Clinically Relevant Concentrations of Propofol Have No Effect on Adenosine Triphosphate–sensitive Potassium Channels in Rat Ventricular Myocytes

This study was approved by the Animal Investigation Committee of Tokushima University (Tokushima, Japan) and was conducted according to the animal use guidelines of the American Physiologic Society (Bethesda, MD).

Details of the experimental design were similar to those of our previous studies. 10,11Sixty-five male Wistar rats (weight, 250–300 g) were anesthetized with ether, and 1.0 IU/g heparin was injected intraperitoneally 30 min before surgery. Myocytes were obtained enzymatically (0.2 mg/ml collagenase and 0.05 mg/ml pronase) with a Langendorff apparatus. All cells used in this experiment were rod-shaped and striated.

Because propofol is highly lipophilic and might be absorbed to plastic syringe and vinyl chloride tubing, we measured actual propofol concentrations in the superfusate. Propofol, dissolved in dimethyl sulfoxide (final concentration < 0.1%), was diluted in superfusate to final five concentrations (1, 3, 10, 30, and 100 μg/ml) and directly applied to myocytes in the glass-bottom plastic cell bath (2-ml volume) at a rate of 2–2.5 ml/min using a plastic syringe (50-ml volume), vinyl chloride tubing (0.8-mm ID; 50-cm length), and syringe pump (Terumo STC-525; Tokyo, Japan); preliminary studies showed that dimethyl sulfoxide had no significant effects on electrophysiologic and flavoprotein fluorescence measurements. At the end of each measurement (mostly 3 min after the start of each concentration of propofol), 2-ml samples of superfusate in the cell bath were collected for determination of propofol concentrations by high-performance liquid chromatography (EICOM, Kyoto, Japan) with electrochemical detector.

Cell-attached and inside-out patch configurations were applied to record the current through single channels via  patch clamp amplifier, as described by Hamill et al.  12In cell-attached configurations, the bathing solution was composed of the following: 140 mm KCl, 10 mm HEPES, 5.5 mm dextrose, and 0.5 mm EGTA. The pipette solution contained 140 mm KCl, 10 mm HEPES, and 5.5 mm dextrose. For inside-out configurations, the bathing solution (intracellular solution) contained 140 mm KCl, 10 mm HEPES, 5.5 mm dextrose, 1 mm MgCl², and 0.5 mm EGTA. The pipette solution (extracellular medium) was of the same composition as that used in cell-attached experiments. The pH of all solutions was adjusted to 7.3–7.4 with KOH. Patch pipettes were pulled with an electrode puller (PP-830; Narishige, Tokyo, Japan) and coated with Sylgard. The resistance of pipettes filled with internal solution and immersed in the Tyrode solution was 3–4 MΩ. Experiments were conducted with solution temperatures of 36 ± 0.5°C. Channel currents were recorded with a patch clamp amplifier (CEZ 2200; Nihon Kohden, Tokyo, Japan) and stored in a personal computer (Aptiva; International Business Machine Corporation, Armonk, NY) with an analog-to-digital converter (DigiData 1200; Axon Instruments, Foster, CA). Sampling frequency of single-channel data was 5 KHz with a low-pass filter (1 KHz). pClamp version 7 software (Axon Instruments) was used for data acquisition and analysis. The open probability (NPo) was determined from current amplitude histograms and was calculated as follows:

where N is the number of channels in the patch and Pn is the integrated channel opening. The NPo of the K^ATPchannels was determined from recordings lasting longer than 60 s and normalized to the NPo value obtained with 25 μm of 2,4-dinitrophenol 10alone (relative channel activity) in the cell-attached configuration. In the inside-out configuration, the NPo of the K^ATPchannels was normalized to the baseline NPo value obtained before propofol at bathing solution without ATP (relative channel activity). Data points obtained were plotted as actual propofol concentrations compared with relative channel activity. The actual propofol concentration needed to induce 50% inhibition of 2,4-dinitrophenol–induced K^ATPchannel activity (EC⁵⁰) was calculated as described in our previous study. 11The Hill coefficient was calculated as described by Ko et al.  13 

The mitochondrial redox state was monitored by recording the fluorescence of flavin adenine nucleotide-linked enzymes in mitochondria and served as an index of mitochondrial K^ATPchannel activity. Myocytes were superfused with bath solution containing 140 mm NaCl, 5 mm KCl, 1 mm CaCl², 1 mm MgCl², and 10 mm HEPES (pH adjusted to 7.3–7.4 with NaOH). Experiments were conducted with solution temperature of 27 ± 1°C. Fluorescence was monitored microscopically (Eclipse TS100; Nikon, Tokyo, Japan) with a digital CCD camera (ORCA; Hamamatsu Photonics, Hamamatsu, Japan) from one cell at a time by focusing on individual myocytes. Fluorescence of single cells was excited for 100 ms every 10 s. Excitation of flavoprotein was produced by Xenon arc lamp filtered at 450–490 nm, and reflected to the objective lens (×40) of the microscope by a dichroic mirror centered at 505 nm. Emitted fluorescence passed through the dichroic mirror to a 520-nm path filter was recorded and stored in a computer. The redox signal images were analyzed for average pixel intensities of regions of interest per myocyte by means of an image processing system (Aquacosmos; Hamamatsu Photonics). We assessed the effects of diazoxide 11alone and in combination with propofol on flavoprotein fluorescence. The change in fluorescence was normalized to the baseline flavoprotein fluorescence value obtained after exposure to 5 μm of 2,4-dinitrophenol at the end of experiments. In each cell, flavoprotein fluorescence was recorded before (diazoxide alone) and with propofol at five different concentrations. The six data points obtained for each cell were plotted as actual propofol concentrations relative to the normalized flavoprotein fluorescence values. The actual propofol concentration needed to induce 50% inhibition of diazoxide-induced flavoprotein oxidation (EC⁵⁰) was calculated in each cell as described in our previous study. 11The Hill coefficient was calculated as described by Ko et al.  13 

Collagenase (Yakult Co., Tokyo, Japan) and pronase (Sigma Chemical Co., St. Louis, MO) were used for enzymatic dissociation. Propofol (2,6-diisopropylphenol; Aldrich Chemical Co., Milwaukee, WI) and diazoxide (Sigma) were dissolved in dimethyl sulfoxide (final concentration < 0.1%) and prepared as stock solutions. 5-Hydroxydecanoic acid 11was purchased from Biomol Research Laboratories (Plymouth Meeting, PA); 2,4-dinitrophenol (Sigma) and glibenclamide (Sigma) were prepared as stock solutions. All other solutions were made freshly each day.

Data are expressed as mean ± SD. Differences between data sets were evaluated by analysis of variance followed by Scheffé test. P < 0.05 was considered significant, and all P  values were two-tailed.

Actual propofol concentrations in the superfusate were 0.4 ± 0.1, 1.3 ± 0.2, 5.6 ± 0.4, 16.7 ± 0.6, and 60.1 ± 3.6 μg/ml at 1, 3, 10, 30, and 100 μg/ml, respectively (n = 6).

We studied whether propofol could block K^ATPchannels from the outside of ventricular myocytes. As shown in figure 1A, K^ATPchannel activity was inhibited by propofol in a concentration-dependent manner (n = 26). The relations between relative K^ATPchannel activity and concentrations of propofol is shown in figure 1B. Propofol inhibited K^ATPchannel activity significantly at actual concentrations of 5.6 μg/ml or greater in the cell-attached configuration. The EC⁵⁰value was 14.2 μg/ml, and the Hill coefficient was 1.2. A decreased NPo of K^ATPchannel activity induced by propofol returned toward baseline values almost completely after propofol washout.

We also studied whether propofol could block K^ATPchannels directly from the cytosolic side of ventricular myocytes. Because K^ATPchannels are inhibited by intracellular ATP, we studied the activities of these channels in the absence of ATP. When inside-out patches were performed, the channel activities were observed at a holding potential of −50 mV (fig. 2A). These channel activities were blocked by 3 mm of ATP (fig. 3in reference 10) and 10 μm of glibenclamide. The inhibitory effect of propofol on K^ATPchannel activities is shown in figure 2A; channel activities are shown to be reactivated after washout of propofol. The open probabilities were suppressed by increased concentrations of propofol (n = 24). As shown in figure 2B, propofol inhibited relative K^ATPchannel activity significantly at actual concentrations of 5.6 μg/ml or greater in the inside-out configurations. The EC⁵⁰value was 11.4 μg/ml, and the Hill coefficient was 1.06. The average recovery of the K^ATPchannel activity after propofol washout was 93 ± 7% of the NPo obtained before application of propofol.

The current–voltage relation observed from 20 patches before and after application of propofol (16.7 μg/ml) is shown in figure 3. The current–voltage curves before and after the application of propofol were linear in the negative potential range, with single channel conductances of 73 ± 2 and 75 ± 1 pS before and after 16.7 μg/ml propofol, respectively. There was no significant difference between the nontreated and the propofol-treated series, suggesting that propofol does not change the K^ATPchannel conductances in the inside-out configuration.

A representative example of the time course of flavoprotein oxidation in cells exposed to diazoxide alone and in combination with propofol is shown in figure 4A. In the presence of 25 μm diazoxide, subsequent exposure to propofol reduced the diazoxide-induced oxidation of the flavoproteins in a concentration-dependent manner (fig. 4A). After washout of propofol, the diazoxide-induced oxidation was reproducible. The relations between propofol concentration and diazoxide-induced flavoprotein oxidation (n = 12), which was normalized to the flavoprotein fluorescence value obtained with 5 μm of 2,4-dinitrophenol at the end of the experiments, is shown in figure 4B. Propofol inhibited flavopro-tein oxidation significantly at actual concentrations of 5.6 μg/ml and greater. The EC⁵⁰value was 14.6 ± 4.7 μg/ml, and the Hill coefficient was 0.28.

The principal findings of the current study are that propofol has no effect on K^ATPchannel activities at clinically relevant concentrations (< 2 μm), while it inhibits both sarcolemmal K^ATPchannel activity without changing the single-channel conductance and diazoxide-induced flavoprotein fluorescence, which correlates with mitochondrial oxidation and depolarization, at very high, nonclinical concentrations (> 5.6 μg/ml; 31 μm).

Propofol is widely used with fentanyl to achieve safe anesthesia in patients who have coronary artery disease and are undergoing a wide variety of surgical procedures. It is also used for sedation in the intensive care units after coronary artery surgery. Coetzee et al.  1reported that propofol (2 or 6 μg/ml; 11 or 34 μm) failed to provide functional benefit on the reperfused pig myocardium after left anterior descending coronary artery occlusion. However, accumulating evidence suggests that propofol provides cardioprotection in the presence of ischemia–reperfusion injury. 2–7,14–16It has been revealed that propofol (1–10 μm) acts as an antioxidant by reacting with free radicals. 2Moreover, there is clinical evidence that propofol in therapeutic doses attenuates ischemia–reperfusion-induced lipid peroxidation in patients undergoing elective peripheral surgery using a tourniquet. 14Recent study has also demonstrated in isolated rat heart that propofol (25 or 50 μm) attenuates the mechanical derangement and lipid peroxidation induced by exogenously applied hydrogen peroxide and that it preserves ATP content. 3Propofol (6–560 μm) has also been shown to inhibit L-type Ca²⁺currents at supratherapeutic concentrations in isolated guinea pig and rat ventricular myocytes 4,5and decrease transsarcolemmal Ca²⁺influx at concentrations of 30 μm or greater in isolated ferret ventricular myocytes. 15These effects are of potential benefit in reducing the severity of the ischemia–reperfusion injury because lipid peroxidation and Ca²⁺overload are associated with myocardial stunning and ischemic reperfusion injury. 17,18In addition, 100 or 35 μm propofol has been reported to delay or attenuate the onset and magnitude of ischemic injury in isolated ischemia-reperfused rat hearts. 7,16Propofol has been shown to have an endothelium-dependent vasodilatory effect on subepicardial coronary arteries in rats. 6 

In the heart, K^ATPchannels open to induce several protective responses during ischemia and reperfusion. 8It has been thought that activation of sarcolemmal K^ATPchannels protects the ischemic myocardium by shortening the action potential duration. 19As another possible mechanism of cardioprotective action of K^ATPchannels, it was recently suggested that mitochondrial rather than sarcolemmal K^ATPchannels might play an important role in the protection of myocardium during ischemia. 20–22However, the importance of sarcolemmal versus  mitochondrial K^ATPchannel opening during myocardial ischemia is still a matter of controversy. Thus, in the current study, we evaluated and compared the effects of propofol on sarcolemmal and mitochondrial K^ATPchannel activities. In the experiments of Park et al.  6and Mathur et al. , 7demonstrating the cardioprotective effects of propofol, 1 or 10 μm glibenclamide had no effect on the cardioprotection associated with propofol treatment. These results suggest that K^ATPchannels are not involved in the cardioprotective effects of propofol. In the current study, propofol inhibited sarcolemmal K^ATPcurrents and reduced diazoxide-induced flavoprotein fluorescence significantly at greater than 5.6 μg/ml (31 μm). Plasma concentrations of propofol necessary to prevent autonomic responses in humans are 4 μg/ml (22 μm) for major surgery and 3 μg/ml (17 μm) for nonmajor surgery. 23In another report, the propofol blood concentrations at which 50% of patients do not respond to verbal command and to skin incision are 3.3 μg/ml (18 μm) and 15.2 μg/ml (85 μm), respectively. 24After an intravenous induction dose, the peak plasma concentration of propofol is as high as 44 μm, whereas plasma concentrations typically range from 10 to 20 μm for maintenance of anesthesia. 16In addition, because protein binding of propofol exceeds 95%, free fractions of propofol are less than 2 μm. 16Thus, the concentrations of propofol that significantly inhibited K^ATPchannel activities in the current study are very high, nonclinical concentrations, suggesting that propofol has no effect on K^ATPchannels at clinically relevant concentrations (< 2 μm).

Our study had several limitations. First, in our cell-attached configurations, we used 2,4-dinitrophenol to simulate ischemia. Because 2,4-dinitrophenol inhibits mitochondrial ATP synthesis, we were able to observe marked opening of K^ATPchannels. However, activation of K^ATPchannels by 2,4-dinitrophenol may differ from activation by ischemia or hypoxia. 13Second, the high extracellular potassium concentration (140 mm) and resulting membrane depolarization in the patch clamp experiment might have altered the behavior of the channel and the sensitivity of propofol. 13Third, we studied the effects of propofol in isolated rat ventricular myocytes. The effects of propofol on rat myocardium may be different from those on human myocardium. Therefore, we should be careful in extending the current results to the human heart.

In summary, propofol had no effect on the sarcolemmal K^ATPchannel activities in patch clamp configurations and the mitochondrial flavoprotein fluorescence induced by diazoxide at clinically relevant concentrations (< 2 μm), whereas it significantly inhibited both K^ATPchannel activities at very high, nonclinical concentrations (> 5.6 μg/ml; 31 μm).