Advertisement
Commentary Free access | 10.1172/JCI16122
Institute of Molecular Cardiobiology, Johns Hopkins University, Baltimore, Maryland 21205, USA.
Phone: (410) 955-2776; Fax: (410) 955-7953; E-mail: marban@jhmi.edu.
Find articles by Marbán, E. in: JCI | PubMed | Google Scholar
Published July 15, 2002 - More info
The golden age of single-channel electrophysiology was punctuated by the discovery of a curious class of potassium-selective channels in the surface membranes of heart cells (1, 2). These channels, silent under normal metabolic conditions, become robustly active when submembrane ATP is depleted. The opening of such KATP channels was soon recognized to underlie the loss of cellular excitability during metabolic stress (3, 4) and (later, in an in vivo correlate) to mediate the so-called ST segment elevation, an electrocardiographic change diagnostic of transmural injury during prolonged coronary ischemia (5). While these channels were discovered in heart cells, they are broadly-distributed throughout the body; their most obvious physiological role is in the pancreas, where they transduce changes in glucose concentration into alterations of β cell excitation and insulin secretion (6). Ironically, the role of KATP channels in the heart, where they were first discovered, remains mysterious. In the present issue and in recent work published elsewhere, new surprises emerge regarding the identities and the roles of KATP channels in the vasculature.
The general molecular structure of KATP channels was clarified by the pioneering work of the Bryan, Aguilar-Bryan (7) and Seino laboratories (8). KATP channels turn out to be complex hetero-octamers of four subunits encoded by sulfonylurea receptor (Sur) genes, surrounding a central pore made up of four Kir6-encoded K channel subunits (9). Different tissues express different permutations of the three possible Surs (Sur1, and the Sur2 splice variants, Sur2A and Sur2B) and two pore-forming subunits (Kir6.1 and Kir6.2). The conventional surface KATP channels in the heart are formed by Sur2 and Kir6.2, whereas those in pancreatic β cells consist of Sur1 and Kir6.2. KATP channels have a rich pharmacology (10), including a number of venerable vasodilators (e.g., the KATP channel agonists diazoxide and minoxidil) and anti-diabetic agents (notably the KATP channel blocker glybenclamide), which were in clinical use long before their targets were recognized.
KATP channels have received much attention recently due to two related puzzles. First, a spirited controversy rages regarding the role of KATP channels in cardioprotection against ischemia. Drugs that open KATP channels clearly confer resistance to ischemia, while blockers prevent the development of ischemic preconditioning (11). The controversy lies in the mechanism of the cardioprotection and in the precise identity of the channels responsible. Much evidence points to mitochondria as the site of action (11), but the mitochondrial site of action came into question when Kir6.2 knockout mice were found to lack surface, but not mitochondrial KATP channels, and to be resistant to ischemic preconditioning (12).
The second puzzle surrounded the physiological role of the Kir6.1 subunit, which is found in many tissues at the mRNA and protein levels. Little was known about Kir6.1 until it was recently knocked-out by Miki and coworkers (13). The loss of Kir6.1 produced a dramatic, discrete phenotype: The mice developed spontaneous bouts of coronary vasospasm and ST elevation, which often proved fatal due to conduction block. These mice reproduced key features of vasospastic or Prinzmetal angina, a syndrome of sudden coronary vasoconstriction without underlying atherosclerosis. Vascular smooth muscle cells from Kir6.1 knockout mice were found to have no detectable KATP currents and to lack all vasodilatory responses to KATP channel agonists.
In the present issue, Chutkow et al. (14) further clarify the identity of vascular smooth muscle KATP channels by showing that Sur2 knockout mice also develop a Prinzmetal phenotype and loss of KATP channels in vascular muscle, while manifesting the expected lack of KATP currents in cardiac cells. The inescapable conclusion of these two studies (13, 14) is that the KATP channels of vascular smooth muscle consist of one or another splice variant of Sur2 along with Kir6.1. That particular combination of subunits thus represents a promising target for the development of novel anti-anginal compounds. The new results give reason to wonder whether nicorandil, a remarkably effective anti-anginal drug commonly used in Japan and Europe (but sadly absent from the US pharmacopeia), might preferentially activate Sur2 and Kir6.1 channels.
The concordant findings between these two recent papers raise a number of critical questions and leave others unanswered. Some of these relate to the relationships among known K+ channels. For instance, are the vascular KATP channels that disappear with knockout of Sur2 or Kir6.1 truly conventional KATP channels, or might they be identical to the previously described nucleotide diphosphate–sensitive (KNDP) channels (15)? More extensive characterization of the pharmacological and single-channel properties of the wild-type channels will be required to sort out this question. Likewise, the identity of the mitochondrial KATP channels that have been implicated in cardiac preconditioning remains in doubt. Miki et al. argue that mitochondrial KATP channels are not encoded by Kir6.1, as evidenced by a preserved mitochondrial flavoprotein response to the cardioprotective KATP channel agonist diazoxide (13). Since flavoprotein responses also remain intact in Kir6.2 knockout mice (12), it may be that neither of these subunits contributes to the channel activity found in mitochondria. Indeed, a role for Sur2, the known partner for Kir6.2 in cardiac tissue, seems unlikely, based on the distinctive pharmacological profile of mitochondrial KATP channels (10), but direct tests in the present knockouts would provide additional welcome information.
It is also unclear, given the wide distribution of KATP channels throughout the vasculature, why Kir6.1 and Sur2 knockout mice exhibit a primary coronary phenotype. Coronary vessels may be unusually prone to episodic vasoconstriction, but it remains to be seen if a similar pattern of events might occur in other organs — although perhaps with relatively benign effects. However, perhaps the most urgent clinical question to arise from this work concerns how faithfully the two models reproduce human Prinzmetal angina. The sex distribution of affected mice in the present report (14) does not mirror the female preponderance in humans. Moreover, conduction disturbances and sudden death are not typical features of the clinical syndrome. Do these terminal events arise simply from the severity of ischemia, or does the knockout of KATP channels per se confer susceptibility to unstable cardiac impulse transmission? Additional studies with pharmacological vasoconstrictors in wild-type mice, along the lines of those reported by Miki et al. (13), would presumably help in sorting out this important question.