|Home | About | Journals | Submit | Contact Us | Français|
Mitochondrial volume regulation depends on K+ movement across the inner membrane and a mitochondrial Ca2+-dependent K+ channel (mitoKCa) reportedly contributes to mitochondrial K+ uniporter activity. Here we utilize a novel KCa channel activator, NS11021, to examine the role of mitoKCa in regulating mitochondrial function by measuring K+ flux, membrane potential (ΔΨm), light scattering, and respiration in guinea-pig heart mitochondria. K+ uptake and the influence of anions were assessed in mitochondria loaded with the K+ sensor PBFI by adding either the chloride (KCl), acetate (KAc) or phosphate (KH2PO4) salts of K+ to energized mitochondria in a sucrose-based medium. K+ fluxes saturated at ~10mM for each salt, attaining maximal rates of 172±17, 54±2.4 and 33±3.8 nmol K+/min/mg in KCl, KAc or KH2PO4, respectively. NS11021 (50nM) increased the maximal K+ uptake rate by 2.5- fold in the presence of KH2PO4 or KAc, and increased mitochondrial volume, with little effect on ΔΨm. In KCl, NS11021 increased K+ uptake by only 30% and did not increase volume. The effects of NS11021 on K+ uptake were inhibited by the KCa toxins charybdotoxin (200nM) or paxilline (1μM). 50nM NS11021 increased the mitochondrial respiratory control ratio (RCR) in KH2PO4, but not in KCl; however, above 1μM, NS11021 decreased RCR and depolarized ΔΨm. A control compound lacking KCa activator properties did not increase K+ uptake or volume, but had similar nonspecific (toxin insensitive) effects at high concentrations. The results indicate that activating K+ flux through mitoKCa mediates a beneficial effect on energetics that depends on mitochondrial swelling with maintained ΔΨm.
Increased mitochondrial volume and an improvement in oxidative phosphorylation have been implicated in the mechanism of ischemic or pharmacological preconditioning [1, 2], but the links between altered K+ flux and the functional effects are not well understood [3, 4]. K+ is the principal monovalent cation in the cytoplasm and, owing to the large electrochemical driving force, is consequently an important ion mediating changes in volume when the K+ conductance of the inner membrane is altered. Because of the restricted matrix space, the mitochondria have been referred to as “perfect osmometers”, hence, the coordinated uptake of both cations and anions, along with water, mediates swelling . Early work stressed the existence of selective electrophoretic pathways for cation uptake and anion transport associated with mitochondrial swelling [5, 6], and Garlid and collaborators  have provided supporting evidence that the major volume regulation involves K+ cycling, consisting of electrophoretic K+ influx and electroneutral K+ efflux by the K+/H+ exchanger. Although electroneutral, K+ extrusion via K+/H+ antiporter requires the proton gradient .
The interaction between mitochondrial volume changes and altered energetics in the context of cardioprotection is unclear, and currently under debate [1, 2, 8]. At least three interdependent regulatory processes may be involved, including i) K+ uptake, ii) matrix volume changes, and iii) mitochondrial energetic alterations (involving respiration, ΔΨm, and the NADH produced by the TCA cycle). Enhanced mitochondrial K+ influx through ATP-dependent (mitoKATP)  or Ca2+-dependent (mitoKCa) K+ channels has been proposed as a potential mediator of protection, and the concomitant increase in mitochondrial volume could alter electron flow through the respiratory chain [10, 11]. In this regard, it has been proposed that the increase in matrix volume may enhance ATP/ADP exchange with the cytoplasm by bringing the inner and outer mitochondrial membranes closer together [1, 2, 8].
Increases in cytoplasmic Ca2+ during excitation-contraction coupling or hormonal stimulation result in an increase in mitochondrial matrix Ca2+ and this is an important factor regulating oxidative phosphorylation. Halestrap and collaborators [12, 13] reported that matrix Ca2+ inhibits a pyrophosphatase, resulting in a rise in matrix pyrophosphate and stimulation of K+ influx in liver or heart mitochondria. Alternatively, Brown and Brand  proposed a Ca2+-mediated increase of matrix volume operating through the well known activating effect of Ca2+ on tricarboxylic acid cycle dehydrogenases, in turn stimulating respiration and increasing ΔΨm, thus increasing the electrochemical driving force for K+ uniport activity [10, 14]. More recently, we proposed that elevated matrix Ca2+ could open large conductance Ca2+-activated K+ channels present in the inner membrane of cardiac mitochondria . Similarly, KCa type channels have been reported in mitochondria from brain [16–18] and cultured human glioma cells . However, the nonspecific effects of high concentrations of KCa activators (e.g. NS1619) on mitochondrial energetics [20, 21] have complicated interpretation of the role of mitoKCa in cardioprotection.
In the present work, we examine the effects of a novel, more potent, KCa channel opener NS11021  and its inactive form (NS13558) on mitochondrial K+ fluxes, matrix volume, and ΔΨm in energized mitochondria, and also evaluate their impact on mitochondrial respiration. We show that in the nanomolar concentration range, NS11021 accelerates the initial rate of K+ uptake and potentiates mitochondrial swelling in the presence of permeable anions, while in the μM range, it leads to energetic deterioration. The beneficial effects correlate with expansion of the matrix space and are sensitive to the KCa channel inhibitors charybdotoxin or paxilline, while the nonspecific effects are not. Moreover, the control compound does not recapitulate the effects of NS11021 on K+ uptake or volume, but does reproduce the toxin-insensitive nonspecific actions.
The protocol of mitochondrial isolation for small sample tissue (2–4 g of fresh tissue) described in Mela and Seitz (1979)  was followed with slight modifications. Briefly, a 300g weight guinea-pig was euthanized and the heart quickly excised. After harvesting the heart, the whole procedure was performed on ice. The heart was immediately immersed into ~10ml of isolation solution (IS) consisting of 75mM sucrose, 225mM mannitol, and 0.01mM EGTA neutralized with Trizma buffer at pH 7.4. After washing, the top of the heart corresponding to the aorta, the pulmonaries, and the atria were discarded, and immersed into fresh 20ml of IS and minced. After the tissue pieces settled, the entire supernatant was discarded and fresh IS (5 ml) was added, and the mixture was transferred to a hand homogenizer. Proteinase (0.8 mg, bacterial, type XXIV, Sigma, formerly called Nagarse) was added just before starting the homogenization procedure. The whole homogenization procedure took no longer than 14min in two steps of ~7min (each with 5ml addition of fresh IS). The homogenate was carefully transferred after each step to a polycarbonate centrifuge tube. After 5 min 480 g centrifugation to discard unbroken tissue and debris, the supernatant was centrifuged at 7700 g for 10 min to sediment the mitochondria. The mitochondrial pellet was washed twice with IS and the last one with suspension solution (IS without EGTA) at 7700 g for 5min each. An average of 10 mg mitochondrial protein/ml was obtained from one guinea-pig heart with this procedure.
Respiratory Control Ratios (RCR; ratio of state 3 over state 4 respiration with glutamate + malate as the substrate) of 10 to 20 were obtained using this method (see Fig. 1). The quality of the mitochondrial preparation was confirmed using electron microscopy. Electron micrographs showed that the Nagarse method yielded a more representative sample of mitochondria with less damage and fragmentation, as confirmed by isolating mitochondria simultaneously from two halves of the same heart in the presence or in the absence of Nagarse (data not shown).
Mitochondrial protein was determined using the bicinchonic acid method, BCA™ protein assay kit (Pierce, IL).
Mitochondrial respiration was assayed at 37°C in a closed chamber of 0.35 ml containing (in mM): 250 sucrose (or 137 KCl), 2 KH2PO4, 0.5 EGTA, 2.5 MgCl2, 20 HEPES at pH 7.1, and 100 to 200 μg of mitochondrial protein. The O2 concentration in the chamber was monitored by means of a fiber optic O2 sensor (Ocean Optics, Inc; probe tip diameter =1 mm) in which the fluorescence emission at 600nm of a ruthenium compound is quenched by O2. The fluorescence signal was calibrated using different mixtures of O2 and N2 and was linear in the range 1–60% O2. To calculate the ATP/O ratio, aliquots of 100 μM of ADP were added and the difference between the O2 signal recorded before and after the ADP addition was quantified upon return to state 4 (i.e., when all of the ADP was converted to ATP).
Mitochondrial swelling, ΔΨm, and NAD(P)H were monitored simultaneously with a spectrofluorometer (Photon Technology, Inc.) in an experimental solution containing (in mM): 250 sucrose (or 137 KCl), 0.5 EGTA, 2.5 MgCl2, 20 HEPES, pH 7.1. Mitochondrial swelling was measured as a decrease in the 90° light scattering signal using 520nm excitation [24–26]. ΔΨm was recorded using tetramethylrhodamine methyl ester (TMRM; 100nM) and applying the ratiometric method of Scaduto and Grotyohann . Briefly, this method uses two excitation wavelengths (λexc) at 546nm and 573nm while recording the fluorescence emission (λem) at 590nm. A calibration curve of ΔΨm versus the 573/546nm ratio was constructed to estimate ΔΨm in each experiment. PBFI fluorescence was monitored ratiometrically (340/380nm excitation at 495nm emission; see also next subsection) and NAD(P)H fluorescence was measured at λem=450nm with λexc=340nm.
Freshly isolated mitochondria were loaded with 20μM PBFI-AM for 20min at room temperature with occasional shaking. Excess dye was removed by centrifugation (1.5min × 14,000 g), and the mitochondrial pellet was washed once under the same centrifugation conditions. After resuspension in 225mM mannitol and 75mM sucrose, mitochondria were assayed for K+ uptake in the same medium described for the respiration measurements with the exception that 250mM sucrose was used instead of 137mM KCl. This allowed us to study the initial K+ uptake rate in a controlled fashion. In order to relate the increase in the 340/380nm ratio of PBFI fluorescence to K+ concentration, 1μM of PBFI salt was dissolved in the same assay medium, but in the absence of mitochondria. The increase in PBFI ratio was determined as a function of increasing concentrations of KCl, KAc, or KH2PO4. The conversion factor for each of the K+ salts was obtained from the slope of the plot of PBFI ratio vs K+ concentration.
We used the radioactive tracer method to measure mitochondrial volume [28, 29], in order to further quantify K+ fluxes. This method uses 3H-labeled water (3H2O) and 14C-labeled mannitol (14C mannitol). Mitochondrial volume was determined in four replicates for each K+ salt under the same conditions as the fluorometric experiments measuring K+ uptake. Using the kinetics of mitochondrial volume change as determined by 90° light scattering, we sampled mitochondria before and after addition of the K+ salt for the volume determination. Briefly, 1 μCi of 3H2O and 0.2 μCi of 14C Mannitol were added to 0.5 ml of mitochondrial suspension (2 to 3 mg of mitochondrial protein/ml), and incubated for 3 min. After the incubation, the whole volume was transferred to a microcentrifuge tube and centrifuged at 13,000 × g for 3 min. The whole volume of the supernatant was recovered and counted, and the pellet was dissolved with 0.1 ml of perchloric acid 20% (wt/wt) and also counted (Beckman LS 6000, Beckman Instruments Inc., Fullerton CA).
Several controls were performed in order to account for i) count spillover: mitochondria were incubated with one radioactive tracer at a time, centrifuged, and supernatant and pellet counted in the 14C and 3H channels in duplicate; ii) nonspecific binding of 14C mannitol: mitochondria were subjected to osmotic lysis, and the same centrifugation as well as counting protocol was followed; and iii) quenching of the 14C and 3H tracers introduced by perchloric acid: 0.01 μCi of each radioactive tracer was dissolved in 0.5ml of water; half of the volume was acidified with perchloric acid whereas the other half was counted directly.
The electron microscopy of freshly isolated mitochondria was performed at The Johns Hopkins University School of Medicine Microscope Facility using standard procedures. Briefly, immediately after isolation, an aliquot of mitochondrial suspension (0.6–0.8 mg of mitochondrial protein) was fixed in 2% glutaraldehyde, 2% formaldehyde with 3mM CaCl2 in 0.1M cacodylate buffer for 1 h at room temperature. After washing in 0.1M cacodylate buffer, the pellet was post-fixed in reduced osmium tetroxide for 1 h at 4°C. Pre-embed staining was done with 2% uranyl acetate in dH20. The pellet was then dehydrated in graded ethanol and embedded in Eponate 12 from Ted Pella. Thin sections were cut on a Reichert-Jung Ultracut E microtome and placed on 200 mesh copper grids. The grids were post-stained with uranyl acetate followed by lead citrate and viewed on a Hitachi 7600 TEM. Digital images were collected with an AMT Advantage HR side mount camera.
All representative records shown in this paper were confirmed in at least three independent experiments. Data were analyzed with the software GraphPad Prism (Ver. 3; San Diego, CA) or MicroCal Origin. The statistical significance of the differences between treatments was evaluated with ANOVA using Tukey’s multiple comparison test, and the results presented as the mean ± SEM (95% confidence interval)
NS11021 (1-(3,5-bis-trifluoromethyl-phenyl)-3-(4-bromo-2-(1H-tetrazol-5-yl)-phenyl)-thiourea) and its inactive form NS13558 (a −CH3 added to the tetrazole moiety of NS11021) were a generous gift from NeuroSearch A/S (Denmark). Paxilline was obtained from Biomol International and charybdotoxin recombinant Escherichia coli from Calbiochem (La Jolla, CA). The acetoxymethyl (AM) ester form of the K+ sensitive benzofuran isophthalate (PBFI), and tetramethylrhodamine methyl ester (TMRM), were purchased from Invitrogen (Carlsbad, CA). All other reagents were purchased from Sigma.
Freshly isolated mitochondria from guinea-pig heart were resuspended in an isotonic sucrose-based assay medium, and energized with 5mM each of glutamate-Na+/malate-Na+ (G/M). The attainment of state 4 was assessed by simultaneous monitoring of ΔΨm, NADH, and swelling (90° light scattering) after addition of G/M. In isotonic sucrose medium (250mOsm), ΔΨm increased immediately by ~30–50mV in parallel with reduction of the NADH pool, and low-amplitude swelling (Fig. 1A). On ADP (1mM) addition, the mitochondria exhibit the state 4→ state 3 transition associated with ΔΨm depolarization (~30mV), NADH oxidation, and contraction. Similar changes occurred in KCl-based isotonic medium (data not shown). Independent measurements of mitochondrial respiration following an identical protocol showed an abrupt increase in oxygen consumption following ADP addition in the presence of G/M (Fig. 1B). These observations are in accord with published experimental evidence (e.g. [30, 31]).
K+ uptake was assessed with mitochondria in state 4 (i.e., after energization with 5mM G/M but in the absence of ADP) after the addition of different K+ salts. From first principles, due to the limitations imposed by charge balance, mitochondrial swelling can only occur when both cations and anions are permeable to the mitochondrial inner membrane [5, 32]. Thus, K+ uptake was assessed in the presence of the relatively impermeable Cl− [33, 34], the passively diffusing Ac− -, and the carrier-mediated H2PO4− .
We used the K+-sensitive ratiometric fluorescent dye PBFI to directly and quantitatively measure K+ fluxes in isolated mitochondria. Mitochondria were incubated in the presence of 5mM G/M in isotonic sucrose medium and different concentrations of each potassium salt were added to the cuvette. A reproducible increase in PBFI fluorescence (ratio 340/380) could be detected in mitochondria when subjected to additions of increasing concentrations of KCl, (Fig. 2A), KAc (Fig. 2D), and KH2PO4 (Fig. 2G). As shown in Figure 3A, K+ fluxes saturated above 10mM for each salt and attained maximal rates (in nmol K+/min/mg) of 172±17, 54±2.4 and 33±3.8 in KCl, KAc, and KH2PO4, respectively. The K0.5 was similar in all three cases (in mM): 7±0.9, 6.4±0.5, and 6.0±0.6, respectively.
The specificity of the response of the K+-sensitive probe was assessed by subjecting mitochondria to an addition of 10mM Na+ acetate (NaAc, Fig. 2D). Under conditions in which 10mM KAc gave a clear increase in PBFI fluorescence, no increase could be registered with 10mM NaAc (Fig. 2D), verifying the K+ selectivity of the PBFI. However, a decrease in the light scattering signal occurred after either 10mM KAc or NaAc, indicating that K+ and Na+ could both enter the mitochondria (Fig. 2E).
The light scattering response was clearly different between the K+ salts assayed. Mitochondria only contracted with KCl (Fig. 2B), whereas they swelled transiently with ≥ 5mM KAc (Fig. 2E), and showed sustained swelling with KH2PO4 (Fig. 2H). Under these conditions, mitochondrial volume was also assessed by the radioactive tracer method (see Materials and Methods) with the following results (expressed as μl/mg mitochondrial protein±SEM; n=4): Control (before K+ salt addition) = 1.75±1.8×10−4; KCl = 1.77±3.9×10−4; KAc = 2.15±7.2×10−4; KH2PO4 = 2.54±3.4×10−4. In all three cases, increasing K+ concentration only slightly depolarized ΔΨm (~5–10mV) (Fig. 2C, 2F, 2I).
We then analyzed the effect of NS11021, a novel activator of plasmalemma KCa channels, on mitochondrial K+ fluxes, ΔΨm, and volume, in the presence of the different K+ salts at near maximal K+ uptake rate (Fig. 3B). In the presence of KH2PO4 or KAc, 50nM NS11021 enhanced the rate of K+ uptake by 2.5-fold and increased mitochondrial volume, while a high ΔΨm was maintained (> 190mV) (data not shown). In contrast, in the presence of KCl, 50nM NS11021 increased K+ flux by only 30%, and no change in volume was observed, while ΔΨm was also maintained (> 190mV).
Importantly, the NS11021 activation of K+ uptake was charybdotoxin (ChTx)- and paxilline (Pax)- sensitive; both of which are well known KCa channel blockers . Charybdotoxin (200nM) blocked the increase in K+ flux (Fig. 4A) as well as the increase in matrix volume, at least partially (Fig. 4B) induced by an addition of 10mM or 20mM KH2PO4 in the presence of 50nM NS11021. Paxilline (1 μM) blocked the activation of K+ flux (Fig. 4C) and the increase in volume (Fig. 4D) induced by 50nM NS11021.
The impact of selective activation of KCa channels upon respiration was evaluated by determining the effect of NS11021 on the RCR and the P/O ratio. The RCR (=state3/state4 respiration) was measured under the same conditions that an increase in K+ flux and mitochondrial volume were observed (Figs. 3 and and4).4). 50nM NS11021 significantly increased the RCR by 70%, in the presence of 20mM KH2PO4 (Fig. 5A). In contrast, 1μM NS11021 significantly decreased the RCR. The P/O ratio in the presence of 50nM NS11021 did not differ from the control but decreased in the presence of 1μM NS11021 (mean±SEM: Control = 2.14±0.19; NS50= 2.05±0.11; NS1= 1.48±0.11; n=3).
In the presence of 10mM KCl, the RCR was also higher with 50nM NS11021, but this difference was not statistically significant; however, 1μM NS11021 still decreased RCR (Fig. 5C). Importantly, preincubation with 200nM ChTx (Fig. 5B) or 1μM paxilline (not shown) prevented the effect of 50nM, but not 1μM, NS11021 on RCR (with KH2PO4) confirming that the effects of 50nM NS11021 on K+ uptake, mitochondrial volume, and respiration were all mediated by mitoKCa channels.
Interestingly, the increase in RCR observed was due to a decrease in state 4 respiration rather than an increase in state 3, which was unchanged by 50nM NS11021 (Fig. 5A, insets). 1μM NS11021 significantly increased state 4 respiration without affecting state 3 compared to the control (Fig. 5A, insets), accounting for the decrease in RCR. The decrease in state 4 observed in the presence of 50nM NS11021 was abrogated by ChTx (inset, Fig. 5B). Furthermore, an acute decrease in the rate of state 4 respiration was registered upon addition of KH2PO4 in the case of mitochondria preincubated with 50nM NS11021 (Fig. 1C), in agreement with the increased RCR described above (Fig. 5A and 5B). ΔΨm was not significantly depolarized by 50nM NS11021 (Fig. 6A).
ChTx did not prevent the decrease in RCR induced by high concentrations of NS11021, in the presence of either KH2PO4 or KCl (Fig. 5B and 5D). There was a larger decrease in RCR with 1μM NS11021 in the presence of ChTx, presumably because any beneficial effects of KCa channel activation were blocked, thus enhancing the nonspecific negative effect. Indeed, a decrease of ΔΨm from −193 mV (control) to −160 mV was observed when mitochondria were preincubated with 1μM NS11021, and further depolarization occurred upon addition of the K+ salts (Fig. 6A). An even larger collapse of ΔΨm was observed after preincubating with 2μM (Fig. 6A) or 5μM NS11021 (not shown).
Control experiments were carried out using NS13558, which bears an additional methyl group on the tetrazole moiety of the parent compound. This modification renders the compound inactive with respect to the activation of surface membrane KCa channels (data not shown). Importantly, 50 nM NS13558, in contrast with NS11021, did not increase K+ fluxes (Fig. 7A), nor did it improve RCR (Fig. 7B) with any of the K+ salts. NS13558 also did not increase volume or change ΔΨm at the 50nM concentration (see Fig. S2, Supplementary Material). However, μM amounts of NS13558 produced nonspecific effects similar to those of NS11021, such as marked ΔΨm depolarization and decreased RCR (Fig. 7B).
The effects of NS11021 were then compared with those of the artificial K+ ionophore valinomycin. Preincubation with valinomycin in the picomolar (pM) range, followed by addition of KH2PO4, also induced a moderate (~10mV up to 100pM Val) or more drastic decrease of ΔΨm (~30 to 40mV at > 500pM Val; Fig. 6B), enhanced K+ influx (Fig. 6C) and high-amplitude swelling (Fig. 6D). ΔΨm was not depolarized by Val in the absence of K+ salts (Fig. 6B), but this was not true for NS11021 concentrations >1μM (Fig. 6A), which depolarized ΔΨm even in the absence of K+. The RCR was not affected by Val <100pM; RCR decreased thereafter due to a significant increase in state 4 and a decrease in state 3 respiration after addition of the K+ salt (see Fig. S1, Supplementary Material). These data indicate that it is the specific activation of K+ fluxes through KCa channels by low concentrations of NS11021 that improves energetic performance and that valinomycin cannot substitute for the effect.
The present work demonstrates that the specific activation of ChTx- or Paxilline-sensitive KCa channels with NS11021 (at nM concentrations) improves mitochondrial energetic performance and that this effect is correlated with enhanced K+ uptake and low amplitude swelling without a large change in ΔΨm. Importantly, our findings (summarized in Fig. 8) also demonstrate that the inactive congener NS13558 did not show any of these effects. In contrast, μM concentrations of NS11021 or NS13558 significantly decreased RCR and the P/O ratio, independent of the mitochondrial KCa channel. Surprisingly, the increase in RCR was due to a decrease in state 4 respiration in response to activation of K+ flux, with similar state 3 respiration. In contrast, μM concentrations of NS11021 or NS13558 not only decreased RCR, but induced depolarization of ΔΨm even in the absence of K+, which was insensitive to K+ channel blockers. The latter nonspecific effect occurred in a concentration range similar to that used by others, when examining the behavior of NS1619, a related, but less potent, KCa channel opener [3, 20]. The lack of effect of NS11021 on the P/O ratio (100μM ADP) indicates that the stoichiometry of ATP production as a function of O2 consumption is not substantially altered; therefore, we would not expect a large change in ATP synthesis with metabolic demand, but would perhaps expect less energy wastage when respiration slows, for example, during conditions of ischemia when the mitochondria move closer to state 4. On the contrary, the nonspecific actions of the NS compounds decrease the P/O ratio and reduce metabolic efficiency.
Structural-functional coupling associated with mitochondrial energetics and volume regulation was described early on by Hackenbrock . Based on electron microscopy data, he showed that the mitochondrial inner membrane/matrix compartment is organized as a network that undergoes geometric rearrangement from tightly packed (“condensed”), evident in mitochondria in state 3, to an expanded (“orthodox”) lattice in state 4. Hackenbrock  observed low amplitude swelling (10–15% relative change) on addition of substrate (state 4 respiration), followed by mitochondrial contraction in the presence of ADP (state4→state 3 transition). Moreover, these changes in volume occurred in association with substantial modifications in energetics, notably ΔΨm, redox, and O2 consumption. In the condensed state, it has been argued that matrix proteins are packed so tightly that the mobility of water is partially restricted  and its viscosity is ~15-fold higher than that of the cytosol . Additionally, Hackenbrock suggested that the reticular proteic network of the matrix is physically continuous with the inner mitochondrial membrane, influencing its conformation and functions including ion and metabolite transport [37, 40]. More recently, these ideas have been supported by electron tomographic imaging combined with cryo-microtomy techniques  further demonstrating that the transitions between orthodox and condensed conformations involves extensive remodeling of the cristae by fusion and fission mechanisms [42, 43].
These observations have led to the idea that changes in mitochondrial volume and cristae structural organization could have an impact on oxidative phosphorylation; and enhanced substrate oxidation with matrix volume expansion has been demonstrated . The present findings demonstrate that changes in K+ influx through mitochondrial KCa channels appear to be obligatory for the improvement in RCR. The beneficial effect was facilitated by permeant anions and preservation of ΔΨm was also important: this characteristic distinguished the positive effect of NS11021 from the nonspecific actions of the compound. Since the nonspecific actions were never, in our experiments, inhibited by selective channel inhibitors, they are also unlikely to be responsible for the cardioprotective actions of K+ channel openers.
The degree of anion permeability was directly correlated with the extent of mitochondrial volume increase and the decrease in state 4 respiration. In fact, in the presence of H2PO4−, the volume increase initiated by K+ uptake was 45% larger, as compared with that obtained in the presence of Cl−. However, the change in ΔΨm was small for all three anions after K+ addition (Fig. 2), suggesting that the cation, anion and water redistribution required to maintain charge balance  during adaptation of matrix volume occurs without a significant change in electrochemical driving force.
Several mechanisms have been proposed to explain how small changes in volume could improve mitochondrial energetic performance, especially with respect to the activation of cardioprotective mitochondrial K+ channels [2, 44, 45]. Potential mechanisms include an optimization of the efficiency of nucleotide exchange by closer apposition of the inner and outer mitochondrial membranes , an increase in reactive oxygen species that trigger protective signaling pathways, and inhibition of glycogen synthase kinase-3 β . An enhanced ability of mitochondria to tolerate high ADP loads in the presence of K+ was also noted as an important protective factor linking K+ fluxes with matrix volume regulation . The present findings suggest an additional novel response of mitochondria to the activation of mitoKCa channels – a decrease in state 4, but not state 3, respiration, resulting in an increase in RCR. The mechanism linking the increase in K+ flux, matrix volume, and the paradoxical decrease in state 4 proton leak will require further investigation. The folding and unfolding of mitochondrial matrix cristae is known to be a dynamic process, and our results suggest that KCa-mediated K+ uptake might alter the topology of the membrane, with concomitant changes in respiratory chain function. Consequently, we propose that this effect could reflect a structural reorganization of the mitochondrial inner membrane induced by the K+-mediated increase in volume, perhaps through a mechano-sensitive ion channel controlling the proton leak across the membrane. This would imply that the changes in volume are coupled with changes in the tension on the membrane. Interestingly, this type of system is present in bacteria, which involves the insertion of a mechano-sensitive channel to protect against osmotic shock, and the insertion pathway is evolutionarily conserved in chloroplasts and mitochondria ; however, no such mechano-sensitive channels have been identified in mitochondria thus far.
The KCa specific effects of NS11021 on mitochondrial function were subtle when compared to the effects of the K+ ionophore valinomycin. While Val also enhanced the K+ uptake rate in a concentration-dependent manner, it induced a larger amplitude swelling. Low concentrations (<100 pM) of Val only slightly depolarized ΔΨm (~10mV) but high concentrations (>200pM) induced an abrupt depolarization when KH2PO4 was added (this effect differed from the nonspecific effects of μM concentrations of NS11021, which depolarized ΔΨm even in sucrose medium in the absence of K+ salt) (Fig. 6). Val treatment did not decrease state 4 respiration at any concentrations tested in the range 50pM to 1nM (see Fig. S1, Supplementary Material). Taken together, the results obtained with Val and NS11021, at pM versus nM concentrations, respectively, suggest that the mechanistic response to K+ transport induced by these two compounds is different. Alternatively, their differential effects on unitary K+ conductance could explain the different energetic responses. In addition, the effect of NS11021 to decrease state 4 respiration does not appear to be a feature of mitochondrial KATP channel openers, which dose dependently increase state 4 respiration in a K+-specific manner .
In summary, the present findings support the idea that the selective activation of mitochondrial KCa channels modulates K+ uptake and volume while maintaining ΔΨm. These properties are likely required to confer protection without compromising oxidative phosphorylation during recovery from metabolic stress. In contrast to other proposed mechanisms involving a slight uncoupling effect (increased respiration not coupled to ATP production), increased K+ flux through the KCa channel apparently decreases state 4 respiration, thus improving the respiratory control ratio of the mitochondria.
Figure S1. Effects of valinomycin on mitochondrial respiration before and after addition of K+ salt. Freshly isolated mitochondria from guinea-pig heart were assayed as described in the legend of Figure 2 after preincubation with increasing valinomycin concentrations (25pM to 1nM). Mitochondrial respiration at state 4 before or after the addition of 10mM KH2PO4 (A) and state 3 (after addition of 1mM ADP) were determined. The RCR (state3/state4) is presented in panel C.
Figure S2. Effects of the inactive congener NS13558 of the KCa activator NS11021 on mitochondrial K+ fluxes, volume and ΔΨm. Freshly isolated mitochondria from guinea-pig heart were assayed under similar conditions as described in the legend of Figure 7. After addition of 10mM KCl (A–C), KAc (D–F), or KH2PO4 (G–I) to state 4 mitochondria in the absence (control) or the presence of 50nM NS13558, the PBFI ratio (A,D,G), 90o light scattering (B,E,H), and ΔΨm, (C,F,I) were monitored simultaneously in duplicates.
The technical assistance of Agnes Sidor and Ling Chen is gratefully acknowledged. This work was supported by NIH grants P01HL081427 and R37HL54598, and The Novo Nordisk Foundation, The Aase and Ejnar Danielsen Foundation, and The Danish Medical Research Council (for MG).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.