Electrochemical study of hydrogen peroxide formation in isolated mitochondria

Bioelectrochemistry 85 (2012) 21–28

Contents lists available at SciVerse ScienceDirect

Bioelectrochemistry journal homepage: www.elsevier.com/locate/bioelechem

Electrochemical study of hydrogen peroxide formation in isolated mitochondria Raluca Marcu a, 1, 2, Stefania Rapino a, 2, Mirella Trinei a, Giovanni Valenti b, Massimo Marcaccio b, Pier Giuseppe Pelicci a, Francesco Paolucci b, Marco Giorgio a,⁎ a b

Experimental Oncology Dept. European Institute of Oncology, Via Adamello 16, 20139 Milan, Italy Chemistry Dept. “G. Ciamician”, University of Bologna, Via Selmi 2, 40126 Bologna, Italy

a r t i c l e

i n f o

Article history: Received 13 July 2011 Received in revised form 16 November 2011 Accepted 22 November 2011 Available online 3 December 2011 Keywords: Platinized carbon fiber microelectrodes Amperometry Hydrogen peroxide Mitochondria

a b s t r a c t Mitochondrial respiration generates reactive oxygen species that are involved in physiological and pathological processes. The majority of methods, with exception of electron paramagnetic resonance, used to evaluate the identity, the rate and the conditions of the reactive oxygen species produced by mitochondria, are mainly based on oxidation sensitive markers. Following latest electrochemical methodology, we implemented a novel electrochemical assay for the investigation of aerobic metabolism in preparations of isolated mitochondria through simultaneous measurement of O2 consumption and reactive species production. This electrochemical assay reveals active H2O2 production by respiring mouse liver mitochondria, and shows that ATP synthase activation and moderate depolarization increase the rate of H2O2 formation, suggesting that ATP synthesizing (state 3) mitochondria might contribute to oxidative stress or signaling. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Incomplete reduction of O2 during mitochondrial electron transfer chain (ETC) reactions generates superoxide anion, O2• − that is dismutated by SODs into hydrogen peroxide, H2O2. H2O2 is either degraded inside mitochondria by glutathione peroxidases, peroxiredoxins and catalases or can diffuse outside mitochondria being removed by extramitochondrial scavenging reactions [1,2]. Despite their toxicity, accumulating evidence support a physiological role for reactive oxygen species (ROS) (including: O2• −, H2O2, and hydroxyl radical, OH•) in a variety of cellular processes, such as growth factors/hormones/cytokines intracellular signaling, genes expression regulation and programmed cell death (apoptosis) [3,4]. Among ROS, H2O2 is the best suited to function as a signaling agent since it is diffusible, less reactive and longer-lived than, for instance, O2• − and OH•, and it is therefore involved in several transduction pathways [5]. Since the rate of mitochondrial ROS formation is modulated, roughly speaking, by O2 availability, substrate supply and ATP synthesis [6,7], the emerging picture is that mitochondria

Abbreviations: ETC, electron transfer chain; ROS, reactive oxygen species; SOD, superoxide dismutase; PCE, platinized carbon fiber microelectrodes; RNS, reactive nitrogen species; FCCP, p-trifluoromethoxy carbonyl cyanide phenyl hydrazone. ⁎ Corresponding author. Tel.: + 39 02 94375040; fax: + 39 02 94375990. E-mail address: [email protected] (M. Giorgio). 1 Present address: Mitochondria and Metabolism Center, 815 Mercer Street, Seattle, USA. 2 These authors contributed equally to this work. 1567-5394/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.bioelechem.2011.11.005

generate ROS in a regulated manner, such as to reflect the metabolic activity of the cell and/or to act as an oxygen sensor involved in the transduction of hypoxic signals [8,9]. Nevertheless, assessing the rate of mitochondrial ROS, and in particular H2O2 production, is a complicated issue due in part to inherent difficulties of available technologies for measuring these molecules. Except for electron paramagnetic spin resonance that can directly detect free radicals with unpaired electrons [10], all other methods for ROS detection are indirect. They measure either the reaction product between ROS and various probes or the fingerprints of oxidative stress on different endogenous molecules (DNA, lipids, proteins, low-molecular-mass antioxidants) [11]. In addition, most traditional assays for H2O2 measurement relay on fluorescent probes that often lack specificity and may result in artefactual signals [12–14]. In the search for other assays to measure ROS, electrochemistry may represent a valid alternative. Oxygen and its reactive species are particularly suited for these kinds of measurements since they can easily exchange electrons at an appropriate potential. Recently, reactive species produced by living cells in response to mechanical challenges have been measured directly and selectively through electrochemistry at platinized carbon fiber microelectrodes (PCE) [15,16]. To overcome the limitations of current methodologies for mitochondrial H2O2 measurement, we implemented an electrochemical assay for the simultaneous evaluation of O2 consumption metabolism and H2O2 production in suspensions of isolated mitochondria. Here we show the results obtained with this approach on mouse liver mitochondria preparations in the presence of different energetic substrates, respiratory inducers and inhibitors.

22

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28

2. Material and methods 2.1. Preparation of isolated mitochondria Mitochondria were purified from mouse liver through standard differential centrifugation method. 129Sv 3–6 months old male mice were killed by cervical dislocation, the liver was rapidly explanted and homogenized in ice-cold MTC isolation buffer (250 mM sucrose, 10 mM Tris HCl, 1 mM EGTA, pH 7.4). The liver homogenate was centrifuged for 10 min at 600 g for the removal of nuclei and cell debris and the supernatant centrifuged at 7000 g for 10 min. The mitochondria containing pellet was washed twice and finally resuspended in a minimal volume of MTC isolation buffer. All steps were performed at 4 °C. Protein concentration was measured using the biuret method. Experiments were performed in MTC assay buffer (125 mM KCl, 10 mM MOPS-Tris, 1 mM inorganic phosphate, 100 μM EGTA, pH 7.4).

2.2. Instrumentation for electrochemical measurements of O2 metabolism Simultaneous measurements of O2 consumption and H2O2 production by isolated mitochondria were achieved integrating an oxygraph apparatus (Hansatech Instruments, Norfolk, England) with a system for the simultaneous detection of H2O2 (Fig. 1A). The electrochemical measurement of O2 was performed at the Clark type electrode by reducing O2 at a reduction potential of −0.7 V vs. Ag/AgCl reference electrode. For H2O2 detection, the chamber of the oxygraph was adapted in order to allocate a 10 μm PCE, an Ag/AgCl reference electrode and a Pt counter electrode. Electrochemical detection of H2O2 was performed by oxidizing this molecule at an oxidation potential of 0.6 V vs. Ag/AgCl reference electrode provided by a bipotentiostat (CH Instruments, Austin, TX). The choice of using 0.6 V to detect H2O2 was based on the fact that, although the plateau for H2O2 oxidation was reached at 0.4 V in PBS solution [17], in the presence of the mitochondrial suspension we observed a maximal current response for H2O2 at 0.6 V vs. Ag/AgCl (Supplementary Fig. 1). The two electrochemical cells for O2 and, respectively, H2O2 detection were separated through a polytetrafluoroethylene (PTFE) membrane covering the Clark type electrode that assures the electrical insulation between the electrolytic solution above the Clark type electrode and the oxygraph chamber solution. The selectivity of the [O2] measurement, at Clark type electrode, is ensured by this design of the analytical: the Clark type electrode is separated in fact from the mitochondrial suspension by the O2 permeable and O2 selective PTFE membrane. For each measurement, we tested the integrity and impermeability of the PTFE shield verifying that the two electrochemical cells (one including the Clark type electrode and the other including the platinized carbon fiber/Pt CE / Ag/AgCl QRE) were isolated electrically. The reaction mixture was stirred by a magnetic follower that abolished local gradients of O2 and H2O2. The chamber was isolated from the external environment through an O2 impermeable plastic cap, without gas phase left between the assay solution and the cap. For the preparation of the PCE, briefly: carbon fibers (Cytec Carbon Fibers, Greenville) were aspired inside borosilicate glass capillaries (Clark Electromedical Instruments, 1.2 mm O.D. × 0.7 mm I.D.) and pulled with a micropipette puller (P-97, Sutter Instruments, U.S.A.) into two electrodes. The carbon fiber protruding from the glass was insulated by electrochemical deposition of a poly (oxy-phenylene) polymer. The surface of the tip was polished by grinding on 0.3 μm alumina paper and then platinized by reducing hydrogen hexachloroplatinate in the presence of lead acetate [15]. Each measurement was performed using a single PCE that was calibrated at the end with known concentrations of H2O2. Before measurements, the electrodes were left to equilibrate for 30 min in MTC assay buffer.

Fig. 1. Electrochemical instrumentation for simultaneous measurement of H2O2 and O2 A. Scheme of the apparatus used for the simultaneous detection of H2O2 and O2. B. Amperometric response of successive addition of H2O2 at PCE. A 0.6 V vs. Ag/AgCl potential was applied at PCE to oxidize H2O2. Measurements were performed in absence of catalase (black line), in presence of 20 nM catalase (blue line) and 200 nM catalase (red line). C. Simultaneous detection of O2 concentration (in the same chamber as B) by Clark type electrode. A − 0.7 vs. Ag/AgCl potential was applied to Clark electrode to reduce O2. The measurements were performed in 20 nM catalase (blue line), and 200 nM catalase (red line). The amperometric current values are converted in O2 concentration in the present plot. Where indicated by the arrows, known concentrations of H2O2 (0.5–5 μM) were added.

2.3. Data analysis for amperometric measurements of H2O2 Once a potential has been applied, PCEs undergo an equilibration phase due to diffusion effects, slow partial oxidation, and modification of the active sites of the platinized surface [18]. The outcome of these equilibration processes is a progressive and continuous reduction of the oxidation current registered at the electrode. The equilibration curve thus generated can be described by an exponential function (Eq. (1)) that we have used to extrapolate the baseline: y ¼ y0 þ A1  ð1−expð−x=t1 ÞÞ þ A2  ð1−expð−x=t2 ÞÞ

ð1Þ

A1, A2, t1, and t2 are dimensionless parameters, specific for every curve, that are determined through the fitting procedure. Thus, any specific electrical event can be revealed by subtracting the baseline from

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28

the total measured signal. In particular, addition of mitochondria induces further surface modification of the electrode, which undergoes another equilibration phase that is also accounted by the above fitting procedure. In addition, we have used appropriate correction factors in order to account for the loss of sensitivity of the platinized electrode during the measurements. The electrode was not affected by sucrose or different substrates or phosphate concentrations but some drugs used to modulate mitochondrial respiration interacted with the electrode, generating an instantaneous increase or drop in the electric current as soon as they were added, but without other effect on the signal measured next. In particular, drugs dissolved in ethanol (FCCP, rotenone, antimycin, oligomycin) produced similar increase in the oxidation current, most probably due to the oxidation and adsorption of ethanol to the platinum surface of the electrode. Other substances, like KCN, aminotriazole and tetramethyl-p-phenylenediamine/ascorbate instead had a strong direct effect on the electrode precluding their use during measurements. 2.4. Immunoblotting Samples from total liver homogenates and mitochondrial fractions were lysed on ice in the following solution: 20 mM Tris, 0.5% NaDOC, 1% Triton (0.5% for the lysis of mitochondria), 0.1% SDS, 150 mM NaCl, 1 mM EDTA and protease inhibitors (0.1 mM PMSF, 1 μg/ml leupeptin, 1 μg/ml aprotinin), pH 7.4. Equal amounts of proteins from each sample were first electrophoresed on polyacrylamide gels and then transferred onto nitrocellulose membranes. Membranes were blocked with 1% BSA in TBS-Tween and then probed with antibodies against SOD1, SOD2, GPx (Abcam), catalase (Biomol), and COX IV (Invitrogen).

23

that micromolar concentrations of H2O2 could be detected in an approximately linear response manner for the entire range of concentrations tested (0.5–5 μM). The H2O2 signal increased after each single addition of H2O2, and, in the absence of catalase, remained stable and summed upon sequential additions, indicating the accumulation of H2O2 in the assay chamber. The response of platinized carbon electrode to single additions of H2O2 was instantaneous, and reflected in real time the concentration of H2O2 in the assay chamber. This allowed us, for instance, to test the kinetics of H2O2 degradation by catalase, a reaction resulting in O2 production that was also kinetically measured by the Clark type electrode. In the presence of catalase, while still increasing after each addition, the H2O2 signal was not stable in time and rapidly disappeared, reflecting the specific scavenging of H2O2 by catalase (Fig. 1B, blue line). In contrast, fluorescent probes such as Amplex Red or dichlorofluorescein diacetate are irreversibly oxidized by H2O2, leading to a constantly increasing fluorescent signal (Supplementary Fig. 2), and therefore the dynamics of H2O2 levels due to the presence of scavenging enzymes cannot be appreciated [12,13]. Notably, the Clark type electrode, simultaneously to the decay of the PCE signal, revealed an O2 formation rate, according to a stoichiometry of 1 mol O2 generated for every 2 mol of H2O2 degraded, in agreement with the reaction mechanism of catalases [19] (Fig. 1C, blue line). As expected, the H2O2 scavenging by catalase increased when higher concentrations of this enzyme were used (Fig. 1B and C, red lines). These results validate the combination of a Clark type electrode and a PCE for the simultaneous detection of H2O2 and O2 in biological samples.

3.2. Electrochemical measurements reveal H2O2 scavenging in liver mitochondrial preparations

2.5. Enzymatic activity measurements Catalase and glutathione peroxidase activity in isolated mitochondria preparations were measured with the Amplex Red Catalase Assay Kit (Invitrogen) and the Glutathione Peroxidase Cellular Activity Assay Kit (Sigma-Aldrich), respectively. 3. Results 3.1. Electrochemical instrumentation for simultaneous measurement of O2 and H2O2 The study was performed using an assay chamber (Fig. 1A) bearing a Clark type electrode for O2 detection, separated from the mitochondria preparation by a polytetrafluoroethylene (PTFE) membrane, while a 10 μm platinized carbon fiber electrode, directly immersed in the mitochondria preparation, was used for the highly sensitive detection of H2O2 (see Materials and methods for details). Two independent potentiostats were allowed to control separately the potential of the two working electrodes: the O2 signal was typically measured at −0.7 V vs. Ag/AgCl, while H2O2 was detected at an oxidation potential of 0.6 V vs. Ag/AgCl. In fact, while the diffusion controlled oxidation of H2O2 is already observed at ≥0.4 V in PBS solutions [17], in the presence of mitochondrial suspension we observed a large shift of the signal towards higher potentials. The measuring potential (0.6 V) was therefore chosen in order to increase the sensitivity of the PCE still maintaining a good selectivity towards H2O2 vis-à-vis other oxidizable species potentially present in the preparation, such as nitric oxide NO• and peroxynitrite ONOO − radicals (vide infra). To test the sensitivity and specificity of the system we measured H2O2 and O2 signals obtained following the addition of known amounts of H2O2 into the assay chamber, in the absence or presence of catalase. Results, shown upon normalization in Fig. 1B, revealed

Therefore, we analyzed with the two electrode-chamber respiration and H2O2 production of freshly isolated mouse liver mitochondria, in the presence of different respiratory substrates and inhibitors of mitochondrial respiration. Whenever the mitochondrial electron transfer chain operates within the assay chamber, the electrodes should reveal the reduction in O2 concentration due to active mitochondrial respiration, and, in principle, the formation of H2O2, that diffuses outside the mitochondria. It is known that both inside and outside the mitochondria H2O2 is efficiently buffered by scavenging enzymes, mainly catalases and glutathione peroxidases (Gpx) [11]. Since the PCE showed net H2O2 concentration after scavenging, we verified the presence of these enzymes in our mitochondrial preparations and, as expected, we confirmed their presence by Western blotting (Fig. 2A). Moreover, analysis of total H2O2 scavenging together with catalase and Gpx enzymatic activities validated an effective H2O2 scavenging activity in isolated mouse liver mitochondrial preparations. Specifically, we found catalase and Gpx activities of 80 ± 0.3 U/mg protein and 0.58 ± 0.05 mU/mg protein, respectively (values are mean of 5 independent experiments ± SE), suggesting a significant contribution of catalases in the scavenging process. Indeed, testing for electrode response to pulses of H2O2 in mitochondrial suspensions revealed transient H2O2 signals and accumulating O2 signals (Fig. 2B and C), indicating that catalases were actively involved in the removal of H2O2. The electrochemical measurements allowed us to appreciate the contribution of the scavenging system in the dynamics of H2O2 production and persistence. From this point of view, electrochemical measurements provide a comprehensive description of the overall process since they allow for quantifications of both generation and scavenging processes. Among the constituents of the antioxidant defense system, the removal of H2O2 can be attributed mainly to the activity of catalases as the decomposition of exogenously added H2O2 is accompanied by O2 release with a stoichiometry of 2:1 [20].

24

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28

Fig. 2. Contribution of scavenging enzymes present in the mitochondrial preparation to the levels of H2O2 measured at platinized carbon fiber electrodes. A. Catalase, glutathione peroxidase, and superoxide dismutase I and II were probed by Western blotting in total liver homogenates, supernatants obtained after centrifugation of the homogenate and mitochondrial fractions. Actin and mitochondrial cytochrome c oxidase subunit IV were used as markers of different fractions. B. Amperometric response of successive addition of H2O2 (1–3 μM) at PCE in the presence of non-energized mitochondria. A 0.6 V vs. Ag/AgCl potential was applied at PCE to oxidize H2O2 C. O2 concentration measured by the Clark type electrode, coincidently to H2O2 additions A − 0.7 vs. Ag/AgCl potential was applied to Clark electrode to reduce O2.

3.3. Electrochemical measurements detect H2O2 production above scavenging levels by state 3 mitochondria Addition of mitochondria into the assay chamber resulted in a signal of the PCE that was set to zero with respect to the electrode equilibration baseline (see Materials and methods for details). The subsequent addition of energetic substrates, such as glutamate/malate (5/2.5 mM) or succinate (5 mM) that provide electrons to complex I and complex II of ETC respectively, stimulated stable respiration (Fig. 3A upper panel) without inducing any sensible change in the signal measured by the PCE (Fig. 3A lower panel). Nevertheless, induction of state 3 respiration by addition of 200 μM ADP to mitochondria energized with either glutamate/malate or succinate (as revealed by the massive and transient increase of O2 consumption: 1–1.5 min long, see Fig. 3B upper panels), resulted in a synchronized increase of the PCE signal (Fig. 3B lower panels). Despite the presence of scavenging enzymes, including catalase as previously indicated, the PCE signal increased progressively as long as state 3

lasted. Yet, once all the ADP was phosphorylated by ATP synthase and O2 consumption resumed to state 4 levels, the PCE signal simultaneously diminished and eventually reached the levels attained before state 3 induction. Notably, inhibition of ATP synthase activity with oligomycin suppressed the rise in the PCE signal induced by ADP (Fig. 3B, red lines), indicating that ATP synthase activity was required for this phenomenon. The rise in signal was directly associated with state 3 mitochondrial respiration since increasing state 3 duration by using higher amounts of ADP also prolonged the duration of the PCE peak signal (Fig. 3C). Next, we verified whether the state 3-associated PCE signal could be abolished by increasing scavenging levels. We found that addition of high concentrations of exogenous catalase (1 μM) could not remove completely the signal but reduced the peak intensity by about 50% (Fig. 4). Thus we questioned the identity of the residual signal detected in the presence of catalase. It has been reported that 0.6 V amperometric measurements with PCEs are also sensitive to the detection of some reactive nitrogen species (RNS), such as nitric oxide, NO•, and peroxynitrite, ONOO − [18]. Experiments performed at 0.4 V can exclude the contribution of NO• radicals to the amperometric signal since only H2O2 and ONOO − are oxidized at this potential. Control experiments at 0.4 V on isolated mitochondria, confirmed the boost of the PCE signal measured upon ADP addition, although with a reduced current intensity (Supplementary Fig. 3). However, when exogenous H2O2 was added to a mitochondrial suspension, the signal measured at 0.4 V was significantly lower than the signal measured at 0.6 V (Supplementary Fig. 1 and Section 2.2), indicating that the plateau for H2O2 oxidation was not reached at 0.4 V as previously reported for PBS solutions [17] and that a tissue extract can change the behavior of the electrode with respect to pure solutions. Therefore the reduced PCE signal measured at 0.4 V might be due to a decreased sensitivity to H2O2 detection at this potential, rather than the lack of detection of NO• radicals. In conclusion, although we cannot exclude the contribution of some RNS to the amperometric signal, our findings indicate that ATP synthase activity influences O2• − formation, leading to the progressive increase in the concentration of H2O2 measured at the PCE above scavenging levels. Interestingly, during state 3, the PCE signal was higher when mitochondria were supplied with glutamate/malate than with succinate, despite the fact that succinate addition resulted in increased O2 consumption compared to glutamate/malate. We estimated the H2O2 amount produced during state 3 respiration by subtracting from the total PCE signal the residual signal measured in the presence of exogenous catalase. Thus, the rates of H2O2 production by mitochondria respiring on glutamate/malate or succinate were 0.293 ± 0.09 μmol/ L/min and 0.088 ± 0.05 μmol/L/min, respectively. As our measurements of the corresponding O2 consumption rates were 25.07 ± 2.93 μmol/L/min with glutamate/malate and 35 ± 0.22 μmol/L/min with succinate, we obtained a significantly higher H2O2 production/ O2 consumption ratio (i.e. the percentage of O2 that is reduced partially to H2O2) for glutamate/malate than for succinate (1.18% ± 0.62 compared to 0.25% ± 0.14, p b 0.001). Therefore, the formation rate of H2O2 in state 3 does not strictly correlate with the overall amount of O2 consumed. 3.4. Moderate depolarization stimulates mitochondrial H2O2 production Subsequently we examined the mechanisms underlying state 3 mitochondrial H2O2 formation. Regardless of the effect on respiration, ATP synthase activation affects transmembrane proton gradient. Thus, to investigate the correlation between mitochondrial transmembrane potential and H2O2 production, we measured PCE signals after stimulation of mitochondria with p-trifluoromethoxy carbonyl cyanide phenyl hydrazone (FCCP), that dissipates the proton gradient across the mitochondrial inner membrane and maximizes

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28

25

Fig. 3. Electrochemical analysis of H2O2 in mouse liver mitochondria. A. O2 concentration and H2O2 amperometric response obtained with Clark type and PCE electrodes (upper and lower panels, respectively). Mitochondria were energized with glutamate/malate (black line) or succinate (red line). B. O2 concentration and H2O2 amperometric current values (upper and lower panels, respectively) measured in state 3 and state 4 mitochondria upon induction with 200 μM ADP (black line), and addition of oligomycin (red line). Mitochondria in MTC assay buffer were energized with glutamate/malate (left panels) or succinate (right panels) C. O2 concentration and H2O2 amperometric current values (upper and lower panels, respectively) measured in glutamate/malate energized mitochondria upon addition of 200 μM ADP (black line) or 400 μM ADP (red line). Traces shown in panels A–C are representative of 5 independent experiments each performed on a different mitochondrial preparation. The potentials for H2O2 and, respectively, O2 detection were 0.6 V vs. Ag/AgCl and − 0.7 V vs. Ag/AgCl.

mitochondrial respiration (uncoupled state). Addition of 10–30 nM FCCP to energized mitochondria moderately increased the respiration rate with respect to basal (state 4) respiration, and resulted in a transient (2–3 min long) increase in the PCE signal (Fig. 5A and B, black lines). High concentrations of FCCP (1 μM) maximized the respiration rate (15.72 ± 2.3 nmol/ml/min for glutamate/malate or 52.75 ± 7.3 nmol/ml/min for succinate) and also generated a PCE signal that decreased slowly (Fig. 5A and B, red lines). However, inducing the collapse of the mitochondrial membrane potential by uniportmediated monovalent cation influx [21] (by additions of A23 in a KCl-based medium plus EDTA, or NaCl in a sucrose-based medium plus EDTA, to mitochondria) doubled the respiration rate without generating any PCE signal (data not shown). Therefore, it appears that a moderate depolarization, rather than the collapse of the mitochondrial membrane potential, is effective in sustaining H2O2 formation from isolated mitochondria as seen upon ADP stimulation. Measurements of H2O2 production by isolated mouse liver mitochondria with the Amplex Red/HRP or DCFDA revealed an increase in

the fluorescence rate of the dyes by hyperpolarized mitochondria. However, we observed a significant contribution to the fluorescent signal from mitochondria also in the absence of respiratory substrates most probably due to unspecific side reactions (see Supplementary Fig. 4). Furthermore, we have investigated the effects of antimycin A's addition to energized mitochondria on electrode signals; inhibition of

Fig. 4. Specificity of H2O2 signals measured for state 3 and state 4 isolated mitochondria H2O2 amperometric signals measured for state 3 and state 4 mitochondria energized with glutamate/malate and upon addition of 200 μM ADP, in the absence (black line) or presence (red line) of 1 μM bovine liver catalase. Traces are representative of 5 independent experiments each performed on a different mitochondrial preparation. A 0.6 V vs. Ag/AgCl potential is applied to PCEs for H2O2 oxidation.

Fig. 5. Mitochondrial H2O2 production and membrane depolarization. O2 concentration (A) and H2O2 amperometric current values (B) measured in glutamate/malate energized mitochondria upon depolarization with FCCP. Where indicated by the arrows, 0 nM (blue line), 30 nM (black line) or 1 μM (red line) FCCP were added. Traces shown in these panels are representative of 3 independent experiments performed on different mitochondrial preparations.

26

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28

respiration with antimycin A should increase ROS from complex III. The current from the platinized carbon fiber microelectrode rapidly increase although, differently to the signal obtained by oxidation sensitive dyes, stabilized (see Fig. 6).

4. Discussion Up to now ROS have been indirectly investigated in living samples, by virtue of their oxidative effects on endogenous biological or exogenously added synthetic markers. Experiments using such approaches have described reduced ROS formation from state 3 and uncoupled mitochondria, while supporting the idea that hyperpolarization favors electron leakage from ETC [22,23]. In our experimental conditions, measurements of H2O2 production by isolated mouse liver mitochondria with the Amplex Red/HRP method revealed no change in the fluorescence rate of the dye (or eventually a faint increase, difficult to be quantified, see Supplementary Fig. 4A) upon ADP addition compared to state 4 respiring mitochondria. Although this phase appeared to be synchronous with state 3 respiration its precise duration could not be appreciated in the absence of a system for the simultaneous measurement of O2 consumption. Moreover, we observed a significant contribution to the fluorescent signal from mitochondria in the absence of respiratory substrates (Supplementary Fig. 4B), most probably due to unspecific side reactions of HRP [12]. Since fluorescent probes such as Amplex Red are irreversibly oxidized by ROS (Supplementary Fig. 2), those events related for instance to H2O2 scavenging cannot be appreciated and, therefore, the dynamic behavior of H2O2 cannot be investigated. The electrochemical approach reported here may represent an alternative tool for the measurement of mitochondrial O2 metabolism. Since ROS generated by mitochondria are directly linked to mitochondrial respiration, it is important to measure both ROS production and O2 consumption at the same time. A synchronized measurement of ROS production and O2 consumption is therefore crucial to evaluate the efficiency of O2 reduction during aerobic metabolism. The amperometric assay we implemented allows for the simultaneous detection of O2 and H2O2 through the usage of two different electrodes that are part of the assay chamber: a Clark type electrode for O2 and a PCE for H2O2 detection. In conditions where scavenging enzymes were present, our electrochemical measurements on isolated mitochondria showed that

Fig. 6. Mitochondrial H2O2 production upon antimycin A administration. O2 concentration (inset) and H2O2 amperometric current values (large plot) measured in energized mitochondria upon inhibition of complex III with antimycin A. in inset where indicated by the arrows antimycin A were added.

the induction of state 3 respiration upon ATP synthase activation would result in H2O2 production above scavenging levels, at a rate depending on the respiratory substrate. However, state 4 respiration alone did not result in detectable H2O2 production, thus production rate could not be measured. Moreover, the transition from state 3 to state 4 respiration was followed by the removal of the H2O2 produced while ATP synthase was functioning. We are aware of the discrepancy between these data and several others associating reduced mitochondrial ROS production with transient mitochondrial depolarization [22,23] or with uncoupling [24]. However, results from the calibration tests and the experiments performed at different potential or with oligomycin would exclude that the PCE generates artifacts. At mechanistic level, blocking proton influx by oligomycin would lead to hyperpolarization and increase the “reduction state” of electron transfer chain sites favoring superoxide formation. Nevertheless in our system, upon oligomycin addition, while the Clark type electrode revealed reduced O2 consumption, the platinized carbon fiber microelectrode, sensitive to H2O2, revealed only a bare increase of the amperometric signal that upon stabilization became comparable with respect to that one registered before oligomycin addition, indicating that H2O2 was not accumulating enough to be detected by the electrode. We believe this discrepancy may arise from the peculiarity of the electrochemical approach. Molecular probes irreversibly react with H2O2 (see Supplementary Fig. 2) and compete with scavenging kinetics whereas the electrode detects both increasing and decreasing H2O2 concentrations after scavenging. Indeed, we believe the electrochemical approach to ROS metabolism deserves attention by scientist interested in biological application of electrochemical tools because of these features. Although NO production by isolated mitochondria was excluded by our and other's measurements [25], RNS could also, in principle, contribute to the amperometric signal. However, at least a fraction of the amperometric peak induced by ADP appears to be catalase sensitive and due to effective H2O2 accumulation. Simultaneous recordings of O2 and H2O2 allowed us to estimate that only a small percent of O2 consumed by mitochondria gives rise to H2O2 production (1.18% for glutamate/malate and 0.25% for succinate), in agreement with previous reports [7]. As rotenone, which stops ETC at complex I, prevented the increase of H2O2 signal in glutamate/malate energized mitochondria, an active ETC is required for the ADP induced H2O2 generation. However, rotenone treatment of mitochondria, in the presence of succinate, partially allowed the spike of H2O2 (approximately one fifth of the amount revealed in the absence of rotenone) without affecting O2 consumption (Supplementary Fig. 5). These results suggest that the electron transfer reactions occurring between complex I and ubiquinone are responsible at least in part for the electron leakage leading to O2• − and then H2O2 production by mitochondria stimulated with ADP. Indeed, several evidence indicates complex I as a major site of O2• − production from the reaction between O2 and the fully reduced flavin mononucleotide center or by reverse electron transport from the reduced ubiquinone [26,27]; complex III also generates O2• − as a result of auto-oxidation of ubisemiquinone, an intermediate compound formed during the ubiquinone cycle [28,29]. How ATP synthase activation affects the rate of O2• − production by the redox centers of complexes I and III, favoring persistent H2O2 release from mitochondria, is unclear. The effect on mitochondrial trans-membrane potential could play a role in H2O2 generation; in fact, similar amperometric signals could be recapitulated by minimal dosage of uncouplers, but other processes such as membrane fluidity [30] or conformational changes in ETC proteins [31] might also modulate the rate of electron leakage from ETC. Recent data indicate that the supramolecular organization of respiratory complexes into

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28

supercomplexes (respirasomes) favors the optimal functioning of mitochondria, with possible effects also on mitochondrial ROS production [32,33]. An interesting hypothesis could be that membrane depolarization associated with ATP synthesis or conformational changes occurring to FoF1-ATPase during ADP phosphorylation modifies the architecture of the I/III supercomplex, altering its reactivity toward O2 and enhancing O2• − production. Regardless of its molecular basis, ATP synthesis-synchronized ROS production above the scavenging barrier may be relevant if occurring in vivo. This event, in fact, can be linked to both physiological [34] and pathological [35] ROS generating processes. Thus ATP synthase activation can in principle signal, through H2O2, the energetic state of mitochondria to the surrounding cellular components. This phenomenon can contribute to oxidative stress as well explaining why intense oxidative phosphorylation, as it occurs during exercise, increases levels of intracellular ROS [36]. In conclusion, our study tested platinized electrodes on isolated mouse liver mitochondria revealing a signal peak upon ATP synthase activation, which is likely due to H2O2 formation above scavenging levels, whereas, under basal respiration conditions the accumulation of H2O2 was prevented by scavenging enzymes that, at equilibrium, buffer to the minimum the concentration of H2O2 in the mitochondrial suspension. These findings support the use of an electrochemical approach to investigate mitochondrial ROS metabolism, and the importance of the interaction between H2O2 production and scavenging in the propagation of oxidative signals that reflect the metabolic state of mitochondria. Acknowledgments We thank Christian Amatore for all the support in preparing the electrodes, Costanza Savino and Massimo Stendardo for technical assistance, and Paola Dalton for help in writing the manuscript. This research was supported by the European School of Molecular Medicine, MIUR, University of Bologna, Fondazione CARISBO and FUV, Fondazione Umberto Veronesi. Appendix A. Supplementary data Supplementary data to this article can be found online at doi:10. 1016/j.bioelechem.2011.11.005. References [1] D.C. Wallace, A mitochondrial paradigm of metabolic and degenerative diseases, aging, and cancer: a dawn for evolutionary medicine, Annu. Rev. Genet. 39 (2005) 359–407. [2] M. Valko, D. Leibfritz, J. Moncol, M.T. Cronin, M. Mazur, J. Telser, Free radicals and antioxidants in normal physiological functions and human disease, Int. J. Biochem. Cell Biol. 39 (2007) 44–84. [3] S. Orrenius, V. Gogvadze, B. Zhivotovsky, Mitochondrial oxidative stress: implications for cell death, Annu. Rev. Pharmacol. Toxicol. 47 (2007) 143–183. [4] J.R. Stone, S. Yang, Hydrogen peroxide: a signaling messenger, Antioxid. Redox Signal. 8 (2006) 243–270. [5] M. Giorgio, M. Trinei, E. Migliaccio, P.G. Pelicci, Hydrogen peroxide: a metabolic by-product or a common mediator of ageing signal, Nat. Rev. Mol. Cell Biol. 8 (2007) 722–728. [6] D.L. Hoffman, D. L., P.S.J. Brookes, Oxygen sensitivity of mitochondrial reactive oxygen species generation depends on metabolic conditions, Biol. Chem. 284 (2009) 16236–16245. [7] M.P. Murphy, How mitochondria produce reactive oxygen species, Biochem. J. 417 (2009) 1–13. [8] D.L. Hoffman, J.D. Salter, P.S. Brookes, Response of mitochondrial reactive oxygen species generation to steady-state oxygen tension: implications for hypoxic cell signaling, Am. J. Physiol. Heart Circ. Physiol. 292 (2007) H101–H108. [9] E.L. Bell, T.A. Klimova, J. Eisenbart, C.T. Moraes, M.P. Murphy, G.R. Budinger, N.S. Chandel, The Qo site of the mitochondrial complex III is required for the transduction of hypoxic signaling via reactive oxygen species production, Cell Biol. 177 (2007) 1029–1036. [10] L. Valgimigli, G.F. Pedulli, M. Paolini, Measurement of oxidative stress by EPR radical-probe technique, Free Radic. Biol. Med. 31 (2001) 708–716.

27

[11] B. Halliwell, J.M.C. Gutteridge, Free Radicals in Biology and Medicine, third ed. Oxford University Press Inc., New York, 2007. [12] T.V. Votyakova, I.J. Reynolds, Detection of hydrogen peroxide with Amplex Red: interference by NADH and reduced glutathione auto-oxidation, Arch. Biochem. Biophys. 431 (2004) 138–144. [13] A. Gomes, E. Fernandes, J.L.J. Lima, Fluorescence probes used for detection of reactive oxygen species, Biochem. Biophys. Methods 65 (2005) 45–80. [14] B. Halliwell, M. Whiteman, Measuring reactive species and oxidative damage in vivo and in cell culture: how should you do it and what do the results mean? Br. J. Pharmacol. 142 (2004) 231–255. [15] S. Arbault, N. Sojic, D. Bruce, C. Amatore, A. Sarasin, M. Vuillaume, Oxidative stress in cancer prone xeroderma pigmentosum fibroblasts, real-time and single cell monitoring of superoxide and nitric oxide production with microelectrodes, Carcinogenesis 25 (2004) 509–515. [16] C. Amatore, S. Arbault, C. Bouton, J.C. Drapier, H. Ghandour, A.C. Koh, Real-time amperometric analysis of reactive oxygen and nitrogen species released by single immunostimulated macrophages, ChemBioChem 9 (2008) 1472–1480. [17] C. Amatore, S. Arbault, C. Bouton, K. Coffi, J.C. Drapier, H. Ghandour, Y. Tong, Monitoring in real time with a microelectrode the release of reactive oxygen and nitrogen species by a single macrophage stimulated by its membrane mechanical depolarization, ChemBioChem 7 (2006) 653–661. [18] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications, John Wiley and Sons Inc., New York, 2001. [19] N. Oshino, R. Oshino, B. Chance, The characteristics of the “peroxidatic” reaction of catalase in ethanol oxidation, Biochem. J. 131 (1973) 555–563. [20] B. Chance, N. Oshino, Kinetics and mechanisms of catalase in peroxisomes of the mitochondrial fraction, Biochem. J. 122 (1971) 225–233. [21] A. Nicolli, A. Redetti, P. Bernardi, P. The, K + conductance of the inner mitochondrial membrane. A study of the inducible uniport for monovalent cations, J. Biol. Chem. 26 (1991) 9465–9470. [22] A.A. Starkov, G. Fiskum, Regulation of brain mitochondrial H2O2 production by membrane potential and NAD(P)H redox state, J. Neurochem. 86 (2003) 1101–1107. [23] 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. [24] S.A. Mookerjee, A.S. Divakaruni, M. Jastroch, M.D. Brand, Mitochondrial uncoupling and lifespan, Mech. Ageing Dev. 131 (2010) 463–472. [25] Y.M. Tay, K.S. Lim, F.S. Sheu, A. Jenner, M. Whiteman, K.P. Wong, B. Halliwell, Do mitochondria make nitric oxide? No? Free. Radic. Res. 38 (2004) 591–599. [26] M.S. King, M.S. Sharpley, J. Hirst, Reduction of hydrophilic ubiquinones by the flavin in mitochondrial NADH:ubiquinone oxidoreductase (Complex I) and production of reactive oxygen species, Biochemistry 48 (2009) 2053–2062. [27] A.J. Lambert, J.A. Buckingham, H.M. Boysen, M.D. Brand, Diphenyleneiodonium acutely inhibits reactive oxygen species production by mitochondrial complex I during reverse, but not forward electron transport, Biochim. Biophys. Acta 1777 (2008) 397–403. [28] J.F. Turrens, Mitochondrial formation of reactive oxygen species, J. Physiol. 552 (2003) 335–344. [29] D. Han, E. Williams, E. Cadenas, Mitochondrial respiratory chain-dependent generation of superoxide anion and its release into the intermembrane space, Biochem. J. 353 (2001) 411–416. [30] P. Schonfeld, L. Wojtczak, Fatty acids as modulators of the cellular production of reactive oxygen species, Free Radic. Biol. Med. 45 (2008) 231–241. [31] C. Batandier, X. Leverve, E. Fontaine, Opening of the mitochondrial permeability transition pore induces reactive oxygen species production at the level of the respiratory chain complex I, J. Biol. Chem. 279 (2004) 17197–17204. [32] I. Wittig, H. Schagger, Supramolecular organization of ATP synthase and respiratory chain in mitochondrial membranes, Biochim. Biophys. Acta 1787 (2009) 672–680. [33] M. Frenzel, H. Rommelspacher, M.D. Sugawa, N.A. Dencher, Ageing alters the supramolecular architecture of OxPhos complexes in rat brain cortex, Exp. Gerontol. 45 (2010) 563–572. [34] D.B. Zorov, M. Juhaszova, S.J. Sollott, Mitochondrial ROS-induced ROS release: an update and review, Biochim. Biophys. Acta 1757 (2006) 509–517. [35] M.R. Duchen, Mitochondria and Ca(2 +)in cell physiology and pathophysiology, Cell Calcium 28 (2000) 339–348. [36] S.K. Powers, M.J. Jackson, Exercise-induced oxidative stress: cellular mechanisms and impact on muscle force production, Physiol. Rev. 88 (2008) 1243–1276.

Raluca Marcu PhD, is a physicist, currently a post-doctoral fellow in the Mitochondria and Metabolism Center, Seattle, USA. She got her PhD in Nanotechnology at the European Institute of Oncology in Milan.

28

R. Marcu et al. / Bioelectrochemistry 85 (2012) 21–28 Stefania Rapino PhD, after the PhD graduation in chemistry at the University of Bologna, is now a senior post-doc at the European Institute of Oncology.

Mirella Trinei PhD, is joining, as associated researcher, the Cardioncology unit of the European Institute of Oncology in Milan.

Giovanni Valenti PhD, is a post-doctoral fellow in the Electrochemistry Group of Francesco Paolucci at the University of Bologna.

Massimo Marcaccio PhD, is a lecturer. He graduated in chemistry at the University of Bologna. After obtaining his PhD in Bologna, he spent 3 years as post-doc at Bristol University, he then went back to Bologna where he got the present position.

Pier Giuseppe Pelicci MD PhD, is a full professor of Pathology at the University of Milan and the director of the Experimental Oncology Department at the European Institute of Oncology in Milan. He graduated in Perugia and specialized at the Columbia University in New York. He is the author of more than 300 publications and contributed to the characterization of the: molecular mechanisms of fusion-protein induced leukemias; physiological function of leukemia-associated translocation genes; mechanisms of HDAC-inhibitor sensitivity of leukemias; Shc protein functions in signal transduction ad life span determination in mammals. Francesco Paolucci PhD, is a full Professor. After graduating in chemistry at the University of Bologna, he spent 5 years as a researcher at the CNR in Padua. Then, after spending one year as a visiting scientist at the University of Southampton, he obtained a research position at the University of Bologna, where he became an associate professor in 2001.

Marco Giorgio, PhD, received a B.Sc. degree in Biology. He spent two years as post-graduate training in the Department of Genetics at the Research Institute for Molecular Biology-IRBM in Pomezia, Italy, working on molecular biology and biochemistry then he studied cancer models at the transgenic laboratory of the “Regina Elena” Institute, Rome. He spent two years at the Dept of Human Genetics at the Memorial Sloan-Kettering Cancer Center, New York, USA, to work on transgenic model of cancer and aging as PhD. fellow. After the PhD degree in Biotechnology he joined the European Institute of Oncology of Milan where, in the Department of Experimental Oncology, he is currently a staff scientist studying bioenergetics and molecular mechanisms of degenerative diseases.