Role of Potassium Channels in Isoflurane- and Sevoflurane-induced Attenuation of Hypoxic Pulmonary Vasoconstriction in Isolated Perfused Rabbit Lungs

The experimental protocol was approved by the Tottori University Faculty of Medicine Laboratory Animal Care Committee. Fifty-six female Japanese white rabbits (weighing 1.8–2.7 kg) were anesthetized with pentobarbital (20 mg/kg intravenously) and ketamine (30 mg/kg intramuscularly). Heparin was injected intravenously (500 U/kg) before surgery. After local anesthesia with 1.0% lidocaine and tracheostomy, the lungs were mechanically ventilated with a ventilator (model 681; Harvard Apparatus, Natick, MA) using room air at a rate of 40 strokes per minute, a tidal volume of 10 ml/kg, and a positive end-expiratory pressure of 2 cm H²O until the pulmonary artery and left atrium were cannulated, after which the lungs were ventilated at the ventilator settings with a warm humidified gas mixture of 21% O²and 5% CO², and balance N². The right common carotid artery was dissected, and a blood sample of approximately 70 ml was collected. After sternotomy, the pulmonary artery and left atrium were cannulated via  the right and left ventricles and were secured in place with sutures. The heart, lungs, and mediastinal structures were removed en bloc  and placed in a humidified chamber at a temperature of 37–38°C. The isolated lungs were perfused throughout the study at a flow rate of 40 ml · kg⁻¹· min⁻¹using a peristaltic pump (Harvard 1215; Harvard Apparatus) with a mixture of autologous whole blood with a physiological salt solution (PSS). The hematocrit of the perfusate was adjusted to approximately 15%. PSS contained 119 mm NaCl, 4.7 mm KCl, 1.17 mm MgSO⁴, 22.61 mm NaHCO³, 1.18 mm KH²PO⁴, and 3.2 mm CaCl². The following were added to each 100 ml of PSS stock solution: 100 mg dextrose, 20 mU insulin, and 3 g bovine serum albumin (Sigma Chemical Co., St. Louis, MO). The pH of the perfusate was adjusted to 7.35–7.45 by the addition of sodium bicarbonate. Indomethacin (Wako Pure Chemical Industries, Osaka, Japan) was added to the mixture at a dose of 20 μg/ml to reduce the production of vasoactive factors that can potentially produce lung injury. After removal from the chest cavity, the lungs were gently inflated several times until obvious atelectasis had cleared away.

Pulmonary arterial (P^pa) and venous (P^pv) pressures were continuously monitored by transducers connected to amplifiers (model 2238; San-ei, Tokyo, Japan). A flow probe (model FF-050T; Nihon Kohden, Tokyo, Japan) connected to an electromagnetic flow meter (MFV-3100; Nihon Kohden) was placed in the perfusion circuit for continuous monitoring of blood flow. P^pvwas adjusted to 5 mmHg to ensure that P^pvexceeded the mean airway pressure. Zero level for pressures was assigned to the bottom of the lung in the chamber. All signals were digitized using an analog-to-digital converter (DigiData 1200; Axon Instruments, Foster City, CA) and were analyzed using commercially available software (Axograph version 3.0; Axon Instruments).

After 30 min of stabilization, the 56 isolated rabbit lung preparations were divided into eight groups (n = 6 each in isoflurane groups and n = 8 each in sevoflurane groups) based on the anesthetic and potassium-channel inhibitor administered in the study. The concentrations of potassium-channel inhibitors in the perfusate were all between 0.5 and 2 orders of magnitude greater than those reported producing half block of the respective channels. 18,19In the control-isoflurane (control-ISO) and control-sevoflurane (control-SEVO) groups, no potassium-channel inhibitors were added. In the glibenclamide–isoflurane (Glib-ISO) and glibenclamide–sevoflurane (Glib-SEVO) groups, the lungs were pretreated with 10⁻⁵m glibenclamide (Wako Pure Chemical Industries). In the iberiotoxin–isoflurane (IbTX-ISO) and iberiotoxin–sevoflurane (IbTX-SEVO) groups, the lungs were pretreated with 10⁻⁷m iberiotoxin (Wako Pure Chemical Industries), a highly selective K^Ca-channel inhibitor. In the 4-aminopyridine–isoflurane (4AP-ISO) and 4-aminopyridine–sevoflurane (4AP-SEVO) groups, the lungs were pretreated with 10⁻³m 4-aminopyridine (Wako Pure Chemical Industries), a K^V-channel inhibitor. The inhibitors were freshly prepared in a small volume of N , N -dimethylformamide (Wako Pure Chemical Industries) and PSS mixture for glibenclamide and PSS mixture for 4-aminopyridine and iberiotoxin just before each of the experiments. The inhibitors were added into the reservoir at the beginning of stabilization, and their concentrations were calculated on the basis of the total volume of the perfusate. Isoflurane (Abbott Laboratories, North Chicago, IL) or sevoflurane (Maruishi Pharmaceutical, Osaka, Japan) was administered using isoflurane (Acoma Medical Industry, Tokyo, Japan) or sevoflurane (Penlon, Abing, Oxon, UK) vaporizers. The concentrations of isoflurane and sevoflurane in inspired gas were monitored with an anesthetic agent monitor (Anesthetic Gas Monitor Type 1304; Brüel & Kjær, Nærum, Denmark).

The HPV response was induced by switching the inspired gas to a hypoxic gas mixture (3% O², 5% CO², balance N²) for 5 min. This period of 5 min was determined as a sufficient time to obtain the maximum HPV response based on our previous study. 8Based on the previous report from another laboratory that the anesthetic concentration in the perfusate equilibrated with that in the lungs within 10 min, 6the lungs were ventilated for 15 min with each concentration of volatile anesthetics in a normoxia condition, then for 5 min with the hypoxic gas mixture containing the same concentration of anesthetics. Three concentrations of isoflurane (0.5, 1, or 2 minimum alveolar concentration [MAC]) or sevoflurane (0.5, 1, 1.62 MAC) were administered randomly to each perfused lung. After the 5-min HPV response test, a fresh normoxic gas mixture was administered for 8 min to washout the anesthetics and allow P^pato return to the baseline level.

The pulmonary vascular response to HPV was expressed as the difference in P^pa(ΔP^pa) before and after 5 min of hypoxic stimuli. The inhibitory effect of anesthetics was expressed as a percentage of inhibition (%R^max), which represented the value of HPV responses in the presence of anesthetic (R) over the value of HPV response in the absence of anesthetic (R^max). The inhibitory effects of anesthetics on HPV were compared by the concentrations in MAC at which 50% of depression of HPV (ED⁵⁰) occurred. The relation between the concentration in MAC units and response in %R^maxwas analyzed by the Wagner equation 20for sigmoid curves:

where Q is the value of R/(R^max− R) at 1 MAC.

The lungs were used to estimate the tissue wet-to-dry weight ratio. After recording the wet weight of the lung tissue, the lungs were placed in a drying oven at 60°C for 2 weeks and reweighed. Wet-to-dry weight ratio was calculated using the formula:

All data are presented as mean ± SD. Differences within groups were analyzed by the one-way analysis of variance with repeated measures, followed by post hoc  analysis with a Scheffé test (Statview 4.5, Abacus Concepts, Berkeley, CA). ED⁵⁰was compared by analysis of variance and post hoc  analysis with a Scheffé test. ΔP^pain the absence of volatile anesthetics among groups was compared by analysis of variance and post hoc  analysis with Scheffé test. Significance was determined when P  was less than 0.05.

The hematocrit of the perfusate of each group ranged between 14.4 ± 0.5% to 15.2 ± 1.0%. Because there were no differences in pH or oxygen and carbon dioxide partial pressures among groups at different time intervals, we averaged the data of the eight groups. The mean values of pH, oxygen partial pressure, and carbon dioxide partial pressure were 7.37 ± 0.08, 150.1 ± 6.0 mmHg, and 37.4 ± 2.0 mmHg, respectively, during normoxia, and 7.38 ± 0.07, 30.6 ± 3.4 mmHg, and 37.3 ± 1.4 mmHg, respectively, during hypoxia. The lung wet-to-dry weight ratio of each group ranged between 5.0 ± 0.4 and 5.3 ± 0.3, and there was no difference among groups. The effects of isoflurane and sevoflurane on pulmonary vascular pressures and the HPV response in the absence andpresence of potassium-channel inhibitors are shown in table 1(isoflurane) and table 2(sevoflurane).

The K^V-channel inhibition by 4-aminopyridine caused a significant increase in P^pa. The baseline P^pain the 4AP-ISO group was significantly higher than that in the control-ISO group (16.8 ± 4.1 vs.  11.3 ± 0.9 mmHg, P < 0.05). P^pawas significantly higher in the 4AP-SEVO group than in the control-SEVO group (15.5 ± 1.2 vs.  12.1 ± 1.4 mmHg, P < 0.05). There were no significant differences in P^paamong other groups. ΔP^pawas significant augmented by K^V- and K^Ca-channel inhibition (4-aminopyridine and iberiotoxin), but was not affected by K^ATP-channel inhibition (glibenclamide) (fig. 1).

Isoflurane did not affect P^pain normoxia in the absence of potassium-channel inhibitors (table 1). Sevoflurane induced a slight but significant dilation of pulmonary vessels in a dose-dependent fashion: P^padecreased from 12.1 ± 1.4 to 11.1 ± 0.6 mmHg (P < 0.05) with the increase of sevoflurane concentration from 0 to 1.62 MAC in the absence of potassium-channel inhibitors (table 2). The effect of isoflurane on P^paduring normoxia was changed by K^V-channel inhibition with 4-aminopyridine but was not affected by K^Ca- and K^ATP-channel inhibition. P^paincreased in a dose-dependent fashion from 16.8 ± 4.1 to 19.1 ± 3.0 mmHg (P < 0.01) with the increase of isoflurane concentration from 0 to 2 MAC when K^Vchannels were inhibited by 4-aminopyridine (table 1). The effect of sevoflurane on P^pawas not influenced by potassium-channel inhibitions (table 2).

Change in P^pawas suppressed in a dose-dependent fashion in all isoflurane groups (table 1). The dose–response curve shifted to the right by K^V-channel inhibition in the 4AP-ISO group and to the left by K^Ca-channel inhibition in the IbTX-ISO group (fig. 2). The HPV response was inhibited even by 0.5 MAC isoflurane in the IbTX-ISO group, an effect not observed in the control-ISO group. ED⁵⁰obtained from each curve was 0.93 ± 0.08 MAC in the control-ISO group, 0.91 ± 0.22 MAC in the Glib-ISO group, 1.15 ± 0.10 MAC in the 4AP-ISO group, and 0.59 ± 0.18 MAC in the IbTX-ISO group. ED⁵⁰was significantly higher in the 4AP-ISO group (P < 0.05) and lower in the IbTX-ISO group (P < 0.05) than in the control-ISO group, but was not different in the Glib-ISO group from that in the control-ISO group.

Change in P^pawas suppressed in a dose-dependent fashion in all sevoflurane groups (table 2). There was no significant difference in the dose–response curve among sevoflurane groups (fig. 3). There was no significant difference in the dose–response curve between control-ISO and control-SEVO groups. ED⁵⁰calculated from each curve was 0.82 ± 0.21 MAC in the control-SEVO group, 0.87 ± 0.25 MAC in the Glib-SEVO group, 0.96 ± 0.24 MAC in the 4AP-SEVO group, and 0.76 ± 0.17 MAC in the IbTX-SEVO group.

The new findings of this study were as follows: (1) the inhibition of K^Vchannels by 4-aminopyridine treatment and K^Cachannels by iberiotoxin treatment augmented the HPV response; (2) the isoflurane-induced attenuation of HPV was suppressed by K^V-channel inhibition with 4-aminopyridine, augmented by K^Ca-channel inhibition with iberiotoxin, and not affected by K^ATP-channel inhibition with glibenclamide; and (3) the sevoflurane-induced attenuation of HPV was not affected by any of the potassium-channel inhibitors investigated in the current study.

There is ample evidence suggesting that potassium channels play important roles in regulating pulmonary vascular tensioin. 1,2,21In the current study, baseline P^paincreased significantly in 4-aminopyridine groups compared with that in control groups, suggesting that K^Vchannels regulate the basal pulmonary vascular tension in normoxia. The change in alveolar oxygen tension is the primary stimulus for triggering HPV. Although the mechanisms underlying this unique response are unclear, a number of studies suggest that some kinds of potassium channels in pulmonary vessels are sensitive to oxygen. 3,22–24Hypoxia can inhibit K^Vchannels and result in depolarization and increase of Ca²⁺entry through voltage-dependent Ca²⁺channels, and vasoconstriction. 25,26Activation of K^Cachannels may cause membrane hyperpolarization and inactivation of the voltage-gated Ca²⁺channels and result in vasodilation. 13It has been proposed that K^Cachannels appear to serve as a negative feedback pathway to regulate the degree of membrane depolarization and hence vasoconstriction caused by vasoconstrictors. 19,27Increases in intracellular Ca²⁺(by hypoxia) may activate K^Cachannels, which would increase potassium efflux, counteract the depolarization and constriction caused by vasoconstrictors. 27This mechanism would also predict that blockage of K^Cachannels should cause depolarization and vessel constriction. K^ATPchannels are reported to mediate, at least in part, the hypoxic vasodilatation in systemic vascular smooth cells. 28However, K^ATPchannels in pulmonary vessels are normally closed and not opened by hypoxia, and glibenclamide does not affect HPV. 29The current results are consistent with these findings and indicate that K^Vand K^Cachannels may play an important role in the HPV response because inhibition of both channels significantly augmented the HPV response, whereas the inhibition of K^ATPchannel had no significant effect on the HPV response. However, enhanced HPV by inhibition of K^ATPchannel by glibenclamide in the conscious state was reported in an in vivo  study. 11The different results may be a result of different models (in vivo  vs.  in vitro ) used in these studies. Although K^ATPchannels in pulmonary vessels are normally closed in isolated lung preparations, 29they may be in an open status in vivo .

The effects of isoflurane on basal pulmonary vascular tension are consistent with our previous data 8and those of other reports. 6,30Although sevoflurane did not significantly change the pulmonary vascular tension in our previous study, 8it dilated the pulmonary vessels slightly but significantly dose-dependently in the current study. The reason may be that the dilation effect of sevoflurane is subtle and can only be statistically detected with a high concentration of sevoflurane. The other possibility is that we used a higher perfusion flow (vs.  30 ml · kg⁻¹· min⁻¹), and the lungs were isolated and placed in the warm chamber in the current study (vs.  left in the chest cavity in the previous study).

Although K^Vchannels in mesenteric vessels are not affected by isoflurane, 174-aminopyridine–sensitive K^Vchannels are reversibly suppressed by clinically relevant concentrations of isoflurane in the canine coronary artery. 14The results in the current study indicate that the pulmonary vascular effect of isoflurane was changed by K^V-channel inhibition both in normoxia and hypoxia, indicating that isoflurane regulates pulmonary vascular tension through K^Vchannels. The right-side shift of the dose–response curve by treatment with 4-aminopyridine indicates that isoflurane-induced HPV attenuation was partly abolished, suggesting that isoflurane inhibits HPV through the initiatory K^V-channel pathway, because K^V-channel inhibition is thought to be a primary step of the HPV initiation. 2,31Although it is unknown if isoflurane can open or inhibit K^Vchannels in the lung, based on the current results, there is a possibility that isoflurane may strengthen K^V-channel blockage of the K^V-channel inhibitor; as a result, isoflurane caused vasoconstriction in normoxia and inhibited the HPV in a weaker extent in the presence of 4-aminopyridine.

Volatile anesthetics inhibit K^Cachannels in cerebral arteries 15and aortic endothelial cells 32but activate or have no effect on K^Cachannels in small mesenteric arteries. 17The current results, with the dose–response curve shifted to the left side by iberiotoxin treatment, indicate that the HPV inhibitory effect of isoflurane was potentiated by K^Ca-channel inhibition, suggesting that isoflurane modulates HPV through K^Cachannels. It also suggests that isoflurane has a different effect on K^Cachannels compared with K^Vchannels. It is highly possible that isoflurane may activate K^Cachannels because the inhibition of these channels in the absence of volatile anesthetics can potentiate the HPV response.

Activation or inhibition of K^ATPchannels changes the pulmonary vascular dilative or constrictive responses during anesthesia. 30,33In the current study, however, the effect of both isoflurane and sevoflurane on HPV was not influenced by K^ATP-channel inhibition. These different results may be a result of the different models (in vivo  vs.  in vitro ), as discussed previously. Both studies by Seki et al.  33and Fujiwara and Murray 30are in vivo  studies. The current results are consistent with those of another in vitro  study indicating that the K^ATPchannel is not important in maintaining low pulmonary vascular tone in the isolated rat lungs. 34 

The inhibitory effect of sevoflurane is similar to that of isoflurane in the absence of potassium-channel inhibitors. This is consistent with our previous study. 8However, the effect of sevoflurane on pulmonary vascular tension was not significantly affected by these potassium-channel inhibitors in both normoxia and hypoxia (we even increased the group size to 8 per group), suggesting that sevoflurane may regulate pulmonary vascular tension and inhibit HPV through other mechanisms rather than through the currently investigated potassium channels. This is not surprising, because sevoflurane has its own specific chemical structure and properties and therefore may affect the pulmonary vessels in a different fashion. The different effects of sevoflurane from other anesthetics on vessels have been demonstrated previously in the mesenteric, 35coronary arteries, 36and thoracic aorta. 37 

The perfusion flow rate used in the current study was 40 ml · min⁻¹· kg⁻¹, 25–35% of normal pulmonary blood in normal living rabbit. We used this flow rate because our previous study indicated that even a much lower flow (30 ml · min⁻¹· kg⁻¹) can produce a significant HPV response, 8and we consider that higher flow rates should be avoided in the isolated lung preparations, because high flow or high pressure may produce lung edema. Lung edema was not observed in the current study, as confirmed by the wet-to-dry lung weight ratios. Another concern about the current study is the use of the cyclooxygenase inhibitor, because it has been reported that cyclooxygenase inhibition potentiated the magnitude of HPV in both conscious and anesthetized states in a chronically instrumented canine model. 10,11However, cyclooxygenase inhibition may not affect the HPV response in in vitro  models. 1,8,38Cyclooxygenase inhibition did not alter ED⁵⁰and the slope of the dose–response curve of sevoflurane in perfused rabbit lungs 8and the HPV response in isolated pulmonary arteries. 1,38 

Although we demonstrated that potassium channels might play a role in HPV initiation and inhibition by isoflurane, we cannot conclude that the volatile anesthetics have a direct effect (activation or inhibition) on these potassium channels because of a lack of electrophysiological evidence. It is also possible that isoflurane may alter the downstream steps, including Ca²⁺influx. Volatile anesthetics also inhibit HPV through Ca²⁺channels. 12The relation between potassium channels and the inward Ca²⁺current in the effect of volatile anesthetics on HPV remains to be investigated.

In summary, K^Vchannel inhibition with 4-aminopyridine potentiated the HPV response and attenuated the isoflurane-induced HPV inhibition. K^Cachannel inhibition with iberiotoxin also potentiated the HPV response but enhanced the isoflurane-induced HPV inhibition. K^ATPchannel inhibition with glibenclamide had no effect on HPV. The sevoflurane-induced attenuation of HPV was not affected by any of the potassium-channel inhibitors tested in the current study. Our results suggest that K^Caand K^Vchannels may play an important role in the initiation of HPV, isoflurane may modulate the HPV response partially through K^Caand K^Vchannels, and sevoflurane may attenuate HPV through other pathways rather than through the currently investigated potassium channels.

The authors thank Yasuaki Kawai, M.D., Ph.D. (Professor, Department of Physiology, Tottori University Faculty of Medicine, Japan), and Yoshimi Inagaki, M.D., Ph.D. (Associate Professor, Department of Anesthesia and Reanimatology, Tottori University Faculty of Medicine, Japan), for instructive suggestions during revision of the manuscript.