± 0.03 in the presence of 5 µM bupivacaine (fig. 7).

The main finding of this work is that negative inotropic actions of clinically relevant concentrations of bupivacaine are mediated by effects on Ca²⁺ sensitivity. Effects on Ca²⁺ loading or Ca²⁺ handling by bupivacaine do not seem to be relevant at these concentrations.

Any reduction in sodium influx will reduce the amount of intracellular Ca²⁺ available to evoke contraction.²¹  Therefore, it seems obvious to suspect negative inotropy of local anesthetics related to an interaction with their main target, the sodium channels.²²  Since the relationship between maximum upstroke velocity of the action potential and the sodium current is nearly linear,¹⁹  maximum upstroke velocity was used as a surrogate for peak sodium current. We did not find evidence that negative inotropy of 5 µM bupivacaine is connected to Na⁺ block, as this concentration left maximum upstroke velocity unaffected. Bupivacaine used in a similar concentration of 3.5 and 5 µM reduced force of contraction and maximum upstroke velocity by about 40 to 50% in guinea pig papillary muscles.^4,23  However, in another study that used the same cardiac preparation, 4 µM bupivacaine depressed force by about 40%, while maximum upstroke velocity was only marginally reduced.²⁴  The finding that bupivacaine in concentrations ~5 µM only marginally depressed maximum upstroke velocity was also found in other studies,^25,26  but simultaneous effects on force were not reported. In isolated rabbit atria, 1.5 µM bupivacaine depressed force by about 60%.²⁷  It should also be noted that during extensive washout (1 h), maximum upstroke velocity almost completely recovered, whereas force did not (compare fig. 1, A and B). In skinned preparations, decrease of maximum force by bupivacaine was not significant, and force could be fully restored after washing, arguing for intracellular accumulation of bupivacaine in intact cells. Taken together, our findings as confirmed by published data strongly suggest that bupivacaine effects on maximum upstroke velocity and force do not necessarily go hand in hand, questioning whether negative inotropy of low micromolar concentrations relates to sodium channel block.

Relevant for the negative inotropy could be a block of L-type Ca²⁺ current, as suspected from early studies that used maximum upstroke velocity of slow action potential as a surrogate parameter for Ca²⁺ currents or more direct current measurements using the sucrose gap technique.^28,29  Two patch clamp studies with bupivacaine showed block of L-type Ca²⁺ current only with concentrations above 10 µM,³⁰  or about 50% of block with concentrations as low as 5 µM.³¹  In this context, it is noteworthy that L-type Ca²⁺ current notoriously exhibits rundown,³²  hence drug effects can be estimated only in comparison with time-matched controls, especially when long exposure times are required for establishing a drug equilibrium. The two aforementioned studies do not provide such information. Our own data suggest that low micromolar concentrations of bupivacaine do not affect L-type Ca²⁺ current and therefore do not contribute to negative inotropy via block of this current.

There are some findings indicating the importance of the β-adrenoceptor signaling pathway for bupivacaine’s negative inotropy. Interestingly, force development was depressed by bupivacaine with an almost identical concentration dependency as measured for norepinephrine release.³³  The depression of the norepinephrine release may contribute to in vivo effects of bupivacaine. Data presented in fig. 1 were recorded in the absence, those in Fig. 2 in the presence of phenoxybenzamine. We would not expect relevant contribution of release inhibition of norepinephrine by bupivacaine to the negative inotropic effects in vitro, since pretreatment with phenoxybenzamine does not abolish bupivacaine negative inotropy.

Many local anesthetics can inhibit binding to β-adrenoceptors.³⁴  For methodologic reasons, only data for human β₂-adrenoceptors are available. The negative logarithm of the half-maximum inhibitory concentration in M value for bupivacaine is 4.1.³⁴  If we would extrapolate that value to β₁-adrenoceptors (assumed to be the prominent β-adrenoceptor population in guinea pig hearts), we would expect from the application of Schild’s law^16,35  that 5 µM bupivacaine would produce a negligible (not more than 0.02 log unit) leftward shift of the concentration–response curve for isoprenaline. Our experimental findings are in line with this prediction.

More important, bupivacaine in concentrations as low as those used in our study (3.5 µM) suppress both basal and catecholamine-stimulated 3′,5′-cyclic adenosine monophosphate production in human lymphocytes.¹⁰  The latter effect is in a noncompetitive manner, closely resembling the findings of our study, namely a reduced maximum inotropic response by catecholamines in the presence of bupivacaine.¹⁰  We could not confirm the results obtained in lymphocytes. One reason for this discrepancy may relate to differences in β-adrenoceptor subtypes involved in the 3′,5′-cyclic adenosine monophosphate signaling in cardiomyocytes and lymphocytes. While in mice and guinea-pig cardiomyocytes, β₁-AR but not β₂-AR mediate catecholamine response on force and on L-type Ca²⁺-current,^36,37  lymphocytes are exclusively dependent on β₂-adrenoceptors.³⁸  However, in contrast to lymphocytes, basal 3′,5′-cyclic adenosine monophosphate concentration was also unaffected by bupivacaine (5 µM). Therefore, marked differences exist between lymphocytes and cardiomyocytes regarding 3′,5′-cyclic adenosine monophosphate regulation. Negative inotropy of 5 µM bupivacaine cannot be explained by interactions with β₁-adrenoceptors or 3′,5′-cyclic adenosine monophosphate generation.

Since block of sodium or Ca²⁺ currents could not serve to explain the negative inotropic effect of low (micromolar) concentrations of bupivacaine,³⁰  we suspected impairment of intracellular Ca²⁺ handling to play a role. Millimolar concentrations of bupivacaine bind to ryanodine receptors⁸  and can affect Ca²⁺ regulation in skeletal muscle.³⁹  Here we found that 5 µM bupivacaine reduced cell shortening but not Ca²⁺ transients. This is in line with a study in ferret cardiomyocytes, where 10 µM bupivacaine impaired cell shortening without any effect on Ca²⁺ transients.⁶  This also shows that the minor, nonsignificant effects of 5 µM bupivacaine on sodium and Ca²⁺ currents are indeed negligible concerning inotropy, as they apparently do not influence intracellular Ca²⁺ concentration.

While millimolar concentrations of bupivacaine shift the Ca²⁺ tension curve to the left,⁴⁰  more clinically relevant concentrations (10 to 100 µM) may result in the opposite effect, therefore desensitizing the muscle to Ca²⁺ and reducing maximum tension.⁹  With a 0.1 log unit shift and reduction of maximum force in ex vivo experiments, we could reproduce the findings of Mio et al.⁹  While this shift might seem small, it is still likely to have a profound impact on cardiac contractility.⁴¹  In other experimental work, a reversal of the negative inotropic effects of ropivacaine was effectively achieved by the application of the Ca²⁺ sensitizer levosimendan.⁴² 

From our previous work, we estimate about sixfold accumulation of bupivacaine in cardiomyocytes.⁴³  Therefore, in our experiments, myofilaments of intact cells should be exposed to higher concentrations of bupivacaine than skinned fibers. Interestingly, the effect of bupivacaine on maximum force is less pronounced in skinned muscle preparations and disappears after washout, which indicates some kind of nonreversible accumulation of bupivacaine in intact cells. However, we would not expect drastic differences, since the shift of the Ca²⁺ tension curve with 100 µM bupivacaine was only slightly more pronounced compared with 10 µM.⁹ 

We provide evidence that negative inotropy by low concentrations of bupivacaine is mediated by a reduction in Ca²⁺ sensitivity of the myofilaments. Current strategies of acute treatment of cardiodepression caused by local anesthetics are based on application of catecholamines, thereby increasing transsarcolemmal Ca²⁺ influx and intracellular Ca²⁺ release, but these two were not affected by low concentrations of bupivacaine. Instead, bupivacaine decreased Ca²⁺ sensitivity. We do not know whether in such a scenario catecholamines decrease Ca²⁺ sensitivity further and whether the decrease in maximum β-adrenoceptor–evoked inotropy by low concentrations of bupivacaine may pose a clinically relevant problem. Nevertheless, limiting the maximum inotropic effect of isoprenaline could suggest that catecholamines may not be perfectly suited to treat bupivacaine-induced cardiodepression. In theory, Ca²⁺ sensitizers⁴⁴  could be useful. From a more general perspective, our findings underscore the relevance of myofilament Ca²⁺ sensitivity for drug-induced negative inotropy.

Heart tissue of different species might react differently to bupivacaine, and translatability of the findings to an in vivo situation in humans has, of course, to be tested. We provide evidence that the negative inotropic action of bupivacaine is mediated via interaction with myofilaments. However, it should be noted that other mechanisms are also conceivable. Bupivacaine interacts with lipid membranes, modifying their fluidity.⁴⁵  Impairment of mitochondrial bioenergetics could contribute to negative inotropy of bupivacaine. No direct measurements of possible bupivacaine effects on mitochondrial function were performed in our experiments and therefore cannot be completely excluded as a mechanism. However, previous data show that bupivacaine concentrations necessary for a reduction in mitochondrial function were 50 times higher than the concentrations used in our experiments.¹¹  On the other hand, low concentrations of bupivacaine similar to those we have used here showed strong yet transient effects on mitochondria morphology and reduced metabolic activity.⁴³ 

We cannot provide a reasonable explanation for why inotropic responses by isoprenaline are not reduced in isolated cardiomyocytes. Isolated cardiomyocytes contracting in an unloaded fashion are rather fragile, especially when challenged with catecholamines. Possibly, discrimination of small differences may be limited in such a model.

Our study shows that the reduction in myofilament sensitivity to Ca²⁺ caused by low micromolar concentrations of bupivacaine might be sufficient to induce a significant negative inotropic effect. Effects of bupivacaine on major ion currents are negligible at these concentrations.

The authors thank Romy Kempe, Constanze Fischer, Trautlinde Thurm, and Annegret Häntzschel, technical assistants, Department of Pharmacology and Toxicology, Medical Faculty Carl Gustav Carus, Dresden, Germany, for providing valuable assistance. The authors thank Silke Skytte Johannsen, Ph.D., SJ Consulting, Hamburg, Germany, for excellent help with statistical analysis.

The study was supported in part by an intramural grant (MedDrive program, Dresden, Germany) funded by the Medical Faculty Carl Gustav Carus, Technical University Dresden, Dresden, Germany.

The authors declare no competing interests.