Cardioprotective Effects of Propofol and Sevoflurane in Ischemic and Reperfused Rat Hearts : Role of KATPChannels and Interaction with the Sodium–Hydrogen Exchange Inhibitor HOE 642 (Cariporide)

Male Sprague Dawley rats (250–300 g) were purchased from Charles-River Canada Ltd (St Constant, Quebec, Canada) or Harland Sprague-Dawley Inc (Indianapolis, IN). The animals were maintained in the Health Sciences Animal Care Facility of the University of Western Ontario in accordance with the guidelines of the Canadian Council on Animal Care (Ottawa, Ontario, Canada).

Rats were killed by decapitation, and the hearts were immediately excised and placed in cold Krebs-Henseleit buffer to stop contractions. Hearts were gently squeezed to remove any blood to prevent clotting. The hearts were picked up by the aorta and prepared for retrograde perfusion using a modified Langendorff method at a constant flow rate of 10 ml/min using a peristaltic pump. The perfusion fluid (p  H 7.4; temperature, 37°C) was Krebs-Henseleit buffer that contained 120 mM NaCl, 4.63 mM KCl, 1.17 mM KH²PO⁴, 1.25 mM CaCl², 1.2 mM MgCl², 20 mM NaHCO³, and 8 mM glucose. The buffer was vigorously gassed with 95% O²/5% CO²before drug addition. Coronary pressure was measured via  a side arm of the perfusion cannula connected to a pressure transducer (Spectramed P23XL, Oxnard, CA). A latex water-filled balloon fixed to a pressure transducer was inserted through the mitral valve into the left ventricle for the determination of left ventricular developed pressure. Rates of pressure development and relaxation (positive and negative dP/dt, respectively) were calculated with a differentiator. Left ventricular end-diastolic pressure was adjusted to approximately 5 mmHg before the start of the experiment by adjusting the volume in the intraventricular balloon with the aid of a micrometer-equipped syringe. Hearts were electrically paced at a rate of 325 beats/min with a stimulator. Pacing was also maintained during the ischemic period. All determinations of ventricular and coronary pressures were obtained on-line on a Pentium 586 computer using a Biopac data analysis system (Biolynx Scientific Equipment, Montreal, Quebec, Canada).

Hearts were initially equilibrated for 15 min, after which either glyburide (10 μM) or its drug vehicle dimethyl sulfoxide (100 μl in 1.2 l perfusate) was added for an additional 5 min. Hearts were then exposed to either propofol (6.2 μg/ml [35 μM]), sevoflurane (delivery and concentration discussed in the following section), pinacidil (1 μM), HOE 642 (5 μM), propofol in combination with HOE 642, or control buffer for an additional 15 min. The glyburide or vehicle was present during this 15 min of drug treatment. Control hearts were perfused with vehicle-containing buffer for 20 min after the 15-min equilibration period. Six hearts were studied in each of the experimental groups.

At the end of the 15-min drug-treatment period, hearts were rendered globally ischemic by stopping the flow for 60 min (zero-flow ischemia), after which reperfusion at the normal flow rate was initiated for an additional 60 min. Recording of left ventricular end-diastolic pressure was taken during the 1-h ischemia. The respective drug treatment was maintained throughout the reperfusion period.

Sevoflurane was injected directly into sealed 4-l glass bottles, each containing 1 l of preoxygenated perfusate (p  H 7.39 ± 0.02, carbon dioxide partial pressure 30 ± 5 mmHg, and oxygen partial pressure 818 ± 31 mmHg) as described by Boban et al.  16A gas-tight syringe equipped with a gas-tight stainless steel valve (Hamilton 1000 series syringe, Hamilton GTS valve H86560; VWR, Mississauga, Ontario, Canada) was used for aspiration and injection of the volatile anesthetic. Specific volumes of sevoflurane were injected immediately into the designated sealed glass bottle using a double stopcock system to avoid leakage and volatilization of the anesthetics before and during injection. The sealed perfusate was stirred continuously to facilitate equilibration of the volatile anesthetics between the liquid and gas phases.

To ensure that sevoflurane concentrations were maintained throughout the complete perfusion period, samples of perfusate were collected at the aortic outlet just before ischemia and 60 min after reperfusion for determination of anesthetic concentrations by gas chromatography using a Varian 3300 gas chromatograph (Mississauga, Ontario, Canada) equipped with a J+W Scientific DB-1 High Resolution Gas Chromatography Column (VWR). Effective volume percent concentrations were calculated as 2.15 vol% for sevoflurane (0.304 mM) before ischemia and 2.14 vol%(0.303 mM) at the end of reperfusion using Krebs-Ringer's solution/gas partition coefficients of 0.36 for sevoflurane at 37°C and 1 atm as described previously. 8 

At the end of the reperfusion period, hearts were clamped with Wollenberger tongs precooled in liquid nitrogen, removed from the cannula, and stored in liquid nitrogen until enzymatic determination for energy metabolites, as previously described. 17Studies were also conducted to determine basal HEP levels before ischemia after the addition of various drug combinations.

All values are given as mean ± SD. One-way analysis of variance and Tukey's post-test for multiple comparisons were used to determine the effects of glyburide, vehicle, propofol, sevoflurane, and HOE 642 on ischemic contracture and metabolite content. Repeated-measures analysis of variance and Tukey's post-test were used to compare the effects on hemodynamic function at each baseline (BL) period and each time interval during postischemic reperfusion. Differences were considered significant at P < 0.05.

Table 1shows no significant differences among the experimental groups with respect to BL left ventricular developed pressure after the initial 15-min equilibration period (BL1). In addition, BL2 shows there was no significant effect of vehicle nor glyburide exposure in any of the groups. BL3 represents the effect of the vehicle or glyburide in the presence of either control buffer, sevoflurane, propofol, or the combination of propofol with HOE 642. Propofol treatment resulted in a 27.3% reduction in left ventricular developed pressure, although this did not reach statistical significance from BL1 or BL2. The combination of propofol and HOE 642 resulted in a 32.6% reduction in left ventricular developed pressure that was significantly lower than that for BL1 and BL2. BL3 for propofol plus HOE 642 was significantly less than the BL3 for controls and glyburide groups; however, there was no further significant differences in the values of BL3 between the remaining groups.

The effect of the drugs on the baseline values of left ventricular pressure development (+dP/dt) and left ventricular relaxation (−dP/dt) parallelled the responses observed with left ventricular developed pressure (data not shown). The left ventricular end-diastolic pressure was initially set at 5 mmHg and was unaffected by any of the drug exposures (data not shown).

Sevoflurane and propofol treatment resulted in a reduction in coronary perfusion pressure (9.5% and 28.3%, respectively); however, there was no significant difference between the groups in the value of the coronary perfusion pressure during the preischemic period (data not shown).

Figure 1depicts the recovery of left ventricular developed pressure during reperfusion. The BL values shown are BL3 from table 1and represent the value of left ventricular developed pressure after 15-min exposure to the agents noted, which were present during the entire reperfusion period. The BL values (BL3 in table 1) were reduced in the hearts exposed to propofol, sevoflurane, and propofol plus HOE 642, making interpretation of recovery difficult. To normalize the data, we calculated the recovery of left ventricular developed pressure as a percentage of the BL value obtained just before ischemia (BL3) for each individual heart. Thus, the recovery of left ventricular developed pressure function in figures 2–5is in the presence of ongoing effect of the agents. The left ventricular end-diastolic pressure was 5 mmHg for all groups just before ischemia; therefore, the recovery of end-diastolic pressure was not normalized, and it is presented as the actual values recorded in figures 2–5.

The effect of propofol and HOE 642 on recovery of left ventricular developed pressure and elevation in left ventricular end-diastolic pressure is presented in figure 2. These results show that propofol improved recovery of left ventricular developed pressure compared with control hearts after 10 min of reperfusion until the end of the reperfusion period. Hearts treated with HOE 642 showed significantly greater recovery than controls at all times during reperfusion, as well as hearts treated with propofol alone for the first 15 min. The group that received the combination of propofol and HOE 642 showed a greater recovery than all other groups for the final 5 min of reperfusion. Propofol, HOE 642, and a combination of these drugs significantly attenuated the increase in left ventricular end-diastolic pressure compared with controls during the entire reperfusion period. For the first 10 min of reperfusion, the groups that received HOE 642 had significantly less increase in left ventricular end-diastolic pressure than did the propofol group. Thus, although recovery associated with propofol and HOE 642 was identical at 60 min of reperfusion, there was clearly a different profile because HOE 642 provided a dramatic recovery of left ventricular developed pressure (preischemic value, 134.6%) as early as 5 min compared with the delay in recovery associated with propofol. In addition, propofol did not attenuate the increase in left ventricular end-diastolic pressure to the same degree as HOE 642 for the initial 10 min of reperfusion.

Figure 3demonstrates the effect of K^ATPchannel opener pinacidil, HOE 642, or sevoflurane on recovery of left ventricular developed pressure and elevation in left ventricular end-diastolic pressure. The left ventricular developed pressure in the sevoflurane group was significantly greater than in controls from 15 min until the end of reperfusion. The left ventricular developed pressure in the pinacidil group was significantly higher than in controls from 20 min until the end of reperfusion. Neither pinacidil nor sevoflurane demonstrated the rapid recovery of left ventricular developed pressure as provided by HOE 642. With respect to left ventricular end-diastolic pressure, groups treated with HOE 642 and sevoflurane significantly attenuated the elevation of left ventricular end-diastolic pressure compared with controls throughout reperfusion. The left ventricular end-diastolic pressure for pinacidil was significantly less than that for controls during the final 50 min of reperfusion.

As shown in figure 4, pretreatment with glyburide had no effect on the ability of propofol to improve recovery of function. On its own, glyburide had no effect on recovery of left ventricular developed pressure, nor did it prevent the elevation in left ventricular end-diastolic pressure.

Figure 5displays the effect of glyburide pretreatment on the functional recovery provided by sevoflurane. As shown in figure 3, sevoflurane improved the recovery of left ventricular developed pressure above that in controls from 15 min until the end of reperfusion. Glyburide had no significant effect on recovery of left ventricular developed pressure. However, glyburide pretreatment attenuated the recovery of left ventricular developed pressure afforded by sevoflurane. The left ventricular developed pressure of hearts exposed to both glyburide and sevoflurane was not significantly different from that of controls for the final 25 min of reperfusion. Glyburide pretreatment also eliminated the protective influence of sevoflurane on the increase in left ventricular end-diastolic pressure observed in reperfusion (fig. 5). As expected, glyburide pretreatment completely eliminated all of the effects of pinacidil on the ischemic reperfused rat hearts (data not shown).

As summarized in table 2, recovery as a percentage of BL3 of the rates of left ventricular pressure development (+dP/dt) and left ventricular relaxation (−dP/dt) parallelled the responses observed with left ventricular developed pressure in all treatment groups. There was no significant differences in coronary perfusion pressure between any of the treatment groups during reperfusion (table 3).

Figure 6demonstrates the effect of cardioprotective treatments (see figs. 2 and 3) on the contracture profile during ischemia per se , that is, before the restoration of flow. Propofol, sevoflurane, HOE 642, and the combination of propofol and HOE 642 significantly reduced the maximum left ventricular end-diastolic pressure reached during ischemia. In addition, these agents significantly delayed the time to reach peak elevation of left ventricular end-diastolic pressure. As shown in the bottom panel of figure 6, the propofol and HOE 642 combination provided a significant additive delay in the time to maximum left ventricular end-diastolic pressure. Interestingly, pinacidil treatment failed to alter the contracture profile when compared with controls.

Figure 7shows the effect of glyburide pretreatment on contracture development during ischemia in hearts exposed to propofol and sevoflurane. Glyburide had no effect on either maximum left ventricular end-diastolic pressure or the time to reach peak left ventricular end-diastolic pressure (data not shown). Moreover, glyburide pretreatment did not affect the influence of propofol on ischemic contracture. However, glyburide did abolish the effects of sevoflurane on maximum left ventricular end-diastolic pressure and the time to reach peak left ventricular end-diastolic pressure.

Table 4summarizes the energy metabolite content in the rat hearts at the end of 60 min of reperfusion. Treatment with propofol, sevoflurane, HOE 642, and propofol plus HOE 642 all resulted in significant sparing of ATP, total HEP, and energy charge compared with controls. Glyburide pretreatment did not affect the metabolite content as compared with controls; however, it eliminated the preservation of ATP, HEP, and energy charge in rat hearts treated with sevoflurane. Moreover, glyburide did not influence the effects of propofol on energy metabolite content.

None of the treatments that preserved ATP or total HEP contents in the reperfused myocardium had any direct effects on these parameters before initiating ischemia (BL3). Thus, basal preischemia values (μM/g dry weight, mean ± SD; n = 4) for these groups for ATP and HEP, respectively, were as follows: controls, 24.1 ± 2.3 and 50.1 ± 2.8; propofol, 27.7 ± 5.6 and 53.7 ± 11.9; sevoflurane, 24.5 ± 5.4 and 49.3 ± 11.6; HOE 642 alone, 24.1 ± 5.6 and 48.5 ± 11.6; propofol plus HOE 642, 21.2 ± 4.8 and 44.4 ± 12.8.

The goal of the first part of our study was to clarify the effects of propofol on the ischemic reperfused myocardium. Our study shows that propofol at clinically relevant concentrations (35 μM) delays the onset and magnitude of ischemic contracture produced by ischemia per se  before reperfusion, allows complete recovery of left ventricular developed pressure during reperfusion after 60 min of ischemia compared with controls, attenuates the increase in left ventricular end-diastolic pressure associated with reperfusion, and preserves HEP levels in reperfused hearts. In addition, we have demonstrated that these effects were not influenced by the pretreatment of glyburide at concentrations that have been shown to effectively inhibit K^ATP, 18,19suggesting that propofol does not provide cardioprotection via  the K^ATPchannel.

This study also included an evaluation of the interaction between propofol and HOE 642 in the ischemic myocardium. Although a number of NHE inhibitors have been developed, HOE 642 is of particular interest in that it specifically inhibits the NHE-1 isoform, the predominant NHE subtype in the heart that provides excellent cardioprotection. 20Moreover, HOE 642 has recently undergone clinical evaluation in a multicentered international study in high-risk patients with acute coronary syndromes, including those undergoing coronary artery bypass surgery. We have proposed that NHE inhibition represents a potentially safe and effective adjunct in cardiac surgery 21and is therefore of significant importance to the anesthesiologist. We recently reported the interaction between isoflurane, sevoflurane, and sufentanil with HOE 642 8; however, the combination with propofol has not been previously studied. Our results suggest distinct and separate mechanisms of protection elicited by these drugs. HOE 642 and propofol provided an equal magnitude of recovery of left ventricular developed pressure and end-diastolic pressure; however, the onset of recovery with HOE 642 was significantly faster than with propofol-treated hearts. Both treatments attenuated the onset and magnitude of contracture during ischemia and reperfusion and preserved ATP content at the end of the reperfusion period. The combination of propofol and HOE 642 provided a superior recovery of left ventricular developed pressure at the end of reperfusion and further delayed the onset of peak contracture, suggesting an additive benefit of separate cardioprotective mechanisms. Indeed, with this drug combination, function after reperfusion was higher than baseline values before ischemia. These results are similar to those obtained with the combination of isoflurane or sevoflurane with HOE 642, 8suggesting that the excellent protection reflects the net effect of two distinct mechanisms. The underlying mechanism for the > 100% recovery in function is unknown at present, but a plausible explanation may involve a sensitizing effect in the reperfused myocardium; however, this needs to be assessed with further studies.

Recently, the K^ATPchannel has been identified as a potentially distinct mechanism for the myocardial protection observed with isoflurane based on the ability of glyburide to attenuate the isoflurane-induced cardioprotection. 12,13Moreover, glyburide has been demonstrated to eliminate the ATP-preserving effect of isoflurane on ischemic reperfused myocardium. 22In the present study, we report for the first time that glyburide pretreatment significantly attenuated the cardioprotection associated with sevoflurane but not propofol. Glyburide eliminated the HEP-sparing effect in sevoflurane-treated hearts after reperfusion. Glyburide also nullified the protective influence of sevoflurane on ischemic contracture and the recovery of left ventricular end-diastolic pressure during reperfusion. Glyburide pretreatment significantly attenuated the recovery of left ventricular developed pressure observed with sevoflurane, but did not abolish the cardioprotection. As noted previously, in view of the fact that we used a glyburide concentration that would be expected to completely block the K^ATPchannel, 18,19the findings suggest that the the K^ATPchannel is unlikely to represent the sole mechanism of protection produced with sevoflurane. In addition, our study confirms that a K^ATPchannel agonist, pinacidil, provides functional cardioprotection in the ischemic reperfused heart, as demonstrated by other investigators. 14,15The profile of recovery was somewhat different to that produced by sevoflurane in that the recovery of left ventricular developed pressure and end-diastolic pressure in pinacidil-treated hearts was slightly delayed in onset as compared with sevoflurane. Moreover, pinacidil was unable to mimic the HEP-sparing effect of sevoflurane, nor did pinacidil attenuate the onset and magnitude of peak ischemic contracture, as did sevoflurane. It should be noted that in our previous study, sevoflurane failed to significantly preserve HEPs as compared with controls. 8The values of ATP content in sevoflurane-treated hearts in this study were similar to those of the previous report; however, the ATP content of controls was significantly lower in the present study. This was surprising but may be a result of the somewhat higher pacing frequency of 325 beats/min used in the present study compared with 300 beats/min used previously, resulting in greater ATP depletion. It is interesting that pinacidil alone failed to preserve HEPs despite the ability of glyburide to reverse sevoflurane-induced preservation of ATP content. Taken together, these findings suggest that although K^ATPchannel activation likely plays an important role in sevoflurane-induced cardioprotection, other mechanisms may contribute to these effects. The failure of glyburide to modulate any parameters in propofol-treated groups suggests that the K^ATPchannel is not involved in the cardioprotection by propofol. The underlying mechanisms for the protective effects of propofol are not known with certainty. However, propofol has been shown to possess antioxidant properties in vitro  23and has been demonstrated to inhibit lipid peroxidation induced by oxidative stress in isolated organelles. 24,25In addition, propofol inhibits the trans-sarcolemmal calcium current in ventricular myocytes. 26–28This is important because lipid peroxidation and calcium overload are associated with myocardial stunning and ischemic reperfusion injury. 29,30However, Coetzee 11reported that propofol failed to protect the pig heart from ischemic reperfusion injury induced by left anterior descending coronary artery occlusion. Recent studies have now demonstrated that propofol attenuates the mechanical derangements and lipid peroxidation induced by hydrogen peroxide and preserves the ATP content. 10Moreover, there has been clinical evidence that propofol reduces lipid peroxidation in ischemic reperfusion injury. 31In addition, a high concentration (100 μM) of propofol has been reported to attenuate ischemic contracture, mechanical dysfunction, lactate dehydrogenase release, and histologic damage in isolated ischemia-reperfused rat hearts. 9Thus, the additive benefits of propofol and HOE 642 may be the result of the combined antioxidant effects of propofol and the reduced calcium overload associated with HOE 642 treatment. Another potential contribution has been suggested in a study involving assessment of coronary artery vasoactivity in which the dilatory effect of propofol was attenuated with nitric oxide synthase and cyclooxygenase inhibitors. 32 

In summary, our study using the isolated rat heart suggests that K^ATPchannel activation, NHE inhibition, and antioxidant effects may all represent distinct pathways of myocardial protection that can be exploited under clinical settings by the anesthesiologist. Our study should be interpreted with some degree of caution, particularly because it was performed using rat hearts, which may not be completely applicable to human tissue. Moreover, we did not study concentration–response relationships for each drug but, instead, relied on concentrations that have been established to produce the relevant effect for the respective agent. It cannot be excluded that different concentrations of these agents could produce other effects. Taken together, however, it is nonetheless attractive to suggest that the different agents used may provide additive protection if used in combination; however, further studies are required to verify these observations in this model of ischemic injury as well as in other animal species.

The authors thank Dr. Wolfgang Scholz, formerly of Hoechst-Marion-Roussel, Frankfurt, Germany, for the generous gift of HOE 642 (cariporide), and Dr. Adrian W. Gelb for his interest in this study and for helpful discussion.