Mitochondrial modulation of Ca²⁺ sparks and transient K_Ca currents in smooth muscle cells of rat cerebral arteries

Abstract

Mitochondria sequester and release calcium (Ca²⁺) and regulate intracellular Ca²⁺ concentration ([Ca²⁺]_i) in eukaryotic cells. However, the regulation of different Ca²⁺ signalling modalities by mitochondria in smooth muscle cells is poorly understood. Here, we investigated the regulation of Ca²⁺ sparks, Ca²⁺ waves and global [Ca²⁺]_i by mitochondria in cerebral artery smooth muscle cells. CCCP (a protonophore; 1 μm) and rotenone (an electron transport chain complex I inhibitor; 10 μm) depolarized mitochondria, reduced Ca²⁺ spark and wave frequency, and elevated global [Ca²⁺]_i in smooth muscle cells of intact arteries. In voltage-clamped (−40 mV) cells, mitochondrial depolarization elevated global [Ca²⁺]_i, reduced Ca²⁺ spark amplitude, spatial spread and the effective coupling of sparks to large-conductance Ca²⁺-activated potassium (K_Ca) channels, and decreased transient K_Ca current frequency and amplitude. Inhibition of Ca²⁺ sparks and transient K_Ca currents by mitochondrial depolarization could not be explained by a decrease in intracellular ATP or a reduction in sarcoplasmic reticulum Ca²⁺ load, and occurred in the presence of diltiazem, a voltage-dependent Ca²⁺ channel blocker. Ru360 (10 μm), a mitochondrial Ca²⁺ uptake blocker, and lonidamine (100 μm), a permeability transition pore (PTP) opener, inhibited transient K_Ca currents similarly to mitochondrial depolarization. In contrast, CGP37157 (10 μm), a mitochondrial Na⁺–Ca²⁺ exchange blocker, activated these events. The PTP blockers bongkrekic acid and cyclosporin A both reduced inhibition of transient K_Ca currents by mitochondrial depolarization. These results indicate that mitochondrial depolarization leads to a voltage-independent elevation in global [Ca²⁺]_i and Ca²⁺ spark and transient K_Ca current inhibition. Data also suggest that mitochondrial depolarization inhibits Ca²⁺ sparks and transient K_Ca currents via PTP opening and a decrease in intramitochondrial [Ca²⁺].

Intracellular calcium (Ca²⁺) signals regulate a diverse array of physiological functions including secretion, contraction, transcription, and apoptosis (Berridge et al. 2000; Duchen, 2000). To control Ca²⁺-dependent physiological processes, cells have developed a variety of mechanisms that regulate the intracellular free Ca²⁺ concentration ([Ca²⁺]_i) (Berridge et al. 2000). Although it has been known for many years that mitochondria can transport Ca²⁺, recent evidence has suggested that these organelles are important physiological modulators of intracellular Ca²⁺ signalling (Berridge et al. 2000; Duchen, 2000).

Mitochondria sequester Ca²⁺ from the cytosol via a low-affinity, high-capacity inner membrane uniporter (Bernardi, 1999; Gunter et al. 2000). The driving force for Ca²⁺ uptake is created by the mitochondrial membrane potential (ΔΨ_m), which is generated by the extrusion of protons (H⁺) by the electron transport chain and is ∼150–200 mV more negative than the cytosol (Bernardi, 1999; Gunter et al. 2000). Conversely, mitochondrial Ca²⁺ release can occur via a number of mechanisms, including the inner membrane Na⁺–Ca²⁺ exchanger and the permeability transition pore (PTP) (Bernardi, 1999). The spatial location of mitochondria nearby sites of intracellular Ca²⁺ release exposes these organelles to elevated microdomains of [Ca²⁺]_i and establishes communication with the sarcoplasmic/endoplasmic reticulum (SR/ER) (Rizzuto et al. 1999).

In arterial smooth muscle cells, several intracellular Ca²⁺ signalling modalities occur that differ in respect to spatial localization, temporal kinetics, and physiological function (Jaggar et al. 2000). Ca²⁺ sparks are spatially restricted [Ca²⁺]_i transients that occur due to the opening of multiple SR ryanodine-sensitive Ca²⁺ release (RyR) channels (Jaggar et al. 2000). In smooth muscle cells, Ca²⁺ sparks induce a local micromolar [Ca²⁺]_i elevation and activate several large-conductance Ca²⁺-sensitive K⁺ (K_Ca) channels to elicit a transient outward K⁺ current (Jaggar et al. 2000; Perez et al. 2001; ZhuGe et al. 2002), which has also been termed a ‘spontaneous transient outward current’ or ‘STOC’ (Benham & Bolton, 1986). In arteries at physiological levels of pressure, inhibition of Ca²⁺ sparks or K_Ca channels leads to membrane depolarization, activation of voltage-dependent Ca²⁺ channels, an elevation in arterial wall [Ca²⁺]_i and constriction (Nelson et al. 1995; Jaggar, 2001). Ca²⁺ waves are propagating, asynchronous intracellular [Ca²⁺]_i elevations that contribute Ca²⁺ for contraction (Kasai et al. 1997; Sward et al. 2002), although Ca²⁺ waves that do not induce vasoconstriction have also been observed (Miriel et al. 1999). In arterial smooth muscle cells, Ca²⁺ wave frequency is increased by RyR channel activation and vasoconstrictors that elevate inositol trisphosphate (Iino et al. 1994; Miriel et al. 1999; Jaggar & Nelson, 2000; Heppner et al. 2002). Global [Ca²⁺]_i is the cytosolic [Ca²⁺] of arterial smooth muscle cells that results from Ca²⁺ entry through voltage-dependent Ca²⁺ channels and Ca²⁺ release from intracellular stores (for review see Jaggar et al. 2000). An elevation in global [Ca²⁺]_i leads to vasoconstriction (Jaggar, 2001).

Global [Ca²⁺]_i transients elevate intramitochondrial [Ca²⁺] in arterial smooth muscle cells, suggesting that mitochondria function to reduce [Ca²⁺]_i in this tissue (Drummond & Tuft, 1999; Monteith & Blaustein, 1999; Gurney et al. 2000; Szado et al. 2003). Similarly, agents that depolarize mitochondria prolong the decay of [Ca²⁺]_i transients and elevate spatially averaged [Ca²⁺]_i (Kamishima et al. 2000; Kamishima & Quayle, 2002; Wang et al. 2003). Recent studies also demonstrate that electron transport chain blockers activate Ca²⁺ sparks in pulmonary artery smooth muscle cells (Wang et al. 2003) and modify agonist-induced Ca²⁺ waves in tail artery smooth muscle cells (Sward et al. 2002). Collectively, these studies suggest that mitochondria regulate local and global Ca²⁺ signalling events in smooth muscle cells.

In the present study, we investigated the regulation of local and global [Ca²⁺]_i signalling modalities by mitochondria in cerebral artery smooth muscle cells. Mitochondrial depolarization with rotenone or CCCP blocked Ca²⁺ sparks and Ca²⁺ waves and elevated global [Ca²⁺]_i in smooth muscle cells of intact arteries. In voltage-clamped smooth muscle cells, mitochondrial depolarization reduced transient K_Ca currents by decreasing Ca²⁺ spark frequency, amplitude, and spatial spread, and by reducing the effective coupling of Ca²⁺ sparks to K_Ca channels. Data also suggest that mitochondrial depolarization inhibits Ca²⁺ sparks and transient K_Ca currents by inducing PTP opening and a decrease in intramitochondrial [Ca²⁺].

Methods

Tissue preparation

Procedures involving animals were approved by the Animal Care and Use Committee policies at the University of Tennessee. Sprague-Dawley rats (200–250 g) of either sex were killed by peritoneal injection of a sodium pentobarbital overdose (150 mg kg⁻¹). The brain was removed and placed into ice-cold (4°C), oxygenated (21% O₂–5% CO₂), physiological saline solution (PSS) containing (mm): 119 NaCl, 4.7 KCl, 24 NaHCO₃, 1.2 KH₂PO₄, 1.6 CaCl₂, 1.2 MgSO₄, 0.023 EDTA and 11 glucose. Posterior cerebral and cerebellar arteries (50–200 μm in diameter) were removed, cleaned of connective tissue and maintained in ice-cold PSS. Individual smooth muscle cells were dissociated from cerebral arteries using an enzyme procedure similar to that previously described (Jaggar, 2001).

Confocal Ca²⁺ imaging

Arterial segments (1–2 mm in length) or isolated smooth muscle cells were placed into Hepes-buffered PSS containing (mm): 134 NaCl, 6 KCl, 2 CaCl₂, 1 MgCl₂, 10 Hepes and 10 glucose (pH 7.4, NaOH) containing 10 μm fluo-4 AM and 0.05% pluronic F-127 for 1 h or 15 min, respectively, at 22°C. Arterial segments or cells were then placed into Hepes-buffered PSS for 30 min to allow indicator de-esterification. Smooth muscle cells were imaged using a Noran Oz laser scanning confocal microscope (Noran Instruments, Middleton, WI, USA) and a × 60 water immersion objective (NA = 1.2) attached to a Nikon TE300 microscope by illuminating with a krypton-argon laser at 488 nm and capturing emitted light >500 nm. Confocal imaging of isolated voltage-clamped (−40 mV) smooth muscle cells was done using a Hepes-buffered PSS (described above) bath solution. Smooth muscle cells in intact arterial segments were imaged in a 30 mm K⁺ bath solution containing (mm): NaCl 110; KCl 30; Hepes 10; CaCl₂ 2; MgCl₂ 1; and glucose 10 (pH 7.4, NaOH) (see Jaggar et al. 1998; Cheranov & Jaggar, 2002 for similar procedure). 30 mm K⁺ depolarizes arterial smooth muscle cells to ∼−40 mV (Harder, 1980), which is similar to the voltage used in patch-clamp experiments in this study. Images (56.3 μm × 52.8 μm) were recorded every 8.3 ms (120 images s⁻¹) in isolated cells or every 16.7 ms (60 images s⁻¹) in intact arteries. For imaging smooth muscle cells in arteries, at least two different representative areas of the same segment were scanned for at least 10 s under each condition. The same part of the artery was scanned only once to avoid any laser-induced changes in Ca²⁺ signalling, and the effects of drugs were measured in paired experiments. In experiments where confocal microscopy was used in combination with patch clamp, electrophysiological and fluorescence measurements were synchronized using a light emitting diode placed above the recording chamber that was triggered during acquisition. Each isolated smooth muscle cell was imaged for at least 10 s under each condition. Ca²⁺ sparks were detected in smooth muscle cells using custom analysis software and manual analysis (see Cheranov & Jaggar, 2002 for detailed description). Detection of Ca²⁺ sparks was performed by dividing an area 1.54 μm (7 pixels) × 1.54 μm (7 pixels) (i.e. 2.37 μm²) in each image (F) by a baseline (F₀) which was determined by averaging 10 images without Ca²⁺ spark activity. The entire area of each image was analysed to detect Ca²⁺ sparks. A Ca²⁺ spark was identified as a local increase in F/F₀ that was greater than 1.2. The calculated root mean square of baseline fluorescence noise is 0.06. Ca²⁺ spark spatial spread was calculated as the full width at half-maximal amplitude (FWHM). Ca²⁺ waves were analysed by placing 2.2 × 2.2 μm boxes in individual smooth muscle cells and refer to a change in F/F₀ of > 1.2 that propagated for at least 20 μm. Global Ca²⁺ fluorescence was calculated from the same images used for Ca²⁺ spark analysis and was the mean pixel value of 100 different images acquired over 10 s. Changes in local or global [Ca²⁺]_i were calculated using the pseudo ratio method (Cheng et al. 1993):

where K is the apparent affinity of fluo 4 for Ca²⁺ (770 nm; Woodruff et al. 2002), R is the fractional fluorescence increase (F/F₀), and [Ca²⁺]_rest is the cytosolic [Ca²⁺]_i at F₀. In isolated cells, Ca²⁺ spark amplitude in control and CCCP was calculated as the local [Ca²⁺]_i increase (i.e. Δ[Ca²⁺]_spark=[Ca²⁺]_spark−[Ca²⁺]_global) within a 1.54 × 1.54 μm region. Arterial wall [Ca²⁺]_rest used for Ca²⁺ calibrations was measured using a photomultiplier tube (Ionoptix, Milton, MA, USA) and the ratiometric Ca²⁺ indicator, fura-2 (see below for detailed methodology). Elevation of extracellular K⁺ from 6 to 30 mm increased arterial wall [Ca²⁺]_i from 92 ± 6 nm to 193 ± 8 nm (n= 7).

Fura-2 imaging

Isolated cerebral artery smooth muscle cells were incubated with the ratiometric fluorescent Ca²⁺ indicator fura-2 AM (2 μm) and 0.02% pluronic F-127 for 20 min, followed by a 15 min wash. All experiments in isolated cells were performed using a 6 mm K⁺ Hepes-buffered bath solution (composition described above). Smooth muscle cells were alternately exited at 340 or 380 nm using a PC-driven hyperswitch (Ionoptix). Background corrected ratios were collected every 1 s at 510 nm using a Dage MTI integrating CCD camera (Ionoptix). [Ca²⁺]_SR was estimated by rapidly applying a high concentration of caffeine (10 mm), a RyR channel activator, and measuring the amplitude of the [Ca²⁺]_i transient (i.e. Δ[Ca²⁺]_i). [Ca²⁺]_i concentrations were calculated using the following equation (Grynkiewicz et al. 1985):

where R is the 340/380 nm ratio, R_min and R_max are the minimum and maximum ratios determined in Ca²⁺-free and saturating Ca²⁺ solutions, respectively, S_f2/S_b2 is the Ca²⁺-free/Ca²⁺-replete ratio of emissions at 380 nm excitation, and K_d is the dissociation constant for fura-2 (224 nm, Grynkiewicz et al. 1985). R_min, R_max, S_f2 and S_b2 were determined at the end of experiments and in separate experiments by increasing the Ca²⁺ permeability of smooth muscle cells with ionomycin (10 μm), and perfusing cells with a high Ca²⁺ (50 mm) or Ca²⁺-free (no added Ca²⁺, 10 mm EGTA) solution.

TMRM imaging

Isolated cerebral artery smooth muscle cells were incubated with TMRM (1 μm) for 15 min followed by a 15 min wash. TMRM is a cationic fluorescent potentiometric indicator that accumulates in mitochondria due to their negative ΔΨ_m. All experiments in isolated cells were performed using a 6 mm K⁺ Hepes-buffered bath solution (composition described above). Smooth muscle cells were excited with 535 nm light and background corrected fluorescence intensity was collected every 2 s at 610 nm using a Dage MTI integrating CCD camera (Ionoptix).

Patch-clamp electrophysiology

Isolated cells were allowed to adhere to a glass coverslip in the bottom of a chamber for 10 min prior to experimentation. Potassium currents were measured using either the conventional whole-cell or perforated-patch configuration of the patch-clamp technique using an Axopatch 200B amplifier (Axon Instruments, Union City, CA, USA). Bath solution was 6 mm K⁺ Hepes-buffered PSS (composition described above). For perforated-patch experiments, the pipette solution contained (mm): 110 potassium aspartate, 30 KCl, 10 NaCl, 1 MgCl₂, 10 Hepes, 0.05 EGTA (pH 7.2 with KOH). For conventional whole-cell experiments, the pipette (i.e. intracellular) solution contained (mm): 140 KCl, 1.9 MgCl₂, 0.037 CaCl₂, 0.1 EGTA, 10 Hepes, 2 Na₂ATP (pH 7.2 with KOH); the calculated free Ca²⁺ and free Mg²⁺ concentrations of this solution are 100 nm and 1 mm, respectively (WEBMAXC, Stanford University, CA, USA). The perforated-patch configuration was used, unless stated otherwise. All experiments were performed with a holding potential of −40 mV, unless stated otherwise. Membrane currents were recorded with a sample rate of 2.5 kHz and filtered at 1 kHz. In experiments where the activity of single K_Ca channel currents was measured, Ca²⁺ sparks, and hence transient K_Ca currents, were abolished with thapsigargin, a SR Ca²-ATPase inhibitor (Nelson et al. 1995). In each patch under each condition, single K_Ca channel activity (NP_o) was calculated from 5 min of continuous gap-free data using Fetchan 6 (Axon Instruments). NP_o was calculated from the following equation: NP_o= (t_ii)/T, where t_i is the time at each channel level i and T is the total time of analysis. Transient K_Ca current analysis was performed off-line using methodology described elsewhere (Cheranov & Jaggar, 2002). A transient K_Ca current was defined as the simultaneous opening of three or more K_Ca channels.

Statistical analysis

Values are expressed as mean ± standard error of the mean. Student's t tests were used for comparing paired or unpaired data, and Student-Newman-Keuls test was used for comparing multiple data sets. First-order polynomial linear fits were used to calculate statistical correlation between the amplitude of Ca²⁺ sparks and evoked transient K_Ca currents (Origin, OriginLab Corp., Northampton, MA, USA). ANCOVA of first-order polynomial best fits was used to compare amplitude correlation data sets (Graphpad Prizm, San Diego, CA, USA). P < 0.05 was considered significant.

Chemicals

Unless otherwise stated, all chemicals used in this study were obtained from Sigma Chemical Company (St Louis, MO, USA). Papain was purchased from Worthington Biochemical Co. (Lakewood, NJ, USA), fura-2 AM, fluo-4 AM and pluronic F-127 from Molecular Probes (Eugene, OR, USA), bongkrekic acid and Ru360 from Calbiochem (La Jolla, CA, USA) and CGP37157 from Tocris Cookson Inc. (Ellisville, MO, USA).

Results

Rotenone and CCCP depolarize mitochondria in cerebral artery smooth muscle cells

Mitochondrial Ca²⁺ uptake is driven by the negative mitochondrial potential (Gunter et al. 2000). Rotenone, an electron transport chain complex I inhibitor, and CCCP, a protonophore, depolarize mitochondria by inducing H⁺ retention and by permeabilizing the inner membrane to H⁺, respectively. To evaluate the time course of mitochondrial depolarization caused by rotenone and CCCP in cerebral artery smooth muscle cells, mitochondrial potential was measured using the fluorescent indicator TMRM.

Rotenone and CCCP gradually decreased TMRM fluorescence, which reached a plateau ∼0.4 of control after approximately 5 min (Fig. 1). In contrast, oligomycin B (1 μm), an F₁F₀-ATP synthase blocker, did not alter TMRM fluorescence over the same time course (Fig. 1). These data indicate that rotenone and CCCP depolarize mitochondria in cerebral artery smooth muscle cells via a mechanism that is independent of a decrease in mitochondrial ATP synthesis.

Figure 1. Rotenone and CCCP depolarize mitochondria in cerebral artery smooth muscle cells.

Original recordings illustrating the mean percentage change in TMRM fluorescence in isolated arterial smooth muscle cells following application (indicated by arrow) of rotenone (10 μm), CCCP (1 μm) or oligomycin B (1 μm). Each trace represents the mean fluorescence change of 7, 10, and 11 cells for rotenone, CCCP and oligomycin B, respectively.

Mitochondrial depolarization inhibits Ca²⁺ sparks and waves and elevates global [Ca²⁺]_i in smooth muscle cells of intact cerebral arteries

To investigate the regulation of intracellular Ca²⁺ signalling modalities by mitochondria, smooth muscle cells within the wall of cerebral arteries were imaged in control and 5 min after sustained application of rotenone (10 μm), CCCP (1 μm) or oligomycin B (1 μm). To compare confocal imaging data with electrophysiological recordings presented in this study, Ca²⁺ sparks in smooth muscle cells of arterial segments were measured in an extracellular solution containing 30 mm K⁺ (see Jaggar et al. 1998; Cheranov & Jaggar, 2002 for similar procedure). Thirty millimolar K⁺ depolarizes arterial smooth muscle cells to ∼−40 mV (Harder, 1980).

Rotenone decreased mean Ca²⁺ spark frequency and amplitude to ∼0.66 and ∼0.75 of control, respectively, and reduced mean Ca²⁺ wave frequency to ∼0.24 of control (Fig. 2A–C). In contrast, rotenone increased mean global [Ca²⁺]_i in the same cells from ∼193 nm (determined with fura-2 in separate experiments; see Methods) to 255 nm (Fig. 2A). In a separate series of experiments, CCCP similarly inhibited Ca²⁺ sparks and Ca²⁺ waves, and elevated global [Ca²⁺]_i (Fig. 2B and C). In contrast, oligomycin B did not alter Ca²⁺ sparks, Ca²⁺ waves, or global [Ca²⁺]_i (Fig. 2B and C). These data suggest that mitochondrial depolarization inhibits Ca²⁺ sparks and Ca²⁺ waves and elevates global [Ca²⁺]_i in cerebral artery smooth muscle cells via a mechanism that does not involve a decrease in ATP.

Figure 2. Regulation of Ca²⁺ sparks, Ca²⁺ waves and global [Ca²⁺]_i in smooth muscle cells of intact cerebral arteries by rotenone, CCCP and oligomycin B.

A, average fluorescence (100 of 600 images) over 10 s of two different 56.3 μm × 52.8 μm areas of the same cerebral artery in control and 5 min after application of rotenone (10 μm). The locations of Ca²⁺ sparks that occurred during 10 s are indicated by white boxes (1.54 μm × 1.54 μm). Representative localized F/F₀ changes over time are illustrated below respective images and labelled accordingly. Average relative effects on Ca²⁺ spark (B) and wave frequency (C) and global [Ca²⁺]_i (D) of a 5 min application of rotenone, CCCP or oligomycin B. Rotenone decreased mean Ca²⁺ spark frequency from 1.18 ± 0.23 to 0.75 ± 0.12 Hz, Ca²⁺ spark amplitude (F/F₀) from 1.31 ± 0.01 to 1.24 ± 0.01, and wave frequency from 0.39 ± 0.07 to 0.09 ± 0.06 Hz (n= 6 arteries). CCCP decreased mean Ca²⁺ spark frequency from 1.05 ± 0.25 to 0.33 ± 0.01 Hz, spark amplitude (F/F₀) from 1.40 ± 0.01 to 1.25 ± 0.01, and wave frequency from 0.31 ± 0.04 to 0.04 ± 0.02 Hz (n= 6 arteries). Oligomycin B did not alter mean Ca²⁺ spark frequency (control, 1.03 ± 0.12; oligomycin B, 1.00 ± 0.38 Hz), spark amplitude (control, 1.34 ± 0.02; oligomycin B, 1.37 ± 0.02), or wave frequency (control, 0.31 ± 0.05; oligomycin B, 0.30 ± 0.04 Hz (n= 4 arteries). *P < 0.05 using Students t test.

Rotenone and CCCP block transient K_Ca currents in isolated arterial smooth muscle cells

In arterial smooth muscle cells, a Ca²⁺ spark activates several K_Ca channels, resulting in a transient K_Ca current (Jaggar et al. 2000). If mitochondrial depolarization inhibits Ca²⁺ sparks, then this should also decrease transient K_Ca currents.

At −40 mV, rotenone (10 μm) and CCCP (1 μm) similarly reduced transient K_Ca current frequency and amplitude when using either the perforated-patch configuration, or the conventional whole-cell configuration with 2 mm ATP included in the pipette solution (Fig. 3A and C). For example, when using the perforated-patch configuration, rotenone decreased mean transient K_Ca current frequency from ∼0.74 to 0.15 Hz in 5 min, or to ∼0.28 of control (Fig. 3A and C and 10B). Rotenone also reduced mean transient K_Ca current amplitude from ∼15.1 to 10.2 pA, or to ∼0.69 of control over the same time course. Ten minutes after rotenone or CCCP application, transient K_Ca currents were almost completely abolished (Fig. 3A, also see Fig. 10C). In contrast, oligomycin B (1 μm) did not inhibit transient K_Ca currents over the same time course. Five minutes after application of oligomycin B, transient K_Ca currents were unaltered, although prolonged application (∼15 min) ultimately decreased the frequency and amplitude of these events (Figs 3B and C). Simultaneous application of CCCP and oligomycin B, which prevents mitochondrial consumption of ATP during depolarization (Duchen, 2000), blocked transient K_Ca currents similarly to when CCCP was applied alone (Fig. 3C). These results suggest mitochondrial depolarization inhibits transient K_Ca currents in arterial smooth muscle cells via a mechanism that does not involve a decrease in cytosolic ATP.

Figure 3. Regulation of transient K_Ca currents in isolated voltage-clamped arterial smooth muscle cells by rotenone, CCCP and oligomycin B.

A and B, original recordings of transient K_Ca currents in cerebral artery smooth muscle cells voltage-clamped at −40 mV using the perforated-patch configuration. Rotenone (10 μm, A) reduced transient K_Ca current frequency and amplitude, whereas oligomycin B (1 μm, B) did not alter transient K_Ca currents. C, average relative changes in transient K_Ca current frequency (left panel) and amplitude (right panel) when compared with control of a 5 min application of: rotenone (10 μm, p-p, n= 7), rotenone (10 μm, w-c, n= 10), CCCP (1 μm, p-p, n= 6), CCCP (1 μm, w-c, n= 7), CCCP (1 μm) + oligomycin B (1 μm, p-p, n= 7), and oligomycin B (1 μm) at 5 and 15 min (p-p, n= 6 for each). p-p, perforated-patch; w-c, conventional whole-cell configuration. *P < 0.05 using Students t test.

Figure 10. Mitochondrial PTP blockers reduce rotenone inhibition of transient K_Ca currents.

A, original recording of transient K_Ca currents in a cerebral artery smooth muscle cell voltage-clamped at −40 mV. Bongkrekic acid (10 μm) reduced inhibition of transient K_Ca currents by rotenone (10 μm). B, frequency histogram illustrating the time course of transient K_Ca current inhibition by rotenone applied alone (from Fig. 3A, black bars) or in the presence of bongkrekic acid (from Fig. 10A, hatched bars). C, average relative effects on transient K_Ca current frequency (left panel) and amplitude (right panel) of bongkrekic acid (BA, 10 μm, n= 3), cyclosporin A (CsA, 1 μm, n= 4), rotenone (10 μm, n= 5) alone, or rotenone applied in the presence of bongkrekic acid (10 μm, n= 6) or cyclosporin A (1 μm, n= 7). *P < 0.05 using Student-Newman-Keuls test.

CCCP alters Ca²⁺ spark spatial and temporal properties and attenuates the effective coupling of Ca²⁺ sparks to K_Ca channels

To examine mechanisms by which mitochondrial depolarization inhibits Ca²⁺ sparks and transient K_Ca currents, Ca²⁺ spark spatial-temporal properties and the signalling relationship between sparks and transient K_Ca currents were measured in the same voltage-clamped (−40 mV) cells in control and 5 min following application of CCCP.

CCCP decreased Ca²⁺ spark amplitude (ΔCa²⁺]_i) from 671 ± 67 nm in control to 433 ± 41 nm, reduced spatial spread (FWHM) from 3.8 ± 0.2 μm to 3.1 ± 0.1 μm, and accelerated the rate of decay (t_1/2) from 72 ± 4 ms to 55 ± 4 ms (P < 0.05 for each). In addition, CCCP reduced the slope of the amplitude correlation between Ca²⁺ sparks and evoked transient K_Ca currents from 0.018 to 0.011 or to 0.61 of control (P < 0.05, Fig. 4A and B). In contrast, CCCP increased mean global [Ca²⁺]_i in smooth muscle cells from ∼193 to 253 ± 32 nm. These data suggest that in voltage-clamped arterial smooth muscle cells, mitochondrial depolarization alters the spatial and temporal properties of Ca²⁺ sparks, reduces the effective coupling of Ca²⁺ sparks to K_Ca channels, and elevates global [Ca²⁺]_i.

Figure 4. CCCP attenuates the coupling relationship between Ca²⁺ sparks and transient K_Ca currents.

A, original simultaneous recordings of Ca²⁺ sparks and transient K_Ca currents in the same voltage-clamped (−40 mV) cerebral artery smooth muscle cell. The black trace illustrates whole cell K⁺ current. Red and green traces illustrate fluorescence changes (F/F₀) measured in two different areas of the cell where Ca²⁺ sparks occurred. Traces show activity in control and 5 min after CCCP (1 μm) and oligomycin B (1 μm). B, scatter plot of Ca²⁺ spark and evoked transient K_Ca current amplitude at −40 mV obtained in the same cells in control (black), and after CCCP (red, n= 5 cells). Ca²⁺ spark amplitude is calculated as the local elevation in [Ca²⁺]_i (i.e. Δ[Ca²⁺]_i) reported by fluo-4. First-order polynomial linear fits for control and CCCP data are illustrated with confidence bands. CCCP reduced the slope of the amplitude correlation between a spark and the evoked transient K_Ca current from 0.018 ± 0.001 to 0.011 ± 0.001 (P < 0.05). Ca²⁺ spark and associated transient K_Ca current amplitudes were significantly correlated for control and CCCP (P < 0.05 for each), but CCCP reduced the correlation coefficient (r) from 0.61 to 0.32.

In the absence of Ca²⁺ sparks, rotenone and CCCP activate K_Ca channels

K_Ca channels are exposed to micromolar [Ca²⁺]_i during a spark (Perez et al. 2001; ZhuGe et al. 2002). A CCCP-induced decrease in the effective coupling of Ca²⁺ sparks to K_Ca channels suggests that mitochondrial depolarization inhibits K_Ca channels. To investigate this hypothesis, the regulation of K_Ca channel activity by mitochondrial depolarization was measured using the perforated-patch clamp configuration. To block Ca²⁺ sparks and transient K_Ca currents, experiments were performed in cells pretreated with thapsigargin, a SR Ca²⁺ ATPase inhibitor (Jaggar et al. 2000).

At 0 mV, rotenone increased K_Ca channel activity (NP_o) ∼2.6-fold (Fig. 5A and B). In a separate series of experiments, CCCP increased K_Ca channel activity ∼2-fold (Fig. 5B). Thus, in the absence of Ca²⁺ sparks, mitochondrial depolarization elevates K_Ca channel activity, presumably by elevating global [Ca²⁺]_i (see Figs 2 and 6).

Figure 5. Rotenone and CCCP activate K_Ca channels.

A, original recordings of single K_Ca channels measured at 0 mV using the perforated-patch configuration. Ca²⁺ sparks, and thus transient K_Ca currents, were abolished with thapsigargin (100 nm). Rotenone increased K_Ca channel activity. ‘C’ indicates closed level. B, average relative effect of rotenone (10 μm) or CCCP (1 μm) on K_Ca channel activity (NP_o). Rotenone increased mean NP_o from 0.017 ± 0.003 to 0.044 ± 0.024 (n= 7 cells). CCCP increased mean NP_o from 0.021 ± 0.005 to 0.042 ± 0.016 (n= 5 cells). *P < 0.05 using Students t test.

Figure 6. Regulation of SR Ca²⁺ load by CCCP, rotenone and oligomycin B.

A and B, original traces illustrating the regulation of intracellular Ca²⁺ concentration and caffeine (10 mm)-induced [Ca²⁺]_i transients in single cerebral artery smooth muscle cells by CCCP (1 μm, A) or oligomycin B (1 μm, B). C, average relative caffeine-induced [Ca²⁺]_i transients in control (two applications 5 min apart), and five, 10 and 15 min after CCCP (1 μm, n= 9 cells), rotenone (10 μm, n= 5 cells) or oligomycin B (1 μm, n= 11 cells). *P < 0.05, when compared with control using Student-Newman-Keuls test.

Regulation of sarcoplasmic reticulum Ca²⁺ load by rotenone, CCCP and oligomycin B

In HeLa and HEK293 cells, mitochondria supply the ER with a local source of Ca²⁺ for sequestration (Arnaudeau et al. 2001). If a similar process exists in smooth muscle cells, mitochondrial depolarization may disrupt this supply mechanism, which would decrease SR Ca²⁺ load and inhibit Ca²⁺ sparks and transient K_Ca currents (Cheranov & Jaggar, 2002). To investigate this hypothesis, the regulation of global [Ca²⁺]_i and SR Ca²⁺ load by CCCP (1 μm), rotenone (10 μm), and oligomycin B (1 μm) was measured in isolated cells.

Rotenone and CCCP elevated [Ca²⁺]_i, whereas oligomycin B did not change [Ca²⁺]_i (Fig. 6). Five minutes after application of either CCCP, rotenone, or oligomycin B, the mean amplitude of caffeine-induced [Ca²⁺]_i transients were unaltered (Fig. 6). However, 10 and 15 min after application of CCCP, rotenone, or oligomycin B, caffeine-induced [Ca²⁺]_i transients were significantly smaller when compared with control (Fig. 6). Because transient K_Ca currents (Figs 3 and 4) were reduced to ∼0.25 of control after a 5 min application of rotenone or CCCP, but SR Ca²⁺ load was unchanged, these data suggest that mitochondrial depolarization does not inhibit Ca²⁺ sparks and transient K_Ca currents due to a decrease in SR Ca²⁺ load.

Mitochondrial depolarization inhibits transient K_Ca currents independently of voltage-dependent Ca²⁺ channels

Voltage-dependent Ca²⁺ channel blockers inhibit Ca²⁺ sparks and transient K_Ca currents in arterial smooth muscle cells (Jaggar et al. 1998; Cheranov & Jaggar, 2002). Conceivably, mitochondrial depolarization may inhibit Ca²⁺ sparks and transient K_Ca currents by blocking voltage-dependent Ca²⁺ channels. To investigate this hypothesis, rotenone was applied in the presence of diltiazem (50 μm), a voltage-dependent Ca²⁺ channel blocker. Diltiazem reduced mean transient K_Ca current frequency to ∼0.46 of control, but did not change mean transient K_Ca current amplitude (Fig. 7A and B). In the same cells in the presence of diltiazem, rotenone reduced mean transient K_Ca current frequency and amplitude to ∼0.25 and ∼0.5 of control, respectively (Fig. 7A and B). These results suggest mitochondrial depolarization inhibits transient K_Ca currents via a mechanism that is independent of Ca²⁺ entry via voltage-dependent Ca²⁺ channels.

Figure 7. Rotenone blocks transient K_Ca currents when applied in the continued presence of diltiazem, and H₂O₂ activates transient K_Ca currents.

Average relative effects of diltiazem (50 μm, n= 6), rotenone (10 μm) applied in the continued presence of diltiazem (50 μm, n= 6), or H₂O₂ (100 μm, n= 7) on transient K_Ca current frequency (A) and amplitude (B). *P < 0.05 using using Student-Newman-Keuls test.

H₂O₂ activates transient K_Ca currents

In arterial smooth muscle cells, mitochondrial depolarization generates reactive oxygen species (ROS) (Michelakis et al. 2002). To investigate if mitochondrial depolarization inhibits Ca²⁺ sparks and transient K_Ca currents via an increase in ROS, the regulation of transient K_Ca currents by H₂O₂ was examined. At −40 mV, H₂O₂ (100 μm) increased mean transient K_Ca current frequency ∼1.6-fold, but did not alter mean transient K_Ca current amplitude (Fig. 7A and B). These data suggest mitochondrial depolarization does not inhibit transient K_Ca currents due to an increase in H₂O₂.

Differential regulation of transient K_Ca currents by inhibitors of mitochondrial Ca²⁺ uptake and release

Mitochondrial depolarization reduces the driving force for Ca²⁺ uptake, leading to a decrease in intramitochondrial [Ca²⁺] (for reviews see Gunter et al. 2000; Duchen, 2000). Conceivably, rotenone and CCCP could inhibit transient K_Ca currents by decreasing intramitochondrial [Ca²⁺]. To investigate this hypothesis, the regulation of transient K_Ca currents by pharmacological blockers of mitochondrial Ca²⁺ uptake and release was measured. Ru360 (10 μm), a selective mitochondrial Ca²⁺ uniporter blocker that reduces [Ca²⁺]_mito (Matlib et al. 1998), decreased mean transient K_Ca current frequency and amplitude to ∼0.42 and ∼0.64 of control, respectively (Fig. 8B). In contrast, CGP37157 (10 μm), a selective mitochondrial Na⁺/Ca²⁺ exchange blocker that elevates [Ca²⁺]_mito (Cox & Matlib, 1993), increased mean transient K_Ca current frequency and amplitude ∼1.82- and ∼1.31-fold, respectively (Fig. 8A and B). In addition, activation of transient K_Ca currents by CGP37157 was reversed by Ru360 (10 μm) (Fig. 8A and B). These results show that inhibitors of mitochondrial Ca²⁺ uptake and release differentially regulate transient K_Ca currents in arterial smooth muscle cells, and suggest that mitochondrial potential regulates Ca²⁺ sparks and transient K_Ca currents by modulating intramitochondrial [Ca²⁺].

Figure 8. Differential regulation of transient K_Ca currents by Ru360 and CGP37157.

A, original recording obtained in a voltage-clamped (−40 mV) cerebral artery smooth muscle cell illustrating activation of transient K_Ca currents by CGP37157 (10 μm) and inhibition by Ru360 (10 μm). B, average relative effects on transient K_Ca current frequency (left panel) and amplitude (right panel) of Ru360 (n= 6), CGP37157 (n= 9), and Ru360 applied in the continued presence of CGP37157 (n= 5). *P < 0.05 using Students t test.

Lonidamine, a PTP opener, blocks transient K_Ca currents

In addition to reducing the driving force for mitochondrial Ca²⁺ uptake, mitochondrial depolarization leads to PTP opening (Bernardi, 1999). Conceivably, mitochondrial depolarization may inhibit Ca²⁺ sparks and transient K_Ca current due to PTP opening. To investigate this hypothesis, the regulation of transient K_Ca currents by lonidamine, a PTP opener (Ravagnan et al. 1999), was studied. At −40 mV, lonidamine (100 μm) reduced transient K_Ca current frequency and amplitude to ∼0.28 and ∼0.64 of control, respectively (Fig. 9A and B). These results suggest that PTP opening inhibits transient K_Ca currents in arterial smooth muscle cells.

Figure 9. Lonidamine inhibits transient K_Ca currents.

A, original recording illustrating inhibition of transient K_Ca currents by lonidamine (100 μm) in a cerebral artery smooth muscle cell voltage-clamped at −40 mV. B, average effects of lonidamine on transient K_Ca current frequency and amplitude (n= 10). *P < 0.05 using Students t test.

PTP blockers reduce rotenone inhibition of transient K_Ca currents

We sought to determine if mitochondrial depolarization inhibits transient K_Ca currents due to PTP opening. To investigate this hypothesis, the regulation of transient K_Ca currents by rotenone was determined in the absence and presence of the PTP blockers bongkrekic acid or cyclosporin A. PTP blockers were applied 5 min prior to rotenone application.

In the absence of a PTP blocker, a 10 min application of rotenone essentially abolished transient K_Ca currents and reduced frequency and amplitude to ∼0.04 and ∼0.42 of control, respectively (see Fig. 10B and C). Transient application (<15 min) of bongkrekic acid (10 μm) or cyclosporin A (1 μm) did not alter transient K_Ca current frequency and amplitude (Fig. 10A and B). However, 15 min after application of cyclosporin A or bongkrekic acid, transient K_Ca current frequency and amplitude were slightly reduced (Fig. 10C). In contrast to the marked inhibitory effects of rotenone when applied alone, when applied in the presence of either bongkrekic acid or cyclosporin A, rotenone (10 μm) was a less effective transient K_Ca current inhibitor (Fig. 10A–C). In the presence of bongkrekic acid, rotenone reduced transient K_Ca current frequency and amplitude to ∼0.42 and ∼0.71 of control, respectively, after 10 min (Fig. 10C). Similarly, in the presence of cyclosporin A, rotenone reduced transient K_Ca current frequency and amplitude to ∼0.43 and ∼0.62 of control, respectively (Fig. 10C). Thus, both PTP blockers reduced rotenone inhibition of transient K_Ca currents. These data suggest that mitochondrial depolarization-induced PTP opening contributes to transient K_Ca current inhibition.

Discussion

In this study, the regulation of local and global [Ca²⁺]_i signalling modalities by mitochondria was investigated in cerebral artery smooth muscle cells. The following novel findings are presented: (1) mitochondrial depolarization inhibits Ca²⁺ sparks and Ca²⁺ waves and elevates global [Ca²⁺]_i in smooth muscle cells of intact arteries; (2) in voltage-clamped smooth muscle cells, mitochondrial depolarization reduces transient K_Ca current frequency and amplitude by decreasing Ca²⁺ spark frequency, amplitude, spatial spread, and the effective coupling of Ca²⁺ sparks to K_Ca channels; (3) inhibition of transient K_Ca currents by mitochondrial depolarization does not occur via a decrease in intracellular ATP or SR Ca²⁺ load and occurs in the presence of a voltage-dependent Ca²⁺ channel blocker; (4) blockers of the mitochondrial Ca²⁺ uniporter or mitochondrial Na⁺–Ca²⁺ exchanger inhibit or activate transient K_Ca currents, respectively; (5) lonidamine, a PTP opener, blocks transient K_Ca currents; and (6) inhibition of transient K_Ca currents by mitochondrial depolarization is attenuated by PTP blockers. Collectively, these data suggest mitochondria regulate local and global Ca²⁺ signals in arterial smooth muscle cells and mitochondrial depolarization leads to PTP opening and a decrease in intramitochondrial [Ca²⁺] that inhibits Ca²⁺ sparks and transient K_Ca currents.

Mitochondrial regulation of Ca²⁺ sparks and transient K_Ca currents

Data in this study are consistent with the prior observation that CCCP rapidly blocked spontaneous transient inward currents (STICs) in rabbit portal vein myocytes (Greenwood et al. 1997). A STIC occurs due to the activation of a number of Ca²⁺-activated Cl⁻ channels by a Ca²⁺ spark (Jaggar et al. 2000). In contrast, cyanide, an electron transport chain complex IV blocker, activated Ca²⁺ sparks and STICs in pulmonary artery smooth muscle cells (Wang et al. 2003). The contrasting results obtained in systemic and pulmonary artery smooth muscle cells may occur due to different mitochondrial physiology in these preparations (e.g. see Michelakis et al. 2002), or the effects of inhibiting different electron transport chain complexes on Ca²⁺ signalling.

In the present study, mitochondrial depolarization reduced Ca²⁺ spark frequency, amplitude, and spatial spread, and accelerated spark decay. Collectively, these findings suggest that mitochondrial depolarization reduces RyR channel activity. These results also suggest mitochondrial Ca²⁺ uptake does not directly contribute to spark decay, consistent with previous evidence that sparks decay primarily due to diffusion (Jaggar et al. 2000; Cheranov & Jaggar, 2002). Mitochondrial depolarization decreased transient K_Ca current amplitude via two mechanisms: (1) a reduction in Ca²⁺ spark amplitude and spatial spread, and (2) a decrease in the effective coupling of Ca²⁺ sparks to K_Ca channels. Ca²⁺ sparks of smaller amplitude and spatial spread will have a lower impact on K_Ca channel activity and activate fewer channels, and therefore induce smaller amplitude transient K_Ca currents. Since K_Ca channels are typically exposed to micromolar [Ca²⁺]_i during a Ca²⁺ spark (Perez et al. 2001; ZhuGe et al. 2002), mitochondrial depolarization presumably reduces effective coupling to Ca²⁺ sparks by decreasing the micromolar Ca²⁺-sensitivity of K_Ca channels. In the absence of Ca²⁺ sparks, rotenone and CCCP activated K_Ca channels, presumably by inducing a voltage-independent nanomolar elevation in global [Ca²⁺]_i.

Mitochondrial depolarization leads to opening of the PTP, a non-selective, high conductance complex comprised of several components including the voltage-dependent anion channel, the adenine nucleotide translocase (ANT) and cyclophilin D (Duchen, 2000). PTP opening leads to the release of mitochondrial factors less than 1.5 kDa, including Ca²⁺ and cytochrome c (Bernardi, 1999). Bongkrekic acid blocks the ANT and stabilizes the closed state of the PTP. Cyclosporin A inhibits the PTP by binding to cyclophilin D (Bernardi, 1999), but also inhibits calcineurin (protein phosphatase 2B) (Bernardi, 1999; Duchen, 2000). Acute application (<15 min) of bongkrekic acid and cyclosporin A did not inhibit transient K_Ca currents, but prolonged application (>15 min) decreased the frequency and amplitude of these events. One explanation for this inhibitory effect is that PTP blockers ultimately decrease cytosolic ATP, particularly since the time course of block was similar to that with oligomycin B.

In arterial smooth muscle cells, mitochondrial depolarization elevates ROS (Michelakis et al. 2002). Rather than inhibiting transient K_Ca currents, exogenous application of H₂O₂ activated these events, similarly to effects in cat tracheal smooth muscle cells (Bauer et al. 1997). H₂O₂-activation of transient K_Ca currents was presumably due to RyR channel and/or K_Ca channel activation (Boraso & Williams, 1994; Barlow & White, 1998). However, oxidative stress also leads to PTP opening (e.g. see Teshima et al. 2003), raising the question of why H₂O₂ did not block transient K_Ca currents? One explanation is that in cardiac myocytes, significant mitochondrial depolarization does not occur until tens of minutes after H₂O₂ exposure (Teshima et al. 2003). Conceivably, prolonged exposure to H₂O₂ that leads to PTP opening may inhibit Ca²⁺ sparks and transient K_Ca currents in arterial smooth muscle cells. Regardless, 5 min after rotenone or CCCP application transient K_Ca currents were blocked, whereas 5 min after H₂O₂ application transient K_Ca currents were activated, arguing against the possibility that mitochondrial depolarization blocked transient K_Ca currents due to an increase in ROS.

Mitochondrial regulation of Ca²⁺ waves and global Ca²⁺

Mitochondrial depolarization blocked spontaneous Ca²⁺ waves, which occur due to RyR channel activation under the conditions used in our experiments (Jaggar et al. 1998). In colonic smooth muscle cells, CCCP also blocked IP₃-evoked [Ca²⁺]_i transients (McCarron & Muir, 1999). However, in tail artery smooth muscle cells electron transport chain blockers increased the frequency, but reduced the amplitude of agonist-induced Ca²⁺ waves (Sward et al. 2002). Collectively, these studies indicate that mitochondrial depolarization modifies Ca²⁺ waves in smooth muscle, but the spatial and temporal changes that occur may depend on the relative contribution of RyR channels and IP₃-gated Ca²⁺ release channels to the events.

Mitochondrial depolarization elevated global [Ca²⁺]_i in agreement with previous studies performed in vascular and non-vascular smooth muscle cells (Taggart et al. 1997; Kamishima et al. 2000; Kamishima & Quayle, 2002; Wang et al. 2003). Data in the present study suggest that mitochondrial depolarization elevates global [Ca²⁺]_i via two mechanisms: (1) a voltage-independent pathway, and (2) via inhibition of Ca²⁺ sparks and transient K_Ca currents, which would lead to membrane depolarization, activation of voltage-dependent channels, and Ca²⁺ influx (Jaggar et al. 2000). Mechanisms by which mitochondrial depolarization leads to a voltage-independent elevation in global [Ca²⁺]_i are unclear, but may involve the elimination of a mitochondrial Ca²⁺ sink. In addition, mitochondrial depolarization may release or generate mitochondrial mediators that stimulate Ca²⁺ influx and/or release and/or inhibit Ca²⁺ removal. In pulmonary and mesenteric artery smooth muscle cells, metabolic inhibition with cyanide elevates global [Ca²⁺]_i by stimulating Ca²⁺ release from the SR and mitochondria (Wang et al. 2003). In cerebral artery smooth muscle cells, mitochondrial depolarization inhibited Ca²⁺ sparks and Ca²⁺ waves, suggesting that global [Ca²⁺]_i did not increase due to RyR channel activation.

Transient and sustained mitochondrial depolarizations may regulate Ca²⁺ signalling in arterial smooth muscle

In cardiac muscle, local SR Ca²⁺ release leads to a transient elevation in intramitochondrial Ca²⁺ concentration termed a ‘Ca²⁺ mark’ (Pacher et al. 2002) and a transient mitochondrial depolarization termed a ‘flicker’ (Duchen et al. 1998). Flickers have been attributed to a number of mechanisms, including PTP opening (Huser & Blatter, 1999), Ca²⁺ influx via the Ca²⁺ uniporter (Duchen et al. 1998), and H⁺ entry via the F₁F₀-ATPase (Buckman & Reynolds, 2001). Recently, flickers have been described in Bufo marinus stomach smooth muscle cells, where mitochondria flicker independently of one another (O'Reilly et al. 2003). In smooth muscle cells, individual flickers last for up to 30 s and range from less than 10 mV to greater than 100 mV (O'Reilly et al. 2003). Conceivably, the regulation of RyR channel activity by mitochondrial flickers may be an ongoing feedback process that continually regulates Ca²⁺ spark generation via mechanisms described here, i.e. via intramitochondrial [Ca²⁺]. In a recent study that measured the spatial arrangement of the SR and mitochondria within venous smooth muscle cells, TMRE fluorescence could not be observed within the vicinity of a Ca²⁺ spark release site (Singh et al. 2003). The authors suggested that local Ca²⁺ released from the SR may be sequestered into mitochondria, leading to depolarization and a loss of TMRE fluorescence at these sites (Singh et al. 2003). Conceivably, in arterial smooth muscle cells mitochondrial Ca²⁺ uptake during a spark could lead to depolarization and the initiation of a localized negative-feedback mechanism that reduces Ca²⁺ spark occurrence.

Several stimuli may regulate local and global Ca²⁺ signalling by inducing sustained mitochondrial depolarization. For example, mitochondria may act as an oxygen sensor (Waypa et al. 2001; Michelakis et al. 2002). In systemic artery smooth muscle cells, hypoxia depolarizes mitochondria (Michelakis et al. 2002). Similarly, vasoconstrictors may regulate smooth muscle Ca²⁺ signalling, in part, by modulating mitochondrial potential. Vasoconstrictor-stimulated Ca²⁺ release elevates mitochondrial [Ca²⁺] in smooth muscle cells (Monteith & Blaustein, 1999; Szado et al. 2003), which would lead to mitochondrial depolarization (Bernardi, 1999; Duchen, 2000). Vasoconstrictors also inhibit Ca²⁺ sparks (Jaggar & Nelson, 2000; Mauban et al. 2001). Conceivably, a vasoconstrictor-induced mitochondrial depolarization may contribute to Ca²⁺ spark inhibition by the mechanisms we describe in this study. Thus, mitochondria may regulate Ca²⁺ signalling and arterial diameter under resting conditions and in response to diverse stimuli.

Mechanisms by which mitochondrial depolarization may inhibit Ca²⁺ sparks and transient K_Ca currents

Mitochondria regulate the activity of both IP₃-gated Ca²⁺ release channels and RyR channels, but regulation appears to be cell specific. For example, mitochondrial depolarization accelerates IP₃-mediated Ca²⁺ waves in cortical astrocytes (Duchen, 2000), but blocks IP₃-evoked Ca²⁺ release in colonic smooth muscle cells (McCarron & Muir, 1999). In cardiac and skeletal muscle, inhibition of mitochondrial Ca²⁺ uptake stimulates Ca²⁺ sparks, suggesting that mitochondria normally suppress RyR channel activity in these tissues (Pacher et al. 2002; Isaeva & Shirokova, 2003). Mechanisms by which mitochondrial depolarization inhibits RyR and K_Ca channels in arterial smooth muscle cells remain to be established, but both local and global regulatory mechanisms must be considered. Mitochondria are spatially located within 20 nm of the ER/SR in many cell types, including arterial smooth muscle cells (Rizzuto et al. 1999; Duchen, 2000; Szado et al. 2003). Although slight differences occur between subtypes, RyR channels typically exhibit a bell-shaped Ca²⁺ dependence and are activated by 1–10 μm Ca²⁺ and inhibited by 1–10 mm Ca²⁺ (Fill & Copello, 2002). Due to the bell-shaped Ca²⁺ dependence of RyR channels, an increase or a decrease in [Ca²⁺]_i can activate or inhibit Ca²⁺ sparks depending on the concentration that is reached. Conceivably, in smooth muscle cells mitochondria may normally buffer, and thus lower, cytosolic Ca²⁺ in the immediate vicinity of RyR channels. Mitochondrial depolarization may remove this localized buffering, leading to high local [Ca²⁺]_i, RyR channel inhibition and a decrease in Ca²⁺ spark occurrence. Alternatively, mitochondria may normally stimulate Ca²⁺ sparks by releasing Ca²⁺ within the vicinity of RyR channels. Mitochondrial depolarization would remove this stimulatory mechanism, which would also decrease RyR channel activity and inhibit Ca²⁺ sparks. Whether mitochondrial depolarization elevates or reduces the local cytosolic [Ca²⁺]_i in the vicinity of RyR channels in smooth muscle cells remains to be determined. If mitochondria stimulate Ca²⁺ sparks by releasing Ca²⁺ within the vicinity of RyR channels, data suggest that this does not occur due to Ca²⁺ released via the PTP or the Na⁺–Ca²⁺ exchanger. Acute application of PTP blockers did not alter transient K_Ca currents and CGP37157 activated transient K_Ca currents. However, data with CGP37157 support a localized regulatory mechanism. CGP37157 elevates intramitochondrial [Ca²⁺] (Cox & Matlib, 1993), which would reduce the buffering capacity of mitochondria and increase the driving force for mitochondrial Ca²⁺ efflux. Both of these effects may elevate local cytosolic Ca²⁺ in the vicinity of RyR channels.

Mitochondrial depolarization not only reduces the driving force for Ca²⁺ uptake, but leads to PTP opening and mitochondrial Ca²⁺ efflux (Gunter & Pfeiffer, 1990; Bernardi, 1999; Duchen, 2000). PTP blockers may reduce depolarization-induced inhibition of transient K_Ca currents by blocking this mitochondrial Ca²⁺ efflux pathway and attenuating the loss of intramitochondrial Ca²⁺. However, PTP opening may also release intramitochondrial components that inhibit RyR and/or K_Ca channels. One potential RyR/K_Ca channel inhibitor released upon PTP opening is cytochrome c (Bernardi, 1999). Cytochrome c blocks Ca²⁺-dependent inhibition of IP₃-gated Ca²⁺ release channels (Boehning et al. 2003) and activates voltage-dependent K⁺ channels in pulmonary artery smooth muscle cells (Platoshyn et al. 2002). Thus mitochondrial depolarization may inhibit RyR and K_Ca channels via local and global mechanisms.

Physiological relevance of mitochondrial regulation of local and global Ca²⁺ signalling in arterial smooth muscle cells

Mitochondria generate metabolic energy, directly and indirectly regulate Ca²⁺ signalling (Rizzuto et al. 1999), play a central role in the survival and programmed death of eukaryotic cells (Gunter et al. 2000; Duchen, 2000), and function as an oxygen sensor (Michelakis et al. 2002; Waypa et al. 2002). Therefore, changes in mitochondrial potential that regulate local and global Ca²⁺ signalling in arterial smooth muscle cells may be important under both physiological and pathophysiological conditions and directly and indirectly regulate diverse cellular functions, including contractility and apoptosis.

In summary, our data indicate that mitochondria regulate local and global Ca²⁺ signals in cerebral artery smooth muscle cells. The data suggest that mitochondrial depolarization inhibits Ca²⁺ sparks and transient K_Ca currents via PTP opening and a reduction in intramitochondrial [Ca²⁺]. These findings suggest that changes in mitochondrial potential may regulate a range of Ca²⁺-dependent cellular functions by modulating local and global Ca²⁺ signalling modalities.