Effect of Volatile Anesthetics on the Force–Frequency Relation in Human Ventricular Myocardium: The Role of the Sarcoplasmic Reticulum Calcium-release Channel

Left ventricular myocardium was obtained from 18 patients suffering from end-stage heart failure. The patients underwent cardiac transplantation as a result of dilated (n = 10) or ischemic (n = 8) cardiomyopathy. Nonfailing left ventricular myocardium was obtained from eight multiorgan donors whose hearts could not be transplanted for technical reasons. All organ donors showed a normal left ventricular pump function as shown by echocardiography (ejection fraction > 55%), and none of the donors had any pathologic cardiovascular history. Because brain death may lead to a pronounced depression of cardiac performance, all patients with brain death–induced cardiac dysfunction were excluded. 15The study was approved by the institutional review board on human research of the University Hospital, Aachen. All patients gave written informed consent. In case of the organ donors, consent was obtained from the relatives.

Immediately after surgical resection, the tissue was placed into bathing solution (2.5 mm CaCl², 119.8 mm NaCl, 1.04 mm MgCl², 5.36 mm KCl, 22.6 mm NaHCO³, 0.42 mm NaH²PO⁴, 5.05 mm glucose, 0.28 mm ascorbic acid, 0.05 mm Na²EDTA, pH 7.4, gassed with 5% CO²–95% O², room temperature) and transferred to the laboratory. In parallel to muscle fiber direction, thin myocardial strips were prepared under stereomicroscopic control. The diameters of the muscle preparations were less than 0.6 mm to avoid core hypoxia and were comparable to those used by other investigators in similar experiments. 16The muscle preparations were transferred into an organ bath filled with prewarmed (37°C) bathing solution and electrically stimulated at 1 Hz with rectangular impulses (5-ms duration, 5–10% above threshold voltage). After an equilibration period of 30 min, the muscles were stretched by increasing the resting tension from 1 mN stepwise by 0.5 mN until the muscle length providing maximal active force generation was reached. Table 1shows the main mechanical characteristics of the muscle preparations. In a first group of muscle preparations, the concentration-dependent negative inotropic effect of the volatile anesthetics was studied (10–12 muscle preparations from 5 donor hearts and 22–23 muscle preparations from 13 failing hearts). The anesthetics were added to the carbogen using vaporizers (Dräger [Lübeck, Germany] for halothane and sevoflurane or Ohmeda [Herts, United Kingdom] for isoflurane) and bubbled through the bathing solution. Gas chromatography measurements revealed that equilibration of the anesthetics with the bathing solution was complete within 5 min. The concentration of the anesthetics was increased in steps of 0.5 minimum alveolar concentration (MAC) up to 3 MAC for isoflurane and halothane. In the case of sevoflurane, 2.5 MAC was the maximal concentration that could be delivered by the vaporizer. If force of contraction after wash-out of the anesthetic was less than 5% lower than the predrug value, the next anesthetic was studied. Halothane, isoflurane, and sevoflurane were studied in a random order. In a second group of muscle preparations, the effect of the anesthetics on the FFR was studied (10–12 muscle preparations from 5 donor hearts and 19–24 muscle preparations from 13 failing hearts). The stimulation frequency was increased from 0.5 to 3 Hz at control conditions, at 1.5 MAC, and at 2.5 MAC of each anesthetic. In 12 muscle preparations from 5 donor hearts and 20 muscle preparations from 13 failing hearts, the effect of ryanodine (10 nm) and verapamil (100 nm) on the FFR was studied for comparison. The effect of low Ca²⁺concentrations in the bathing solution (1.5 and 0.5 mm) on the FFR was studied in 10 muscle preparations from 3 failing hearts.

To exclude that hypoxia occurs in the preparations, at the end of the experiment the muscle strips were kept in a condition with a high oxygen demand (i.e. , a high stimulation frequency). After force of contraction had reached steady state, the gas bubbling (95% O², 5% CO²) was changed to 80% O²plus 15% N²and 5% CO²for a period of 30 min. All muscle preparations showing a decline in force of contraction of more than 5% during this period were excluded from the study (< 5% of all muscle preparations). 17 

Immediately after explantation, the myocardium was cut into small pieces, immersed in liquid nitrogen, and stored at −80°C until use. ³H-ryanodine binding experiments were performed as previously described. 18Briefly, 70 mg myocardium of each heart were homogenized three times for 15 s with an Ultraturrax (IKA, Staufen, Germany) and afterward with 25 bursts with a glas potter at 1,100 rpm in 3 ml ice-cold buffer A: 1 m KCl, 20 mm HEPES, and 1 mm EGTA at pH 7.4. For saturation binding experiments, the free Ca²⁺concentration was adjusted to 10⁻⁵m according to the calculations of Fabiato. 19Radioligand binding was started by addition of 150 μl of the heart homogenates (100 μg protein) to 100 μl buffer A containing 0.3–40 nm ³H-ryanodine. The reaction was stopped by the addition of 5 ml ice-cold buffer A, immediately followed by rapid filtration through Whatman GF/C filters. The filters were washed twice with 5 ml buffer A and counted after the addition of 5 ml Hionic Fluor (Canberra-Packard, Dreieich, Germany). Nonspecific radioactivity was determined by the addition of 10 μm unlabeled ryanodine into the binding assays and was approximately 15% at 1 nm ³H-ryanodine. To investigate Ca²⁺-dependent ³H-ryanodine binding, 100 μg protein was incubated with 12 nm ³H-ryanodine in the presence of 15 different Ca²⁺concentrations ranging from 10⁻⁹to 10⁻¹m. Ca²⁺ions were buffered with 1 mm EGTA and adjusted to the respective final free concentrations according to Fabiato. 19All other steps were performed as described previously. To study the effect of volatile anesthetics on Ca²⁺-dependent binding, the membrane preparation was gassed for 15 min with anesthetic in nitrogen or with pure nitrogen for control to avoid oxidative alterations of membrane lipids. The radioligand binding experiments were performed in anesthetic–nitrogen atmosphere or pure nitrogen, respectively. As determined by gas chromatography, equilibration of anesthetic concentration in the medium was achieved within 15 min. In these experiments, the assays were performed in the presence of 10 different Ca²⁺concentrations ranging from 10⁻⁸to 10⁻⁵m.

Protein content of the homogenates was measured by the method of Bradford 20using γ-globulin as standard.

³H-ryanodine (68.4–74.8 Ci/mmol) was obtained from NEN DuPont de Nemours (Bad Homburg, Germany), and ryanodine was obtained from Calbiochem-Novabiochem (Bad Soden, Germany). All chemicals were of analytical or best commercial grade available. Deionized water was used throughout.

Saturation curves, Ca²⁺-dependent ³H-ryanodine binding curves, and Scatchard plots 21were fitted with computer-assisted least square regression analysis using software designed by Graphpad (San Diego, CA). Data are expressed as mean ± SD. For EC⁵⁰and EC⁴⁰values, 95% confidence intervals are given. Statistical significance was determined with the unpaired Student t  test (for comparison of two groups) or by one-way analysis of variance with Newman-Keuls test for comparison of multiple groups. A P  value < 0.05 was considered statistically significant.

A total of 34 muscle preparations from 5 donor hearts (in 3 donors only frozen tissue was obtained) and 68 preparations from 16 failing hearts were studied. Length, diameter, and resting force at the maximal active force generation of the preparations did not differ between nonfailing and failing myocardium, whereas the active force at a stimulation frequency of 1 Hz and the ratio between active and resting force were significantly higher in nonfailing myocardium (table 1).

Figure 1(top) shows the concentration-dependent negative inotropic effect of all three anesthetics in left ventricular myocardium of five organ donors. The negative inotropic effect of halothane was more pronounced than that of isoflurane or sevoflurane. At a physiologic rate (1 Hz), 0.72 MAC halothane reduced contractile force by 40% (EC⁴⁰), whereas approximately 1.4 MAC isoflurane or sevoflurane was needed to elicit the same effect (P < 0.05;table 2). This difference was even more pronounced at a stimulation frequency of 3 Hz. Interestingly, only the first part of the concentration–response curve of halothane, at concentrations up to 1.5 MAC, was steeper compared with the other compounds. At greater than 1.5 MAC, the contractile force decreased parallel to the decrease with isoflurane and sevoflurane. In failing human myocardium, at a stimulation frequency of 1 Hz, the negative inotropic effect of halothane was also most pronounced (fig 1, bottom). Approximately 0.5 MAC halothane was sufficient to reduce contractile force by 40% (table 2). Similar as in nonfailing myocardium, the negative inotropic potency of isoflurane and sevoflurane was lower (P < 0.05;table 1). Again, only the initial part of the concentration–response curve of halothane was steeper. At concentrations greater than 1.5 MAC, force of contraction decreased parallel in all three anesthetics. At 3 Hz, the negative inotropic effect of all three anesthetics was similar. The concentration–response curves were nearly superimposable. Forty percent reduction of contractile force was reached at equianesthetic concentrations of halothane, isoflurane, and sevoflurane (nonsignificant;table 2). Interestingly, the negative inotropic effect of halothane was much less pronounced at 3 Hz compared with 1 Hz.

Figure 2shows the influence of the anesthetics on the FFR. In nonfailing myocardium and in the absence of anesthetics (baseline), the FFR was positive. All three anesthetics reduced contractile force equally at all stimulation frequencies, with halothane being the strongest compound, but none of them changed the positive shape of the FFR. In contrast, in failing myocardium, the FFR was differentially affected by the anesthetics. In the absence of anesthetics, the FFR was negative in failing myocardium. As previously mentioned, the negative inotropic effect of halothane was relatively strong at 1 Hz, whereas it was much less pronounced at 3 Hz. As a result, the shape of the FFR became positive in the presence of halothane, although the baseline FFR was negative. The restoration of the positive shape of the FFR could be demonstrated at 1.5 and 2.5 MAC halothane. In contrast, isoflurane and sevoflurane did not alter the negative shape of the FFR.

In nonfailing myocardium, neither verapamil nor ryanodine changed the positive shape of the FFR, although both drugs showed a strong negative inotropic effect (fig. 3). In failing myocardium, verapamil also did not alter the (negative) shape of the FFR, but blockade of the SR by ryanodine reduced contractile force preferentially at low stimulation frequencies. Similar to halothane, ryanodine restored the positive shape of the FFR. In failing myocardium, lowering the availability of Ca²⁺for the contraction by low Ca²⁺concentrations in the bathing solutions (1.5 and 0.5 mm) did not change the shape of the FFR despite the pronounced negative inotropic effect (table 3).

Binding of the radioligand ³H-ryanodine to left ventricular homogenates showed saturation characteristics (fig. 4A), and the Scatchard plot (fig. 4B) was linear, indicating that ³H-ryanodine bound to one distinct receptor binding site. Similar to our previous observation, 18the ³H-ryanodine binding site density was the same in nonfailing (86 ± 11 fmol/mg protein; n = 8) and failing myocardium (91 ± 30 fmol/mg protein; n = 11; nonsignificant).

A representative example of Ca²⁺-dependent ³H-ryanodine binding to the human Ca²⁺-release channel is shown in figure 4C. ³H-ryanodine binding showed a bell-shaped pattern with an increase in specific binding at submicromolar Ca²⁺concentrations and a decrease at Ca²⁺concentrations higher than 0.5 mm.

Because it has been shown that increasing specific ³H-ryanodine binding at submicromolar Ca²⁺concentrations reflects increasing open probability of the channel and the role of decreasing ³H-ryanodine binding at Ca²⁺concentrations in the millimolar range is still unknown, in the following experiments we concentrated on Ca²⁺concentrations ranging from 10⁻⁸to 10⁻⁵m. Ca²⁺-dependent ³H-ryanodine binding to the SR Ca²⁺-release channel from nonfailing and failing human myocardium showed a similar pattern, and EC⁵⁰values of the increase in specific ³H-ryanodine binding were the same in nonfailing and failing myocardium (table 4). Figure 5shows the influence of the halogenated anesthetics (1.5 MAC) on the Ca²⁺-dependent ³H-ryanodine binding. Halothane shifted the curve toward lower Ca²⁺concentrations. The extent of this effect was similar in failing and nonfailing myocardium. In contrast, isoflurane and sevoflurane (1.5 and 2.5 MAC) did not alter Ca²⁺-dependent ³H-ryanodine binding. In addition, the maximal binding of ryanodine was not changed by either compound. Table 4shows the statistical analysis of the binding data. Halothane at 2.5 MAC did not provoke a further shift of the binding curve to the left. Both in nonfailing and failing myocardium, the maximal effect of halothane on the SR Ca²⁺-release channel was already reached at 1.5 MAC.

The current study provides a direct comparison of the negative inotropic effects of halothane, isoflurane, and sevoflurane in nonfailing and failing ventricular myocardium at different stimulation frequencies. We demonstrated that in both failing and nonfailing myocardium, halothane exerts the most pronounced negative inotropic effect. In failing myocardium, halothane and ryanodine restored the positive shape of the FFR, whereas isoflurane and sevoflurane did not change FFR. Only halothane activated the myocardial Ca²⁺-release channel of the SR, which offers an explanation for the differential effect of volatile anesthetics on the FFR and for the more pronounced negative inotropic effect of halothane.

As recently reported, 18Ca²⁺-dependent specific ³H-ryanodine binding can be used as a probe for the activation of the human cardiac SR Ca²⁺-release channel. The plant alkaloid ryanodine only binds to the open state of the channel. 22Accordingly, a good correlation between channel activation by Ca²⁺(i.e. , opening of the channel) and increased specific ³H-ryanodine binding at submicromolar Ca²⁺concentrations has been reported during various conditions. 22–24Although ³H-ryanodine binding studies are an indirect method to assess channel activation, there is increasing evidence that Ca²⁺-dependent ³H-ryanodine binding reflects Ca²⁺-induced opening of the channel. 22 

As expected, specific ³H-ryanodine binding was absent at nanomolar Ca²⁺concentrations, indicating that the channel is closed during these conditions. At submicromolar Ca²⁺concentrations, ³H-ryanodine binding increased reflecting enhanced open probability of the channel. The Ca²⁺concentration eliciting the half-maximal ³H-ryanodine binding can be used as a measure of the Ca²⁺sensitivity of the channel. 21The increase of Ca²⁺-dependent ³H-ryanodine binding could be observed at lower Ca²⁺concentrations when 1.5 MAC halothane was present, indicating that halothane increased the Ca²⁺sensitivity. Halothane at 3.0 MAC did not provoke a further leftward shift of the Ca²⁺-dependant ³H-ryanodine binding curve, indicating that relatively low halothane concentrations already elicit the maximal effect. In contrast, isoflurane and sevoflurane did not change the Ca²⁺sensitivity of the SR Ca²⁺-release channel.

The increased Ca²⁺sensitivity of the channel in the presence of halothane will increase the average open probability of the channel 25as a response to physiologic variations in intracellular Ca²⁺concentrations. It has been suggested that this effect leads to a decrease in the ability of the SR to store and accumulate Ca²⁺and thus to a Ca²⁺depletion of the SR. 3 

Numerous studies in various mammalian species demonstrated a pronounced, concentration-dependent, and reversible cardiodepressant effect of all halogenated anesthetics. 5,26,27Because important physiologic differences exist between animal myocardium and human cardiac tissue, 28especially with regard to the FFR, 29there is a need for studies on the inotropic effect of the anesthetics in human myocardium. Only very recently was an extensive work on the differential negative inotropic effect of the most frequently used volatile anesthetics on isolated atrial human myocardium published. 13It showed that the negative inotropic effect of halothane was more pronounced than that of isoflurane, sevoflurane, and desflurane. Some properties of electrical and mechanical function in ventricular myocardium, however, might differ from the behavior of atrial myocardium. Action potential duration 30and contraction times 31have been shown to be shorter in atrial myocardium, and isometric force development is lower than in ventricular myocardium. 31Moreover, altered Ca²⁺homeostasis in failing human myocardium 32might have an influence on the negative inotropic mechanism of volatile anesthetics. Our data demonstrate that, in failing and nonfailing human ventricular myocardium, the negative inotropic effect of halothane is more pronounced than that of isoflurane and sevoflurane, which depress myocardial contractility to a similar extent. Interestingly, the slope of the concentration–response curve of halothane was steepest at low concentrations up to 1.5 MAC, resulting in a higher negative inotropic potency (lower EC⁴⁰value). At concentrations greater than this, contractile force declined parallel with the negative inotropic effect of isoflurane and sevoflurane. Because 1.5 MAC halothane exerts the maximal effect on the Ca²⁺sensitivity of the SR Ca²⁺-release channel, these data suggest that the initial steep decline of contractile force is mainly caused by activation of the SR Ca²⁺-release channel and subsequent Ca²⁺depletion of the SR. At concentrations greater than 1.5 MAC, no further activation of the SR Ca²⁺-release channel occurs, and all anesthetics reduce force of contraction with a similar slope of the concentration–response curve, probably by inhibition of the Ca²⁺inward current via  the L-type Ca²⁺channel. 33Isoflurane and sevoflurane do not interact with the SR Ca²⁺-release channel and therefore do not show the steep decrease in contractile force at low concentrations of the anesthetics.

Schmidt et al.  12reported that halothane restores the positive FFR in human ventricular myocardium and proposed that prevention of diastolic Ca²⁺overload by the antagonistic effects of halothane on L-type Ca²⁺channels or activation of the SR Ca²⁺-release channel might have caused the restoration of the FFR. Moreover, a decrease of the Ca²⁺sensitivity of myofilaments might contribute to the phenomenon, but the exact molecular mechanism remained unclear. In concordance with the findings of Schmidt et al. , 12we also found that the positive shape of the FFR is restored by halothane in failing human myocardium. The comparison with other volatile anesthetics and with the effect of verapamil and ryanodine allow further conclusions to be drawn on the intracellular mechanism. With respect to inhibition of the Ca²⁺inward current via  the L-type Ca²⁺channel, no major differences have been reported between halothane, isoflurane, and sevoflurane. 33Because only halothane restored the positive shape of the FFR in failing human myocardium, the interaction of the anesthetics with L-type Ca²⁺channels cannot account for this effect of halothane. Accordingly, the Ca²⁺antagonist verapamil did not influence the shape of the FFR in failing myocardium. Similarly, the low availability of Ca²⁺itself cannot account for the positive shape of the FFR in the presence of halothane. At low Ca²⁺concentrations in the bathing solution (1.5 or 0.5 mm), the FFR was still negative in failing myocardium.

A decrease in Ca²⁺sensitivity of the myofilaments also cannot explain why halothane, but not isoflurane, alters the FFR in failing myocardium. Halothane and isoflurane depressed the Ca²⁺sensitivity of human ventricular myofilaments to a similar extent. 4 

A more promising explanation for the restoration of the positive FFR by halothane is its interaction with the SR Ca²⁺-release channel of the SR. As demonstrated by our ³H-ryanodine radioligand studies, only halothane, but not isoflurane or sevoflurane, increased the Ca²⁺sensitivity of the Ca²⁺-release channel, leading to an increased open probability of the channel and Ca²⁺depletion of the SR. The Ca²⁺depletion of the SR lowers contractile force, particularly at low stimulation frequencies at which time for Ca²⁺reuptake into the SR is not limited, and thus the contribution of intracellular Ca²⁺cycling via  the SR to contractile force is high. Because the activity of the SR Ca²⁺–adenosine triphosphatase is reduced in failing human myocardium, 34the contribution of SR Ca²⁺cycling to contractile force becomes less important at higher stimulation frequencies with decreased time for reuptake of Ca²⁺into the SR. 35This could explain why, in failing human myocardium, halothane decreased contractile force preferentially at low stimulation frequencies, resulting in the restoration of the FFR. The hypothesis is strongly supported by our observation that ryanodine also restored the positive shape of the FFR in failing myocardium. Ryanodine keeps the SR Ca²⁺-release channel in a 40% subconductance state and thereby blocks the repetitive reuptake and release of Ca²⁺by the SR, resulting in Ca²⁺depletion of the intracellular Ca²⁺stores. 22The similarity of halothane and ryanodine with respect to their interaction with the SR and their effect on the FFR in failing myocardium strongly suggests that halothane restores the FFR by increasing the open probability of the SR Ca²⁺-release channel, thereby depleting the sarcoplasmic reticular Ca²⁺stores, leading to a cardiodepressant effect particularly at low stimulation frequencies. Because reduction of the heart rate has been suggested to improve myocardial contractility in patients with heart failure, 10knowledge about the influence of halogenated anesthetics on the FFR might help to improve the clinical outcome of anesthetized patients with depressed myocardial function.

Although the negative inotropic effect of halothane was also more pronounced than that of isoflurane and sevoflurane in nonfailing human myocardium, the positive shape of the FFR was not altered by any anesthetic. Because the activity of the SR Ca²⁺–adenosine triphosphatase was not depressed, intracellular Ca²⁺cycling via  the SR strongly contributed to the active force development at all stimulation frequencies. Thus, inhibitors of the sarcoplasmic reticular function are expected to reduce contractile force at all rates. Indeed, our data demonstrate that both ryanodine and halothane depress contractile force to a similar extent at all stimulation frequencies. Accordingly, in nonfailing myocardium, the positive shape of the FFR was maintained in the presence of all drugs.

Extrapolation of our data to the situation in vivo  should be performed with caution. Although changes in the shape of the FFR in vitro  go hand-in-hand with respective alterations of the FFR in  vivo, 7the effect of volatile anesthetics on frequency-dependent force development might be influenced by a number of additional factors beyond changes of intrinsic myocardial contractility. Changes of cardiac performance during the administration of anesthetics may also depend on modifications of the physiologic variations in sympathetic and parasympathetic tone, venous return, and vascular tone.

Moreover, the number of specimens from nonfailing hearts was small because of the limited access to appropriate tissue samples. However, in the absence of volatile anesthetics, baseline force of contraction and the shape of the FFR was very similar to results previously reported in similar tissue samples, 8and the variability within our population of muscle preparations was not extraordinarily high. Thus, we suggest that the data obtained from the limited number of specimens investigated here represent the typical contractile behavior of nonfailing myocardium studied during similar conditions.