Date Published: July 29, 2014
Publisher: Public Library of Science
Author(s): Ewa Soltysinska, Bo Hjorth Bentzen, Maria Barthmes, Helle Hattel, A. Brianne Thrush, Mary-Ellen Harper, Klaus Qvortrup, Filip J. Larsen, Tomas A. Schiffer, Jose Losa-Reyna, Julia Straubinger, Angelina Kniess, Morten Bækgaard Thomsen, Andrea Brüggemann, Stefanie Fenske, Martin Biel, Peter Ruth, Christian Wahl-Schott, Robert Christopher Boushel, Søren-Peter Olesen, Robert Lukowski, Hossein Ardehali.
Mitochondrial potassium channels have been implicated in myocardial protection mediated through pre-/postconditioning. Compounds that open the Ca2+- and voltage-activated potassium channel of big-conductance (BK) have a pre-conditioning-like effect on survival of cardiomyocytes after ischemia/reperfusion injury. Recently, mitochondrial BK channels (mitoBKs) in cardiomyocytes were implicated as infarct-limiting factors that derive directly from the KCNMA1 gene encoding for canonical BKs usually present at the plasma membrane of cells. However, some studies challenged these cardio-protective roles of mitoBKs. Herein, we present electrophysiological evidence for paxilline- and NS11021-sensitive BK-mediated currents of 190 pS conductance in mitoplasts from wild-type but not BK−/− cardiomyocytes. Transmission electron microscopy of BK−/− ventricular muscles fibres showed normal ultra-structures and matrix dimension, but oxidative phosphorylation capacities at normoxia and upon re-oxygenation after anoxia were significantly attenuated in BK−/− permeabilized cardiomyocytes. In the absence of BK, post-anoxic reactive oxygen species (ROS) production from cardiomyocyte mitochondria was elevated indicating that mitoBK fine-tune the oxidative state at hypoxia and re-oxygenation. Because ROS and the capacity of the myocardium for oxidative metabolism are important determinants of cellular survival, we tested BK−/− hearts for their response in an ex-vivo model of ischemia/reperfusion (I/R) injury. Infarct areas, coronary flow and heart rates were not different between wild-type and BK−/− hearts upon I/R injury in the absence of ischemic pre-conditioning (IP), but differed upon IP. While the area of infarction comprised 28±3% of the area at risk in wild-type, it was increased to 58±5% in BK−/− hearts suggesting that BK mediates the beneficial effects of IP. These findings suggest that cardiac BK channels are important for proper oxidative energy supply of cardiomyocytes at normoxia and upon re-oxygenation after prolonged anoxia and that IP might indeed favor survival of the myocardium upon I/R injury in a BK-dependent mode stemming from both mitochondrial post-anoxic ROS modulation and non-mitochondrial localizations.
Ischemia and post-ischemic reperfusion induce cardiac damage . Brief episodes of ischemia prior to or after the prolonged ischemic insult have been shown to be very effective in rendering the heart less susceptible to damage , . The infarct-reducing effects via ischemic pre-conditioning (IP) are mediated by multiple factors, which, at least in part, signal through potassium channels present at the inner mitochondrial membrane (IMM) of the cardiomyocyte (CM) . Opening of mitochondrial KATP (mitoKATP) channels has been shown to modulate mitochondrial bioenergetics and to inhibit formation of the mitochondrial permeability transition pore (mPTP), which is a major determinant of irreversible injury during reperfusion when oxygen is reintroduced and the IMM potential is regenerated –. However, their cardio-protective roles in the mitochondrion have also been questioned . In addition to mitoKATP, intracellular Ca2+-activated potassium channels of big-conductance (BK)  first recognized in a human glioma cell line  have been identified in mouse CMs  at the IMM (mitoBK) –, but not in the sarcolemma , . Experiments using isolated ventricular CMs from rat and guinea pig hearts suggest that opening of mitoBK prevents mitochondrial matrix Ca2+ overload and attenuates reactive oxygen species (ROS) production, pathological hallmarks for mPTP formation and cell death, upon reperfusion of the heart –. MitoBK activation may thus confer cardio-protection in a manner similar to but independent of mitoKATP, but to this end its function at physiological conditions is unclear. Physiological and pharmacological properties of mitoBK channels purified from cardiac mitochondria indicate that their properties are comparable to plasma membrane BK channels , –. Importantly, BK channel openers (NS1619 and NS11021) applied at the onset of reperfusion are cardio-protective independent of changes in KATP activity or hemodynamic alterations –. Cardio-protection was sensitive to the BK blocker paxilline and could be observed for 24 hours upon pretreatment . Additionally, accessory β1-subunits of BK that modulate kinetics and Ca2+-sensitivity of the channel were identified in mitochondria of CMs and their knock-down abolished infarct-limiting effects of the phosphodiesterase-5 inhibitor sildenafil . A combination of yeast two-hybrid and immune-cytochemical assays provided further evidence for an IMM complex enclosing BK β1-subunits and cytochrome c oxidase I indicating direct recruitment of multimeric BK channel proteins to the oxidative phosphorylation machinery . A number of additional studies have reported the involvement of BK in the signaling pathway of agents leading to cardio-protection. This list includes among others adenosine , κ-opoid receptor agonist , β-estradiol , and desflurane , . These agents protected the heart from I/R injuries when applied prior to ischemia, and the protection was attenuated by co-administration of toxins, such as paxilline, inhibiting BK, thereby strongly supporting the notion that pro-survival pathways promote BK activity. However, these conclusions have been recently challenged since the widely used BK opener NS1619 might have additional effects unrelated to mitoBK channels –. Furthermore, IP or anesthetic preconditioning (APC) protocols efficiently protected BK-negative hearts from I/R injury , although, beneficial effects of APC were sensitive to paxilline in the absence of BK. Finally, it has been suggested that protection via BK and NS1619 at I/R is a function of intrinsic cardiac neurons , but it remained largely elusive how the protective signals would be transmitted to the CM.
We first confirmed that the BK channel is detectable in mitochondrial protein fractions isolated from BK wild-type (BK+/+) hearts and CMs, whereas mitochondrial fractions derived from the respective BK−/− cells/organs remained BK negative (Fig. 1A). As compared to control tissues (e.g. cerebellum (not shown)) BK protein expression levels in heart and CM mitochondria were very low, however, the molecular weight of the immuno-positive bands indicate that mitoBK channels have a similar molecular identity of approx. 120 kDa as canonical BKs usually present in the plasma membrane of cells. To examine the electrophysiological and pharmacological properties of the mitoBK we studied mitoplasts, generated by osmotic swelling. The quality of these purifications i.e. the strongly availability of contaminant mitochondrial outer membranes was confirmed by high magnification microscopy (Figure S1A in File S1). Under the given experimental conditions (s. Methods) we observed ionic conductances of 4 magnitudes in BK+/+ mitoplasts with a different relative frequency during 9 minutes of recording: ∼60 pS (35%), ∼120 pS (80%), ∼190 pS (55%) and ∼370 pS (15%) (Fig. 1B). Importantly, only the ∼190 pS conductance was absent from BK−/− mitoplasts, whereas all other magnitudes were still present. Therefore, subsequent analyses focused on the ∼190 pS conductance. We separated the patch-clamp recordings into 3 identical intervals of 3 min duration. After the first control phase resembling 3 min of baseline measurement, NS11021 (10 µmol/l) was applied with the external solution, which was exchanged by paxilline (100 nmol/l in external solution) 3 min later (Fig. 1C). NS11021 increased the relative frequency of the ∼190 pS conductance by 12%, whereas paxilline superfusion had the opposite effect reducing the relative frequency of the respective conductance to 6% in BK+/+ mitoplasts. In BK negative mitochondria neither NS11021 nor paxilline had an effect strongly supporting the notion that mitoBK channels are the product of the KCNMA1 gene targeted by our global BK channel knockout strategy . For a better characterization of the mitoBK we examined phases of single channel activity. These in-depth analyses of representative traces revealed a slope conductance of 190 pS and an open probability of 0.79 at +80 mV for the murine mitoBK (Fig. 1D and E). Importantly, these data complement very recent findings by Singh and co-workers , and together provide strong evidence for mitoBK channels encoded by the murine KCNMA1 gene in cardiac myocytes.