Extremely anoxia-tolerant animals, such as freshwater turtles, survive anoxia and reoxygenation without sustaining tissue damage to their hearts. In contrast, for mammals, the ischemia–reperfusion (IR) injury that leads to tissue damage during a heart attack is initiated by a burst of superoxide (O2·) production from the mitochondrial respiratory chain upon reperfusion of ischemic tissue. Whether turtles avoid oxidative tissue damage because of an absence of mitochondrial superoxide production upon reoxygenation, or because the turtle heart is particularly protected against this damage, is unclear. Here, we investigated whether there was an increase in mitochondrial O2· production upon the reoxygenation of anoxic red-eared slider turtle hearts in vivo and in vitro. This was done by measuring the production of H2O2, the dismutation product of O2·, using the mitochondria-targeted mass-spectrometric probe in vivo MitoB, while in parallel assessing changes in the metabolites driving mitochondrial O2· production, succinate, ATP and ADP levels during anoxia, and H2O2 consumption and production rates of isolated heart mitochondria. We found that there was no excess production of in vivo H2O2 during 1 h of reoxygenation in turtles after 3 h anoxia at room temperature, suggesting that turtle hearts most likely do not suffer oxidative injury after anoxia because their mitochondria produce no excess O2· upon reoxygenation. Instead, our data support the conclusion that both the low levels of succinate accumulation and the maintenance of ADP levels in the anoxic turtle heart are key factors in preventing the surge of O2· production upon reoxygenation.

Anoxia-tolerant freshwater turtles regularly experience reoxygenation after hypoxia following a dive (Williams and Hicks, 2016) and after anoxia following winter hibernation (Herbert and Jackson, 1985; Ultsch and Jackson, 1982), without sustaining appreciable tissue damage (Bundgaard et al., 2018; Wasser et al., 1992). Isolated turtle hearts are also exceptionally tolerant of anoxia and reoxygenation, maintaining function (Bundgaard et al., 2018; Milton et al., 2007; Wasser et al., 1992). This is remarkable, because reperfusion of hypoxic or ischemic mammalian tissue with oxygenated blood causes a burst of mitochondrial superoxide (O2·) production, which initiates a cascade of oxidative damage that leads to extensive ischemia–reperfusion (IR) injury (Kowaltowski et al., 2009). This phenomenon has been extensively studied in mammals as it can occur clinically, for example following removal of a blood clot with current treatments for heart attack or stroke, or as a consequence of surgical procedures, such as organ transplantation (Martin et al., 2019; Murphy and Steenbergen, 2008; Yellon and Hausenloy, 2007). The ability of turtles and other anoxia-tolerant animals to survive anoxia and reoxygenation without sustaining oxidative tissue damage has been described as one of the three core adaptations that underlie anoxia tolerance (Bickler and Buck, 2007). In turtles, this tolerance has been attributed to (1) relatively high levels of antioxidant enzymes (Willmore and Storey, 1997a) that could effectively scavenge O2· and H2O2 (Bickler and Buck, 2007; Galli and Richards, 2014; Hermes-Lima and Zenteno-Savín, 2002; Hermes-Lima et al., 2015) and/or (2) absence or suppression of the generation of O2· by the mitochondrial respiratory chain in turtles upon reoxygenation after anoxia (Bundgaard et al., 2019a; Milton et al., 2007; Nayak et al., 2009; Pamenter et al., 2007).

Hypothesis 1, that high antioxidant levels are responsible for the absence of oxidative damage in turtles, is based on the observation that antioxidant systems are relatively highly expressed in turtle tissues compared with other ectotherms and similar to mammals, based on comparisons of values reported in the literature (Hermes-Lima and Zenteno-Savín, 2002; Hermes-Lima et al., 2015; Rice et al., 1995; Willmore and Storey, 1997a). However, no formal comparisons have been made in parallel between species, and the activities of the antioxidant enzymes glutathione peroxidase, catalase, superoxide dismutase and glutathione reductase in the heart all decrease by 20–50% when turtles are exposed to anoxia for 20 h at 5°C (Willmore and Storey, 1997a,b). This is paradoxical if antioxidant defences are responsible for preventing oxidative damage upon reoxygenation.

Instead, in support of hypothesis 2, our previous work has shown that the central factors that drive excess production of mitochondrial O2· and H2O2 after ischemia and reperfusion in mammals might not be present to the same extent in turtles, even after prolonged anoxia. The mechanism that is thought to drive O2· and H2O2 production during reperfusion involves accumulation of succinate and depletion of adenine nucleotides during ischemia (Chouchani et al., 2014; Murphy, 2009). This leads to a highly reduced mitochondrial ubiquinone pool and an elevated proton motive force, which together, upon the return of oxygen, cause reverse electron transport (RET) at complex I that drives extensive production of O2· and H2O2 (Fig. 1). Isolated cardiac turtle mitochondria can produce O2· and H2O2 by RET when incubated in vitro with excess succinate (Bundgaard et al., 2019b). It has previously been shown that turtles do accumulate some succinate in the heart with anoxia (Buck, 2000; Bundgaard et al., 2019b; Hochachka et al., 1975). However, the level of succinate accumulated appears to be two to three orders of magnitude lower in the turtle after anoxia than after ischemia in the mouse (Bundgaard et al., 2019b). Furthermore, in contrast to mammals (Martin et al., 2019), turtles are able to maintain adenine nucleotide levels during anoxia (Galli et al., 2013; Stecyk et al., 2009), partly because they are able to enter rapid metabolic depression with anoxia and switch to anaerobic glycolysis (Bickler and Buck, 2007; Jackson, 2002). In particular, maintenance of ADP may enable dissipation of the mitochondrial proton motive force through activation of the F0F1-ATP synthase, thereby preventing RET upon reperfusion. Together, these findings suggest that the conditions for extensive ROS production upon reperfusion may not be present after anoxia in turtles. Consistent with this, the production of mitochondrial H2O2in vitro does not increase with anoxia and reoxygenation in turtle brain slices or neuronal cultures (Milton et al., 2007; Nayak et al., 2009; Pamenter et al., 2007). However, whether mitochondrial H2O2 production occurs in the turtle heart in vivo upon anoxia followed by reoxygenation is not known.

Fig. 1.

Model of the main mechanistic elements driving reverse electron transfer in the mitochondria during reoxygenation. Electrons from oxidation of accumulated succinate moves in the reverse direction from CII to CI due to a high proton motive force (ΔΨm), set by inhibition of CV owing to depletion of the substrate ADP. The direction of the electron flow is marked with the dashed line. We investigated the conditions of this mechanism in turtles exposed to anoxia and reoxygenation by assessing (1) in vivo H2O2 production with MitoB, (2) succinate accumulation and (3) ADP and ATP concentrations and the effect of inhibition of CV on the H2O2 production. CI, complex I (NADH dehydrogenase); CII, complex II (succinate dehydrogenase); CV, complex V (F0F1-ATP synthase); SOD, superoxide dismutase.

Fig. 1.

Model of the main mechanistic elements driving reverse electron transfer in the mitochondria during reoxygenation. Electrons from oxidation of accumulated succinate moves in the reverse direction from CII to CI due to a high proton motive force (ΔΨm), set by inhibition of CV owing to depletion of the substrate ADP. The direction of the electron flow is marked with the dashed line. We investigated the conditions of this mechanism in turtles exposed to anoxia and reoxygenation by assessing (1) in vivo H2O2 production with MitoB, (2) succinate accumulation and (3) ADP and ATP concentrations and the effect of inhibition of CV on the H2O2 production. CI, complex I (NADH dehydrogenase); CII, complex II (succinate dehydrogenase); CV, complex V (F0F1-ATP synthase); SOD, superoxide dismutase.

Here, for the first time, we set out to measure changes in the production of mitochondrial O2· and H2O2in vivo during reoxygenation of an anoxia-tolerant animal, the red-eared slider turtle (Trachemys scripta elegans). To do this, we used the mass-spectrometric probe MitoB, which specifically measures the in vivo levels of H2O2 within mitochondria that are formed from the dismutation of O2·. The use of MitoB has previously been validated in Drosophila melanogaster (Cochemé et al., 2011, 2012), mice (Mus musculus) (Chouchani et al., 2013, 2014; Logan et al., 2014b; Lopez-Fabuel et al., 2016) and brown trout (Salmo trutta) (Salin et al., 2015, 2017, 2018), and can thus be used in a wide range of animal models and experimental settings (Logan et al., 2014a; Murphy, 2016). We hypothesised that turtles do not produce a burst of O2· upon reoxygenation after anoxia, so that the MitoP/B ratio would be unchanged upon reoxygenation. A corollary is that this may arise owing to the limited accumulation of succinate and/or preservation of ADP levels during anoxia that prevent RET upon reoxygenation. We tested these hypotheses by measuring the MitoP/B ratio in vivo and in vitro while in parallel assessing changes in succinate and ATP and ADP levels during anoxia. We found no elevation in MitoP/B upon reoxygenation after anoxia in turtle hearts in vitro and in vivo and this was associated with far lower changes in succinate and ADP during anoxia compared with mammalian tissues. Thus, a major aspect of the ability of anoxia-tolerant turtles to tolerate hypoxia is the lack of ROS production upon reoxygenation, which is due in large part to the lack of dramatic changes in succinate and ADP during anoxia.

Chemicals

All chemicals were purchased from Sigma-Aldrich. MitoB [(3-hydroxybenzyl)triphenylphosphonium bromide] was dissolved in 60°C ethanol to a concentration of 5 mmol l−1 and diluted to 0.5 mmol l−1 with saline (0.9% NaCl/H2O w/v). MitoB is rapidly distributed to the mitochondria when administered to animals (Cochemé et al., 2012; Logan et al., 2014a). MitoB is converted to MitoP by reaction with H2O2. The amount of MitoP generated is thus proportional to the amount of H2O2, the concentration of MitoB and the exposure time. To eliminate the dependence on the concentration of MitoB, MitoP can be expressed as the MitoP/MitoB ratio. The MitoP/MitoB ratio then reflects the levels of H2O2 and can be used to assess relative changes in H2O2 in the mitochondria. Importantly, as the reaction of MitoB with H2O2 is 107 times slower than the other peroxidase reactions of the mitochondria, MitoB does not affect physiological H2O2 levels (Cochemé et al., 2012; Salin et al., 2017). Oligomycin, an inhibitor of F0F1-ATP synthase (complex V), was used as 5 mg ml−1 stock in ethanol, further diluted to 0.25 mg ml−1 in saline.

Animals

Male and female red-eared slider turtles [Trachemys scripta elegans (Wied-Neuwied 1839)] were purchased from NASCO (Fort Atkinson, WI, USA) and kept for several months before being used for experiments. Turtles were kept at 25°C in freshwater aquaria with free access to basking platforms on a 12 h:12 h light:dark cycle and fed three times a week with dry pellets and raw krill at either Manchester University, UK, or Aarhus University, Denmark. MitoB, oligomycin and pentobarbital were all injected into the supravertebral venous sinus with a 23-gauge syringe (Mans, 2008). A total of 32 turtles were used in this study, and the specific details about the experimental groups are given in the descriptions below. All procedures in Manchester were carried out in accordance with The Animals (Scientific Procedures) Act, 1986. Local ethical approval was granted by The University of Manchester Animal Welfare Ethical and Review Board, and all experiments were conducted within the UK Home Office laws for animal care and experimentation. All experiments in Denmark were conducted in accordance with the laws for animal care and experimentation in Denmark under the permit 2015-15-0201-00544.

Experimental procedure

In vivo anoxia and reoxygenation

Turtles (mean±s.e.m. body mass per group in parentheses) were injected with 1 nmol MitoB per gram body mass (nmol g−1) as ∼0.2% of body mass (v/w) of a 0.5 mmol l−1 solution (equal to ∼500–700 µl), left to equilibrate for 5 min in shallow water and then exposed to either 3 h anoxia/1 h reoxygenation (320.5±15.4 g), 3 h anoxia (293.7±21.8 g) or 4 h normoxia (324.7±21.5 g) (n=5 for all groups). Anoxic exposure was conducted in water-filled aquaria, continuously bubbled with nitrogen gas to keep the oxygen tension <2%, monitored by an EcoSense ODO200 oxygen meter. To prevent pulmonary ventilation, turtles were kept submerged by a metal mesh to prevent surface breathing. Normoxic and reoxygenating turtles were kept in tanks with shallow water (∼2 cm), allowing them easy access to air breathing. Experiments were conducted at 21°C. Turtles were euthanized by an overdose of pentobarbital (50 mg kg−1), injected into the supravertebral sinus. When the corneal reflex was gone (after ∼5 min), the turtles were decapitated by guillotine and the brain was destroyed by freezing in liquid nitrogen. Anoxic turtles were prevented from breathing by pressing their heads into their shells before removing them from the water and keeping them pressed in until decapitation after anaesthesia. Tissues were sampled as described below in the section ‘Tissue sampling’.

In vivo inhibition of F0F1-ATP synthase during reoxygenation

We assessed the effect of inhibition of F0F1-ATP synthase (complex V) with oligomycin during reoxygenation on in vivo H2O2 production. We compared the MitoP/B ratio after 3 h anoxia and 1 h reoxygenation in turtles injected with 0.5 mg kg−1 oligomycin (n=4, 315.5±18.3 g) or vehicle (n=4, 345.5±23.6 g). Following the same procedure as described above in the ‘In vivo anoxia and reoxygenation’ section, we injected 0.25 mg ml−1 oligomycin into the supravertebral venous sinus with a 23-gauge syringe to a final concentration of 0.5 mg kg−1 oligomycin. To inject the drug, the heads of the comatose anoxic turtles were pressed into the shell and the nostrils were covered with a wet paper tissue to prevent them from breathing when briefly removed from the aquaria. The anoxic turtles remained comatose during handling and after transfer back to the anoxic aquaria. In parallel, a control group was injected with vehicle. After a total of 3 h anoxia, turtles were reoxygenated for 1 h and tissues were sampled as described in the section ‘Tissue sampling’.

In vitro inhibition of F0F1-ATP synthase during reoxygenation

Next, we tested the effect of inhibiting F0F1-ATP synthase on H2O2 production in turtle heart strips, based on the procedure described previously (Joyce et al., 2014; Overgaard et al., 2005). Six turtles (424.5±59.7 g) were injected with MitoB into the supravertebral sinus. MitoB was allowed to distribute within the turtle for 15 min, before euthanasia with pentobarbital as described above. The ventricle was quickly dissected out and placed in ice-cold turtle Ringer’s solution (100 mmol l−1 NaCl, 25 mmol l−1 NaHCO3, 1 mmol l−1 NaH2PO4, 2.5 KCl, 1 mmol l−1 MgSO4, 2 mmol l−1 CaCl2 and 10 mmol l−1 glucose). The ventricle was then cut into ∼1×1×10 mm slices from base to apex, avoiding areas with connective tissue. Six strips were mounted on glass rods connected to force transducers with surgical silk thread and immersed in turtle Ringer’s solution kept at 21°C bubbled with 98% air, 2% CO2 (normoxia) delivered by a gas-mixing pump (Wösthoff, Bochum, Germany), which resulted in pH 7.6–7.8. The Ringer’s buffer was not exchanged during the experiment to mimic ischemic conditions. The strips were mounted in two different setups: two strips were mounted on a 305C Dual-Mode muscle lever connected to a 701C electrical stimulator (Aurora Scientific, ON, Canada), and four strips were mounted in separate setups connected to force transducers (Statham UC 2, Oxnard, CA, USA) and stimulated with Grass SD 9 stimulators (Quincy, MA, USA) connected to platinum electrodes. The two setups were used in parallel to be able to investigate all six experimental treatments in each turtle. The length of the strips and stimulation intensity from the stimulators were then adjusted to give a maximal isometric force output and the strips were allowed to equilibrate for 30 min in normoxia. The strips were then exposed to four treatments, which were randomised between chambers: (1) 90 min normoxia (normoxic control), (2) 30 min normoxia, followed by 60 min anoxia (anoxic control), (3) 60 min anoxia (98% N2, 2% CO2)/30 min reoxygenation (A/R) or (4) A/R with 10 µmol l−1 oligomycin (A/R+omy). Oligomycin was added 15 min before reoxygenation to allow the drug to equilibrate with the heart strip. After the experiment, the strips were removed from the setup, frozen in liquid N2 and stored at −80°C until analysed for MitoP/B, [ATP], [ADP] and [succinate]. The efficiency of oligomycin inhibition of complex V was assessed in a fifth strip exposed to 10 µmol l−1 oligomycin in turtle Ringer’s solution for 30 min in a 2 ml Eppendorf tube placed in a heating block at 25°C. The strip was frozen in liquid nitrogen and stored at −80°C until assayed for F0F1-ATP synthase activity.

Tissue sampling

For collecting tissues after in vivo experiments, after confirming absence of all reflexes (corneal reflex, limb withdrawal, and head and jaw tone) following pentobarbital administration, the plastron was cut open with a bone-saw. The heart ventricle (∼1 g) was removed and all connective tissue was dissected away. The blood and tissues from anoxic turtles had the characteristic dark colour of deoxygenated haemoglobin and myoglobin, indicating that they were only minimally reoxygenated before sampling of tissues. All samples were snap-frozen in liquid nitrogen and stored at −80°C before further experiments.

Succinate content

Frozen heart ventricle samples were weighed into Precellys tubes prefilled with ceramic beads (CK28-R; Stretton Scientific Ltd, Derbyshire, UK), and 25 µl mg−1 extraction solution (30% acetonitrile, 50% methanol and 20% water) was added together with 100 pmol internal standard (13C4-succinate; Sigma-Aldrich). Samples were lysed using a Precellys® 24 tissue homogeniser (Bertin Corp, Rockville, MD, USA; 6500 rpm, 15 s×2) and centrifuged twice (17,000 g, 10 min, 4°C), retaining the supernatant at each step. The supernatant was transferred to autosampler vials and stored at −80°C until liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis on an LCMS-8060 mass spectrometer (Shimadzu) with a Nexera X2 UHPLC system (Shimadzu) as described previously (Prag et al., 2022) and quantified from a standard curve run under the same conditions. Briefly, samples were separated with a SeQuant ZIC-HILIC column (MerckMillipore) with a ZIC-HILIC guard column. A flow rate of 200 μl min−1 was run with buffer A (10 mmol l−1 ammonium bicarbonate) and buffer B (100% acetonitrile), and a run gradient of 0–0.1 min, 80% MS buffer B; 0.1–4 min, 80–20% B; 4–10 min, 20% B; 10–11 min, 20–80% B; and 11–15 min, 80% B. The instrument was run in negative ion mode with multiple reaction monitoring, and spectra were recorded using LabSolutions software (Shimadzu).

ATP/ADP assay

The ATP and ADP content of heart ventricle samples was determined using a bioluminescence-based luciferase assay (Strehler, 1974), as described in detail elsewhere (Bundgaard et al., 2019b). Briefly, frozen tissue samples were homogenised in perchloric acid (3% v/v HClO4, 2 mmol l−1 Na2EDTA, 0.5% Triton X-100), and samples and standards were then pH neutralised with 2 mol l−1 KOH, 2 mmol l−1 Na2EDTA and 50 mmol l−1 MOPS. For ADP measurements, neutralised sample was mixed with 250 μl ATP sulfurylase assay buffer [20 mmol l−1 Na2MoO4, 5 mmol l−1 GMP, 0.2 U ATP sulfurylase (New England Biolabs), in buffer 100 mmol l−1 Tris-HCl, 10 mmol l−1 MgCl2, pH 8.0], to degrade ATP, incubated for 30 min at 30°C and then heat-inactivated for 5 min at 100°C. Samples and standards were then diluted in buffer (100 mmol l−1 Tris, 2 mmol l−1 Na2EDTA, 50 mmol l−1 MgCl2, pH 7.75) in luminometer tubes. To convert ADP to ATP, pyruvate kinase (6 U ml−1) and phosphoenolpyruvate (100 mmol l−1) were added to one set of samples for ADP measurements and incubated in the dark for 30 min at 25°C. ATP was measured by bioluminescence using an AutoLumat LB-953-Plus luminometer (Berthold, Germany) by addition of luciferase/luciferin solution (7.5 mmol l−1 DTT, 0.4 mg ml−1 BSA, 1.92 μg luciferase ml−1, 120 μmol l−1d-luciferin), delivered via auto injection, protected from light and measured for 1 min.

ROS production with MitoB

In vivo mitochondrial H2O2 production was measured as the ratio between the oxidised product MitoP and the mitochondria-targeted H2O2 probe MitoB (MitoP/B) by LC-MS/MS, as described in detail elsewhere (Cochemé et al., 2012). Briefly, ∼100 mg of frozen tissue was homogenised in 60% (v/v) acetonitrile and 0.1% (v/v) formic acid in HPLC-grade H2O and the homogenate was spiked with internal standards (100 pmol d15-MitoB and 50 pmol d15-MitoP kindly provided by Richard C. Hartley, University of Glasgow, UK) to generate standard curves before centrifugation and re-extraction of the pellet. Standard curves were prepared in parallel with analysis of the samples. The supernatants were filtered, dried and re-dissolved in 20% (v/v) acetonitrile and 0.1% (v/v) formic acid in HPLC-grade H2O. The samples were then centrifuged and transferred to 1.5 ml autosampler vials for LC-MS/MS determination of MitoP and MitoB.

Complex V activity assay

Complex V activity was assayed as the oligomycin-sensitive decrease in NADH in the presence of rotenone and KCN in turtle heart homogenate as described in detail previously (Bundgaard et al., 2019a). Briefly, ∼50 mg tissue was homogenised by bullet blending at 50 Hz for 10 min at 4°C in a Thermo tissue lyser in 25 mmol l−1 potassium phosphate buffer, pH 7.2, and homogenate was added to 100 mmol l−1 Tris, 50 mmol l−1 KCl, 2 mmol l−1 MgCl2, 0.2 mmol l−1 EDTA, pH 8.0, with 1 mmol l−1 phosphoenolpyruvate, 2.5 mmol l−1 ATP, 2 U ml−1 lactate dehydrogenase and 4 U ml−1 pyruvate kinase. Background NADH oxidation was prevented by addition of 10 µmol l−1 rotenone, 300 nmol l−1 antimycin A and 2 mmol l−1 KCN.

Aconitase activity

Aconitase is an enzyme in the citric acid cycle, and the mitochondrial matrix isoform is very susceptible to oxidation by O2· and thus serves as a sensitive tissue marker of oxidative damage (Gardner et al., 1994). Total aconitase activity (i.e. both cytosolic and matrix isoforms) was assayed based on the coupled enzyme assay described by Gardner et al. (1994), where aconitase converts citrate to isocitrate coupled with the decarboxylation to α-ketoglutarate with the reduction of NADP to NADPH by isocitrate dehydrogenase. Turtle hearts were homogenised (by bullet blending at 50 Hz for 10 min at 4°C in a Thermo tissue lyser) in 50 mmol l−1 Tris-HCl buffer pH 7.4 with 0.6 mmol l−1 MnCl2. Then, 20 µmol l−1 dl-fluorocitrate was added to the isolation buffer to protect the oxidation-sensitive iron–sulfur cluster. Aconitase activity was assayed immediately in 50 mmol l−1 Tris-HCl buffer pH 7.4 with 0.6 mmol l−1 MnCl2, 5 mmol l−1 sodium citrate, 0.2 mmol l−1 NADP and 0.4 U ml−1 type IV isocitrate dehydrogenase from porcine heart as the appearance of NADPH at 340 nm at 25°C on a microplate reader (SpectraMax® iD3, Molecular Devices, CA, USA) using ε340 NADPH=6.22×103 (mol l−1)−1 cm−1. Background rate was measured in the presence of 100 µmol l−1 dl-fluorocitrate.

In vitro H2O2 production and consumption

The in vitro capacity for H2O2 consumption and production was assessed on isolated mitochondria from turtle (n=3, 233±44.4 g), green iguana (n=3, 310±42.4 g) and mouse (n=3, 26.3±2.9 g) hearts. For turtles and green iguanas, animals were anaesthetised with 50 mg kg−1 pentobarbital injected into the supravertebral sinus or tail vein, respectively. Mice were killed by cervical dislocation. Heart tissue samples were stored in STE buffer [250 mmol l−1 sucrose, 5 mmol l−1 TRIS and 1 mmol l−1 EGTA, pH 7.4 with 0.5% (w/v) fatty-acid free BSA] on ice for ∼30 min before mitochondrial isolation. Mitochondria were isolated by differential centrifugation as described in detail elsewhere (Bundgaard et al., 2019a,b). Briefly, tissue was homogenised in a Potter–Elvehjem homogeniser on ice in STE buffer, and the homogenate was then centrifuged at 700 g for 5 min at 4°C. The supernatant was then filtered through cheesecloth and centrifuged twice at 10,000 g for 10 min at 4°C, before the final pellet was resuspended in STE without BSA. Protein content was determined with a BCA assay using BSA as a standard.

H2O2 consumption

H2O2 consumption was measured based on the method described by Munro et al. (2016). Isolated mitochondria were added to a 1 ml Eppendorf tube containing respiration buffer [70 mmol l−1 sucrose, 220 mannitol, 2 mmol l−1 EGTA, 5 mmol l−1 MgCl2, 5 mmol l−1 KH2PO4, 10 mmol l−1 Hepes, pH 7.4 with 0.3% (w/v) BSA on the day] equilibrated to 21°C (on an Eppendorf thermomixer) with 2.5 µmol l−1 H2O2 and either: (1) no substrate; (2) 2.5 mmol l−1 malate, 10 mmol l−1 glutamate and 1 mmol l−1 ADP (Mal/Glu/ADP); or (3) 20 mmol l−1 succinate. After 0, 2.5, 5, 7.5 and 10 min, a 180 µl aliquot was removed and mixed with 20 µl quenching solution [250 µmol l−1 Amplex UltraRed, 50 U ml−1 horseradish peroxidase (HRP) and 250 U ml−1 superoxide dismutase] in a black 96-well microplate to convert all H2O2 remaining in the solution to water and producing the fluorescent product resorufin. After 10 min, fluorescence was measured at 544 nm excitation/590 nm emission on a microplate reader (SpectraMax® iD3, Molecular Devices). H2O2 consumption was standardised with a standard curve of 0, 1 and 2.5 µmol l−1 H2O2 in respiration buffer that was prepared in parallel.

H2O2 production

Mitochondrial H2O2 production was measured as previously described (Bundgaard et al., 2019a,b) on an Oroboros O2K high-resolution respirometer fitted with a FluoO2K module with filters specific for Amplex UltraRed (Oroboros Instruments, Innsbruck, Austria). The machines were calibrated with respiration buffer equilibrated with air at 21°C every day.

Mitochondria (62.5 µg ml−1) were added to each of the two chambers, followed by the components for the Amplex UltraRed assay (10 µmol l−1 Amplex UltraRed, 5 U ml−1 HRP and 25 U ml−1 superoxide dismutase) and the fluorescent signal was calibrated with injections of 0.1 µmol l−1 H2O2. Two protocols were run in parallel to match the conditions for the H2O2 consumption assay. In one chamber, 2.5 mmol l−1 malate and 5 mmol l−1 glutamate were added to initiate non-phosphorylating, leak respiration, before addition of 1 mmol l−1 ADP to initiate phosphorylating respiration. Cytochrome c (10 µmol l−1, equine) was then injected to assess the integrity of the outer membrane, which did not elicit an increase in respiration rate above 10% in any experiment. In the other chamber, 20 mmol l−1 succinate was injected to initiate maximal H2O2 production via RET, before addition of 0.5 µmol l−1 rotenone and 1 mmol l−1 ADP to stimulate phosphorylating respiration.

Statistics

Significant differences were determined by one- or two-way ANOVA with Tukey's or Šidák's multiple comparisons test for normally distributed data (Prism 9, Graphpad Software, San Diego, CA, USA), and experimental groups were considered significantly different when P<0.05. Statistical comparisons and P-values are given in the Results below. All data are shown as means±s.e.m.

We assessed mitochondrial ROS production in vivo in the hearts of red-eared slider turtles (T. s. elegans) under three conditions: normoxia, anoxia and reoxygenation after anoxia (Fig. 2A). To see whether reoxygenation of anoxic tissues led to an increase in the in vivo production of mitochondrial O2·, we measured the proportion of its dismutation product H2O2 using the conversion of MitoB to MitoP. The MitoP/MitoB ratio in hearts from turtles exposed to anoxia and reoxygenation was similar to that from turtles exposed to anoxia alone, and both were lower than the ratio in hearts from normoxic controls (one-way ANOVA, P<0.001; Fig. 2B). This result is consistent with the lack of O2 as substrate for O2· generation during anoxia, and thus lowered H2O2 production. When controlling for the different durations of anoxia and reoxygenation (Fig. 2A), the rate of MitoB to MitoP conversion was not elevated in turtles exposed to reoxygenation after anoxia compared with anoxia alone (Tukey's multiple comparisons test, P=0.9949; Fig. 2C). The activity of the O2·-sensitive citric acid cycle enzyme aconitase was also not different between the experimental groups (one-way ANOVA, P=0.8919; Fig. 2D), suggesting that it was not differentially inactivated by O2·. Together, these findings show that there is no excessive production of mitochondrial O2· in the turtle heart upon reoxygenation after anoxia in vivo.

Fig. 2.

In vivo ROS production in turtle hearts after anoxia and reoxygenation. (A) Experimental groups. Turtles were injected with 1 nmol g−1 MitoB and allowed to equilibrate in normoxia for 5 min, before exposing turtles to 4 h normoxia, 3 h anoxia or 3 h anoxia followed by 1 h reoxygenation in normoxia. (B) In vivo H2O2 production measured by the MitoP/MitoB ratio and (C) MitoP/MitoB ratio per hour in turtle hearts. (D) Total aconitase activity as a measure of oxidative damage. Data are means±s.e.m., n=5. Significantly different P-values are shown; for statistical details, see Results.

Fig. 2.

In vivo ROS production in turtle hearts after anoxia and reoxygenation. (A) Experimental groups. Turtles were injected with 1 nmol g−1 MitoB and allowed to equilibrate in normoxia for 5 min, before exposing turtles to 4 h normoxia, 3 h anoxia or 3 h anoxia followed by 1 h reoxygenation in normoxia. (B) In vivo H2O2 production measured by the MitoP/MitoB ratio and (C) MitoP/MitoB ratio per hour in turtle hearts. (D) Total aconitase activity as a measure of oxidative damage. Data are means±s.e.m., n=5. Significantly different P-values are shown; for statistical details, see Results.

In mammalian hearts, the key factors driving excessive production of mitochondrial O2· upon reperfusion are thought to be accumulation of succinate and depletion of adenine nucleotides (Chouchani et al., 2014). So, we next assessed whether these factors change during anoxia in the turtle hearts. We found a minor apparent ∼2-fold elevation in [succinate] upon anoxia to 10±2.3 ng mg−1, which did not achieve significance (Tukey's multiple comparisons test, P=0.2647), and which was restored to normoxic levels upon reoxygenation (Fig. 3A). This contrasts dramatically with the ∼10-fold elevation in [succinate] in the mouse heart upon ischemia to ∼400 ng mg−1 (Chouchani et al., 2014; Martin et al., 2019). There were no significant changes in [ATP] and [ADP] in the turtle heart upon anoxia (two-way ANOVA, P=0.6022; Fig. 3B). The ATP/ADP ratio (Fig. 3C) did decrease significantly in the turtle heart upon anoxia compared with normoxia (Tukey's multiple comparisons test, P=0.0486), probably owing to inactivation of oxidative phosphorylation. The ATP/ADP ratio was restored after 1 h reoxygenation. The lack of change in [ATP] and [ADP] contrasts markedly with the dramatic decreases in these upon ischemia in the mouse heart (Chouchani et al., 2014; Martin et al., 2019). Together, these results suggest that there are major metabolic differences in the turtle heart during anoxia compared with the hearts of mammals, and that the key drivers of mitochondrial O2· production upon reoxygenation do not accumulate during anoxia in turtles. This feature underlies the lack of oxidative damage to their hearts upon reoxygenation after anoxia.

Fig. 3.

Minimal changes in [succinate] and adenine nucleotides with anoxia and reoxygenation in turtle hearts. (A) Absolute succinate concentration in turtle hearts. The statistical differences between the anoxia and normoxia groups and the anoxia and reoxygenation groups were P=0.2647 and P=0.2875, respectively. (B) Absolute adenine nucleotide concentration and (C) ATP/ADP ratio in turtle hearts. Experimental groups as described in Fig. 2, data are means±s.e.m., n=5. Significantly different P-values are shown; for statistical details, see Results.

Fig. 3.

Minimal changes in [succinate] and adenine nucleotides with anoxia and reoxygenation in turtle hearts. (A) Absolute succinate concentration in turtle hearts. The statistical differences between the anoxia and normoxia groups and the anoxia and reoxygenation groups were P=0.2647 and P=0.2875, respectively. (B) Absolute adenine nucleotide concentration and (C) ATP/ADP ratio in turtle hearts. Experimental groups as described in Fig. 2, data are means±s.e.m., n=5. Significantly different P-values are shown; for statistical details, see Results.

Next, we wanted to test whether the turtles' ability to maintain ADP during anoxia was solely responsible for preventing mitochondrial O2· production upon reoxygenation. The presence of ADP upon reoxygenation would allow the mitochondrial proton motive force to be dissipated through the action of the F0F1-ATP synthase and thus prevent mitochondrial O2· production by RET. To test this hypothesis, we assessed the effect of inhibition of the F0F1-ATP synthase on mitochondrial O2· production after anoxia and reperfusion in vivo. To do this, we injected anoxic turtles with the F0F1-ATP synthase inhibitor oligomycin 15 min prior to reoxygenation (Fig. 4A), which led to a significant decrease in F0F1-ATP synthase activity (Student's t-test, P<0.001; Fig. 4B). [ADP] did not change significantly (Šidák's multiple comparisons test, P=0.9936; Fig. 4C), whereas [ATP] (Fig. 4C) and the ATP/ADP ratio (Fig. 4D) were significantly lower in oligomycin-treated turtles compared with controls (Šidák's multiple comparisons test and Student's t-test, P=0.005 and P=0.0326, respectively). This suggests that the lower ATP levels with oligomycin were due to further degradation to purine nucleotide degradation products. Interestingly, [ATP] and the ATP/ADP ratio were around 2-fold lower in turtle hearts with oligomycin than in controls after 1 h of reoxygenation (Fig. 4C,D), similar to the fold-reduction in anoxia (Fig. 3B,C), which suggests that turtles are able to maintain 50% of normal ATP levels by glycolysis, even during normoxia. We found no difference in the MitoP/MitoB ratio (Student's t-test, P=0.8531; Fig. 4E), nor total aconitase activity (Student's t-test, P=0.1840; Fig. 4F) upon reoxygenation of oligomycin-treated turtles compared with controls. This suggests that maintenance of ADP levels is not solely responsible for preventing excessive mitochondrial O2· production in turtles after anoxia and reoxygenation, but that other factors, such as the low accumulation of succinate during anoxia, may also contribute.

Fig. 4.

Inhibition of F0F1-ATP synthase with oligomycin does not increase in vivo ROS production in turtle hearts after anoxia and reoxygenation. (A) Experimental groups. Turtles were injected intravenously with 1 nmol g−1 MitoB and allowed to equilibrate in normoxia for 5 min, before exposing turtles to 3 h anoxia followed by 1 h reoxygenation in normoxia. Turtles were injected intravenously with 0.5 mg kg−1 oligomycin or vehicle 15 min before reoxygenation. (B) Oligomycin sensitive F0F1-ATP synthase (CV) activity of turtle heart homogenate. (C) Absolute adenine nucleotide concentration and (D) ATP/ADP ratio. (E) In vivo H2O2 production measured by MitoP/B ratio. (F) Total aconitase activity as a measure of oxidative damage. Data are means±SEM, n=4. Significantly different P-values are shown; for statistical details, see Results.

Fig. 4.

Inhibition of F0F1-ATP synthase with oligomycin does not increase in vivo ROS production in turtle hearts after anoxia and reoxygenation. (A) Experimental groups. Turtles were injected intravenously with 1 nmol g−1 MitoB and allowed to equilibrate in normoxia for 5 min, before exposing turtles to 3 h anoxia followed by 1 h reoxygenation in normoxia. Turtles were injected intravenously with 0.5 mg kg−1 oligomycin or vehicle 15 min before reoxygenation. (B) Oligomycin sensitive F0F1-ATP synthase (CV) activity of turtle heart homogenate. (C) Absolute adenine nucleotide concentration and (D) ATP/ADP ratio. (E) In vivo H2O2 production measured by MitoP/B ratio. (F) Total aconitase activity as a measure of oxidative damage. Data are means±SEM, n=4. Significantly different P-values are shown; for statistical details, see Results.

Anoxia in vivo in the turtle differs from ischemia in mammals, because in ischemia there is no blood flow whereas in anoxia there is efflux of lactate from the heart, enabling glycolysis to continue (Gruszczyk et al., 2022). Thus, metabolism may differ between the two systems. To assess this, we injected turtles with MitoB in vivo and then, to mimic ischemia, we prepared heart slices and exposed the heart strips to anoxia and reoxygenation in vitro (Fig. 5A). The MitoP/MitoB ratio did not vary significantly between the experimental groups (one-way ANOVA, P=0.9890; Fig. 5B), and although [succinate] was slightly higher in the anoxic in vitro incubation compared with normoxia (Tukey's multiple comparisons test, P=0.1189; Fig. 5C), ATP concentration (Fig. 5D) and ATP/ADP ratios (Fig. 5E) decreased during anoxia compared with the control (Šidák's and Tukey's multiple comparisons test, P=0.1658 and P=0.2345, respectively). However, these small, non-significant changes were similar to those in vivo in the anoxic turtle (Fig. 2). We repeated these experiments in the presence of oligomycin (Fig. 5A), which inhibited the activity of the F0F1-ATP synthase (Student's t-test, P<0.001; Fig. 5G), but this did not affect the outcome (Fig. 5B–E). Thus, even under conditions mimicking ischemia, with or without artificial inhibition of the F0F1-ATP synthase with oligomycin, the metabolic changes are similar to anoxia in vivo and turtle heart tissue does not produce excessive O2· after anoxia and reoxygenation. The relative force of contraction of the heart strips recovered after anoxia and reoxygenation was similar in all groups (Fig. 5F). This suggests that there was no functional effect of anoxia and reoxygenation on turtle hearts regardless of the presence of oligomycin.

Fig. 5.

In vitro incubation of turtle heart strips with the F0F1-ATP synthase inhibitor oligomycin does not increase ROS production measured by MitoB/P. (A) Experimental groups. Turtles were injected intravenously with MitoB and allowed to equilibrate for 15 min before euthanasia and mounting of heart strips on force transducers. After stabilisation of heart strips in turtle Ringer’s solution at 25°C, they were exposed to either (1) 90 min normoxia (normoxic control), (2) 30 min normoxia, followed by 60 min anoxia (anoxic control), (3) 60 min anoxia (98% N2, 2% CO2)/30 min reoxygenation (A/R) or (4) A/R with 10 µmol l−1 oligomycin. Oligomycin was added after 45 min of anoxia as indicated. (B) MitoB/P ratio, (C) absolute succinate concentration, (D) absolute adenine nucleotide concentration and (E) ATP/ADP ratio of turtle heart strips. (F) Relative force of contraction of heart strips during the experiment. (G) ATP synthase activity of a heart strip incubated with 10 µmol l−1 oligomycin in parallel to the experimental strips for 30 min. Data are means±s.e.m., n=6. Significantly different P-values are shown; for statistical details, see Results.

Fig. 5.

In vitro incubation of turtle heart strips with the F0F1-ATP synthase inhibitor oligomycin does not increase ROS production measured by MitoB/P. (A) Experimental groups. Turtles were injected intravenously with MitoB and allowed to equilibrate for 15 min before euthanasia and mounting of heart strips on force transducers. After stabilisation of heart strips in turtle Ringer’s solution at 25°C, they were exposed to either (1) 90 min normoxia (normoxic control), (2) 30 min normoxia, followed by 60 min anoxia (anoxic control), (3) 60 min anoxia (98% N2, 2% CO2)/30 min reoxygenation (A/R) or (4) A/R with 10 µmol l−1 oligomycin. Oligomycin was added after 45 min of anoxia as indicated. (B) MitoB/P ratio, (C) absolute succinate concentration, (D) absolute adenine nucleotide concentration and (E) ATP/ADP ratio of turtle heart strips. (F) Relative force of contraction of heart strips during the experiment. (G) ATP synthase activity of a heart strip incubated with 10 µmol l−1 oligomycin in parallel to the experimental strips for 30 min. Data are means±s.e.m., n=6. Significantly different P-values are shown; for statistical details, see Results.

Finally, we tested whether the antioxidant capacity is higher in turtle mitochondria compared with a mammal (mouse) or an anoxia-intolerant reptile (green iguana) (Fig. 6). We found that isolated turtle heart mitochondria did not scavenge H2O2 more effectively than heart mitochondria from mice or green iguanas under conditions for state III respiration (malate, glutamate and ADP) or RET (succinate alone) (Šidák's multiple comparisons test, P=0.6774, 0.0893, 0.7982 and 0.0894, respectively). Moreover, all three animals tested produced H2O2 to a similar degree when incubated with succinate alone (two-way ANOVA, P=0.3091), showing that both turtle and green iguana mitochondria have the capacity to produce H2O2 by RET in vitro. It just seems to be the case that at least turtle hearts are able to avoid producing H2O2in vivo.

Fig. 6.

In vitro H2O2 consumption and production rates in isolated mitochondria from mouse, turtle and green iguana. H2O2 consumption rates were measured by an Amplex UltraRed microplate based fluorescent assay and H2O2 production rates were measured with an Oroboros O2K respirometer fitted with a FluoO2K module with filters specific for Amplex UltraRed. State III respiration was measured with malate, glutamate and ADP and reverse electron transfer (RET) was measured with succinate alone. Data are means±s.e.m., n=3. Significantly different P-values are shown; for statistical details, see Results.

Fig. 6.

In vitro H2O2 consumption and production rates in isolated mitochondria from mouse, turtle and green iguana. H2O2 consumption rates were measured by an Amplex UltraRed microplate based fluorescent assay and H2O2 production rates were measured with an Oroboros O2K respirometer fitted with a FluoO2K module with filters specific for Amplex UltraRed. State III respiration was measured with malate, glutamate and ADP and reverse electron transfer (RET) was measured with succinate alone. Data are means±s.e.m., n=3. Significantly different P-values are shown; for statistical details, see Results.

Turtles are able to return from complete anoxia without sustaining oxidative tissue damage to their hearts (Bickler and Buck, 2007; Bundgaard et al., 2018; Milton et al., 2007; Wasser et al., 1992), in contrast to mammals, which encounter severe oxidative damage after ischemia and reperfusion (Martin et al., 2019; Murphy and Steenbergen, 2008; Yellon and Hausenloy, 2007). In this study, we investigated whether turtles avoid oxidative damage to their hearts because they do not produce excessive O2· and H2O2 after anoxia and reoxygenation in vivo, and whether this is because they are able to avoid the metabolic changes that are thought to drive reperfusion damage in mammals.

We show that there is no increase in the MitoP/MitoB ratio (Fig. 2B,C) or decrease in the O2·-sensitive marker enzyme aconitase (Fig. 2D) in turtle hearts after 3 h anoxia and 1 h reoxygenation in vivo, which shows that there is no burst of mitochondrial H2O2 and O2· production upon reoxygenation of anoxic turtle hearts. In comparison, extensive H2O2 production is observed in mouse hearts exposed to 30 min ischemia at 37°C and 15 min of reperfusion. This is reflected in ratios of ∼0.125 MitoP/MitoB per hour after 30 min ischemia and 15 min reperfusion, compared with ∼0.0125 MitoP/MitoB per hour after 45 min of normoxia (Chouchani et al., 2014), and in an ∼50% decrease in aconitase activity in mice with ischemia and reperfusion (Bulteau et al., 2005; Chouchani et al., 2014; Gardner, 2002). Absence of mitochondrial O2· in the turtle heart in vivo is supported by the findings in turtle brain slices and neuronal cell cultures where there was no excessive production of H2O2 after exposure to anoxia and reoxygenation in vitro (Milton et al., 2007; Nayak et al., 2009; Pamenter et al., 2007). Furthermore, studies have found that other downstream cellular markers of oxidative stress, such as lipid peroxidation (Willmore and Storey, 1997a) and the GSH/GSSG ratio (Willmore and Storey, 1997b), only increase slightly or not at all in turtle tissues after anoxia and reoxygenation in vivo. This is in keeping with the remarkable tolerance of turtle hearts to maintain function after in situ anoxia or ischemia and reoxygenation (Bundgaard et al., 2018; Wasser et al., 1992). Together, this shows that turtle hearts avoid oxidative reoxygenation damage because there is no burst of mitochondrial O2· upon reoxygenation after anoxia.

We next considered how mitochondria are thought to produce O2· upon reperfusion. Electrons delivered from accumulated succinate led to a highly reduced ubiquinone pool, and low levels of F0F1-ATP synthase substrate (ADP) prevent dissipation of the proton motive force through ATP synthesis. Together, this leads to RET from accumulated succinate to complex I where superoxide is produced (Fig. 1) (Chouchani et al., 2014; Murphy, 2009). Here, we show that the factors driving O2· and H2O2 in mammals, succinate accumulation and ADP depletion, are largely unaffected in the turtle heart after anoxia in vivo (Fig. 3) compared with those in the ischemic mammalian heart, where oxidative damage can occur (Chouchani et al., 2014; Martin et al., 2018). Anoxic incubation of heart strips also caused a modest, but insignificant increase in [succinate] and transient decrease in ATP levels (Fig. 5). Even longer exposure (14–28 days) to anoxia at 4–5°C also only causes a modest, albeit statistically significant, increase in succinate levels in turtles (Bundgaard et al., 2019b) and in the closely related freshwater turtle Chrysemys picta (Buck, 2000), suggesting that although succinate accumulates in the turtle heart, it never reaches the high levels driving oxidative damage after ischemia in mammalian hearts.

Similarly, although there was a significant ∼50% decrease in the ATP/ADP ratio with both in vivo and in vitro anoxia, this was restored to normoxic levels in the group with 1 h reoxygenation (Figs 3 and 5). Only in the experimental groups that were treated with oligomycin did ATP levels stay low after reoxygenation (Figs 4 and 5), consistent with a lower ATP production when F0F1-ATP synthase is inhibited. This was associated with little loss of adenine nucleotides during anoxia in turtle hearts. This is consistent with the fact that ADP is not degraded in turtles hearts with prolonged anoxia (Bundgaard et al., 2019b; Galli et al., 2013; Kelly and Storey, 1988; Stecyk et al., 2009) and reflects the turtles' remarkable ability to maintain energy balance during anoxia (Bickler and Buck, 2007; Bundgaard et al., 2019b; Jackson, 2002; Stecyk et al., 2009). It likely also reflects the difference between ischemia, where absence of blood flow not only decreases oxygen levels but also causes acidification and accumulation of other end products (Gruszczyk et al., 2022), and whole-animal anoxia, where the circulation remains active to remove end-products and provide metabolic fuel. Together, this shows that the factors necessary to drive excessive production of O2· and H2O2 in the mammalian heart upon reperfusion after ischemia are not present after anoxia in turtle hearts.

We considered two possible explanations for these findings: (1) succinate levels never accumulate to levels that are high enough to drive harmful O2· and H2O2 production, or (2) turtle mitochondrial antioxidant capacity is so high that excess production of O2· and H2O2 would be instantly buffered under these conditions, and therefore not be reflected in an increase in the MitoP/MitoB ratio. In support of the first explanation, we have shown here that even with artificial inhibition of F0F1-ATP synthase, [succinate] is never high enough to drive extensive production of O2· and H2O2 in turtle heart tissue (Figs 35), and therefore preservation of ATP levels alone is not enough to explain why turtles avoid oxidative reoxygenation damage. In this study, we were unable to elevate succinate levels to those observed in mammals using turtle tissue, but previous work in isolated mitochondria demonstrated that turtle mitochondria exposed to mmol l−1 concentrations of succinate do generate ROS via RET (Bundgaard et al., 2019b). This highlights the importance of the turtles' ability to avoid extensive succinate accumulation during anoxia. This likely reflects the turtle's ability to rapidly enter metabolic depression when they encounter anoxia and their ability to shift their energy metabolism to anaerobic glycolysis, sequestering accumulated lactate in their shells (Jackson, 2002). Succinate may also be excreted from the cardiomyocytes during anoxia via the monocarboxylate transporter 1 (Prag et al., 2020), although succinate did not accumulate in anoxic heart strips (Fig. 5) and does not accumulate in the blood with prolonged anoxia in turtles (Buck, 2000).

Furthermore, we tested hypothesis 2 and showed that the antioxidant capacity of isolated turtle mitochondria was comparable to that in mouse mitochondria but higher than that in green iguana mitochondria (Fig. 6), suggesting that turtle heart mitochondria do not have an exceptionally high antioxidant capacity compared with mammals or other reptiles. Thus, although antioxidants may contribute to regulation of mitochondrial ROS levels in vivo, high levels of antioxidants are unlikely to be the main mechanism that limits ROS production upon reoxygenation compared with mammals.

Finally, the absence of reoxygenation damage even in the presence of oligomycin in turtle hearts could be due to a third explanation: that there are other intrinsic aspects of turtle metabolism and mitochondrial physiology that contribute to prevent O2· and H2O2 production upon reoxygenation as well as low succinate accumulation and preservation of ADP levels. This could include activation of uncoupling factors of the inner mitochondrial membrane, e.g. via uncoupling proteins or KATP-channels (Pamenter et al., 2008; Staples and Buck, 2009), direct regulation of the electron transport chain complexes via post-translational modifications such as S-nitrosation of complex I (Bundgaard et al., 2018; Chouchani et al., 2013) or phosphorylation of complex II (Mathers et al., 2016), or through regulation of the highly stable supercomplex interactions between complexes I and III in turtles (Bundgaard et al., 2019a). Decreases in mitochondrial respiration rate (Bundgaard et al., 2019a; Galli et al., 2013; Pamenter et al., 2016) and capacity for H2O2 production (Bundgaard et al., 2019a; Milton et al., 2007) with prolonged anoxia in turtles further points to intrinsic regulation of mitochondrial physiology with anoxia.

In conclusion, we show that the low rates of mitochondrial O2· production is the main factor responsible for the absence of tissue damage upon reoxygenation in turtles. This is likely because of a low total accumulation of succinate and the lack of degradation of adenine nucleotides during anoxia in the turtles. These factors would remove the potential for RET to occur upon reoxygenation in anoxia-tolerant turtles.

The authors would like to thank Heidi Meldgaard Jensen for her help with animal care at Zoophysiology, Aarhus University.

Author contributions

Conceptualization: A.B., A.M.J., G.L.J.G., M.P.M., A.F.; Methodology: A.B., A.V.G., H.A.P., C.W., A.M., A.M.J., M.P.M.; Formal analysis: A.B.; Investigation: A.B., A.V.G., H.A.P., C.W., A.M., I.M.R., G.L.J.G.; Resources: G.L.J.G., M.P.M., A.F.; Data curation: A.B., A.V.G., H.A.P.; Writing - original draft: A.B.; Writing - review & editing: A.B., A.V.G., H.A.P., C.W., I.M.R., A.M.J., G.L.J.G., M.P.M., A.F.; Visualization: A.B.; Supervision: A.M.J., M.P.M., A.F.; Project administration: A.B., M.P.M., A.F.; Funding acquisition: A.B., M.P.M., A.F.

Funding

This work was supported by a research grant (34435) from VILLUM FONDEN to A.B., an Aarhus University Research Foundation (NOVA grant AUFF-E-2016-9-37) to A.F. and by the Medical Research Council UK (MC_UU_00028/4) and a Wellcome Trust Investigator award (220257/Z/20/Z) to M.P.M. Deposited in PMC for release after 6 months.

Data availability

All relevant data can be found within the article.

Bickler
,
P. E.
and
Buck
,
L. T.
(
2007
).
Hypoxia tolerance in reptiles, amphibians, and fishes: life with variable oxygen availability
.
Annu. Rev. Physiol.
69
,
145
-
170
.
Buck
,
L. T.
(
2000
).
Succinate and alanine as anaerobic end-products in the diving turtle (Chrysemys picta bellii)
.
Comp. Biochem. Physiol. B Biochem. Mol. Biol.
126
,
409
-
413
.
Bulteau
,
A.-L.
,
Lundberg
,
K. C.
,
Ikeda-Saito
,
M.
,
Isaya
,
G.
and
Szweda
,
L. I.
(
2005
).
Reversible redox-dependent modulation of mitochondrial aconitase and proteolytic activity during in vivo cardiac ischemia/reperfusion
.
Proc. Natl. Acad. Sci. USA
102
,
5987
-
5991
.
Bundgaard
,
A.
,
James
,
A. M.
,
Joyce
,
W.
,
Murphy
,
M. P.
and
Fago
,
A.
(
2018
).
Suppression of reactive oxygen species generation in heart mitochondria from anoxic turtles: the role of complex I S-nitrosation
.
J. Exp. Biol.
221
,
jeb174391
.
Bundgaard
,
A.
,
Qvortrup
,
K.
,
Rasmussen
,
L. J.
and
Fago
,
A.
(
2019a
).
Turtles maintain mitochondrial integrity but reduce mitochondrial respiratory capacity in the heart after cold acclimation and anoxia
.
J. Exp. Biol.
222
,
jeb200410
.
Bundgaard
,
A.
,
James
,
A. M.
,
Gruszczyk
,
A. V.
,
Martin
,
J.
,
Murphy
,
M. P.
and
Fago
,
A.
(
2019b
).
Metabolic adaptations during extreme anoxia in the turtle heart and their implications for ischemia-reperfusion injury
.
Sci. Rep
9
,
2850
.
Chouchani
,
E. T.
,
Methner
,
C.
,
Nadtochiy
,
S. M.
,
Logan
,
A.
,
Pell
,
V. R.
,
Ding
,
S.
,
James
,
A. M.
,
Cochemé
,
H. M.
,
Reinhold
,
J.
,
Lilley
,
K. S.
et al. 
(
2013
).
Cardioprotection by S-nitrosation of a cysteine switch on mitochondrial complex I
.
Nat. Med.
19
,
753
-
759
.
Chouchani
,
E. T.
,
Pell
,
V. R.
,
Gaude
,
E.
,
Aksentijević
,
D.
,
Sundier
,
S. Y.
,
Robb
,
E. L.
,
Logan
,
A.
,
Nadtochiy
,
S. M.
,
Ord
,
E. N. J.
,
Smith
,
A. C.
et al. 
(
2014
).
Ischaemic accumulation of succinate controls reperfusion injury through mitochondrial ROS
.
Nature
515
,
431
-
435
.
Cochemé
,
H. M.
,
Quin
,
C.
,
Murphy
,
M. P.
,
Hartley
,
C.
,
Partridge
,
L.
,
Michael
,
S.
,
McQuaker
,
J.
,
Cabreiro
,
F.
,
Logan
,
A.
,
Prime
,
T. A.
et al. 
(
2011
).
Measurement of H2O2 within living Drosophila during aging using a ratiometric mass spectrometry probe targeted to the mitochondrial matrix
.
Cell Metab.
13
,
340
-
350
.
Cochemé
,
H. M.
,
Logan
,
A.
,
Prime
,
T. A.
,
Abakumova
,
I.
,
Quin
,
C.
,
McQuaker
,
S. J.
,
Patel
,
J. V.
,
Fearnley
,
I. M.
,
James
,
A. M.
,
Porteous
,
C. M.
et al. 
(
2012
).
Using the mitochondria-targeted ratiometric mass spectrometry probe MitoB to measure H2O2 in living Drosophila
.
Nat. Protoc
7
,
946
-
958
.
Galli
,
G. L. J.
,
Richards
,
J. G.
(
2014
).
Mitochondria from anoxia-tolerant animals reveal common strategies to survive without oxygen
.
J. Comp. Physiol. B Biochem. Syst. Environ. Physiol.
184
,
285
-
302
.
Galli
,
G. L. J.
,
Lau
,
G. Y.
and
Richards
,
J. G.
(
2013
).
Beating oxygen: chronic anoxia exposure reduces mitochondrial F1FO-ATPase activity in turtle (Trachemys scripta) heart
.
J. Exp. Biol
216
,
3283
-
3293
.
Gardner
,
P. R.
(
2002
).
Aconitase: sensitive target and measure of superoxide
.
Methods Enzymol.
349
,
9
-
23
.
Gardner
,
P. R.
,
Nguyen
,
D. D.
and
White
,
C. W.
(
1994
).
Aconitase is a sensitive and critical target of oxygen poisoning in cultured mammalian cells and in rat lungs
.
Proc. Natl. Acad. Sci. USA
91
,
12248
-
12252
.
Gruszczyk
,
A. V.
,
Casey
,
A. M.
,
James
,
A. M.
,
Prag
,
H. A.
,
Burger
,
N.
,
Bates
,
G. R.
,
Hall
,
A. R.
,
Allen
,
F. M.
,
Krieg
,
T.
,
Saeb-Parsy
,
K.
et al. 
(
2022
).
Mitochondrial metabolism and bioenergetic function in an anoxic isolated adult mouse cardiomyocyte model of in vivo cardiac ischemia–reperfusion injury
.
Redox Biol.
54
,
102368
.
Herbert
,
C. V.
and
Jackson
,
D. C.
(
1985
).
Temperature effects on the responses to prolonged submergence in the turtle Chrysemys picta bellii. I. Blood acid-base and ionic changes during and following anoxic submergence
.
Physiol. Zool.
58
,
655
-
669
.
Hermes-Lima
,
M.
and
Zenteno-Savín
,
T.
(
2002
).
Animal response to drastic changes in oxygen availability and physiological oxidative stress
.
Comp. Biochem. Physiol. C Toxicol. Pharmacol.
133
,
537
-
556
.
Hermes-Lima
,
M.
,
Moreira
,
D. C.
,
Rivera-Ingraham
,
G. A.
,
Giraud-Billoud
,
M.
,
Genaro-Mattos
,
T. C.
and
Campos
,
É. G.
(
2015
).
Preparation for oxidative stress under hypoxia and metabolic depression: revisiting the proposal two decades later
.
Free Radic. Biol. Med.
89
,
1122
-
1143
.
Hochachka
,
P. W.
,
Owen
,
T. G.
,
Allen
,
J. F.
and
Whittow
,
G. C.
(
1975
).
Multiple end products of anaerobiosis in diving vertebrates
.
Comp. Biochem. Physiol. B Biochem.
50
,
17
-
22
.
Jackson
,
D. C.
(
2002
).
Hibernating without oxygen: physiological adaptations of the painted turtle
.
J. Physiol
543
,
731
-
737
.
Joyce
,
W.
,
Gesser
,
H.
and
Wang
,
T.
(
2014
).
Purinoceptors exert negative inotropic effects on the heart in all major groups of reptiles
.
Comp. Biochem. Physiol. A Mol. Integr. Physiol.
171
,
16
-
22
.
Kelly
,
D. A.
and
Storey
,
K. B.
(
1988
).
Organ-specific control of glycolysis in anoxic turtles
.
Am. J. Physiol. Regul. Integr. Comp. Physiol
255
,
R774
-
R779
.
Kowaltowski
,
A. J.
,
de Souza-Pinto
,
N. C.
,
Castilho
,
R. F.
and
Vercesi
,
A. E.
(
2009
).
Mitochondria and reactive oxygen species
.
Free Radic. Biol. Med.
47
,
333
-
343
.
Logan
,
A.
,
Cochemé
,
H. M.
,
Li Pun
,
P. B.
,
Apostolova
,
N.
,
Smith
,
R. A. J.
,
Larsen
,
L.
,
Larsen
,
D. S.
,
James
,
A. M.
,
Fearnley
,
I. M.
,
Rogatti
,
S.
et al. 
(
2014a
).
Using exomarkers to assess mitochondrial reactive species in vivo
.
Biochim. Biophys. Acta Gen. Subj.
1840
,
923
-
930
.
Logan
,
A.
,
Shabalina
,
I. G.
,
Prime
,
T. A.
,
Rogatti
,
S.
,
Kalinovich
,
A. V.
,
Hartley
,
R. C.
,
Budd
,
R. C.
,
Cannon
,
B.
and
Murphy
,
M. P.
(
2014b
).
In vivo levels of mitochondrial hydrogen peroxide increase with age in mtDNA mutator mice
.
Aging Cell
13
,
765
-
768
.
Lopez-Fabuel
,
I.
,
Le Douce
,
J.
,
Logan
,
A.
,
James
,
A. M.
,
Bonvento
,
G.
,
Murphy
,
M. P.
,
Almeida
,
A.
and
Bolaños
,
J. P.
(
2016
).
Complex I assembly into supercomplexes determines differential mitochondrial ROS production in neurons and astrocytes
.
Proc. Natl. Acad. Sci. USA
113
,
13063
-
13068
.
Mans
,
C.
(
2008
).
Venipuncture techniques in chelonian species
.
Lab. Anim.
37
,
303
-
304
.
Martin
,
J. L.
,
Gruszczyk
,
A. V.
,
Beach
,
T. E.
,
Murphy
,
M. P.
and
Saeb-Parsy
,
K.
(
2018
).
Mitochondrial mechanisms and therapeutics in ischaemia reperfusion injury
.
Pediatr. Nephrol.
34
,
1167
-
1174
.
Martin
,
J. L.
,
Costa
,
A. S. H.
,
Gruszczyk
,
A. V.
,
Beach
,
T. E.
,
Allen
,
F. M.
,
Prag
,
H. A.
,
Hinchy
,
E. C.
,
Mahbubani
,
K.
,
Hamed
,
M.
,
Tronci
,
L.
et al. 
(
2019
).
Succinate accumulation drives ischaemia-reperfusion injury during organ transplantation
.
Nat. Metab.
1
,
966
-
974
.
Mathers
,
K. E.
,
McFarlane
,
S. V.
,
Zhao
,
L.
and
Staples
,
J. F.
(
2016
).
Regulation of mitochondrial metabolism during hibernation by reversible suppression of electron transport system enzymes
.
J. Comp. Physiol. B Biochem. Syst. Environ. Physiol.
187
,
227
-
234
.
Milton
,
S. L.
,
Nayak
,
G.
,
Kesaraju
,
S.
,
Kara
,
L.
and
Prentice
,
H. M.
(
2007
).
Suppression of reactive oxygen species production enhances neuronal survival in vitro and in vivo in the anoxia-tolerant turtle Trachemys scripta
.
J. Neurochem.
101
,
993
-
1001
.
Munro
,
D.
,
Banh
,
S.
,
Sotiri
,
E.
,
Tamanna
,
N.
and
Treberg
,
J. R.
(
2016
).
The thioredoxin and glutathione-dependent H2O2 consumption pathways in muscle mitochondria: involvement in H2O2 metabolism and consequence to H2O2 efflux assays
.
Free Radic. Biol. Med.
96
,
334
-
346
.
Murphy
,
M. P.
(
2009
).
How mitochondria produce reactive oxygen species
.
Biochem. J.
417
,
1
-
13
.
Murphy
,
M. P.
(
2016
).
Understanding and preventing mitochondrial oxidative damage
.
Biochem. Soc. Trans.
44
,
1219
-
1226
.
Murphy
,
E.
and
Steenbergen
,
C.
(
2008
).
Mechanisms underlying acute protection from cardiac ischemia-reperfusion injury
.
Physiol. Rev.
88
,
581
-
609
.
Nayak
,
G.
,
Prentice
,
H. M.
and
Milton
,
S. L.
(
2009
).
Role of neuroglobin in regulating reactive oxygen species in the brain of the anoxia-tolerant turtle Trachemys scripta
.
J. Neurochem.
110
,
603
-
612
.
Overgaard
,
J.
,
Wang
,
T.
,
Nielsen
,
O. B.
and
Gesser
,
H.
(
2005
).
Extracellular determinants of cardiac contractility in the cold anoxic turtle
.
Physiol. Biochem. Zool.
78
,
976
-
995
.
Pamenter
,
M. E.
,
Richards
,
M. D.
and
Buck
,
L. T.
(
2007
).
Anoxia-induced changes in reactive oxygen species and cyclic nucleotides in the painted turtle
.
J. Comp. Physiol. B Biochem. Syst. Environ. Physiol.
177
,
473
-
481
.
Pamenter
,
M. E.
,
Shin
,
D. S.-H.
,
Cooray
,
M.
and
Buck
,
L. T.
(
2008
).
Mitochondrial ATP-sensitive K+ channels regulate NMDAR activity in the cortex of the anoxic western painted turtle
.
J. Physiol.
586
,
1043
-
1058
.
Pamenter
,
M. E.
,
Gomez
,
C. R.
,
Richards
,
J. G.
and
Milsom
,
W. K.
(
2016
).
Mitochondrial responses to prolonged anoxia in brain of red-eared slider turtles
.
Biol. Lett.
12
,
20150797
.
Prag
,
H. A.
,
Gruszczyk
,
A. V.
,
Huang
,
M. M.
,
Beach
,
T. E.
,
Young
,
T.
,
Tronci
,
L.
,
Nikitopoulou
,
E.
,
Mulvey
,
J. F.
,
Ascione
,
R.
,
Hadjihambi
,
A.
et al. 
(
2020
).
Mechanism of succinate efflux upon reperfusion of the ischemic heart
.
Cardiovasc. Res.
117
,
1188
-
1201
.
Prag
,
H. A.
,
Aksentijevic
,
D.
,
Dannhorn
,
A.
,
Giles
,
A. V.
,
Mulvey
,
J. F.
,
Sauchanka
,
O.
,
Du
,
L.
,
Bates
,
G.
,
Reinhold
,
J.
,
Kula-Alwar
,
D.
et al. 
(
2022
).
Ischemia-selective cardioprotection by malonate for ischemia/reperfusion injury
.
Circ. Res.
131
,
528
-
541
.
Rice
,
M. E.
,
Lee
,
E. J. K.
and
Choy
,
Y.
(
1995
).
High levels of ascorbic acid, not glutathione, in the CNS of anoxia-tolerant reptiles contrasted with levels in anoxia-intolerant species
.
J. Neurochem.
64
,
1790
-
1799
.
Salin
,
K.
,
Auer
,
S. K.
,
Rudolf
,
A. M.
,
Anderson
,
G. J.
,
Cairns
,
A. G.
,
Mullen
,
W.
,
Hartley
,
R. C.
,
Selman
,
C.
and
Metcalfe
,
N. B.
(
2015
).
Individuals with higher metabolic rates have lower levels of reactive oxygen species in vivo
.
Biol. Lett.
11
,
20150538
.
Salin
,
K.
,
Auer
,
S. K.
,
Villasevil
,
E. M.
,
Anderson
,
G. J.
,
Cairns
,
A. G.
,
Mullen
,
W.
,
Hartley
,
R. C.
and
Metcalfe
,
N. B.
(
2017
).
Using the MitoB method to assess levels of reactive oxygen species in ecological studies of oxidative stress
.
Sci. Rep.
7
,
41228
.
Salin
,
K.
,
Villasevil
,
E. M.
,
Anderson
,
G. J.
,
Auer
,
S. K.
,
Selman
,
C.
,
Hartley
,
R. C.
,
Mullen
,
W.
,
Chinopoulos
,
C.
and
Metcalfe
,
N. B.
(
2018
).
Decreased mitochondrial metabolic requirements in fasting animals carry an oxidative cost
.
Funct. Ecol.
32
,
2149
-
2157
.
Staples
,
J. F.
and
Buck
,
L. T.
(
2009
).
Matching cellular metabolic supply and demand in energy-stressed animals
.
Comp. Biochem. Physiol. A Mol. Integr. Physiol.
153
,
95
-
105
.
Stecyk
,
J. A. W.
,
Bock
,
C.
,
Overgaard
,
J.
,
Wang
,
T.
,
Farrell
,
A. P.
and
Pörtner
,
H.-O.
(
2009
).
Correlation of cardiac performance with cellular energetic components in the oxygen-deprived turtle heart
.
Am. J. Physiol. Regul. Integr. Comp. Physiol
297
,
R756
-
R768
.
Strehler
,
B. L.
(
1974
).
Adenosine-5′-triphosphate and creatine phosphate
:
determination with luciferase
. In Methods of Enzymatic Analysis, 2nd edn (ed. H. U. Bergmeyer), pp.
2112
-
2126.
Academic Press.
Ultsch
,
G. R.
and
Jackson
,
D. C.
(
1982
).
Long-term submergence at 3°C of the turtle Chrysemys picta bellii in normoxic and severely hypoxic water III. Effects of changes in ambient pO2 and subsequent air breathing
.
J. Exp. Biol.
97
,
87
-
99
.
Wasser
,
J. S.
,
Meinertz
,
E. A.
,
Chang
,
S. Y.
,
Lawler
,
R. G.
and
Jackson
,
D. C.
(
1992
).
Metabolic and cardiodynamic responses of isolated turtle hearts to ischemia and reperfusion
.
Am. J. Physiol.
262
,
437
-
443
.
Williams
,
C. L.
and
Hicks
,
J. W.
(
2016
).
Continuous arterial PO2 profiles in unrestrained, undisturbed aquatic turtles during routine behaviors
.
J. Exp. Biol.
219
,
3616
-
3625
.
Willmore
,
W. G.
and
Storey
,
K. B.
(
1997a
).
Antioxidant systems and anoxia tolerance in a freshwater turtle Trachemys scripta elegans
.
Mol. Cell. Biochem.
170
,
177
-
185
.
Willmore
,
W. G.
and
Storey
,
K. B.
(
1997b
).
Glutathione systems and anoxia tolerance in turtles
.
Am. J. Physiol.
273
,
219
-
225
.
Yellon
,
D. M.
and
Hausenloy
,
D. J.
(
2007
).
Myocardial reperfusion injury
.
New Engl. J. Med. Rev.
357
,
1121
-
1135
.

Competing interests

The authors declare no competing or financial interests.