Comparative Contractile Effects of Halothane and Sevoflurane in Rat Aorta

The study was approved by the Kyoto University Animal Use Committee. Male Wistar rats (weight, 250–350 g) were anesthetized with pentobarbital, 50 mg/kg given intraperitoneally, and killed by exsanguination. The descending part of the thoracic aorta was isolated and cut into rings 5–6 mm wide. The endothelium was removed by rotating the rings around a rough-surfaced needle. The absence of functional endothelium was confirmed by the inability of 10 μM acetylcholine to inhibit phenylephrine contraction in each ring in tension experiments. In the subsequent experiments throughout, a supramaximal dose (10 μM) of verapamil was included in the bathing solution. This dose of verapamil completely blocked KCl-induced contraction and [Ca²⁺]ⁱresponse in rat aorta, 7,16,17proving that VOCCs were functionally inhibited. As the effect of the anesthetic agents was observed in the continuous presence of verapamil, it should have been free from the influence of the minor nonspecific effects of verapamil, 17such as suppression of Na⁺or K⁺channels.

The aortic rings were mounted vertically in organ baths filled with 10 ml Krebs–Ringer’s solution aerated with a 95% O²/5% CO²gas mixture. Isometric tension was recorded with a force displacement transducer (San-ei, Tokyo, Japan). The rings were placed in the bath with an optimal resting tension of 3 g for 60 min to equilibrate. They were then incubated for 10 min with 10 μM verapamil and contracted with 300 nM phenylephrine. This dose of phenylephrine was shown in our preliminary study to elicit ≈60% of the maximal contraction and was considered to allow enhancement and suppression of the contraction induced by the anesthetic agents. Phenylephrine induced a sustained contraction, which reached a stable level within 15 min (fig. 1A). Twenty minutes after addition of phenylephrine, halothane or sevoflurane (1–4%) was introduced into the gas aerating the baths. Halothane and sevoflurane were vaporized using Fluotec 3 (Ohmeda, West Yorkshire, United Kingdom) and MK III-ST (Acoma, Tokyo, Japan), respectively, and their concentrations in the gas were adjusted while monitoring with an Atom 303 anesthetic agent monitor (Atom, Tokyo, Japan). In a previous study from our laboratory in which the concentration of halothane and sevoflurane in the bath solution was assayed by gas chromatography, the anesthetic agents were proven to reach stable levels within 10 min after introduction into the gas. 1,18Anesthetic agents were maintained for 15 min and then discontinued, after which recovery from their effects were observed for 20 min. The maximum tension increment after introduction of the anesthetic agents was expressed as a percent of the maximum contraction induced by phenylephrine before addition of anesthetic agents in each ring.

In some experiments with halothane, the composition of Krebs–Ringer’s solution was modified to control the movement of Ca²⁺. Reference contraction was obtained in each ring with 300 nM phenylephrine in the presence of 10 μM verapamil. After phenylephrine had been washed out and the tension had returned to its basal level, modified experiments were conducted. Modifications to the Krebs–Ringer’s solution were made, by replacing 2 mM Ca²⁺with 1 mM EGTA or by adding either 30 μM SK&F 96365 (blocker of VICCs), 20 μM ryanodine (blocker of Ca²⁺release channel of the SR), or 0.1 μM thapsigargin (inhibitor of the SR Ca²⁺-adenosine triphosphatase). All solutions contained 10 μM verapamil. In either experiment, the increase of maximum tension after addition of halothane was expressed as a percent of the reference contraction induced by phenylephrine.

To estimate the amount of unidirectional influx of Ca²⁺into the smooth muscle, we used ⁴⁵Ca²⁺as a tracer ion. Experiments were performed using a method described previously 7,19with some modifications. Aortic rings were prepared and deprived of endothelium as described earlier. They were then cut open into strips and bathed in HEPES-buffered saline solution (HBSS) containing 10 μM verapamil. Verapamil was included in the solution throughout the experiment. The strips were allowed to equilibrate for 60 min in the solution with continuous aeration with air at 37°C. The strips were then transferred to HBSS containing 300 nM phenylephrine. After 10 min, 0–4% halothane was introduced to the solution through the aeration line. Halothane was vaporized and its concentration monitored as described earlier. The strips were exposed to halothane for up to 20 min. Every 5 min during the course, the rings were transferred to another vial filled with HBSS containing 1 μCi/ml ⁴⁵Ca²⁺in addition to phenylephrine and halothane. After exposure of the strips to ⁴⁵Ca²⁺for 2 min, they were bathed in 2 mM EGTA-containing HBSS at 0°C for 50 min to remove extracellular ⁴⁵Ca²⁺and then blotted and weighed. Intracellular ⁴⁵Ca²⁺was extracted by incubating the strips in 1.5 ml of 5 mM EDTA at room temperature overnight. ⁴⁵Ca²⁺in the solution was analyzed with a liquid scintillation counter after adding 3 ml of scintillation cocktail (Scintisol EX-H; Dojindo, Kumamoto, Japan). Data are expressed as the estimated total amount of influx of Ca²⁺(μmol) per kilogram aorta (wet weight). The number of minutes for each data point represents the time from the beginning of stimulation with phenylephrine until removal from the ⁴⁵Ca²⁺solution. Therefore, strips were exposed to ⁴⁵Ca²⁺for the last 2 min of the time indicated. Zero minutes means that strips were exposed to ⁴⁵Ca²⁺without stimulation by phenylephrine. The same experiment was conducted with sevoflurane (0–4%), but its effect was analyzed only at the time point of 20 min, i.e. , after 10 min of exposure to the anesthetic agent.

The Krebs–Ringer’s solution was composed of (in mM) NaCl 120, KCl 5.0, CaCl²2.0, MgCl²1.0, NaHCO³25.0, and dextrose 5.5; the p  H of the solution was 7.3–7.4 when the solution was aerated with a 95% O²/5% CO²gas mixture. The HEPES-buffered saline solution was composed of (in mM) NaCl 140, KCl 5.0, CaCl²2.0, MgCl²1.0, dextrose 5.5, HEPES 5.0 and p  H-adjusted to 7.30–7.35 with NaOH. The drugs used were halothane (Takeda Pharmaceutical Co., Osaka, Japan); sevoflurane (Maruishi Pharmaceutical Co., Osaka, Japan); phenylephrine, ryanodine, thapsigargin, and phentolamine (Sigma Chemical Co., St. Louis, MO); acetylcholine (Daiichi Pharmaceutical Co., Tokyo, Japan); verapamil, EGTA, and EDTA (Nacalai Tesque, Kyoto, Japan), and SK&F 96365 (BIOMOL; Plymouth Meeting, PA).

Data for anesthetic-induced contraction were expressed as median values with 25th and 75th percentiles because the maximum tension increment after anesthetic agents that we measured could not be lower than zero and did not follow a normal distribution. The other data were distributed normally and are expressed as mean ± SD. For comparison between two groups, t  test was used. For comparison among more than two groups, one-way analysis of variance was used. When data were not distributed normally or when variances were not equal among groups, the Kruskall-Wallis test was used. In either test, if significance (P < 0.05) was detected among groups, each group was compared with the control using Dunnett’s test for post hoc  multiple comparison. Differences at P < 0.05 were considered statistically significant.

For a control tension experiment, rat aortic rings pretreated with verapamil were exposed to halothane or sevoflurane (1%, 2%, or 4%; n = 4 each) at their basal tension. No contraction was observed in any of the rings examined (data not shown). We then tried to increase the concentration of anesthetic agent immediately by replacing the bath solution with already saturated solution, a method described by Kakuyama et al.  1to obtain a more definite contraction by halothane. Again, halothane or sevoflurane at 1%, 2%, or 4% failed to evoke any contraction (n = 4 each; data not shown).

Phenylephrine at 300 nM evoked a sustained contraction (fig. 1A), which averaged 0.844 g (SD = 0.209 g, n = 144) and did not differ significantly among groups. In this setting when halothane was introduced into the solution, the precontraction by phenylephrine was enhanced continuously (≤ 2%) or transiently (4%;fig. 1B). Sevoflurane 1–4%, failed to cause any contraction or relaxation (fig. 1C). When the maximum tension increment after use of anesthetic agents was plotted, halothane caused contraction in a linear, dose-dependent manner, whereas sevoflurane had no significant effect (fig. 1D).

The experiment on influx of ⁴⁵Ca²⁺revealed a stable increase in influx of Ca²⁺by phenylephrine, which developed within 10 min (fig. 2A). As this influx occurred while VOCCs were blocked, it should be through VICCs activated by phenylephrine. Halothane 1% induced a marked and sustained increase in influx of Ca²⁺, showing a time course consistent with halothane-induced contraction (fig. 2A). Halothane 4%, however, which induced the largest contraction, caused only a slight and transient increase in influx of Ca²⁺(fig. 2A). Ten minutes after introduction, the effects of varied concentrations (1–4%) of halothane were compared (fig. 2B). Influx of Ca²⁺was significantly larger in the 1% and 2% halothane groups than in the control (0%). In contrast, 10 min of exposure to sevoflurane 1–4% did not change influx of Ca²⁺(fig. 2B). In the subsequent experiments, the mechanism of halothane-induced contraction and influx of Ca²⁺was investigated.

To verify the role of influx of Ca²⁺in halothane-induced contraction, tension experiments were conducted in the absence of extracellular Ca²⁺. Five minutes after the bath solution was replaced by Ca²⁺-free Krebs–Ringer’s solution, 300 nM phenylephrine was added to the bath, which induced a transient contraction (fig. 3), possibly attributable to release of Ca²⁺from the SR. As the contraction was not sustained, we had to start halothane when the tension was near the basal level. Halothane 1% or 4% induced virtually no change in the tension in any ring (n = 4 each;fig. 3).

To confirm that halothane-induced contraction and influx of Ca²⁺were mediated by VICCs, the effect of SK&F 96365, a blocker of VICCs, 11was examined. In vascular smooth muscle, its blocking effect on VICCs has been shown in patch–clamp and contraction studies. 15,20,21Although it also may have a weak VOCC-blocking action, 21it should not matter in the presence of the maximal dose of verapamil. Contraction was induced by 300 nM phenylephrine 10 min after adding SK&F 96365 to the bath, which was only transient as in Ca²⁺-free solution (fig. 4A), suggesting that SK&F 96365 effectively blocked influx of Ca²⁺via  VICCs without interrupting release of Ca²⁺. Again, halothane 1% or 4% induced no change in tension (fig. 4A). The effect of SK&F 96365 was also examined in experiments on the influx of ⁴⁵Ca²⁺. SK&F 96365 30 μM was included in the HBSS from the beginning of the equilibrium period throughout the experiment. Influx of Ca²⁺was estimated at 0, 10, or 20 min after addition of 300 nM phenylephrine, either with or without exposure to 1% halothane for the last 10 min. Phenylephrine did not significantly increase influx of Ca²⁺over the time course (fig. 4B). Moreover, halothane did not increase influx of Ca²⁺at 20 min compared with the control (fig. 4B).

In another series of tension experiments, we used ryanodine and thapsigargin to block release of Ca²⁺by halothane. Ryanodine is an open-channel blocker of Ca²⁺-induced Ca²⁺release channel, which could abolish halothane-induced release of Ca²⁺in vascular smooth muscle cells, 1,2,4,22whereas thapsigargin inhibits Ca²⁺-adenosine triphosphatase of the SR, 23both resulting in depletion of stored Ca²⁺. When ryanodine or thapsigargin was added to the solution, a slowly developing contraction was induced, which has been explained by increased leakage of Ca²⁺from the SR or by influx of Ca²⁺attributable to SR depletion. 24,25After 40 min, when the contraction was stable, 300 nM phenylephrine and then halothane were applied to the ring (fig. 5A). Treatment with ryanodine, however, enhanced contraction by halothane up to 2%, but reduced it at 4% (figs. 5A and 5B). Thus, the maximal contraction was obtained at 2% (fig. 5B). Treatment with thapsigargin gave essentially the same result with a similar concentration–response curve (fig. 5B).

To examine the possibility that halothane specifically acts on α-receptors, we conducted an experiment on the influx of ⁴⁵Ca²⁺without phenylephrine. Phentolamine, an α-receptor antagonist, was included at 10 μM in the HBSS from the beginning of the equilibrium period. Influx of Ca²⁺was estimated either without any treatment or after 10 min of exposure to 1% halothane. In either group, the rings were incubated in solutions containing ⁴⁵Ca²⁺for the last 2 min. Estimated influx of Ca²⁺was significantly higher in halothane-exposed strips (44.6 ± 17.5 μmol/kg, n = 16, P < 0.05) than in control strips (33.6 ± 7.0 μmol/kg, n = 16). The 1% halothane-evoked fraction of influx of Ca²⁺was 11.0 μmol/kg, much less than that in the presence of phenylephrine (27.1 μmol/kg). This appeared to be consistent with the result of the tension experiment, in which halothane alone could not induce contraction without precontraction.

In rat aortic rings pretreated with verapamil, halothane at 0.5–4.0% enhanced phenylephrine-evoked contraction in a linear, dose-dependent manner, whereas 1–4% sevoflurane had no effect. The result with halothane is compatible with a previous study in which halothane enhanced norepinephrine-induced contraction in verapamil-treated rat aorta. 5Although the study showed a simultaneous increase in [Ca²⁺]ⁱby using a Ca²⁺indicator dye, this technique does not define the mechanism of increased [Ca²⁺]ⁱ, which might be induced by anesthetic agents either through influx of Ca²⁺, 7release of Ca²⁺, 4,6,26inhibited extrusion of cytosolic Ca²⁺, 27,28or decreased uptake of cytosolic Ca²⁺into the SR. 4We performed an experiment on the influx of ⁴⁵Ca²⁺because it was a useful method for evaluating influx of Ca²⁺alone, unaffected by movement of Ca²⁺inside the cell. 19The results of our experiment on the influx of ⁴⁵Ca²⁺demonstrated that halothane 1% and 2% increased the influx of Ca²⁺, which had been submaximally preactivated by phenylephrine. The pathway of influx of Ca²⁺should be VICCs because VOCCs had been functionally inhibited by verapamil. Halothane 1% induced an even larger influx of Ca²⁺(27.1 μmol/kg) than phenylephrine (15.5 μmol/kg) despite that it caused only ≈10% of phenylephrine-induced contraction. This puzzling disproportion may be attributable to their opposite effects on Ca²⁺sensitivity of the contractile protein, i.e. , sensitization with α-receptor stimulation 16,17and desensitization with halothane. 3,29In contrast, influx of Ca²⁺induced by 4% halothane was unexpectedly low, yielding a curious parabolic-shaped dose–response curve different from that of contraction. In addition, 1–4% sevoflurane had no effect on influx of Ca²⁺, which was consistent with its tension result.

We postulated that the transient contraction induced by 4% halothane was attributable to release of Ca²⁺and tried to confirm this hypothesis. In previous studies, contraction by halothane has been ascribed to release of Ca²⁺from the SR, because the contraction was resistant to removal of extracellular Ca²⁺4,29or was abolished by ryanodine. 1–4,22In this study, 1% or 4% halothane did not evoke contraction in Ca²⁺-free solution; however, this did not necessarily prove that halothane did not release Ca²⁺from the SR, because soaking the ring in EGTA-containing solution should have reduced the basal [Ca²⁺]ⁱand depleted intracellular Ca²⁺stores, 30which is an untoward condition to observe release of Ca²⁺by halothane. We therefore performed an experiment in normal solution using drugs that deplete the store of Ca²⁺. Ryanodine and thapsigargin reduced the contraction induced by 4% halothane but not that by 1% and 2% halothane. Although it has been suggested that smooth muscle cells may have a store of Ca²⁺that is resistant to ryanodine but sensitive to inositol 1,4,5-trisphosphate, 31thapsigargin is considered to deplete both types of store. 23Therefore our results suggest that release of Ca²⁺from the SR plays an important role in contraction by halothane at 4% but not at concentrations of ≤ 2%. After depletion of the store of Ca²⁺, the dose–response curve of contraction became more similar to that of halothane-induced influx of ⁴⁵Ca²⁺and could be considered to reflect the activity of influx of Ca²⁺more directly than that without depletion of the store. A minor discrepancy between the experiments on tension and influx of ⁴⁵Ca²⁺may be pointed out; 4% halothane still could cause contraction after depletion of the store, whereas it could not cause a large influx of Ca²⁺. This finding can be explained, however, by the different way of measurement; contraction was measured at its maximal point and influx of Ca²⁺at the fixed time, 10 min after use of halothane.

To confirm that halothane-induced influx of Ca²⁺was mediated by VICCs, we examined the effect of SK&F 96365, a blocker of VICCs, in the experiments on tension and influx of ⁴⁵Ca²⁺. SK&F 96365 blocked the sustained phase of contraction induced by phenylephrine to evoke a phasic contraction similar to that in a Ca²⁺-free solution. Contraction induced by 1% or 4% halothane was also abolished. It was considered, however, that this condition may not be favorable for contraction by halothane, because it should deplete intracellular Ca²⁺by blocking Ca²⁺replenishment as in Ca²⁺-free solution. We then tested SK&F 96365 in experiments on the influx of ⁴⁵Ca²⁺. This completely abolished influx of Ca²⁺by phenylephrine and 1% halothane. This is an additional finding that supports that halothane activates VICCs. SK&F 96365 is a nonspecific blocker of VICCs. 11As subtype-specific blocker is not yet established, we could not determine the subtype of VICCs sensitive to halothane.

The mechanism of VICC activation by halothane was not fully examined in this study. Halothane should not act on the α-receptor as an agonist because 1% halothane was able to activate influx of Ca²⁺even in the presence of phentolamine. In addition, it is unlikely that halothane augments the binding of phenylephrine to the α-receptor, because it has been shown that halothane did not increase, but rather inhibited, the binding of radioligands to α-receptors obtained from rat brain 32or rabbit myometrium. 33It seems, however, that activation of VICCs by halothane is not independent of the signal transduction system from the α-receptor to VICCs, because the amount of influx of Ca²⁺activated by halothane was higher in the presence of phenylephrine. Halothane might enhance activation of the signaling pathway distal to the level of the receptor, i.e. , G protein, intracellular second messengers, or the channel protein itself. Another possibility is that halothane evokes release of Ca²⁺, thus depleting stored Ca²⁺and stimulating influx of Ca²⁺via  a depletion-operated mechanism. This has been suggested to account for the activated pathway of influx of Ca²⁺induced by halothane in canine airway smooth muscle. 34This may not be a primary mechanism of VICC activation in rat aorta, however, considering that abolishment of release of Ca²⁺by ryanodine or thapsigargin did not inhibit contraction by 1–2% halothane. It is necessary to elucidate the signal transduction pathway from the α-receptor to the channel before further conclusions can be made.

In previous studies using mesenteric arteries, low concentrations (< 0.75%) of halothane were shown to induce contraction by stimulating release of Ca²⁺even in a resting state. 1,2,4In contrast, in the current study using rat aorta, release of Ca²⁺was detected only when rings were precontracted in the presence of verapamil and the concentration of halothane was increased to 4%. The different responses might be ascribable to the different Ca²⁺regulatory machinery in these two tissues. The vascular smooth muscle from different size of vessels and tissue bed can display different forms of pharmacomechanical coupling mechanisms, such as relative roles of VICCs, VOCCs, and the SR as the source of Ca²⁺, sensitivity of the SR to halothane, the content of Ca²⁺in the SR, and the Ca²⁺-buffering capacity of the cytosol. Although influx of Ca²⁺has not been examined directly in studies using mesenteric arteries, some of the results have raised the possibility that halothane stimulates influx of Ca²⁺through the cell membrane. 1,4Therefore, it is possible that influx of Ca²⁺induced by halothane is not a specific phenomenon in aortic tissues but a common underlying mechanism in other vascular tissues. In contrast, some studies that have monitored [Ca²⁺]ⁱin A7r5, a cultured smooth muscle cell line established from rat aorta, have shown that halothane suppressed rather than stimulated vasopressin-activated influx of Ca²⁺. 35,36The discrepancy might be attributable to the absence of VOCC blockers in their studies in addition to the differences in the cell type, vasoconstrictor, or technique used. The current study appears to be the first to demonstrate directly the activation of the VICCs by halothane in vascular smooth muscle.

The effect of halothane on influx of Ca²⁺was similar to that of isoflurane shown in our previous study. 7Isoflurane induced influx of Ca²⁺resistant to nifedipine and verapamil, and its effect was not linearly dose-dependent, showing a maximal effect at 2.3%. Most anesthetic agents may exert their effects on VICCs through common mechanisms. The distinct feature of contraction by halothane is that release of Ca²⁺also contributed to its effect, unlike with isoflurane. In contrast, sevoflurane seems completely devoid of these effects within a clinically relevant concentration range (1–4%). Although sevoflurane has been shown to suppress KCl-induced contraction in coronary arteries, 18vasodilating action was abolished by verapamil in this study. Therefore, it can be assumed that the direct vascular effect of sevoflurane is mediated mainly by suppression of VOCCs and that it is inactive with the other contractile mechanisms found for halothane or isoflurane.

Halothane is suggested to have dual contractile effects on verapamil-treated precontracted rat aorta, with sustained contraction at up to 2% attributable mainly to influx of Ca²⁺via  VICCs and transient contraction at 4% attributable mainly to release of Ca²⁺from the SR. In contrast, sevoflurane had no contractile effect. Although these anesthetic agents have VOCC-blocking action in common, they are different in exerting contractile effects. The contractile effect of halothane is weak in rat aorta, hardly visible in normal conditions. The overall consequence of this effect on the other vascular tissues and the systemic circulation may be more profound, however.