Anoxia halts oxidative phosphorylation (OXPHOS) causing an accumulation of reduced compounds in the mitochondrial matrix which impedes dehydrogenases. By simultaneously measuring oxygen concentration, NADH autofluorescence, mitochondrial membrane potential and ubiquinone reduction extent in isolated mitochondria in real-time, we demonstrate that Complex I utilized endogenous quinones to oxidize NADH under acute anoxia. 13C metabolic tracing or untargeted analysis of metabolites extracted during anoxia in the presence or absence of site-specific inhibitors of the electron transfer system showed that NAD+ regenerated by Complex I is reduced by the 2-oxoglutarate dehydrogenase Complex yielding succinyl-CoA supporting mitochondrial substrate-level phosphorylation (mtSLP), releasing succinate. Complex II operated amphidirectionally during the anoxic event, providing quinones to Complex I and reducing fumarate to succinate. Our results highlight the importance of quinone provision to Complex I oxidizing NADH maintaining glutamate catabolism and mtSLP in the absence of OXPHOS.

GABA is catabolized in the mitochondrial matrix through the GABA shunt, encompassing transamination to succinic semialdehyde followed by oxidation to succinate by the concerted actions of GABA transaminase (GABA-T) and succinic semialdehyde dehydrogenase (SSADH), respectively. Gamma-hydroxybutyrate (GHB) is a neurotransmitter and a psychoactive drug that could enter the citric acid cycle through transhydrogenation with α-ketoglutarate to succinic semialdehyde and d-hydroxyglutarate, a reaction catalyzed by hydroxyacid-oxoacid transhydrogenase (HOT). Here, we tested the hypothesis that the elevation in matrix succinate concentration caused by exogenous addition of GABA, succinic semialdehyde or GHB shifts the equilibrium of the reversible reaction catalyzed by succinate-CoA ligase towards ATP (or GTP) hydrolysis, effectively negating substrate-level phosphorylation (SLP). Mitochondrial SLP was addressed by interrogating the directionality of the adenine nucleotide translocase during anoxia in isolated mouse brain and liver mitochondria. GABA eliminated SLP, and this was rescued by the GABA-T inhibitors vigabatrin and aminooxyacetic acid. Succinic semialdehyde was an extremely efficient substrate energizing mitochondria during normoxia but mimicked GABA in abolishing SLP in anoxia, in a manner refractory to vigabatrin and aminooxyacetic acid. GHB could moderately energize liver but not brain mitochondria consistent with the scarcity of HOT expression in the latter. In line with these results, GHB abolished SLP in liver but not brain mitochondria during anoxia and this was unaffected by either vigabatrin or aminooxyacetic acid. It is concluded that when mitochondria catabolize GABA or succinic semialdehyde or GHB through the GABA shunt, their ability to perform SLP is impaired.

Encysted embryos (cysts) of the crustacean Artemia franciscana exhibit enormous tolerance to adverse conditions encompassing high doses of radiation, years of anoxia, desiccation and extreme salinity. So far, several mechanisms have been proposed to contribute to this extremophilia, however, none were sought in the lipid profile of the cysts. Here in, we used high resolution shotgun lipidomics suited for detailed quantitation and analysis of lipids in uncharacterized biological membranes and samples and assembled the total, mitochondrial and mitoplastic lipidome of Artemia franciscana cysts. Overall, we identified and quantitated 1098 lipid species dispersed among 22 different classes and subclasses. Regarding the mitochondrial lipidome, most lipid classes exhibited little differences from those reported in other animals, however, Artemia mitochondria harboured much less phosphatidylethanolamine, plasmenylethanolamines and ceramides than mitochondria of other species, some of which by two orders of magnitude. Alternatively, Artemia mitochondria exhibited much higher levels of phosphatidylglycerols and phosphatidylserines. The identification and quantitation of the total and mitochondrial lipidome of the cysts may help in the elucidation of actionable extremophilia-affording proteins, such as the 'late embryogenesis abundant' proteins, which are known to interact with lipid membranes.

Provision of NAD+ for oxidative decarboxylation of alpha-ketoglutarate to succinyl-CoA by the ketoglutarate dehydrogenase complex (KGDHC) is critical for maintained operation of succinyl-CoA ligase yielding high-energy phosphates, a process known as mitochondrial substrate-level phosphorylation (mSLP). We have shown previously that when NADH oxidation by complex I is inhibited by rotenone or anoxia, mitochondrial diaphorases yield NAD+, provided that suitable quinones are present (Kiss G et al., FASEB J 2014, 28:1682). This allows for KGDHC reaction to proceed and as an extension of this, mSLP. NAD(P)H quinone oxidoreductase 1 (NQO1) is an enzyme exhibiting diaphorase activity. Here, by using Nqo1-/- and WT littermate mice we show that in rotenone-treated, isolated liver mitochondria 2-methoxy-1,4-naphtoquinone (MNQ) is preferentially reduced by matrix Nqo1 yielding NAD+ to KGDHC, supporting mSLP. This process was sensitive to inhibition by specific diaphorase inhibitors. Reduction of idebenone and its analogues MRQ-20 and MRQ-56, menadione, mitoquinone and duroquinone were unaffected by genetic disruption of the Nqo1 gene. The results allow for the conclusions that i) MNQ is a Nqo1-preferred substrate, and ii) in the presence of suitable quinones, mitochondrially-localized diaphorases other than Nqo1 support NADH oxidation when complex I is inhibited. Our work confirms that complex I bypass can occur by quinones reduced by intramitochondrial diaphorases oxidizing NADH, ultimately supporting mSLP. Finally, it may help to elucidate structure-activity relationships of redox-active quinones with diaphorase enzymes.

The oxidation of proline to pyrroline-5-carboxylate (P5C) leads to the transfer of electrons to ubiquinone in mitochondria that express proline dehydrogenase (ProDH). This electron transfer supports Complexes CIII and CIV, thus generating the protonmotive force. Further catabolism of P5C forms glutamate, which fuels the citric acid cycle that yields the reducing equivalents that sustain oxidative phosphorylation. However, P5C and glutamate catabolism depend on CI activity due to NAD+ requirements. NextGen-O2k (Oroboros Instruments) was used to measure proline oxidation in isolated mitochondria of various mouse tissues. Simultaneous measurements of oxygen consumption, membrane potential, NADH, and the ubiquinone redox state were correlated to ProDH activity and F1FO-ATPase directionality. Proline catabolism generated a sufficiently high membrane potential that was able to maintain the F1FO-ATPase operation in the forward mode. This was observed in CI-inhibited mouse liver and kidney mitochondria that exhibited high levels of proline oxidation and ProDH activity. This action was not observed under anoxia or when either CIII or CIV were inhibited. The duroquinone fueling of CIII and CIV partially reproduced the effects of proline. Excess glutamate, however, could not reproduce the proline effect, suggesting that processes upstream of the glutamate conversion from proline were involved. The ProDH inhibitors tetrahydro-2-furoic acid and, to a lesser extent, S-5-oxo-2-tetrahydrofurancarboxylic acid abolished all proline effects. The data show that ProDH-directed proline catabolism could generate sufficient CIII and CIV proton pumping, thus supporting ATP production by the F1FO-ATPase even under CI inhibition.