Redox-Optimized ROS balance and the relationship between mitochondrial respiration and ROS

Abstract

The Redox-Optimized ROS Balance [R-ORB] hypothesis postulates that the redox environment [RE] is the main intermediary between mitochondrial respiration and reactive oxygen species [ROS]. According to R-ORB, ROS emission levels will attain a minimum vs. RE when respiratory rate (VO₂) reaches a maximum following ADP stimulation, a tenet that we test herein in isolated heart mitochondria under forward electron transport [FET].

ROS emission increased two-fold as a function of changes in the RE (~400 to ~900 mV·mM) in state 4 respiration elicited by increasing glutamate/malate (G/M). In G/M energized mitochondria, ROS emission decreases two-fold for RE ~500 to ~300 mV·mM in state 3 respiration at increasing ADP. Stressed mitochondria released higher ROS, that was only weakly dependent on RE under state 3. As a function of VO₂, the ROS dependence on RE was strong between ~550 and ~350 mV·mM, when VO₂ is maximal, primarily due to changes in glutathione redox potential. A similar dependence was observed with stressed mitochondria, but over a significantly more oxidized RE and ~3-fold higher ROS emission overall, as compared with non-stressed controls.

We conclude that under non-stressful conditions mitochondrial ROS efflux decreases when the RE becomes less reduced within a range in which VO₂ is maximal. These results agree with the R-ORB postulate that mitochondria minimize ROS emission as they maximize VO₂ and ATP synthesis. This relationship is altered quantitatively, but not qualitatively, by oxidative stress although stressed mitochondria exhibit diminished energetic performance and increased ROS release.

Introduction

How mitochondria attain a reliable energy provision while keeping ROS within physiological limits compatible with signaling is a relevant question linked with the relationship between respiration and ROS. The importance of this question cannot be overstated because of the central role that mitochondrial energetics/redox functions play in the pathogenesis of numerous human disorders [reviewed in [1, 2]]. Also unbeknownst to us is how the dependence of ROS emission on respiration is influenced by oxidative stress. This is also a fundamental question because it is becoming increasingly clear that mitochondrial dysfunction under oxidative stress underlies aging [3, 4] and acute or chronic disease conditions such as metabolic disorders (obesity, diabetes) [5], cancer [6, 7], neurodegeneration [8], and cardiovascular disease [9–11]. The role of mitochondrial functions pertains not only to pathology but also to vital physiological activities, such as exercise and nutrition [12].

“Uncoupling to survive” [13], or its parental “mild uncoupling” [MU] [14, 15], and the recent Redox-Optimized ROS balance [R-ORB] [16], are hypotheses attempting to explain at a mechanistic level the link between mitochondrial respiration and ROS levels. A main difference between MU and R-ORB is the proposed intermediary between respiration-ROS: mitochondrial membrane potential [ΔΨ_m] or redox environment [RE], respectively.

The MU hypothesis postulates that in energized mitochondria, mild uncoupling, i.e. slight drop in ΔΨ_m, decreases mitochondrial ROS production. Originally, it was proposed that MU was mediated by proteins that regulate the proton permeability of inner mitochondrial membrane [15, 17]. Evidence of mild-uncoupling has been obtained in isolated mitochondria under both forward [18] and reverse electron transport modes of respiration [14]. In this hypothetical framework, ΔΨ_m would be the main mediator between the rate of respiration and ROS. That a slight decrease in ΔΨ_m should also decrease ROS is a major prediction of the MU hypothesis but this has been verified only in isolated mitochondria, not in intact cells[16]. Indeed, MU elicited an increase rather than a decrease in ROS in living cardiomyocytes from guinea pig loaded with two ROS probes simultaneously. The R-ORB hypothesis was introduced in part to account for this discrepancy.

The R-ORB hypothesis postulates that [16]: i) the extent of ROS imbalance is determined by the RE; ii) ROS levels attain a minimum at intermediate values of RE; iii) ROS overflow can occur at both extremes of RE, i.e. highly reduced or highly oxidized, but through completely different mechanisms [Fig. 1]. According to R-ORB, the net flux of ROS released from mitochondria depends upon ROS production by the respiratory chain and ROS scavenging, the balance of which is altered at both extremes of RE. When very reduced [Fig. 1, right hand arm], the RE corresponds to slow electron flow through the respiratory chain thus increasing the probability of generating the free radical superoxide, irrespective of high antioxidant availability. Instead, when the RE is rather oxidized [Fig. 1, left hand arm], the compromised scavenging capacity becomes rate-controlling, thus ROS overflow occurs.

Figure 1. The Redox-Optimized ROS Balance [R-ORB] hypothesis.

R-ORB postulates that ROS levels [as the net result of production and scavenging] depend on the intra-cellular and -mitochondrial RE. It also proposes that there is a minimum level of ROS emission when mitochondria maximize their energetic output. According to R-ORB, a decrease in ROS levels does not require compromising the efficiency of mitochondrial energy transduction [e.g. mild ΔΨ_m uncoupling] but instead it proposes that under high energy demand, and despite large respiratory rates, ROS emission levels will be kept to a minimum by ROS scavenging systems. Recently, this contention received experimental support [20].

Physiological ROS signaling [denoted between dashed lines] occurs within a range close to the minimum of the overall [red] curve that corresponds to intermediate values of the RE. Oxidative stress can happen at either extreme of RE, either highly reduced or highly oxidized, but governed by completely different mechanisms.

Reprinted by permission from Aon, Cortassa, O’Rourke (2010) Biochim Biophys Acta 1797, 865–877.

Another major difference between MU and R-ORB concerns the consequences on mitochondrial energetics. To decrease ROS levels, MU proposes uncoupling, i.e. decreasing ΔΨ_m, the main driving force for ATP synthesis and ion and substrate movement in mitochondria. This mainly applies to fully energized mitochondria, i.e. close to state 4, where ΔΨ_m is maximal and according to MU more prone to produce ROS. In contrast, with its focus on the RE, R-ORB predicts that ROS emission levels will reach a minimum when mitochondria maximize their energetic output [i.e. maximal state 3 respiration] which corresponds to more oxidized values of RE, as compared with the low respiratory flux in state 4. Thus, R-ORB encompasses both states 4 and 3 of respiration and, potentially, includes MU [Fig. 1, right branch of red curve] in a more encompassing view.

In the present work, we test whether mitochondrial ROS emission levels tend to a minimum when respiration attains a maximum, a major tenet of the R-ORB hypothesis. This question begs the dual functional role of mitochondria comprising consistency in energy supply and tuning of ROS to non-harming levels, compatible with signaling. Moreover, we also put to test the postulate that ROS imbalance depends on the RE, in this case changed as a function of respiratory substrates and ADP. We perform these studies with isolated guinea pig heart mitochondria in the absence or the presence of oxidative stress.

Materials and Methods

Mitochondrial isolation

Procedures for the isolation and handling of mitochondria from guinea pig heart were performed as described [16, 19]. Respiratory Control Ratios [state 3 / state 4 respirations with 5mM glutamate + malate] of 8 or higher were obtained using this method.

Assay of mitochondrial respiration

Freshly isolated mitochondria were assayed for respiration as described [20]. Briefly, a high throughput automated 96-well extracellular flux analyzer [Seahorse Bioscience XF96, Billerica, MA] was utilized, and a medium [buffer B] containing [in mM]: 137 KCl, 2 KH₂PO₄, 0.5 EGTA, 2.5 MgCl₂, 20 HEPES at pH 7.2 and 37°C, in the presence of 0.2% fatty acid free BSA. Respiration was evaluated with substrates of complex I [glutamate/malate, G/M, 5mM each]. Mitochondrial protein was determined using the bicinchronic acid method, BCA™ protein assay kit [Thermo Scientific, IL]. VO₂ was determined in parallel with the same mitochondrial preparation utilized for the fluorescent determination of other redox and energetic variables.

Other bioenergetic variables and ROS detection

NAD[P]H, hydrogen peroxide [H₂O₂], GSH, 90° light scattering, and ΔΨ_m were determined as described [16, 19], and monitored simultaneously with a wavelength scanning fluorometer [Quantamaster, Photon Technology, Inc.] utilizing the same medium above for measuring respiration [excluding BSA], and a multidye program for simultaneous online monitoring of different fluorescent probes. H₂O₂ was detected using the Amplex Red kit from Invitrogen [Carlsbad, CA].

Estimation of the redox environment [RE]

The RE comprises two factors, the half-cell reduction potential [Nernst potential, see eq. 2 below] of the redox couple i, E_i, e.g. NADH/NAD⁺, and the concentration of the reduced species i of that couple, e.g. NADH. The RE is calculated according to [21, 22]:

$$\mathit{\text{Redox environment}}={\sum }_{i=1}^{n(\mathit{\text{couple}})}{E}_i{[\mathit{\text{reduced species}}]}_i$$ [eq. 1]

Equation 1 represents the summation of the product of the redox potential times the concentration of the reduced species of the linked redox couples present in the system. Redox couples include those involved in electron transport [NADH/NAD⁺] and in antioxidant pathways [NADPH/NADP⁺, GSH/GSSG, Trx[SH]₂/TrxSS]. In this work, we will estimate the RE as a function of NADH/NAD⁺ and GSH/GSSG redox couples.

In the mitochondrial matrix, NADH at ~1mM and GSH at 1–2mM represent the two major redox couples contributing to the RE, as compared with ~0.1mM NADPH and ~0.01–0.05 mM Trx [20, 22–25].

Redox potential determination

The Nernst equation was used to estimate the redox potential of each redox couple. This equation relates the change in free energy of the transfer of electrons to the voltage of an electrochemical cell, and can be expressed as:

$$\mathrm{\Delta }E=\mathrm{\Delta }E°-\frac{RT}{nF}\mathit{\text{ln}}Q$$ [eq. 2]

ΔE° is the electromotive force under standard conditions that corresponds to the difference in standard reduction potentials of the two half-cells involved in the process; n is the number of electrons exchanged in the chemical process; F is the Faraday constant; R, the universal gas constant; T, temperature (Kelvin scale). Q stands for the quotient of the concentration of reduced over oxidized species from the redox couple. Standard redox potentials, ΔE° [see eq. 2], used were −320 mV for NADH/NAD⁺ and −240 mV for 2 GSH/GSSG [21].

To calculate the redox potential using eq. 2 we need to determine the redox status of the couple, i.e. the relative concentration of reduced and oxidized species.

NAD(P)H and GSH determination

NAD(P)H was monitored by exciting mitochondrial suspensions at 340 nm and collecting the emission at 450 nm. The NAD(P)H signal was calibrated with addition of 1mM KCN [100%] and 6µM FCCP [0%±5%–10%] at the end of the experiment [e.g. Fig. 2C]. Under certain experimental conditions FCCP addition may only account for 90% to 95% of NAD(P)H due to the existence of NADPH and bound-NADH; the latter may vary between mitochondrial preparations within the specified range and may imply longer time to achieve full oxidation.

Figure 2. Substrate-induced redox transitions under pseudo steady state conditions.

Freshly isolated mitochondria from guinea pig heart were monitored simultaneously for NAD(P)H and H₂O₂ emission with Amplex Red [ARed] as a function of increasing concentrations of G/M [A] or ADP after energization with 5mM G/M [B]. The concentration-response change in specific rate of H₂O₂ emission as a function of G/M [C] or ADP [D] was determined. Measurements of H₂O₂ were performed during the periods marked, as examples, by thickening of the NAD(P)H traces topped by short horizontal lines in panels A and B. These corresponded to the late phase of the ADP phosphorylation response. ROS emission levels below 80µM ADP could not be accurately measured thus not represented in panel D. F/F0 represents the fluorescent signal normalized with respect to the initial fluorescence value, i.e. before substrate addition. At the end of the experiment, 50nM H₂O₂ (A) or 1mM KCN and 6µM FCCP (B) were added for calibrating the ARed and NAD(P)H signals, respectively. See M&M for further explanation.

GSH was quantitated in supernatants from mitochondria undergoing states 4 and 3 respiration after trichloroacetic acid precipitation as described elsewhere[25]. Briefly, after neutralizing to pH 7.4 the trichloroacetic acid supernatants, we utilized non-enzymatic fluorescent labeling of thiols with the probe N-[4-[7-diethylamino-4-methylcoumarin-3-yl]phenyl]maleimide; picomole levels of GSH can be detected with this method [25]. The rate of reaction of the thiol fluorescent probe with GSH is linear within the range of 20–100 pmol [see [25], for further details]. We determined values ranging between 2.0 to 1.0 mM GSH for control, and 1.0 to 0.5 mM GSH for stressed mitochondria from guinea pig heart. The matrix concentration of GSH was estimated considering a mitochondrial volume of 2µl/mg of mitochondrial protein according to our previous determination [19]. A 1 to 2 mM mitochondrial matrix GSH pool appears reasonable considering the 2.7mM intracellular GSH determined before in guinea pig ventricular cardiomyocytes [26].

GSH was also monitored as glutathione S-bimane, GSB [λ_exc=390nm; λ_em =480nm]. GSB is the fluorescent adduct obtained after the reversible reaction of monochlorobimane, MCB, with GSH catalyzed by glutathione S-transferase [27, 28] in mitochondria loaded with 20µM MCB, as described [16].

Knowing the relative (%) value of NAD(P)H or the GSH concentration, and the total pool of each one of them present in the mitochondrial matrix, the concentration value (in mM) of each species can be estimated under specified conditions. In this work, we assumed 1mM for the total pool of NADH, and determined 2mM or 1mM for GSH under control, non-stressed, or stressed conditions, respectively (see below). The values of RE reported herein (e.g. Fig.6) are in agreement with values obtained by computational modeling [21].

Figure 6. Mitochondrial respiration and H₂O₂ emission as a function of RE.

Mitochondria loaded with 20µM MCB concentration [non-stressed: A, C] or 50µM MCB plus 1µM H₂O₂ addition [stressed: B, D] were assayed in parallel for redox variables [NAD(P)H, GSH, H₂O₂] by fluorometry [A, B] and respiration, VO₂ [C, D] as a function of ADP concentration (from 10 to 800 µM) with high throughput Seahorse equipment as described [see M&M]. In panels C and D displayed are the state 3 [black squares] and uncoupled with 50µM dinitrophenol [empty circles] VO₂ values. The RE determined from GSH and NAD(P)H was calculated as described under M&M.

Protocol for stressing mitochondria

Freshly isolated mitochondria were stressed by loading with 50µM MCB, higher than the 20µM MCB used for monitoring GSH. At 50µM, MCB will partially deplete the available GSH pool without affecting respiration. In Figure S3 we show that within the range 20 to 50µM MCB does not affect respiration, and in Figure 5B that it partially depletes ~50% and ~25% the GSH and NAD(P)H pools, respectively [compare with Fig. 4B]. MCB [50µM]-loaded mitochondria were further challenged by oxidative stress with 1µM H₂O₂ when assayed in the fluorometer for energy/redox variables or for VO₂ determination, under otherwise identical conditions utilized for control, non-stressed, mitochondria.

Figure 5. RE as a function of substrate addition in stressed mitochondria.

Mitochondria loaded with higher MCB concentration [50µM] were assayed as described in Figure 3 but further stressed with 1µM H₂O₂ before starting the titration with substrates (see M&M). NAD(P)H and GSH abundance, H₂O₂ emission and the RE were determined as described in the legend of Figure 4. The RE was estimated as described under M&M. The arrows in panels C and D indicate increasing substrate addition.

Figure 4. RE as a function of substrate addition in non-stressed mitochondria.

The experiment was carried out with MCB-loaded [20µM] mitochondria as described in Figure 3. The relative abundance of NAD(P)H (%) and GSH (mM) and H₂O₂ emission were monitored simultaneously as a function of G/M [A, state 4 respiration] or 5mM G/M plus different concentrations of ADP [B, state 3 respiration], and the RE as a function of G/M [C] and ADP [D] calculated as described under M&M. From the data in panel B, the RE in the abscissa of panel D was calculated; the additional symbols in panel D correspond to interpolated values, e.g. GSH at 400µM ADP. The NAD(P)H signal was calibrated as described under M&M. The arrows in panels C and D indicate increasing substrate addition.

Results

Two main postulates of the R-ORB hypothesis are: i) the extent of ROS imbalance depends on the RE, and ii) ROS levels attain a minimum when respiration is maximal and the RE exhibits intermediate values. ROS increase under more reducing (i.e. right arm in Fig. 1) or severe oxidizing conditions (i.e. left arm in Fig.1) although by different mechanisms [16]. Since normal substrate consumption changes mitochondrial RE, in the first section we analyze the response of mitochondrial ROS emission as a function of substrate concentration.

Cyclic changes in the rate of H₂O₂ release from mitochondria during energetic transitions: pseudo steady state and time-dependent behavior

Pseudo steady state behavior

Mitochondria dynamically modulate matrix redox potential, membrane potential (ΔΨ_m), and the pH gradient across the inner membrane when supplied with substrates [21, 29]. In de-energized (substrate-depleted) mitochondria, redox and electrochemical gradients can be increased in a graded manner in response to increasing concentrations of respiratory substrate (Figs. 2 and S1). Dissipation of ΔΨ_m and a decrease in NADH redox potential occurs during ATP synthesis initiated by ADP, ion or metabolite transport, or activation of the ROS scavenging systems. In addition, oxidation of the main components of the antioxidant system (GSH/GSSG, TrxSH2/TrxSS) can occur if NADPH levels become limiting [21, 25].

The relationship between mitochondrial NAD(P)H and H₂O₂ emission (Amplex Red [ARed] oxidation) with increasing concentrations of G/M in the absence of ADP is shown in Figure 2. H₂O₂ emission increases more than two fold over the range of G/M from 0.17 – 12mM [Figs. 2A and 2C].

Similarly, we challenged fully energized mitochondria in the presence of 5mM G/M, with pulses of increasing ADP concentration. Figure 2B shows that the NAD(P)H signal attains lower values concomitant to a decrease in the rate of change of ARed fluorescence as a function of increasing ADP. An almost linear decrease in H₂O₂ emission occurs as a function of ADP that levels off at 0.8mM [Fig. 2D].

The results shown indicate that mitochondrial ROS emission augments or diminishes in parallel with a progressively more reduced or oxidized redox status as a function of G/M or G/M+ADP, respectively.

Time-dependent behavior

Figure 3 depicts the phase relationship between different redox variables [NAD(P)H, GSH, ROS] during the state 4→3 transition elicited by successive ADP pulses. The mitochondrial energetic transition is clearly indicated by the orthodox→condensed shift in mitochondrial volume [30], as probed by 90° light scattering (LS) [19]. Following ADP addition, mitochondria contract (LS increases), and then swell again as ADP is consumed (LS decreases). Interestingly, the instantaneous rate of H₂O₂ emission (i.e. the first derivative [dARed/dt] of the ARed signal) decreases to a minimum at the peak of mitochondrial contraction and OxPhos rate, as indicated by NADH oxidation. The instantaneous rate of ROS release increases again when ADP is consumed (Figs. 3 and S2), as can be judged from the mitochondrial increase in volume and NADH reduction.

Figure 3. Phase relationship between light scattering and redox variables during time-dependent redox/energetic transitions triggered by ADP pulses.

Freshly isolated heart mitochondria from guinea pig were loaded with 20µM MCB for GSH detection as described under M&M. MCB-loaded mitochondria [100–200µg mitochondrial protein] were resuspended in a cuvette in the assay mixture described under M&M, energized with 5mM G/M, and then subjected to pulses of increasing ADP concentration [from 10µM to 800µM]. Light scattering [LS], NAD(P)H, GSH, and the instantaneous rate of ROS emission [measured as the first derivative of the ARed signal, dARed/dt], were determined simultaneously. The axis for each variable is placed immediately below its corresponding label.

We also monitored simultaneously the status of the GSH pool by loading mitochondria with the fluorescent probe MCB. Remarkably, the peak of GSH coincides with the minimum in H₂O₂ emission rate, which starts to decrease concomitantly with the rise of the ROS scavenger. The troughs in GSH reflect its maximal utilization during H₂O₂ scavenging. Electrons from NADH and Krebs cycle intermediates are not only directed to the respiratory chain but also to the scavenging systems, for example through pyridine nucleotide transhydrogenase activity and NADP⁺-dependent isocitrate dehydrogenase to NADPH, which drives glutathione and thioredoxin reductases coupled to their respective peroxidases, glutathione peroxidase and peroxiredoxin [26, 31, 32].

Taken together, the data reveal that when mitochondria maximize their energetic output, ROS emission declines to a minimum, and increases thereafter as ADP is consumed. This behavior was verified both at pseudo steady state and during dynamic transitions in respiratory state.

ROS emission as a function of the RE

In the previous section, we showed that ROS release minima occur with more oxidized NAD(P)H redox state and high OxPhos rate, and that ROS flow increases under more reduced NAD(P)H redox states corresponding to low respiration both under steady (Fig. 2) and time-dependent (Fig. 3) states. Next, we monitored H₂O₂ emission as a function of GSH and NAD(P)H that were monitored simultaneously to calculate the mitochondrial RE from these two major contributors.

Figure 4 depicts the steady state levels of GSH and NAD(P)H as a function of G/M at state 4 respiration (no ADP added) [Fig. 4A] and at state 3 respiration with 5mM G/M energized mitochondria challenged with consecutive pulses of increasing ADP concentration [Fig. 4B]. We can clearly see that ROS emission levels are augmented or diminished following a more reduced or oxidized RE elicited by G/M or ADP, respectively. H₂O₂ emission varies by approximately two-fold over the entire range of RE explored [Fig. 4C, 4D].

In terms of the R-ORB hypothesis, the data described above supports the relationship between ROS and RE on the right arm of the schematic plot displayed in Fig. 1, i.e., from intermediate to reduced values of the RE. To explore this relationship under conditions in which the GSH pool is more oxidized, we stressed mitochondria by loading them with 50µM MCB, which will partially deplete the available GSH pool without affecting respiration [see M&M]. Under conditions in which MCB-loaded mitochondria were further challenged by oxidative stress with 1µM H₂O₂, ROS emission levels were about 2-fold higher than in non-stressed mitochondria in response to both G/M and ADP additions [compare panels C and D between Figs. 4 and 5]. ROS emission behaved independently from G/M but not from ADP levels, and remained at high levels over the range of G/M concentrations studied [Fig. 5C], but declined as ADP concentration increased [Fig. 5D]. Interestingly, the range of variation of the RE was larger in non-stressed as compared with stressed mitochondria (i.e. ~200 to ~900 vs. ~200 to ~550 mV·mM) (Figs. 4C, 4D and 5C, 5D). The difference in GSH level was the main underlying factor responsible for the relative oxidation of the mitochondrial RE under oxidative stress (compare panel B in Figs. 4 and 5).

The results indicate that under non-stressful conditions mitochondrial ROS emission levels exhibit a clear dependence on the RE in a certain range of values: from ~400 to ~900 mV·mM in state 4 respiration (Fig.4C) and ~500 to ~300 mV·mM in state 3 respiration (Fig. 4D). However, under oxidative conditions, although mitochondrial ROS emission was high, it became independent of RE under state 4 and only weakly dependent on RE under state 3 respiration (Fig. 5C, 5D).

Respiration and ROS emission as a function of the RE

Concerning the central issue of the relationship between energetic and redox aspects of mitochondrial function, a main prediction of the R-ORB is that ROS emission levels will attain a minimum when mitochondrial energetic output is maximal [16]. We directly tested this proposal by assessing, in parallel, the rates of mitochondrial respiration [VO₂], ROS emission, and the RE.

Figure 6 shows the behavior of VO₂ and H₂O₂ emission as a function of the RE varied as a function of ADP concentration. It can be clearly seen that H₂O₂ emission tends to lower values at maximal VO₂ [Fig. 6A]. The strong dependence of ROS emission on RE occurs in the restricted range comprised between ~500 and ~400 mV·mM, when respiratory flux is highest, remaining independent thereafter. This behavior reflects the prototypical flux (e.g. respiration, ATP synthesis) vs. force (e.g. RE, phosphorylation potential) relationship exhibited by energy transducers (see Discussion). As compared with non-stressed controls, stressed mitochondria displayed a similar dependence of H₂O₂ emission on the RE but in a significantly more oxidized range: ~300 and ~200 mV·mM and higher levels of ROS (~3-fold at maximal emission) [Fig. 6B].

We also evaluated the relationship between VO₂ and H₂O₂ emission vs. ΔΨ_m, under both non-stressed [Fig. 7A] and stressed [Fig. 7B] conditions. Under state 3 respiration, ROS emission varied 2- and 1.5-fold in a rather narrow ~7 mV span of ΔΨ_m in control and stressed mitochondria, respectively. Likewise the RE, the dependence between ROS and ΔΨ_m was strongest at maximal VO₂. Importantly, ROS emission was not sensitive to ΔΨ_m changes within the range corresponding to state 4 respiration [Figs. 7A, 7B and S1, S5]. This observation is relevant because MU predicts that the sensitivity of H₂O₂ emission to ΔΨ_m will be highest under state 4, when mitochondria are maximally energized.

Figure 7. Mitochondrial respiration and H₂O₂ emission as a function of ΔΨ_m.

Experiments were performed as described in the legend of Figure 6 and the data for non-stressed (A) and stressed (B) mitochondria represented as a function of ΔΨ_m that was monitored simultaneously with NAD(P)H and GSH, and in parallel with H₂O₂ or respiration. See Figure S5 for a representative experiment where ΔΨ_m can be monitored at states 4 and 3 respiration.

Energetic performance, assessed by measuring VO₂ vs. ADP, was markedly altered in oxidatively-stressed mitochondria as compared to control, non-stressed, mitochondria. V_max was 15% lower (p<0.05 stressed vs. non-stressed), and the Km for ADP was non-significantly decreased [Fig. 6C, 6D; black data points], while uncoupled VO₂ [Fig. 6C, 6D; empty circles] decreased by 33% in stressed mitochondria (p<0.01 stressed vs. non-stressed).

Taken together, the data presented are in agreement with the idea that when mitochondrial respiration is maximal, ROS efflux is strongly dependent on RE and trends to a minimum. At respiratory rates below maximum, the dependence between ROS and the RE appears to be weaker. Although, in general, stressed mitochondria follow a similar behavior, the RE was significantly more oxidized, ROS levels were higher and respiration lower.

Discussion

The main contribution of this work is to show that mitochondria are able to maximize their energetic output while keeping ROS emission to a minimum. This major principle of R-ORB converges on the idea that mitochondria have evolved in such a way that optimization of redox and energetic functions concur [Fig. 6]. Under pseudo steady state and time-dependent conditions triggered by ADP pulses, the rate of ROS emission also decreases to a minimum at maximal rate of OxPhos [Fig. 3]. However, this capacity is partially lost under oxidative stress where mitochondria decrease their energetic output concomitantly with a more oxidized RE and increased ROS emission [Fig. 6]. R-ORB was initially postulated based on data obtained with guinea pig heart mitochondria, however we do not expect that this choice will influence the validity of the findings that we report herein as applied to other species [20].

The concerted actions of GSH/thioredoxin [Trx] scavenging systems in mitochondria are essential for keeping minimal levels of H₂O₂ emission, especially during state 3 respiration, when the energetic output is maximal [20]. The well-orchestrated function of the two antioxidant systems working continuously to offset ROS production are chiefly responsible for the minimum in ROS emission attained at intermediate values of RE and also for the inability of mitochondria to limit ROS overflow under stressful oxidant conditions. State 3 respiration transients, induced by ADP additions, show that minima in the rate of H₂O₂ emission occur concomitantly with increase in the GSH fluorescence signal. Apparently, this response of the GSH pool [Fig. 3] differs from the global decrease in GSH observed during state 3 respiration by direct assay in mitochondrial supernatants [Fig. 4B]. A possible explanation would be that MCB, the fluorescent probe of GSH, is reporting a local pool of this molecule, different from the bulk phase. A drop in the dynamic balance between the rate of ROS production by the respiratory chain and the rate of ROS scavenging under maximal OxPhos, may explain a transient recovery of the local GSH pool reported by the probe.

Mitochondrial ROS emission and Redox-Energy transduction under normal and stressful conditions

The dependence of mitochondrial ROS efflux on the RE follows the general sigmoidal-curve shape exhibited by energy transducers, i.e. quasi linear flux-force relation within a certain span, and independence at extreme low or high values of the driving force [33]. The range where ROS and RE exhibit dependence is wider under state 4 than state 3 respiration for non-stressed mitochondria (Fig. 4C and 4D). However, under stress, mitochondria display H₂O₂ emission levels about 2-fold higher than in its absence (compare Figs. 4C,D and 5C,D), and exhibit no relation (Fig. 5C) or a weak one (Fig. 5D) between ROS vs. RE (Fig. 5C and 5D). Apparently, oxidative stress shifts the RE toward the more oxidized range where the ROS efflux is insensitive to the RE in state 4 (compare Figs. 4C and 5C) and less sensitive in state 3 respiration (compare Figs. 4D and 5D). A decrease in mitochondrial matrix GSH of ~50% is mainly responsible for the shift (compare Figs. 4B and 5B) and the higher ROS emission, i.e. reduced antioxidant capacity. Understanding the energetic and redox behavior of mitochondria under oxidative stress is a significant biological problem because of the increasingly recognized fact that aging and chronic diseases such as diabetes and obesity involve the progressive deterioration of the redox status of cells and the organism [9, 22, 34].

Concerning energetics, the respiratory flux was lower under oxidative stress than in its absence (Fig. 6A and 6B) although ΔΨ_m was similar (Fig.S1A, C). The latter result suggests that the NADH-electron-donor capacity to respiration might have been diminished under stress likely due to redirection of electrons to the antioxidant systems. Indeed, the relationship between respiration and ROS is altered by oxidative stress, resulting in decreased mitochondrial energetic performance and higher levels of ROS emission (Figs. 6–8).

Figure 8. Effect of oxidative stress on mitochondrial respiration, H₂O₂ emission, and RE.

The plot displays the summary of the response of respiration (black traces) and ROS emission in mitochondria stressed with MCB and exogenously added H₂O₂ (gray traces) plus further addition of the uncoupler FCCP (dashed-dotted line). Continuous lines correspond to the absence of stress whereas dashed lines belong to mitochondria under stressed conditions. Black arrows indicate the direction of change in VO₂ and ROS elicited by stress. Notice the shift toward more oxidized RE in the curves corresponding to stressful conditions. The thick gray arrow pointing to the left denotes pathological conditions arising, e.g. from chronic diseases, where severe stress will affect both energetic (e.g. ΔΨ_m, ADP consumption) and redox (e.g. NAD(P)H, GSH, Trx) functions thus increased mitochondrial ROS emission and higher cytoplasmic ROS levels [38].

Present data suggest that under high energy demand, e.g. exercise, in the absence of additional oxidative stress, mitochondria will function at relatively more reduced RE, i.e. from right arm to the minimum ROS rather than on the left arm [Figs. 1, 8]. However, a shift to more oxidized REs will happen under pathological conditions thus displacing mitochondrial function toward the left arm of the plot in Fig. 1 (see Fig. 8). This was indeed the case for heart mitochondria isolated from diabetic guinea pigs as shown by recent data from our laboratory (Tocchetti, Stanley, Sivakumaran, O’Rourke, Paolocci, Cortassa, Aon, submitted). Likewise, an effective shift to higher ROS release was observed previously [16], when heart mitochondria were stressed under similar conditions and were subjected to increasing levels of uncoupling (thick gray arrow pointing toward the dash-dot line in Fig. 8). In fact, the redox stress introduced by FCCP in already stressed mitochondria [Fig. S3 A, B] was responsible for further RE oxidation and higher ROS emission[16].

Role of compartmentation and the release of ROS from mitochondria

As redox/energetic functions of mitochondria are intermingled, so are the mitochondrial and cytoplasmic REs. A dynamic interaction between the mitochondrial RE and energetic function may underlie mitochondrial dysfunction in aging and disease. For instance, in burn injury, mitochondrially-targeted antioxidant treatment with the peptide SS-31 improves mitochondrial redox status, restoring OxPhos coupling and ATP synthesis rate in skeletal muscle[35].

As a dynamic metric, the RE is a function of the different redox couples accounting for both their redox potential and the concentration of the reduced species [21, 22]. The RE exhibits its own dynamics in different but interdependent cellular compartments, e.g. mitochondria and cytoplasm, mediated by exchange of ROS and redox-related components such as GSH. Compartmentation plays a significant role in the control of ROS levels, the RE, and dynamic behavior [21]. Furthermore, and since GSH/Trx scavenging systems appear to act concertedly and continuously to keep low rates of H₂O₂ emission from mitochondria [20], duplication of antioxidant defense systems in multiple compartments can be viewed as an efficient salvage mechanism to avoid or to reduce oxidative bursts [21].

The phenomenological concept of “mitohormesis” refers to the association observed between increased mitochondrial metabolism and ROS formation with the induction of an adaptive response that culminates in increased stress resistance, antioxidant defense and extended life span [36, 37]. This concept is similar to what R-ORB proposes as levels of ROS compatible with physiological signaling, due to moderate increases around the minimum toward more reduced or oxidized RE [see Fig. 1, gray area around the minimum]. Conceptually, mitohormesis points out that not all increase in ROS is negative, but it can be positive instead. However, mitohormesis does not specify under which conditions the adaptive response will happen, and how this is compatible with mitochondria supplying energy reliably and consistently, while at the same time keeping ROS levels within non detrimental boundaries compatible with physiological signaling.

The intermediary between respiration and ROS

Considering the role of ΔΨ_m and RE as intermediaries between mitochondrial respiration and ROS, we should keep in mind that redox and energetic functions are inextricably linked. An example is NADH and its electron-donating role to both the respiratory chain and the antioxidant systems through transhydrogenase conversion into NADPH. Electric [ΔΨ_m] and redox [NADH] potentials build up after G/M addition to non-energized mitochondria, and are dissipated following ADP addition [Figs. S1, S5]. One could argue that electric and redox potentials could both explain mitochondrial ROS modulation, since they are difficult to untangle. However, the following arguments favor RE over ΔΨ_m as a candidate intermediate between respiration and ROS. First, ROS emission levels varied strongly as a function of the RE over a 200 mV·mM range, in part due to the 20mV drop in the redox potential of GSH:GSSG [Fig. 6], whereas ΔΨ_m only changed 7 mV for the same span [Fig. 7]. A larger rather than a narrow range of variation in a control parameter is more in agreement with the robust function of a system. According to R-ORB, mitochondria optimize both maximal energetic output and minimum ROS emission, thus a co-variation of ROS and ΔΨ_m is expected at maximal VO₂ [Fig. 7]. Second, as a driving force, the RE accounts both for ROS production in the respiratory chain and the ROS scavenging capacity of mitochondria, while ΔΨ_m only accounts for changes in ROS production. Third, in state 4 respiration, the condition under which MU was formulated, no modification in ROS emission occurs for a similar range of variation in ΔΨ_m of ~6 mV when compared to state 3 respiration [Figs.7, S1, S5]. This important observation indicates that the shape of the relationship between ROS emission and RE or ΔΨ_m is dependent upon the type of intervention, i.e. exponential with uncouplers [14, 16, 17], or sigmoidal-shaped with natural substrates (G/M, ADP) as described herein.

Conclusions

Defining the relationship between mitochondrial respiration and ROS emission is crucial to understanding compatibility between reliable energy supply, minimization of oxidative damage, and ROS signaling. In the framework of the R-ORB hypothesis, this work tested the critical idea of whether the RE or ΔΨ_m is the main intermediary between respiration and ROS under normal and stressful conditions. The main finding shows that ROS emission is a strong function of the RE, which works as a robust and encompassing controller of both production and scavenging of ROS. The evidence presented supports the idea that the R-ORB hypothesis, in terms of optimization, reconciles maximal energetic output with minimal ROS likely compatible with signaling.

We conclude that conceptually R-ORB offers a broader framework for describing the relationship between respiration and ROS, since it may account for both physiological and pathological conditions, while including MU as a special case under certain conditions, i.e. the behavior corresponding to the right hand arm of the plot in Figure 1.

Highlights.

• Mitochondrial H₂O₂ emission tends to a minimum at maximal O₂ consumption.

• Mitochondrial H₂O₂ emission is a strong function of the redox environment (RE).

• ROS emission from stressed mitochondria is higher and weakly dependent on RE.

• The RE appears to be the link between mitochondrial respiration and ROS.

Abbreviations

ARed

Amplex Red

ΔΨ_m

mitochondrial membrane potential

FET

Forward Electron Transport

G/M

glutamate/malate

GSH

reduced glutathione

GSSG

oxidized form of GSH

H₂O₂

hydrogen peroxide

LS

light scattering

MCB

monochlorobimane

MU

mild uncoupling

OxPhos

oxidative phosphorylation

RE

redox environment

R-ORB

Redox-Optimized ROS Balance

ROS

reactive oxygen species

Trx

thioredoxin

VO₂

respiratory rate