INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ACTIVATION of the ATP-sensitive K⁺ (K_ATP) channels may be an endogenous protective mechanism against cardiac damage during myocardial ischemia (9). K_ATP channels have been also known to play an important role in ischemic preconditioning (21); brief periods of prior ischemia protect the heart against subsequent prolonged ischemia (16). It was suggested recently that mitochondrial rather than sarcolemmal K_ATP channels might be primarily responsible for the cardioprotection of ischemic preconditioning (18, 28).

Ketamine was shown to block K_ATP channels in rat ventricular myocytes (13). The importance of this observation, with respect to a potential link between K_ATP channel activation and ischemic preconditioning, is that ketamine may inhibit the cardioprotective effects of K_ATP channels, presenting a potential peril to patients who have coronary artery disease and are undergoing a variety of surgical procedures. In the preceding studies of this issue, ketamine abolished the cardioprotection induced by preconditioning in isolated rat hearts (14) and in rabbit hearts in vivo (15). However, no studies to date specifically address the precise mechanism of this prevention of ischemic preconditioning by ketamine. Furthermore, even if one might speculate that ketamine may block ischemic preconditioning via inhibition of K_ATP channels, the relevance and specificity of these channels have not been clarified.

Therefore, the purpose of this study was to determine whether ketamine exerts a direct effect on K_ATP channels in isolated rabbit ventricular myocytes, and to determine whether ketamine prevents ischemic preconditioning via K_ATP channel inhibition in rabbit heart slices model of simulated myocardial ischemia and reperfusion. Specifically, we investigated whether ketamine can prevent mitochondrial K_ATP channel-induced preconditioning.

In the present study, we confirmed that ketamine inhibits K_ATP channel activities in cell-attached and inside-out patches of rabbit ventricular myocytes. We also found that ketamine inhibited the cardioprotective effect of preconditioning by anoxia, pinacidil, and diazoxide. Our results provide first direct evidence that ketamine prevents ischemic preconditioning via inhibition of sarcolemmal or mitochondrial K_ATP channels.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Heart Preparation

Electrophysiological Recording and Data Analysis

Channel activity. The threshold for judging the open-state K_ATP channels was set at half single channel amplitude. The open probability (P_o) was calculated using the formula
where t_j is the time spent at current levels corresponding to j = 0, 1, 2,..., N (superscript) channels in the open state, T_d is the duration of the recording, and N is the number of channels active in the patch. The number of channels in a patch was estimated by dividing the maximum current observed, during an extended period at zero ATP, by the mean unitary current amplitude. P_o was calculated over 60-s records. The degree of channel activity was assessed by digitizing segments of current records and expressed as NP_o. The ketamine-inhibitory curve was obtained by fitting a conventional Hill function to the experimental data using a nonlinear least-squares method (8).

Single-channel kinetics. Single-channel kinetics were analyzed with methods following those of Han et al. (8). Briefly, single-channel recordings showing no obvious overlapped channel activity were used in the analysis. Opening and closing events were detected using a standard one-half amplitude threshold procedure. Dwell-time distribution histograms were constructed from single-channel current recordings of more than 60 s. All currents were filtered at 0.2-5 kHz and digitally sampled at 2-20 kHz. Single or multiexponential fits of the open or closed time distribution histograms were done using Fetchan and P-Stat software programs (Axon Instruments). Bins, corresponding to the time of underestimated events assumed equal to twice the digitization time, were excluded from fitting.

Measurement of MDA Content and LDH Release

Preparation of heart slices. After anesthesia, a dissected heart was mounted on a Langendorff apparatus and perfused retrogradely with oxygenated normal Tyrode solution for 5-6 min until all signs of blood were removed. Thin (0.4-0.5 mm thick) slices of heart were prepared using a Stadie-Riggs microtome and were stored in an ice-cold modified Cross-Taggart medium containing (in mM) 130 NaCl, 10 KCl, 1.5 CaCl₂, 5 glucose, and 20 Tris · HCl (pH 7.4). The average weight of heart slices used was 59.6 ± 2.9 mg (n = 251).

Experimental protocol. The protocols are summarized in Fig. 1. After the equilibration period (20 min), heart slices were subjected to a 5-min treatment interval (series 4-20). During this treatment interval, heart slices were alternatively pretreated with 50 µM pinacidil (series 13-16), 100 µM diazoxide (series 17-20), or with anoxic preconditioning by bubbling the organ bath with 95% N₂-5% CO₂ instead of 100% O₂ to simulate ischemia (series 8-12). These pretreatments were also repeated in the presence of 50 µM glibenclamide (series 10, 14, and 18), 100 µM 5-hydroxydecanoate (5-HD, series 11, 15, and 19), or of 100 µM ketamine to determine the role of K_ATP channels and ketamine (series 12, 16, and 20). Heart slices that served as time control (series 1) and vehicle (series 2) or those subjected only to the subsequent anoxic insult alone (series 3) received no treatment during this initial 5-min interval. Because glibenclamide, pinacidil, and diazoxide were dissolved in DMSO at a final concentration <0.1%, the vehicle used for the experiment was physiological solution with DMSO at a final concentration <0.1%. At the end of the treatment (time: 5 min) and after a washout, a 15-min interval followed during which no treatments were administered. Heart slices in all groups (except series 1 and 2) were then made anoxic by replacing 100% O₂ with 95% N₂-5% CO₂ in the buffer for 30 min, followed by a 60-min oxygenated period. LDH release and MDA formation were evaluated at the end of the 60-min oxygenated period (time: 110 min).

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Experimental protocol for rabbit heart slices. Heart slices were equilibrated in a bath for at least 30 min before application of one of the protocols. Excluding time control and vehicle groups (A), all heart slices were made anoxic for 30 min (solid bars). B: nonpreconditioned heart slices were exposed to 30 min anoxia alone. C: in the absence of oxygen, heart slices were underwent anoxic preconditioning alone or in the presence of 50 µM glibenclamide, 100 µM 5-hydroxydecanoate (5-HD), or 100 µM ketamine. D: in the presence of oxygen, heart slices were treated with 50 µM pinacidil or 100 µM diazoxide in the presence and absence of 50 µM glibenclamide, 100 µM 5-HD, or 100 µM ketamine for 5 min. Time point corresponding to data reported in Fig. 4 and 5 is denoted by arrows.

Measurement of LDH release. For the measurement of LDH release, heart slices were homogenized in 2 ml of distilled water, and the tissue homogenate was centrifuged at 2,000 g for 5 min. The pellet was discarded, and the supernatant was saved. LDH activity was determined in the supernatant and incubation medium using a LDH assay kit (Asan Pharm; Kyunggee-do, Korea). Final values were expressed as a percentage of the total LDH released from heart slices.

Lipid peroxidation assay. Lipid peroxidation measured as MDA was estimated with thiobarbituric acid as previously described (7). In the heart homogenate, the MDA content was expressed per milligram of protein. The protein content was determined according to the method of Bradford (2). Heart slices were homogenized in an ice-cold 1.15% KCl (5% wt/vol). A 0.2-ml aliquot of homogenate was mixed with 50 µl of 8.1% sodium dodecyl sulfate and incubated for 10 min at room temperature. Acetic acid (375 µl, 20%, pH 3.5) and 375 µl of thiobarbituric acid (0.6%) were added. The mixture was heated for 60 min in a boiling water bath. The samples were allowed to cool at room temperature. After addition of n-butanol and pyridine (15:1, 1.25 ml), the contents were vigorously vortexed and centrifuged at 1,000 rpm for 5 min. The absorbance of the upper colored layer was measured at 535 nm and 520 nm with a spectrophotometer (Hitachi, U-2000) and compared with freshly prepared 1,1,3,3-tetraethoxypropane standards. Final values were expressed as the relative percentage of the control group.

Solutions and Drugs

The PO₂ was measured using a blood gas analyzer (Nova Stat Profile-3, Nova Bomedical; Waltham, MA). The oxygenation of the incubation medium was maintained by a continuous flow of 100% O₂ to obtain a PO₂ between 27.2 and 31.5 kPa. For the induction of simulated anoxia, the medium was bubbled with 95% N₂-5% CO₂; the PO₂ was ~0.23 kPa. The anoxia-reoxygenation experiment carried out at 37°C.

Ketamine was obtained from Yuhan (Seoul, Korea). Pentobarbital sodium was obtained from Hanlim Pharmaceutical (Gyeonggi-do, Korea). All other reagents used in this study were obtained from Sigma (St. Louis, MO). Ketamine and 5-hydroxydecanoate (5-HD) were dissolved directly in experimental solutions for each experiment. 2,4-Dinitrophenol (2,4-DNP), pinacidil, diazoxide, and glibenclamide were dissolved in DMSO, which in its final concentration did not exceed 0.1%. At this concentration, DMSO did not affect the levels of LDH release and MDA formation or K_ATP channel activity.

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

To investigate whether ketamine affects K_ATP channel activities in intact ventricular myocytes, we studied the effects of ketamine using the cell-attached configuration (Fig. 2A, left). In these experiments, the potential in the patch pipette was held at 0 mV, and thus the membrane potential was held at the resting potential, which was assumed to be about 70 mV with 5.4 mM K⁺ in the extracellular solution. K_ATP channel activity was not observed from cell-attached patches, and only the inward rectifier K⁺ (I_K1) channels were recorded, which was identified from its current-voltage (I-V) relations showing a slope conductance of about 27 pS for the inward current. After an addition of 2,4-DNP (0.2 mM) in bath solution, I_K1 channel depressed gradually, and K_ATP channels having large amplitude were activated within 5 min. The I-V relations for such a channel showed a mean single-channel conductance of about 77 pS and the reversal potential of +70 mV in cell-attached patches, recorded with high KCl concentration solution in the pipette and high NaCl concentration solution in the bath (data not shown). The result suggests that inhibition of ATP synthesis by 2,4-DNP results in the channel activation in the cell-attached patch. In this patch, at least six channel levels were revealed after exposure to 2,4-DNP. Under these experimental conditions, the effect of ketamine on the development of K_ATP channel was tested. In the continuous presence of 0.2 mM 2,4-DNP, addition of 1 mM ketamine reduced NP_o from 1.25 to 0.25. After washout of ketamine, a dramatic increase in channel activity was observed (NP_o = 1.47). The channel induced by 2,4-DNP was inhibited by 50 µM glibenclamide, indicating that it was the K_ATP channel. The same result was observed in all examined patches (n = 15; Fig. 2A, right).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   A: left, effect of ketamine on single channel current of the ATP-sensitive K+ (KATP) channel induced by 2,4-dinitrophenol (DNP) in cell-attached patches. Holding potential was 0 mV. DNP, ketamine, and glibenclamide were added to the bath solution for the periods indicated by bars. Dashed line indicates the zero current level. Data were sampled at 20 kHz and filtered at 2 kHz. Right, histogram showing the effect of ketamine on the KATP channel activity. Data are presented as means ± SE for NPo for the following conditions: 0.2 mM 2,4-DNP alone (a), additional application of 1 mM ketamine (b), washout of 1 mM ketamine (c), and 50 mM glibenclamide in the presence of 0.2 mM 2,4-DNP (d). * Statisically signifcant (P < 0.05). B: left, effect of ketamine on the KATP channel in the inside-out patch membrane exposed to symmetrical 140 mM K+ held at a membrane potential of 60 mV. Solution exchange protocol for ATP and ketamine is shown above current traces. Dashed line indicates the zero current level. Data were sampled at 20 kHz and filtered at 2 kHz. Right, histogram showing the effect of ketamine on the KATP channel activity. Data are presented as means ± SE for NPo for the following conditions: control (a), 1 mM ketamine alone (b), washout of 1 mM ketamine (c), and 500 mM ATP alone (d). * Statistically significant (P < 0.05).

We also studied whether ketamine could inhibit K_ATP channels directly from the intracellular side of ventricular myocytes. When inside-out patches were formed in the ATP-free solution, the activation of the K_ATP channel was observed (Fig. 2B, left). K_ATP channel activity was rapidly and reversibly inhibited by 500 µM ATP. The application of 1 mM ketamine to the bath induced a marked and reversible decrease in K_ATP channel activity. NP_o was reduced from 2.25 to 0.15. This inhibition by ketamine was partially reversible after removal of ketamine (NP_o = 1.42). The same result was observed in 19 other patches (Fig. 2B, right).

Figure 3A shows the concentration-dependent effects of ketamine on K_ATP channel activity. Ketamine inhibited K_ATP channel activity at a concentration starting as low as 1 µM and exhibited further inhibitory effects in a concentration-dependent manner. K_ATP channel activity was maximally inhibited by 1 mM ketamine. To obtain the internal solution concentration of ketamine ([ketamine]_i)-K_ATP channel activity relationship, we determined the effect of one concentration of ketamine from each inside-out patch (Fig. 3B). A plot of relative channel activity as a function of [ketamine]_i, using data obtained from a total of 17 patches, was fitted to the Hill equation using the nonlinear least-square method as described previously (8). The half-maximal inhibitory [ketamine]_i (K_d) was 78.3 ± 15.5 µM. The Hill coefficient was 1.0 ± 0.1. 

View larger version (30K): [in this window] [in a new window]   Fig. 3.   A: representative tracing showing a concentration-dependent inhibition of KATP channel activity by ketamine in the inside-out patch held at 50 mV. Bars above the trace indicate the application of various test solutions. Dashed line indicates the zero current level. Data were sampled at 20 kHz and filtered at 1 kHz. B: concentration-response relationship for the inhibitory action of ketamine. Each data point was obtained from 7 patches. Data were fitted to the Hill equation: relative channel activity = 1/{1 + ([ketamine]i/Kd)n} using the nonlinear least-squares method, where [ketamine]i is ketamine concentration in internal solution, Kd is the ketamine concentration at the half-maximal inhibition; n is the Hill coefficient. Data are presented as means ± SE. C: distribution of open and closed times before (top) and during application of 50 µM ketamine (bottom). Histograms of the open and closed time within bursts were analyzed from the current records at 5 kHz. Histograms of the burst and interburst duration were analyzed from the current records at 0.2 kHz. Membrane potential was 50 mV. Time constants of interburst duration histograms were fitted to two exponentials (fast and slow), the others were fitted to single exponentials. o, Open-time constant; c, closed-time constant; b, burst time constant; c1, c2, fast and slow components of interburst time constant, respectively. D: histogram showing the pooled data (means ± SE) for b and c. * Significant (P < 0.05) different from control value (n = 7 patches).

The open and closed time distributions were analyzed to determine the effect of ketamine on the kinetic properties of the channel (Fig. 3C). Ketamine (50 µM) had no effect on the fast open and closed kinetics within the burst. In the presence of 50 µM ketamine, the exponential time constants of the distribution of open (_o) and closed times (_c) were 1.38 and 0.30 ms, respectively. These values were similar to those in the control solution (_o = 1.40 ms and _c = 0.31 ms). On the other hand, 50 µM ketamine decreased the burst duration and increased the interburst duration. The time constant of the burst duration (_b) was markedly decreased from 21.01 to 11.58 ms by ketamine. It seems the interburst duration histograms were fitted to two exponential functions. In these histograms, the exponential time constants of both the fast (_c1 = 145 ms) and slow (_c2 = 2,769 ms) components were increased by 50 µM ketamine. The values of _c1 and _c2 were 74 and 293 in the control solution. A quantitative analysis of this and six other patches is shown in Fig. 3D.

To further determine any interaction between the ketamine-induced K_ATP channel inhibition and ischemic preconditioning in the heart slice model of simulated myocardial ischemia, the effects of ketamine on the LDH release and lipid peroxidation were evaluated at the end of the 60-min postanoxia in pretreated heart slices with anoxia, pinacidil, or diazoxide. In Figs. 4A and 5A, the LDH release and MDA formation in the time-control group were 7 ± 3% and 220 ± 110 pmol/mg protein, respectively. Vehicle itself did not affect the LDH release and MDA formation under any experimental conditions. In the nonpreconditioned group, the LDH release and MDA formation were significantly increased compared with the time control and vehicle groups. In addition, 50 µM glibenclamide, 100 µM 5-HD, and 100 µM ketamine had no effect on the LDH release and MDA formation. In Figs. 4B and 5B, pretreatment of the heart slices with a single 5-min anoxic interval resulted in a significant reduction in the LDH release and MDA formation. The beneficial effects of 5 min anoxia were prevented by the dose of glibenclamide, 5-HD, and ketamine. Pretreatment with 50 µM pinacidil (Figs. 4C and 5C) or 100 µM diazoxide (Figs. 4D and 5D) also significantly reduced the LDH release and MDA formation to that in the anoxic preconditioning group, with this effect being prevented by glibenclamide, 5-HD, and ketamine, as in the anoxic preconditioning group.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Lactate dehydrogenase (LDH) release of rabbit heart slices 60 min after a 30-min anoxic insult in time control (n = 66 slices), vehicle (n = 14 slices), nonanoxic preconditioning (non-AP) (n = 58 slices), non-AP with vehicle (n = 10 slices), non-AP with glibenclamide (Glib, 50 µM, n = 10 slices), non-AP with 5-HD (100 µM, n = 10 slices), non-AP with ketamine (n = 12 slices), AP with vehicle (n = 10 slices), AP (n = 54 slices), AP with Glib (n = 48 slices), AP with 5-HD (n = 35 slices), AP with ketamine (n = 75 slices), pinacidil (Pin, 50 µM, n = 54 slices), Pin with Glib (n = 25 slices), Pin with 5-HD (n = 55 slices), Pin with ketamine (n = 67 slices), diazoxide (Diaz, 100 µM, n = 39 slices), Diaz with Glib (n = 15 slices), Diaz with 5-HD (n = 61), and Diaz with ketamine (n = 42 slices). Heart slices exposed to the preconditioning by AP (B), Pin (C), and Diaz (D) released significantly less LDH compared with nonpreconditioned heart slices (A) where they were exposed to the 30-min anoxia only. Note that the preconditioning-induced protection was blocked by 50 µM Glib, 100 µM 5-HD, and 100 µM ketamine. Data are presented as means ± SE.  P < 0.05 compared with time control and vehicle. * P < 0.05 compared with non-AP. #P < 0.05 compared with AP. NS, differences are not statistically significant.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Malondialdehyde (MDA) formation of rabbit heart slices 60 min after a 30-min anoxic insult in time control (n = 69 slices), vehicle (n = 14 slices), non-AP (n = 40 slices), non-AP with vehicle (n = 10 slices), non-AP with Glib (n = 10 slices), non-AP with 5-HD (n = 10 slices), non-AP with ketamine (n = 12 slices), AP with vehicle (n = 10 slices), AP (n = 60 slices), AP with Glib (n = 51 slices), AP with 5-HD (n = 42 slices), AP with ketamine (n = 80 slices), Pin (n = 46 slices), Pin with Glib (n = 32 slices), Pin with 5-HD (n = 64 slices), Pin with ketamine (n = 79 slices), and Diaz (n = 43 slices), Diaz with Glib (n = 32 slices), Diaz with 5-HD (n = 55), and Diaz with ketamine (n = 28 slices). A brief treatment with AP (B), Pin (C), and Diaz (D) before the 30 min of anoxia significantly reduced MDA formation. Protection was blocked by glibenclamide, 5-HD, and ketamine. Data are presented as means ± SE.  P < 0.05 compared with time control and vehicle. * P < 0.05 compared with non-AP. #P < 0.05 compared with AP.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Ischemic preconditioning is a phenomenon in which brief episodes of ischemia and reperfusion protect the heart against subsequent lethal ischemia (16). K_ATP channels have been proposed to be the end effector of this protection (5, 21). Several studies of the effects of anesthetics on the K_ATP channel activities in the heart have been performed. Volatile anesthetics, including sevoflurane, isoflurane, and desflurane, have been shown to activate K_ATP channels and have cardioprotective effects (20, 22-24). In contrast, intravenous anesthetics, including ketamine and thiobarbiturates, have been shown to inhibit K_ATP channel activities if applied to cell-attached or inside-out patches (13, 26), suggesting that ketamine and thiobarbiturates may attenuate the cardioprotective effects of K_ATP channels during ischemia and reperfusion. These anesthetics should be used with caution in patients at risk for myocardial ischemia during the intraoperative and perioperative period. Induction of anesthesia with ketamine is associated with increase in cardiac output, arterial blood pressure, and heart rate (10). Because of these cardiovascular-stimulating effects, ketamine is used in all types of anesthesia in patients whose cardiac performance must be maintained or increased. Therefore, it is important to evaluate any interaction between ketamine and ischemic preconditioning.

Ko et al. (13) studied the effects of ketamine on single rat ventricular myocytes and reported a concentration-dependent inhibitory effect of ketamine on K_ATP channel activities. The present study, using isolated rabbit ventricular myocytes, confirms that ketamine inhibits K_ATP channel activities without affecting the channel conductance during simulated ischemia in the cell-attached and inside-out patches. The concentration dependence of ketamine for the K_ATP channel inhibition indicates a first-order reversible reaction between a receptor and ketamine molecule. The findings that ketamine has a stereoselective effect on the myocardium (14, 15) provide further evidence that ketamine's actions are receptor mediated. Because the sulfonylurea receptor is the only receptor known to block K_ATP channels (5), it may be possible target for ketamine. Taken together, these results suggest that ketamine inhibits the K_ATP channel activities in a membrane-delimited manner. In addition, we further investigated the effects of ketamine on the kinetic properties of the channel and found that ketamine decreases the burst duration and increases the interburst duration without affecting the fast open and closed kinetics within burst. Thus it seems that ketamine does not influence the rapid, open, and closed times within the burst. Rather, it may decrease the number of openings within the burst (Fig. 2) and may produce a decrease in the burst durations and an increase in interburst intervals as shown in Fig. 3C, resulting in the decrease in the channel activity. The analysis of amplitude histogram indicated that the channel activity was decreased but the amplitude of unitary current was not affected by ketamine (data not shown).

Ketamine was recently shown to block ischemic preconditioning in isolated rat hearts (14) and in rabbits in vivo (15). On the basis of these findings, however, it is not possible to determine the exact mechanism of the blockade of ischemic preconditioning by ketamine. Although one might assume that K_ATP channels are most likely involved in the effects of ketamine on ischemic preconditioning from the findings that activation of K_ATP channels is involved in the cardioprotection induced by ischemic preconditioning (21, 25) and that ketamine also blocks the K_ATP channels in isolated cells (13), there is no direct evidence linking ketamine and K_ATP channel in ischemic preconditioning. In the present study, several lines of evidence indicate that ketamine prevents the cardioprotective effect of ischemic preconditioning through K_ATP channel inhibition. First, ketamine was able to block the protective effect of preconditioning when it was present during the preconditioning anoxia. Second, prior exposure of heart slices to the K_ATP channel agonist pinacidil before the prolonged anoxia caused a decrease in the level of hypoxia-induced LDH release and MDA formation. This protective effect of pinacidil was blocked by ketamine. The data provide direct evidence that K_ATP channel inhibition is the mechanism of the blockade of ischemic preconditioning by ketamine.

The protection by K_ATP channel was initially thought to be via sarcolemmal K_ATP channels (12). However, recent studies have shown that the mitochondrial K_ATP channel is responsible for the cardioprotection of ischemic preconditioning (17); diazoxide, a selective mitochondrial K_ATP channel opener in cardiac myocytes, reduces myocardial injury, and 5-HD, a selective mitochondrial K_ATP channel inhibitor, eliminates protection from diazoxide and ischemic preconditioning (21, 27, 29). In the present study, several lines of evidence indicate that mitochondrial K_ATP channel is present and is activated during preconditioning anoxia to mediate its cardioprotective effect and is inhibited by ketamine. First, 5-HD was able to block the protective effect of preconditioning when it was present during the preconditioning anoxia. Second, prior exposure of the heart slices to diazoxide before the prolonged anoxia caused a decrease in the levels of LDH release and MDA formation. Third, the cardioprotective effect of diazoxide was blocked not only by 5-HD but also by ketamine. The results provide first demonstration that ketamine abolishes the cardioprotective effects of ischemic preconditioning through inhibition of mitochondrial K_ATP channels. In a different set of experiments, we also observed that intravenous anesthetic pentobarbital sodium prevented the cardioprotection in a manner similar to ketamine (data not shown).

Inherent to models of preconditioning is variability in the methods used to simulate ischemia. In the present study, ischemia was simulated by anoxia, which was achieved by substitution of nitrogen for oxygen. To simulate preconditioning (brief initial preconditioning anoxia), heart slices were exposed to 5 min of anoxia, reoxygenated for 15 min, and then incubated in the presence of continuous anoxia for 30 min to induce injury (second sustained anoxia). Although many studies use a duration of anoxia up to 90 min, 30 min has been used in human models (3, 20). This anoxic period may result in a reversible injury without cell death. Similar to preconditioning in human models, in this study prior anoxic preconditioning lessened the injury by 30-min anoxia.

The current study demonstrated that heart slices from the rabbit in which preconditioning has been conclusively shown in the intact hearts (1) and in in vivo (15) also exhibited preconditioning. The data provided important insight by indicating that heart slices can be also be preconditioned and that the cardioprotective mechanism of preconditioning may be exerted, at least in part, at the level of the cardiac myocytes in the intact heart and in in vivo. The use of this heart slice model for preconditioning and a protocol identical to that employed in preconditioning of isolated perfused heart or in vivo experiments would facilitate cellular characterization of this phenomenon and enable quantitative determination of the extent of cardioprotection by preconditioning. In addition, there are advantages of this experimental model to investigate the mechanism of the blockade of ischemic preconditioning by ketamine. First, the systemic and most humoral side effects of ketamine on ischemic preconditioning can be excluded. Thus the direct myocardial effect of ketamine on ischemic preconditioning can be assessed. Second, the exact concentrations of ketamine and K_ATP channel modulators can be determined. This allows quantification of the extent of cardioprotection or anoxic injury during various interventions.

In this study, ketamine was tested at concentrations ranging from 1 to 1,000 µM. In inside-out patches, ketamine inhibited the K_ATP channel activities at micromolar range with the half-maximal inhibitory concentration of 78.3 µM. Pedraz et al. (19) reported a peak plasma concentration of 63 µM after administration of 10 mg/kg ketamine by intravenous bolus injection in the rabbit. The clinical concentrations of ketamine are up to ~100 µM after induction and 10 µM during maintenance of anesthesia (4). Furthermore, the peak plasma concentration of ketamine is 3-60 µM (11) when patients are anesthetized with intravenous injection of ketamine 2 mg/kg. Combining our observations and these data, it is clear that the concentrations of ketamine used in this study are of physiological relevance and may be sufficient to inhibit K_ATP channels in clinical situations.

In conclusion, the current study demonstrated that ketamine inhibited K_ATP channel activities in rabbit ventricular myocytes and blocked the cardioprotective effect of preconditioning by anoxia, pinacidil, or diazoxide in the heart slices of the rabbit at concentrations as they occur during clinical ketamine anesthesia. These data indicate that inhibition of sarcolemmal or mitochondrial K_ATP channel may contribute, at least in part, to the mechanism of the blockade of ischemic preconditioning by ketamine. To our knowledge, this is the first study specifically address the mechanism of the blockade of ischemic preconditioning by ketamine.