Effects of Sevoflurane on Contractile Responses and Electrophysiologic Properties in Canine Single Cardiac Myocytes 

After the approval of the local animal committee, adult mongrel dogs on either sex weighing 8–12 kg were anesthetized with ketamine (25 mg/kg, intramuscular injection). Heparin (5,000 U) and pentobarbital (20 mg/kg) were administered intravenously, and the chest was opened. Hearts were excised rapidly, and Tyrode solution (37 degrees Celsius) was perfused via the coronary artery for 15 min. Tyrode solution contained (in mM): NaCl 135, KCl 4, Na²HPO⁴0.33, Dextrose 10, HEPES 10, MgCl²6H²O 1, and CaCl²1.8 (pH 7.4 by NaOH). Calcium-free Tyrode solution containing 0.05% collagenase (type 1, Sigma, St. Louis, MO) was perfused for 30 min. Hearts were minced into small pieces and agitated in calcium-free Tyrode solution containing 0.05% collagenase and 0.006% trypsin [6](Sigma) until single cells were released. Single cells were resuspended in calcium-free solution and used within 6 h for experiments.

Sevoflurane was administered by bubbling 100% Oxygen²into Tyrode solution through a vaporizer (Sevotec 3, Ohmeda, West Yorkshire, UK). The concentration of vaporized sevoflurane was monitored by an anesthetic gas analyzer (CAPNOMAC, Datax, Helsinki, Finland). The gas mixture was sampled between the vaporizer and the glass filter for bubbling. All the experiments were performed at 37 degrees Celsius.

Effects of sevoflurane on contractile response, membrane potential, membrane ionic currents, and intracellular cyclic adenosine monophosphate (cAMP) concentration were studied over a range of clinically relevant vaporized concentrations (1.0, 2.0, and 4.0%). [7] The values obtained were expressed as mean plus/minus SD. Statistical analyses were performed first by one-way analysis of variance (ANOVA) and, if indicated, were determined between two groups by using t test. A probability value of P < 0.05 was considered significant.

Dispersed cells were placed in a perfusion chamber (37 degrees Celsius) on the stage of an inverted microscope (IMT-2, Olympus, Tokyo, Japan) equipped with Nomarski optics. A hydraulic micromanipulator (MO-103, Narishige, Tokyo, Japan) was used to position a glass microelectrode with tip resistance of 3–5 M omega on the membrane of single ventricular cell. The pipette solution contained (in mM): potassium gluconate 140, ATP 2, HEPES 5, and EGTA 0.15 (pH 7.2 by KOH). We used this pipette solution for the measurements of I^Cabecause the delayed rectified potassium current was very small and negligible.

Contraction of single cell was initiated by extracellular field stimulation (0.2 Hz, 15–20 V, 1 ms). Unloaded cell shortening was measured on-line using a video system. A miniature video camera (FCD-725, Olympus, Tokyo, Japan) was attached to the camera port of the inverted microscope, and the output was fed to a video-edge detector (Crescent Electronics, Sandy, UT), which monitored changes in position of either the right or left cell edge. Records of cell shortening were digitized by the digital recording system (RP-880, NF Electronic Instruments, Yokohama, Japan) and stored on tape. Action potentials induced by intracellular stimulation were recorded with the same digital recording system. To examine the effects of sevoflurane on transmembrane Calcium²+ flux, the cells were bathed in high-Potassium sup +(26 mM) Tyrode solution containing 10 sup -6 isoproterenol. Under these conditions, the membrane potential was depolarized to -40 mV, which inactivated the rapid inward Sodium⁴+ current.

Whole-cell voltage clamp technique was employed to measure ionic currents by using the patch clamp amplifier (Yale Mk-V, New Haven, CT). All data were digitized and recorded by the digital recording system on tape. To measure I^Na, Cadmium²+(0.2 mM) was added to the Tyrode solution to block I^Ca. To achieve an adequate voltage clamp of I^Na, the conventional whole cell recording method, [8] which permits a lower series resistance, was used instead of the perforated patch technique. For this reason, we also excluded cells that exhibited peak control I^Nalarger than 5 nA, because the error in voltage control is proportional to the magnitude of the measured current.

To maintain the cytoplasmic constituents and to avoid “run-down” of Calcium²¹currents, we have used the nystatin perforated patch method as described by Horn and Marty [9] for whole-cell I^Carecordings. Nystatin (Sigma) was dissolved in dimethyl sulfoxide (5 mg/ml) and kept in the freezer. This stock solution was used within 3 days and diluted 1:250 with the pipette solution to give a final concentration of 200 micro gram/ml. This pipette solution containing nystatin was sonicated to ensure complete solubilization. Then, the electrode was filled with this solution. After forming the gigaohm seal with suction, the capacity transient in response to -10 mV pulses was monitored continuously. Within 10–15 min, the capacitative transient increased to a steady level, and the access resistance was decreased (approximately, to 10–30 M omega). Some cells remained stable for more than 1 h in acceptable experiments.

Isolated canine ventricular cells were suspended in Tyrode solution at a density of 1–2 mg protein per 1 ml. The protein concentration of the suspension was estimated by the method of Lowery [10] with bovine serum albumin as the standard. Sevoflurane was vaporized by 100% Oxygen²and bubbled into the cell suspension for 10 min. At the end of the period, 450 micro liter of bubbled cell suspension was sampled. Then 25 micro liter of 2N-HCl and 25 micro liter of 100 mM EDTA were added to the mixture, and the incubation tube was immersed in boiling water for 5 min to extract cAMP, as described by Hazeki and Ui. [11] After a brief centrifugation of the acidified suspension at 3,000 rpm for 10 min, the supernatant was subject to the radio immunoassay by means of cAMP assay kits (Yamasa Syouyu, Chiba, Japan). The cAMP level was expressed as picomoles per milligram of protein.

Sevoflurane dose-dependently decreased contraction induced by electrical stimulation (1.0%, 65.7 plus/minus 6.2; 2.0%, 37.8 plus/minus 5.2, 4.0%, 18.2 plus/minus 4.2% of control; n = 10) in single canine ventricular cells (Figure 1). These depressant effects were maximal within 5–7 min and recovered to control level in 10–15 min after washout of sevoflurane. The durations of cell shortenings were approximately 700 ms, and it was consistent with our previous study with multicellular preparation [5] but was longer than the duration of rat ventricular myocytes. [12].

As shown in Figure 2and presented in Table 1, sevoflurane did not alter the resting membrane potential in normal Tyrode solution. On the other hand, sevoflurane decreased the plateau height of action potentials and shortened the duration of action potentials. We measured the length of action potentials at 30% and 90% of repolarization. Thirty percent of repolarization reflected the plateau phase of the action potential and Calcium²+ influx; on the other hand, 90% reflected the duration of the action potential, and this was used to avoid the interference of artifact caused by the stimulation. At 30% repolarization, the time was reduced to approximately 56% of control by 2.0% sevoflurane. These results suggest that sevoflurane is likely to alter Calcium²+ homeostasis. In high-Potassium⁴Tyrode solution, sevoflurane had no effects on the resting membrane potential. However, plateau height of the action potential was markedly reduced, suggesting an anesthetic-induced depression of voltage-activated Calcium²+ entry (Figure 3and Table 2). Moreover, the duration of the action potential was markedly reduced.

I^Cawas obtained by applying depolarizing pulses from a holding potential of -30 mV to prevent sodium current being elicited. We recognized the current as I^Ca, because it was blocked by applying Cadmium²+(0.2 mM) or nifedipine (10⁶M). As shown in Figure 4, 2% sevoflurane decreased peak I^Caby 38%. Neither the threshold nor the the reversal potential was altered by sevoflurane. The current-voltage relationship for peak I^Cais shown in Figure 4(B). Peak I^Cawas measured at 0 mV test potentials, and the amplitude of peak I^Cawas inhibited by sevoflurane in a dose-dependent manner, as shown in Figure 5. It also appears that, in high concentration of sevoflurane (4.0%), peak currents were inhibited (1.0%, 78.2 + 7.8; 2.0%, 63.4 + 8.9; 4.0%, 45.6 + 12.2% of control; n = 10) to a greater extent than currents at the end of pulse (1.0%, 76.3 plus/minus 5.4; 2.0%, 65.5 plus/minus 6.8; 4.0%, 60.8 plus/minus 6.2% of control), suggesting that sevoflurane may affect the inactivation of I sub Ca. Furthermore, the effect of sevoflurane on I^Cawas not voltage-dependent. To examine whether a use-dependent phenomenon occurred with sevoflurane, I^Cawere measured in the absence of sevoflurane with test depolarization once every 5 s. Depolarizations were then discontinued, sevoflurane was applied, and after a 10 min quiescent period, depolarizing test pulses were resumed. Under these conditions, I^Ca, in the presence of sevoflurane, (4.0%) did not show a decrease in the peak I^Ca, which is consistent with a non-use-dependent reduction in I^Ca(Figure 6). As shown in Figure 6(A), inactivation of I^Caappeared to be less pronounced in the presence of sevoflurane. This inhibition of peak I^Cawas greater than that at the end of pulse.

The effects of sevoflurane on sodium current (I^Na) were examined by applying a series of test pulses from a holding potential of -70 mV to more positive potentials. I^Nawas recognized as it was blocked by TTX (3 x 10⁵M). I^Nawas decreased only by 4.0% sevoflurane (1.0%, 98.6 plus/minus 3.1; 2.0%, 92.6 plus/minus 5.8; 4.0%, 85.2 plus/minus 6.3% of control; n = 5).

Because I^Cain the heart can be modulated by cAMP, we tested the possibility that sevoflurane may depress cAMP leading to a subsequent decrease in I^Caby measuring cAMP activity with radioimmunoassay. cAMP levels were decreased only with 4.0% sevoflurane (control, 1.07 plus/minus 0.16; 1.0% sevoflurane, 1.06 plus/minus 0.17; 2.0%, 1.05 plus/minus 0.16; 4.0%, 0.85 plus/minus 0.09 pmol/mg protein; n = 5). This suggested a negligible role for the depression of cAMP by sevoflurane.

We have used several techniques to examine the mechanical and electrophysiologic effects of sevoflurane on isolated cardiac myocytes.

The video-edge detector technique has been used by a number of laboratories to study the mechanical properties of single cardiac myocytes. [12].

Sevoflurane decreased the plateau and shortened the duration of action potential in Tyrode solution with a more pronounced effect in high Potassium sup + solution. Slow action potentials recorded from partially depolarized myocytes result primarily from the entrance of Calcium²+ into the cell, presumably via L-type Calcium²+ channels. These results are similar to those of halothane, isoflurane, and enflurane in multicellular preparations. [13,14] Furthermore, the voltage clamp technique allowed us to clarify the ionic currents affected at clinically relevant anesthetic concentrations. Our findings are that I^Cabut not I^Nawas affected. Bosnjak et al. reported that both halothane and isoflurane produced a similar degree of inhibition of I^Caelicited by the voltage clamp studies in single guinea pig ventricular cells. [15] Our study is the first to show that sevoflurane produces a dose-dependent depression of peak Iodine^Calciumwithout shifting the current-voltage relationship for Calcium sup 2+ channel activation. On the other hand, some changes in the kinetics of Iodine^Calciuminactivation were noted at high dose of sevoflurane. Further studies will be required to identify the biophysical characteristics of these effects. Therefore, the negative inotropic effect of sevoflurane on the myocardium are probably related, at least in part, to the depression of Iodine^Calciumsimilar to those of other inhalational agents. [16–18] In addition, the fact that sevoflurane did not show the use-dependent block (accentuation of drug inhibition by repetitive membrane depolarization) indicated that the inhibitory blocking mechanism of Iodine^Calciumis different from the effect of thiopental [19] or D600, an organic Calcium²+ channel blocker. [20] However, these results do not exclude the possibility that the inhibitory effects on myocardial contractility by sevoflurane may be due to effects on Sodium sup +/Calcium²+ exchange or intracellular Calcium²+ stores. [4] The effects of sevoflurane on the action potential are consistent with a reduction in Iodine^Calcium. In our observation, canine ventricular myocytes had little delayed outward potassium current, and further examination is required. Moreover, Iodine^Sodiumwas depressed only by 4.0% sevoflurane, indicating that cardiac Sodium sup + channel depression has a relatively minor or negligible role in the negative inotropic effect of the anesthetic.

We reasoned that, if sevoflurane diminishes cAMP levels, this action might account for the anesthetic-induced depression of both Iodine^Calcium[21] and myocyte contractility. However, use of a radioimmunoassay indicated that cAMP levels were unchanged except for the high concentration of sevoflurane (4.0%) employed in this study. These results differ from those in response to halothane. [22].

In summary, our study indicates that the decrease of myocardial contractile response in single canine cardiac cells is related, at least in part, to the inhibition of Iodine^Calciumat the sarcolemma. The effects of sevoflurane on intracellular cAMP levels and Iodine^Sodiumwere minimal.

The authors thank Dr. Hamid I. Akbarali, for his comments.