Modulation of intracellular chloride channels by ATP and Mg2+

Biochimica et Biophysica Acta 1797 (2010) 1300–1312

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / b b a b i o

Modulation of intracellular chloride channels by ATP and Mg2+ Viera Kominkova a, Lubica Malekova a, Zuzana Tomaskova a, Peter Slezak b, Adam Szewczyk c, Karol Ondrias a,⁎ a b c

Institute of Molecular Physiology and Genetics, Centre of Excellence for Cardiovascular Research, Slovak Academy of Sciences, 83334 Bratislava, Slovakia Institute of Normal and Pathological Physiology, Centre of Excellence for Cardiovascular Research, Slovak Academy of Sciences, 83334 Bratislava, Slovakia Nencki Institute of Experimental Biology, 02-093 Warsaw, Poland

a r t i c l e

i n f o

Article history: Received 16 October 2009 Received in revised form 3 February 2010 Accepted 26 February 2010 Available online 4 March 2010 Keywords: Mitochondrial membrane Lysosomes Mitochondrial chloride channel ATP Mg2+ Bilayer lipid membrane

a b s t r a c t We report the effects of ATP and Mg2+ on the activity of intracellular chloride channels. Mitochondrial and lysosomal membrane vesicles isolated from rat hearts were incorporated into bilayer lipid membranes, and single chloride channel currents were measured. The observed chloride channels (n = 112) possessed a wide variation in single channel parameters and sensitivities to ATP. ATP (0.5–2 mmol/l) modulated and/or inhibited the chloride channel activities (n = 38/112) in a concentration-dependent manner. The inhibition effect was irreversible (n = 5/93) or reversible (n = 15/93). The non-hydrolysable ATP analogue AMP-PNP had a similar inhibition effect as ATP, indicating that phosphorylation did not play a role in the ATP inhibition effect. ATP modulated the gating properties of the channels (n = 6/93), decreased the channels' open dwell times and increased the gating transition rates. ATP (0.5–2 mmol/l) without the presence of Mg2+ decreased the chloride channel current (n = 12/14), whereas Mg2+ significantly reversed the effect (n = 4/4). We suggest that ATP-intracellular chloride channel interactions and Mg2+ modulation of these interactions may regulate different physiological and pathological processes. © 2010 Elsevier B.V. All rights reserved.

1. Introduction The production of ATP through oxidative phosphorylation requires a large electrochemical driving force for protons across the mitochondrial inner membrane. This implies that the mitochondrial inner membrane should be generally impermeable to ions other than H+. However, it has become clear in recent years that a variety of selective and non-selective ion channels are present in the outer or inner mitochondrial membranes [1–4]. Mitochondrial channels are involved in apoptosis. The Bcl-2 family of proteins regulates apoptosis by controlling the release of mitochondrial cytochrome c via the Bax/Bak channel. The activation of Bax and Bak upon the induction of apoptosis involves their oligomerisation (Bax oligomers up to hexamers), integration into the mitochondrial membrane (Bax) and formation of non-selective channels/lipidic pores or channels poorly selective for chloride [5– 9]. A mitochondrial apoptosis-induced channel (MAC) was found in proteoliposomes prepared from mitochondrial outer membranes that Abbreviations: BLM, bilayer lipid membrane; P-open, open probability of single channel; Erev, reversal potential; AMP-PNP, adenosine 5′-(β,γ-imido) triphosphate; DIDS, 4, 4′-diisothiocyanatostilbene-2, 2′-disulfonic acid disodium salt; PTP, permeability transition pore ⁎ Corresponding author. Institute of Molecular Physiology and Genetics, Slovak Academy of Sciences, Vlarska 5, 833 34 Bratislava, Slovakia. Tel.: + 421 2 54774102; fax.: + 421 2 54773666. E-mail address: [email protected] (K. Ondrias). 0005-2728/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.bbabio.2010.02.031

were isolated from cells undergoing apoptosis [10]. This channel had high conductance and multiple substates [11]. The Bcl-2 protein family regulates apoptosis in part by controlling the formation of MAC, which is a putative cytochrome c release channel induced early in the intrinsic apoptotic pathway. MAC is preferentially associated with oligomeric Bax. Both MAC and Bax have similar single channel activities modified by cytochrome c [12]. The voltage-dependent anion channel (VDAC), with homology to bacterial porins, forms an outer mitochondrial membrane pore with a diameter of 2.5–3.0 nm. The VDAC channel is partially anion-selective, allowing the passage of ions and metabolites [1]. It plays a role in apoptotic mitochondrial membrane permeabilisation by interacting with Bcl-2 family members and is required for the pro-apoptotic activity of Bax in the absence of Bak [13]. Bax and Bak have been shown to interact with permeability transition pore (PTP), VDAC1, VDAC2 or adeninenucleotide translocator [14]. There have been reports of peripheral-type benzodiazepine receptor localisation to the mitochondrial outer membrane and anion channels in heart mitochondria mitoplasts blocked by ligands of mitochondrial benzodiazepine receptor [15,16]. A translocase of the outer membrane of mitochondria (TOM) and its partner on the inner membrane, TIM, can form large conductance channels [17]. Several calcium, potassium and chloride channels have been observed in inner mitochondrial membranes [1–4,18,19]. Mitochondrial Ca2+ uniporter [20], Ca2+ uptake (RaM) [21], PTP and calcium channelryanodine receptors (RyR1) may participate in Ca2+ mitochondrial

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

influx and/or efflux [22]. ATP (mitoKATP) and calcium-regulated potassium channels, identified in the inner mitochondrial membrane, have cytoprotective properties [3,23,24], in which regulation of the mitochondrial ROS production might be involved [25]. Studies of mitochondrial swelling have revealed that an inner mitochondrial membrane anion channel (IMAC) is active under certain conditions. IMAC fluxes of a wide variety of anions, ranging from small singly charged ions, such as Cl−, to multi-charged anions, such as citrate, superoxide anion or ATP, have been observed [26,27]. Using patch-clamp and bilayer lipid membrane methods, anionselective channels having several single channel properties have been observed in the inner mitochondrial membrane [2,4,16,28–42]. The most-studied channel is an anion-selective 108 pS channel, which has been proposed to correspond to IMAC [34,35]. Several anion channels having different single channels properties and modulation have also been observed [29,33,36–39]. Recently, De Marchi et al. [40] characterised a “maxi” mtCl channel in the inner membrane of mitochondria of a colon tumour cell line. The channel was voltage-dependent, with a conductance of ∼ 400 pS in the symmetrical 150 mmol/l KCl; this is almost half of the conductance of the PTP. The first intracellular chloride channel localised to the mitochondria (CLIC4, or p64H1), incorporated in the BLM, had low conductances and was poorly selective for Cl−/K+ [43]. Some mitochondrial transport proteins, such as the phosphate carrier protein [44], the uncoupling protein from brown adipose tissue mitochondria [45] or the pro-apoptotic Bax, have been reported to form channels selective for chloride [46]. Despite numerous reports of mitochondrial chloride channel properties, their molecular origins and functional significances are still not fully understood. It has been suggested that inner mitochondrial membrane anion channels are superoxide-permeable and are

1301

responsible for the extrusion of superoxide anions from the mitochondrial matrix [47–49]. Changes in ΔΨm, repetitive transient depolarisations of a single mitochondrion, have been observed in several types of cells [47,50]. A role of IMAC in the ΔΨm oscillations is supported by observations that inhibitors of IMAC inhibited mitochondrial ΔΨm oscillations [47,49]. As a result of decreased blood flow, cells in ischemic tissues are depleted of oxygen and metabolic nutrients and lose their capacity for ATP production. Due to energy deprivation, destructive processes are activated, culminating in cell injury and death [51,52]. Cell injuries resulting from ATP depletion are mediated by multiple factors, involving Ca2+ overload, the generation of free radicals and the activation of hydrolytic enzymes [53,54]. However, the roles of mitochondrial or other intracellular chloride channels in these processes are not fully understood. ATP has been reported to influence the function of anion (chloride) channels. Ahern and Laver [55] described an ATP-dependent inhibition of the 75–105 pS anion channel from the sarcoplasmic reticulum of rabbit skeletal muscle. Ballarin and Sorgato [33] observed an anion channel from yeast mitochondria that was blocked with ATP. An ATP-induced inhibition of a plasma membrane CIC-1 chloride channel was shown to be controlled by oxidation and reduction [56,57]. Concentrations of Mg2+ and/or ATP may change or decrease during several physiological and pathological processes, including myocardial infarction and ischemia. However, the ATP–chloride channel–Mg2+ interactions in these processes, particularly with the involvement of heart intracellular chloride channels, are not fully understood. Therefore, in the present work, we investigated the influence of ATP and Mg2+ on the single channel properties of mitochondrial and lysosomal chloride channels.

Fig. 1. Representative traces from three experiments of the regular (A) and ragged (B) chloride single channels. Effect of pH on the regular (C) and ragged (D) chloride channels. The voltage was 0 mV. The lines on the left mark the closed state of the channels.

1302

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

2. Materials and methods 2.1. Chemicals Lipids were from Avanti Polar Lipids (Alabaster, AL, USA). Protease inhibitors were from Roche Diagnostics GmbH (Mannheim, Germany). All other chemicals were purchased from Sigma-Aldrich (Germany). Stock solutions of ATP-Na2 or ADP–Na salt (100 mmol/l) were prepared fresh each day and buffered with Tris (200 mmol/l) to prevent a change of pH in the solutions after the ATP or ADP application. Simultaneously with the application of ATP or ADP, appropriate amounts of CaCl2 and MgCl2 were added to restore constant concentrations of free Ca2+ and Mg2+. The free Ca2+, Mg2+ and ATP concentrations were calculated using the Winmaxc32 program. A 100-mmol/l stock solution of DIDS was made in DMSO. The final concentration of DMSO in solution was b0.1%. 2.2. Isolation of submitochondrial particles (membrane vesicles) from crude and purified mitochondria Crude preparation of rat heart mitochondria and submitochondrial particles (inner membrane vesicles) from crude mitochondria were isolated from the hearts of male Wistar rats [58–60] as described in

our previous study [37]. In brief, heart ventricles were immersed in the ice cold isolation buffer (in mmol/l: 50 sucrose, 200 mannitol, 5 KH2PO4, 1 EGTA and 5 MOPS, plus 0.2% BSA (pH 7.3 adjusted with KOH)); this was homogenised and centrifuged at 750 ×g for 15 min, the supernatant was centrifuged at 7000 ×g for 10 min and the resulting pellet was resuspended in the isolation buffer (in further steps without EGTA and BSA) and immediately used for preparation of inner membrane vesicles. The mitochondrial suspension was sonicated 8× for 15 s at 35 kHz on ice and then centrifuged at 10,000 ×g for 10 min. The obtained supernatant was again centrifuged at 100,000 ×g for 30 min. The final membrane pellet was resuspended in a EGTA- and BSA-free isolation buffer, frozen in liquid N2 and stored at −70 °C in small aliquots until use for reconstitution into a bilayer lipid membrane (BLM). All procedures were performed at 4 °C, and the buffers contained the protease inhibitors (in µmol/l) 1 leupeptin, 1 pepstatin, 1 aprotinin, 1000 benzamidine and 200 Pefabloc SC. Since the crude mitochondria contained a significant fraction of lysosomal membranes [38], we assume that the vesicles isolated from the crude mitochondria were derived from mitochondrial and lysosomal membranes. Percoll-purified rat heart mitochondria and submitochondrial particles (membrane vesicles) from the purified mitochondria were isolated from the crude mitochondria as described in our previous

Fig. 2. Three opening levels of chloride current of the ragged channels (two experiments, a and b) at three different time resolutions (A). The lines on the left mark the closed state of the channels. Histograms of the number of points (B) and observations (C) obtained from experiment a. The voltage was 0 mV. The “number of points” histogram shows an amplitude histogram of every digitalised data point in the data file. The “number of observations” histogram shows an amplitude histogram of every channel in the data file, where the number of observations means the number of channels.

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

1303

Fig. 3. Correlations between the reversal potential and conductance (A), the reversal potential and the single channel amplitude at 0 mV (B) and the open probability (P-open) and the channel conductance (C) as well as the relationship between the open dwell time τ2 and the open probability (P-open) (D) of the regular chloride channel. The data (D) were fitted by a modified hyperbolic function, f = a − b / (1 + c ⁎ x) ^ (1/d), a = 0.956, b = 1.34, c = 0.127, d = 0.466.

study [38]. All procedures were approved by the State Veterinary and Nutritive Administration of Slovak Republic. 2.3. Bilayer lipid membrane measurement BLM was formed, and the single chloride current was measured using a BC-525C amplifier (Warner Instrument, Hamden, CT, USA), filtered at 1000 Hz and sampled at 4000 Hz, as described in [37]. The composition of the solutions, when vesicles were isolated from the crude mitochondria, were (buffer A in mmol/l): trans chamber: 50 KCl, 2 MgCl2, 0.4 CaCl2, 1 EGTA and 10/5 Hepes/Tris, 7.4 pH; and cis chamber: 250 KCl, 2 MgCl2, 0.1 CaCl2 and 10/5 Hepes/Tris, 7.4 pH. The composition of the solutions, when vesicles were isolated from the Percoll-purified mitochondria, were (buffer B in mmol/l): trans chamber: 50 KCl, 0.1 CaCl2, 0.3 EGTA and 10/5 Hepes/Tris, 7.4 pH; and cis chamber: 250 KCl, 0.1 CaCl2, 0.3 EGTA and 10/5 Hepes/Tris, 7.4 pH. The membrane vesicles were added to the cis chamber. A KCl gradient from 600 to 800 mmol/l cis versus 50 mmol/l trans was used to facilitate the fusion of the vesicles into the BLM. After the fusion of the vesicle, the osmotic gradient and excessive proteins were removed by perfusion of the cis chamber. All of the voltages reported here refer to the trans side, and the cis side was grounded. Under our conditions, the positive current that increased with the application of positive voltages meant a flux of chloride anions from the cis to the trans side. ATP, ADP, adenosine 5′-(β,γ-imido) triphosphate tetralithium salt (AMP-PNP) or 4, 4′-diisothiocyanatostilbene-2, 2′-disulfonic acid disodium salt (DIDS) were added to the cis and/or trans chamber,

followed by continuous stirring for 1 min. The channel conductance and the reversal potential (Erev) were calculated from the current– voltage relationship within the voltage range of ±30 mV. The timedependence of the single channel open probability (P-open) was determined as the average of the 10-s records for 3 min and calculated from the ratio of the (open time)/(total time) intervals. The open dwell time was fitted by a two-order exponential function, giving the dwell times τ1 and τ2. A rough relative ionic selectivity Cl−/K+ was calculated from the reversal potential Erev, neglecting the permeabilities of the other ions, according to the modified GHK equation [61]. Channels having high non-regular conductances, non-stable baselines and promiscuous channels were excluded from our experiments [41]. All procedures were carried out at 22–24 °C. Results from independent experiments are expressed as means ± standard deviations (SD), except for the open probability (P-open) and the open dwell time τi, which are expressed as medians and lower (Q1) and upper (Q3) quartiles (Q 1 –Q 3 ); n refers to the number of independent experiments. 3. Results 3.1. Chloride channels derived from crude mitochondria vesicles After the incorporation of vesicles into the BLM, we observed and evaluated the chloride channels in 93 experiments. Some of the channel behaviour appeared to be burst-like because periods of silence separating periods of activity also occurred. Since the channels had various single channel properties, they were divided into two

1304

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

main groups, which we named the regular chloride channels (observed in 71/93 experiments) and the ragged chloride channels (observed in 22/93 experiments). We defined the regular chloride channels as the channels having stable and constant open and closed levels and chloride currents in the voltage range from −20 to +20 mV. We defined the ragged chloride channels as the high flickering channels with stable baselines but not having stable and constant open levels and chloride currents in the voltage range from −30 to + 30 mV. The comparison of the single channel currents of the regular and ragged chloride channels is shown in Fig. 1. We observed that the regular channels were blocked by the application of HCl to the cis side (n = 10/11), which decreased the pH from 7.4 to 5.8 (Fig. 1C). On the other hand, the non-regular channels were not blocked by decreasing the pH from 7.4 to 5.8 (n = 2/2) (Fig. 1D). The ragged channels had one to three current opening levels. Two or three opening levels of the ragged channels and current substates were observed in 12/22 experiments, and 1 opening level was seen in 6/22 experiments. In 4/22 experiments the substates were not observed. An example of 2 experiments of the opening of the ragged channels into three levels and at 3 resolutions is shown in Fig. 2. The 3 open levels were clearly seen at the low time resolution; however, at the high resolution, non-constant open levels were observed (Fig. 2A). From the experimental data of the “a” experiment (Fig. 2A), we evaluated points and observations histograms (Fig. 2B and C). The 3 levels of openings are evident from the histograms. The single channel properties of the regular channels were evaluated from 57 experiments. From the I–V relationship (at the voltages ±20 mV or ±30 mV), the conductance and Erev were calculated. The mean conductance was G = 117 ± 31 pS, and the reversal potential was Erev = − 20 ± 5 mV (n = 57). However, the ranges of G and Erev were wide: G ranged from 62 pS to 220 pS, and Erev ranged from −10 mV to −32 mV. There was no correlation between Erev and the channel conductance (r = 0.253, P = 0.0618, Fig. 3A). The single channel currents, measured at 0 mV, were mostly in the range of 1.3–3.6 pA, with mean i0 = 2.3 ± 0.8 pA (Fig. 3B). There was no correlation between the chloride current and the channel conductance (data not shown), but a correlation was observed between the Erev and the channel amplitude (r = −0.741, P b 0.0001, Fig. 3B). The median P-open of the regular channels was 0.893 (0.8252–0.931). There was no correlation between the P-open and the channel conductance (r = 0.247, P = 0.1089, Fig. 3C). The open dwell time of the regular channels also had a heterogeneous distribution. The dwell time distribution of the single channel was fitted by the second-order exponential function, giving two characteristic open dwell times of τ1 and τ2. The channels had different open dwell time distributions. The median open dwell times, τ1 and τ2, of the regular chloride channels were 2 ms (1.35–5) and 23 ms (17–39), respectively. There was no correlation between the mean open time τ1 and the channel conductance (data not shown), but there was a hyperbolic relationship between τ2 and P-open (Fig. 3D). The voltage dependence of the single chloride current and Popen of regular channels at symmetric 250 mmol/l KCl showed slight rectification at b−40 mV, but P-open decreased significantly with negative voltages (Fig. 4). The single channel properties of the ragged channels obtained from 6/93 experiments, in which only one chloride channel was observed, were also heterogeneous. These chloride channels, in comparison to the regular channels, had lower P-open, conductance and Erev and shorter open dwell times (Table 1). 3.1.1. ATP effects on chloride channels We observed four different effects of ATP on the chloride channels: (1) it reversibly or (2) irreversibly blocked the channels, (3) it modulated their gating kinetics or (4) it did not have an effect. An example of the reversible blocking effect of ATP is shown in Fig. 5. ATP (0.5 mmol/l) inhibited the channel activity (Fig. 5A). The inhibitory

Fig. 4. (A) Voltage dependence of single chloride current at asymmetric 250/50 mmol/ l KCl (circles) and symmetric 250 mmol/l KCl (triangles; 173 pS), and (B) voltage dependence of P-open at symmetric 250 mmol/l KCl (n = 2, mean ± SD).

effect of ATP on P-open was time-dependent (Fig. 5B). ATP (0.52 mmol/l) inhibited the single channel activity of regular (n = 13/71) and ragged chloride channels (n = 11/22) (Fig. 5C). The mean conductance of the inhibited channels was 114 ± 44 pS, Erev = −19 ± 3 mV (n = 24/93) and the calculated average permeability ratio was Cl−/K+ = 3.3. The ATP inhibition effect was reversible in 7/10 of the regular and 8/10 of the ragged chloride channels. The single channel properties of the reversibly (n = 15/93) and irreversibly (5/93) inhibited chloride channels varied and were comparable with the chloride channels not affected by ATP (Figs. 3A and 5C). ATP (0.5–2 mmol/l) decreased the single channel amplitude by 10–30% (data not shown). The side dependence of the ATP reversible inhibition effects was studied on the same regular chloride channels. ATP (0.5–2 mmol/l) inhibited the channels from the one side of BLM (cis) and not from the other side in 5/5 experiments. From this, we assumed that most of the channels

Table 1 Single channel properties of ragged chloride channels.

1 2 3 4 5 6

G [pS]

Erev [mV]

I0 mV [pA]

P-open

τ1 [ms]

τ2 [ms]

99 93 88 105 95 92

− 16 − 19 − 17 − 19 − 28 − 20

1.7 2.0 1.3 2.2 2.7 1.7

0.218 0.614 0.491 0.428 0.581 0.285

1.0 1.9 0.6 1.1 1.5 1.8

3.6 7.7 2.6 3.9 5.4 4.0

G — conductance, Erev — reversal potential obtained from I–V curve within the ± 20 mV, I0 mV — single chloride channel current measured at 0 mV, P-open — open probability of ragged chloride channels, τ1 and τ2 — open dwell times of ragged chloride channels fitted by two-order exponential function.

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

1305

Fig. 5. Effect of ATP on the single chloride channel current. The application of ATP (0.5 mmol/l) to the cis side of BLM was followed by a cis reperfusion. The voltage was 0 mV. The lines on the left mark the closed state of the channels (A). Time-dependence of the inhibitory effect of ATP (0.5 mmol/l) on the single chloride channel open probability (P-open) (B). Correlation between the reversal potential and the conductance of the channels influenced by ATP (C); regular (full circles) and ragged (open circles) chloride channels inhibited by ATP, and the chloride channels for which the gating dynamics was influenced by ATP (triangles).

were oriented in the same direction in the BLM. Because the concentration-dependent inhibitory effect of ATP varied significantly among different chloride channels, the concentration-dependence curve of inhibition was very scattered. Therefore, we use the concept that the channels were modulated or inhibited within a range of ATP concentrations. In order to investigate whether phosphorylation of the chloride channels was involved in the effect of ATP, a non-hydrolysable ATP, AMP-PNP, was used. During the procedure, after the inhibitory effect of ATP, the ATP was removed by perfusion of the cis solution and AMPPNP was applied on the same chloride channel. In some experiments, after the inhibitory effect of AMP-PNP, the AMP-PNP was removed by perfusion and ATP was applied. An example is shown in Fig. 6. The application of 0.5–1 mmol/l AMP-PNP reversibly inhibited the chloride channel. After the AMP-PNP was removed by reperfusion, a subsequent application of 0.5–1 mmol/l ATP to the same channel caused its reversible inhibition. In 5/8 experiments, neither ATP nor AMP-PNP influenced the activity of the chloride channels. However, in 3/8 experiments, ATP and AMP-PNP reversibly inhibited the same chloride channels. In order to investigate whether the reversible inhibitory effect of ATP is specific, the effect of ADP on the same chloride channels was investigated in comparison. During the procedure, after observing the inhibitory effect of ATP, the ATP was removed by perfusion of the cis solution and ADP was applied on the same chloride channel. In 2/2 experiments, an application of 0.5–2 mmol/l ATP reversibly inhibited the regular chloride channel, but subsequent application of ADP to the

same channel did not inhibit it (Fig. 7). ATP reversibly decreased P-open, but ADP did not (Fig. 7B and C). ATP slightly decreased the single channel amplitude, whereas ADP had a pronounced effect on it (Fig. 7A). ADP (2 mmol/l) decreased the single channel amplitude to 36.0 ± 2.5% (n = 2). In 2/2 experiments, ATP blocked ragged chloride channels, and subsequent application of ADP had the same inhibitory effect as ATP (data not shown). In 6/93 experiments, ATP (0.5–2 mmol/l) increased the gating transition rate of the chloride channels. Fig. 8A shows representative single channel current traces before and after the application of ATP and 100 µmol/l DIDS to the cis side of the BLM. The chloride channel blocker DIDS blocked the channel influenced by ATP. The addition of 2 mmol/l ATP to the cis side decreased the channel open dwell time, τ2, to 12 (10–18) ms but did not affect the τ1 value (Fig. 8B). The Popen of the chloride channels (with changed kinetics) was in the range of 0.85–0.95 under control conditions, and it slightly decreased by 11 ± 5% after the application of 2 mmol/l ATP (data not shown). The application of 2 mmol/l ATP decreased the single channel current by 20% (Fig. 8C). Those channels that were dynamically influenced by ATP had a mean conductance of 120 ± 11 pS, a single channel chloride current of 3.7 ± 0.4 pA at 0 mV, Erev = −31 ± 2 mV (Fig. 5C) and a calculated average permeability ratio of Cl−/K+ = 9.5. The Erev values and the permeability ratio, Cl−/K+, for the channels dynamically modified by ATP were significantly higher than for the channels inhibited by ATP. DIDS, an inhibitor of chloride channels, was applied on the channels that were reversibly blocked and dynamically modified by ATP, and this

1306

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

Fig. 6. Effects of AMP-PNP and ATP on the single chloride channel current. There was an application of AMP-PNP to the cis side of BLM at 0.5 and 1 mmol/l concentrations, followed by a cis reperfusion, and an application of ATP to the cis side of BLM at 0.5 and 1 mmol/l concentrations, followed by a cis reperfusion. The voltage was 0 mV. The lines on the left mark the closed state of the channels (A). Time-dependence of the inhibitory effect of AMP-PNP and ATP (0.5 and 1 mmol/l) on the single chloride channel open probability (P-open) (B).

was used to test the functionality and orientation of the studied chloride channel. DIDS (100 µmol/l) inhibited the channels from the cis side within 2–5 min. The inhibitory effect was irreversible. 3.2. Chloride channels derived from Percoll-purified mitochondria vesicles To study the effects of Mg2+ on ATP-modulated chloride channels, the single channel properties were measured in the absence of Mg2+ and in the symmetrical ∼40 nmol/l Ca2+ (buffer B). The channels, observed in 19 experiments, had various single channel properties similar to those observed for the channels derived from vesicles isolated from crude mitochondria. Some of the channel behaviour appeared to be burst-like, with periods of silence separating periods of activity (10/19). The conductance (130 ± 27 pS) and Erev (−27 ± 7 mV) of the chloride channels varied from 68 pS to 164 pS and from − 16 mV to − 38 mV, respectively. The experimentally measured current amplitude at 0 mV was i0 = 3.09 ± 0.8 pA. The channel activity was characterised by the median open probability of Popen = 0.538 (0.231–0.716). The open dwell time distribution was best fitted with a second-order exponential function, resulting in two open dwell times, τ1 = 1.1 ms (0.76–1.74) and τ2 = 2.8 ms (1.98–3.7). 3.2.1. Effects of ATP and Mg2+ on chloride channels Na2ATP (0.5–2 mmol/l) inhibited the chloride channels by: decreasing the single channel current and the channel conductivity (12/14); decreasing the single channel current and decreasing P-open (4/14); and decreasing the single channel current and subsequently blocking the channels (4/14) (Figs. 9, 10). The intensity of the Na2ATP effect to decrease the channel current varied among the channels. The effect was reversible and side- and voltage-dependent (Figs. 9, 10).

The reversibility and side dependence of the ATP effect are shown in Fig. 9. An addition of 0.5–2 mmol/l of Na2ATP to the cis side decreased the current, and, after the cis perfusion, the current returned to the original value; subsequent addition of ATP to the trans side was more effective at decreasing the current than that from the cis side. An addition of 0.5–2 mmol/l of ATP to the cis side slightly increased the current, but subsequent addition to the trans side pronouncedly decreased the current (data not shown). Na2ATP (2 mmol/l) decreased the single channel amplitude to 62 ± 11% (n = 9) and to 38 ± 11% (n = 4) at 0 mV and +30 mV, respectively. In the experiments when solutions contained 2 mmol/l MgCl2, ATP decreased the current by 10–30% only (Figs. 5 and 8). Therefore, we studied the effects of MgCl2 on the ATP-induced current decrease. ATP (2 mmol/l) decreased the channel current, and the effect was voltage-dependent. The I–V relationship of the control channel without ATP was linear in the range from 0 mV to + 30 mV (Fig. 10). However, in the presence of 2 mmol/l ATP at the cis side, the I–V became hyperbolic, which means that the ATP inhibition effect increased with increasing positive voltage. MgCl2 (2 and 5 mmol/l), in a concentration-dependent manner, significantly reversed the ATP effect. Upon increasing the concentration of Mg2+, the I–V relationship was less hyperbolic (Fig. 10). 4. Discussion 4.1. Origin and single channel properties of the chloride channels The isolation we used for crude mitochondria also contained, besides mitochondrial membranes, lysosomal membranes and possibly membranes from peroxisomes and endosomes having similar density as mitochondria [38]. Each of the other membrane contaminations (i.e.,

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

1307

Fig. 7. Effects of ATP and ADP on the regular single chloride channel current. There was an application of ATP to the cis side of BLM at 0.5 and 2 mmol/l concentrations, followed by a cis reperfusion, and an application of ADP to the cis side of BLM at 0.5 and 2 mmol/l concentrations, followed by a cis reperfusion. The voltage was 0 mV. The lines on the left mark the closed state of the channels (A). Time-dependence of the inhibitory effect of ATP (B) and ADP (C) at 0.5–2 mmol/l concentrations on the single chloride channel open probability (P-open).

plasmalemma, Golgi, endoplasmic reticulum and nuclear membrane) was less than 2.2% [38]. The mitochondria isolated at the Percoll gradient had no detectable band of lysosomal markers in a western blot [38]. We may assume that most of the observed chloride channels were from mitochondria. However, the purity of the vesicle preparation is critical, as vesicles from different membranes may have different probabilities of fusing into BLMs. So, a minor contamination of the vesicles having high fusion probabilities may misrepresent the channel data. Therefore, we cannot exclude the possible contamination of the mitochondrial vesicles from other membranes. The observed chloride channels had a broad range of single channel behaviour and parameters. We cannot explain these variations due to the limited knowledge of mitochondrial chloride channels, as the molecular identities of these channels are unknown. It may be assumed that the observed channels permeable for Cl−/K+ N 1 and having gating properties belong to at least three different categories of proteins. First, some of the observed chloride channels that develop naturally for purposes completely distinct from ion transport may accidentally exhibit channel properties. An example may be the TOM–TIM complex [62]. The second “category” would reflect proteins with a distinct main physiological function but an auxiliary function of, e.g., a chloride channel. Some of the membrane carriers may belong to this category [44,45]. The third category comprises the already annotated genes with chloride channel properties. However, the single channel properties of the third category channels in our experimental conditions are not known. Only two chloride channels have been reported so far to be

present in mitochondria. The chloride channel, isolated from tobacco, belonged to the voltage-dependent chloride channel (ClC) family, ClCNt. It was enriched in plant mitochondrial membranes [63]. The first intracellular chloride channel identified, known alternately as mtCLIC or CLIC4 (p64H1, or p64 homologue 1), was localised to the mitochondria [64]. Purified recombinant mammalian CLIC4 protein in BLMs formed poorly selective, redox-regulated ion channels [43]. Under oxidising and reducing conditions and in 500/50 mmol/l KCl solutions, the channels had conductances of 8.9. pS (Cl−/K+ = 1.4) and 14.8 pS (Cl−/K+ = 1.5), respectively. These channels also displayed several substates [43]. Since possible multi-oligomerisation or coupling of the CLIC4 channels is unknown, we cannot address the possible connection between the observed chloride channels and CLIC4. Our observed chloride channels had a conductance higher than that for CLIC4 and higher than those of some of the reported chloride channels from cardiac mitoplasts, liver and yeast mitochondria [31–33,43]. The single channel properties of the chloride channels observed in our study were different from the reported VDAC, which has a very high conductance (N1 nS) and a poor Cl−/K+ selectivity (∼b2) and closes to lower conducting states when either positive or negative potential is applied [65]. We did not observe high subconductance jumps of the regular or ragged chloride current, as was reported for VDAC. Our observed chloride channels had a higher conductance than some of the reported chloride channels from cardiac mitoplasts, liver and yeast mitochondria [31–33]. In numerous studies, the 108 pS channel (at symmetrical 150 mmol/l KCl) has been detected in mitochondria in patch-clamp

1308

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

Fig. 8. Modulation of mitochondrial chloride channel kinetics by ATP. (A) Representative recordings of single chloride channel current under control conditions and after the addition of ATP (0.5–2 mmol/l) to the cis side followed by the inhibitory effect of 100 µmol/l DIDS. (B) Open dwell time histogram of the single chloride channel before (dashed line; τ1 = 4 ms, τ2 = 22 ms) and after the application of 2 mmol/l ATP (full line; τ1 = 2 ms, τ2 = 2 ms). (C) Single channel current amplitude histogram before (dashed line) and after the application of 1 mmol/l ATP (full line; leak current was not subtracted). After the application of ATP, the data points were collected for a longer time than in the case before ATP application.

experiments on mitoplasts from different tissues, including heart [28– 30,34,35,66]. The channel has been proposed to correspond to IMAC. The mean conductances of our observed channels were obtained at asymmetric 250/50 KCl. In two experiments, the conductance of 173 pS was obtained at the symmetrical 250 mmol/l KCl (Fig. 4). From this value, we estimated a conductance of 104 pS for symmetrical 150 mmol/l KCl, which is similar to that found for IMAC. Our observed regular channels were inhibited by low pH, as were IMAC and tributyltin [37], which is one of the specific inhibitors of IMAC [67]. Therefore, we assume that most of the observed channels in our study could originate from inner mitochondrial membranes and correspond to IMAC. However, in our study, the chloride channels were observed in the solutions containing MgCl2, which blocks the IMAC channel [26]. More studies are needed to clarify this observation. The differences between the observed regular and ragged and the reported promiscuous [41] intracellular chloride channels indicate that there are at least three main groups of proteins permeable for chloride in mitochondrial membranes. One group

with classical single channel properties is probably the group of regular chloride channel proteins. The ragged and promiscuous groups of chloride channels may involve proteins that pass chloride under specific conditions, as, for example, mitochondrial carriers or BAX [5,44,45,68]. Compared to the regular chloride channels, the ragged chloride channels had lower Popen values, shorter open dwell time, lower Erev values, lower Cl−/K+ selectivity and pronounced flickering. The mitochondria of eukaryotes contain typically between 35 and 55 different carriers [69]. Some of them have been reported to form chloride channels [44,45,68]. Reported homodimers and the three-fold symmetry of mitochondrial carriers in the inner mitochondrial membrane as well as the homotrimer symmetry of the CLIC4 chloride channel [69–74] might be connected with the observed three-level channels (Fig. 2). Mitochondrial proteins that are ready to pass chloride at different mitochondrial functions, as pro-apoptotic BAX forming different oligomeric structures [5,6] or cooperative channel coupling, may contribute to the channel diversity as found for RyR calcium channels [75].

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

1309

Fig. 9. (A) Single chloride channel current under control conditions and after the addition of ATP (2 mmol/l) to the cis side followed by the reperfusion of the cis side and the subsequent addition of 2 mmol/l ATP to the trans side (channel conductance 140 pS, Erev = ∼−40 mV, Cl−/K+ = ∼ 98). (B) Number of observations versus the single channel amplitude of the current traces of the selected current intervals in which the channel was open at different Na2ATP concentrations in trans (Control — full line; 0.5, 1 and 2 mmol/ l Na2ATP are the short dash line, dash-dot line and dotted line, respectively). (C) Single channel amplitude at different Na2ATP concentrations in the cis and trans solutions.

4.2. Effects of ATP and Mg2+ The main observation of our study is that ATP, within a physiological range of concentrations, reversibly interacted with intracellular chloride channels, modified their single channel properties and had effects that were modulated by Mg2+. Concerning the regular channels, the reversible inhibition effect of ATP was sidedependent, similarly to that found for other compounds in our previous studies. The chloride channel blockers DIDS, 5-nitro-2(phenylpropylamino)-benzoate (NPPB) and 2′,4′,6′-trihydroxy-3-(4hydroxyphenyl (phloretin) and 5.8 pH inhibited most of the channels from the cis side, similarly to ATP, whereas bongkrekic acid (BKA), atractyloside (ATR), carboxyatractyloside (CAT) and H2S/HS− inhibited the channels from the opposite trans side of the BLM [37,38,42]. The side-dependent effect of ATP suggests the physiological or pathological significance of the interaction. There are a few studies of ATP effects on intracellular chloride channels. ATP was reported to inhibit an anion channel current in brown fat mitochondria [30], where the effect of ATP was similar to the effect of ADP and was partly reversible. The anion 45 pS channel from yeast mitochondria was blocked by ATP [33]. ATP inhibited the Ca2+-activated anion channel in the sarcoplasmic reticulum of skeletal muscle [55]. We observed that the interactions of ATP with intracellular chloride channels were very complex. From the similar reversible inhibitory effects of ATP and AMP-PNP, it may be suggested that the ATP effect is independent of channel phosphorylation and might occur

at an adenine-nucleotide binding site. In the case of regular chloride channels, we observed that ATP and ADP interacted differently with these channels, which may indicate different binding sites for ATP and ADP. However, in the case of the two ragged chloride channels studied, in which ATP and ADP blocked the same channels, the conclusion about the binding sites is not clear. The side-dependent, reversible inhibitory effect of ATP on the same regular chloride channels may indicate that the ATP binding site was accessible from one side of the channel only. However, in the case of the ragged channels reversibly inhibited by ATP, the conclusion about the binding sites is unclear. A few chloride channels were irreversibly blocked by ATP. In this case, we may suppose that there are other factors that complementarily contribute to the effect and might be responsible for the irreversible block. We observed that ATP increased the gating transition rates of some of the regular chloride channels, which had different single channel properties in comparison to the channels inhibited by ATP. It may be supposed that these channels have different molecular identities than the channels blocked by ATP. However, their molecular identities and their physiological role within the interaction are not known. In the solutions without MgCl2, ATP decreased the single chloride channel current and likewise its conductivity, whereas Mg2+ significantly reversed this effect. These results indicate that ATP interacts with the chloride channels, and Mg2+ may interact with the channels and/or with ATP. Since the molecular identity of the observed channels is unknown, the ATP or Mg2+ binding sites cannot be evaluated. We may speculate that the negatively charged ATP binds

1310

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

Fig. 10. (A) Single chloride channel current under control conditions and after the addition of ATP (2 mmol/l) to the cis side followed by the addition of MgCl2 (2 and 5 mmol/l) to the cis side, measured at + 30 mV (channel conductance 111 pS, Erev = − 28 mV, Cl−/K+ = 6.9). (B) I–V relationship of the control channel (open cycles), 2 mmol/l ATP (cis) (full circles) and after the addition of 2 mmol/l (triangles) and 5 mmol/l (squares) MgCl2 to the cis side where the 2 mmol/l of ATP was present.

to the positively charged sides of the chloride channel, thus not excluding binding to or very close to a permeable pore and therefore partially obstructing the passage of chloride ions. After the addition of Mg2+, the ATP formed MgATP, pulling ATP out of the assumed binding site or pore. Mg2+ may bind to negatively charged sides, thus influencing the structure of the channel and/or ATP binding sites. The increased ATP inhibition effect on the channel current at the increased positive voltages (trans minus cis) may indicate that negatively charged ATP was additionally “attached” into a binding site by electrical force and that the ATP-channel interaction is voltagedependent. However, since ATP binds strongly to Mg2+, there is a possibility that the binding sites of ATP are different than the binding sites of MgATP, resulting in the changes of the channel structure. 4.3. Functional significances of the mitochondrial chloride channels The magnesium ion, Mg2+, is an important divalent cation in cells, modulating numerous protein functions including the activities of channels or ATP-related reactions [76–80]. Some diseases, as Parkinson's disease [81], hypertension [82] or diabetes [83], are also related to changes of Mg2+ concentration. It has been suggested that mitochondria serve as the intracellular Mg2+ store, accumulating and/or releasing Mg2+ and that mitochondrial depolarisation triggers Mg2+ release from mitochondria [84–87]. We may assume that the MgATP– chloride channel interactions observed in our study play a role in these processes. The result that ATP reversibly blocked some chloride channels indicates that these channels are blocked at physiological ATP and that they are activated when the ATP concentration decreases. Chloride channels are involved in many physiological and pathological processes including heartbeat, ischemia, myocardial infarction and in the intrinsic mitochondrial apoptotic pathway [2–4,38,88]. The concentration of ATP decreases in many physiological and pathological processes where tissues are depleted of oxygen or metabolic

nutrients [51–54]. In these processes, the ATP–chloride channel interactions observed in our study may play a role. It has been suggested that IMAC is involved in the repetitive transient depolarisations of a single mitochondrion, in which reactive oxygen species (ROS) induce the opening of IMAC [47,48,89] and mitochondrial membrane depolarisation is related to ATP depletion [90]. A decrease in ATP production has been implied with increases in ROS and apoptosis [91], and chloride channel blockers have been shown to inhibit apoptosis, including a ROS-dependent apoptosis [92,93]. From this, we may speculate that the decrease in ATP concentration may activate mitochondrial chloride channels, thereby decreasing the membrane potential, activating the efflux of superoxide anion from the matrix and modulating apoptosis. Generally, the functional significance of mitochondrial chloride channels or proteins functioning as chloride channels in certain conditions is still mostly unknown. Some inner mitochondrial membrane carriers, as IMAC, VDAC or BAX, can function as membrane channels passing chloride. Since these proteins are involved in apoptosis, inner membrane potential regulation, superoxide release or mitochondrial morphogenesis [89,94–96], the significance and details of their function as chloride channels should be the focus of further studies. In conclusion, the observed results suggest the following: mitochondrial membranes and/or intracellular membranes contain different types of chloride channels that are distinctly influenced by ATP, and the ATP–chloride channel interaction was modulated by Mg2+ ions, which may regulate physiological and pathological processes.

Acknowledgements Financial support by the Slovak Science Grant Agency VEGA 2/ 0150/10 is gratefully acknowledged.

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

References [1] T.K. Rostovtseva, W. Tan, M. Colombini, On the role of VDAC in apoptosis: fact and fiction, J. Bioenerg. Biomembr. 37 (2005) 129–142. [2] B. O'Rourke, Mitochondrial ion channels, Annu. Rev. Physiol. 69 (2007) 19–49. [3] A. Szewczyk, W. Jarmuszkiewicz, W.S. Kunz, Mitochondrial potassium channels, IUBMB Life 61 (2009) 134–143. [4] M. Zoratti, U. De Marchi, E. Gulbins, I. Szabò, Novel channels of the inner mitochondrial membrane, Biochim. Biophys. Acta 1787 (2009) 351–363. [5] P.H. Schlesinger, A. Gross, X.-M. Yin, K. Yamamoto, M. Saito, G. Waksman, S.J. Korsmeyer, Comparison of the ion channel characteristics of proapoptotic BAX and antiapoptotic BCL-2, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 11357–11362. [6] B. Antonsson, S. Montessuit, B. Sanchez, J.C. Martinou, Bax is present as a high molecular weight oligomer/complex in the mitochondrial membrane of apoptotic cells, J. Biol. Chem. 276 (2001) 11615–11623. [7] V. Mikhailov, M. Mikhailova, D.J. Pulkrabek, Z. Dong, M.A. Venkatachalam, P. Saikumar, Bcl-2 prevents Bax oligomerization in the mitochondrial outer membrane, J. Biol. Chem. 276 (2001) 18361–18374. [8] J.C. Sharpe, D. Arnoult, R.J. Youle, Control of mitochondrial permeability by Bcl-2 family members, Biochim. Biophys. Acta 1644 (2004) 107–113. [9] J. Skommer, D. Wlodkowic, A. Deptala, Larger than life: mitochondria and the Bcl2 family, Leuk. Res. 31 (2007) 277–286. [10] L.M. Dejean, S. Martinez-Caballero, K.W. Kinnally, Is MAC the knife that cuts cytochrome c from mitochondria during apoptosis? Cell Death Differ. 13 (2006) 1387–1395. [11] E.V. Pavlov, M. Priault, D. Pietkiewicz, E.H. Cheng, B. Antonsson, A novel, high conductance channel of mitochondria linked to apoptosis in mammalian cells and Bax expression in yeast, J. Cell Biol. 155 (2001) 725–731. [12] M.L. Dejean, S. Martinez-Caballero, L. Guo, C. Hughes, O. Teijido, T. Ducret, F. Ichas, S.J. Korsmeyer, B. Antonsson, E.A. Jonas, K.W. Kinnally, Oligomeric Bax is a component of the putative cytochrome c release channel MAC, mitochondrial apoptosis-induced channel, Mol. Biol. Cell 16 (2005) 2424–2432. [13] H. Yamagata, S. Shimizu, Y. Nishida, Y. Watanabe, W.J. Craigen, Y. Tsujimoto, Requirement of voltage-dependent anion channel 2 for pro-apoptotic activity of Bax, Oncogene 28 (2009) 3563–3572. [14] E.H. Cheng, T.V. Sheiko, J.K. Fisher, W.J. Craigen, S.J. Korsmeyer, VDAC2 inhibits BAK activation and mitochondrial apoptosis, Science 301 (2003) 513–517. [15] R.R. Anholt, P.L. Pedersen, E.B. De Souza, S.H. Snyder, The peripheral-type benzodiazepine receptor. Localization to the mitochondrial outer membrane, J. Biol. Chem. 261 (1986) 576–583. [16] K.W. Kinnally, D.B. Zorov, Y.N. Antonenko, S.H. Snyder, M.W. McEnery, H. Tedeschi, Mitochondrial benzodiazepine receptor linked to inner membrane ion channels by nanomolar actions of ligands, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 1374–1378. [17] S.M. Grigoriev, C. Muro, L.M. Dejean, M.L. Campo, S. Martinez-Caballero, K.W. Kinnally, Electrophysiological approaches to the study of protein translocation in mitochondria, Int. Rev. Cytol. 238 (2004) 227–274. [18] I. Szabò, J. Bock, A. Jekle, M. Soddemann, C. Adams, F. Lang, M. Zoratti, E. Gulbins, A novel potassium channel in lymphocyte mitochondria, J. Biol. Chem. 280 (2005) 12790–12798. [19] A. Szewczyk, J. Skalska, M. Głąb, B. Kulawiak, D. Malińska, I. Koszela-Piotrowska, W.S. Kunz, Mitochondrial potassium channels: from pharmacology to function, Biochim. Biophys. Acta 1757 (2006) 715–720. [20] T.E. Gunter, L. Buntinas, G. Sparagna, R. Eliseev, K. Gunter, Mitochondrial calcium transport: mechanisms and functions, Cell Calcium 28 (2000) 285–296. [21] G.C. Sparagna, K.K. Gunter, S.S. Sheu, T.E. Gunter, Mitochondrial calcium uptake from physiological type pulses of calcium. A description of the rapid uptake mode, J. Biol. Chem. 270 (1995) 27510–27515. [22] G. Beutner, V.K. Sharma, L. Lin, S.Y. Ryu, R.T. Dirksen, S.S. Sheu, Type 1 ryanodine receptor in cardiac mitochondria: transducer of excitation–metabolism coupling, Biochim. Biophys. Acta 1717 (2005) 1–10. [23] W. Xu, Y. Liu, S. Wang, T. McDonald, J.E. Van Eyk, Cytoprotective role of Ca2+activated K+ channels in the cardiac inner mitochondrial membrane, Science 298 (2002) 1029–1033. [24] H.T. Facundo, M. Fornazari, A.J. Kowaltowski, Tissue protection mediated by mitochondrial K+ channels, Biochim. Biophys. Acta 1762 (2006) 202–212. [25] A. Andrukhiv, A.D. Costa, I.C. West, K.D. Garlid, Opening mitoKATP increases superoxide generation from complex I of the electron transport chain, Am. J. Physiol. Heart Circ. Physiol. 291 (2006) H2067–H2074. [26] A.D. Beavis, K.D. Garlid, The mitochondrial inner membrane anion channel. Regulation by divalent cations and protons, J. Biol. Chem. 262 (1987) 15085–15093. [27] A.D. Beavis, Properties of the inner membrane anion channel in intact mitochondria, J. Bioenerg. Biomembr. 24 (1992) 77–90. [28] M.C. Sorgato, B.U. Keller, W. Stühmer, Patch-clamping of the inner mitochondrial membrane reveals a voltage-dependent ion channel, Nature 330 (1987) 498–500. [29] Y.N. Antonenko, K.W. Kinnally, S. Perini, H. Tedeschi, Selective effect of inhibitors on inner mitochondrial membrane channels, FEBS Lett. 285 (1991) 89–93. [30] T. Klitsch, D. Siemen, Inner mitochondrial membrane anion channel is present in brown adipocytes but is not identical with the uncoupling protein, J. Membr. Biol. 122 (1991) 69–75. [31] K.A. Hayman, T.D. Spurway, R.H. Ashley, Single anion channels reconstituted from cardiac mitoplasts, J. Membr. Biol. 136 (1993) 181–190. [32] Y.N. Antonenko, D. Smith, K.W. Kinnally, H. Tedeschi, Single-channel activity induced in mitoplasts by alkaline pH, Biochim. Biophys. Acta 1194 (1994) 247–254.

1311

[33] C. Ballarin, M.C. Sorgato, An electrophysiological study of yeast mitochondria. Evidence for two inner membrane anion channels sensitive to ATP, J. Biol. Chem. 270 (1995) 19262–19268. [34] J. Borecky, P. Jezek, D. Siemen, 108-pS channel in brown fat mitochondria might be identical to the inner membrane anion channel, J. Biol. Chem. 272 (1997) 19282–19289. [35] P. Schönfeld, I. Sayeed, R. Bohnensack, D. Siemen, Fatty acids induce chloride permeation in rat liver mitochondria by activation of the inner membrane anion channel (IMAC), J. Bioenerg. Biomembr. 36 (2004) 241–248. [36] I. Koszela-Piotrowska, K. Choma, P. Bednarczyk, K. Dolowy, A. Szewczyk, W.S. Kunz, L. Malekova, V. Kominkova, K. Ondrias, Stilbene derivatives inhibit the activity of the inner mitochondrial membrane chloride channels, Cell. Mol. Biol. Lett. 12 (2007) 493–508. [37] L. Malekova, V. Kominkova, M. Ferko, P. Stefanik, O. Krizanova, A. Ziegelhoffer, A. Szewczyk, K. Ondrias, Bongkrekic acid and atractyloside inhibits chloride channels from mitochondrial membranes of rat heart, Biochim. Biophys. Acta 1767 (2007) 31–44. [38] L. Malekova, J. Tomaskova, M. Novakova, P. Stefanik, J. Kopacek, B. Lakatos, S. Pastorekova, O. Krizanova, A. Breier, K. Ondrias, Inhibitory effect of DIDS, NPPB, and phloretin on intracellular chloride channels, Pflugers Arch. 455 (2007) 349–357. [39] Z. Tomaskova, J. Gaburjakova, A. Brezova, M. Gaburjakova, Inhibition of anion channels derived from mitochondrial membranes of the rat heart by stilbene disulfonate-DIDS, J. Bioenerg. Biomembr. 39 (2007) 301–311. [40] U. De Marchi, I. Szabò, G.M. Cereghetti, P. Hoxha, W.J. Craigen, M. Zoratti, A maxichloride channel in the inner membrane of mammalian mitochondria, Biochim. Biophys. Acta 1777 (2008) 1438–1448. [41] K. Ondrias, L. Malekova, O. Krizanova, Potassium-chloride promiscuous channels in mitochondrial membranes, Gen. Physiol. Biophys. 27 (2008) 38–44. [42] L. Malekova, O. Krizanova, K. Ondrias, H2S and HS−donor NaHS inhibits intracellular chloride channels, Gen. Physiol. Biophys. 28 (2009) 190–194. [43] H. Singh, R.H. Ashley, CLIC4 (p64H1) and its putative transmembrane domain form poorly selective, redox-regulated ion channels, Mol. Membr. Biol. 24 (2007) 41–52. [44] K. Herick, R. Kramer, H. Luhring, Patch clamp investigation into the phosphate carrier from Saccharomyces cerevisiae mitochondria, Biochim. Biophys. Acta 1321 (1997) 207–220. [45] S.G. Huang, M. Klingenberger, Chloride channel properties of the uncoupling protein from brown adipose tissue mitochondria: a patch-clamp study, Biochemistry 35 (1996) 16806–16814. [46] P.H. Schlesinger, A. Gross, X.-M. Yin, K. Yamamoto, M. Saito, G. Waksman, S.J. Korsmeyer, Comparison of the ion channel characteristics of proapoptotic BAX and antiapoptotic BCL-2, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 11357–11362. [47] M.A. Aon, S. Cortassa, E. Marban, B. O'Rourke, Synchronized whole cell oscillations in mitochondrial metabolism triggered by a local release of reactive oxygen species in cardiac myocytes, J. Biol. Chem. 278 (2003) 44735–44744. [48] S. Cortassa, M.A. Aon, R.L. Winslow, B. O'Rourke, A mitochondrial oscillator dependent on reactive oxygen species, Biophys. J. 87 (2004) 2060–2073. [49] M.A. Aon, S. Cortassa, B. O'Rourke, Mitochondrial oscillations in physiology and pathophysiology, Adv. Exp. Med. Biol. 641 (2008) 98–117. [50] M.R. Duchen, A. Leyssens, M. Crompton, Transient mitochondrial depolarizations reflect focal sarcoplasmic reticular calcium release in single rat cardiomyocytes, J. Cell Biol. 142 (1998) 975–988. [51] R.B. Jennings, K.A. Reimer, The cell biology of acute myocardial ischemia, Annu. Rev. Med. 42 (1991) 225–246. [52] U. Dirnagl, C. Iadecola, M.A. Moskowitz, Pathobiology of ischaemic stroke: an integrated view, Trends Neurosci. 22 (1999) 391–397. [53] B.F. Trump, I.K. Berezesky, Calcium-mediated cell injury and cell death, FASEB J. 9 (1995) 219–228. [54] K. Zhang, J.M. Weinberg, M.A. Venkatachalam, Z. Dong, Glycine protection of PC12 cells against injury by ATP-depletion, Neurochem. Res. 28 (2003) 893–901. [55] G.P. Ahern, D.R. Laver, ATP inhibition and rectification of a Ca2+-activated anion channel in sarcoplasmic reticulum of skeletal muscle, Biophys. J. 74 (1998) 2335–2351. [56] B. Bennetts, G.Y. Rychkov, H.-L. Ng, C.J. Morton, D. Stapleton, M.W. Parker, B.A. Cromer, Cytoplasmic ATP-sensing domains regulate gating of skeletal muscle ClC1 chloride channels, J. Biol. Chem. 280 (2005) 32452–32458. [57] X.D. Zhang, P.Y. Tseng, T.Y. Chen, ATP inhibition of CLC-1 is controlled by oxidation and reduction, J. Gen. Physiol. 132 (2008) 421–428. [58] G.P. Brierley, M.H. Davis, D.W. Jung, Intravesicular pH changes in submitochondrial particles induced by monovalent cations: relationship to the Na+/H+ and K + /H+ antiporters, Arch. Biochem. Biophys. 264 (1988) 417–427. [59] P. Paucek, V. Yarov-Yarovoy, X. Sun, K.D. Garlid, Inhibition of the mitochondrial KATP channel by long-chain acyl-CoA esters and activation by guanine nucleotides, J. Biol. Chem. 271 (1996) 32084–32088. [60] E.L. Holmuhamedov, L. Wang, A. Terzic, ATP-sensitive K+ channel openers prevent Ca2+ overload in rat cardiac mitochondria, J. Physiol. 519 (1999) 347–360. [61] K.A. Hayman, T.D. Spurway, R.H. Ashley, Single anion channels reconstituted from cardiac mitoplasts, J. Membr. Biol. 136 (1993) 181–190. [62] C. Muro, S.M. Grigoriev, D. Pietkiewicz, K.W. Kinnally, M.L. Campo, Comparison of the TIM and TOM channel activities of the mitochondrial protein import complexes, Biophys. J. 84 (2003) 2981–2989. [63] C. Lurin, J. Guclu, C. Cheniclet, J.P. Carde, H. Barbier-Brygoo, C. Maurel, CLC-Nt1, a putative chloride channel protein of tobacco, colocalizes with mitochondrial membrane markers, Biochem. J. 348 (2000) 291–295.

1312

V. Kominkova et al. / Biochimica et Biophysica Acta 1797 (2010) 1300–1312

[64] E. Fernandez-Salas, M. Sagar, C. Cheng, S.H. Yuspa, W.C. Weinberg, p53 and tumor necrosis factor alpha regulate the expression of a mitochondrial chloride channel protein, J. Biol. Chem. 274 (1999) 36488–36497. [65] M. Colombini, Voltage gating in the mitochondrial channel, VDAC, J. Membr. Biol. 111 (1989) 103–111. [66] M.C. Sorgato, O. Moran, V. DePinto, B.U. Keller, W. Stühmer, Further investigation on the high-conductance ion channel of the inner membrane of mitochondria, J. Bioenerg. Biomembr. 21 (1989) 485–496. [67] M.F. Powers, A.D. Beavis, Triorganotins inhibit the mitochondrial inner membrane anion channel, J. Biol. Chem. 266 (1991) 1725–17256. [68] M. Schwarz, A. Gross, T. Steinkamp, U.I. Flugge, R. Wagner, Ion channel properties of the reconstituted chloroplast triose phosphate/phosphate translocator, J. Biol. Chem. 269 (1994) 29481–29489. [69] E.R. Kunji, The role and structure of mitochondrial carriers, FEBS Lett. 564 (2004) 239–344. [70] H. Hackenberg, M. Klingenberg, Molecular weight and hydrodynamic parameters of the adenosine 5′-diphosphate–adenosine 5′-triphosphate carrier in Triton X-100, Biochemistry 19 (1980) 548–555. [71] C.S. Lin, H. Hackenberg, E.M. Klingenberg, The uncoupling protein from brown adipose tissue mitochondria is a dimer. A hydrodynamic study, FEBS Lett. 113 (1980) 304–306. [72] R. Kotaria, J.A. Mayor, R.S. Kaplan, Oligomeric state of wild-type and cysteine-less yeast mitochondrial citrate transport proteins, J. Bioenerg. Biomembr. 31 (1999) 543–549. [73] E. Pebay-Peyroula, C. Dahout-Gonzalez, R. Kahn, V. Trézéguet, G.J.-M. Lauquin, G. Brandolin, Structure of mitochondrial ADP/ATP carrier in complex with carboxyatractyloside, Nature 426 (2003) 39–44. [74] Y.F. Li, D.F. Li, Z.H. Zeng, D.C. Wang, Trimeric structure of the wild soluble chloride intracellular ion channel CLIC4 observed in crystals, Biochem. Biophys. Res. Commun. 343 (2006) 1272–1278. [75] S.O. Marx, K. Ondriaš, A.R. Marks, Coupled gating between individual skeletal muscle Ca2+ release channels, Science 281 (1998) 818–821. [76] M.L. Mayer, G.L. Westbrook, P.B. Guthrie, Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones, Nature 309 (1984) 261–263. [77] L. Nowak, P. Bregestovski, P. Ascher, A. Herbet, A. Prochiantz, Magnesium gates glutamate-activated channels in mouse central neurones, Nature 307 (1984) 462–465. [78] H. Matsuda, A. Saigusa, H. Irisawa, Ohmic conductance through the inwardly rectifying K channel and blocking by internal Mg2+, Nature 325 (1987) 156–159. [79] P. Brion, E. Westhof, Hierarchy and dynamics of RNA folding, Annu. Rev. Biophys. Biomol. Struct. 26 (1997) 113–137. [80] Y.H. Ko, S. Hong, P.L. Pedersen, Chemical mechanism of ATP synthase. Magnesium plays a pivotal role in formation of the transition state where ATP is synthesized from ADP and inorganic phosphate, J. Biol. Chem. 274 (1999) 28853–28856.

[81] M. Yasui, T. Kihira, K. Ota, Calcium, magnesium and aluminium concentrations in Parkinson's disease, Neurotoxicology 13 (1992) 593–600. [82] R.M. Touyz, E.L. Schiffrin, Activation of the Na+–H+ exchanger modulates angiotensin II-stimulated Na+-dependent Mg2+ transport in vascular smooth muscle cells in genetic hypertension, Hypertension 34 (1999) 442–449. [83] R. Lopez-Ridaura, W.C. Willett, E.B. Rimm, S. Liu, M.J. Stampfer, J.E. Manson, F.B. Hu, Magnesium intake and risk of type 2 diabetes in men and women, Diabetes Care 27 (2004) 134–140. [84] G.P. Brierley, M. Davis, D.W. Jung, Respiration-dependent uptake and extrusion of Mg2+ by isolated heart mitochondria, Arch. Biochem. Biophys. 253 (1987) 322–332. [85] G.A. Rutter, N.J. Osbaldeston, J.G. McCormack, R.M. Denton, Measurement of matrix free Mg2+ concentration in rat heart mitochondria by using entrapped fluorescent probes, Biochem. J. 271 (1990) 627–634. [86] M. Salvi, A. Bozac, A. Toninello, Gliotoxin induces Mg2+ efflux from intact brain mitochondria, Neurochem. Int. 45 (2004) 759–764. [87] T. Kubota, Y. Shindo, K. Tokuno, H. Komatsu, H. Ogawa, S. Kudo, Y. Kitamura, K. Suzuki, K. Oka, Mitochondria are intracellular magnesium stores: investigation by simultaneous fluorescent imagings in PC12 cells, Biochim. Biophys. Acta 1744 (2005) 19–28. [88] G. Heimlich, J.A. Cidlowski, Selective role of intracellular chloride in the regulation of the intrinsic but not extrinsic pathway of apoptosis in Jurkat T-cells, J. Biol. Chem. 281 (2006) 2232–2241. [89] M.A. Aon, S. Cortassa, B. O'Rourke, Mitochondrial oscillations in physiology and pathophysiology, Adv. Exp. Med. Biol. 641 (2008) 98–117. [90] E.Y. Wu, M.T. Smith, G. Bellomo, D. Di Monte, Relationships between the mitochondrial transmembrane potential, ATP concentration, and cytotoxicity in isolated rat hepatocytes, Arch. Biochem. Biophys. 282 (1990) 358–362. [91] J.E. Ricci, N. Waterhouse, D.R. Green, Mitochondrial functions during cell death, a complex (I–V) dilemma, Cell Death Differ. 10 (2003) 488–492. [92] X. Wang, N. Takahashi, H. Uramoto, Y. Okada, Chloride channel inhibition prevents ROS-dependent apoptosis induced by ischemia–reperfusion in mouse cardiomyocytes, Cell. Physiol. Biochem. 16 (2005) 147–154. [93] S. Tanabe, X. Wang, N. Takahashi, H. Uramoto, Y. Okada, HCO3-independent rescue from apoptosis by stilbene derivatives in rat cardiomyocytes, FEBS Lett. 579 (2005) 517–522. [94] R. Kumarswamy, S. Chandna, Putative partners in Bax mediated cytochrome-c release: ANT, CypD, VDAC or none of them? Mitochondrion 9 (2009) 1–8. [95] A. Autret, S.J. Martin, Emerging role for members of the Bcl-2 family in mitochondrial morphogenesis, Mol. Cell 36 (2009) 355–363. [96] Z. Tomaskova, K. Ondrias, Mitochondrial chloride channels—what are they for? FEBS Lett. (2010), doi:10.1016/j.febslet.2010.01.035.