Effects of Volatile Anesthetics on Sarcolemmal Calcium Transport and Sarcoplasmic Reticulum Calcium Content in Isolated Myocytes

All experimental procedures were reviewed and approved by the Animal Care and Use Committee of the Mayo Foundation (Rochester, MN). Protocols were completed in accordance with National Institutes of Health guidelines and in accordance with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources Commission on Life Sciences, National Research Council). Adult male ferrets (mean weight, 1.2 kg; Marshall Farms, North Rose, NY) were anesthetized with 3% halothane in 100% oxygen, and their hearts were rapidly excised. The aorta was then cannulated and the heart perfused with physiologic salt solution (PSS) bubbled with 100% oxygen. The composition of this solution was 117.5 mm NaCl, 5 mm KCl, 1.0 mm MgCl, 1.0 mm CaCl², 5 mm MOPS, 20 mm Na-acetate, and 10 mm glucose (pH 7.40).

Ferret ventricle was chosen because the balance between SERCA, NCX, and PMCA is similar to human ventricle. 20,28Functionally, this results in both rest potentiation of SR Ca²⁺release and rest decay of the SR Ca²⁺content.

To isolate single myocytes, the right ventricular free wall was removed en bloc  and rinsed for 5 min in PSS containing no added Ca²⁺. The tissue was then minced into small pieces and gently stirred in a Teflon beaker containing a dispersion solution 29composed of 75 mm NaCl, 2.4 mm KCl, 1.0 mm MgCl, 10 mm HEPES, 58 mm sucrose, 10 mm dextrose, 5 mm NaHCO³, and 2.5 mm l-glutamic acid (pH 6.96) with 160 μg/ml type II collagenase and 0.6 mg/ml bovine serum albumin. The solution was decanted and replaced four times at 15-min intervals. Myocytes were harvested by low-speed centrifugation and then resuspended in culture medium (Medium 199).

Single isolated myocytes were loaded for 2–3 min with the acetoxymethyl ester of the fluorescent Ca²⁺indicator fluo-3 (Molecular Probes, Eugene, OR) at an extracellular concentration of 1 μm. Fluo-3 was chosen for the intracellular Ca²⁺indicator because it undergoes an approximately 100-fold increase in fluorescence on Ca²⁺binding. This property makes it particularly well suited for measuring the kinetics of Ca²⁺transients because it improves the signal-to-noise ratio. The rate and extent of shortening of fluo-3–loaded cells was not different from unloaded controls. A comparison of fura-2 and fluo-3 in pilot studies indicated that the two indicators yielded similar results and that the fluo-3 signals were less noisy.

Cells were subsequently rinsed with Medium 199 and rested for 30 min to allow complete deesterification of the fluo-3. For individual experiments, an aliquot of cells was transferred to a small plastic chamber (90-μl volume) mounted on the stage of the inverted microscope and allowed to adhere to a laminin-coated coverslip. The adhering myocytes were then superfused with PSS delivered to the chamber by gravity at a rate of 5 ml/min. Cells were electrically stimulated by a pair of platinum electrodes to reach a steady state level of intracellular Ca²⁺. The electrodes carried square-wave pulses of 5-ms duration, at a voltage 10% greater than threshold and at a rate of 0.25 Hz. All experiments were conducted at 23°C. At this temperature, the extrusion of fluo-3 from ferret ventricular myocytes is minimal compared with that which occurs at 37°C (unpublished observation, September 4, 1998).

Fluo-3 fluorescence was measured by a fluorometer (C&L Instruments, Hummelstown, PA) mounted on an inverted microscope (TE300; Nikon, Melville, NY). The fluo-3 was excited with 485 nm light from a xenon bulb, and Ca²⁺-dependent fluorescence emission was detected at 535 nm. Dye photobleaching was minimized by restricting the power level of the light source and by preventing exposure to light when data were not being acquired.

In all experiments an in-line calibrated anesthetic vaporizer was used to add the appropriate concentration of isoflurane to the solution that bathed the preparation. Concentrations of volatile anesthetics in the gas over the bathing solution were monitored by Raman spectroscopy (Ohmeda Rascal II, Madison, WI) and were also verified in the chamber solution by gas chromatography (5880A; Hewlett-Packard, Palo Alto, CA). 30The 1 minimum alveolar concentration (MAC) value for the three anesthetics at 23°C (isoflurane = 0.87%, halothane = 0.48%, sevoflurane = 1.53%) was calculated from published values measured at 37°C in the ferret (halothane and isoflurane) 31and rat (sevoflurane). 32The MAC values at different temperatures are essentially constant when expressed as aqueous phase concentrations. 32 

In high concentration (10 mm), caffeine is known to release all SR Ca²⁺content and prevent functional reuptake of Ca²⁺. 33When a myocyte bathed in PSS is exposed to caffeine, the intracellular [Ca²⁺] rapidly increases to a peak and then decreases exponentially because of the action of the NCX (approximately 86%) and PMCA (approximately 14%). 20Uptake of Ca²⁺into the mitochondria contributes relatively little to the rate of decline. 19 

In some experiments, the NCX was inhibited by bathing the cells in 0 Na⁺–0 Ca²⁺solution, which contained no added Na⁺(LiCl substituted for NaCl) and 1 mm EGTA. 34When caffeine is applied to the myocytes bathed in this solution, PMCA is the predominant mechanism responsible for the decline of the Ca²⁺transient. In another set of experiments, we inhibited the PMCA by exposing cells to 5-(and-6) carboxyeosin diacetate (succinimidyl ester; Molecular Probes). Carboxyeosin has been shown to block cardiac PMCA more effectively than ionic substitution (0 Na⁺, 10 mm Ca²⁺) in ferret and rabbit ventricular myocytes. 19,35,36A 10-mm stock solution in dimethyl sulfoxide was added to the control PSS to achieve a final concentration of 20 μm. Cells were loaded with carboxyeosin by exposing them to this solution for 15 min. They were then incubated in normal PSS without stimulation for 15 min to allow for deesterification of carboxyeosin. We performed only a single contracture in each of these cells (no time control) because the resting fluorescence did not return to baseline after caffeine application in the presence of carboxyeosin.

Signals were digitized and recorded at a sampling rate of 200 Hz. The least squares method was used to fit a single exponential curve (Y = Y⁰+ ae^−kt) to the declining phase of the data.

Individual measurements obtained in the presence of an anesthetic agent were normalized to the value obtained in the same cell in the absence of the anesthetic. For time control cells, the same protocol was followed except that they were not exposed to anesthetic. For statistical analysis, measurements in anesthetic-treated cells were compared with time controls, except in cells treated with carboxyeosin, where this was not possible. All measurements are reported as mean ± SD. Statistical significance (P < 0.05) was determined using the one-way analysis of variance with Dunnett test for multiple comparisons versus  control.

Figure 1shows representative records from four cells used in these experiments. The left half of the figure shows normal Ca²⁺transients recorded during electrical stimulation followed by a prolonged transient obtained during the application of caffeine. We used the peak of the caffeine-induced transient to monitor the Ca²⁺content of the SR, and the rate of decline as an index of the rate of Ca²⁺extrusion from the cell. 33,34Caffeine apparently opens the Ca²⁺release channel in the SR, resulting in the prolonged transients. In this situation, the intracellular [Ca²⁺] slowly declines because of the continued function of the NCX and PMCA mechanisms that move Ca²⁺out of the cell.

The right half of the figure repeats this sequence in the absence of anesthetic (time control, fig. 1A) and in the presence of halothane (fig. 1B), isoflurane (fig. 1C), and sevoflurane (fig. 1D), each at 1 MAC. Control experiments showed that repeated exposures to caffeine were well tolerated. However, we felt the use of time controls was necessary to account for the possibility of cell rundown and that the level of fluo-3 might decrease over the course of the experiment because of photobleaching or loss from the cell.

Figure 2shows the mean data for the relative amplitude of the electrically stimulated (fig. 2A) and caffeine-induced (fig. 2B) Ca²⁺transients. Figure 2Ashows that each anesthetic decreased the peak intracellular Ca²⁺during electrical stimulation. The peak of the electrically stimulated Ca²⁺transient (time control = 100 ± 19%) in 1 MAC anesthetic decreased for halothane (64 ± 12%), isoflurane (82 ± 14%), and sevoflurane (86 ± 17%).

In contrast, the results are different for the caffeine-induced Ca²⁺transient. Figure 2Bshows that total SR Ca²⁺content was unchanged in time controls (94 ± 14%) and decreased in the presence of halothane (60 ± 7%). Of note, it was unchanged in the presence of isoflurane (97 ± 14%) and actually increased in the presence of sevoflurane (110 ± 21%).

We also compared the effect of anesthetics on the fractional release of Ca²⁺from the SR. Fractional release was calculated by dividing the peak of the electrically stimulated transient (amount released) by the peak of the subsequent caffeine-induced transient (amount available). 7,28,37In the absence of anesthetic, the fractional release is high (very close to 100%;fig. 2C) during the conditions of the experiment. The slow rate of electrical stimulation (0.25 Hz) allows more SR Ca²⁺release channels to recover between beats (increases the electrically stimulated transient), and a certain amount of Ca²⁺leaks from the SR during the 30-s rest period before the application of caffeine (decreases the caffeine-induced transient).

During these circumstances, the fractional Ca²⁺release (time control = 102 ± 22%) remained unchanged in the presence of halothane (104 ± 24%). In contrast, both isoflurane (84 ± 27%) and sevoflurane (79 ± 21%) decreased the relative amount of available Ca²⁺that was released by electrical stimulation.

Only halothane prolonged the duration of the electrically stimulated intracellular Ca²⁺transient compared with time controls, indicating that it decreases the reuptake of Ca²⁺from the myoplasm by the SR. Figure 3Ashows the relative effects of each anesthetic on the half-time of the electrically stimulated Ca²⁺transient. The greatest effect in percentage terms was seen in the presence of halothane (133 ± 31%). Isoflurane (123 ± 29%) and sevoflurane (114 ± 36%) did not produce a statistically significant prolongation compared with the time controls (97 ± 23%). The half-time of relaxation of the Ca²⁺transient was 160 ± 55 ms during time control conditions and increased to 252 ± 107 ms in the presence of halothane (P < 0.05). Again, there was not a statistically significant prolongation of the electrically stimulated transient by isoflurane (170 ± 74 ms) or sevoflurane (155 ± 68 ms).

When a myocyte is exposed to caffeine in the presence of normal PSS, the function of the SR is inhibited by caffeine. 33However, the intracellular [Ca²⁺] declines because both NCX and PMCA mechanisms are active in extruding Ca²⁺from the cell. Table 1shows the effects of inhibition of NCX (0 [Na⁺] and 0 [Ca²⁺]) and PMCA (carboxyeosin) compared with control conditions (PSS) on the caffeine-induced Ca²⁺transients. We did not attempt to inhibit mitochondrial uptake of Ca²⁺because it does not appear to make a substantial contribution to the rate of decline of intracellular Ca²⁺during twitches or caffeine contractures. 19,33 

When the myocyte is bathed in normal PSS, the rate of decline decreased in the presence of isoflurane and sevoflurane, but not halothane. Figure 3Bshows that, during these circumstances, the duration of the caffeine-induced Ca²⁺transient (time control = 103 ± 10%) is unchanged in the presence of halothane (102 ± 23%) but increases in the presence of isoflurane (121 ± 14%) and sevoflurane (133 ± 16%). This suggests that isoflurane and sevoflurane inhibit Ca²⁺efflux from the cell. Figure 4shows representative traces illustrating the effect of no anesthetic (simply repeating the contracture;fig. 4A), halothane (fig. 4B), isoflurane (fig. 4C), and sevoflurane (fig. 4D) on the rate of decline of the caffeine-induced Ca²⁺transient. During these circumstances, only isoflurane and sevoflurane significantly inhibited Ca²⁺extrusion.

To further clarify this issue, we examined the effect of isoflurane and sevoflurane on the duration of the caffeine-induced transient when NCX was inhibited in a solution lacking Na⁺and Ca²⁺(0 Na⁺and 0 Ca²⁺). Figure 5Ashows the mean data from these experiments. Although halothane had no effect, both isoflurane and sevoflurane slowed the rate of decline. The relative half-time of decline was 120 ± 53% in the presence of isoflurane and 126 ± 22% in the presence of sevoflurane, compared with 89 ± 14% for time controls. These results suggest that both isoflurane and sevoflurane inhibit PMCA.

In another set of experiments, we examined the effects of isoflurane and sevoflurane when only NCX was active in extruding Ca²⁺from the cell. In this case, we inhibited the surface membrane Ca²⁺pump with carboxyeosin. 19,35Carboxyeosin has been shown to more effectively block the PMCA than ionic substitution in ferret ventricular myocytes. 19 Figure 5Bsummarizes the results of these experiments. Although halothane decreased the half-time of decline of the caffeine-induced transient (half-time = 1.57 ± 0.44 s⁻¹), it does not appear to actually increase the rate of extrusion because the resting level of Ca²⁺increased during these circumstances. On the other hand, isoflurane (half-time = 2.0 ± 0.30 s⁻¹) and sevoflurane (half-time = 2.1 ± 0.41 s⁻¹) had no effect on the duration of the caffeine-induced contracture when compared with controls (half-time = 2.1 ± 0.38 s⁻¹) and had no apparent effect on resting Ca²⁺. This result is consistent with the interpretation that isoflurane and sevoflurane prevent Ca²⁺extrusion by inhibiting the PMCA. Assuming that carboxyeosin causes 100% inhibition of the PMCA, we can use the rate constants of the decline of the caffeine-induced calcium transient to estimate the approximate inhibition caused by 1 MAC anesthetic. In this case, sevoflurane caused 88% inhibition and isoflurane 56%.

The interpretation of our experimental data must take into account the ability of individual volatile anesthetics to affect multiple sites within the myocardial cell in different ways. The interplay between the capacity of each system involved in Ca²⁺homeostasis and the degree to which it is affected by a particular volatile anesthetic determines the amount of Ca²⁺stored in and released from the SR. In this study we found that halothane, isoflurane, and sevoflurane decrease the electrically stimulated intracellular Ca²⁺transient, but that only isoflurane and sevoflurane tend to preserve the Ca²⁺content of the SR. The latter observation is somewhat surprising, because all three anesthetics have been reported to inhibit I^Ca9,38and therefore might be expected to decrease SR Ca²⁺content in the absence of other antagonistic effects. However, our data also indicate that isoflurane and sevoflurane inhibit Ca²⁺extrusion from the cell by the PMCA. This effect is important because it apparently allows the SR to accumulate relatively more Ca²⁺than it would otherwise.

The ability of volatile anesthetics to inhibit I^Cais important in explaining our results for two reasons. First, the relatively small amount of Ca²⁺that enters the cell in this way triggers the release of a much greater amount of Ca²⁺from the SR in a process known as Ca²⁺-induced Ca²⁺release (CICR), and fractional Ca²⁺release can be affected by interventions that alter this process. Second, the relatively small amount of Ca²⁺that makes up I^Cacan accumulate in the SR during some circumstances and contribute to the larger pool of Ca²⁺released from the SR in subsequent beats. For example, this process probably contributes substantially to the positive force–frequency relation seen in some species (rabbit, ferret, and human, but not rat or mouse). 20 

Therefore, the volatile anesthetics could potentially affect fractional Ca²⁺release or SR Ca²⁺content by affecting I^Ca. In the presence of isoflurane and sevoflurane, the most likely explantation for the observed decrease in the electrically stimulated Ca²⁺transient despite normal or increased SR Ca²⁺content is that CICR is impaired when Ca²⁺current decreases. 37This effect on fractional release is what would be expected solely because of inhibition of surface membrane Ca²⁺current, and has been documented for nifedipine in rat ventricular myocytes. 7 

Our observation that the SR Ca²⁺content can be preserved (isoflurane) or increased (sevoflurane) suggests that other mechanisms must be involved, because an agent that only decreased I^Cawould be expected to reduce SR Ca²⁺content. Among the three major mechanisms responsible for controlling the free intracellular [Ca²⁺] and balancing systolic Ca²⁺influx, SERCA sequesters Ca²⁺in the SR for later release during systole, whereas NCX and PMCA decrease the amount of Ca²⁺available for release by transporting Ca²⁺out of the cell. During a normal twitch, SERCA is responsible for transporting most (65–92%) of the cytosolic Ca²⁺, NCX removes less (7–30%), and the PMCA still less (1–3%). 33,34Clearly, volatile anesthetic interactions with any or all of these processes could affect the SR Ca²⁺content.

At equilibrium, by definition the amount of Ca²⁺entering the cell each beat equals the amount leaving. Interventions that inhibit I^Ca, for example, the volatile anesthetics, could possibly decrease the Ca²⁺transient by two mechanisms. First, less Ca²⁺could be available for release (I^cais the main determinant of SR Ca²⁺content), and second, the trigger mechanism could be less effective (decreased fractional release, inhibition of CICR). Indeed, it has been consistently observed that the amount of Ca²⁺released from the SR during electrical stimulation decreases in the presence of volatile anesthetics (halothane, enflurane, isoflurane, and sevoflurane) and the peak free intracellular Ca²⁺concentration declines. 3,7,11,39 

However, halothane clearly affects Ca²⁺homeostasis differently from isoflurane and sevoflurane, and the most likely explanation is that, of the anesthetics tested in this study, only halothane opens the Ca²⁺release channel. 40-42Halothane apparently causes Ca²⁺to leak into the cytoplasm and depletes SR Ca²⁺content. 22-24On the other hand, halothane does not appear to alter fractional Ca²⁺release even though it decreases I^Ca, indicating that it may enhance CICR. This facilitation of SR Ca²⁺release (CICR) after abrupt exposure to halothane (and enflurane) appears to be responsible for the ability of these anesthetics to cause a short-lived increase in intracellular Ca²⁺. 26It has been suggested that, in the steady state situation, halothane may have offsetting effects on Ca²⁺current and CICR. 7 

Our results are consistent with those of Davies et al. , 7who reported the effects of halothane, isoflurane, and sevoflurane on fractional Ca²⁺release and SR Ca²⁺content in isolated rat ventricular myocytes. They also found that halothane reduced SR Ca²⁺content but not fractional release, and that isoflurane and sevoflurane decrease fractional Ca²⁺release. In contrast, they found that isoflurane decreased and sevoflurane maintained SR Ca²⁺content. They proposed that the effects of sevoflurane were caused by the combination of inhibition of Ca²⁺current and inhibition of CICR or suppression of Ca²⁺extrusion from the cell. Our results indicate that suppression of Ca²⁺efflux is the more likely of the latter two mechanisms.

Our experiments indicate that both isoflurane and sevoflurane inhibit Ca²⁺extrusion from the cell by affecting the PMCA. We found that when both NCX and PMCA are functioning, the rate of decline of the Ca²⁺transient is slower in the presence of isoflurane or sevoflurane, but not halothane. When only NCX was inhibited, both isoflurane and sevoflurane still slowed the rate of decline, indicating an effect on PMCA. Consistent with this interpretation, when only PMCA was inhibited, the rate of decline was unchanged by these anesthetics.

We should point out that when the function of the SR is inhibited by rapid application of caffeine, PMCA apparently assumes a greater role in Ca²⁺transport, especially in ferret myocytes. 19,20It is also noteworthy that the relative contribution of PMCA to this process in ferret heart is similar to that in human heart (and much higher than in rat heart). 19 

Volatile anesthetics have been shown to decrease PMCA activity in neural cells, 43and it has been postulated that this effect may be at least partly responsible for alteration of the sensitivity to volatile anesthetics in disease states and with aging. Mammalian PMCAs are encoded by four separate genes, and additional isoforms are generated by alternative splicing of RNA. 44,45The expression of PMCAs is determined by the stage of development, tissue, and cell type. PMCAs 1 and 4 are expressed in most tissues, but PMCAs 2 and 3 are found mainly in the brain and striated muscle. 44,45 

Inhibition of PMCA activity by isoflurane and sevoflurane would be expected to maintain or increase SR Ca²⁺content if SR function (SERCA and Ca²⁺release channel) was not affected to any appreciable extent. Evidence from other studies supports the hypothesis that isoflurane and sevoflurane exert relatively little influence on the SR. For example, isoflurane has been reported to have no effect 24,40,41or to make the SR less leaky, 27and sevoflurane has also been reported to leave SR Ca²⁺content intact. 7 

Because both isoflurane and sevoflurane are known to decrease L-type Ca²⁺current, our results suggest that inhibition of PMCA activity counteracts this and any affect on SR function to leave SR content intact (isoflurane) or actually increased (sevoflurane) during the conditions of these experiments. These results are significant because they help to explain the decreased depression of myocardial contractility seen with isoflurane and sevoflurane when compared with halothane.