Preconditioning by Isoflurane Induces Lasting Sensitization of the Cardiac Sarcolemmal Adenosine Triphosphate–sensitive Potassium Channel by a Protein Kinase C-δ–mediated Mechanism

The animal use and experimental protocols of this study were approved by the Animal Use and Care Committee of the Medical College of Wisconsin (Milwaukee, Wisconsin).

Ventricular myocytes were isolated from hearts of adult male Wistar rats (150–250 g) by enzymatic dissociation with 0.5 mg/ml collagenase type II (Invitrogen, Carlsberg, CA) and 0.25 mg/ml protease XIV (Sigma-Aldrich, St. Louis, MO) as described previously.24After isolation, the myocytes were stored in the Tyrode solution at 20°–22°C and used for patch clamp experiments within 5 h.

The modified Tyrode solution had the following composition: 132 mm NaCl, 5 mm KCl, 2 mm MgCl², 0.1 mm CaCl², 5 mm HEPES, 5 mm glucose, and 20 mm taurine, adjusted to pH 7.4 with NaOH. The pipette solution contained 60 mm K-glutamate, 50 mm KCl, 10 mm HEPES, 1 mm CaCl², 11 mm EGTA, and 0.5 mm K²ATP, at pH 7.2 adjusted with KOH. The bath solution contained 132 mm N -methyl-d-glucamine, 2 mm MgCl², 1 mm CaCl², 5 mm KCl, and 10 mm HEPES at pH 7.4 adjusted with HCl. Nisoldipine (Miles-Pentex, West Haven, CT) was added to the external solution at 200 nm to block the L-type Ca channels. A 10-mm stock of pinacidil and a 100-mm stock of levcromakalim were prepared in dimethyl sulfoxide (DMSO). The K^ATPchannel blocker glibenclamide was also prepared in DMSO as a 1-mm stock. After dilution in the recording buffer, the final concentrations of DMSO were 0.05, 0.01, and 0.1% for pinacidil, levcromakalim, and glibenclamide, respectively. In control experiments, we tested whether 0.1% DMSO has an effect on rat I^KATP. DMSO alone did not activate I^KATPwhen present in the bath solution during 1-h-long time course experiments, and the whole cell I^KATPelicited by pinacidil was not affected by DMSO. For the latter experiments, the stock of pinacidil was made in 0.1N HCl, and DMSO was applied to the cells when pinacidil-activated current reached the steady state level. Chelerythrine, an isoform-nonselective PKC inhibitor, was dissolved in distilled water to make a 5-mm stock solution. Peptide inhibitors of PKC-ϵ (KIEI 1-1; KAI Pharmaceuticals, San Francisco, CA) and PKC-δ (deltaV1.1/Antennapeida carrier; a gift from Daria Mochly-Rosen, Ph.D., Professor and Chair, Department of Molecular Pharmacology, Stanford University School of Medicine, Stanford, California) were prepared as 10-μm stock solutions in distilled water. The PKC-δ peptide activator (KAD1-1), PKC-ϵ peptide activator (KAE1-1), and inactive carrier peptide (C1), all from KAI Pharmaceuticals, were prepared as 10-μm stock solutions in distilled water. All stock solutions were stored at −20°C and thawed immediately before experiments. The drugs were applied to myocytes in the bath solution, except chelerythrine, which was applied in the pipette solution. Isoflurane (Abbott Laboratories, North Chicago, IL) was dispersed in the bath solution by sonication and delivered to a recording chamber via  perfusate from the airtight glass syringes. At the end of each experiment, samples of isoflurane-containing solution were collected from the recording chamber, and the concentration of isoflurane in the perfusate was analyzed by gas chromatography (Shimadzu, Kyoto, Japan). The average concentration of isoflurane used in this study was 0.56 ± 0.1 mm, equivalent to 1.2 vol% at 22°C.

The I^KATPwas recorded in the whole cell configuration of the patch clamp technique using an EPC-7 amplifier (List, Darmstadt-Eberstadt, Germany), and a Digidata 1322A interface (Axon Instruments, Foster City, CA). The pClamp9 software (Axon Instruments) was used for data acquisition and analysis. Pipettes were pulled from the borosilicate glass (Garner Glass, Claremont, CA) using a PC-84 puller (Sutter, Novato, CA) and were heat polished with an MF-83 microforge (Narishige, Tokyo, Japan). Resistance of the patch pipettes ranged from 1.5 to 2.5 MΩ. Experiments were performed in a recording chamber mounted on the stage of inverted Olympus IMT2 microscope (Tokyo, Japan). Only healthy-looking, rod-shaped, and quiescent myocytes with distinct cross-striations were chosen for patch clamp experiments. After a gigaohm seal was formed, the patch of membrane inside the pipette tip was ruptured, and the whole cell configuration was established. The series resistance was then electronically adjusted to obtain the fastest possible capacitance transient without causing ringing.

The membrane holding potential was set at −40 mV, and the whole cell I^KATPwas monitored over time during a 200-ms depolarizing pulse to 0 mV applied every 15 s. Current amplitude was measured at the end of each voltage step and was normalized to cell capacitance to calculate current density, reported in pA/pF. This way, comparison of results obtained from different cells was possible. To assure equilibration of ATP between the pipette solution and the cytosol, at least 30 min was allowed at the beginning of each experiment, before application of pinacidil or levcromakalim. Current was measured at the point of steady state activation.

Data were analyzed using Clampfit 9 software (Axon Instruments) and Origin 7 software (OriginLab, Northampton, MA). Results are expressed as mean ± SD. Statistical analysis was performed using paired and unpaired Student t  tests. Differences were considered significant at P < 0.05.

Previous studies in isolated guinea pig cardiomyocytes have demonstrated that isoflurane potentiates whole cell I^KATPby sensitizing the channel to opening by pinacidil.22In the current study, the mechanism of isoflurane sensitization was further investigated in rat ventricular myocytes. I^KATPwas measured in the presence of the cardiac K^ATPchannel opener pinacidil (5 μm), and the magnitude of pinacidil-activated I^KATPwas used as an indicator of channel readiness to open. Extracellularly applied pinacidil elicited an outward time-independent current that was sensitive to blockade by glibenclamide (1 μm), confirming identity of the I^KATP. Under control anesthetic-free conditions, pinacidil elicited I^KATPwith density of 3.8 ± 3.7 pA/pF (n = 11; fig. 1A). However, when isoflurane was present before and during application of pinacidil (fig. 1B), opening of the channel was greatly enhanced, and I^KATPwas increased to 17.2 ± 9.5 pA/pF (n = 6). This confirmed our previous findings in guinea pigs that showed sensitization and increased ability of the sarcK^ATPchannel to open in the presence of isoflurane.22 

The above experiments tested immediate effects of isoflurane on the sarcK^ATPchannel in rat myocytes. However, to date, no studies have addressed a possibility of persisting effects of volatile anesthetics on the sarcK^ATPchannel. Such lasting effects could explain the cardioprotection memory of APC. Therefore, we hypothesized that isoflurane induces a prolonged sensitization of the sarcK^ATPchannel, which persists even after anesthetic withdrawal. To test this hypothesis, the following experiments were performed. Voltage clamped myocytes were exposed to isoflurane for 10 min. Anesthetic was then washed out before application of pinacidil. Two different washout periods were tested: 10 and 30 min. A 10-min washout was sufficient to remove all anesthetic, which was confirmed by repeated gas chromatography measurements. After cell exposure to isoflurane and 10 min washout, the density of pinacidil-elicited I^KATPcontinued to be significantly higher (15.6 ± 11.3 pA/pF, n = 12; fig. 2A) compared with the anesthetic-free control. Even after 30 min of anesthetic washout, the sarcK^ATPchannel sensitization was still present, although somewhat attenuated (11.8 ± 3.9 pA/pF, n = 6; fig. 2B). These results showed that channel sensitization by isoflurane persists even after anesthetic withdrawal (fig. 3). To confirm this finding, we used another cardiac-specific K^ATPchannel opener, levcromakalim (10 μm). In control experiments, the magnitude of levcromakalim-activated I^KATPwas 18.3 ± 5.8 pA/pF (n = 6; not shown). After cell pretreatment with isoflurane and 10 min washout of anesthetic, the density of levcromakalim-elicited I^KATPwas significantly increased to 24.5 ± 3.5 pA/pF (n = 6; not shown), indicating that sensitization is not exclusive to pinacidil but can be extended to other K^ATPchannel openers. These results showed that after isoflurane exposure and its washout, the sarcK^ATPchannel remains sensitized to opening, suggesting a possible memory to previous anesthetic exposure.

To elucidate the mechanism of the isoflurane-induced memory, we investigated whether the PKC-mediated signaling is responsible for the prolonged channel sensitization. In these experiments, we first used chelerythrine (5 μm), an isoform-nonselective inhibitor of PKC.25When the sensitization experiments were conducted in the continued presence of chelerythrine, the density of pinacidil-activated I^KATPwas decreased to 6.6 ± 4.6 pA/pF (n = 11; fig. 4A) and therefore was significantly lower than the current recorded without PKC inhibition (fig. 2). Chelerythrine alone, applied under control anesthetic-free conditions, did not affect pinacidil-activated I^KATP. Current density in these experiments was 4.0 ± 5.8 pA/pF (n = 6; fig. 4B). These results suggested that PKC is involved in prolonged channel sensitization by isoflurane.

To determine which PKC isoform mediates isoflurane-induced channel sensitization, we used specific peptide inhibitors of PKC-ϵ and PKC-δ. These isoform-specific inhibitors of PKC translocation were applied in the extracellular solution before addition of isoflurane. As shown in figure 5A, after pretreatment with peptide inhibitor of PKC-δ (100 nm), isoflurane sensitization was abolished, and the pinacidil-activated I^KATPdecreased significantly (7.7 ± 5.4 pA/pF, n = 12). By contrast, PKC-ϵ peptide inhibitor (200 nm) had no effect on isoflurane sensitization, and pinacidil-elicited current was 14.8 ± 9.6 pA/pF (n = 12; fig. 5B). These results suggest that PKC-δ mediates isoflurane-induced channel sensitization. As a positive control, we used specific peptide activators of PKC-δ and PKC-ϵ (200 nm) in place of isoflurane. Interestingly, both PKC-δ and PKC-ϵ activators were able to induce sarcK^ATPchannel sensitization. With PKC-δ activator, the density of I^KATPwas 18.9 ± 7.2 pA/pF (n = 12; fig. 6A). With PKC-ϵ activator, the density of I^KATPwas 18.6 ± 11.1 pA/pF (n = 10; fig. 6B). In addition, when PKC-δ and PKC-ϵ activators were administered before isoflurane exposure and washout, there was no additive effect, and pinacidil-elicited I^KATPwas 15.6 ± 11.0 pA/pF (n = 7) and 13.5 ± 3.1 (n = 4), respectively (data not shown). As a negative control, we used an inactive peptide carrier C1 (200 nm), which did not affect isoflurane-induced channel sensitization, and under these conditions, pinacidil-elicited I^KATPwas 15.2 ± 9.4 pA/pF (n = 5; fig. 7). This suggested that saturation of the effect is reached by the PKC-δ or PKC-ϵ activator and isoflurane alone. These results indicate that although both PKC-δ and PKC-ϵ activation are able to sensitize the sarcK^ATPchannel, activation of PKC-δ seems to be the event mediating isoflurane-induced sensitization of the sarcK^ATPchannel.

In the current study, we used an in vitro  model of single isolated cardiac myocyte to investigate the mechanism of isoflurane-induced APC, in particular, the memory phase of early APC. We focused on the sarcK^ATPchannel and tested the hypothesis that isoflurane produces lasting modulation of this channel. Our results show that in rat ventricular myocytes, isoflurane sensitizes the sarcK^ATPchannel to opening as evidenced by the potentiation of I^KATP, which persists even after anesthetic withdrawal. This suggests that after exposure to anesthetic and its removal, the channel remains primed to opening, indicating a memory of the channel and/or signaling elements upstream. Furthermore, we demonstrated that the mechanism of isoflurane-induced memory involves specific activation of PKC-δ.

A distinct characteristic of preconditioning phenomenon is the memory phase when cardioprotection continues despite removal of the preconditioning stimulus. An early memory phase of APC, which starts immediately after anesthetic exposure and may continue up to 3 h, was the major interest of this study. The mechanism of early memory involves posttranslational modifications of the cellular elements rather then changes in the protein expression26and seems similar in all modes of cardiac preconditioning including APC. It involves activation of G protein–coupled receptors, protein kinases, reactive oxygen species, and the K^ATPchannels. However, thus far, no studies have shown whether and how the putative end-effectors of preconditioning such as K^ATPchannels are modulated during the memory phase of APC. In the current study, we showed that after APC, sarcK^ATPchannel “remembers” previous exposure to anesthetic and remains sensitized to opening. Therefore, during potentially lethal ischemia–reperfusion injury, the primed channel could open more readily and/or to a greater extent and thereby exert its protective effects more efficiently. A similar effect was demonstrated in rabbit cardiomyocytes where agents that are able to precondition the myocardium, adenosine and phorbol 12-myristate 13-acetate, primed the sarcK^ATPchannel to opening, thus decreasing the latency of channel opening, shortening action potential duration during metabolic inhibition, and tending to delay the onset of myocyte hypercontracture.12In addition, it has been demonstrated that in beating guinea pig cardiomyocytes preconditioned by a brief episode of ischemia, opening of the sarcK^ATPchannels during subsequent prolonged ischemia is greatly enhanced, and myocytes tolerate hypoxia better.11 

The exact mechanism by which preconditioning primes the sarcK^ATPchannel to opening is not known but may involve multiple factors. In guinea pig cardiomyocytes, ischemic preconditioning increased the trafficking of sarcK^ATPchannels thereby up-regulating the number of channels in the plasma membrane.11Other studies indicated importance of the channel protein phosphorylation by PKC as a major mechanism of channel priming.24,27Results from our study support the latter hypothesis. PKC has been reported to potentiate sarcK^ATPchannel opening.16,17In the intact rabbit and human cardiomyocytes at low intracellular ATP, activation of PKC by phorbol 12,13-didecanoate elicited whole cell I^KATP.17Also, application of constitutively active PKC to excised inside-out membrane patches increased the sarcK^ATPchannel activity by reducing its sensitivity to ATP inhibition.16This effect was abolished in the presence of active protein phosphatase 2A, suggesting that direct phosphorylation of the channel by PKC is responsible for the potentiation of channel opening.16A phosphorylation site responsible for the PKC-induced modulation of sarcK^ATPchannel was found to be threonine residue T180 on the Kir6.2 subunit, the channel pore.28 

Despite ample evidence demonstrating a direct interaction of PKC with sarcK^ATPchannel, thus far, there are no studies to show which isoform of PKC is involved in preconditioning-induced modulation of the sarcK^ATPchannel. The heart expresses several different isoforms of PKC: conventional (α and β), novel (δ, ϵ, and η), and atypical (λ and ζ).29Of these isoforms, only PKC-δ and -ϵ are indicated to be crucial for the cardioprotection afforded by APC. Volatile anesthetics were shown to activate δ and ϵ isoforms by inducing specific translocation of these isoforms to various cellular locations.20,21,24,30–33This translocation enables a close proximity of the PKC and its substrates and ensures the specificity of their interaction.34Specific translocation sites of δ and ϵ isoforms include sarcolemma and mitochondria, where two putative end-effectors of preconditioning, the sarcolemmal and mitochondrial K^ATPchannels, are located. PKC-δ was shown to translocate to the mitochondria and PKC-ϵ was shown to translocate to the sarcolemma in isolated rat hearts preconditioned by isoflurane in vitro .20However, in the hearts from rats preconditioned by isoflurane in vivo , PKC-δ translocated to the sarcolemma and PKC-ϵ translocated to the mitochondria.21Also, in the model of isolated rat trabeculae, sevoflurane induced translocation of PKC-δ to the sarcolemma, whereas PKC-ϵ distribution did not change.30Similar results were demonstrated in human atrial tissue samples after APC by sevoflurane, where PKC-δ translocated to the sarcolemma and PKC-ϵ translocated to the mitochondria, nuclei, and intercalated discs.32Findings from these and other studies24are contradictory, and this is most likely because of differences in the species used, preconditioning agents, and the type of experimental preparation. These conflicting results reflect the controversy and difficulty in drawing clear-cut conclusions regarding the specific role of each PKC isoform in the preconditioning phenomenon (including ischemic preconditioning and other forms of pharmacologic preconditioning).

In our study, inhibition of PKC-δ abolished isoflurane-induced sensitization of the sarcK^ATPchannel, but inhibition of PKC-ϵ had no effect. However, when activators of PKC-δ and PKC-ϵ were used instead of isoflurane, they sensitized the channel to the same extent as isoflurane. When PKC-δ and PKC-ϵ activators were used together with isoflurane, no additive effect was observed. Based on these findings, we conclude that activation of PKC-δ most likely mediates isoflurane effects. Activation of PKC-ϵ can also sensitize the sarcK^ATPchannel, but based on experiments using the peptide inhibitor of PKC-ϵ, it seems that this isoform does not mediate isoflurane effects on the sarcK^ATPchannel. That PKC-δ or PKC-ϵ activator and isoflurane do not have an additive effect might be explained by the saturation of the effect reached by the PKC-δ or PKC-ϵ activator and isoflurane alone. Takeishi et al.  35demonstrated that specific isoforms of PKC respond differently to different stimuli (hypoxia, ischemia, oxidative stress, angiotensin II). Therefore, one possible explanation is that isoflurane specifically activates and translocates PKC-δ but not PKC-ϵ to the membrane. Another explanation is that activation of PKC-ϵ by the peptide activator may indirectly affect the sarcK^ATPchannel via  other pathways, such as mitochondrial pathway, by initiating cross-talk between mitochondria and sarcK^ATPchannels. Evidence supporting this possibility was given by Aizawa et al. ,24who demonstrated that priming action of PKC-ϵ activator on the sarcK^ATPchannel is blocked by the coadministration of 5-hydroxydecanoate, a mitochondrial K^ATPchannel blocker.

Opening of the cardiac sarcK^ATPchannels during ischemia–reperfusion reduces action potential duration and decreases calcium overload, thus being beneficial for cell survival. However, excessive sarcK^ATPchannel activation might predispose the heart to lethal arrhythmias.36Our study demonstrates that isoflurane exposure sensitizes the sarcK^ATPchannel to opening via  a PKC-mediated mechanism. In this sensitized state, the sarcK^ATPchannel would open more readily during ischemia–reperfusion and exert its protective effects more efficiently. Furthermore, PKC may have additional beneficial effects. It has been demonstrated recently in adult rat myocytes and African green monkey kidney cells (COS-7 cells) expressing recombinant K^ATPchannels that prolonged activation of PKC may prevent excessive and thus detrimental opening of the K^ATPchannels by down-regulating the number of channels in the plasma membrane.37This negative feedback involves channel internalization. Therefore, PKC may have multiple, complex effects on the K^ATPchannel. By potentiating channel activation through phosphorylation and by regulating channel trafficking, PKC may tightly regulate activity of the sarcK^ATPchannel during periods of ischemia and reperfusion.37 

In conclusion, we report a novel finding that APC by isoflurane induces a prolonged potentiation of the sarcK^ATPchannel activity that lasts even after withdrawal of isoflurane. Our results indicate that the δ isoform of PKC, rather than the ϵ isoform, is the mediator of isoflurane effects.

The authors thank Wai-Meng Kwok, Ph.D. (Associate Professor, Departments of Anesthesiology and Pharmacology, Medical College of Wisconsin, Milwaukee, Wisconsin), and Martin Bienengraeber, Ph.D. (Assistant Professor, Departments of Anesthesiology and Pharmacology, Medical College of Wisconsin), for helpful discussions. They also thank Kei Aizawa, M.D. (Research Fellow, Department of Anesthesiology, Medical College of Wisconsin), for assistance in myocyte isolation, as well as Mary Ziebell (Research Technologist, Department of Anesthesiology, Medical College of Wisconsin) for isoflurane measurements.