Reactive oxygen species generation by reverse electron transfer at mitochondrial complex I under simulated early reperfusion conditions

Abstract

Ischemic tissues accumulate succinate, which is rapidly oxidized upon reperfusion, driving a burst of mitochondrial reactive oxygen species (ROS) generation that triggers cell death. In isolated mitochondria with succinate as the sole metabolic substrate under non-phosphorylating conditions, 90 % of ROS generation is from reverse electron transfer (RET) at the Q site of respiratory complex I (Cx-I). Together, these observations suggest Cx-I RET is the source of pathologic ROS in reperfusion injury. However, numerous factors present in early reperfusion may impact Cx-I RET, including: (i) High [NADH]; (ii) High [lactate]; (iii) Mildly acidic pH; (iv) Defined ATP/ADP ratios; (v) Presence of the nucleosides adenosine and inosine; and (vi) Defined free [Ca²⁺]. Herein, experiments with mouse cardiac mitochondria revealed that under simulated early reperfusion conditions including these factors, total mitochondrial ROS generation was only 56 ± 17 % of that seen with succinate alone (mean ± 95 % confidence intervals). Of this ROS, only 52 ± 20 % was assignable to Cx-I RET. A further 14 ± 7 % could be assigned to complex III, with the remainder (34 ± 11 %) likely originating from other ROS sources upstream of the Cx-I Q site. Together, these data suggest the relative contribution of Cx-I RET ROS to reperfusion injury may be overestimated, and other ROS sources may contribute a significant fraction of ROS in early reperfusion.

Keywords: Complex-I; Ischemia; Mitochondria; Reactive oxygen species; Reperfusion; Reverse electron transfer.

Fig. 1

Baseline ROS Generation by Cx-I RET. Mouse heart mitochondria were incubated as described in the methods, with ROS generation measured by pHPA fluorescence. (A):pHPA fluorescent traces (individual traces in gray, average in black). Where indicated, succinate, rotenone, and H₂O₂ were added. (B): Quantitation of H₂O₂ generation rates from traces as shown in A. Individual data points are superimposed on bars showing means ± SEM. (C): As in A, but with use of S1QEL and subsequently S3QEL instead of rotenone. (D): As in B. ***p < 0.0005 vs. baseline condition. NS: not significant.

Fig. 2

Impact of NADH on ROS. (A): Schematic showing the incubation system to generate NADH from PDH, with export of acetyl-CoA from mitochondria by carnitine, thus avoiding engagement of the TCA cycle. (B): NADH fluorescence in isolated mitochondria (λ_EX 340 nm/λ_EM 460 nm), with sequential additions of pyruvate, carnitine, ADP, and rotenone. Representative trace from 3 similar experiments. (C):pHPA fluorescent traces (individual traces in gray, average in black). Where indicated, pyruvate, carnitine, succinate, S1QEL, S3QEL and H₂O₂ were added sequentially. (D): Quantitation of H₂O₂ generation rates from traces as shown in C. Individual data points are superimposed on bars showing means ± SEM. **p < 0.005 vs. baselne condition. NS: not significant.

Fig. 3

Impact of Lactate and Acidic pH on ROS. (A):pHPA fluorescent traces (individual traces in gray, average in black). Where indicated, pyruvate, carnitine, succinate, lactate, S1QEL, S3QEL and H₂O₂ were added sequentially. (B): Quantitation of H₂O₂ generation rates from traces as shown in A. Individual data points are superimposed on bars showing means ± SEM. (C): As in A, but at pH 6.8. (D): As in B. *p < 0.05, **p < 0.005, vs. baseline condition. NS: not significant.

Fig. 4

Impact of Ischemia-Like ATP/ADP on ROS. (A): Schematic showing the creatine kinase (CK) and phosphocreatine/creatine (PCr/Cr) buffer system. (B):pHPA fluorescent traces. Where indicated, substrates, S1QEL, S3QEL and H₂O₂ were added sequentially. Two independent traces are shown for each ATP/ADP condition (color coded), with the data offset vertically by 20 units for clarity. (C): Quantitation of H₂O₂ generation rates from traces as shown in B. Individual data points are superimposed on bars (color-coded to match B) showing means ± SEM. Individual data points are superimposed on bars showing means ± SEM. *p < 0.05, **p < 0.005, vs. baseline condition. NS: not significant.

Fig. 5

Impact of Purine Nucleosides and Ca²⁺on ROS. (A): Effect of adenosine or inosine on ROS generation (original pHPA traces not shown). Graph shows rates of H₂O₂ generation calculated from traces, with the ischemia-like ATP/ADP condition from previous figure (red bar) forming the baseline condition for addition of nucleosides. (B): Quantitation of H₂O₂ generation rates under different prevailing free [Ca²⁺] conditions. The baseline condition herein represents the presence of both adenosine and inosine from panel A. Individual data points are superimposed on bars showing means ± SEM. *p < 0.05, **p < 0.005, ***p < 0.0005, vs. baseline condition. NS: not significant.

Fig. 6

Overall Comparison of Conditions (A): Comparison of all conditions studied herein, with developments to the model system proceeding left-to-right. Blue bars show S1QEL-sensitive component of ROS, green the S3QEL-sensitive, and gray the component not attributable to either source. The bars show absolute values of H₂O₂ (see Y-axis) with standard errors, while numbers within the bars show the percentage contribution of each component to the total. (B): Visualization of the overall and contributing components of ROS generation in the original (succinate only) and final (simulated early reperfusion) conditions. Each colored box represents 1 unit of ROS, with the original condition assigned to 100 units. In the final model, overall ROS is 56 % of that seen in the original model. Contribution of each component is color-coded as per panel A, and % contributions scaled appropriately. Values are ±95 % confidence intervals.