Mitochondrial responses to anoxia-reoxygenation exposure in crucian carp (Carassius carassius)
Gigi Y. Lau, Lucie Gerber, Anette Johansen, Helge-Andre Dahl, May-Kristin Torp, Georgina Bates, Michael P. Murphy, Lars Eide, Kåre-Olav Stensløkken, Göran E. Nilsson, Sjannie Lefevre
TL;DR
Crucian carp mitochondria produce less harmful oxygen byproducts during oxygen deprivation and recovery, helping them survive in low-oxygen environments.
Contribution
The study reveals tissue-specific mitochondrial adaptations in crucian carp that reduce oxidative stress during anoxia-reoxygenation.
Findings
Crucian carp mitochondria emit up to 4-fold less hydrogen peroxide than common carp.
Heart mitochondria show a 40% reduction in Complex II-mediated flux after anoxia-reoxygenation.
Tissue-specific increases in protein and mtDNA oxidation suggest varied oxidative stress responses.
Abstract
The crucian carp (Carassius carassius) is one of the most anoxia-tolerant vertebrates. While physiological underpinnings of its ability to withstand O2 deprivation are well studied, the ability to tolerate the return to normoxia is still enigmatic. Such reoxygenation is associated with detrimental oxidation damage in other organisms, where mitochondria play a central role in the damaging effects. This leads to the question whether mitochondrial adaptations play a central role in the anoxia and reoxygenation tolerance of crucian carp. We here address whether mitochondria from crucian carp circumvent the negative effects of anoxia–reoxygenation exposure, namely the generation of reactive oxygen species (ROS) and subsequent oxidative stress. Crucian carp brain and heart mitochondria generated up to 4-fold less hydrogen peroxide (H2O2; a major ROS) compared with the closely related,…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8- —Norges Forskningsrådhttp://dx.doi.org/10.13039/501100005416
- —UiO:Life Science Convergence Environments
- —Erasmus+
- —Medical Research Councilhttp://dx.doi.org/10.13039/501100000265
- —University of Oslohttp://dx.doi.org/10.13039/501100005366
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsPhysiological and biochemical adaptations · Cancer, Hypoxia, and Metabolism · Mitochondrial Function and Pathology
INTRODUCTION
In aquatic environments, oxygen levels can fluctuate between hyperoxic and hypoxic episodes of varying duration and severity. Among these environments are lakes and ponds that freeze over during winter. A thick layer of snow-covered ice will block both photosynthetic activity and O_2_ diffusion from the air into the water. These factors combined with respiration of the organisms present cause oxygen levels to gradually decline during winter to near or complete depletion of dissolved O_2_ until it is slowly replenished in the spring as ice melts. Such conditions can lead to one of the most extreme naturally occurring settings of O_2_ limitation. Unlike ischemia–reperfusion in mammals, these conditions do not impose sudden oxidative stress but allowed for the evolution of physiological strategies to survive extended periods of low oxygen and subsequent reoxygenation. Species inhabiting these environments offer models to investigate adaptations for surviving cycles of anoxia and subsequent reoxygenation. Studying their mitochondria in this context provides insight into natural adaptations for anoxia tolerance and energy efficiency, offering a complementary perspective to research focused on ischemia–reperfusion. The crucian carp (Carassius carassius) and some species of freshwater turtles of the Trachemys and Chrysemys genera, which all overwinter in frozen-over ponds, are known for their remarkable anoxia tolerance (Nilsson and Lutz, 2004). Yet, the physiological strategies employed to combat anoxia–reoxygenation can be markedly different among anoxia-tolerant animals and plasticity in response to O_2_ limitation typically develops over hours to days (Bundgaard et al., 2020a; Fago, 2022; Lefevre and Nilsson, 2023; Lutz and Nilsson, 1997). For example, freshwater turtles drastically decrease metabolic rate to maintain energy balance, i.e. ATP turnover (Nilsson and Lutz, 2004), whereas crucian carp stay relatively active (albeit reducing swimming activity) and maintain brain function (Lefevre et al., 2017; Nilsson, 2001) and cardiac output (Stecyk et al., 2004) during anoxic periods. Furthermore, crucian carp heart mitochondria maintain morphology and membrane potential, even under anoxic conditions (Scott et al., 2024). Among the crucian carp's unique strategies is having the highest liver glycogen levels measured in teleosts to support glycolysis during the anoxic period (Vornanen, 1994) and then generating ethanol as an alternative glycolytic end product to lactate, which can be excreted across the gills (Johnston and Bernard, 1983). In combination with a somewhat reduced energy demand, this unique ability to support high glycolytic activity allows the crucian carp to maintain ATP turnover in anoxia throughout the winter period. This set of adaptive traits (Lefevre and Nilsson, 2023) makes crucian carp a model organism of choice to study natural mechanisms to cope with cycles of anoxia and reoxygenation. While many aspects of how crucian carp can tolerate extended anoxic periods have been examined, how they are able to deal with reoxygenation stress after anoxia is far less studied.
Reoxygenating a previously oxygen-deprived tissue presents a unique challenge. While recovering aerobic respiration rate as soon as possible would allow mitochondria to resume normoxic ATP production levels, this could come at the potential cost of hazardous generation rates of reactive oxygen species (ROS), specifically at the mitochondrial complex I ROS generation site (Murphy, 2009). If this excess of produced ROS is not scavenged by antioxidant mechanisms, it can cause oxidative damage to tissues. The excess ROS production is therefore considered the culprit for ischemia–reperfusion injury in mammalian models (Chouchani et al., 2016; Wu et al., 2018), where the recovery of O_2_ supply to previously anoxic tissue leads to tissue damage and is associated with a range of medical pathologies (Eltzschig and Eckle, 2011). Specifically, during the ischemic phase, the oxidation of reduced equivalents (NADH, FADH_2_) by the mitochondrial electron transport system (ETS) is slowed down owing to O_2_ limitation and causes an immediate generation of fumarate from malate and subsequent reversal of complex II (succinate dehydrogenase, SDH) activity, which causes accumulation of the Kreb's cycle intermediate succinate (Chouchani et al., 2014; Prag et al., 2019). When the tissue is re-perfused (and thereby reoxygenated), ETS activity resumes and the accumulated succinate is rapidly oxidized by complex II. However, this rapid oxidation leads to reverse electron transport (RET) from the reduced ubiquinol pool into the complex I (NADH dehydrogenase) ROS generation site, causing a release of the ROS superoxide that is converted to hydrogen peroxide (H_2_O_2_) (Chouchani et al., 2014; Yin et al., 2021). This stimulation of ROS generation can overwhelm the existing ROS scavenging capacity, such that ROS react with a range of biomolecules (e.g. proteins, DNA and lipids) to induce oxidative damage and potentially activating cell death pathways (Bergamini et al., 2004; Li and Jackson, 2002).
Given the unique ability of crucian carp to endure natural cycles of anoxia–reoxygenation, studying their adaptive strategies may provide us with insights into mechanisms for alleviating or avoiding the ischemia–reperfusion injury described above. The same driver for ROS production during ischemia–reperfusion injury (i.e. accumulated succinate) has been detected in multiple tissues in crucian carp during anoxia (Dahl et al., 2021; Scott et al., 2024), although the succinate levels accumulated in anoxia vary significantly between tissues, with mammalian-like levels in the liver and low levels in the heart (Dahl et al., 2021; Lefevre and Nilsson, 2023). It is possible that suppressing succinate-driven mitochondrial ROS generation prevents oxidative damage in sensitive tissues, such as the heart or brain (Andrienko et al., 2017; Chouchani et al., 2014; Yin et al., 2021). Freshwater turtles exposed to anoxia–reoxygenation did not show an increase in lipid peroxidation products except for an increase in thiobarbituric acid reactive substance (TBARS) levels in the kidney (Willmore and Storey, 1997), indicating that the extent of oxidative damage in this anoxia-tolerant species is likely limited. This may relate to a much lower accumulation of succinate in turtles compared with mammals (mice) and crucian carp (Bundgaard et al., 2019; Dahl et al., 2021; Lefevre and Nilsson, 2023). Additionally, previous studies have shown that heart mitochondria from normoxic crucian carp have lower H_2_O_2_ flux when compared with other ectotherms, including anoxia-tolerant turtles (Bundgaard et al., 2019), but also anoxia-sensitive mice (Gerber et al., 2024). It can be hypothesised that the crucian carp generally have low ROS generation rates which helps to prevent high ROS generation rates when they do encounter anoxia–reoxygenation events. Nonetheless, signs of brain cell death have been observed in the crucian carp after 7 days of anoxia with 1 day of reoxygenation, which was followed by increased mRNA abundance of proliferating cell nuclear antigen, a marker for neurogenesis (Lefevre et al., 2017). This observation suggests that anoxic crucian carp are unable to completely avoid damaging effects of reoxygenation and that repair processes are necessary. Hence, the question remains as to what degree crucian carp experience and/or can protect themselves from ROS-driven oxidative damage following anoxia–reoxygenation exposure and what role mitochondria have in this process.
Using a combination of in vitro and in vivo approaches, the present study aimed to investigate whether crucian carp show differences in mitochondrial function that are consistent with mitochondria being more efficient under O_2_-limiting conditions as well as being able to limit oxidative damage. First, we compared the function of isolated mitochondria between crucian carp and the closely related common carp (Cyprinus carpio), which shows some tolerance to hypoxia but not extended anoxia and is unable to convert lactate to ethanol as an anaerobic end product (Van Ginneken et al., 1996). This interspecies comparison was chosen to investigate the hypothesis that crucian carp inherently have more O_2_-efficient mitochondria, where we predicted that crucian carp mitochondria would show a higher mitochondrial phosphorylation efficiency and activity of mitochondrial complexes, higher oxidative scavenging capacity and lower H_2_O_2_ generation, when compared with common carp mitochondria. Although these experiments were performed under normoxic conditions and primarily reflect intrinsic, baseline differences between species, they provide a foundation for understanding how crucian carp mitochondria may be inherently poised to tolerate low-oxygen conditions. Second, we investigated whether there was mitochondrial plasticity in crucian carp exposed to 5 days of anoxia or 5 days of anoxia followed by 1 day of reoxygenation. Although compensatory mechanisms during the 24 h recovery period may reverse transient functional changes (Bezawork-Geleta et al., 2017), we predicted that crucian carp would show a persistent reduction in complex II activity (i.e. enzymatic activity and mitochondrial respiratory capacity or flux) as an adaptative response to anoxia–reoxygenation exposure, as suggested by previous studies in other hypoxia-tolerant species (Devaux et al., 2019; Lau et al., 2020; Munro et al., 2025). Finally, we investigated whether crucian carp were able to minimize mitochondrial H_2_O_2_ generation and hence tissue accumulation and expected that there would be limited oxidative tissue damage after whole-animal anoxia–reoxygenation exposure.
MATERIALS AND METHODS
All chemicals, unless otherwise specified, were purchased from Sigma-Aldrich.
Animal care
Crucian carp [Carassius carassius (Linnaeus 1758)] were collected in September–October from Tjernsrudtjernet, a pond near Oslo (N59.92182, E10.60947; Norway), using baited nylon net cages. Common carp [bred ornamental koi (Cyprinus carpio Linnaeus 1758)] were purchased from a local supplier (Tropex). Fish were housed at the in vivo aquatic facilities at the Department of Biosciences, University of Oslo, in semi-closed recirculating tanks supplied with dechlorinated tap water. Fish were maintained at 12–14°C with a 12 h light:12 h dark photoperiod and fed carp pellets daily (Tetra Pond, Melle, Germany). Fish were acclimated to lab conditions for a minimum of 2 weeks and were fasted 24 h immediately prior to anoxia–reoxygenation exposure. The holding temperature was lowered over a couple of days to 8°C, 1 week prior to anoxia exposure. All experimental protocols were approved by the Norwegian Food Safety Authority (FOTS permit ID 16063), in accordance with the animal welfare law of Norway (‘Dyrevelferdsloven’) and instructions on use of animals for research (‘Forskrift om bruk av dyr i forsøk’).
Preparation of isolated mitochondria
The mitochondria isolation procedure described here was used for normoxic crucian carp, normoxic common carp and crucian carp exposed to anoxia–reoxygenation (exposure A). All glassware was chilled prior to starting the isolation procedure. Fish were euthanized with a sharp blow to the head and spinal severance, after which the brain and heart were quickly dissected out and submerged in ice-cold isolation buffer (250 mmol l^−1^ sucrose, 5 mmol l^−1^ Tris, 1 mmol l^−1^ EGTA, pH 7.4; heart tissue with an additional 0.1% bovine serum albumin). In a Petri dish with ice-cold isolation buffer, the tissues were diced with a razor blade into smaller chunks and rinsed with fresh isolation buffer. The diced heart tissue was added to a 2 ml Eppendorf tube with metal beads and quickly homogenized for 5 s on 25 Hz using the Qiagen TissueLyser before transferring it to a Dounce homogenizer. Diced brain tissue was homogenized in a Dounce homogenizer after rinsing with isolation buffer. After homogenization, the tissue was allowed to settle and rinsed with fresh isolation buffer. The tissue suspensions were then centrifuged at 2420 rpm (1100 g) for 5 min, after which the supernatant was transferred to a new centrifuge tube and subjected to two high-speed spins at 9150 rpm (16,000 g) for 10 min each to pellet mitochondria. The final pellet was resuspended in MiR05 buffer (0.5 mmol l^−1^ EGTA, 3 mmol l^−1^ MgCl_2_·6H₂O, 60 mmol l^−1^ K-lactobionic acid, 20 mmol l^−1^ taurine, 10 mmol l^−1^ KH_2_PO_4_, 20 mmol l^−1^ HEPES, 110 mmol l^−1^ sucrose and 1 g l^−1^ bovine serum albumin, pH 7.1) and protein concentration was determined via the Bradford method (Bradford, 1976).
High-resolution respirometry protocols and calculations
The Oroboros oxygraph high-resolution respirometer and fluorometer was used to simultaneously measure mitochondrial respiration and H_2_O_2_ emission. The O_2_ sensors were calibrated daily at the running temperature (12°C) with air-saturated MiR05 buffer for normoxia and with yeast for anoxia. Emission of H_2_O_2_ was assessed fluorometrically using Amplex Ultrared (Invitrogen) with the Oroboros fluorometer (excitation LED at 525 nm with filter set optimized for Amplex Ultrared as per manufacturer's instructions), set at 1000 gain and polarization voltage of 750 mV.
Two protocols were used. Both protocols were set up the same way, where an isolated mitochondria sample was first injected into the Oroboros chamber (0.08 mg protein for heart and 0.4 mg protein for brain). The following were then added in order into the 2 ml chamber for simultaneous measurement of H_2_O_2_ emission: 10 µmol l^−1^ of Amplex Ultrared, 2 U of horseradish peroxidase and 10 U of superoxide dismutase. For calibration of the fluorescence signal, H_2_O_2_ was titrated into the chamber to a final volume of 0.2 µmol l^−1^.
For protocol 1, NADH-generating substrates (5 mmol l^−1^ pyruvate, 1 mmol l^−1^ malate and 10 mmol l^−1^ glutamate; PMG) were added to stimulate complex I-fuelled state 2 respiration, followed by addition of 31 µmol l^−1^ ADP for assessment of ADP/O ratio with NADH generating substrates. Linear extrapolations of O_2_ concentrations during state 3 respiration (in the presence of ADP) and state 4 respiration (after ADP depletion) were used to calculate the consumed O_2_. The total O_2_ uptake was taken as the difference in O_2_ concentrations at these intercepts and then used to divide 31 µmol l^−1^ ADP to give the ADP/O ratio (nmol ADP/nmol O) (Chance and Williams, 1955). This step was followed by the addition of a saturating level of ADP (1 mmol l^−1^) to stimulate complex I-fuelled state 3 respiration. Then, 10 mmol l^−1^ succinate was added to stimulate complex II-fuelled state 3 respiration rate. Thereafter, 2.5 µmol l^−1^ oligomycin was added to enter a state 4 leak respiration and carbonyl cyanide 4-(trifluoromethyoxy) phenylhydrazone (FCCP) was titrated in 0.5 µmol l^−1^ steps until the mitochondria were fully uncoupled to obtain ETS capacity/maximal respiration rate. After the membrane was uncoupled, 0.5 µmol l^−1^ rotenone was added to inhibit complex I to obtain rotenone-inhibited respiration rate, followed by 5 mmol l^−1^ malonic acid to inhibit complex II to obtain malonic acid-inhibited respiration rate. Finally, 2.5 mmol l^−1^ antimycin A was added to inhibit complex III and obtain residual oxygen consumption rate. Using protocol 1, complex II flux and capacity were assessed as follows.
Complex II flux was calculated as the difference between complex I- and complex I+II-fuelled state 3 respiration rates in coupled mitochondria, normalized to FCCP respiration rates:
where PMG+D respiration represents complex I-fuelled state 3 respiration in coupled mitochondria, PMGS+D respiration represents complex I+II-fuelled state 3 respiration in coupled mitochondria and FCCP respiration represents ETS maximal capacity.
Complex II capacity was calculated as the difference between rotenone (CI)-inhibited and malonic acid (CII)-inhibited respiration rates in uncoupled mitochondria, normalized to FCCP respiration rates:
where rotenone respiration represents rotenone (CI)-inhibited respiration in uncoupled mitochondria, malonate respiration represents malonate (CII)-inhibited respiration in uncoupled mitochondria and FCCP respiration represents ETS maximal capacity.
For protocol 2, only the complex II substrate succinate (10 mmol l^−1^) was added, to stimulate complex II-fuelled state 2 respiration rate, followed by 1 mmol l^−1^ ADP to stimulate complex II-fuelled state 3 respiration rate. Thereafter, 2.5 µmol l^−1^ oligomycin was added to enter state 4 leak respiration, 0.5 µmol l^−1^ rotenone was added to inhibit complex I and FCCP was then titrated in 0.5 µmol l^−1^ steps until the mitochondria were fully uncoupled to obtain FCCP-uncoupled respiration rate. Finally, 2.5 mmol l^−1^ antimycin A was added to inhibit complex III and obtain residual oxygen consumption rate.
Anoxia–reoxygenation exposure and tissue sampling
General protocol for anoxia exposure
Crucian carp (39.44±1.86 g) of both sexes were placed in 25 l buckets (up to 12 fish in each bucket) which were filled with dechlorinated tap water and sealed with airtight lids. These exposure buckets were submerged in a water bath with connected chillers to maintain temperature at 8°C. The lids contained built-in tubing with an air stone. The tubing was connected either to an air pump or to N_2_ gas supply. Fish were left in the buckets with air bubbling for 24 h prior to the start of the experiment to recover from handling. Fire sting fibre-optic O_2_ probes (Pyroscience GmBH, Aachen, Germany) were inserted into a second opening in the bucket lid to monitor O_2_ levels inside the bucket for the duration of the experiment. To initiate the anoxic exposure, N_2_ gas was bubbled into the bucket and anoxia (<0.1% air saturation; the detection limit) was then maintained for 5 days. To reoxygenate the bucket after 5 days of anoxia, N_2_ gas supply was substituted for aeration and the O_2_ levels in the bucket returned to normoxic levels within 1–2 h. Fish were netted out and euthanized via concussion followed by spinal severance. Several exposures were performed for measurements of mitochondrial function, in vivo H_2_O_2_ levels and oxidative damage in anoxia–reoxygenation (see below).
Exposure A for measurement of mitochondrial function in anoxia–reoxygenation
Multiple anoxia–reoxygenation exposures were performed, with only a few fish exposed at a time, over several weeks (for a total of N=18 fish; n=6 per treatment). Fish were sampled after (1) 5 days of normoxia (N), (2) 5 days of anoxia without reoxygenation (A) and (3) 5 days of anoxia with 1 day of reoxygenation (AR1d). Sampled brain and heart were used for measurements of mitochondrial function and enzymatic activities of complexes in anoxia–reoxygenation.
Exposure B for detection of mitochondrial H2O2 levels in vivo in anoxia–reoxygenation following MitoB probe injection
MitoB is a mitochondria-targeted ratiometric probe used to detect matrix H_2_O_2_ levels in vivo (Logan et al., 2014). A pilot study was conducted prior to the main experiment to optimize MitoB use for this study, by determining the resident time of the MitoB probe in tissues of crucian carp (showing concentration of MitoB and MitoP in tissues after 12 h; Fig. S1A). We determined that animals had to be injected intraperitoneally with MitoB 3 h before the end of the anoxia exposure, such that there would still be a detectable amount of MitoB present in the tissue at the 6 h reoxygenation time point. For exposure B, a smaller 10 l bucket was placed inside the same 25 l buckets used in exposures A and C. This smaller inner bucket had two large mesh windows on the side such that anoxic fish could be removed from the outer exposure bucket while still being submerged in 4 l anoxic water in the smaller bucket. This set-up minimized disturbance to the fish and prevented air exposure of the fish prior to anaesthesia for injection of the MitoB probe. Six fish designated either for normoxic control (N) or for anoxia–reoxygenation (AR) exposure were placed in the smaller bucket inside the exposure bucket. Three hours before the end of the 5 days of anoxia exposure, the exposure bucket was opened and the inner bucket with fish was gently removed. Submerged anoxic fish were lightly anaesthetized with 50 mg l^−1^ benzocaine, briefly removed one by one for injection of MitoB and immediately returned into the larger exposure bucket. The exposure bucket was then resealed for the rest of the exposure (any lost water volume was replaced with anoxic water). Two fish were removed from each of the two control (5 days of normoxia, N) and two AR (5 days of anoxia, A; 5 days of anoxia with 3 h reoxygenation, AR3h; 5 days of anoxia with 6 h reoxygenation, AR6h) set-ups at three time points and sampled. This exposure was performed twice (for a total of N=36 fish; n= 6 per treatment and time-point combination). Sampled brain, heart, liver and gill tissues were flash frozen in liquid N_2_ and stored at −80°C until sample preparation for mass spectrometry.
Exposure C for measurements of oxidative damage in anoxia–reoxygenation
The same sampling timeline (without removing fish to inject MitoB) as described in exposure B (see above) was used for characterization of oxidative damage. Fish (total N=48; n=12 per treatment) were sampled after (1) 5 days of normoxia (N), (2) 5 days of anoxia (A), (3) 5 days of anoxia with 3 h reoxygenation (AR3h) and (4) 5 days of anoxia with 6 h reoxygenation (AR6h). Brain, heart, liver and gill tissues were sampled, snap frozen in liquid nitrogen and stored at −80°C until analysis for markers of protein and mitochondrial DNA (mtDNA) damage from individual samples (n=3–4 per treatment for protein damage and n=6 per treatment for mtDNA damage).
Total oxidative scavenging capacity
The methodology for assessing the total oxidant scavenging capacity (TOSC) of mitochondrial pellets was modified from the methodology for measuring TOSC of tissue homogenate (Lau et al., 2019). Frozen isolated mitochondrial pellets from the brain and heart of both normoxic crucian and common carp were freeze-thawed three times. Fifty µl of assay buffer (1 U ml^−1^ horseradish peroxidase, 0.1 mmol l^−1^ Amplex Ultrared and 50 mmol l^−1^ sodium citrate, pH 6.0) were added to 50 µl diluted sample or catalase standards (0 to 9 U) into the spectrophotometer plate well and the absorbance was pre-read at 570 nm. Fifty µl H_2_O_2_ (diluted in cold sodium citrate) was then added to each well and the end-point absorbance was measured after 10 min. The inverse of the pre- and post-absorbance was plotted against the catalase activity units to form the standard curve, where activities were in the linear range between 0.07 to 1.125 U of catalase activity. The final TOSC activity was then expressed as µmol H_2_O_2_ min^−1^ (which is equivalent to 1 U of catalase activity) and normalized to the protein concentration of the mitochondrial pellet sample assayed via the Bradfords assay (Bradford, 1976).
Enzymatic activity assays of mitochondrial complexes
Maximal enzyme activities of mitochondrial complexes I to V were measured on whole tissue homogenates from heart and brain of normoxic crucian carp and common carp, as well as from crucian carp exposed to anoxia-reoxygenation (exposure A). Whole tissues were homogenized in 500 µl of 50 mmol l^−1^ potassium phosphate buffer (KPi; pH 7.8) using the Qiagen Tissuelyser for 2 min at 25 Hz. The homogenate was then centrifuged at 17,000 g for 10 min at 4°C. The supernatant was then removed and the pellet resuspended in 100 µl KPi and kept frozen at −80°C until the day of assay. The method for assaying enzymatic activity of the complexes is described by Galli et al. (2013). Briefly, complex I activity was detected as rotenone-sensitive reduction of 5,5′-dithiobis(2-nitroenzoic acid) or DCPIP monitored at 600 nm; complex II activity was detected as the reduction of DCPIP in the presence of rotenone and antimycin A monitored at 600 nm; complex III activity was monitored in the presence of rotenone as the reduction of cytochrome c with electron donor ubiquinol monitored at 550 nm; complex IV activity was measured as the oxidation of reduced cytochrome c monitored at 550 nm; and complex V activity was measured as oligomycin-sensitive oxidation of NADH at 340 nm with pyruvate kinase and lactate dehydrogenase present. All enzyme activities were assayed at 25°C. Protein concentration was determined according to Bradford (1976). Complex enzyme activity (in nmol min^−1^ mg^−1^ protein) was normalized to citrate synthase (CS) activity (in nmol min^−1^ mg^−1^ protein).
Measurement of in vivo mitochondrial H2O2 generation via MitoP/MitoB ratio
For the analytical processing of tissue samples to obtain MitoP/MitoB ratios via LC-MS/MS, we followed the methods and protocols previously described (Lau et al., 2019) with some modifications. Approximately 20–50 mg of frozen tissue (brain, heart, gill and liver from crucian carp exposed to anoxia–reoxygenation; exposure B) was homogenized in 210 µl ice-cold 60% acetonitrile/0.1% formic acid spiked with internal standards (10 µmol l^−1^ d_15_-MitoB/10 µmol l^−1^ d_15_-MitoP) in the Qiagen Tissuelyzer II (2 min at 25 Hz). The homogenate was then centrifuged at 16,000 g for 10 min at 4°C, after which the supernatant was transferred to a new 1.5 ml microcentrifuge tube. The tissue was resuspended again in 200 µl 60% acetonitrile/0.1% formic acid, homogenized again, followed by centrifugation at 16,000 g for 10 min at 4°C. After combining the supernatant from the two centrifugation steps, the sample was vortexed for 10 s and incubated at 4°C for 30 min. After incubation, samples were centrifuged at 16,000 g for 10 min and the supernatant was filtered with the Millipore centrifugal filter plate (0.45 µm hydrophilic, low protein binding Durapore membrane; Merck) and centrifuged at 3000 g for 10 min. The filtrate was collected and subsequently dried in a Speed Vac concentrator (Concentrator Plus, Eppendorf; Hamburg, Germany). The dried sample was then resuspended in 250 µl of 20% acetonitrile/0.1% formic acid, vortexed for 5 min, then centrifuged at 16,000 g for 10 min. Two hundred µl of the suspension was analysed by LC-MS/MS using a TripleQuad mass spectrometer (Lau et al., 2019), at the MRC Mitochondrial Biology Unit (University of Cambridge, UK). Briefly, the mass spectrometer was connected in series to an I-class Aquity LC system (Waters). Samples (stored in an autosampler at 4°C) were injected into a flow-through needle and RP-UPLC at 40°C using an Acquity UPLC^®^ BEH C18 column (1×50 mm, 1.7 µm; Waters) with a Waters UPLC filter (0.2 µm). Mass spectrometer buffers A (95% water/5% acetonitrile/0.1% formic acid) and B (90% acetonitrile/10% water/0.1% formic acid) were infused at 200 µl min^−1^ using the following gradients: 0–0.3 min, 5% B; 0.3–3 min, 5–100% B; 3–4 min, 100% B, 4–4.10 min, 100–5% B; 4.10–4.60 min, 5% B. Eluants were diverted to waste at 0–1 min and 4–4.60 min and compounds were detected in multiple reactions monitoring in positive ion mode. For quantification, the following transitions were used: MitoB, 397>183; d15-MitoB, 412>191; MitoP, 369>183; d15-MitoP, 384>191. Standard curves for MitoB (0 to 10 pmol) and MitoP (0 to 10 pmol) were prepared using crucian carp liver tissue (Fig. S1B) and used to determine the amount of MitoB and MitoP in each sample. Deuterated internal standards were used to normalize samples to account for variation in sample volume. The peak area of MitoB, MitoP, internal standard of samples and standard curves were quantified using the MassLynx 4.1 software.
Quantification of oxidative damage via western blot analyses
Detection of carbonyl groups for protein damage
Carbonyl groups introduced into protein by oxidative reactions were detected and quantified using Abcam's Oxidized Protein Western Blot Detection Kit (ab178020). Preparation of extracts were performed following the manufacturer's protocol. Total protein content was determined using Pierce™ BCA Protein Assay Kit (23225, Thermo Fisher Scientific, Rockford, IL, USA). The resulting sample extracts were used for detection of both protein damage (detection of carbonyl groups) and lipid peroxidation (detection of 4-hydroxynonenal; 4-HNE).
For the carbonyl assay, two aliquots for each brain, heart, gill and liver samples (exposure C) were used. One was treated with DNPH Derivatization Solution and the other served as a negative control (treated with Derivatization Control Solution, see method verification in Fig. S2). DNP-derivatized protein samples from liver (10 µg protein/well), heart (2 µg protein/well), gill (5 µg protein/well) and brain (10 µg protein/well) were separated using NuPAGE Bis-Tris polyacrylamide gels (4 to 12% gradient) for 50 min at 200 V using NuPAGE MOPS SDS running buffer (Invitrogen) under reducing conditions (i.e. 50 mmol l^−1^ DTT added prior loading). Proteins were then transferred onto 0.45 µm LF-PVDF membrane (Amersham) using 2× NuPAGE transfer buffer at 20 V (0.5A) for 20 min with Bio-Rad Trans-Blot Turbo system for semi-dry transfer. Membranes were blocked for 1 h at room temperature in Intercept (TBS) Blocking Buffer (LI-COR). For the detection of DNP-derivatized proteins the blots were incubated overnight with anti-DNP (1:5000, rabbit, ab178020) diluted in Intercept T20 (TBS) Antibody Diluent (LI-COR). The membrane was then washed three times for 5 min each in TBS-T (Thermo Scientific Pierce 20× tris buffered saline, 1% Tween-20) before incubation with IRDye 800CW (1:10,000, goat anti-rabbit IgG) secondary antibody diluted in Intercept (TBS) Blocking Buffer for 1 h at room temperature. Membranes were then washed three times for 5 min in TBS-T. After a final wash with TBS, the membranes were air dried and imaged using the Odyssey CLx imager (LI-COR) at an excitation wavelength of 800 nm. Beta-actin was used as a loading control and for normalization of the relative abundance of carbonyl-proteins. Membranes were incubated separately with β-actin antibody (1:5000, mouse monoclonal, ab170325) as described above but only for 1 h at room temperature and detected with IRDye 680RD (1:10,000, Goat anti-mouse IgG) secondary antibody (with excitation wavelength of 700 nm). Quantification of the carbonyl groups was performed using Image Studio Lite (version 5.2.5). See blots used for quantification in Fig. S2.
Detection of 4-hydroxynonenal for lipid peroxidation
4-hydroxynonenal (4-HNE) modifications of protein were assessed in crucian carp liver (50 µg protein/well), heart (20 µg protein/well), brain (50 µg protein/well) and gills (25 µg protein/well) exposed to anoxia–reoxygenation (exposure C) via western blotting as described above using an anti-4 HNE antibody (anti-rabbit, Abcam 46545, 1:2500). This antibody only detected 4-HNE on reduced samples (treated with 100 mmol l^−1^ DTT) and had a detection limit of 1 µg of 4-hydroxynonenal modified bovine serum albumin (4-HNE BSA, ab194193). See blots used for quantification and method verification in Fig. S3.
Quantification of mitochondrial DNA damage via qPCR
The quantification of mitochondrial DNA (mtDNA) damage was performed according to a previously published method and is based on the ability of any modification in the DNA to inhibit TaqI restriction digestion (Wang et al., 2016). First, DNA was extracted from whole brain, heart, gill and liver tissues from crucian carp exposed to anoxia–reoxygenation (exposure C) with the Qiagen DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer's protocol. The isolated DNA concentration was then quantified using a Nanodrop spectrophotometer (Thermo Fisher Scientific) and subsequently diluted to a concentration of 6 ng µl^−1^ per aliquot. PCR primers were designed to target the T^CGA TaqI restriction enzyme sites of the crucian carp mitochondrial genome (Genbank JQ911695; Cheng et al., 2012). Forward (5′-TTACATAGTGCCCCCTTTGG-3′) and reverse (5′-GGGTTTTTGGCAGAGGTTTT-3′) primers at final concentration of 0.5 µmol l^−1^ were mixed with a total of 1.8 ng µl^−1^ of mtDNA and 5 µl Power SYBR Green PCR Master mix (Thermo Fisher Scientific) according to their standard protocol. For each sample, two sets of triplicates were analysed, in which one set was treated with 0.025 U of TaqI DNase whereas the other remained untreated. The final volume per well was 10 µl and was analysed on a qPCR 7900HT Fast Real-Time PCR system SDS2.3 (Thermo Fisher Scientific) using the following programme: 1 cycle at 65°C for 15 min, 1 cycle of 94°C for 10 min, 40 cycles of 94°C for 10 s followed by 60°C for 60 s. The cycle number (Ct) for each of the samples were derived from absolute quantification analysis and the difference in cycle number was calculated between the treated and untreated triplicates for each sample (ΔCt). The DNA damage frequency was calculated as 2^−ΔCT^ and normalized to normoxic control values for visualisation.
Statistical analysis
Differences in mitochondrial O_2_ consumption and H_2_O_2_ emission data between species and mitochondrial substrates were analysed with two-way ANOVAs, followed by Šidák's multiple comparison tests. Differences in TOSC and mitochondrial enzymatic activities between species were analysed with unpaired t-tests. For the MitoP/MitoB dataset, as each treatment sampling time point was compared with a normoxic control time point, two-way ANOVAs were used to assess the effect of anoxia–reoxygenation exposure (normoxic control versus anoxia–reoxygenation exposure within each time point) and sampling group (anoxia versus AR3h versus AR6h within both the treated group and the control group). For the oxidative damage, mitochondrial complex activity and the complex II datasets from anoxia–reoxygenated crucian carp, one-way ANOVAs followed by Tukey's multiple comparison tests were used to assess the effect of anoxia–reoxygenation exposure (normoxic control versus anoxia–reoxygenation exposure). The output statistics and P-values from all analyses are provided in Tables S1–S3 and S5–S9.
RESULTS
Comparison of isolated mitochondrial function between crucian carp and common carp
We first compared crucian carp and common carp isolated brain and heart mitochondria fuelled by complex I (CI) substrates (pyruvate, malate, glutamate) or complex II (CII) substrate (succinate) in state 2 (substrate only) and 3 (substrate with ADP) (for two-way ANOVAs results see Table S1). Complex I-fuelled state 2 respiration was lower in crucian carp brain, but not heart, compared with common carp, whereas complex I-fuelled state 3 respiration was lower in both brain and heart mitochondria from crucian carp (Fig. 1A,B). For brain mitochondria (Fig. 1A) in state 2, CII-fuelled respiration was higher than CI-fuelled respiration in crucian carp, but not in common carp. Conversely, in state 3, CII-fuelled respiration was lower than CI-fuelled respiration in common carp, but not in crucian carp. For heart mitochondria (Fig. 1B) in state 2, CII-fuelled respiration was higher than CI-fuelled respiration in both species.
Mitochondrial function in isolated mitochondria of normoxic crucian carp and common carp. (A–D) Brain (A,C) and heart (B,D) complex I [CI, protocol 1 with the NADH-generating substrates, pyruvate (5 mmol l−1), malate (1 mmol l−1) and glutamate (10 mmol l−1)] or II [CII, protocol 2, with only the complex II substrate succinate (10 mmol l−1)] fuelled mitochondrial respiration rates (A,B) and H2O2 emission rates (C,D) in state 2 (substrate only, no ADP) and state 3 (substrate with ADP) of normoxic crucian carp and common carp. Statistical differences were assessed with two-way ANOVAs followed by Šidák's multiple comparisons, with species differences within a substrate/state combination indicated by hashtags (#) and differences between substrates within a species/state combination indicated by letters (dissimilar lowercase letters for crucian carp and dissimilar uppercase letters for common carp). See results from statistical analyses in Table S1. Data are presented as means±s.e.m. with individual points. Sample size n=3–6 per group. The corresponding free radical leak (defined as H2O2 emission normalized to respiration) is shown in Fig. S4.
Crucian carp brain and heart mitochondria in state 2 had lower H_2_O_2_ emission compared with common carp, regardless of the substrate (Fig. 1C,D). Importantly, addition of the reoxygenation-relevant CII substrate succinate led to higher H_2_O_2_ emission in both brain and heart of common carp, but not in crucian carp. For brain mitochondria in state 3, there was no difference in H_2_O_2_ emission between species, regardless of the substrate (Fig. 1C). Conversely, for heart mitochondria in state 3, H_2_O_2_ emission remained higher in common carp compared with crucian carp when fuelled with CI, but not CII substrates, as the CII substrates led to lower H_2_O_2_ emission in common carp compared with the CI substrates (Fig. 1D). The corresponding free radical leak (defined as H₂O₂ emission normalized to respiration) is shown in Fig. S4.
In addition to differences in H_2_O_2_ emission rates, phosphorylation efficiency (ADP/O) with NADH-generating substrates was higher in brain mitochondria of crucian carp compared with common carp, whereas there were no differences in heart mitochondria (Fig. 2A). Crucian carp also showed higher TOSC in both brain and heart mitochondrial pellets when compared with common carp (Fig. 2B).
*Mitochondrial phosphorylation efficiency and oxidative scavenging capacity of normoxic crucian carp and common carp. (A,B) Mitochondrial phosphorylation efficiency (ADP/O ratio) with NADH-generating substrates assessed in isolated mitochondria (protocol 1) (A) and total oxidant scavenging capacity (TOSC) assessed in mitochondrial pellet (B) from brain and heart of normoxic crucian carp and common carp. Data are compared with unpaired t-tests with species differences indicated by asterisks (*P<0.05; *P<0.01). See results from statistical analyses in Table S2. Data are presented as means±s.e.m., with individual points. Sample size n=3–7 per group.
When comparing enzyme activities of mitochondrial complexes, crucian carp complex capacity levels from both brain and heart homogenate were all higher than those from common carp (Fig. 3), except for CI in brain.
*Maximal enzymatic activity of mitochondrial complexes of normoxic crucian carp and common carp. (A–J) Relative enzymatic activity of mitochondrial complexes I–V in brain (A–E) and heart (F–J) tissue homogenates of normoxic crucian carp and common carp. The activity of complexes (in nmol min−1 mg−1 protein) were normalized to citrates synthase (CS) activities (in nmol min−1 mg−1 protein). Statistical differences were assessed using unpaired t-tests with differences indicated by asterisks (*P<0.05; **P<0.01; **P<0.001). See results from statistical analyses in Table S3. Data are presented as means±s.e.m. with individual data points. Sample size n=6 per group.
Response of crucian carp to anoxia-reoxygenation exposure
Mitochondrial respiration and complex activities
We investigated changes to both complex II respiratory flux (in coupled mitochondria with excess ADP, state 3) and capacity (in uncoupled mitochondria with the uncoupler FCCP and complex I inhibitor rotenone) by high resolution respirometry (Fig. 4). Brain mitochondria isolated from crucian carp exposed to 5 days of normoxia, 5 days of anoxia and 5 days of anoxia with 1 day of reoxygenation showed no differences in complex II flux (Fig. 4A). A significant decrease in complex II capacity was found after 5 days of anoxia, which was partially recovered after 1 day of reoxygenation (Fig. 4B). In contrast, heart mitochondria from crucian carp exposed to 5 days of anoxia with 1 day of reoxygenation showed significantly decreased complex II flux compared with fish under normoxia and anoxia (Fig. 4C), but no differences in complex II capacity between the treatment groups (Fig. 4D).
Changes in mitochondrial complex II flux and capacity in crucian carp exposed to anoxia-reoxygenation. (A–D) Mitochondrial complex II flux (A,C) and capacity (B,D) expressed as % of ETS capacity in brain (A,B) and heart (C,D) mitochondria of crucian carp exposed to 5 days of normoxia (N), 5 days of anoxia (A) and 5 days of anoxia with 1 day of reoxygenation (AR1d). Fish were from exposure A. Flux was assessed using protocol 1 as the difference between complex I- and complex I+II-fuelled state 3 respiration in coupled mitochondria normalized to FCCP respiration (see Eqn 1). Capacity was assessed using protocol 1 as the difference between rotenone (CI)-inhibited and malonic acid (CII)-inhibited respiration in uncoupled mitochondria normalized to FCCP respiration (see Eqn 2). Respiration values used for calculations are provided in Table S4. Differences were assessed by one-way ANOVAs followed by Tukey's multiple comparison tests with differences indicated by letters. See results from statistical analyses in Table S5. Data are presented as means±s.e.m. with individual data points. Sample size n=4–6 per group.
To investigate the underlying cause of differences in flux versus capacity (Fig. 4), the maximal enzymatic activity of mitochondrial complexes was measured in heart and brain tissue homogenate (Fig. 5). No differences in brain enzymatic activity for the five mitochondrial complexes were found between treatment groups (Fig. 5A–E). In contrast, heart complex II and V showed a significant decrease in enzymatic activity after 5 days of anoxia with 1 day reoxygenation compared with 5 days of anoxia (Fig. 5G,J).
Maximal enzymatic activity of mitochondrial complexes in crucian carp exposed to anoxia–reoxygenation. (A–J) Relative enzymatic activity of mitochondrial complexes I–V in brain (A–E) and heart (F–J) tissue homogenate of crucian carp exposed to 5 days of normoxia (N), 5 days of anoxia (A) and 5 days of anoxia with 1 day of reoxygenation (AR1d). Fish were from exposure A. Complex activity was normalized to citrate synthase (CS) activity. Differences were assessed by one-way ANOVAs followed by Tukey's multiple comparison tests with differences indicated by letters. See results from statistical analyses in Table S6. Data are presented as means±s.e.m. with individual data points. Sample size n=4–6 per group.
In vivo mitochondrial H2O2 generation in tissues
The in vivo generation of mitochondrial H_2_O_2_ was quantified by assessing changes in the ratio of MitoP to MitoB (MitoP/MitoB) in tissues over the course of anoxia and reoxygenation. In liver (Fig. 6D), but not brain, heart and gill tissues (Fig. 6A–C), there was a significant effect of the anoxia treatment (i.e. relative to normoxic controls; Table S7). There were no differences between sampling time points in either the normoxic controls or the anoxia-exposed treatment groups.
In vivo mitochondrial H2O2 generation in crucian carp exposed to anoxia-reoxygenation. (A–D) Tissue MitoP/MitoB in brain (A), heart (B), gill (C) and liver (D) of crucian carp exposed to normoxia throughout the experiment as a time control (N), 5 days of anoxia (A), 5 days of anoxia with 3 h of reoxygenation (AR3 h) and 5 days of anoxia with 6 h of reoxygenation (AR6h). Fish were from exposure B. Differences were assessed by two-way ANOVAs followed by Šidák's multiple comparisons with significant treatment effect indicated with P-value. See results from statistical analyses in Table S7. Data are presented as means±s.e.m. with individual data points. Sample size n=5–6 per group.
Oxidative damage in tissues
We quantified markers of protein oxidation, DNA damage and lipid peroxidation over the course of anoxia and reoxygenation. In the brain, heart and gill tissues, there was a significant effect of experimental treatment on levels of protein carbonylation, a marker of protein damage (Fig. 7). Protein carbonylation in the brain reached maximal values after 3 h of reoxygenation (AR3h) (Fig. 7A). In the gills, however, the levels of protein carbonylation continued to increase up to 6 h of reoxygenation (AR6h), with intermediate oxidation at anoxia and 3 h reoxygenation (Fig. 7C).
Protein oxidative damage in crucian carp exposed to anoxia-reoxygenation. (A–D) Relative abundance of protein carbonylation (DNP-derivatized protein) normalized to β-actin in brain (A), heart (B), gill (C) and liver (D) tissue of crucian carp exposed to 5 days of normoxia (N), 5 days of anoxia (A), 5 days of anoxia with 3 h of reoxygenation (AR3h) and 5 days of anoxia with 6 h of reoxygenation (AR6h). Fish were from exposure C. One-way ANOVAs followed by Tukey's multiple comparison tests (differences indicated by letters). See results from statistical analyses in Table S8 and blots used for quantification in Fig. S2. Data are presented as means±s.e.m. with individual data points. Sample size n=3–4 per group.
In the brain and gills, there were no differences in the level of mtDNA damage between treatment groups (Fig. 8A,C). In the heart, however, there was a significantly higher level of mtDNA damage in both the anoxic (A) and 6 h reoxygenated (AR6h) treatment groups compared with the normoxic controls, whereas the levels at 3 h reoxygenation (AR3h) were intermediate (Fig. 8B). In the liver, mtDNA damage was significantly higher in the anoxic treatment group compared with the normoxic controls, whereas the AR3h and AR6h treatment groups had intermediate values (Fig. 8D). As an attempt to investigate potential lipid oxidative damage, we quantified levels of the lipid peroxidation product 4-hydroxynonenal (4-HNE), but there were no detectable levels of 4-HNE in any of the tissues analysed (Fig. S3).
Mitochondrial DNA (mtDNA) damage in crucian carp exposed to anoxia–reoxygenation. (A–D) Relative change in mtDNA (2−ΔCT) normalized to normoxic control values in brain (A), heart (B), gill (C) and liver (D) tissues of crucian carp exposed to 5 days of normoxia (N), 5 days of anoxia (A), 5 days of anoxia with 3 h of reoxygenation (AR3h) and 5 days of anoxia with 6 h of reoxygenation (AR6h). Fish were from exposure C. One-way ANOVAs followed by Tukey's multiple comparison tests (differences indicated by letters). See results from statistical analyses in Table S9. Original (non-normalized) data are shown in Table S10. Data are presented as means±s.e.m. with individual data points. Sample size n=6 per group.
DISCUSSION
In this study, we investigated the role of mitochondria, focusing on brain and heart, in the ability of crucian carp to tolerate extreme fluctuations in oxygen availability and in avoiding associated reoxygenation-mediated damage. Crucian carp mitochondrial respiration and H_2_O_2_ generation were compared with that of common carp mitochondria, revealing that the crucian carp mitochondria produce less H_2_O_2_, have a higher total oxidant scavenging capacity and a relatively high maximal activity and coupling efficiency (ADP/O) of biochemical complexes. Crucian carp mitochondria were also remarkably resistant to generate H_2_O_2_ in vivo in response to reoxygenation and therefore, the observed increase in reoxygenation-associated oxidation of protein and mtDNA damage in tissues is probably not caused by mitochondrial ROS generation.
Mitochondrial adaptations unique to crucian carp
We first compared mitochondrial function between crucian carp and the closely related common carp, to determine whether crucian carp has a unique strategy that is linked to its exceptional anoxia tolerance or whether it is a carp family trait (Figs 1–3). In a previous study, it was shown that heart mitochondria from crucian carp have significantly lower H_2_O_2_ emission rates under state 2 conditions with succinate (but without ADP) when compared with other ectotherms (Bundgaard et al., 2020b). Lower H_2_O_2_ release rates in crucian carp were also found under complex V (F_1_F_0_-ATPase) inhibition and linked to reduced complex II capacity when compared with an endotherm (Gerber et al., 2024). We thus predicted that differences in mitochondrial function would be related to complex II, owing to its role in RET during anoxia–reoxygenation. Our results indicate that both brain and heart mitochondria in crucian carp generally had lower respiration rates than in the common carp in oxidative phosphorylation state (state 3), regardless of the substrate (Fig. 1A,B). Additionally, brain mitochondria in crucian carp showed higher phosphorylation efficiency (ADP/O) with NADH-generating substrates when compared with that in common carp, wheras there was no species difference in heart mitochondria (Fig. 2A). Similarly, no difference in mitochondrial efficiency (ATP-linked O_2_ consumption with the CI substrates malate and pyruvate and CII substrate succinate) was observed in heart mitochondria from crucian carp compared with those from mice (Gerber et al., 2024). These tissue and species differences in phosphorylation efficiency suggest that crucian carp brain mitochondria may inherently convert energy more efficiently than brain mitochondria in the common carp. Interestingly, while species differences in mitochondrial respiration rates were observed in a phosphorylating state (state 3), species differences in mitochondrial H_2_O_2_ emission rates were mainly observed in the non-phosphorylating/leak state (state 2). Indeed, crucian carp heart and brain mitochondria consistently had lower rates of H_2_O_2_ emission in state 2, regardless of the substrates, when compared with common carp (Fig. 1C,D). As net ROS emission from isolated mitochondria is a result of both ROS generation and scavenging, we also assessed whether there was a difference in ROS scavenging capacity using the TOSC assay on isolated mitochondrial pellets. Crucian carp showed higher total oxidant scavenging capacity than common carp in both brain and heart mitochondria (Fig. 2B), suggesting that the overall lower H_2_O_2_ emission may be a result of higher ROS scavenging capacity in mitochondria, although a more mechanistic approach will be required to confirm a causal link. A lower mitochondrial H_2_O_2_ release rate (5- to 15-fold, depending on mitochondrial state) in heart mitochondria of crucian carp was also previously associated to higher total antioxidant capacity (twofold) in heart tissue homogenate when compared with mice (Gerber et al., 2024). It could also be due to lower gross ROS production from the ETS as supported by the higher capacities at most ETS complexes in heart and brain mitochondria of crucian carp (Fig. 3). For future studies, it would be interesting to investigate the impact of the redox-dependent peroxide production in mitochondria of crucian and common carp found in this study (Fig. 1CD) on the spare respiratory capacity (see review by Marchetti et al., 2020) in primary neurons and cardiomyocytes provided with different substrates. Our comparison of the crucian carp with the common carp accentuates the crucian carp's unique strategy at the mitochondrial level to tolerate anoxia, as both species possess similar genomic backgrounds, sharing the same whole-genome duplication with goldfish and other cyprinids (Chen et al., 2019; Fagernes et al., 2017; Valencia-Pesqueira et al., 2025). That the common carp remains less anoxia tolerant than crucian carp is likely to be due to differences in selective pressure and lack of acquisition of the ‘right’ random mutations of duplicated genes.
Mitochondrial plasticity in anoxia–reoxygenation
We then investigated whether crucian carp exposed to anoxia–reoxygenation show plastic mitochondrial responses that would further help to alleviate reoxygenation stress (Figs 4 and 5), expecting that there would be a decrease in complex II activities (i.e. enzymatic activity and/or mitochondrial flux or capacity) in both brain and heart mitochondria in response to anoxia–reoxygenation exposure. Crucian carp exposed to anoxia–reoxygenation showed no differences in brain mitochondrial complex enzymatic activities (Fig. 5A–E) but after 1 day of reoxygenation, both complexes II and V in heart tissue showed decreases in activity (Fig. 5G,J). Other comparative studies in hypoxia-tolerant vertebrates support a link between reduced complex II function and hypoxia (–reoxygenation) tolerance: cardiac mitochondria from naked mole-rats exhibit lower complex II-fuelled mitochondrial respiration than those from mice (Lau et al., 2020); in epaulette sharks (Hemiscyllium ocellatum and Chiloscyllium punctatum), complex II activity and catalytic efficiency was significantly decreased during hypoxia-reoxygenation in the more hypoxia-tolerant H. ocellatum (Devaux et al., 2019); and across African mole-rat species, the downregulation of complex II-fuelled mitochondrial respiration correlates with hypoxia tolerance (Munro et al., 2025). In the respirometry data, we also observed that brain complex II capacity was reduced during anoxia (Fig. 4B), whereas heart complex II flux was reduced during reoxygenation (Fig. 4C), which could be a tissue-specific mechanism to counteract RET and associated ROS production caused by excessive succinate accumulation. Although we did observe plastic mitochondrial responses in crucian carp, it is possible that there were more transient changes in mitochondrial function not captured after 24 h of reoxygenation, as compensatory mechanisms could have occurred during this period and restored function (Bezawork-Geleta et al., 2017).
Low in vivo accumulation of H2O2 in response to anoxia–reoxygenation
To investigate whether in vivo mitochondrial H_2_O_2_ levels change in response to anoxia–reoxygenation, we used the ratiometric MitoB probe (Cochemé et al., 2012; Logan et al., 2014; Murphy et al., 2022). Comparing MitoP/MitoB in normoxic crucian carp tissue, the liver had the highest MitoP/MitoB ratio (average 0.39) and the gill tissue had the lowest MitoP/MitoB ratio (average 0.092) (Fig. 6C,D). This may indicate tissue variation in mitochondrial H_2_O_2_ generation in normoxic animals. There was a general decrease in MitoP/MitoB ratio with anoxia–reoxygenation exposure compared with normoxic controls at the same sampling point (although this was only significant in the liver; Fig. 6D). The decrease of H_2_O_2_ generated during anoxia was expected as O_2_ is required for ROS generation and thus mitochondrial H_2_O_2_ generation is predictably limited. Unexpectedly, at the two reoxygenation time points, there were no increases in mitochondrial H_2_O_2_ generation either (Fig. 6). There are several possible explanations for this observation. While there is succinate accumulation in anoxia (Dahl et al., 2021), striking tissue differences were observed: mammalian-like levels were found in the crucian carp liver (∼2600 µmol l^−1^), whereas levels in the heart were much lower (∼650 µmol l^−1^). That is just 1/5 of that measured in ischemic mice hearts (∼3500 µmol l^−1^) (Bundgaard et al., 2019). Even lower levels of accumulated succinate (∼100 µmol l^−1^) were observed in anoxic freshwater turtles (Bundgaard et al., 2019). Thus, it is possible that these levels of accumulated succinate are low enough to avoid RET-driven ROS generation in hearts of these anoxia-tolerant animals. Additionally, the other drivers for ROS generation during reoxygenation may be absent in anoxic crucian carp. As in anoxic turtles, the adenylate pool, and most importantly ADP, is preserved in anoxia-exposed crucian carp and the breakdown of adenine nucleotides to xanthine and hypoxanthine only occurred at low levels, compared with that observed in murine models (Bundgaard et al., 2019, 2023; Dahl et al., 2021). Thus, there is ADP available for complex V to use for ATP production while harvesting the H^+^ gradient that would otherwise become excessively high and lead to an accumulation of electrons that react with O_2_ to generate ROS. The redox status of NADH/NAD^+^ also appeared to be relatively stable over anoxia–reoxygenation in crucian carp tissues (Dahl et al., 2021).
Oxidative stress
Despite the low accumulation of ROS (as assessed by H_2_O_2_ levels in vitro and in vivo), our results indicate that crucian carp exhibit changes in some markers of oxidative stress, as we observed tissue-specific increases in protein carbonylation (Fig. 7) and mtDNA damage (Fig. 8) levels. This tissue specificity is consistent with what was observed when comparing the metabolome, proteome and phosphoproteome of various tissues in anoxia-exposed and reoxygenated crucian carp (Dahl et al., 2021; Johansen et al., 2023, 2024). It has been proposed that succinate efflux from ischaemic murine heart potentially aids in preventing large increases in ROS production upon reperfusion (Prag et al., 2021). In anoxic crucian carp, blood plasma succinate levels are higher than in brain and heart but lower than in liver, indicating transport from presumably sensitive tissues such as the brain and heart to the liver, which may have higher antioxidant capabilities (Dahl et al., 2021). There are several possible explanations for the presence of oxidative damage, but lack of detected mitochondrial H_2_O_2_ levels, observed in this study. It is possible that H_2_O_2_ is not the main ROS driver of oxidative damage in anoxia-reoxygenated crucian carp and that other types of ROS could be responsible for the oxidative damage observed. Both reactive nitrogen species (RNS) or reactive sulfur species (RSS) can act on similar targets as ROS (e.g. targeting cysteine thiols) and can be more volatile than ROS (Cortese-Krott et al., 2017; Olson, 2020). Whereas anoxic ROS generation should be inhibited owing to limitation in O_2_, it is possible that RSS increases during anoxia (Olson, 2020) and acts as a signalling molecule to affect downstream targets. In fact, both carbonylation levels (Koike and Ogasawara, 2016) and mtDNA damage have been observed to be modulated by levels of RNS or RSS (Ballinger et al., 2000; Shackelford et al., 2021). Whether or not anoxia-tolerant organisms such as crucian carp generate RSS and RNS during anoxia–reoxygenation exposure has not been investigated, but it would be worthwhile exploring this considering the observed anoxia-induced increases in the level of mtDNA damage in heart and liver and increases in protein carbonylation at re-oxygenation. Oxidative damage products can serve as signalling molecules to provide feedback to initiate the appropriate cellular response (Bundgaard et al., 2024; Holmström and Finkel, 2014; Sies, 2017) and there is some evidence that redox signalling is involved in the reoxygenation phase in crucian carp: there was an increase in brain proliferating cell nuclear antigen transcription levels after 7 days of reoxygenation from anoxia, which followed an increase in cell death observed after 1 day of reoxygenation (Lefevre et al., 2017). This potential stimulation of cell cycle regulation, which could indicate tissue recovery after oxidative damage, could potentially be under redox regulation (Sarsour et al., 2009). Whether there is an adaptive role of changes in oxidative damage markers and redox signalling that is pertinent to the response of anoxia-tolerant animals to environmental O_2_ variability requires further investigation. Here, we did not investigate the role of mtDNA repair per se, but the levels of mtDNA damage were decreasing towards normoxic levels by 3 h reoxygenation, especially in the liver. Studies from mouse hearts undergoing ischemia and subsequent reperfusion in an ex vivo (Langendorff) model demonstrated rapid induction of mtDNA damage upon reperfusion that was removed within 60 min. The repair protein OGG1 was found to be important for removal (Bliksøen et al., 2015) and this orthologue is present in crucian carp as well (NCBI gene ID 132097372 in genome GCF_963082965.1; ccar_ub18-g40225 in GCA_047456465.1; Valencia-Pesqueira et al., 2025) and thus might be responsible for rapid removal. Thus, the reduction in mtDNA damage in liver (from anoxic levels) might be caused by increased mtDNA repair activity, driven by the strong metabolic aberrations observed early in the reoxygenation period in crucian carp (Dahl et al., 2021).
Conclusion
This study further supports previous findings, indicating that adaptations of mitochondrial functions play key roles in the anoxia survival of crucian carp. We show here that crucian carp mitochondria have low levels of ROS production (indicated by low H_2_O_2_ levels) even during reoxygenation and that they inherently release less H_2_O_2_ than mitochondria of the closely related but anoxia-intolerant common carp. The low ROS production could partly be explained by higher total oxidant scavenging capacity. Suppression of complex II activity is another likely contributor to the absence of significant ROS production after anoxic succinate accumulation. Still, there were tissue-specific indications of some oxidative damage, including protein carbonylation and mtDNA damage, but these are probably effectively repaired as revealed by the fact that the crucian carp in nature tolerates year after year of anoxia and reoxygenation.
Supplementary Material
10.1242/jexbio.251839_sup1Supplementary information
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Andrienko, T. N., Pasdois, P., Pereira, G. C., Ovens, M. J. and Halestrap, A. P. (2017). The role of succinate and ROS in reperfusion injury – A critical appraisal. J. Mol. Cell. Cardiol. 110, 1-14. 10.1016/j.yjmcc.2017.06.01628689004 PMC 5678286 · doi ↗ · pubmed ↗
- 2Ballinger, S. W., Patterson, C., Yan, C. N., Doan, R., Burow, D. L., Young, C. G., Yakes, F. M., Van Houten, B., Ballinger, C. A., Freeman, B. A. et al. (2000). Hydrogen peroxide- and peroxynitrite-induced mitochondrial DNA damage and dysfunction in vascular endothelial and smooth muscle cells. Circ. Res. 86, 960-966. 10.1161/01.RES.86.9.96010807868 · doi ↗ · pubmed ↗
- 3Bergamini, C., Gambetti, S., Dondi, A. and Cervellati, C. (2004). Oxygen, reactive oxygen species and tissue damage. Curr. Pharm. Des. 10, 1611-1626. 10.2174/138161204338466415134560 · doi ↗ · pubmed ↗
- 4Bezawork-Geleta, A., Rohlena, J., Dong, L., Pacak, K. and Neuzil, J. (2017). Mitochondrial complex II: at the crossroads. Trends Biochem. Sci. 42, 312-325. 10.1016/j.tibs.2017.01.00328185716 PMC 7441821 · doi ↗ · pubmed ↗
- 5Bliksøen, M., Baysa, A., Eide, L., Bjørås, M., Suganthan, R., Vaage, J., Stensløkken, K. O. and Valen, G. (2015). Mitochondrial DNA damage and repair during ischemia-reperfusion injury of the heart. J. Mol. Cell. Cardiol. 78, 9-22. 10.1016/j.yjmcc.2014.11.01025446179 · doi ↗ · pubmed ↗
- 6Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. 10.1016/0003-2697(76)90527-3942051 · doi ↗ · pubmed ↗
- 7Bundgaard, A., James, A. M., Gruszczyk, A. V., Martin, J., Murphy, M. P. and Fago, A. (2019). Metabolic adaptations during extreme anoxia in the turtle heart and their implications for ischemia-reperfusion injury. Sci. Rep. 9, 2850. 10.1038/s 41598-019-39836-530808950 PMC 6391391 · doi ↗ · pubmed ↗
- 8Bundgaard, A., Ruhr, I. M., Fago, A. and Galli, G. L. J. (2020 a). Metabolic adaptations to anoxia and reoxygenation : New lessons from freshwater turtles and crucian carp. Curr. Opin. Endocr. Metab. Res. 11, 55-64. 10.1016/j.coemr.2020.01.002 · doi ↗