Mitochondrial hydrogen peroxide production as determined by the pyridine nucleotide pool and its redox state

Biochimica et Biophysica Acta 1817 (2012) 1879–1885

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbabio

Mitochondrial hydrogen peroxide production as determined by the pyridine nucleotide pool and its redox state☆ Alexandra V. Kareyeva, Vera G. Grivennikova, Andrei D. Vinogradov ⁎ Department of Biochemistry, School of Biology, Moscow State University, Moscow 119991, Russian Federation

a r t i c l e

i n f o

Article history: Received 20 February 2012 Received in revised form 27 March 2012 Accepted 29 March 2012 Available online 5 April 2012 Keywords: Respiratory complex I Dihydrolipoamide dehydrogenase Hydrogen peroxide Pyridine nucleotides Mitochondria

a b s t r a c t The rates of NADH-supported superoxide/hydrogen peroxide production by membrane-bound bovine heart respiratory complex I, soluble pig heart dihydrolipoamide dehydrogenase (DLDH), and by accompanying operation of these enzymes in rat heart mitochondrial matrix were measured as a function of the pool of pyridine nucleotides and its redox state. Each of the activities showed nontrivial dependence on nucleotide pool concentration. The NAD+/NADH ratios required for their half maximal capacities were determined. About half of the total NADH-supported H2O2 production by permeabilized mitochondria in the absence of stimulating ammonium could be accounted for by DLDH activity. The significance of the mitochondrial NADH-dependent hydrogen peroxide production under physiologically relevant conditions is discussed. This article is part of a Special Issue entitled: 17th European Bioenergetics Conference (EBEC 2012). © 2012 Elsevier B.V. All rights reserved.

1. Introduction Mitochondria isolated from various tissues produce hydrogen peroxide when incubated aerobically in the presence of oxidizable substrates [1–3]. The specific rates of the mitochondrial H2O2 production published in numerous reports vary significantly, being dependent on particular tissue, substrate-electron donor (see, for example, Refs. [4–6]), ionic composition, and pH, and also on the assay methods employed. Despite these variations a general phenomenon is evident: coupled mitochondria (state 4 respiration) show high rate of hydrogen peroxide formation, whereas much lower H2O2 production is seen during uncoupled or state 3 respiration unless some specific inhibitors of the respiratory chain are present [1,7,8]. An obvious explanation for this phenomenon is that the rate of H2O2 formation in the mitochondrial matrix in the presence of NAD+-dependent substrates or succinate directly correlates with the redox state of the NADH/NAD+ couple, the actual electron donor/acceptor pair for the site(s) where oneor two-electron reduction of oxygen proceeds (see Ref. [9] for the contrary). It is very hard, if not impossible; to gain direct information on the nature and properties of the enzymes that generate hydrogen peroxide and/or superoxide from the data obtained using intact mitochondria. When external H2O2 released by respiring mitochondria is measured in either state 4 or state 3, the net rate is a complex function

Abbreviations: DLDH, dihydrolipoamide dehydrogenase; ROS, reactive oxygen species; SOD, superoxide dismutase; SMP, submitochondrial particles ☆ This article is part of a Special Issue entitled: 17th European Bioenergetics Conference (EBEC 2012). ⁎ Corresponding author. Tel./fax: + 7 495 939 13 76. E-mail address: [email protected] (A.D. Vinogradov). 0005-2728/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbabio.2012.03.033

of multiple parameters such as permeation/exchange of the respiratory substrate/product, activities of corresponding matrix located dehydrogenases and “antioxidant” enzymes (superoxide dismutase, glutathione reductase, glutathione peroxidase, catalase), glutathione content, intramitochondrial concentrations of NADH/NAD+ and NADPH/NADP+, permeability of the mitochondrial membrane for hydrogen peroxide, and at last, possibly at least, on the activities of actual generators. More explicit information on the respiratory chain componentmediated H2O2/O2 (conventionally called ROS) production have been accumulated from studies using inside–out submitochondrial particles (SMP) essentially devoid of the matrix-located generators and “antioxidant” enzymes [10–14]. About two-thirds of the succinatesupported superoxide generation by coupled SMP via the reverse electron transfer pathway is blocked by rotenone [14] or NADH-OH [15], a specific complex I nucleotide binding site-directed inhibitor [16], whereas the residual, strongly stimulated by Antimycin A reaction is catalyzed by complex III [17–19]. This suggests that complex I is a major contributor to the respiratory chain-mediated superoxide production. Purified preparations of complex I are also known to catalyze NADH-supported superoxide formation [20–22]. Almost equal superoxide production by SMP is seen during succinatesupported reverse electron transport or in the presence of optimal (50 μM) NADH and rotenone [23]. Both, reverse electron transfersupported and forward NADH oxidation-supported superoxide generation by SMP are inhibited by high (millimolar) concentration of NADH and low (μmolar) concentration of NAD+ [14]. A fraction of matrix located proteins have been shown to catalyze NADH-supported, ammonium-stimulated H2O2 production [24]. Dihydrolipoamide dehydrogenase (DLDH), E-3 component of several lipoamide-dependent enzymes, has been identified

1880

A.V. Kareyeva et al. / Biochimica et Biophysica Acta 1817 (2012) 1879–1885

as the protein catalyzing the latter reaction [25] in agreement with previously described α-oxoglutarate dehydrogenase as a potential source of mitochondrially generated hydrogen peroxide [26,27]. In the presence of stimulatory ammonium at fixed NADH concentration (50 μM), most of the total H2O2 production (up to 90%) by alamethicin-permeabilized rat heart mitochondria is catalyzed by DLDH [28]. To get greater insight into the physiological significance of complex I- and DLDH-mediated activities, it seemed of interest to determine how hydrogen peroxide production depends on the total NAD + (NADH) pool concentration and on its redox state. In this paper, data on these dependences are reported along with those measured for alamethicin-permeabilized mitochondria, i.e. under the conditions where these enzymes conjointly operate within their natural protein environment and where the pool as well as the redox state of the NAD +/NADH pair can be deliberately varied [29].

2. Materials and methods Rat heart mitochondria were isolated essentially as described [30]. Bovine heart submitochondrial particles (SMP) were prepared, coupled by treatment with oligomycin, and activated according to our standard protocol [31,32]. Rat heart mitochondria were permeabilized [29] in the standard reaction mixture by preincubation with alamethicin (20 μg/ ml) and 5 mM MgCl2 for 1 min. Pig heart DLDH was obtained from Calzyme (U.S.A.). Rat heart mitochondria and bovine heart SMP were used for comparison to be a compromise between the requirements for intact right-side-out and notoriously inside-out preparations, respectively, assuming that no significant difference in the properties and specific activities exist for either species. Separate experiments showed that all catalytic properties of commercially available DLDH were the same as those previously described for the enzyme purified from bovine heart mitochondrial matrix [25]. Bovine heart complex I was purified according to Hatefi [33]. The standard reaction mixture for all assays except for those described in Figs. 1 and 4 was composed of 0.25 M sucrose, 10 mM KCl, 0.1 mM EDTA-KOH, and 5 mM potassium phosphate, pH 7.5 (other additions are indicated in the legends to the tables and figures). H2O2 generation was monitored at 30 °C at 572 nm (formation of resorufin, ε572–600 = 54 mM− 1 cm− 1 [34]) in the standard reaction mixture supplemented with 20 μM Amplex Red, horseradish peroxidase (2 units/ml), and superoxide dismutase (SOD) from bovine erythrocytes (6 units/ml). Note should be made that in agreement with an earlier report [35], freshly prepared NADH solution contained significant hydrogen peroxide, which is increased upon storage. A substantial amount of rapid resorufin formation was detected when NADH, particularly its high concentrations, was added to the assay mixture in the absence of added SMP or DLDH. SMP- and DLDH-catalyzed hydrogen peroxide generation was initiated by the addition of the enzymes to the assay system. Superoxide production was measured as SOD-sensitive acetylated cytochrome c (10 μM) reduction at 550 nm (ε550 = 20 mM− 1 cm− 1) as described before [14]. The succinate oxidase activity of SMP was monitored as fumarate formation measured as an increase in absorption at 278 nm (ε278 = 0.3 mM − 1 cm − 1). Separate experiments showed that the same specific succinate oxidase activity is detected when measured either amperometrically with an oxygen-sensitive electrode or by photometric assay at 278 nm. Protein content was determined by the biuret procedure. Amplex Red was from AnaSpec, Inc. (U.S.A.); other fine chemicals were from Sigma-Aldrich (U.S.A.). NADH-OH was kindly provided by Dr. Gary Cecchini (Molecular Biology Division, VA Medical Center and Department of Biochemistry & Biophysics, University of California, San Francisco).

Fig. 1. Inhibition of NADH oxidase and NADH-supported superoxide production catalyzed by SMP. SMP (2 mg protein/ml) were preincubated and assayed in the mixture composed of 0.25 M sucrose, 50 mM Tris-Cl (pH 8.0), and 0.2 mM EDTA for 15 min with NADH-OH (concentrations are indicated on abscissa) and their NADH oxidase (open symbols) and superoxide-sensitive NADH:acetylated cytochrome c reductase (filled symbols) activities were assayed at 30 °C. Gramicidin (0.05 μg/ml) was added to NADH oxidase assay mixture. NADH (50 μM), acetylated cytochrome c (10 μM) and rotenone (5 μM) were added when superoxide production was measured. The reactions were initiated by the additions of SMP (100 μg protein/ml). Hundred percent activities correspond to 0.85 μmol of NADH oxidized and 1.5 nmol of cytochrome c reduced per min per mg protein. NADHOH used was not 100% pure and its concentrations shown in abscissa can be used only as approximate measure of complex I content in SMP (0.11 nmol/mg protein).

3. Results Three instrumental approaches were used in this study to discriminate between complex I- and DLDH-mediated mitochondria H2O2 production: (i) use of NADH-OH as the specific tight inhibitor of all NADH-dependent complex I activities [15,16]; (ii) ammonium as the specific activator of DLDH-catalyzed reaction [25], and (iii) alamethicininduced permeabilization of intact rat heart mitochondria, a procedure which allows intentional variation of low molecular mass constituents in the mitochondrial matrix [29]. Several preliminary experiments were carried out to test the suitability of these approaches and to exclude possible nonspecific effects on complex I- and DLDH-mediated activities. First, SMP-catalyzed NADH oxidase and NADH-supported superoxide production was titrated by NADH-OH (Fig. 1). Either activity was equally prevented upon NADH-OH increase, showing the titration pattern characteristic for a single site-directed high affinity (Ki ≤enzyme concentration) inhibition. Table 1 depicts the effects of NADH-OH and NH4Cl on the specific H2O2/O2 generating activities of various preparations measured at optimal (50 μM) NADH concentration. Ammonium affected neither SMP-bound nor purified complex I-catalyzed reactions, whereas about 6-fold stimulation of DLDH-catalyzed activity was seen for either soluble purified enzyme or for permeabilized mitochondria. Interestingly, slight stimulation of SMP-catalyzed reaction by ammonium was observed (Table 1, sample 1). This stimulation was much more prominent (about 4fold) in the presence of NADH-OH, when complex I-mediated activity was prevented (sample 2). These observations suggested that bound DLDH is present in SMP preparations. Indeed, NADH:lipoamide oxidoreductase activity was detected in SMP. We were unable to remove the residual NADH-supported, NADH-OH insensitive, ammonium stimulated hydrogen peroxide generating activity by repeated washing of SMP under various conditions. Studies on DLDH which is present in the inner mitochondrial membranes were not further elaborated in the present studies and are under current investigation in our laboratory. As shown in Table 1, ammonium did not affect complex I-mediated activity and NADH-

A.V. Kareyeva et al. / Biochimica et Biophysica Acta 1817 (2012) 1879–1885 Table 1 Hydrogen peroxide production by inside-out submitochondrial particles, soluble dihydrolipoamide dehydrogenase, purified complex I and permeabilized mitochondria. Enzyme preparation

1 2 3 4 5 6 7 8

SMP +NADH-OHc Purified complex I +NADH-OHc DLDH +NADH-OHc Permeabilized mitochondria +NADH-OHc

Specific rate of H2O2 formationa nmol/min/mg of protein (pH 7.5, 30 °C) − NH4Cl

+ NH4Clb

4.0 ± 0.3 0.36 ± 0.1 15.8 ± 0.4 1.0 ± 0.2 100 ± 10 100 ± 10 5.4 ± 0.5 2.9 ± 0.3

5.3 ± 0.3 1.7 ± 0.1 15.8 ± 0.4 1.0 ± 0.2 610 ± 46 610 ± 46 22.0 ± 0.6 21.5 ± 0.6

a

NADH (50 μM) as the substrate. 30 mM NH4Cl was added to the standard assay mixture. c Preincubated with NADH-OH (5 nmol/mg of protein) at ambient temperature for 1 min. b

OH did not inhibit DLDH-catalyzed reaction. Next, the NADH-OH-sensitive fraction of hydrogen peroxide producing activity of SMP (complex Imediated activity) was measured as a function of pool (NAD+ +NADH) concentration with NADH only and at two different NAD+/NADH ratios (Fig. 2A). Qualitatively similar behavior was seen at any NAD+/NADH ratio: the dependences exhibited an optimum in the range of 50–100 μM of the nucleotide and showed strong sensitivity to the product/substrate ratio. This is more clearly seen from the data presented in Fig. 2B, where the activity was measured as a function of NAD+/NADH ratio at optimal (50 μM, curve 1) and high (500 μM, curve 2) pool concentrations. The halfmaximal activities observed upon variation of NAD+/NADH ratio (externally poised redox potential) were different for optimal and high (NAD+ +NADH) pool (0.2 and 0.05, respectively). The results of the same experimental series for the soluble DLDH are depicted in Fig. 3. To simplify comparison, only the data obtained in the presence of activating ammonia are shown, because the same stimulatory effect (about six-fold increase in the presence of 30 mM NH4Cl) was seen at any pool or fractional NADH and NAD + concentration, although in agreement with our previous data [25] the apparent Km for DLDH NADH-supported activity was slightly changed (6 and 17 μM in the absence and presence of ammonium,

1881

respectively). Essentially the same as for complex I activity, patterns of the NAD +/NADH ratio dependences were obtained at different nucleotide pool concentrations (50 and 500 μM) with half-maximal activity corresponding to 0.17 and 0.06 for NAD +/NADH ratios, respectively. It has been shown that the rates of ROS production by mitochondria [2], SMP [23], and complex I [21] linearly depend on oxygen concentration. The results presented in Fig. 4 show that this is also true for the DLDHcatalyzed reaction. The following experimental approach was used. Because succinate oxidation catalyzed by SMP proceeds as a zero-order reaction of oxygen consumption, its residual concentration during succinate oxidation is strictly linearly dependent on time as shown in Fig. 4A (line 1). The time course of H2O2 formation by DLDH added to the mixture containing SMP oxidizing succinate in the presence of rotenone and NADH-OH (added to avoid complex I-mediated reactions) was followed (Fig. 4A, curve 2). Because oxygen consumption capacity of succinate oxidase of SMP is much higher than that of added DLDH, the amount of oxygen consumed in the H2O2-producing reaction could be neglected. When the rate of DLDH-catalyzed H2O2 production was plotted as a function of the residual oxygen concentration (Fig. 4B), a straight line corresponding to a first-order reaction was obtained. If the initial rate of DLDH-catalyzed reaction was measured in reaction mixture bubbled with pure oxygen for 10 min, the activity significantly increased, reasonably corresponding to first-order dependence. Next, the dependences of NADH-supported H2O2 production on pool concentration and fractional reduction of nucleotides were determined for permeabilized mitochondria where both “major” generators are present within their natural environment. Fig. 5 shows the relative contribution of complex I and DLDH to the overall hydrogen peroxide formation in the absence of stimulatory ammonium. At optimal NADH concentration, mitochondria produced 6 nmol/min/mg of H2O2, a value which was gradually decreased upon increase in NADH concentration. About half of the total production resulted from DLDH activity at any nucleotide pool concentration (as evident from partial inhibition by NADH-OH). When DLDH was stimulated by ammonium, the total activity increased 5-fold and only slight inhibitory effect of NADH-OH was seen (Fig. 5B). No significant difference in the nucleotide redox state dependence of non-stimulated and ammonium-stimulated reactions was observed. The data on sensitivity of complex I- and DLDHmediated NADH-supported hydrogen peroxide production to the

Fig. 2. Nucleotide concentrations and their redox state dependences of NADH-supported hydrogen peroxide production by SMP-bound complex I. Reactions were initiated by addition of SMP (100 μg/ml) to standard reaction mixture supplemented with rotenone (5 μM). (A) Total added nucleotide concentrations are indicated on the abscissa; figures on the curves are NAD+/NADH ratios. The NADH-OH insensitive activities of SMP preincubated with NADH-OH (5 nmol/mg protein) (about 10% of the total, see Table 1) were subtracted. (B) Redox state dependence of reactions measured at optimal (50 μM, open circles) and high (500 μM, filled circles) nucleotide pool.

1882

A.V. Kareyeva et al. / Biochimica et Biophysica Acta 1817 (2012) 1879–1885

Fig. 3. Nucleotide concentrations (A) and their redox state dependences (B) of ammonium-stimulated hydrogen peroxide production by soluble dihydrolipoamide dehydrogenase. Reactions were initiated by addition of DLDH (2 μg/ml). Figures on curves are NAD+/NADH ratios. Open and filled symbols in (B) correspond to pool concentrations of (NAD+ + NADH) of 50 and 500 μM, respectively.

nucleotide pool and its redox state are summarized in Table 2. Some data reported by other groups are also presented for comparison. 4. Discussion When no electron acceptors other than oxygen are present, both NADH-reduced membrane-bound complex I and soluble DLDH produce O2/H2O2 and show qualitatively similar and expected kinetic pattern that is non-hyperbolically dependency on NADH concentration. The turnover numbers in oxygen reduction for either enzyme are several orders of magnitude lower than those in their major catalytic activities,

reversible NADH:ubiquinone and NADH: lipoamide oxidoreductase reactions, respectively. Thus, it is safe to assume that at any NADH and NAD+ concentrations all enzyme–substrate (product) complexes are in true equilibrium. In the simplest model, if only a single oxygen-reactive enzyme-bound species and a single NADH (NAD +) binding site exist and concentration of NAD + is negligible, at least two reduced forms of the enzyme are expected to be present at any NADH concentration, those being free reduced protein and reduced protein–NADH complexes. Their particular concentrations are not simple hyperbolic functions of NADH concentration. Relative affinities of NADH to oxidized and reduced enzyme as well as intermolecular equilibrium

Fig. 4. Oxygen dependence of NADH-supported, ammonium-stimulated hydrogen peroxide formation by soluble dihydrolipoamide dehydrogenase. (A) Actual tracings of fumarate formation (curve 1) and H2O2 production by a mixture of SMP and soluble DLDH (curve 2). The standard reaction mixture contained no sucrose and was supplemented with rotenone (5 μM), NADH-OH (5 nmol/mg protein), NADH (200 μM), ammonium chloride (30 mM), ethanol (100 mM), alcohol dehydrogenase (60 units/ml), semicarbazide (10 mM), and gramicidin (0.05 μg/ml). A mixture of SMP (65 μg/ml), DLDH (2 μg/ml), and succinate (5 mM) was added, the cuvette was covered to avoid oxygen diffusion (no air was left above the solution) and zero-order formation of fumarate was followed as an increase in absorption at 278 nm (line 1). After all oxygen was consumed as was evident from abrupt termination of succinate oxidation, the same experiment was carried out in the presence of hydrogen peroxide registration system (Amplex Red, horseradish peroxidase, see Material and Methods section), and resorufin formation was recorded (curve 2). Control experiments with no added DLDH were performed to take into account hydrogen peroxide formation by SMP (about 15% of that catalyzed by the soluble added DLDH) and SMP-catalyzed reaction (curve 3) was subtracted from the total (SMP + DLDH) one. (B) First derivative of difference between curve 2 and curve 3 (panel A) (rate of DLDH-catalyzed H2O2 production) as a function of oxygen concentration, calculated from curve 1 (panel A). Point at 750 μM oxygen corresponds to the initial rate of H2O2 formation measured in the reaction mixture bubbled with pure oxygen for 10 min. The actual oxygen concentration in this sample was determined as shown in panel A, line 1, as the amount of fumarate formed during complete oxygen consumption.

A.V. Kareyeva et al. / Biochimica et Biophysica Acta 1817 (2012) 1879–1885

1883

Fig. 5. Nucleotide concentration dependences of NADH-supported hydrogen peroxide production by permeabilized mitochondria in the absence (A) or in the presence (B) of 30 mM ammonium. Standard reaction mixture was supplemented with rotenone (5 μM). NADH-OH (5 nmol/mg protein) was present where indicated (open symbols). Rat heart mitochondria (25 μg protein/ml) were preincubated with alamethicin (20 μg/ml) and 5 mM MgCl2 for 1 min. The reactions were initiated by simultaneous addition of NADH and NAD+. Figures on the curves are NAD+/NADH ratios.

between protein-bound NADH and redox cofactor are the parameters that set fractional content of the reduced enzyme species. It seems likely that their oxygen reactivities are more or less different for different reduced enzyme forms. Although this study was not intended to analyze the mechanism of O2/H2O2 production by complex I and DLDH, the shapes of the curves (Figs. 2A and 3A) for NADH-supported hydrogen peroxide formation by either enzyme merit brief comments. The most likely oxygen-reactive protein-bound component is fully reduced flavin (FMNH− in complex I and FADH− in DLDH) [21,23], although flavosemiquinone and ubisemiquinone have also been suggested as superoxide-producing species in complex I [22]. In contrast to auto-oxidizable free reduced flavins [38,39], flavoproteins are remarkable in their extremely variable oxygen reactivity and product species (superoxide anion and hydrogen peroxide) [39–41]. Isolated mammalian complex I have been reported to produce only superoxide [21], whereas the membrane-bound complex (SMP) [42] and DLDH [43] are capable of H2O2 and superoxide production. The first-order nature of oxygen reduction catalyzed by isolated complex I [21], SMP [23], mitochondria [2,44], and DLDH (Fig. 4B)

suggests that no specific oxygen-binding sites are involved in mitochondrial one- or two-electron reduction of oxygen. If flavin or its oxygen-reactive site (4a position in the isoalloxazine ring) is not exposed to the solvent, the first-order oxidation rate constant reflects a diffusion of oxygen through a single channel or several “channels” arising due to the protein dynamics. Either “pathway” for oxygen may be different in free and substrate-loaded reduced protein. Simultaneous formation of superoxide anion and hydrogen peroxide [42] suggests that no unique mechanism of oxygen reduction operates during the reaction. Interestingly, the inhibition of complex I-catalyzed hydrogen peroxide formation by high NADH concentrations (Fig. 2A), (also seen for DLDH, Fig. 3A) is significantly less pronounced than that previously reported for its superoxide generation activity [14]. At present no simple explanation for this puzzling phenomenon can be offered. Both complex I and DLDH activities are very sensitive to NAD +/ NADH ratio (Table 2). Half-maximal activities at 50 μM pool concentration were observed at the ratios of 0.2 and 0.17 for complex I and DLDH, respectively. As these “half-maximal” ratios depend on the total

Table 2 Mitochondrial hydrogen peroxide production at variable pool and redox state of pyridine nucleotides. Enzyme system

Maximal rate observeda (2 electron-equivalents/min/mg) · 109

(NAD+ + NADH) pool μM

Complex I (SMP)

Complex I (bovine heart SMP) Rat heart mitochondria

2.8 1.9 115 100 650 660 7.2 6.7 26 33 0.5 0.2

Complex I (rat brain SMP) Purified bovine heart complex I

Not reported 25

50 500 50 500 50 500 50 500 50 500 Variable Intramitochondrial nucleotides buffered by acetoacetate/oxobutyrate couple 1000 Variable 30–200 μM

DLDH (pig heart soluble enzyme) +NH4Cl Rat heart permeabilized mitochondria +NH4Cl

a b c

At optimal (30–50 μM) NADH concentration. Ratio corresponding to half maximal activity. Calculated from the original data published as activity versus NADH:NAD+ redox potential graphs.

(NAD+/NADH)0.5 ratiob 0.20 0.05 0.17 0.06 0.16 0.06 0.07 0.07 0.10 0.05 0.2 0.01c 15c 0.1c

Reference This paper This paper

This paper

[13] [36] [37] [21]

1884

A.V. Kareyeva et al. / Biochimica et Biophysica Acta 1817 (2012) 1879–1885

pool concentration, they are certainly not true redox potential of the protein-bound flavin (as one would estimate using the Nernst equation and well defined NAD+/NADH couple redox potential); rather, as noted above, they are complex functions of the nucleotides binding constants and intermolecular redox equilibrium. The “redox state buffer” dependence of complex I-mediated reaction has been previously reported for mitochondria from rat heart [36], rat brain [37], bovine heart submitochondrial particles [13], and isolated bovine heart complex I [21]. Intramitochondrial NAD+/NADH ratio is evidently a crucial parameter for superoxide/hydrogen peroxide activity. Average content of NADH+ NAD + in heart mitochondria (values reported by different groups using different extraction procedure [45–47]) is in a range of 5–7 nmol/mg protein. The water content in the mitochondrial matrix space is variable being dependent on osmolarity, ionic composition, and metabolic state. About 1 μl of water/mg of mitochondrial protein is an average value reported (with considerable variation) by different groups [48–50]. This corresponds to the molar nucleotide pool concentration of 5–7 mM. A considerable part of the intramitochondrial NADH (and NADPH) is protein-bound, as evident from the strong increase in the fluorescence quantum yield, blue shift of its maximum [45,51], and heterogeneous distribution of mitochondrial fluorescence lifetimes [52,53]. We could not find data on partial distribution of free and protein-bound NAD+ in the mitochondrial matrix, but it seems likely that in intact mitochondria complex I is always saturated by its substrate/product. The steady-state level of 1.7 for the mitochondrial NAD +/NADH ratio in starved rat liver cells incubated with 5 mM lactate has been reported [54]. There are no reasons to believe that this ratio is considerably lower in heart. Indeed, a significant fraction of pyridine nucleotides are oxidized in vivo in brain and kidney [51], resting myocytes [55], and in perfused heart [56] under normoxic conditions. With an assumption that nucleotide-binding sites of complex I and DLDH are in mitochondria saturated by the NAD +/NADH couple, our data on hydrogen peroxide generation by permeabilized mitochondria in the absence of stimulating ammonia (Fig. 5 and Table 2) suggest that the mitochondrial H2O2 production in heart under normoxic metabolically active physiological conditions proceeds at the rate that is less than 20% of their maximal capacity. Mitochondria are a source and a sink for intracellular hydrogen peroxide [57]. Their contribution to the overall production is undoubtedly tissue-specific, and as has recently been pointed out [58], the widely circulated in current literature statement that “mitochondria are the major source of cellular ROS production” should be considered with great precautions. Availability of oxygen is the other crucial factor for both complex I and DLDH hydrogen peroxide production due to their apparent firstorder dependences (see Refs. [21,23] for complex I). Considering the physiological significance of mitochondrial hydrogen peroxide production, a paradox is evident: mitochondria are capable of significant ROS generation only when their pyridine nucleotide pool is highly reduced; on the other hand, this state arises only under strongly hypoxic conditions when ROS production is expected to be significantly decreased due to the first-order nature of oxygen reactivity. A burst of mitochondrial ROS production is, however, expected during anoxia/normoxia transition when anaerobically deactivated complex I [59] is only slowly acquiring its NADH oxidizing capacity. The last point to be briefly discussed concerns the nature and physiological relevance of DLDH stimulation by ammonium (Refs [24,25] and Fig. 5). More than half of the overall hydrogen peroxide formation by permeabilized mitochondria is catalyzed by DLDH in the absence of ammonium, and this proportion is not significantly changed upon variation of the pool of nucleotides (Fig. 5A). Most of the ammonium stimulated overall production (about 5-fold increased) is catalyzed by DLDH (almost no sensitivity to NADH-OH was seen (Fig. 5B)). The apparent affinity of DLDH for ammonium in its stimulatory effect is quite low (higher than millimolar range

[25]). On the other hand, the stimulation is highly specific: guanidine chloride (10 mM), methyl-, dimethyl-, trimethyl-amines (30 mM each), lysine, arginine, and spermine (10 mM each) do not stimulate DLDH-catalyzed hydrogen peroxide formation (negative results from this laboratory, not shown). It worth noting that we were unable to find in the literature enzymes involved in nitrogen metabolism that have high affinity for ammonium. Whether stimulation of DLDH by ammonium is physiologically relevant would depend on permeability of the inner mitochondrial membrane for ammonium cation or ammonia. Classical data on large amplitude swelling of mitochondria suggest that ammonia diffuses through the inner mitochondria membrane [60]. This does not exclude a possibility that a specific carrier for ammonium exists in mitochondria. Recently aquaporin-8 (AQP8) has been shown to facilitate mitochondrial ammonia transport [61]. Whatever the actual mechanism of mitochondrial ammonium transport is, a possibility of significant concentration gradient between intra- and extramitochondrial compartments cannot be ruled out. Acknowledgements This study was supported by grant 11-04-00916 to ADV and grant 1204-00602 to VGG of the Russian Foundation for Fundamental Research. We are grateful to Dr. G. Cecchini for his kind gift of NADH-OH. Linguistic help of Dr. Richard Lozier in preparation of the manuscript is gratefully acknowledged. References [1] G. Loschen, L. Flohé, B. Chance, Respiratory chain linked H2O2 production in pigeon heart mitochondria, FEBS Lett. 18 (1971) 261–264. [2] A. Boveris, B. Chance, The mitochondrial generation of hydrogen peroxide. General properties and effect of hyperbaric oxygen, Biochem. J. 134 (1973) 707–716. [3] B. Chance, H. Sies, A. Boveris, Hydroperoxide metabolism in mammalian organs, Physiol. Rev. 59 (1979) 527–605. [4] E.B. Tahara, F.D. Navarete, A.J. Kowaltowski, Tissue-, substrate-, and site-specific characteristics of mitochondrial reactive oxygen species generation, Free Radic. Biol. Med. 46 (2009) 1283–1297. [5] A.J. Lambert, H.M. Boysen, J.A. Buckingham, T. Yang, A. Podlutsky, S.N. Austad, T.H. Kunz, R. Buffenstein, M.D. Brand, Low rates of hydrogen peroxide production by isolated heart mitochondria associate with long maximum lifespan in vertebrate homeotherms, Aging Cell 6 (2007) 607–618. [6] J. St-Pierre, J.A. Buckingham, S.J. Roebuck, M.D. Brand, Topology of superoxide production from different sites in the mitochondrial electron transport chain, J. Biol. Chem. 277 (2002) 44784–44790. [7] A. Boveris, N. Oshino, B. Chance, The cellular production of hydrogen peroxide, Biochem. J. 128 (1972) 617–630. [8] S.S. Korshunov, V.P. Skulachev, A.A. Starkov, High protonic potential actuates a mechanism of production of reactive oxygen species in mitochondria, FEBS Lett. 416 (1997) 15–18. [9] A.J. Lambert, J.A. Buckingham, M.D. Brand, Dissociation of superoxide production by mitochondrial complex I from NAD(P)H redox state, FEBS Lett. 582 (2008) 1711–1714. [10] P.K. Jensen, Antimycin-insensitive oxidation of succinate and reduced nicotinamide-adenine dinucleotide in electron-transport particles, Biochim. Biophys. Acta 122 (1966) 157–166. [11] J.F. Turrens, A. Boveris, Generation of superoxide anion by the NADH dehydrogenase of bovine heart mitochondria, Biochem. J. 191 (1980) 421–427. [12] K. Takeshige, S. Minakami, NADH- and NADPH-dependent formation of superoxide anions by bovine heart submitochondrial particles and NADH-ubiquinone reductase preparation, Biochem. J. 180 (1979) 129–135. [13] A.I.G. Krishnamoorthy, P. Hinkle, Studies on the electron transfer pathway, topography of iron–sulfur centers, and site of coupling in NADH-Q oxidoreductase, J. Biol. Chem. 263 (1988) 17566–17575. [14] V.G. Grivennikova, A.D. Vinogradov, Generation of superoxide by the mitochondrial complex I, Biochim. Biophys. Acta 1757 (2006) 553–561. [15] V.G. Grivennikova, A.B. Kotlyar, J.S. Karliner, G. Cecchini, A.D. Vinogradov, Redox-dependent change of nucleotide affinity to the active site of the mammalian complex I, Biochemistry 46 (2007) 10971–10978. [16] A.B. Kotlyar, J.S. Karliner, G. Cecchini, A novel strong competitive inhibitor of complex I, FEBS Lett. 579 (2005) 4861–4866. [17] J. Sun, B. Trumpower, Superoxide anion generation by the cytochrome bc1 complex, Arch. Biochem. Biophys. 419 (2003) 198–206. [18] M.Yu. Ksenzenko, A.A. Konstantinov, G.B. Khomutov, A.N. Tikhonov, E.K. Ruuge, Effect of electron transfer inhibitors on superoxide generation in the cytochrome bc1 site of the mitochondrial respiratory chain, FEBS Lett. 155 (1983) 19–24.

A.V. Kareyeva et al. / Biochimica et Biophysica Acta 1817 (2012) 1879–1885 [19] Q. Chen, E.J. Vazquez, S. Moghaddas, C.L. Hoppel, E.J. Lesnefsky, Production of reactive oxygen species by mitochondria. Central role of complex III, J. Biol. Chem. 278 (2003) 36027–36031. [20] A. Galkin, U. Brandt, Superoxide radical formation by pure complex I (NADH: ubiquinone oxidoreductase) from Yarrowia lipolytica, J. Biol. Chem. 280 (2005) 30129–30135. [21] L. Kussmaul, J. Hirst, The mechanism of superoxide production by NADH: ubiquinone oxidoreductase (complex I) from bovine heart mitochondria, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 7607–7612. [22] S.T. Ohnishi, K. Shinzawa-Itoh, K. Ohta, S. Yoshikawa, T. Ohnishi, New insights into the superoxide generation sites in bovine heart NADH:ubiquinone oxidoreductase (complex I): the significance of protein-associated ubiquinone and dynamic shifting of generation sites between semiflavin and semiquinone radicals, Biochim. Biophys. Acta 1797 (2010) 1901–1909. [23] A.D. Vinogradov, V.G. Grivennikova, Generation of superoxide-radical by the NADH:ubiquinone oxidoreductase of heart mitochondria, Biochem. Mosc. 70 (2005) 120–127. [24] V.G. Grivennikova, G. Cecchini, A.D. Vinogradov, Ammonium-dependent hydrogen peroxide production by mitochondria, FEBS Lett. 583 (2008) 1287–1291. [25] A.V. Kareyeva, V.G. Grivennikova, G. Cecchini, A.D. Vinogradov, Molecular identification of the enzyme responsible for the mitochondrial NADH-supported ammonium-dependent hydrogen peroxide production, FEBS Lett. 585 (2011) 385–389. [26] A.A. Starkov, G. Fiskum, C. Chinopoulos, B.J. Lorenzo, S.E. Browne, M.S. Patel, M.F. Beal, Mitochondrial alpha-ketoglutarate dehydrogenase complex generates reactive oxygen species, J. Neurosci. 24 (2004) 7779–7788. [27] L. Tretter, V. Adam-Vizi, Generation of reactive oxygen species in the reaction catalyzed by α-ketoglutarate dehydrogenase, J. Neurosci. 24 (2004) 7771–7778. [28] V.G. Grivennikova, A.V. Kareyeva, A.D. Vinogradov, What are the sources of hydrogen peroxide production by heart mitochondria? Biochim. Biophys. Acta 1797 (2010) 939–944. [29] I.S. Gostimskaya, V.G. Grivennikova, T.V. Zharova, L.E. Bakeeva, A.D. Vinogradov, In situ assay of the intramitochondrial enzymes: use of alamethicin for permeabilization of mitochondria, Anal. Biochem. 313 (2003) 46–52. [30] W.E. Jacobus, V.A. Saks, Creatine kinase of heart mitochondria: changes in its kinetic properties induced by coupling to oxidative phosphorylation, Arch. Biochem. Biophys. 219 (1982) 167–178. [31] A.B. Kotlyar, A.D. Vinogradov, Slow active/inactive transition of the mitochondrial NADH-ubiquinone reductase, Biochim. Biophys. Acta 1019 (1990) 151–158. [32] D.S. Burbaev, I.A. Moroz, A.B. Kotlyar, V.D. Sled, A.D. Vinogradov, Ubisemiquinone in the NADH-ubiquinone reductase region of the mitochondrial respiratory chain, FEBS Lett. 254 (1989) 47–51. [33] Y. Hatefi, Preparation and properties of NADH: ubiquinone oxidoreductase (complex I), Methods Enzymol. 53 (1978) 11–14. [34] M. Zhou, Z. Diwu, N. Panchuk-Voloshina, R.P. Haugland, A stable nonfluorescent derivative of resorufin for the fluorometric determination of trace hydrogen peroxide: applications in detecting the activity of phagocyte NADPH oxidase and other oxidases, Anal. Biochem. 253 (1997) 162–168. [35] M.I. Dolin, Effect of near ultraviolet irradiation on the peroxide content of solutions of oxidized or reduced diphosphopyridine nucleotide, Biochim. Biophys. Acta 63 (1962) 219–221. [36] Y. Kushnareva, A.N. Murphy, A. Andreyev, Complex I-mediated reactive oxygen species generation: modulation by cytochrome c and NAD(P)+ oxidation–reduction state, Biochem. J. 368 (2002) 545–553. [37] A.P. Kudin, N.Y. Bimpong-Buta, S. Vielhaber, C.E. Elger, W.S. Kunz, Characterization of superoxide-producing sites in isolated brain mitochondria, J. Biol. Chem. 279 (2004) 4127–4135. [38] Q.H. Gibson, S.W. Hastings, The oxidation of reduced flavin mononucleotide by molecular oxygen, Biochem. J. 83 (1962) 368–377. [39] V. Massey, Activation of molecular oxygen by flavins and flavoproteins, J. Biol. Chem. 269 (1994) 22459–22462.

1885

[40] M. Dixon, The acceptor specificity of flavins and flavoproteins. III. Flavoproteins, Biochim. Biophys. Acta 226 (1971) 269–284. [41] A. Mattevi, To be or not to be an oxidase: challenging the oxygen reactivity of flavoenzymes, Trends Biochem. Sci. 31 (2006) 276–283. [42] T.Y. Pepelina, R.V. Chertkova, T.V. Ostroverkhova, D.A. Dolgikh, M.P. Kirpichnikov, V.G. Grivennikova, A.D. Vinogradov, Site-directed mutagenesis of cytochrome c: reactions with respiratory chain components and superoxide radical, Biochem. Mosc. 74 (2009) 625–632. [43] V. Massey, S. Strickland, S.G. Mayhew, L.G. Howell, P.C. Engel, R.G. Matthews, M. Schuman, P.A. Sullivan, The production of superoxide anion radicals in the reaction of reduced flavins and flavoproteins with molecular oxygen, Biochem. Biophys. Res. Commun. 36 (1969) 891–897. [44] J.F. Turrens, B.A. Freeman, J.G. Levitt, J.D. Crapo, The effect of hyperoxia on superoxide production by lung submitochondrial particles, Arch. Biochem. Biophys. 217 (1982) 401–410. [45] Y. Avi-Dor, J.M. Olson, M.D. Doherty, N.O. Kaplan, Fluorescence of pyridine nucleotides in mitochondria, J. Biol. Chem. 237 (1962) 2377–2383. [46] R.L. Lester, Y. Hatefi, Studies on the mechanism of oxidative phosphorylation. IV. Pyridine nucleotide binding and its relation to activity in heart mitochondria, Biochim. Biophys. Acta 29 (1958) 103–112. [47] K.B. Jacobson, N.O. Kaplan, Pyridine coenzymes of subcellular tissue fractions, J. Biol. Chem. 226 (1957) 603–613. [48] E. Pfaff, M. Klingenberg, E. Ritt, W. Vogell, Correlation of the unspecific permeable mitochondrial space with the “intermembrane space”, Eur. J. Biochem. 5 (1968) 222–232. [49] A.P. Halestrap, P.T. Quinlan, The intramitochondrial volume measured using sucrose as an extramitochondrial marker overestimates the true matrix volume determined with mannitol, Biochem. J. 214 (1983) 387–393. [50] E.J. Harris, K. van Dam, Changes of total water and sucrose space accompanying induced ion uptake or phosphate swelling of rat liver mitochondria, Biochem. J. 106 (1968) 759–766. [51] B. Chance, P. Cohen, F. Jobsis, B. Schoener, Intracellular oxidation–reduction states in vivo, Science 137 (1962) 499–508. [52] H.D. Vishwasrao, A.A. Heikal, K.A. Kasischke, W.W. Webb, Conformational dependence of intracellular NADH on metabolic state revealed by associated fluorescence anisotropy, J. Biol. Chem. 280 (2005) 25119–25126. [53] K. Blinova, S. Carroll, S. Bose, A.V. Smirnov, J.J. Harvey, J.R. Knutson, R.S. Balaban, Distribution of mitochondrial NADH fluorescence lifetimes: steady-state kinetics of matrix NADH interactions, Biochemistry 44 (2005) 2585–2594. [54] M.E. Tischler, D. Friedrichs, K. Coll, J.R. Williamson, Pyridine nucleotide distributions and enzyme mass action ratios in hepatocytes from fed and starved rats, Arch. Biochem. Biophys. 184 (1977) 222–236. [55] J. Eng, R.M. Lynch, R.S. Balaban, Nicotinamide adenine dinucleotide fluorescence spectroscopy and imaging of isolated cardiac myocytes, Biophys. J. 55 (1989) 621–630. [56] R. Araki, M. Tamura, I. Yamazaki, The effect of intracellular oxygen concentration on lactate release, pyridine nucleotide reduction, and respiration rate in the rat cardiac tissue, Circ. Res. 53 (1983) 448–455. [57] A.A. Starkov, The role of mitochondria in reactive oxygen species metabolism and signaling, Ann. N. Y. Acad. Sci. 1147 (2008) 37–52. [58] G.C. Brown, V. Borutaite, There is no evidence that mitochondria are the main source of reactive oxygen species in mammalian cells, Mitochondrion 12 (2012) 1–4. [59] E. Maklashina, Y. Sher, H.Z. Zhou, M.O. Gray, J.S. Karliner, G. Cecchini, Effect of anoxia/reperfusion on the reversible active/de-active transition of NADH-ubiquinone oxidoreductase (complex I) in rat heart, Biochim. Biophys. Acta 1556 (2002) 6–12. [60] J.B. Chappell, K.N. Haarhoff, The penetration of the mitochondrial membrane by anions and cations, in: E.C. Slater, Z. Kaniuga, L. Wojtchak (Eds.), Biochemistry of Mitochondria, Acad. Press, London, 1967, pp. 75–91. [61] L.R. Soria, E. Fanelli, N. Altamura, M. Svelto, R.A. Marinelli, G. Calamita, Aquaporin-8-facilitated mitochondrial ammonia transport, Biochem. Biophys. Res. Commun. 393 (2010) 217–221.