ISSN 00062979, Biochemistry (Moscow), 2014, Vol. 79, No. 1, pp. 4453. © Pleiades Publishing, Ltd., 2014. Original Russian Text © O. V. Akopova, L. I. Kolchinskaya, V. I. Nosar, V. A. Bouryi, I. N. Mankovska, V. F. Sagach, 2014, published in Biokhimiya, 2014, Vol. 79, No. 1, pp. 5767.

Effect of PotentialDependent Potassium Uptake on Production of Reactive Oxygen Species in Rat Brain Mitochondria O. V. Akopova*, L. I. Kolchinskaya, V. I. Nosar, V. A. Bouryi, I. N. Mankovska, and V. F. Sagach Bogomolets Institute of Physiology, National Academy of Sciences of Ukraine, ul. Bogomol'tsa 4, 01601 Kiev24, Ukraine; Email: a[email protected] Received July 9, 2013 Revision received September 16, 2013 Abstract—The effect of potentialdependent potassium uptake on reactive oxygen species (ROS) generation in mitochon dria of rat brain was studied. It was found that the effect of K+ uptake on ROS production in the brain mitochondria under steadystate conditions (state 4) was determined by potassiumdependent changes in the membrane potential of the mito chondria (ΔΨm). At K+ concentrations within the range of 0120 mM, an increase in the initial rate of K+uptake into the matrix resulted in a decrease in the steadystate rate of ROS generation due to the K+induced depolarization of the mito + chondrial membrane. The selective blockage of the ATPdependent potassium channel (KATP channel) by glibenclamide and 5hydroxydecanoate resulted in an increase in ROS production due to the membrane repolarization caused by partial inhibition of the potentialdependent K+ uptake. The ATPdependent transport of K+ was shown to be ~40% of the poten tialdependent K+ uptake in the brain mitochondria. Based on the findings of the experiments, the potentialdependent transport of K+ was concluded to be a physiologically important regulator of ROS generation in the brain mitochondria and + that the functional activity of the native KATP channel in these organelles under physiological conditions can be an effective tool for preventing ROS overproduction in brain neurons. DOI: 10.1134/S0006297914010076 + Key words: potassium, brain mitochondria, reactive oxygen species, KATP channel

Mitochondria are major consumers of cellular oxy gen and major producers of reactive oxygen species (ROS) [13]. During aerobic metabolism, the respiratory chain of mitochondria uses more than 90% of the cellular oxygen, and about 2% of the consumed oxygen under certain conditions can be converted to ROS, which are products of the incomplete reduction of oxygen [1, 2]. Mitochondrial dysfunction associated with disorders in the energy metabolism of mitochondria because of inhi bition of electron transport or uncoupling of the respira tory chain can result in both hyperproduction of ROS or suppression of their generation in physiologically neces sary amounts. It is also well known that ROS are inducers of opening of the cyclosporinsensitive pore [4] and of cellular apoptosis [5, 6]. This explains the increased attention of researchers to studies on mechanisms of ROS generation by the respiratory chain of mitochondria. The major sites and some main regulatory mechanisms of ROS generation in mitochondria are now known [13, 7]. The greatest amount of ROS is initially generated as superoxide (О2 ) in complexes I and III of the respiratory

chain [13]. More precise localization of the ROS gener ation sites is under discussion in the literature. Within complex I, ROS are most probably generated in the FMNbinding site [7], FeSclusters, N 1α [8], or N 2 [9] and, probably the ubiquinonebinding site [10]. In com plex III, superoxide is generated as a result of autooxida tion of ubisemiquinone CoQ, which is a hypothetical free radical intermediate of the Qcycle [2, 3]. High membrane potential (ΔΨm) and reduced state of pools of pyridine nucleotides (NADH/NAD+) and ubiquinone (CoQH2/CoQ) are the main prerequisites for ROS gener ation [1, 3, 11]. The hyperpolarization of mitochondria in state 4 corresponds to the maximal reduction of redoxactive sites of the respiratory chain [1, 3]. Electron transfer onto oxygen with production of О2 becomes more advanta geous thermodynamically than the electron transport by the redoxpotential gradient, which leads to “electron leakage” onto oxygen with production of superoxide [1, 11, 12]. The generation of ROS within complex I is promot ed by high membrane potential and reduced state of pyri dine nucleotides and ubiquinone, especially under condi

* To whom correspondence should be addressed.

44

EFFECT OF POTASSIUM UPTAKE ON ROS PRODUCTION tions of reverse electron transport [1, 3, 7]. Within com plex III, ROS generation depends on the redoxpotential of cytochrome b hemes and on the steadystate concen tration of ubisemiquinone CoQ [1, 2, 12]. Increase in this concentration due to hyperpolarization of the mem branes and reduction of cytochrome b [12] or inhibition of electron transport by antimycin A leads to autooxida tion of CoQ [2] and high production of ROS [13, 14]. The potentialdependent uptake of K+ is an impor tant physiological modulator of ROS generation in mito chondria [15]. Interest in this process noticeably increased in connection with the demonstration of cyto protective effects of pharmacological openers of the ATP dependent K+channel (K+ATPchannel) [1618]. Numerous different K+channels are known to exist in the inner mitochondrial membrane that are responsible for potentialdependent uptake of K+ [18, 19]. The ATP dependent uptake of K+ is prevalent as compared to other types of K+conductivity, and its bioenergetic effects are the best studied. It is supposed that the cytoprotective effect of K+ATP channel openers, as well as of the entire transport of K+, is based on the phenomenon of “mild uncoupling” of the respiratory chain [20] that is accompanied, in particular, by prevention of hyperproduction of ROS [11, 16, 21]. However, the literature on the influence of K+ATPchannel activators on ROS generation in mitochondria are rather contradictory. Thus, in some works [22, 23] a decrease in ROS generation was shown on activation of the K+ATP channel in heart and liver mitochondria. However, in works of Garlid’s group an activation of the K+ATPchannel in heart mitochondria was shown to increase ROS pro duction due to alkalization of the mitochondrial matrix [24]. According to the literature, both decrease and increase in ROS generation in mitochondria can lead to cytoprotective effects. In the first case, the mechanism of cell protection was caused by suppression of opening of the mitochondrial pore responsible for inducing apopto sis [22], because ROS are inducers of this pore [46]. In the second case, the protective effect was due to activa tion of the mitochondrial isoform of protein kinase C by hydrogen peroxide within the matrix that, finally, also led to inhibition of the mitochondrial pore [25]. Nevertheless, despite a rather satisfactory explanation of mechanisms of the protective effect of the mitochondrial K+ATPchannel openers, it is obvious that the role of the ATPdependent uptake of potassium in the regulation of ROS generation is still unclear. It should be noted that in the majority of works the bioenergetic effects of the ATPdependent uptake of K+ were studied using pharmacological openers of the K+ATP channel [2225]. However, to more adequately under stand physiological functions of the K+ATPchannel and also of other mitochondrial K+channels, it was interest ing to evaluate the contribution of different types of endogenous K+conductivity to bioenergetic effects of BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

45

the potentialdependent uptake of K+ in native isolated mitochondria in the absence of pharmacological openers. The purpose of the present work was to study the influ ence of the potentialdependent uptake of K+ on ROS generation in rat brain mitochondria and to evaluate the contribution of K+ATPchannels to the regulation of ROS generation in these organelles.

MATERIALS AND METHODS Wistar rats with body weight of 200250 g were used. The brain was washed with cooled (4°C) 0.9% KCl solu tion, minced, and homogenized in 5× volume of medium (250 mM sucrose, 20 mM TrisHCl buffer, 1 mM EDTA, 1 mg/ml BSA, pH 7.4). Mitochondria were isolated by centrifugation at 4°C for 7 min at 1000g, then the super natant was centrifuged at 4°C for 15 min at 12,000g. The precipitate was suspended in a small volume of medium without addition of EDTA and BSA and stored on ice at 4°C. The protein content was determined by the Lowry method. The generation of ROS was assessed by changes in the fluorescence of dichlorofluorescein [24]. The mito chondria were loaded with the nonfluorescent penetrat ing probe 2′,7′dihydrodichlorofluorescein diacetate (DCFHDA), which hydrolyzes within the matrix to the nonpenetrating derivative (dihydrofluorescein) and, upon oxidation by mitochondrial ROS forms dichloroflu orescein (DCF). The mitochondrial suspension was loaded with DCFHDA (final concentration 200 μM) for 20 min at 37°C; after loading of the specimen and wash ing off of the external probe by reprecipitation of the mitochondria, the suspension was stored on ice. Aliquots of the suspension (with 1 mg/ml protein) were introduced into the incubation medium, and the DCF fluorescence was recorded with excitation and emission wavelengths of 504 and 525 nm, respectively. From the fluorescent signal amplitude (F), the basal fluorescence (F0) was subtracted. The F0 value was the result of the introduction of the mitochondria into the incubation medium, and it was found by extrapolation of kinetic curves to zero time. The transport of hydrogen ions was assessed by changes in the fluorescence of the pHsensitive probe 2′,7′bis(2carboxyethyl)5(6)carboxyfluorescein (BCECF) caused by changes in pH of the preloaded mitochondrial matrix (pHi) at excitation and emission wavelengths of 509 and 535 nm, respectively. The mito chondria were preloaded with BCECFAM (final con centration 10 μM [26]) for 10 min at 37°C with subse quent reprecipitation of the mitochondria to remove the excess probe. To determine the rate of proton transport, aliquots of the suspension (0.5 mg/ml) were titrated with HCl solution in the incubation medium without addition of succinate in the presence of 5⋅10–6 M rotenone and 1 μM carbonyl cyanide mchlorophenylhydrazone

46

AKOPOVA et al.

(CCCP) with parallel recording of changes in the BCECF fluorescence; changes in pH were determined under the same conditions with a glass microelectrode; the volume of the medium was 1 ml. The amplitude of the pHi changes and the initial rate of proton transport (nmol H+·min–1·mg–1) were determined from the calibration curves. From the fluorescent signal amplitude, as in the case of DCF, the basal fluorescence of mitochondria (F0) at zero time was subtracted. The membrane potential of the mitochondria (1 mg/ml protein) was recorded in the incubation medi um in the presence of 10 μM safranine at excitation and emission wavelengths of 495 and 586 nm, respectively [22]. The difference between the fluorescence of depolar ized and energized mitochondria was found; for depolar ization, 10–6 M CCCP was added. Oxygen consumption was determined under stan dard conditions by polarography in a closed cell using a platinum electrode at 26°C (1.5 mg/ml final protein con centration). In all cases, the recording was started at the moment of introduction of the mitochondria into the incubation medium: 300 mM sucrose, 2 mM TrisHCl buffer (pH 7.4), 5 mM sodium succinate, 1 mM NaH2PO4. Potassium chloride was added in concentrations of 0 120 mM at overall solution osmolarity of 300 mosmol/ liter, which was maintained with sucrose. Calcium chlo ride was added in to 10 nmol/mg protein together with 1 μM cyclosporin A. In the absence of Ca2+, 1 mM EDTA was added into the medium. Blockers of the K+ATPchan nel (glibenclamide and 5hydroxydecanoate) were intro duced at concentrations of 10–5 and 2·10–4 M, respective ly. In the present work, we used sodium succinate and Tris (base) (Fluka, Switzerland) and 2′,7′dihydrodichlo rofluorescein diacetate, safranine, 2′,7′bis(2car boxyethyl)5(6)carboxyfluorescein, glibenclamide, sodium 5hydroxydecanoate, EDTA, rotenone, valino mycin, cyclosporin A, carbonyl cyanide mchlorophenyl hydrazone, and BSA (Sigma, USA). Other reagents were of analytical purity. Solutions were prepared in bidistilled water. The significance of results was assessed with Student’s ttest; p < 0.05 was considered to be statistical ly significant.

RESULTS AND DISCUSSION Taking into account changes in physiological con centrations of potassium in cells (120150 mM [27]), the influence of potentialdependent K+ uptake was studied within the range of its concentrations of 0120 mM. A change in DCF fluorescence (Fig. 1a) was recorded upon achievement of the steadystate rate of respiration (Fig. 1b) in the presence of succinate as a substrate of oxida tion (state 4). It was found that under conditions of con

stant rate of the substrate oxidation, d[O2]/dt = const (Fig. 1b), the DCF fluorescence increased linearly (Fig. 1a). The constant rate of accumulation of the oxidized product DCF, d[DCF]/dt = const (Fig. 1a), indicated the constant rate of oxidation of the probe by ROS released into the matrix. This allowed us to speak about a definite steadystate level of ROS maintained by the ratio between the rates of ROS generation and removal due to their metabolism in mitochondria [1, 3]. In the absence of a substrate, under conditions of rotenoneblocked endogenous respiration, the observed rate of DCF pro duction during the experiment was close to zero (data not presented), and this allowed us to think that the increase in DCF fluorescence was associated with ROS generation due to functioning of the respiratory chain. On replacement of sucrose by medium containing 120 mM KCl, the respiration rate increased (Fig. 1b, curve 2), which corresponded to the increase in poten tialdependent K+ uptake [28] observed by us earlier concurrently with decrease in the rate of ROS generation (Fig. 1a, curve 2). The introduction of Ca2+ (10 nmol/mg) did not noticeably influence the respira tion rate within the range of K+ concentrations studied (Fig. 2a, curve 2), but it sharply increased the rate of ROS production as compared to the calciumfree medi um, independently of the presence of K+ (Fig. 2b, columns 1 and 2). Removal of Ca2+ from the matrix and the medium by addition of the Ca2+ionophore A23187 in the presence of EDTA returned the ROS generation rate to the control level (Fig. 2b, columns 3). Selective chelation of Ca2+ with EGTA under the same conditions led to the same result (Fig. 2b, columns 4). Thus, we detected no influence of endogenous Mg2+ on ROS gen eration, and the observed Ca2+induced increase in ROS generation was probably caused by transfer of Ca2+ into the mitochondrial matrix. It should be noted that despite advances in under standing of ROS generation mechanisms [13, 7, 21], the influence of Ca2+ transport on ROS generation in mito chondria is not explained sufficiently in the literature [29]. We demonstrated earlier that the rate of ROS gener ation in brain mitochondria increased under conditions of activation of cyclic transport of Ca2+ in the absence of K+, whereas with constant ΔΨm it was directly propor tional to the respiration rate [30]. However, the increase in ROS generation observed in the presence of Ca2+ com pared to calciumfree medium (Fig. 2b, columns 1 and 2) cannot be explained by activation of the Ca2+cycle, because the added amount of calcium did not noticeably contribute to the respiration rate (Fig. 2a, curve 2). Increase in ROS production under the influence of Ca2+ has been observed by many authors [22, 29], and it is considered to be a cause of pathophysiological conse quences of Ca2+overloading of mitochondria. Probably this increase, in addition to bioenergetic effects, is caused by activation of some free radical reactions in the matrix BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

EFFECT OF POTASSIUM UPTAKE ON ROS PRODUCTION

a

47

b 250

200

1

150

30

[O2], μM

F504/525, arbitrary units

45

2

1

100

15

2 50

0

2

4

0

6

2

4

Time, min

Fig. 1. Timedependent changes in DCF fluorescence (a) and oxygen concentration in the medium (b) on succinate oxidation by brain mito chondria. The incubation medium consisting of 2 mM TrisHCl buffer (pH 7.4), 5 mM sodium succinate, 1 mM NaH2PO4, and 300 mM sucrose (1) was replaced by 120 mM KCl (2) at overall osmolarity of 300 mosmol/liter. CaCl2 was introduced to 10 nmol/mg protein. Mitochondria were introduced in amounts of 1 (a) and 1.5 mg/ml (b). M ± m, n = 6.

a

b 1 2 3 4 5 6

45 100

1 30

75 F, %

JО2, ngat O/min per mg

2

* 50

*

15

*

* *

*

25

0 0

30

60

120

I

II

KCl concentration, mM

Fig. 2. Influence of potassium on respiration rate (a) and ROS generation (b) in brain mitochondria. Incubation medium: a) 300 mM sucrose, 2 mM TrisHCl buffer (pH 7.4), 5 mM sodium succinate, 1 mM NaH2PO4. The medium was supplemented with 1 mM EDTA (1) or 10 nmol/mg CaCl2 and 1 μM cyclosporin A (2). The concentration of KCl was varied within the range 0120 mM at overall solution osmo larity of 300 mosmol/liter. b) Sucrose (300 mM) (I) was replaced by 120 mM KCl (II); the medium was supplemented with 1 mM EDTA (1), 10 nmol/mg CaCl2 and 1 μM cyclosporin A (25), 1 μM A23187, and 1 mM EDTA (3), with 1 μM A23187 and 1 mM EGTA (4), 2 min after the beginning of the registration, with rotenone (5⋅10–6 M) (5) and valinomycin (10–7 M) (6). M ± m, n = 4; * p < 0.05, significant with respect to column 2 (I); # p < 0.05, significant with respect to column 2 (II).

BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

48

AKOPOVA et al.

[1, 3] or by immediate induction of generation of free radicals in the mitochondrial membrane [31]. Under the conditions of our experiment, on inhibition of the cyclosporinsensitive pore the addition of calcium did not influence the membrane potential in state 4, which corre lated with insignificant contribution of Ca2+ to the respi ration rate (Fig. 2a). We concluded that the increase in the respiration rate with increase in K+ concentration in the medium (Fig. 2a) in both the absence and presence of Ca2+ was caused by the potentialdependent uptake of K+ (Fig. 1b, curve 2; Fig. 2a, curves 1 and 2). Thus, it can be supposed that the decrease in ROS generation by mito chondria in the K+containing medium (Fig. 1a, curves 1 and 2) can also be due to the transport of K+ into the mitochondrial matrix. Considering the low rate of ROS generation in the calciumfree medium (Fig. 2b, columns 1), the DCF fluorescence was later recorded in the pres ence of Ca2+ (10 nmol/mg). According to the literature, the use of succinate as a respiration substrate can lead to a high yield of ROS due to the reverse transport of electrons from succinate with reduction of FMN and the NADH/NAD pool, which are supposed to be sources of superoxide generation in mito chondria [1, 3, 7]. Rotenone, an inhibitor of complex I, blocks reverse electron transport in the ubiquinonebind ing site, which decreases the ROS production dependent on this process [7, 10]. However, the introduction of rotenone did not decrease ROS generation in the brain mitochondria even in the absence of K+; therefore, reverse electron transport is probably not the mechanism of ROS generation under the conditions of our experi ment (Fig. 2b, columns 5). Nevertheless, the introduc tion of valinomycin, which did not influence ROS gener ation in the absence of K+, sharply decreased the rate of ROS generation in the presence of 120 mM K+ (Fig. 2b, columns 6). Therefore, we think that the observed decrease in ROS generation in the K+containing medi um (Fig. 2b, I, II, columns 1 and 2) should be associated with the potentialdependent uptake of K+ into the mito chondrial matrix. Thus, the purpose of our work was to determine the main characteristics of the potential dependent uptake of K+ in brain mitochondria and to study its influence on ROS generation within the whole range of studied K+ concentrations. To characterize the potentialdependent uptake of K+, we recorded the release of protons in the absence of Ca2+ using the pH sensitive probe BCECF (Fig. 3). Cation transfer into the matrix is associated with both an increase in respiration rate and release of stoi chiometric numbers of protons into the medium [32]. The initial rate V0 of K+ uptake was determined in the cal ciumfree medium by the initial rate of changes in BCECF fluorescence found as described by us earlier [28], taking the K+/H+ stoichiometry as 1 : 1 [32]. The experimental results reveal hyperbolic dependence between V0 of proton release and potassium concentra

tion in the medium (Fig. 3a, curve 1). The maximal rate of the potentialdependent K+ uptake, Vmax, was deter mined from the dependence of V0 of H+ transport on the K+ concentration linearized in doublereciprocal coordi nates: 1/V = (K0.5/Vmax)⋅[K+]–1 + 1/Vmax (Fig. 3b, curve 1). We obtained Vmax = 167.0 ± 3.0 nmol⋅min–1⋅mg–1, which is in good correlation with known characteristics of potentialdependent K+ uptake in mitochondria [33]. Changes in matrix pH due to potentialdependent K+ uptake (ΔpHi) were also found in the absence (Fig. 3c, columns 1) and in the presence of Ca2+ (Fig. 3c, columns 2 and 3). The increase in the K+ uptake was accompanied by alkalization of the matrix (Fig. 3c, columns 1), which indicated the accumulation of K+ and binding of K+ in the mitochondrial matrix. Increase in pH is known to promote ROS generation, whereas medium acidification immediately inhibits generation of superoxide [12, 21]. The matrix alkalization increased still more in the pres ence of Ca2+ (Fig. 3c, columns 2) due to the concurrent potentialdependent Ca2+ uptake, which seemed to be a cause of the Ca2+induced increase in ROS production observed by us (Fig. 2b, columns 1 and 2) and by other authors [22, 29]. Under these conditions, the blocker of the K+ATPchannel 5hydroxydecanoate (5HD) was intro duced, and this resulted in relative acidification of the matrix because of partial prevention of the K+ uptake (Fig. 3a, curve 2; Fig. 3c, columns 2 and 3), which corre lates with literature data on the increase in pHi due to ATPdependent uptake of K+ in heart mitochondria [24]. Activation of the K+ATPchannel of heart mitochondria increased the pH of the matrix and enhanced ROS pro duction [24], but did not influence the membrane poten tial of the mitochondria [34]. In our experiments, both ATPdependent and potentialdependent K+ uptake in brain mitochondria also led to increase in pHi (Fig. 3c, columns 1 and 2); however, this was accompanied by noticeable mitochondrial depolarization with increase in K+ concentration in the medium (Fig. 4a, curve 1) that was already shown by us earlier [28]. Introduction of cytochrome c did not affect the ΔΨm value recorded with in 34 min after the introduction of the mitochondria (data not presented). The ΔΨm was decreasing with increase in the rate of the potentialdependent K+ uptake, which was indicated by our observed linear dependence of membrane potential on V0 of K+ transport (Fig. 4b) and rate of K+stimulated respiration [28] according to data of work [35]. Concurrently, the rate of ROS generation (JR) in the brain mitochondria decreased over the whole range of studied K+ concentrations (Fig. 5, curve 1). The introduction of the K+ATPchannel blocker 5HD partially restored ΔΨm and the rate of ROS generation (Fig. 4a; Fig. 5, white circles), but in the K+containing medium they still remained lower than the control level recorded in the absence of K+. The same was observed in the case of glibenclamide (Fig. 5, gray circles). BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

EFFECT OF POTASSIUM UPTAKE ON ROS PRODUCTION

a

120

49

b

1

90

1/V × 102, nmol−1·min·mg

V0, nmol·min−1·mg−1

5

2 60

30

2

4

3 1 2

1

0

40

80

120

0

–30

30

60

KCl concentration, mM –1 1/C, M−1

c

0.20

1 0.16

2 3

0.12 ΔрНi

* *

* 0.08

* 0.04

0

30

60

90

120

KCl concentration, mM Fig. 3. Influence of K+ concentration in the medium on the initial rate of proton release (a) and pH of the matrix of brain mitochondria (c); b) assessment of Vmax of potentialdependent K+ uptake on concentration dependence (a) linearized in doublereciprocal coordinates. Incubation medium: 300 mM sucrose, 2 mM TrisHCl buffer (pH 7.4), 5 mM sodium succinate, 1 mM EDTA, 1 mM NaH2PO4 ((ac), columns 1); instead of EDTA, 10 nmol/mg CaCl2 and 1 μM cyclosporin A were introduced ((c), columns 2 and 3). The KCl concentration was varied within the range 0120 mM. The medium was supplemented with 2·10–4 M 5HD ((a, b), curves 2; (c), columns 3). Mitochondria were introduced to 0.5 mg/ml protein. M ± m, n = 6 (ac); # p < 0.05, significant with respect to columns 1; * p < 0.05, significant with respect to columns 2 (c). c) Amplitude of changes in matrix pH, ΔpHi, in the absence of Ca2+ (columns 1) and in the presence of 10 nmol/mg CaCl2 (columns 2 and 3).

There was a correlation between the decrease in ΔΨm and suppression of ROS generation on the increase in the K+ concentration in the medium. Taking into account the literature data about the dependence of ROS generation BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

on the membrane potential of mitochondria [1, 3, 11, 12], it was supposed that the K+induced depolarization of the membrane due to the potentialdependent K+ uptake in brain native mitochondria could be the main

50

AKOPOVA et al.

a 100

b 100

2 3

ΔΨm, %

75

75 1

50

50

25

25

0

40

80

120

KCl concentration, mM

0

40

80 −1

V0, nmol·min ·mg

120 −1

Fig. 4. Influence of K+ concentration in the medium (a) and of the initial rate of potentialdependent K+ uptake (b) on brain mitochondria membrane potential in control (1), in the presence of 5HD (2, white circles); 3) evaluation of changes in ΔΨm on the supposed 40% inhibi tion of K+ uptake (dotted line). M ± m, n = 6.

cause of the decrease in ROS generation. To test this hypothesis, we studied the influence of membrane depo larization induced by valinomycin and malonate on the rate of ROS generation in the brain mitochondria. It is known that depolarization of the energized mitochondrial membrane in the presence of valinomycin depends on the K+ concentration in the medium, which has to provide K+ uptake into the matrix at a given ΔΨm value [36]. Therefore, in the absence of K+ valinomycin failed to depolarize the mitochondrial membrane (Fig. 2b, I, column 6). To reveal the dependence of ROS gen eration on the depolarizing effect of the potential dependent K+ uptake induced by valinomycin, we studied the dependence of ΔΨm on the K+ concentration in the medium in the presence of valinomycin with concurrent recording the rate of ROS generation. The K+ concentra tions were chosen to provide only partial depolarization of the membrane upon the introduction of valinomycin. Dependences of the membrane potential and the ROS generation rate on malonate concentration without addi tion of K+ were studied similarly. Results of these experi ments are presented in Fig. 6. Thus, the depolarizing effect of the K+ transport in native mitochondria virtually coincided with the effect of the potentialdependent K+ uptake induced by valino mycin (Fig. 6, curve 1, white and gray circles). So, the experiment confirmed the existence of the potential dependent mechanism of ROS regulation under condi tions of K+ accumulation by energized brain mitochon dria. This was also confirmed on inhibition of respiration by malonate (Fig. 6, curve 2), although there are discrep

ancies in the region of low ΔΨm. These discrepancies were probably due to differences in the conditions of ROS gen eration. Thus, in the case of malonate the depolarization was accompanied by inhibition of electron transport and relative acidification of the matrix, whereas in the case of K+ uptake electron transport and also pHi increased. These differences became greater with depolarization of the mitochondrial membrane. As mentioned above, the matrix alkalization as a result of the ATPdependent K+ uptake in the absence of depolarization could lead to increase in ROS production [24], whereas elimination of the transmembrane gradient of pH lowered it [10, 21]. Under conditions of K+ induced depolarization, the concurrent proceeding of both processes – increase in ROS generation owing to the increase in pHi (and in ΔpH) and its decrease owing to decrease in ΔΨm – could finally result in a relative increase in the rate of ROS generation at the same ΔΨm values (Fig. 6, curves 1 and 2). On the contrary, in the case of malonate the inhibition of respiration can lead to decrease in ΔpH with respect to the control and K+ uptake at equal potential values. This also can be an addi tional cause of a relative decrease in ROS generation and of observed differences in the depolarizing action of K+ and malonate on ROS generation at low values of the membrane potentials. Under physiological conditions, the potential dependent K+ uptake into the mitochondrial matrix and its bioenergetic effects are mediated through mitochon drial K+channels [18, 19]. It can be supposed that the different actions of the potentialdependent K+ uptake on BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

EFFECT OF POTASSIUM UPTAKE ON ROS PRODUCTION 100

75 1 JR, %

mitochondrial functions in different types of cells described in the literature [34, 35] could be essentially due to different densities of the distribution of K+chan nels in the mitochondrial membranes [18, 37]. To more fully understand the role of K+channels in the regulation of ROS production by mitochondria, it was necessary to assess their individual contribution to potentialdepend ent K+ transport, oxygen consumption, and regulation of membrane potential of mitochondria. The wellknown cardioprotective effect of K+ATP channel openers is often thought to be associated with regulation of ROS generation in mitochondria [2225]; however, data on the direct action of the K+ATPchannel activators on ROS production are rather contradictory. According to the literature, the density of K+ATPchannel distribution in the membrane of brain mitochondria is higher than in heart and other tissues [37]. Therefore, the bioenergetic effects of ATPdependent K+ transport in mitochondria of neurons were supposed to have some specific features. And it was interesting to determine the

51

50 2 25

0

40

80

ΔΨm, %

Fig. 6. Influence of the membrane potential of brain mitochondr ial on the rate of ROS generation in state 4. Membrane depolar ization was induced by increasing the K+ concentration in the medium within the range 0120 mM (1, gray circles) or by intro duction of valinomycin (1, white circles) or malonate (2). The corresponding values in the absence of K+ were taken as 100%.

100

JR, %

75

2

1

50

25

0

40

80

120

KCl concentration, mM

Fig. 5. Influence of K+ concentration in the medium on ROS gen eration in rat brain mitochondria: 1) control; 2) dependence found on supposed 40% inhibition of K+ uptake (dotted curve, explanation in text); white and gray circles represent the experi mental data in the presence of the K+ATPchannel blockers 5HD and glibenclamide. Incubation medium: 300 mM sucrose, 2 mM TrisHCl buffer (pH 7.4), 5 mM sodium succinate, 1 mM NaH2PO4, 1 μM cyclosporin A. The KCl concentration was var ied within the range 0120 mM at overall solution osmolarity of 300 mosmol/liter. Calcium chloride was introduced to 10 nmol/mg protein. Mitochondria were added to 1 mg/ml pro tein. Blockers of the K+ATPchannel (glibenclamide and 5HD) were introduced to concentrations of 10–5 and 2·10–4 M, respec tively. M ± m, n = 6. The value obtained in the absence of added K+ was taken as 100%.

BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

contribution of the native K+ATPchannel to the regulation of ROS production in brain mitochondria. For this pur pose, we studied the influence of the selective K+ATPchan nel blockers glibenclamide and 5hydroxydecanoate (5 HD) on ROS generation in these organelles. Results of experiments revealed (Fig. 3, a and b, curve 2) that inhibition of the K+ATPchannel by 5HD resulted in a decrease in Vmax of the potentialdependent K+ uptake from 167.0 ± 3.0 to 92.0 ± 2.7 nmol⋅min–1⋅mg–1. This cor responded to the Vmax of the ATPdependent K+ uptake of 72 nmol⋅min–1⋅mg–1, which was ~43% of the potential dependent K+ uptake. We found earlier using inhibition of the K+ATPchannel by Mg⋅ATP with subsequent reacti vation with diazoxide that the contribution of the K+ATP channel to respiration rate on succinate in the presence of 120 mM K+ was 11.5 ± 0.9 ngat O⋅min–1⋅mg–1. The resulting value was close to results of K+ATPchannel inhi bition by glibenclamide and 5HD in the absence of Mg⋅ATP, respectively, 12.0 ± 1.2 and 14.0 ± 1.5 ngat O⋅min–1⋅mg–1 [28], which was ~30% of the maximal rate of K+stimulated respiration of mitochondria (43.4 ± 1.1 ngat O⋅min–1⋅mg–1) and, moreover, suggested an immediate inhibition of the K+ATPchannel by these block ers. To more accurately assess the contribution of K+ATP channels to the potentialdependent rate of K+ uptake, we took into account polarography data on the contribu tion of proton leakage to the respiration rate in state 4. The rate of proton transport on titrating the respiration

52

AKOPOVA et al.

with malonate in the absence of K+ was calculated at the equal value of the depolarizing effect and on considera tion of the stoichiometry H+/O = 6 [38]. The resulting value was subtracted from the maximal rate of the K+ stimulated respiration on the suggestion of the additivity of the H+ and K+cycle contributions to the rate of oxy gen consumption. The resulting Vmax value of the K+ stimulated respiration (28.7 ± 1.2 ngat O⋅min–1⋅mg–1) on consideration of the stoichiometry K+/O = 6 was ~172 nmol K+⋅min–1⋅mg–1. Thus, according to the polarography data, the K+ATPchannel contribution (~12.6 ngat O⋅min–1⋅mg–1) was ~44% of the potential dependent K+ uptake in the brain mitochondria. This result is close to the evaluation obtained with the pHsen sitive probe BCECF. It followed from the dependence of ΔΨm on the rate of K+ transport (Fig. 4b) and also from the dependence of the ROS generation rate on the membrane potential (Fig. 6, curve 1) that the inhibition of the potentialdependent K+ uptake corresponding to the K+ATPchannel contribu tion had to lead to repolarization of the mitochondrial membrane and increase in the rate of ROS generation. Assessment of the repolarizing effect of K+ATPchannel inhibition from the ΔΨm dependence on the rate of K+ transport (Fig. 4b) indicated that the recovery of ΔΨm was most similar to the effect of the K+ATPchannel blocker 5 HD, on the suggestion of inhibition of 40% of the K+ transport (Fig. 4a, white circles (2), dotted line (3)). The assessment of the influence of the repolarizing effect of the brain mitochondrial K+ATPchannel inhibition on ROS production (Fig. 5, dotted line (2)) was also con firmed by results of experiments with blockers of the K+ATPchannel. And the assessment of the effect of the inhibition of K+ transport on ROS generation from the dependence of ROS production on ΔΨm (Fig. 6, curve 1), on the suggestion of 40% inhibition of the K+ uptake was also the most similar to the data obtained with gliben clamide and 5HD (Fig. 5, white and gray circles, dotted line (2)). Similarity of results obtained based on the empirical dependences to direct effects of the K+ATPchan nel blockers indicated the reliability of the assessment of the contribution of the K+ATPchannel to the potential dependent transport of K+. Thus, the experimental results confirmed the hypothesis that in the steadystate the influence of the K+ATPchannel blockers on ROS genera tion in mitochondria was due to the repolarizing effect of the partial inhibition of the K+ uptake. According to the literature, potentialdependent K+ uptake can cause differently directed effects on ROS gen eration [15, 2225]. Thus, in heart mitochondria the di azoxideactivated ATPdependent uptake of K+ resulted in either an increase or a decrease in ROS production [22, 24]. In Garlid’s works, the increase in ROS production was explained by the alkalizing effect of the ATPdepend ent uptake of K+ [24], whereas the decrease in ROS pro duction also observed on activation of the K+ATPchannel

the authors [22, 23] thought to be caused by activation of K+/H+exchange. In our experiment the potential dependent uptake of K+ resulted in a noticeable alkaliza tion of the matrix enhanced by the concurrent uptake of Ca2+ (Fig. 3c, columns 1 and 2). As differentiated from our earlier observed activation of K+/H+exchange on the activation of the K+ATPchannel of liver mitochondria [39], the uptake of K+ in brain mitochondria was not accom panied by activation of K+/H+exchange during the experiment, which could be due to a decrease in the activity of the K+/H+exchanger due to depolarization of the membrane described in the literature [40]. Notwithstanding the increase in pHi and absence of acti vation of K+/H+exchange, the ROS production was decreased over the whole range of studied K+ concentra tions. Under conditions of both the potentialdependent uptake of K+ in native mitochondria and the valino mycininduced uptake of K+, the decrease in ROS release most of all correlated with the depolarization of the mito chondrial membrane. And the regulation of ROS genera tion by the membrane potential was not mediated through the reverse transport of electrons, which required high values of the potential and strongly depended on ΔΨm. Based on results of the experiments, we conclude that the K+induced depolarization of mitochondria under conditions of direct electron transport is the main cause of the decrease in ROS generation in brain mito chondria. The effect of inhibition of the native K+ATP channel of mitochondria confirms this conclusion, because, notwithstanding a decrease in the alkalizing effect of the K+ uptake (Fig. 3c, columns 3), the inhibi tion of the ATPdependent K+ uptake results in an increase in ATP release due to partial repolarization of the mitochondrial membrane. Thus, as differentiated from heart mitochondria, where the increase in potentialdependent K+ uptake due to activation of the K+ATPchannel led to increase in ROS production because of alkalization of the matrix in the absence of changes in ΔΨm [24, 34], in the brain mito chondria the increase in the potentialdependent trans port of this cation led to a decrease in ROS production because of membrane depolarization. The depolarizing effect of K+ and also the decrease in ROS production in the native brain mitochondria were significantly con tributed to by the ATPdependent transport of K+. This finding probably suggests the cellular specificity of bioen ergetic effects of the potentialdependent uptake of K+, and the reasons for it are to be elucidated in further inves tigations. The results of our experiments indicate an important role of the potentialdependent transport of K+ and, in particular, of the K+ATPchannel as modulators of membrane potential and potentialdependent bioener getic effects in brain mitochondria. The regularities found by us suggest that the functional activity of the native K+ATPchannel of brain mitochondria should act as an effective endogenous mechanism for regulation of ROS BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

EFFECT OF POTASSIUM UPTAKE ON ROS PRODUCTION generation and an important element of the neuroprotec tive system of the body responsible for prevention of the development of apoptosis and pathological processes in brain neurons caused by hyperproduction of ROS.

REFERENCES 1. Murphy, M. P. (2009) Biochem. J., 417, 113. 2. Han, D., Williams, E., and Cadenas, E. (2001) Biochem. J., 353, 411416. 3. Andreyev, A. Y., Kushnareva, Y. E., and Starkov, A. A. (2005) Biochemistry (Moscow), 70, 200214. 4. Zorov, D. B., Juhaszova, M., and Sollott, S. (2006) Biochim. Biophys. Acta, 1757, 509517. 5. Kroemer, G., Galluzzi, L., and Brenner, C. (2007) Physiol. Rev., 87, 99163. 6. Skulachev, V. P. (1999) Mol. Aspects Med., 20, 139184. 7. Grivennikova, V. G., and Vinogradov, A. D. (2006) Biochim. Biophys. Acta, 1757, 553561. 8. Kushnareva, Y., Murphy, A. N., and Andreyev, A. (2002) Biochem. J., 368, 545553. 9. Genova, M. L., Ventura, B., Giuliano, G., Bovina, C., Formiggiani, G., Parenti, C. G., and Lenaz, G. (2001) FEBS Lett., 505, 364368. 10. Lambert, A. J., and Brand, M. D. (2004) J. Biol. Chem., 279, 3941439420. 11. Korshunov, S. S., Skulachev, V. P., and Starkov, A. A. (1997) FEBS Lett., 416, 1518. 12. Liu, Sh.S. (1999) J. Bioenerg. Biomembr., 31, 367376. 13. Turrens, J. F., Alexandre, A., and Lehninger, A. L. (1985) Arch. Biochem. Biophys., 237, 408414. 14. Chen, Q., Vazquez, E. J., Moghaddas, S., Hoppel, C. L., and Lesnefsky, E. J. (2003) J. Biol. Chem., 278, 36027 36031. 15. Malinska, D., Mirandola, S. R., and Kunz, W. S. (2010) FEBS Lett., 584, 20432048. 16. Mironova, G. D., Kachaeva, E. V., and Kopylov, A. T. (2007) Vestnik Ros. Akad. Med. Nauk, 2, 3443. 17. Busija, D. W., Lacza, Zs., Rajapakse, N., Shimizu, K., Kis, B., Bari, F., Domoki, F., and Horiguchi, T. (2004) Brain Res. Rev., 46, 282294. 18. O’Rourke, B. (2004) Circ. Res., 94, 420432. 19. Szewczyk, A., Jarmuszkiewicz, W., and Kunz, W. (2009) IUBMB Life, 61, 134143.

BIOCHEMISTRY (Moscow) Vol. 79 No. 1 2014

53

20. Garlid, K. D., and Paucek, P. (2003) Biochim. Biophys. Acta, 1606, 2341. 21. Brand, M. D., Affourtit, Ch., Esteves, T., Green, K., Lambert, A. J., Miwa, S., Pakay, J. L., and Parker, N. (2004) Free Radical Biol. Med., 37, 755767. 22. Facundo, H. T. F., dePaula, J. G., and Kowaltowski, A. J. (2005) J. Bioenerg. Biomembr., 37, 7582. 23. Ferranti, R. F., da Silva, M. M., and Kowaltowski, A. J. (2003) FEBS Lett., 536, 5155. 24. Andrukhiv, A., Costa, A. D., West, I. C., and Garlid, K. D. (2006) Am. J. Physiol., 291, H2067H2074. 25. Costa, A. D. T., and Garlid, K. D. (2008) Am. J. Physiol., 295, H874H882. 26. Jung, D. W., Davis, M. H., and Brierley, G. P. (1989) Anal. Biochem., 178, 348354. 27. Moore, C. L. (1971) Curr. Top. Bioenerg., 4, 191236. 28. Akopova, O. V., Kolchinskaya, L. I., Nosar, V. I., Bouryi, V. A., Mankovska, I. N., and Sagach, V. F. (2013) Biochemistry (Moscow), 78, 8090. 29. Feissner, R. F., Skalska, J., Gaum, W. E., and Sheu, Sh. Sh. (2009) Front. Biosci., 14, 11971218. 30. Akopova, O. V., Kolchinskaya, L. I., Nosar, V. I., Smirnov, A. N., Malysheva, M. K., Mankovska, I. N., and Sagach, V. F. (2011) Ukr. Biokhim. Zh., 83, 4655. 31. Grijalba, M. T., Vercesi, A. E., and Schreier, Sh. (1999) Biochemistry, 38, 1327913287. 32. Massari, S., and Azzone, G. F. (1970) Eur. J. Biochem., 12, 301309. 33. Aon, M. A., Cortassa, S., Wei, A.Ch., Grunnet, M., and O’Rourke, B. (2010) Biochim. Biophys. Acta, 1797, 7180. 34. Kowaltowski, A. J., Seetharaman, S., Paucek, P., and Garlid, K. D. (2001) Am. J. Physiol., 280, H649H657. 35. Cancherini, D., Trabuco, L. G., Reboucas, N. A., and Kowaltowski, A. J. (2003) Am. J. Physiol., 285, F1291 F1296. 36. Chizmadzhev, Y. A. (ed.) (1981) Membranes: Ionic Channels [in Russian], Mir, Moscow. 37. Bajgar, R., Seetharaman, S., Kowaltowski, A. J., Garlid, K. D., and Paucek, P. (2001) J. Biol. Chem., 276, 33369 33374. 38. Brown, G. C. (1992) FASEB J., 6, 29612965. 39. Akopova, O. V., Nosar, V. I., Bouryi, V. A., Mankovska, I. N., and Sagach, V. F. (2010) Biochemistry (Moscow), 75, 11391147. 40. Bernardi, P., and Azzone, G. F. (1983) Biochim. Biophys. Acta, 724, 212223.