Opening of mitochondrial ATP-sensitive potassium channels enhances cardioplegic protection

2001;71:1281-1288 Ann Thorac SurgYoshiya Toyoda, Sidney Levitsky and James D. McCully

protectionOpening of mitochondrial ATP-sensitive potassium channels enhances cardioplegic

http://ats.ctsnetjournals.org/cgi/content/full/71/4/1281on the World Wide Web at:

The online version of this article, along with updated information and services, is located

Print ISSN: 0003-4975; eISSN: 1552-6259. Southern Thoracic Surgical Association. Copyright © 2001 by The Society of Thoracic Surgeons.

is the official journal of The Society of Thoracic Surgeons and theThe Annals of Thoracic Surgery

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

http://ats.ctsnetjournals.org/cgi/content/full/71/4/1281

http://ats.ctsnetjournals.org

Opening of Mitochondrial ATP-Sensitive PotassiumChannels Enhances Cardioplegic ProtectionYoshiya Toyoda, MD, Sidney Levitsky, MD, and James D. McCully, PhDDivision of Cardiothoracic Surgery, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts

Background. Mitochondrial and sarcolemmal ATP-sensitive potassium channels have been implicated incardioprotection; however, the role of these channels inmagnesium-supplemented potassium (K/Mg) cardiople-gia during ischemia or reperfusion is unknown.

Methods. Rabbit hearts (n 5 76) were used for Lange-ndorff perfusion. Sham hearts were perfused for 180minutes. Global ischemia hearts received 30 minutes ofglobal ischemia and 120 minutes of reperfusion. K/Mghearts received cardioplegia before ischemia. The role ofATP-sensitive potassium channels in K/Mg cardioprotec-tion during ischemia and reperfusion was investigated,separately using the selective mitochondrial ATP sensi-tive potassium and channel blocker, 5-hydroxydecanoate,and the selective sarcolemmal ATP-sensitive potassiumchannel blocker HMR1883. Separate studies were per-formed using the selective mitochondrial ATP-sensitivepotassium channel opener, diazoxide, and the nonselec-tive ATP-sensitive potassium channel opener pinacidil.

Results. Infarct size was 1.9% 6 0.4% in sham, 3.7% 60.5% in K/Mg, and 27.8% 6 2.4% in global ischemiahearts (p < 0.05 versus K/Mg). Left ventricular peak-developed pressure (percent of equilibrium) at the end of120 minutes of reperfusion was 91% 6 6% in sham, 92%

6 2% in K/Mg, and 47% 6 6% in global ischemia (p <0.05 versus K/Mg). Blockade of sarcolemmal ATP-sensitive potassium channels in K/Mg hearts had noeffect on infarct size or left ventricular peak-developedpressure. However, blockade of mitochondrial ATP-sensitive potassium channels before ischemia signifi-cantly increased infarct size to 23% 6 2% in K/Mg hearts(p < 0.05 versus K/Mg; no statistical significance [NS] ascompared to global ischemia) and significantly decreasedleft ventricular peak-developed pressure to 69% 6 4%(p < 0.05 versus K/Mg). Diazoxide when added to K/Mgcardioplegia significantly decreased infarct size to 1.5% 60.4% (p < 0.05 versus K/Mg).

Conclusions. The cardioprotection afforded by K/Mgcardioplegia is modulated by mitochondrial ATP-sensitive potassium channels. Diazoxide when added toK/Mg cardioplegia significantly reduces infarct size, sug-gesting that the opening of mitochondrial ATP-sensitivepotassium channels with K/Mg cardioplegic protectionwould allow for enhanced myocardial protection in car-diac operations.

(Ann Thorac Surg 2001;71:1281–9)© 2001 by The Society of Thoracic Surgeons

Cardioplegia continues to be the standard method formyocardial protection in cardiac operations. In pre-

vious reports, we have shown that magnesium-supplemented potassium (K/Mg) cardioplegia signifi-cantly decreases infarct size and significantly enhancespostischemic functional recovery [1–3]. The mechanismsby which K/Mg cardioplegia affords cardioprotectionhave been shown to include the modulation of cytosoliccalcium overload, enhanced preservation and resynthe-sis of high energy phosphates, and the modulation ofnuclear and mitochondrial function [3–5].

The end effector of these mechanisms remains to beelucidated; however, recent investigations by us andother researchers have suggested that ATP-sensitive po-tassium (KATP) channels play an important role in endo-genous cardioprotection such as ischemic precondition-ing and adenosine-enhanced ischemic preconditioning

[6, 7]. The role of KATP channels in the cardioprotectionafforded by K/Mg cardioplegia and during ischemia andreperfusion was unknown.

Under normal conditions the KATP channels are closed,this inhibition occurs by free intracellular ATP ([ATP]i)and Mg21–ATP and is responsive to changes in [ATP]i

produced by glycolysis but not by increases throughapplication of exogenous ATP [8]. The opening of KATP

channels during ischemia occurs as intracellular ATPlevels decrease (not extracellular ATP) and has beenpostulated to reduce the action potential plateau phase[8]. However, Grover and colleagues [9] have shown thatwhen action potential duration is blocked, the cardiopro-tective action (decreased infarct size) of the KATP channelopener cromakalim is maintained, suggesting that theKATP channels play an alternative role such as attenuat-ing intracellular Ca21 accumulation thus providing pro-tection from cellular injury and the effects of stunning.

Two KATP channel subtypes exist in the myocardiumwith one type located in the sarcolemma (sarcKATP) [10]and the other in the inner membrane of the mitochondria(mtKATP) [10]. Garlid and associates [11] have demon-strated that mtKATP channels play an important role in

Accepted for publication Nov 6, 2000.

Address reprint requests to Dr McCully, Division of CardiothoracicSurgery, Beth Israel Deaconess Medical Center, Harvard Institutes ofMedicine, 77 Ave Louis Pasteur, Room 144, Boston, MA 02115; e-mail:[email protected].

© 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00Published by Elsevier Science Inc PII S0003-4975(00)02667-9



cardioprotection, and recent reports suggest that mtKATP

channels may be the site of action mediating the cardio-protective effects of ischemic preconditioning [10, 12].

The purpose of this study was to determine whetherthe cardioprotection afforded by K/Mg cardioplegia wasmodulated by KATP channels and if so, to determine thespecificity of KATP channel modulation on K/Mg cardio-protection, and to determine whether this modulationoccurred during ischemia or during reperfusion. Ourresults indicate that infarct size reduction is primarilymodulated by mtKATP channels during ischemia in K/Mgcardioplegia, whereas sarcKATP channels do not appearto be involved in K/Mg cardioprotection. In addition, weshow that the addition of a selective mtKATP channelopener diazoxide [12, 13] to K/Mg cardioplegia beforeischemia significantly decreases infarct size, suggestingthat the cardioprotection afforded by K/Mg cardioplegiacan be enhanced by selective opening of mtKATP

channels.

Material and Methods

AnimalsNew Zealand White rabbits (n 5 76, 15 to 20 weeks; 3 to4 kg) were obtained from Millbrook Farm, Amherst, MA.All animals were housed individually and provided withlaboratory chow and water ad libitum. All experimentswere approved by the Beth Israel Deaconess MedicalCenter Animal Care and Use Committee and the Har-vard Medical Area Standing Committee on Animals(Institutional Animal Care and Use Committee) andconformed to the U.S. National Institutes of Healthguidelines regulating the care and use of laboratoryanimals (NIH publication no. 5377-3, 1996).

Langendorff PerfusionAll rabbits were anesthetized with ketamine (33 mg/kg)and xylazine (16 mg/kg) and heparin (200 U/kg) intrave-nously through the marginal ear vein. The heart wasexcised and placed in a 4°C bath of Krebs-Ringer solutionequilibrated with 95% O2 and 5% CO2 (pH 7.4 at 37°C),where spontaneous beating ceased within a few seconds.Krebs-Ringer solution contained (in mmol/L) NaCl 100,KCl 4.7, KH2PO4 1.1, MgSO4 1.2, NaHCO3 25, CaCl2 1.7,glucose 11.5, pyruvic acid 4.9, and fumaric acid 5.4.Langendorff retrograde perfusion was performed as pre-viously described [1, 5, 6]. In brief, a Latex ballooncontaining a catheter-tip transducer (Millar Instruments,Inc, Houston, TX) was inserted into the left ventricle. Thevolume of the water-filled balloon was determined at aconstant physiological end-diastolic pressure in a rangeof 5 to 10 mm Hg using a calibrated microsyringe duringequilibrium, and this balloon volume was maintained forthe duration of the experiment. The aorta was cannulatedwith a metal cannula and the heart was subjected toLangendorff retrograde perfusion at a constant pressureof 75 cm H2O at 37°C. Hearts were placed through theright atrium at 180 6 3 beats/min throughout the exper-iment using a Medtronic model 5330 stimulator

(Medtronic, Minneapolis, MN). Hemodynamic variableswere acquired using the PO-NE-MAH digital data acqui-sition system (Gould, Valley View, OH), with an AcquirePlus processor board, and left ventricular pressure anal-ysis software, and were expressed as a percentage ofequilibrium values.

Experimental ProtocolHearts were perfused for 30 minutes to establish equilib-rium hemodynamics. Equilibrium was ceased whenheart rate, coronary flow, left ventricular end-diastolicpressure (LVEDP) and peak developed pressure(LVPDP), which is defined as the difference from the leftventricular systolic pressure to the end-diastolic pressurewere maintained at the same level for three continuousmeasurement periods timed 5 minutes apart. Shamhearts (n 5 8) were perfused without global ischemia at37°C for 180 minutes. Global ischemia hearts (GI; n 5 10)were subjected to 30 minutes of GI and 120 minutes ofreperfusion. Global ischemia was achieved by cross-clamping the perfusion line. The K/Mg hearts (n 5 8)were perfused with normothermic (37°C) K/Mg cardio-plegia (K1, 20 mmol/L, Mg21, 20 mmol/L in Krebs-Ringersolution) for 5 minutes before ischemia.

Effect of KATP Channel Blockers on Infarct Size andFunctional RecoveryTo determine the effects of KATP channel blockers onK/Mg cardioprotection from the persistent drug effect ofKATP channel blockers, sham hearts were perfused sep-arately with the selective mtKATP channel blocker, 5-hy-droxydecanoate [10, 14] (5HD; 200 mmol/L in Krebs-Ringer solution; Sigma Chemical Co, St. Louis, MO) andthe selective sarcKATP channel blocker HMR 1883 [10](HMR; 50 mmol/L in Krebs-Ringer solution; the kind giftof H. C. Englert, Hoechst-Marion-Roussel, Frankfurt,Germany) for 7 minutes before ischemia and for 2 min-utes at the onset of reperfusion (sham 1 5HD-IR, n 5 3;sham 1 HMR-IR, n 5 3).

Role of mtKATP Channels in K/Mg CardioprotectionDuring Ischemia and ReperfusionTo investigate the role of mtKATP channels in K/Mgcardioplegic protection during ischemia and reperfusion,K/Mg hearts were perfused separately with 5HD (200mmol/L in Krebs-Ringer solution) for 2 minutes beforeK/Mg cardioplegia infusion and during the 5 minutes ofcardioplegia infusion before GI (K/Mg 1 5HD-I; n 5 8) orfor 2 minutes at the onset of reperfusion (K/Mg 1 5HD-R;n 5 5) or during both periods (K/Mg 1 5HD-IR; n 5 6).

Role of sarcKATP Channels in K/Mg CardioprotectionDuring Ischemia and ReperfusionTo investigate the role of sarcKATP channels in K/Mgcardioplegic protection during ischemia and reperfusion,K/Mg hearts were perfused separately with HMR (50mmol/L in Krebs-Ringer solution) for 2 minutes beforeK/Mg cardioplegia infusion and during the 5 minutes ofK/Mg cardioplegia infusion before GI (K/Mg 1 HMR-I;

1282 TOYODA ET AL Ann Thorac SurgMITOCHONDRIAL KATP CHANNELS IN CARDIOPLEGIA 2001;71:1281–9



n 5 5) or both before ischemia and for 2 minutes at theonset of reperfusion (K/Mg 1 HMR-IR; n 5 5).

Effect of KATP Channel Openers in Global Ischemiaand K/Mg CardioprotectionTo determine the effects of KATP channel openers onK/Mg cardioplegic protection, K/Mg hearts were per-fused separately with the selective mtKATP channelopener diazoxide [12, 13] (50 mmol/L in Krebs-Ringersolution; Sigma Chemical), or with the nonselective KATP

channel opener pinacidil [12, 15] (50 mmol/L in Krebs-Ringer solution; Sigma Chemical) for 5 minutes before GIin concert with K/Mg cardioplegia (K/Mg 1 diazoxide;n 5 6, K/Mg 1 pinacidil; n 5 4). Separate group of GIhearts were perfused with diazoxide (50 mmol/L inKrebs-Ringer solution) for 5 minutes before GI (GI 1diazoxide; n 5 5) in place of K/Mg cardioplegia. Diazox-ide and pinacidil were dissolved in dimethyl sulfoxide(DMSO, Fisher Scientific Co, Fair Lawn, NJ) before beingadded into Krebs solutions. The final concentration ofDMSO was less than 0.1%.

Measurement of Infarct SizeInfarct size was determined as previously describedusing 1% triphenyl tetrazolium chloride (Sigma Chemi-cal) in phosphate buffer (pH 7.4). The area of left ventricleand the area of infarcted tissue were measured by anindependent blinded observer using computer plani-metry as previously described [1, 6].

Wet Weight/Dry Weight RatiosLeft ventricular tissue samples from all experimentalgroups were weighed (wet weight), and dried at 80°C for24 hours for reweighing (dry weight) and then used forthe determination of wet/dry weight ratios, using previ-ously described methods [1, 6].

Statistical AnalysisStatistical analysis was performed using SAS (version6.12) software package (SAS Institute, Cary, NC). Themean 6 the standard error of the mean for all data wascalculated for all variables. Statistical significance wasassessed using repeated measures analysis of variancewith group as a between subjects factor and time as a

within subjects factor. If this overall test was significant,then one-way analysis of variance was performed atindividual time points, and when significant, post hoccomparisons were made between groups at a time point.Dunnett’s test was used for comparisons between K/Mgand other groups. Bonferroni correction was used forcomparisons between groups other than K/Mg. One-wayanalysis of variance was used for infarct size. Post hoccomparisons of infarct size between K/Mg and K/Mg 1diazoxide, and K/Mg and K/Mg 1 pinacidil were per-formed by least significant comparisons analysis. Statis-tical significance was claimed at p value less than 0.05.

Results

Equilibrium HemodynamicsFollowing equilibrium, LVPDP was 106 6 3.9 mm Hg,LVEDP 6.8 6 0.8 mm Hg, 1dP/dt 1,429 6 67 mm Hg/secor coronary flow 44 6 2.6 mL/min in sham hearts. Nosignificant difference in LVPDP, LVEDP, 1dP/dt, or cor-onary flow was observed within or between groups at theend of equilibrium.

Effect of Global Ischemia and K/Mg Cardioplegia onFunctional Recovery and Infarct SizeThe LVPDP in GI hearts was significantly decreased ( p ,0.05 versus sham and K/Mg) throughout 120 minutes ofreperfusion (Table 1). After 20 minutes of reperfusion (80minutes of perfusion) no significant difference in LVPDPwas observed between sham and K/Mg hearts (Table 1).Similar values were observed for 1dP/dt (results notshown). Coronary flow in K/Mg hearts was significantlyincreased at 120 to 180 minutes of perfusion ( p , 0.05versus GI, no statistical significance as compared tosham). No significant difference in coronary flow wasobserved between sham and K/Mg hearts throughoutreperfusion (results not shown).

The LVEDP in GI hearts was significantly increased( p , 0.05 versus sham and K/Mg) throughout reperfu-sion; no significant difference in LVEDP was observedbetween sham and K/Mg hearts throughout reperfusion(results not shown).

Infarct size was 1.9% 6 0.4% in sham hearts, and was

Table 1. Effect of Global Ischemia, K/Mg Cardioplegia, and KATP Channel Blockers on Left Ventricular Peak DevelopedPressure (% of Equilibrium)

Group

Time (min)

70 80 90 120 150 180

Sham 106 (3.3)a 105 (2.8) 104 (1.6) 99 (2.3) 92 (2.6) 91 (6.0)GI 45 (4.2)a 55 (3.9)a 59 (5.4)a 57 (6.5)a 52 (6.4)a 47 (6.3)a

K/Mg 77 (5.2) 95 (5.0) 100 (3.1) 102 (2.7) 97 (2.0) 92 (1.8)Sham 1 5HD-IR 89 (0.8) 93 (0.4) 92 (1.4) 89 (2.2) 86 (2.7) 84 (4.2)Sham 1 HMR-IR 98 (4.1) 99 (3.3) 100 (3.1) 97 (2.5) 94 (2.4) 91 (2.8)

Left ventricular peak developed pressure, expressed as a percentage of equilibrium values, during 120 minutes of reperfusion after 30 minutes of globalischemia, for sham, global ischemia (GI), magnesium-supplemented potassium cardioplegia hearts (K/Mg), and sham hearts perfused separately withthe selective mitochrondrial ATP-sensitive potassium channel blocker, 5-hydroxydecanoate (sham 1 5HD-IR), and the selective sarcolemmalATP-sensitive potassium channel blocker, HMR1883 (sham 1 HMR-IR). All results are shown as the mean 6 standard error of the mean for each group.Significant differences are shown as a for p , 0.05 versus K/Mg.

1283Ann Thorac Surg TOYODA ET AL2001;71:1281–9 MITOCHONDRIAL KATP CHANNELS IN CARDIOPLEGIA



significantly increased to 27.8% 6 2.4% in GI hearts ( p ,0.05 versus sham). Infarct size in K/Mg hearts wassignificantly decreased to 3.7% 6 0.5% ( p , 0.05 versusGI). No significant difference in infarct size was observedbetween sham and K/Mg hearts (Fig 1).

Effect of KATP Channel Blockers on FunctionalRecovery and Infarct SizeTo test the effect of KATP channel blockers on postisch-emic functional recovery and infarct size, sham heartsreceived, separately, 5HD or HMR (sham 1 5HD-IR;sham 1 HMR-IR). No significant difference in LVPDPwas observed during 70 to 180 minutes of perfusionbetween sham and sham 1 5HD-IR and sham 1 HMR-IR

hearts (Table 1). Similar findings were observed forLVEDP, 1dP/dt, and coronary flow (results not shown).Infarct size was 2.2% 6 0.5% in sham 1 5HD-IR, and 1.7%6 0.2% in sham 1 HMR-IR hearts (NS versus sham).

Role of mtKATP Channels in K/Mg CardioprotectionDuring Ischemia and ReperfusionThe LVPDP in K/Mg 1 5HD-I hearts was significantlydecreased at 80 to 120 minutes of perfusion ( p , .05versus K/Mg), but was not significantly different fromthat observed in K/Mg hearts at 150 to 180 minutes ofperfusion (Table 2). The LVPDP in K/Mg 1 5HD-I heartswas significantly increased as compared to GI heartsthroughout the 120 minutes of reperfusion. The LVPDPin 5HD-IR hearts was significantly decreased as com-pared to K/Mg hearts, but was significantly increased ascompared to GI hearts at 80 to 180 minutes of perfusion.Similar findings were observed for 1dP/dt (results notshown). No significant differences in LVEDP or coronaryflow were observed between K/Mg, K/Mg 1 5HD-I,K/Mg 1 5HD-R, and K/Mg 1 5HD-IR hearts throughout120 minutes of reperfusion (results not shown).

Infarct size was significantly increased to 21.5% 6 1.9%in K/Mg 1 5HD-I hearts and 23.0% 6 1.7% in K/Mg 15HD-IR hearts as compared to K/Mg hearts ( p , 0.05)(Fig 1). No significant difference in infarct size wasobserved between K/Mg 1 5HD-I, K/Mg 1 5HD-IR, andGI hearts. No significant difference in infarct size wasobserved in K/Mg 1 5HD-R (6.5% 6 1.2%) as comparedto K/Mg hearts (NS versus K/Mg; p , 0.05 versus GI).

Role of sarcKATP Channels in K/Mg CardioprotectionDuring Ischemia and ReperfusionAfter 20 minutes of reperfusion (80 minutes of perfusion),no significant difference in LVPDP was observed be-tween K/Mg 1 HMR-I and K/Mg 1 HMR-IR hearts ascompared to K/Mg hearts throughout 120 minutes ofreperfusion (Table 3). After 10 minutes of reperfusion (70minutes of perfusion) LVPDP in K/Mg 1 HMR-I andK/Mg 1 HMR-IR was significantly increased ( p , 0.05)as compared to GI hearts throughout reperfusion. Nosignificant differences in LVEDP, dP/dt, or coronary flowwere observed between K/Mg and K/Mg 1 HMR-I and

Fig 1. Infarct size, expressed as a percentage of left ventricular vol-ume, after 30 minutes of global ischemia and 120 minutes of reper-fusion for sham, global ischemia (GI), magnesium-supplementedpotassium cardioplegia hearts (K/Mg), K/Mg hearts perfused withthe selective mitochondrial ATP-sensitive potassium channel blocker5-hydroxydecanoate (5HD), before global ischemia (K/Mg 15HD-I) or at the onset of reperfusion (K/Mg 1 5HD-R) or duringboth periods (K/Mg 1 5HD-IR), and K/Mg hearts perfused withthe selective sarcolemmal ATP-sensitive potassium channel blockerHMR1883 before global ischemia (K/Mg 1 HMR-I) or both beforeischemia and at the onset of reperfusion (K/Mg 1 HMR-IR). Allresults are shown as the mean 6 standard error of the mean foreach group. Significant differences are shown as * for p , 0.05 ver-sus K/Mg and as ** for p , 0.05 versus GI.

Table 2. Role of mtKATP Channels in K/Mg Cardioprotection During Ischemia and Reperfusion: Left Ventricular PeakDeveloped Pressure (% of Equilibrium)

Group

Time (min)

70 80 90 120 150 180

GI 45 (4.2)a 55 (3.9)a 59 (5.4)a 57 (6.5)a 52 (6.4)a 47 (6.3)a

K/Mg 77 (5.2) 95 (5.0) 100 (3.1) 102 (2.7) 97 (2.0) 92 (1.8)K/Mg 1 5HD-I 69 (3.8)a 79 (3.3)a 85 (3.2)a 87 (2.9)a 83 (3.1) 78 (2.5)K/Mg 1 5HD-R 72 (4.5) 81 (5.4) 91 (5.7) 93 (4.0) 88 (3.2) 82 (3.2)K/Mg 1 5HD-IR 60 (6.7)a 75 (5.9)a 83 (3.7)a 85 (3.0)a 78 (3.6)a 71 (4.4)a

Left ventricular peak developed pressure, expressed as a percentage of equilibrium values, during 120 minutes of reperfusion after 30 minutes of globalischemia for sham, global ischemia (GI), magnesium-supplemented potassium cardioplegia hearts (K/Mg), and K/Mg hearts perfused with the selectivemitochondrial ATP-sensitive potassium channel blocker, 5-hydroxydecanoate (5HD) before global ischemia (K/Mg 1 5HD-I) or at the onset ofreperfusion (K/Mg 1 5HD-R) or during both periods (K/Mg 1 5HD-IR). All results are shown as the mean 6 standard error of the mean for each group.Significant differences are shown as a for p , 0.05 versus K/Mg.




K/Mg 1 HMR-IR hearts throughout reperfusion (resultsnot shown).

No significant difference in infarct size was observedbetween K/Mg 1 HMR-I (5.9% 6 0.9%) and K/Mg 1HMR-IR (6.5% 6 0.6%) hearts as compared to K/Mghearts (NS versus K/Mg; p , 0.05 versus GI) (Fig 1).

Effect of KATP Channel Openers in Global Ischemiaand K/Mg CardioprotectionNo significant difference in LVPDP was observed be-tween K/Mg and K/Mg 1 diazoxide hearts throughout120 minutes of reperfusion (Table 4). In contrast LVPDPin GI 1 diazoxide hearts was significantly decreased ascompared with K/Mg hearts ( p , 0.05) throughoutreperfusion. No significant difference in LVPDP wasobserved between GI and GI 1 diazoxide hearts. LVEDPin GI 1 diazoxide hearts was significantly increased ascompared to K/Mg hearts at 180 minutes of perfusion,and 1dP/dt was significantly decreased at 150 to 180minutes of perfusion ( p , 0.05 versus K/Mg; results notshown). No significant difference in coronary flow wasobserved between K/Mg, K/Mg 1 diazoxide, and GI 1diazoxide hearts throughout perfusion.

The K/Mg hearts perfused with pinacidil (K/Mg 1pinacidil) developed ventricular fibrillation upon reper-fusion that lasted 13.5 6 4.3 minutes. The LVPDP inK/Mg 1 pinacidil hearts was significantly decreased ( p ,0.05 versus K/Mg; Table 4, Fig 2) throughout 120 minutesof reperfusion. Similar values were observed for 1dP/dt

(results not shown). Coronary flow in K/Mg 1 pinacidilhearts was significantly decreased throughout reperfu-sion ( p , 0.05 versus sham and K/Mg).

Infarct size in GI 1 diazoxide hearts was significantlydecreased to 13.6% 6 1.4% as compared to GI hearts ( p ,0.05), but was significantly greater ( p , 0.05) than thatobserved in K/Mg hearts (Fig 3). Infarct size in K/Mg 1diazoxide hearts was significantly decreased to 1.5% 60.4% as compared to K/Mg hearts ( p , 0.05). Infarct sizein K/Mg 1 pinacidil hearts was significantly increased to17.4% 6 6.7% ( p , 0.05 versus K/Mg).

Wet Weight/Dry Weight RatiosThe wet weight/dry weight ratio in K/Mg 1 diazoxidehearts was significantly ( p , 0.05) decreased (5.2 6 0.3)as compared to K/Mg 1 pinacidil hearts (7.4 6 0.9). Noother significant differences were observed within orbetween groups.

Comment

In a recent study, we have shown that there is a separa-tion in the modulation of infarct size and functionalrecovery [6]. We and other investigators have suggestedthat mtKATP channels play an important role in modu-lating infarct size [6, 11–14]. Our data presented hereinprovide further evidence to suggest that mtKATP chan-nels are involved in the mechanism modulating infarctsize reduction and that this modulation occurs primarily

Table 3. Role of sarcKATP Channels on K/Mg Cardioprotection During Ischemia and Reperfusion: Left Ventricular PeakDeveloped Pressure (% of Equilibrium)

Group

Time (min)

70 80 90 120 150 180

GI 45 (4.2)a 55 (3.9)a 59 (5.4)a 57 (6.5)a 52 (6.4)a 47 (6.3)a

K/Mg 77 (5.2) 95 (5.0) 100 (3.1) 102 (2.7) 97 (2.0) 92 (1.8)K/Mg 1 HMR-I 64 (6.7)a 76 (5.7)a 84 (5.6) 92 (2.1) 90 (1.4) 84 (1.4)K/Mg 1 HMR-IR 73 (5.0) 81 (4.6) 87 (2.3) 90 (2.3) 84 (2.9) 79 (3.1)

Left ventricular peak developed pressure, expressed as a percentage of equilibrium values, during 120 minutes of reperfusion after 30 minutes of globalischemia for global ischemia (GI), magnesium-supplemented potassium cardioplegia hearts (K/Mg), and K/Mg hearts perfused with the selectivesarcolemmal ATP-sensitive potassium channel blocker HMR1883 before global ischemia (K/Mg 1 HMR-I) or both before ischemia and at the onset ofreperfusion (K/Mg 1 HMR-IR). All results are shown as the mean 6 standard error of the mean for each group. Significant differences are shown as a forp , 0.05 versus K/Mg.

Table 4. Effect of KATP Channel Openers in Global Ischemia and K/Mg Cardioplegia: Left Ventricular Peak DevelopedPressure (% of Equilibrium)

Group

Time (min)

70 80 90 120 150 180

GI 45 (4.2)a 55 (3.9)a 59 (5.4)a 57 (6.5)a 52 (6.4)a 47 (6.3)a

K/Mg 77 (5.2) 95 (5.0) 100 (3.1) 102 (2.7) 97 (2.0) 92 (1.8)K/Mg 1 diazoxide 77 (3.3) 88 (2.0) 95 (3.2) 95 (3.5) 91 (3.1) 87 (3.8)K/Mg 1 pinacidil 15 (15)a 40 (22)a 66 (11)a 69 (6.9)a 60 (5.6)a 51 (6.3)a

GI 1 diazoxide 52 (7.3)a 69 (6.2)a 76 (4.3)a 75 (3.0)a 67 (2.9)a 58 (3.9)a

Left ventricular peak developed pressure, expressed as a percentage of equilibrium values, during 120 minutes of reperfusion for global ischemia (GI),magnesium-supplemented potassium cardioplegia hearts (K/Mg), GI hearts perfused with the selective mitochrondrial ATP-sensitive potassium channelblocker channel opener, diazoxide, for 5 min before global ischemia (GI 1 diazoxide), and K/Mg hearts perfused separately with diazoxide coincidentwith K/Mg cardioplegia (K/Mg 1 diazoxide) or with the nonselective KATP channel opener pinacidil, coincident with K/Mg cardioplegia (K/Mg 1pinacidil). All results are shown as the mean 6 standard error of the mean for each group. Significant differences are shown as a for p , 0.05 versus K/Mg.




during ischemia. Our results indicate that mtKATP chan-nel blockade before ischemia or both before ischemiaand at the immediate start of reperfusion completelyabolished K/Mg infarct size reduction with infarct size inK/Mg 1 5HD-I and K/Mg 1 5HD-IR being not signifi-cantly different from that observed in GI hearts. Incontrast mtKATP channel blockade at the immediate startof reperfusion (K/Mg 1 5HD-R) had no effect on K/Mginfarct size reduction. Postischemic functional recoverywas significantly decreased in K/Mg hearts only whenmtKATP channels were blocked both before ischemia andat the immediate start of reperfusion.

Previous reports have suggested that mtKATP andsarcKATP channels function separately in the modulationof infarct size and functional recovery and that themtKATP but not sarcKATP channels modulate cell viability[6, 9, 11, 12]. Our data would support this hypothesis asour results indicate that the blockade of sarcKATP chan-nels had no effect on the cardioprotection afforded byK/Mg cardioplegia.

The blockade of sarcKATP channels has been shown toincrease cytosolic calcium ([Ca21]i) accumulationthrough the interaction of a series of receptor mediatedevents [10]. Our results indicate that sarcKATP channelblockade before ischemia (K/Mg 1 HMR-I) or bothbefore ischemia and at the immediate start of reperfusion(K/Mg 1 HMR-IR) had only a transient effect on K/Mgcardioprotection affecting postischemic functional recov-ery only during early reperfusion (70 to 80 minutes ofperfusion). The sarcKATP channel blockade had no effect

on K/Mg infarct size reduction. These results are inagreement with previous reports by us [6] and otherresearchers [9, 11, 12] in which sarcKATP channels wereshown not to modulate cell viability.

In this report sarcKATP channel blockade had only atransient effect on K/Mg postischemic functional recov-ery. This is in contrast to our recent report indicating thatsarcKATP channels modulate postischemic functional re-covery with the modified endogenous cardioprotection ofadenosine-enhanced ischemic preconditioning [6]. Thesedifferences are explained by the mechanisms by whichK/Mg cardioplegia and adenosine-enhanced ischemicpreconditioning provide for cardioprotection [3, 16]. Pre-viously, we have shown that the cardioprotection af-forded by K/Mg cardioplegia occurs through the signifi-cant decrease in [Ca21]i accumulation through inhibitionof L-type Ca21 channels and sarcoplasmic reticulumCa21 release channels [5]. The mechanism of action ofadenosine-enhanced ischemic preconditioning is differ-ent from that of K/Mg and does not involve the amelio-ration of [Ca21]i accumulation through Ca21 channels. Inthis report, we have not measured [Ca21]i, however, inprevious reports we have shown that K/Mg cardioplegiaameliorates [Ca21]i accumulation during global ischemia[3, 5]. We speculate that the effects of sarcKATP channelblockade are masked by the decrease in [Ca21]i accumu-lation with K/Mg cardioplegia.

The blockade of mtKATP channels with 5HD has beenpreviously shown by other investigators to decreasemitochondrial depolarization and permit Ca21 entry intothe mitochondria [10, 17]. Under homeostatic conditionsthe mitochondrial inner membrane (cristae) that contains

Fig 3. Infarct size, expressed as a percentage of left ventricular vol-ume, after 30 minutes of global ischemia and 120 minutes of reper-fusion for global ischemia (GI), magnesium-supplemented potassiumcardioplegia hearts (K/Mg), GI hearts perfused with the selectivemitochondrial ATP-sensitive potassium channel opener diazoxide for5 minutes before global ischemia (GI 1 diazoxide), and K/Mghearts perfused separately with diazoxide added to K/Mg cardiople-gia (K/Mg 1 diazoxide) or with the nonselective KATP channelopener pinacidil added to K/Mg cardioplegia (K/Mg 1 pinacidil).All results are shown as the mean 6 standard error of the mean foreach group. Significant differences are shown as * for p , 0.05 ver-sus K/Mg and as ** for p , 0.05 versus GI.

Fig 2. Left ventricular peak developed pressure, expressed as a per-centage of equilibrium values, during 30 minutes of equilibrium, 30minutes of global ischemia, and 120 minutes of reperfusion forglobal ischemia (GI), magnesium-supplemented potassium cardio-plegia hearts (K/Mg), GI hearts perfused with the selective mito-chondrial ATP-sensitive potassium channel opener diazoxide for 5minutes before global ischemia (GI 1 diazoxide), and K/Mg heartsperfused separately with diazoxide coincident with K/Mg cardiople-gia (K/Mg 1 diazoxide) or with the nonselective KATP channelopener pinacidil coincident with K/Mg cardioplegia (K/Mg 1pinacidil). All results are shown as the mean 6 standard error ofthe mean for each group. Significant differences are shown as * forp , 0.05 versus K/Mg.




the electron transport chain expels protons to the cytosol,creating a charge gradient that provides the passiveenergy for Ca21 influx by the Ca21 uniporter. Increasedmitochondrial Ca21 accumulation destabilizes the innermitochondrial membrane, and causes the inner mem-brane pore to open and permit further cation movement(“futile calcium cycling”) [18]. It has been speculated thatthis futile calcium cycling in the mitochondrion, anenergy-dependent process requiring ATP to transportcalcium against the electrochemical gradient out of themitochondrion, uses needed ATP required for the main-tenance of cell viability [17, 19]. Our results showing thatmtKATP channel blockade either before ischemia or bothbefore ischemia and at the immediate start of reperfusioncompletely abolished K/Mg infarct size reduction wouldsupport this mechanism leading to cellular injury.

The mechanism by which cardioplegia provides forenhanced cardioprotection remains to be elucidatedfully, and previous reports by other researchers haveshown that superior cardioprotection can be achievedthrough the addition of nonselective KATP channel open-ers to K1 or Mg21 cardioplegia [20, 21]. The specificrole of sarcKATP or mtKATP channels in cardioplegiccardioprotection, however, was unknown. In this reportwe have used pinacidil, a nonselective KATP channelopener [12, 15] and diazoxide, a selective mtKATP channelopener [12, 13].

Pinacidil, a nonselective KATP channel opener, hasbeen shown to open both sarcKATP and mtKATP channelsin rabbit ventricular myocytes at concentrations of 50 and100 mmol/L [15], and provides dose-dependent myocar-dial protection when used at concentrations between 10and 200 mmol/L [20]. We have used 50 mmol/L pinacidilwith K/Mg cardioplegia. In our investigation, K/Mg car-dioplegia with pinacidil significantly decreased postisch-emic functional recovery ( p , 0.05 versus K/Mg, NSversus GI) and significantly increased infarct size to17.4% 6 6.7% ( p , 0.05 versus K/Mg, NS versus GI).

The opening of the sarcKATP channels with potassiumchannel openers, such as pinacidil, has been shown todecrease [Ca21]i accumulation by hyperpolarization ofthe sarcolemmal membrane [10]. It is important to notethat all hearts receiving K/Mg cardioplegia and pinacidilin our investigation had ventricular fibrillation immedi-ately upon reperfusion that lasted for approximately 14minutes followed by spontaneous defibrillation. Previousreports have shown that ventricular fibrillation results inincreased [Ca21]i, decreased high energy phosphate, andhypoperfusion of the subendocardium because of highend-diastolic pressure, and significantly contributes tocellular injury and decreased functional recovery [22].

Our results are in agreement with Fagbemi and col-leagues [23], who have also reported that in the isolatedbuffer perfused rabbit heart, all hearts treated withpinacidil exhibited ventricular fibrillation upon reperfu-sion. Our results also agree with Dorman and associates[21] who have shown that SR47063 (50 mmmol/L), anonspecific KATP channel opener, when used with car-dioplegia in the in situ blood perfused pig heart, inducedrefractory arrhythmogenesis. They concluded that the

application of nonspecific KATP channel openers as apretreatment may be problematic in the setting of cardiacsurgery. These results, however, are in contrast with thatof Lawton and colleagues [20] who have shown thatpinacidil alone provides superior protection as comparedto St. Thomas’ Hospital cardioplegic solution in theisolated blood perfused rabbit heart model.

Garlid and coworkers [13] have shown that diazoxidedecreases cell injury in a dose-dependent manner atconcentrations between 1 and 30 mmol/L, whereas con-centrations from 30 to 100 mmol/L diazoxide afford asimilar level of cardioprotection. In this study we haveused 50 mmol/L diazoxide to investigate the role ofmtKATP channels in cardioprotection. We have not useddiazoxide during reperfusion as our results indicate thatinfarct size is modulated by mtKATP channels duringischemia, not reperfusion [6].

Our data (Fig 3) indicate that diazoxide, when usedindependently in GI hearts (GI 1 diazoxide), significantlydecreased ( p , 0.05) infarct size as compared to GIhearts, but that infarct size was significantly greater ( p ,0.05) than that observed in K/Mg hearts. Diazoxide whenadded to K/Mg cardioplegia significantly decreased ( p ,0.05) infarct size to 1.5% 6 0.4% as compared to 3.7% 60.5% in K/Mg hearts.

Recent reports suggest that mitochondrial membranedepolarization caused by K1 entry through the openingof mtKATP channels would reduce mitochondrial Ca21

overload [11, 24]. Subsequently, these events are believedto result in ATP production and cell salvage [10, 24]. Inthis study, we have not measured the action potential ofthe sarcolemmal membrane or the oxidation of flavopro-tein [12–14] as indicators of the activities of sarcKATP ormtKATP channels as we have investigated the role ofthese channels in the whole heart model, not the in vitroisolated cardiomyocyte model; however, this mechanismwould agree with our previous report in which we haveshown that K/Mg cardioplegia ameliorates [Ca21]i accu-mulation during ischemia but had no direct effect onmitochondrial Ca21 accumulation [4]. MitochondrialCa21 accumulation was found to be increased similarlyduring ischemia in both GI and K/Mg hearts in themature rabbit [4]. Although we have not measured mi-tochondrial calcium, the mechanism by which mtKATP

channels afford enhanced cardioprotection has been pre-viously suggested by others to occur through a K1

conductance into the mitochondria leading to the depo-larization of the mitochondrial membrane and resultingin increased mitochondrial matrix volume and improvedrespiration through preservation of electron transportfunction [10, 25].

Our results suggest that the opening of mtKATP chan-nels when used with K/Mg cardioplegia would appear toprovide for additive cardioprotection significantly en-hancing the infarct size reduction afforded by K/Mgcardioplegia alone. In this study we have investigated therole of KATP channels in the isolated buffer-perfusedLangendorff heart model and therefore, the effects ofneutrophils and plasma-borne inflammatory compo-nents on cardioprotection were not assessed. However,




in earlier reports, we have shown that the beneficialeffects of K/Mg cardioplegia are preserved in the in situblood perfused heart model [1, 2]. The role of KATP

channels in the cardioprotection afforded by K/Mg car-dioplegia in the blood perfused model remain to beelucidated. It should be noted that in our model electricaldefibrillation could not be performed and therefore theeffects of ventricular fibrillation on infarct size and post-ischemic functional recovery most likely represent a“worst case” example. In total, our results suggest thatthe cardioprotection afforded by K/Mg cardioplegia ismodulated by KATP channels and that the effect of thesechannels on cardioprotection occurs primarily duringischemia. K/Mg infarct size reduction is primarily mod-ulated by mtKATP channels during ischemia. Our resultsalso indicate that diazoxide when added to K/Mg cardio-plegia would appear to enhance infarct size reduction,suggesting that opening of mtKATP channels with K/Mgcardioplegic protection would allow for enhanced myo-cardial protection in cardiac operations.

This study was supported by the National Institutes of Health(HL29077, HL59542) and the American Heart Association.

References

1. McCully JD, Uematsu M, Parker RA, Levitsky S. Adenosine-enhanced ischemic preconditioning provides enhancedpostischemic recovery and limitation of infarct size in therabbit heart. J Thorac Cardiovasc Surg 1998;116:154–62.

2. McCully JD, Uematsu M, Levitsky S. Adenosine-enhancedischemic preconditioning provides myocardial protectionequal to that of cold blood cardioplegia. Ann Thorac Surg1999;67:699–704.

3. McCully JD, Levitsky S. Mechanisms of in vitro cardiopro-tective action of magnesium on the aging myocardium.Magnes Res 1997;10:157–68.

4. Faulk EA, McCully JD, Hadlow NC, et al. Magnesiumcardioplegia enhances mRNA levels and the maximal veloc-ity of cytochrome oxidase I in the senescent myocardiumduring global ischemia. Circulation 1995;92(Suppl 2):405–12.

5. Tsukube T, McCully JD, Federman M, Krukenkamp IB,Levitsky S. Developmental differences in cytosolic calciumaccumulation associated with surgically induced global isch-emia: optimization of cardioplegic protection and mecha-nism of action. J Thorac Cardiovasc Surg 1996;112:175–84.

6. Toyoda Y, Friehs I, Parker RA, Levitsky S, McCully JD.Differential role of sarcolemmal and mitochondrial ATP-sensitive potassium channels in adenosine enhanced isch-emic preconditioning. Am J Physiol Heart Circ Physiol2000;279:H2694–703.

7. Gross GJ, Auchampach J. Blockade of ATP-sensitive potas-sium channels prevents myocardial preconditioning in dogs.Circ Res 1992;70:223–33.

8. Grover G. Pharmacology of ATP-sensitive potassium chan-

nel (KATP) openers in models of myocardial ischemia andreperfusion. Can J Physiol Pharmacol 1997;75:309–15.

9. Grover G, D’Alonso A, Parham C, Darbenzio R. Cardiopro-tection with the KATP opener cromakalim is not correlatedwith ischemic myocardial action potential duration. J Car-diovasc Pharmacol 1995;26:145–52.

10. Gross GJ, Fryer RM. Sarcolemmal versus mitochondrialATP-sensitive K1 channels and myocardial precondition-ing. Circ Res 1999;84:973–9.

11. Garlid K, Paucek P, Yarov-Yarovoy V, et al. Cardioprotectiveeffect of diazoxide and its interaction with mitochondrialATP-sensitive K1 channels: possible mechanism of cardio-protection. Circ Res 1997;81:1072–82.

12. Liu Y, Sato T, O’Rourke B, Marban E. Mitochondrial ATP-dependent potassium channels: novel effectors of cardiopro-tection? Circulation 1998;97:2463–9.

13. Garlid KD, Paucek P, Yarov-Yarovoy V, Sun X, Schindler PA.The mitochondrial KATP channel as a receptor for potas-sium channel openers. J Biol Chem 1996;271:8796–9.

14. Sato T, O’Rourke B, Marban E. Modulation of mitochondriaATP-dependent K1 channels by protein kinase C. Circ Res1998;83:110–4.

15. Hu H, Sato T, Seharaseyon J, et al. Pharmacological andhistochemical distinctions between molecularly defined sar-colemmal KATP channels and native cardiac mitochondrialKATP channels. Mol Pharmacol 1999;55:1000–5.

16. McCully JD, Levitsky S. Alternatives for myocardial protec-tion: adenosine-enhanced ischemic preconditioning. AnnN Y Acad Sci 1999;874:295–305.

17. Gunter TE, Pfeiffer DR. Mechanisms by which mitochondriatransport calcium. Am J Physiol 1990;258:C755–86.

18. Frolkis VV, Frolkis RA, Mkhitarian LS, et al. Contractilefunction and Ca21 transport system of myocardium in age-ing. Gerontology 1988;34:64–74.

19. Peng CF, Kane JJ, Murphy ML, Straub KD. Abnormal mito-chondrial oxidative phosphorylation of ischemic myocar-dium by calcium chelating agents. J Mol Cell Cardiol 1977;9:897–908.

20. Lawton JS, Sepic JD, Allen CT, Hsia PW, Damiano RJ.Myocardial protection with potassium-channel openers is aseffective as St. Thomas’ solution in the rabbit heart. AnnThorac Surg 1996;62:31–8.

21. Dorman BH, Hebbar L, Zellner JL, et al. ATP-sensitivepotassium channel activation before cardioplegia. Effects onventricular and myocyte function. Circulation 1998;98:II176–83.

22. Kusuoka H, Chacko VP, Marban E. Myocardial energeticsduring ventricular fibrillation investigated by magnetizationtransfer nuclear magnetic resonance spectroscopy. Circ Res1992;71:1111–22.

23. Fagbemi SO, Chi L, Lucchesi BR. Antifibrillatory and profi-brillatory actions of selected class I antiarrhythmic agents.J Cardiovasc Pharmacol 1993;21:709–19.

24. Holmuhamedov EL, Jovanovic S, Dzeja PP, Jovanovic A,Terzic A. Mitochondrial ATP-sensitive K1 channels modu-late cardiac mitochondrial function. Am J Physiol 1998;275:H1567–76.

25. Halestrap A. Regulation of mitochondrial metabolismthrough changes in matrix volume. Biochem Soc Trans 1994;22:522–8.

INVITED COMMENTARY

Potassium-based cardioplegic solutions have long beenused by surgeons to reduce the damaging effects ofischemia. The basic premise has been that depolarizedarrest reduces cardiac work and myocardial energy de-

mands, thereby increasing ischemic tolerance. Thismanuscript by Toyoda and colleagues provides new datathat suggests that other mechanisms may be involved. Ina carefully controlled series of experiments they suggest


© 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00Published by Elsevier Science Inc PII S0003-4975(01)02482-1



2001;71:1281-1288 Ann Thorac SurgYoshiya Toyoda, Sidney Levitsky and James D. McCully

protectionOpening of mitochondrial ATP-sensitive potassium channels enhances cardioplegic

& ServicesUpdated Information

http://ats.ctsnetjournals.org/cgi/content/full/71/4/1281including high-resolution figures, can be found at:

References http://ats.ctsnetjournals.org/cgi/content/full/71/4/1281#BIBL

This article cites 23 articles, 15 of which you can access for free at:

Citationshttp://ats.ctsnetjournals.org/cgi/content/full/71/4/1281#otherarticlesThis article has been cited by 15 HighWire-hosted articles:

Subspecialty Collections

http://ats.ctsnetjournals.org/cgi/collection/myocardial_protection Myocardial protection

following collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

[email protected]: orhttp://www.us.elsevierhealth.com/Licensing/permissions.jsp

in its entirety should be submitted to: Requests about reproducing this article in parts (figures, tables) or

Reprints [email protected]

For information about ordering reprints, please email:


http://ats.ctsnetjournals.org/cgi/content/full/71/4/1281

http://ats.ctsnetjournals.org/cgi/content/full/71/4/1281#BIBL

http://ats.ctsnetjournals.org/cgi/collection/myocardial_protection

http://www.us.elsevierhealth.com/Licensing/permissions.jsp


Opening of mitochondrial ATP-sensitive potassium channels enhances cardioplegic protection

Documents