Opening of mitochondrial K(ATP) channels triggers cardioprotection. Are reactive oxygen species involved?

Ischemic preconditioning, a phenomenon in which brief episodes of ischemia and reperfusion paradoxically protect the heart against subsequent lethal ischemia,1 has been conceptually divided into triggers and mediators/effectors.2,3 The trigger, which acts before the index ischemia, is followed by the protection of mediators/effectors during the lethal ischemia. Known triggers include activation of adenosine receptors, a1-adrenegic receptors and opioid receptors, elevated intracellular Ca, and increased reactive oxygen species (ROS).2 The mitochondrial ATP-dependent potassium channel (mitoKATP) has been proposed to be the mediator of this protection.4 The link between the trigger and effector may be the activation of protein kinases (eg, protein kinase C, tyrosine kinase, and downstream kinases), which may phosphorylate mitoKATP, causing the channel to open early and/or to a greater extent to reduce injury during the lethal ischemia. Interestingly, opening of mitoKATP can also trigger cardioprotection. Hearts treated with the mitoKATP opener diazoxide for a brief period before ischemia had significantly smaller infarction.3,5 The triggering effect from diazoxide can be blocked by protein kinase C5,6 and tyrosine kinase inhibitors,3 suggesting that diazoxide activates protein kinases, acting similarly to other triggers. This effect was lost when ROS scavengers were coadministrated with diazoxide.3 ROS are known to activate protein kinases and act as a trigger.2 Thus, it was proposed by Pain et al3 that the opening of mitoKATP by diazoxide may increase ROS. In this issue of Circulation Research, Forbes et al7 provide a direct demonstration that opening of mitoKATP increases ROS production in isolated rat ventricular myocytes. Using a ROSsensitive fluorescent probe 29,79-dichlorofluorescin (DCF), they showed that diazoxide as well as pinacidil (a nonselective KATP opener) increased DCF fluorescence, implying an elevated ROS production. Furthermore, a selective mitoKATP blocker 5-hydroxydecanoate (5-HD) abolished the increase. The antioxidants N-acetylcysteine or N-mercaptopropionylglycine also blocked the fluorescence increase. Exposing hearts to diazoxide before ischemia improved functional recovery after 20 minutes of global ischemia in their isolated rat heart model, although a classical trigger effect was not tested, because there was no diazoxide washout period before the ischemia. Although there is no question that a mild ROS stress can trigger cardioprotection by activating protein kinases,2 the mechanistic links between mitoKATP opening and ROS are unknown. Most of the knowledge about mitoKATP on mitochondrial energetics and function has been gained from studies on isolated mitochondria from heart and liver. A well-described consequence of mitoKATP opening is matrix swelling.8 Although some studies have shown that mitoKATP openers increase mitochondrial respiration, depolarize mitochondrial membrane potential partially, and decrease Ca uptake into mitochondria,9 others have not observed these effects. Studies from Kowaltowski et al8 have shown no significant changes of respiration, mitochondrial membrane potential, or Ca uptake by opening mitoKATP. Although such studies of isolated mitochondria have proven invaluable for assessing the direct effects of compounds on mitochondrial function, there are several considerations to keep in mind when comparing such results with more intact preparations. First, the choice of medium components is often distinctly unphysiological (eg, the presence of high sucrose, choice of substrates, Ca concentration, ionic composition, and ATP concentration), and critical interactions with other cellular components are missing, including the absence of a cytoskeleton (which apparently is critical for diazoxide-induced protection2) and the lack of coupling with energy-consuming sites (eg, myofibrils and ion transport ATPases). Furthermore, high-oxygen partial pressures (.159 mm Hg) are typically present that are orders of magnitude greater than the in vivo mitochondrial milieu under normoxic conditions ('2 to 3 mm Hg).10 Such a hyperoxic condition will undoubtedly increase oxidative stress. An additional problem is introduced when comparing drug concentrations used in isolated mitochondria with intact cells or hearts where the intracellular concentration of the drug is unknown. Furthermore, diazoxide has been shown to affect mitochondrial membrane potential and energy metabolism nonspecifically in pancreatic b-cells.11 However, the observations that the nonspecific effect was not blocked by glybenclamide11 whereas the protective effect of diazoxide was eliminated with glybenclamide and 5-HD5,7,12,13 and 5-HD also abolished the mitochondrial oxidative effect of diazoxide13 strongly argue against nonspecific effects of diazoxide in the observed cardioprotection and mitochondrial oxidation. Intact isolated myocyte studies are one step closer to the intact system but are still subject to some of the limitations The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From the Maryland Research Laboratories (Y.L.), Otsuka Maryland Research Institute, Rockville, Md, and Institute of Molecular Cardiobiology (B.O’R.), Johns Hopkins University, Baltimore, Md. Correspondence to Yongge Liu, PhD, Maryland Research Laboratories, Otsuka Maryland Research Institute, 9900 Medical Center Dr, Rockville, MD 20850. E-mail [email protected] (Circ Res. 2001;88:750-752.) © 2001 American Heart Association, Inc.