Cytoprotective Role of Ca2+- Activated K+ Channels in the Cardiac Inner Mitochondrial Membrane

See allHide authors and affiliations

Science  01 Nov 2002:
Vol. 298, Issue 5595, pp. 1029-1033
DOI: 10.1126/science.1074360


Ion channels on the mitochondrial inner membrane influence cell function in specific ways that can be detrimental or beneficial to cell survival. At least one type of potassium (K+) channel, the mitochondrial adenosine triphosphate–sensitive K+ channel (mitoKATP), is an important effector of protection against necrotic and apoptotic cell injury after ischemia. Here another channel with properties similar to the surface membrane calcium-activated K+ channel was found on the mitochondrial inner membrane (mitoKCa) of guinea pig ventricular cells. MitoKCa significantly contributed to mitochondrial K+ uptake of the myocyte, and an opener of mitoKCa protected hearts against infarction.

Sustained adenosine triphosphate (ATP) production by mitochondria requires maintenance of a large electrochemical gradient for protons across the mitochondrial inner membrane. This proton motive force is established by active proton pumping by the electron transport chain, producing both a pH gradient (ΔpH) and a mitochondrial transmembrane potential (ΔΨm). Because ΔΨm can be depolarized by energy-dissipating ion flux, the mitochondrial inner membrane was earlier assumed to have a low resting permeability to cations (1). However, it is well established that both divalent (2) and monovalent cation transport pathways (uniporters) are present on the inner membrane and that K+ conductance can be substantial in energized mitochondria (3,4).

A growing body of evidence indicates that mitochondrial ATP-sensitive K+ channels (mitoKATP) are important determinants of resistance to ischemic damage (5, 6) and apoptosis (7) and may be clinically recruitable to prevent or mitigate cardiac or neuronal ischemic injury (8). To determine whether other K+ channel subtypes are also present on the cardiac mitochondrial inner membrane, here we use direct single channel patch-clamp recordings of cardiac mitoplasts and mitochondrial K+ flux measurements to identify mitochondrial Ca2+-activated K+channels (mitoKCa) as a component of the mitochondrial background K+ conductance, and we test whether mitoKCa confers protection against infarction in the intact heart.

Mitoplasts prepared from isolated cardiac myocytes were patch-clamped (9) to identify the major single channel conductances of the inner membrane. In K+ solutions (150 mM K+) containing 512 nM Ca2+, single channel currents with a full unitary conductance of 307 ± 4.6 pS (n = 4 of 17 single channel patches) were often observed, with openings frequently interrupted by transitions to subconductance states ranging from 24 to 161 pS (Fig. 1, A to C). When pipettes were backfilled with the K+ channel toxin charybdotoxin (ChTx; 200 nM) to permit slow diffusion of the toxin into the pipette tip, channel activity disappeared within 30 min, indicative of the probable presence of KCa channels (Fig. 1D). In some patches in 512 nM Ca2+, and particularly at higher bath Ca2+ concentrations, channel activity was too great to identify individual channel openings; in these cases, ensemble average patch currents were analyzed and shown to be reversibly increased by raising Ca2+ in the medium (Fig. 1E). This activation by Ca2+ was eliminated when ChTx was present in the pipette (Fig. 1E). The ChTx-sensitive channels were unaffected by applying the mitoKATP inhibitor 5-hydroxydecanoate (500 μM), excluding a contribution from mitoKATP, which was not expected to be activated under the experimental conditions (Fig. 1D).

Figure 1

ChTx-sensitive channels in cardiac mitoplasts. (A) Single channel recording of a 300-pS channel within minutes of making the gigaohm seal (control) and 30 min later (200 nM ChTx) after toxin diffusion into the tip. (B) Amplitude histograms of the full and subconductance openings for the patch shown in (A). (C) Mean conductance levels for 17 single channel patches. Transitions between levels were commonly observed within the same recordings. (D) Inhibition of mitoplast ensemble average currents (at −60 mV) by ChTx. Paired currents were normalized to initial levels recorded within the first 5 min of the experiment. Currents were reported without correction for leak current across the seal, thus underestimating the fraction of total current attributable to ChTx-sensitive channels. Filled square denotes currents 20 to 30 min after the addition of 5-hydroxydecanoate (5-HD; normalized to pre-5HD control). (E) Effect of increasing bath free [Ca2+] from 512 nM to 40 μM for patches with or without 200 nM ChTx. Pairedt test: *P < 0.05, †P < 0.01, versus control.

To determine whether the KCa activity observed in single-channel recordings contributed significantly to the total K+ influx of intact mitochondria, we performed K+ concentration ([K+]) jump experiments in isolated cardiac myocytes loaded with the K+-selective fluorescent indicator PBFI (9), with the loading protocol optimized to achieve preferential labeling of the mitochondrial matrix (Fig. 2A). Although KCachannels are not believed to be present on the surface membranes of ventricular myocytes, we took the further precaution of permeabilizing the surface membrane with saponin to eliminate potential complications associated with K+ flux across the sarcolemma or signal contamination from cytosolic PBFI during wide-field fluorescence imaging. Using a rapid switching device capable of changing solutions in <2 s (Fig. 2B), we increased bath [K+] from 0 to 5 mM, and we determined net K+ flux into the matrix as the change in the fluorescence excitation ratio of PBFI (340/380 nm). The response to K+ was reversible and reproducible (Fig. 2C) and was accelerated, as expected, by the K+ ionophore valinomycin (3). In paired experiments, 100 nM ChTx slowed the time constant of mitochondrial K+ uptake by 8- to 10-fold (Fig. 2, D and E). The rate of K+ influx was also inhibited by the K+ channel blockers Ba2+ and quinine (Fig. 2E), known inhibitors of surface membrane KCachannels (10, 11) that also block K+ influx into isolated mitochondria (12, 13).

Figure 2

Effects of KCa inhibition on mitochondrial K+ uptake. (A) Pattern of PBFI fluorescence indicated preferential mitochondrial loading of the dye, which was retained upon saponin permeabilization. (B) Rapid exchange of the bathing medium confirmed by superfusion and washout of 100 μM carboxyfluorescein (CF). (C) Concentration jumps from 0 to 5 mM K+ accompanied by a reproducible increase in mitochondrial matrix PBFI excitation ratio (F340/F380 nm), permitting paired experiments within a field containing five to eight myocytes. (D) Inhibition of K+ uptake in the presence of 100 nM ChTx. ChTx slowed K+ uptake by ∼20-fold in the experiment shown. (E) Summary of the effects of KCa channel inhibitors on the time constant of mitochondrial K+ uptake. Inhibition of K+uptake depended on ChTx concentration and was also affected by Ba2+, iberiotoxin, and quinine. (F) The BKCa channel opener NS-1619 accelerated the rate of K+ uptake by more than twofold. Paired ttest: *P < 0.05, †P, < 0.01, ‡P < 0.001 versus control.

As ChTx has been shown to block several KCa subtypes and may also inhibit voltage-gated K+ channels of the Kv1.3 type, we also tested the efficacy of iberiotoxin, which is selective for large conductance KCa channels of the BK subtype (14). Similar to ChTx, iberiotoxin (100 nM), slowed the time constant of K+ influx by about 146%, indicating a contribution from large conductance KCa channels (Fig. 2E). We further confirmed the involvement of the large conductance KCa channel by using the KCa opener NS-1619. This compound opens the BKCa subtype of surface membrane KCa without affecting small- or intermediate-conductance KCa family members (15). Mitochondrial K+ uptake was accelerated about twofold by NS-1619 (Fig. 2F).

To determine whether KCa channel proteins were present in mitochondrial membranes, we performed immunoblot analysis on mitochondria from isolated myocytes (9). Immunostaining of intact myocyte proteins with an antibody against the surface membrane BKCa channel showed at least two major bands of about 55 and 220 kD (Fig. 3A). In contrast, a single protein band of about 55 kD, similar to the predicted size of the α subunit of KCa, was evident for an equal amount of mitochondrial protein.

Figure 3

(A) Immunoblot of mitochondria isolated from cardiomyocytes revealed a major protein band of about 55 kD that specifically bound an antibody directed against the COOH terminus of the BKCa channel (anti-BKCachannel). Larger molecular mass bands were observed in intact cardiomyocytes, possibly indicative of multimeric forms of the protein. (B) Displacement of the anti-BKCa immunoreactive band with antigenic peptide. Arrows denote positions of 55-kD (cardiac) and 80-kD (liver) bands. An increase in background staining was evident when antigen was present. (C) One-dimensional PAGE of liver mitochondrial inner membrane (10 μg) stained with Gel Code blue (left) or antibody to BKCa (right). (D) Enlarged region (pH 5 to 7.5 zone) of a two-dimensional polyacrylamide (pH 3 to 10;10% SDS-PAGE) silver-stained gel (lower) and its corresponding immunoblot (upper), performed sequentially, first with the antibody to BKCa and then with antibody to ATP synthase β chain. Spots identified as BKCa are circled and the ATP synthase β chain is indicated by arrows.

We used a highly purified liver mitochondrial inner membrane preparation to more rigorously confirm the mitochondrial localization of the BKCa immunoreactive protein. On a one-dimensional gel, BKCa antibody staining showed a prominent band at about 80 kD and two fainter bands in the 50- to 75-kD range (Fig. 3C) displaceable by control antigen (Fig. 3B). Further separation of the mitochondrial proteins by two-dimensional polyacrylamide gel electrophoresis (PAGE) (pH 3 to 10; 10% SDS) revealed a plethora of mitochondrial inner membrane proteins by silver staining (Fig. 3D). In the range of pH 5 to 7.5, 40 to 85 kD, the most prominent spot was the mitochondrial ATP synthase β chain (∼56 kD), which we used as a landmark for the subsequent immunoblot analysis (Fig. 3D). The antibody to BKCa specifically labeled a protein spot of ∼80 kD that was clearly separated from the ATP synthase by virtue of its charge differential in the isoelectric focusing dimension (pH 5 to 7.5 gradient).

The opening of the mitoKATP channel plays an important role in protecting hearts against ischemic damage; therefore, we reasoned that activating KCa channels might similarly protect hearts against infarction. We subjected perfused hearts to global ischemia and reperfusion after pretreatment with NS-1619 in the presence or absence of the KCa antagonist paxilline (9, 15). Left ventricular developed pressure (LVP) and heart rate were unchanged during the 5- to 10-min exposure to the KCa channel compounds (Table 1). As expected from the cross-reactivity of the KCa opener with the vascular surface membrane isoform of the channel, NS-1619 initially increased coronary flow by 24% at 10 μM and 66% at 30 μM during exposure. This increase in flow was blocked by coapplication with paxilline, which had no effect on flow in the absence of the opener. We applied global ischemia to eliminate coronary flow as a factor during the ischemic phase [see comments on selectivity in (9)]. No significant difference between groups in heart rate or flow was evident after reperfusion (Table 1). LVP was better preserved in the NS-1619–treated groups; however, this difference did not reach statistical significance. A 5-min preischemic exposure to 30 μM NS-1619 approximately halved the extent of myocardial infarction, a level of protection similar to that reported for the mitoKATP channel opener diazoxide in the same infarction model (16) (Fig. 4). For 10 μM NS-1619, no protection was observed for a 5-min pretreatment, but extending the exposure time to 10 min resulted in protection in four of six hearts (Fig. 4). Paxilline alone had no effect on infarct size, but it completely blocked the protection afforded by 30 μM NS-1619.

Figure 4

Effect on myocardial infarct size. Open circles, infarct size of individual hearts; filled circles, mean and standard errors of the group. Ctrl, untreated controls with 30-min global ischemia and 2-hour reperfusion; NS (10 μM), hearts treated with 10 μM NS-1619 for 10 min before ischemia; NS (30 μM), hearts treated with 30 μM NS-1619 for 5 min before ischemia; PX, hearts treated for 5 min with 1 μM paxilline before ischemia; PX+NS (30 μM), hearts treated with a combination of 1 μM paxilline and 30 μM NS-1619 for 5 min before ischemia. *P < 0.05.

Table 1

Hemodynamic data. Cardiac hemodynamics and coronary flow during control perfusion, NS-1619 application, and post-reperfusion. Details are as described in (9). Values listed under drug and reperfusion were obtained at the end of drug perfusion and at the end of the 2-hour reperfusion, respectively.

View this table:

The results indicate that an isoform of the large conductance KCa exists on the mitochondrial inner membrane and constitutes a large fraction of K+ uniport activity. MitoKCa-mediated current was detectable at cytosolic Ca2+ concentrations in the range of resting Ca2+ in the myocyte (∼200 nM; see Fig. 2) and enhanced at high cytosolic Ca2+ concentrations (Fig. 1E). Because the ChTx-sensitive current in mitoplast-attached patches increased when Ca2+ outside the pipette was raised, the relevant regulatory site for Ca2+ on the channel is likely to face the mitochondrial matrix. MitoKCa would then be activated as matrix Ca2+ rises in response to an increase in the average cytosolic Ca2+ load, such as occurs during an increase in cardiac work or ischemia. Thus, mitoKCa may play an important role both in modulating bioenergetics under physiological conditions and during conditions of Ca2+overload.

The function of mitoKCa may be to improve the efficiency of mitochondrial energy production. K+ is required for optimal functioning of oxidative phosphorylation (17) and may also modulate other mitochondrial functions, such as reactive oxygen species production. The mitochondrial K+ cycle, involving electrophoretic K+ uptake and electroneutral K+/H+ exchange, is important for mitochondrial volume regulation (4); this too can influence substrate oxidation (18). Analogously, the activation of mitoKATP improves ATP production (19), dampens mitochondrial Ca2+ accumulation during ischemia (20,21), and alters the rate of mitochondrial reactive oxygen species production (22, 23), the latter leading to the activation of intracellular signaling pathways (24). The present finding, that activating a completely different class of mitochondrial K+ channel confers a similar degree of protection, independently confirms that mitochondrial K+ influx is an important factor in mitigating injury, a conclusion that has been difficult to prove unequivocally for mitoKATP with diazoxide because of the potentially confounding nonspecific effects of the drug at high concentrations (25).

Mitochondrial K+ uptake is increased by energization (3, 26) and Ca2+ (13) in isolated mitochondria, but the molecular identity of the uniporter is unknown. Interestingly, the size of the mitoKCa channel monomer is similar to other putative mitochondrial K+channels. Several proteins in the same molecular size range (50 to 60 kD) were purified from mitochondria with a quinine affinity column and confer K+ channel activity upon reconstitution (27,28): similarly, a 54-kD protein was tentatively identified as a component of mitoKATP (29). The sensitivity of such reconstituted channels to K+ channel toxins and their potential immunoreactivity with antibodies to BKCa remain to be investigated. Excluding mitoKATP, which would not be active under energized conditions, the present findings are the only ones known to link mitochondrial K+ flux to a specific mitochondrial K+ channel. KCa channels have been reported in glioma cell mitoplasts (30), but their contribution to K+ flux and their physiological role were not explored.

In summary, the present findings identify mitoKCa in cardiac mitochondria, demonstrate that it contributes to mitochondrial K+ uniport conductance, and assign a role for mitoKCa in protection against ischemic injury.

Supporting Online Material

Materials and Methods

  • * To whom correspondence should be addressed. E-mail: bor{at}


Stay Connected to Science

Navigate This Article