MitoBKCa channel is functionally associated with its regulatory β1 subunit in cardiac mitochondria

Abstract

Key points: Association of plasma membrane BK_Ca channels with BK-β subunits shapes their biophysical properties and physiological roles; however, functional modulation of the mitochondrial BK_Ca channel (mitoBK_Ca ) by BK-β subunits is not established. MitoBK_Ca -α and the regulatory BK-β1 subunit associate in mouse cardiac mitochondria. A large fraction of mitoBK_Ca display properties similar to that of plasma membrane BK_Ca when associated with BK-β1 (left-shifted voltage dependence of activation, V_1/2 = -55 mV, 12 µm matrix Ca²⁺ ). In BK-β1 knockout mice, cardiac mitoBK_Ca displayed a low P_o and a depolarized V_1/2 of activation (+47 mV at 12 µm matrix Ca²⁺ ) Co-expression of BK_Ca with the BK-β1 subunit in HeLa cells doubled the density of BK_Ca in mitochondria. The present study supports the view that the cardiac mitoBK_Ca channel is functionally modulated by the BK-β1 subunit; proper targeting and activation of mitoBK_Ca shapes mitochondrial Ca²⁺ handling.

Abstract: Association of the plasma membrane BK_Ca channel with auxiliary BK-β1-4 subunits profoundly affects the regulatory mechanisms and physiological processes in which this channel participates. However, functional association of mitochondrial BK (mitoBK_Ca ) with regulatory subunits is unknown. We report that mitoBK_Ca functionally associates with its regulatory subunit BK-β1 in adult rodent cardiomyocytes. Cardiac mitoBK_Ca is a calcium- and voltage-activated channel that is sensitive to paxilline with a large conductance for K⁺ of 300 pS. Additionally, mitoBK_Ca displays a high open probability (P_o ) and voltage half-activation (V_1/2 = -55 mV, n = 7) resembling that of plasma membrane BK_Ca when associated with its regulatory BK-β1 subunit. Immunochemistry assays demonstrated an interaction between mitochondrial BK_Ca -α and its BK-β1 subunit. Mitochondria from the BK-β1 knockout (KO) mice showed sparse mitoBK_Ca currents (five patches with mitoBK_Ca activity out of 28 total patches from n = 5 different hearts), displaying a depolarized V_1/2 of activation (+47 mV in 12 µm matrix Ca²⁺ ). The reduced activity of mitoBK_Ca was accompanied by a high expression of BK_Ca transcript in the BK-β1 KO, suggesting a lower abundance of mitoBK_Ca channels in this genotype. Accordingly, BK-β1subunit increased the localization of BKDEC (i.e. the splice variant of BK_Ca that specifically targets mitochondria) into mitochondria by two-fold. Importantly, both paxilline-treated and BK-β1 KO mitochondria displayed a more rapid Ca²⁺ overload, featuring an early opening of the mitochondrial transition pore. We provide strong evidence that mitoBK_Ca associates with its regulatory BK-β1 subunit in cardiac mitochondria, ensuring proper targeting and activation of the mitoBK_Ca channel that helps to maintain mitochondrial Ca²⁺ homeostasis.

Keywords: Paxilline; cardiomyocytes; mitochondria; mitochondrial calcium retention capacity; potassium channel.

Figure 1. Preparation of mitoplast from left ventricle cardiomyocytes

A) Healthy rat cardiomyocytes isolated with a Langendorff perfusion apparatus. B) Electron microscopy image of mitoplasts (inner mitochondrial membrane devoid of outer membrane), red arrow. A low percentage of intact swollen mitochondria remained (top-left panel B). C) Mitoplasts labeled with Mitotracker Red visualized with differential interference contrast (DIC) Nomarski microscopy through a 60X objective. The arrow in C indicate a selected mitoplast for patch-clamp. D) DIC image of panel C show a borosilicate patch pipette next to a mitoplast. E) Schematic representation of the preparation of mitoplasts. In an excised patch (inside-out configuration), the bath solution represents the mitochondrial matrix.

Figure 2. Voltage-dependence of MitoBK_Ca channel.

In an excised inside-out patch, the bath solution represents the matrix side of the mitoplast. Channel activity was recorded in symmetric 150 KMeSO₄. A) Single channel currents from a patch that contains at least three channels obtained at 12 µM matrix Ca²⁺ (c=closed state; o=open state). B) The same patch as in A exposed to 5 µM matrix Ca²⁺. Recordings in B show the activity of only one channel at positive membrane potentials. C) Same patch as in A and B exposed to 0.02 µM matrix Ca²⁺, single channel current was observed only at +80 mV. D) Plot of normalized NP_o vs V. Data points were fit to a Boltzmann distribution (continuous lines) with a V_1/2=−55, +42 and +70 mV at 12 (n=8), 5 (n=4), and 0.02 µM Ca²⁺ (n=3), respectively. Z values of 2.2, 1.4 and 1.6 were calculated for 12, 5 and 0.02 µM matrix Ca²⁺, respectively. E) Plots of I vs. V of mitoBK_Ca channels with a slope conductance of 309±19 pS (n=11). Bars represent mean and SEM here and throughout.

Figure 3. Paxilline reduces the open probability of mitoBK_Ca channels.

Single-channel currents of mitoBK_Ca recorded from the same mitoplast in absence (A) and in presence of paxilline (B) at the indicated matrix Ca²⁺ and membrane potential. Paxilline decreased the P_o of the 7 pA channel from 0.77±0.1 to 0.41±0.1 (n=3) as indicated in the current amplitude histograms (C and D). Mean open time (E) and mean closed time (F) observed in 3 different patches. Paxilline augmented significantly the mean closed time of the channel (*p<0.05 Student’s t-test, n=3 hearts).

Figure 4. MitoBK_Ca alpha and auxiliary β1 subunit association in cardiac mitochondria.

A) Immunoprecipitation (IP) of BK_Ca alpha and BK-β1 subunit with an antibody against BK_Ca alpha (see methods). Western blots revealed the presence of GRIM 19 and SERCA as markers of mitochondrial and sarcoplasmic reticulum membranes (4A, lower panels). *The signal of BK-β1 in the whole lysate was detected with chemiluminescent (ECL) substrate Supersignal™ West Femto Maximun sensitivity (see methods). Taken together, these results indicates an association between mitoBK_Ca and its auxiliary β1 subunit. Having demonstrated this, we characterized the biophysical properties of mitoBK_Ca in Wt and β1-KO cardiac mitoplasts. B) Single-channel currents of mitoBK_Ca recorded from Wt mitoplasts. The P_o of the channel was 0.9 ± 0.08, n=4, in average at +80 mV and the indicated matrix Ca²⁺. C) Representative traces of a mitoplast from the β1 KO, were no single-channel current was detected at the indicated voltages and matrix Ca²⁺ (upper traces). Few mitoplasts (5 out of 28, mitoplasts, n=5 hearts) showed mitoBK_Ca channel activity with a P_o of 0.4 ± 0.004, n=3 at +40 mV in the indicated matrix Ca²⁺. D) Plots of I vs. V at 12 µM matrix Ca²⁺ from Wt and β1 KO mitoplasts (data from 6 and 3 patches, respectively). A slope conductance of 302±26 (n=4) and 335 ±15 pS (n=3) was calculated for Wt and β1 KO, respectively. E) Plots of P_o vs. V for mitoBK_Ca channels, data points represent the average of 4 and 3 mitoplasts for Wt and BK-β1 KO, respectively. Bars SEM.

Figure 5. Paxilline reversibly blocks mouse cardiac mitoBK_Ca channel

Single channel current in absence (A) and in presence (B) of paxilline at the indicated voltage and matrix Ca²⁺. C) Current amplitude histogram of the 6 pA channel recorded in A, paxilline reduced the P_o from 0.7±0.1 to 0.2±0.09 (n=3) (D). E) Single channel current recorded after the patch was perfused again with 12 µM matrix Ca²⁺. F) Current amplitude histogram that shows the reappearance of the 6 pA current (P_o=0.4) after washing out paxilline.

Figure 6. Localization of BKDEC and β1 subunit in HeLa cells.

A) Cultured cells co-transfected with BKDEC alpha+β1subunit, loaded with mitotracker red, and DAPI B) β1-flag labeled with an anti-flag C) BKDEC labeled with anti-HA. D) Merge of mitotracker red, BKDEC and β1. E) Cross-correlation index (CCi), relative to mitotracker red signal, calculated from cells overexpressing BKDEC alone (Cci=0.53±0.05) and co-expressed with β1 subunit (Cci=0.77±0.01, n=11 cells from 3 different preparations).*p<0.05.

Figure 7. Mitochondrial calcium uptake is impaired by Paxilline and by genetic ablation of β1-subunit

A) Paxilline reduces the amount of mitochondrial Ca²⁺ necessary to induce formation and opening of mPTP. B) Mitochondria from the β1 KO required less Ca²⁺ than mitochondria from the Wt to trigger the opening of mPTP. C) Calcium retention capacity values obtained in A for paxilline and in B for the β1-KO mitochondria. Bars SEM, (*p<0.05 Student’s t-test). Black downward arrow, addition of mitochondria. Black upward arrows, 10 nmol CaCl₂ pulses. (▼) opening of mPTP.

Figure 8. A different population of mitoBK_Ca in cardiac mitochondria

A) Single channel currents recorded in 0.02 µM matrix Ca²⁺ at the indicated voltages. B) Same patch as in A exposed to 12 µM matrix Ca²⁺. C) Plot of I vs. V, a slope conductance of 284 ± 6 pS (n=7) was calculated. D) Plots of P_o vs. V from the currents observed in A and B, data points represent an average of ≥4 different patches at the indicated matrix Ca²⁺.