Mitochondrial respiration and ROS emission during β-oxidation in the heart: An experimental-computational study

Abstract

Lipids are main fuels for cellular energy and mitochondria their major oxidation site. Yet unknown is to what extent the fuel role of lipids is influenced by their uncoupling effects, and how this affects mitochondrial energetics, redox balance and the emission of reactive oxygen species (ROS). Employing a combined experimental-computational approach, we comparatively analyze β-oxidation of palmitoyl CoA (PCoA) in isolated heart mitochondria from Sham and streptozotocin (STZ)-induced type 1 diabetic (T1DM) guinea pigs (GPs). Parallel high throughput measurements of the rates of oxygen consumption (VO2) and hydrogen peroxide (H2O2) emission as a function of PCoA concentration, in the presence of L-carnitine and malate, were performed. We found that PCoA concentration < 200 nmol/mg mito protein resulted in low H2O2 emission flux, increasing thereafter in Sham and T1DM GPs under both states 4 and 3 respiration with diabetic mitochondria releasing higher amounts of ROS. Respiratory uncoupling and ROS excess occurred at PCoA > 600 nmol/mg mito prot, in both control and diabetic animals. Also, for the first time, we show that an integrated two compartment mitochondrial model of β-oxidation of long-chain fatty acids and main energy-redox processes is able to simulate the relationship between VO2 and H2O2 emission as a function of lipid concentration. Model and experimental results indicate that PCoA oxidation and its concentration-dependent uncoupling effect, together with a partial lipid-dependent decrease in the rate of superoxide generation, modulate H2O2 emission as a function of VO2. Results indicate that keeping low levels of intracellular lipid is crucial for mitochondria and cells to maintain ROS within physiological levels compatible with signaling and reliable energy supply.

Fig 1. Scheme of the two-compartment mitochondrial model including β-oxidation and lipid transport.

The present model encompasses mitochondrial energetic and redox processes, their interactions, and transport between compartments, in addition to energy metabolism, ion transport (H⁺, Ca²⁺, Na⁺, Pi) and β-oxidation (top). O₂^.− is dismutated to H₂O₂ by matrix-localized superoxide dismutase (MnSOD) or transported to the extra-mitochondrial compartment, where it will be scavenged by Cu,ZnSOD. H₂O₂ can either diffuse from the matrix or be scavenged by the large capacity glutathione (GSH) and thioredoxin (Trx) systems, or by catalase (CAT) in the extra-mitochondrial compartment. The palmitate oxidation is linked to the TCA cycle via AcCoA and NADH. Also, NADH and FADH₂ (bound to the electron transfer flavoprotein, ETF) are involved in the interaction between β-oxidation and the respiratory chain. The multiple regulatory interactions involved are not represented for simplicity. Key to symbols: Concentric circles with an arrow across (upper left of the scheme) represent the ΔΨ_m. The scheme is modified from [15,63].

Fig 2. Respiratory and ROS emission fluxes from Sham and diabetic mitochondria as a function of PCoA concentration.

Freshly isolated heart mitochondria from Sham and diabetic GPs were assayed for β-oxidation driven respiration and H₂O₂ emission as described in detail under Materials and Methods. Depicted are the specific rates of O₂ consumption, VO₂, (A, B) and H₂O₂ emission fluxes (C, D) determined under states 4 (no ADP; A, C) and 3 (+1mM ADP; B, D) respiration in Sham and STZ-treated (diabetic) mitochondria, respectively. The specific rates of O₂ consumption and H₂O₂ emission, measured in parallel in the same mitochondrial preparations at different PCoA concentrations (panels A-D), were plotted against each other for Sham and diabetic under states 4 (E) and 3 (F) respiration. N = 12 technical repeats from 3 biological replicates (experiments/hearts) in each Sham or diabetic group. Raw traces of O₂ consumption and H₂O₂ emission from representative experiments with Sham and diabetic mitochondria are shown in Fig C in S1 Text.

Fig 3. Model simulations of mitochondrial respiration and ROS emission fluxes from Sham and diabetic mitochondria vs. PCoA concentration.

Steady state values of the rates of respiration (VO₂) (A, B), H₂O₂ emission (V_H2O2) (C, D) and their relationship, i.e. V_H2O2 vs. VO₂ (E, F) were obtained at the indicated levels of PCoA (in μM) under states 4 (A, C, E) and 3 (B, D, F) respiration. The plots display total VO₂ (i.e., derived from NADH, succinate and FADH₂ driven electron transport). The three curves represent simulations obtained at three different Vmax values of glutathione reductase (both intra- and extra-mitochondrial) as indicated in the legend of the plots. The model parameterization and initial conditions employed in the simulations are described in S1 Text together with the Matlab code for the full computational model.

Fig 4. Experimental and computational redox behavior of mitochondrial NAD(P)H in the presence of PCoA-driven respiration, and its impact on mitochondrial function.

Freshly isolated heart mitochondria from Sham or T1DM GPs were assayed in a fluorimeter in the presence of PCoA at the indicated concentrations, and as described in detail under Materials and Methods. (A) The first arrow in panel A indicates the addition time of mitochondria and 0.5mM malate (to enable the regeneration of the mitochondrial coenzyme A pool), after β-oxidation is triggered with the addition of both PCoA and L-carnitine (second arrow). Subsequent sequential additions of 5mM each of G/M and 1mM ADP were performed to test the state 4→3 transition. Depicted are the results from simultaneous monitoring of NAD(P)H in a typical experiment performed with Sham mitochondria but similar results were observed in mitochondria from diabetic GPs. (B) Model simulations of the transient, time-dependent, behavior of NADH upon successive addition of PCoA followed by glutamate and ADP to test the state 4→3 transition as indicated. The simulations condition mimic the experimental protocol in order to reproduce the experimental time courses depicted in panel A. All other model parameters are the same to those indicated in S1 Text.

Fig 5. Modeling studies of mitochondrial respiration, H₂O₂ emission and ROS scavenging intermediates as a function of PCoA concentration and proton conductance.

The steady state behavior of VO₂ in the model was analyzed for the same range of PCoA as in Fig 3 mimicking those used in the experiments under states 4 (A) and 3 (C) respiration (blue traces) shown in Fig 2A and 2B. Also depicted in black traces are the rates of H₂O₂ emission corresponding to the same simulations. Panels B and D display the steady state concentration of main components of matrix antioxidant systems (GSH, Trx(SH)₂, NADPH) under both states 4 (B) and 3 (D) respiration. The parameterization and initial conditions employed in the simulations are described in S1 Text together with the Matlab code for the full computational model.

Fig 6. Mitochondrial antioxidant defense protein levels and ROS emission in the absence or the presence of inhibitors of the GSH and Trx antioxidant systems in Sham and STZ hearts.

(A, B) Heart tissue from Sham or diabetic GPs was processed and the antioxidant proteins indicated were analyzed by Western Blot as described in Materials and Methods. Left panel shows representative Western Blot analysis of protein abundance. The bar plot on the right panel displays the statistical comparison between the results obtained with heart tissue from the two groups [n = 4, four experiments]. Protein was normalized to total protein abundance in a given lane based on Direct Blue 71 (DB71) staining of the membrane as described [16,21]. Key to symbols: GR, glutathione reductase; Gpx4, glutathione peroxidase 4; Trx2, thioredoxin 2; TrxR2, thioredoxin reductase 2; Nnt, nicotinamide nucleotide transhydrogenase; SOD2, superoxide dismutase 2; Prx3, peroxiredoxin 3. (C, D) Freshly isolated heart mitochondria (100μg mitochondrial protein) from Sham or STZ-treated GP were preincubated without or with 50nM auranofin (AF) plus 10μM 1-chloro-2,4 dinitrobenzene (DNCB) [13,16]. Monitoring of H₂O₂ was performed with an Amplex red assay during state 4 (with [A] 10μM PCoA/0.5mM malate/0.5mM L-carnitine or [B] 5mM each of G/M) and state 3 (+1 mM ADP) of mitochondrial respiration, using a wavelength scanning fluorometer (QuantaMaster; Photon Technology International, Inc.) [13,16] in the same assay medium utilized for high throughput measurements. The specific fluxes of H₂O₂ emission are shown for PCoA (C) and G/M (D) in the absence or the presence of AF + DNCB. (E, F) Depicted are the model simulated rates of ROS emission and O₂ consumption upon inhibition of antioxidant defense activities, glutathione reductase (GR) and thioredoxin reductase (TrxR) to mimic the actions of the experimentally utilized inhibitors DNCB and auranofin, respectively (C, D). The plots display the steady state values of H₂O₂ emission (E) and VO₂ (F) obtained at high (control: 100% GR and TrxR activities, empty bars) and low (15% and 22% of GR and TrxR activities, respectively, grey bars) scavenging capacity, under states 4 and 3 respiration as indicated. For comparative purpose, in panel E are shown the experimental values of Sham depicted in panel C, corresponding to the absence (filled squares) or presence (filled triangles) of inhibitors, under states 4 and 3 respiration as indicated. The simulations correspond to 40μM PCoA with 5x10^-4mM and 0.1mM ADP in states 4 and 3, respectively, while the remaining parameters are described in S1 Text.