Effects of acute hypoxia and reoxygenation on oxygen sensors, respiratory metabolism, oxidative stress, and apoptosis in hybrid yellow catfish “Huangyou-1”

The regulation mechanism of the hybrid yellow catfish “Huangyou-1” was assessed under conditions of hypoxia and reoxygenation by examination of oxygen sensors and by monitoring respiratory metabolism, oxidative stress, and apoptosis. The expressions of genes related to oxygen sensors (HIF-1α, HIF-2α, VHL, HIF-1β, PHD2, and FIH-1) were upregulated in the brain and liver during hypoxia, and recovered compared with control upon reoxygenation. The expressions of genes related to glycolysis (HK1, PGK1, PGAM2, PFK, and LDH) were increased during hypoxia and then recovered compared with control upon reoxygenation. The mRNA levels of CS did not change during hypoxia in the brain and liver, but increased during reoxygenation. The mRNA levels of SDH decreased significantly only in the liver during hypoxia, but later increased compared with control upon reoxygenation in both tissues. Under hypoxic conditions, the expressions of genes related to oxidative stress (SOD1, SOD2, GSH-Px, and CAT) and the activity of antioxidant enzymes (SOD, CAT, and GSH-Px) and MDA were upregulated compared with control. The expressions of genes related to apoptosis (Apaf-1, Bax, Caspase 3, Caspase 9, and p53) were higher than those in control during hypoxic exposure, while the expressions of Bcl-2 and Cyt C were decreased. The findings of the transcriptional analyses will provide insights into the molecular mechanisms of hybrid yellow catfish “Huangyou-1” under conditions of hypoxia and reoxygenation. Overall, these findings showed that oxygen sensors of “Huangyou-1” are potentially useful biomarkers of environmental hypoxic exposure. Together with genes related to respiratory metabolism, oxidative stress and apoptosis occupy a quite high position in enhancing hypoxia tolerance. Our findings provided new insights into the molecular regulatory mechanism of hypoxia in “Huangyou-1.”


Introduction
In natural aquatic environments, the dissolved oxygen content is usually difficult to maintain. The spatiotemporal heterogeneity distribution of dissolved oxygen affects the distribution of aquatic organisms, whose evolutionary strategy involves induced adaptation to an anoxic environment (Herbert and Steffensen 2005;Thomas and Rahman 2009). Teleost fish have evolved complex physiological and biochemical systems to cope with hypoxia and to maintain the oxygen balance in vivo (Herbert and Steffensen 2005;Zhu et al. 2013).
Pelteobagrus fulvidraco and Pelteobagrus vachelli belong to Actinopterygii, Siluriformes, Bagridae, and Pelteobagrus. They are among the most important economic fishes in China. With the expansion of the aquatic market, intensive farming of P. fulvidraco and P. vachelli has become mainstream. During the process of artificial aquaculture, erroneous maintenance practices of the respective personnel often lead to "turning pools." Previous studies have shown that dissolved oxygen concentrations that are either too low or too high can affect the growth, reproduction, and immune function of P. fulvidraco and P. vachelli (Kai et al. 2015;Zhang et al. 2016a, b, c;Zhang et al. 2017). Breeding methods, such as whole male breeding and crossbreeding, have allowed great progress in the improvement of fish species. Hybrid yellow catfish, "Huangyou-1," is a hybrid of P. fulvidraco♀ and P. vachelli ♂ (Qiang et al. 2019a, b;Zhang et al. 2019). It has become an important freshwater aquaculture species in China due to attributes that include good taste, no intermuscular fishbones, and high nutritional value. Many research studies have investigated bacterial infections, temperature tolerance, and hunger stimulus of hybrid yellow catfish, with less research having been done on hypoxia (Qiang et al. 2019a, b;Zhang et al. 2019). We have found that the ability of "Huangyou-1" juveniles to tolerate hypoxia is significantly higher than that of paternal tile-type P. vachelli (Zhang et al. 2017), which makes it a suitable experimental organism to study hypoxia stress in aquatic waters. Information on the mechanism of hypoxia tolerance of "Huangyou-1" catfish could provide theoretical support for further improvement of new breeds.
The hypoxia-inducible factor-1 (HIF-1) transcription factor was discovered in 1992 during an investigation of the expression of the erythropoietin gene (EPO) in the Hep3B cell line (Semenza 2000;Semenza and Wang 1992;Wang and Semenza 1993). HIF-1 is the most important family of transcription factors known to respond to oxygen concentrations in vivo. HIF is a heterodimer composed by HIF-α and HIF-β (Semenza 2000). HIF-β is considered to be an aryl hydrocarbon nuclear translocator that is typically expressed in large quantities in organisms (Bi et al. 2015;Santhakumar et al. 2012). When the oxygen concentration is sufficient, the proline residue of HIF-α is hydroxylated by proline hydroxylases (PHDs), which are recognized by von Hippel-Lindau tumor suppressor (Vhl) proteins, and is degraded by ubiquitination and proteasomes (Maxwell et al. 1999;Niecknig et al. 2012). However, PHDs are unable to hydroxylate HIF-α due to lack of oxygen molecules during hypoxia (Appelhoff et al. 2004). HIF-α rapidly accumulates and binds to HIF-1β to form a heterodimer. These heterodimers bind to hypoxia-responsive elements contained in the promoter regions of target genes, regulating their transcription and initiating a series of biochemical and physical responses . In addition to PHD2, HIF-1 asparaginyl hydroxylase (FIH-1) is also widely recognized as an oxygen-substituted hydroxylase (So et al. 2014). During normoxia, FIH-1 can hydroxylate the asparagyl residue within the C-terminal transactivation domain (C-TAD) (Lando et al. 2002). HIF-1 is important in hypoxic pathways, apoptotic factors, glucose transporter (GLUT1) and glycolysis-related factors (Chen et al. 2001;Ravi et al. 2000;Semenza 2000). To date, studies on hypoxia-inducible factors in fish that have included Ictalurus punctatus, Megalobrama amblycephala, Ctenopharyngodon idellus, and P. vachelli have focused on HIF-1α, HIF-2α, FIH-1, and PHD2 (Geng et al. 2014;Law et al. 2006;Zhang et al. 2016a, b, c;Zhang et al. 2017). There have been relatively fewer studies on HIF-1β and Vhl.
In vertebrates and invertebrates, glycogen metabolism is the primary source of energy, especially in unstable environments. Under normoxia, cells produce ATP through glycolysis and oxidative phosphorylation in mitochondria (Li et al. 2018). However, during hypoxia, both the respiratory chain and aerobic metabolism are inhibited. The enhancement of anaerobic metabolism is one of the energy compensation methods to deal with hypoxia stress (Zhang et al. 2003). The hypoxic activation of HIF-1 promotes ATP production through increased anaerobic glycolysis, which partially compensates for hypoxic cellular energy demands (Fulda and Debatin 2007). It has been suggested that a hypoxia-induced metabolic switch acts to shunt glucose metabolites away from mitochondria to maintain ATP production and to prevent the production of toxic reactive oxygen species (ROS) (Kim et al. 2006;Luo et al. 2006). Studies of the changes of metabolite concentrations and the activity of these key enzymes in respiratory metabolism could provide a way to test the severity of the hypoxic response of organisms.
Fish regulate their energy metabolism in response to hypoxic stress and also catalyze physiological and biochemical reactions to counteract the damage caused by excessive ROS (Lushchak et al. 2005). In general, fish mitochondria are thought to produce less ROS in a normoxia environment. In the case of hypoxia/reoxygenation, the total amount of ROS exceeds the maximum tolerance of the organism, resulting in severe reoxidative stress damage (Zhang et al. 2016a, b, c). ROS are continuously produced in organisms by non-enzymatic and enzymatic reactions. Simultaneously, the ROS are continuously removed by the synergistic action of the antioxidant enzymes and exogenous/endogenous antioxidants. The main antioxidant enzymes in fish are superoxide dismutase (Cu/Zn SOD [SOD1] and Mn-SOD [SOD2]), catalase (CAT), and glutathione peroxidase (GSH-Px) (Leveelahti et al. 2014). SOD 1 and SOD 2 can protect cells from potential ROS damage by converting superoxide anion to hydrogen peroxide (H 2 O 2 ) and H 2 O. H 2 O 2 will eventually decompose to H 2 O and O 2 under the action of CAT and GSH-Px, so that the cells are protected from H 2 O 2 damage.
The study of apoptosis during hypoxia can clarify the adaptation to hypoxia that can occur. Apoptosis is a highly regulated programmed cell death (Poon et al. 2007). When cells are exposed to hypoxia, oxidative phosphorylation of mitochondria is inhibited, resulting in reduced ATP production and the production of a large amount of ROS in the electron transport chain due to the insufficient supply of oxygen molecules to the electron acceptor (Tanaka et al. 2002). Under acute hypoxia, HIF-1 has a proapoptotic role mainly through the ROSdependent pathway (Mansfield et al. 2005). HIF-1 can modulate mitochondrial membrane permeability under hypoxic conditions through regulation of B-cell lymphoma-2 (Bcl-2) protein family members, including Bcl-2, or by increased mitochondrial permeability transition pore (PTP) activity to release mitochondrial cytochrome c (Cyt C) and form an apoptotic complex composed of Cyt C, Apaf-1, and Caspase 9, which initiates a Caspase cascade to activate Caspase 3 (Carmeliet et al. 1998;Mansfield et al. 2005). As a tumor-suppressor gene, p53 plays an important role in regulating cell growth, differentiation, and proliferation. p53 affects the immune response when an organism is exposed to hypoxia or excessive ROS (Moll and Zaika 2001). p53 can regulate the activities of energy metabolism and respiratory metabolism-related enzymes, and coincidentally affects the expressions of the Bcl-2 protein family as well as oxidative phosphorylation-related genes (Erster and Moll 2004;Luo et al. 1998;Riva et al. 2001). Few studies have investigated hypoxiarelated mitochondrial apoptosis pathways in fish, especially acute hypoxic stress.
Based on the above findings, we hypothesized that hypoxia will activate the HIF-1 signaling pathway and cause some important physiological and biochemical changes in the "Huangyou-1" catfish. To test this hypothesis, we analyzed the transcriptional regulation of genes that encode oxygen sensors PHD2,and Vhl) in response to hypoxia and reoxygenation. In addition, we evaluated the energy requirements and antioxidant capacity of hybrid yellow catfish under hypoxic conditions by observing the expression of respiratory metabolism-related genes (HK1, PGK1, PGAM2, PFK, LDH, CS, and SDH) and oxidative stress-related genes (SOD1, SOD2, GSH-Px, and CAT ). We also studied the expression changes of apoptosis-related genes (Apaf-1, Bax, Bcl-2, Cyt C, Caspase 3, Caspase 9, and p53) to evaluate the effect of hypoxia on apoptosis. The findings will contribute to a better understanding of molecular mechanisms of the hypoxia signaling pathway for "Huangyou-1" catfish under conditions of hypoxia and reoxygenation.

Experimental fish
Healthy "Huangyou-1" catfish (5 months old, 9 ± 2.1 cm in length, 12 ± 2.3 g wet weight) were collected from Nanjing Fisheries Research Institute, China. The fish were randomly allocated to five aquaculture glass tanks that each contained a biofilter. The dimensions of each tank (L × W × H) were 0.8 m × 0.55 m × 0.4 m). Tank conditions were as follows: water flow rate was 5 L/min, temperature was 24 ± 1 °C, and pH was 7.5 ± 0.2. The juvenile fish were fed an artificial diet at 9 am and 5 pm every day for 1 month, and were fasted for 24 h before the trial.

Determination of oxygen threshold
One hundred fifty healthy and disease-free "Huangyou-1" catfish were selected and put into three water circulation aquaculture biofilter-equipped aquaria with biofilter set to average. The dissolved oxygen in the water before hypoxia stress started was 7.29 ± 0.40 mg/L, as measured using an HQd Portable Meter equipped with an LDO101 probe (LDO, USA). After the start of the hypoxia experiment, the aerator and water intake were stopped, and the entire glass cylinder was sealed with transparent film and filled with nitrogen. During the experiment, the activity of fish was observed. The fish sought to obtain more oxygen by direct mouth breathing, which is often referred to as "floating head." When the oxygen concentration was less than 0.55 ± 0.06 mg/L, floating head behavior was evident. When the oxygen concentration was less than 0.25 ± 0.05 mg/L, the fish suffocated. Therefore, we defined an oxygen concentration of 0.7 mg/L as the hypoxic condition for the hypoxia challenge in this study.
Acute hypoxia exposure and reoxygenation (recovery) Three hundred "Huangyou-1" catfish were randomly placed into six water recycling aquaculture aquaria that were individually equipped with a biofilter device. The normal oxygen control group (C) and the hypoxic stress recovery group (H and R) were in three parallel groups. Each group contained 50 individuals. Before the experiment, six fish were taken from the C and H groups and the hypoxic stress recovery group as H0 and C0, respectively. The intake and aerator of the hypoxic stress group were closed, and the hypoxic stress experiment was started. Each aquarium received nitrogen for approximately 40 min. The dissolved oxygen level reached 0.7 ± 0.05 mg/L. This level was maintained for 6.5 h. After this period of hypoxia stress, oxygen was introduced into the aquarium. After 30 min, dissolved oxygen recovered to 7.29 ± 0.40 mg/L and this level was maintained for 6.5 h. Seven aquaria corresponded to seven time points of challenged group sampling [hypoxia (H0, H1.5, H4, and H6.5), and reoxygenation (R1.5, R4, and R6.5)]. The other seven normoxia aquaria were set as the control (C0, CH1.5, CH4, CH6.5, CR1.5, CR4, and CR6.5) corresponding to seven time points of sampling. The fish in the hypoxia stress group and the normoxia control group were dissected on an ice tray. Liver, brain, gill, intestine, spleen, heart, head, kidney, and muscle tissues were retrieved for genomic tissue expression analysis. Blood samples were collected from the caudal vein during the dissections. Six experimental fish tissues were mixed in the same glass tank, treated with liquid nitrogen, and stored at − 80 °C for further analysis. The experiment was repeated three times. Two fish were randomly taken from each aquarium and mix into one sample. Groups C and H were repeated thrice with three aquariums.

Enzymatic activity
The changes in indices in all samples were measured using superoxide dismutase (SOD-A001: the xanthine oxidase method), malondialdehyde (MDA-A005: the thiobarbituric acid developing method), catalase (CAT-A007: the ammonium molybdate method), glutathione peroxidase (GSH-Px-A005: the enzymatic recycling method), and lactate dehydrogenase (LDH-A020: the 2,4-dinitrophenylhydrazine developing method) kits (all from Jiancheng Bioengineering, Nanjing, China). The liver tissue samples were washed with ice-cold 0.9% saline (1:9 m/v), weighed, and homogenized with 10 volumes of 0.9% saline. The homogenate was centrifuged at 4 °C for 10 min at 2500 rpm. In this experiment, no related experimental determinations were performed in the brain due to the limited sampling volume of the brain. Extract the blood in a centrifuge tube at 4 °C for 1 h, and then centrifuge tubes at 2500 rpm for 10 min. The enzyme activities and MDA content in the supernatant were determined. The protein concentrations were determined by Coomassie Brilliant Blue staining of crude extracts (Jiancheng Bioengineering). Each sample was measured in triplicate.

qRT-PCR analysis
The High Purity RNA Fast Extract Reagent (BioTeke, Beijing, China) was used to extract total RNA from liver and brain of the "Huangyou-1" catfish. The quality of RNA extracted was determined by spectrophotometry using a model NanoDrop-1000 instrument (Thermo Fisher Scientific, Waltham, MA, USA). Table 1 lists the primer sequences that were used. All primers were validated by the DNA dilution sequence, and the display efficiency exceeded 90%. Single-stranded cDNA was synthesized using HiS-cript™ QRT SuperMix (Vazyme Biotech Inc., Piscataway, NJ, USA). PCR amplification was performed in triplicate using the following cycling parameters: 94 °C for 30 s followed by 40 cycles of 95 °C for 15 s and 55 °C for 1 min, and an extension period at 72 °C for 60 s. To confirm the specificity of the amplification, the dissociation curve was analyzed for amplified products to ensure an obvious amplification peak. The expression level was calculated by the 2 −ΔΔCt method and statistically analyzed. The expression of mRNA in liver and brain tissues after hypoxia was detected by qRT-PCR.

Western blotting analysis
Total protein was extracted from frozen samples using commercial kits (KeyGen Biotech, Nanjing, China). Protein concentrations were determined using a Pierce® BCA Protein Assay Kit (Thermo Fisher Scientific). Proteins were resolved using 12% SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Milople, Bedford, MA, USA) to block 5% albumin bovine V (Solarbio, Beijing, China). The mouse antibody to beta-actin (1:2600 dilution, A5441; Sigma-Aldrich, St. Louis, MO, USA) was used as the internal reference. It and the following antibodies were applied and cultured overnight at 4 °C: with different FIH-1 (1:1000 dilution, D123653; Bioworld Technology, Shanghai, China) and Bax (1:1000 dilution, D220073-0025; Bioworld Technology). Samples were then treated using goat anti-rabbit IgG secondary antibody (SAB, Baltimore, MD, USA) or goat anti-mouse IgG secondary antibody (SAB). Immuno-reactive bands were visualized with a chemiluminescence reagent (PerkinElmer Life Science, Waltham, MA, USA). Densitometry analysis was performed using ImageJ software (NIH, Bethesda, MD, USA).

Statistical analyses
Compared with the control group, the changes of each index of "Huangyou-1" catfish were reflected after hypoxia stress. The experimental results were calculated by single-factor analysis of variance (oneway ANOVA) using SPSS V22.0 software (SPSS Inc., Chicago, IL, USA). The T-test was used to calculate P-value. A P-value < 0.05 was considered statistically significant. The values are expressed by mean ± standard deviation (SD) of triplicate samples.

Tissue distribution of genes
Genes related to oxygen sensors, energy metabolism, oxidative stress, and apoptosis of "Huangyou-1" catfish were generally expressed in tissues that included intestine, liver, muscle, spleen, heart, gill, brain, and kidney. The expressions of HIF-1α, PHD2, CAT , and SOD1 were highest in the liver while that of HIF-1β, HIF-2α, Vhl, FIH-1, HK1, PFK, and PGK1 were highest in the heart. CS, PGAM2, SDH, and SOD2 expressions were highest in the muscle. GSH-Px, Apaf-1, Bax, Caspase 3, and Caspase 9 were abundantly expressed in the spleen. The highest expressions of LDH, p53, and Bcl-2 were in the kidney (Fig. 1).
Temporal expression profiles of related genes of "Huangyou-1" catfish during hypoxia and reoxygenation In the liver, the expressions of HIF-1α, PHD2, HIF-1β, HIF-2α, Vhl, and FIH-1 related to oxygen sensors were increased during hypoxia exposure in "Huangyou-1" catfish. The expressions of HIF-1β, Vhl, and PHD2 peaked at H6.5 (P < 0.01). The expressions of HIF-2α and Vhl returned to basal levels at R6.5. The expressions of HK1, PFK, PGK1, LDH, and PGAM2 related to glucose metabolism were increased under hypoxic exposure, and then returned to their basal levels during reoxygenation. The expressions of PFK, PGK1, LDH, and PGAM2 were significantly higher than the control group at H6.5 (P < 0.05). The expression of CS was almost unchanged under hypoxia, while it increased significantly during reoxygenation and reached the highest at R6.5 (P < 0.05). The expression of SDH in H4 was significantly lower than that in the control group (P < 0.05). SDH expression was increased during reoxygenation and was highest at R1.5 (P < 0.01). The expressions of SOD 1 , SOD 2 , GSH-Px, and CAT related to oxidative stress were increased under hypoxia, and they were not significantly different compared with the control group at R6.5. The expressions of Apaf-1, Bax, and p53 related to apoptosis were increased. Apaf-1 and Bax returned to basal levels when the oxygen supply was restored. The expressions of Caspase 3 and Caspase 9 were increased compared with the control group during hypoxia/ reoxygenation. The expressions of Cyt C and Bcl-2 were decreased in hypoxia and approached the control group during reoxygenation (Fig. 2).

Enzymatic activities
In the liver, the SOD, LDH, and CAT activities and MDA levels were highest at R1.5 compared with the control group during hypoxia. All increased from H0 to R1.5 under hypoxia, and returned to the values of the control group during reoxygenation (P < 0.01) (Fig. 4).
In the serum of "Huangyou-1" catfish, the SOD activity was increased continuously and peaked (P < 0.001) at R1.5 under hypoxia. The activity was restored to normal at R6.5. The LDH activity and level of MDA were increased slightly under hypoxic stimulation. They gradually returned to the values of the control group at R6.5. GSH-Px activity started increasing from H4 and at R4, was higher than that observed in the control group (P < 0.05) (Fig. 5).

Western blotting analysis
In the liver, the amounts of FIH-1 and Bax were upregulated compared with control during hypoxia. The abundance of FIH-1 was decreased from R1.5 compared with control. The abundance of Bax was obviously higher than that in the control (P < 0.01).

Discussion
This study provides the first evidence of the effects of acute hypoxia and reoxygenation on oxygen sensors, respiratory metabolism, oxidative stress, and apoptosis in hybrid yellow catfish "Huangyou-1." qRT-PCR revealed that the genes related to oxygen sensors (HIF-1α, HIF-1β, HIF-2α, PHD2, VHL, and FIH-1) were highly expressed in the brain and heart, similar to P. fulvidraco and Takifugu fasciatus Zhang et al. 2017). However, the expressions of genes related to oxygen sensors were mainly concentrated in the liver, brain, and heart, similar to I. punctatus, Dicentrarchus labrax, and Danio rerio (Geng et al. 2014;Liu et al. 2013;Terova et al. 2008). The different distribution of these genes indicates that they might be related to their particular physiological functions. Because of the high expression of oxygen sensors related genes in the liver, heart, and brain, and the extensive reports of liver and brain in hypoxia, these two tissues were selected as the candidate tissues to assess gene expression under hypoxia and reoxygenation conditions. The induction mechanism of HIF-α has been confirmed and widely reported in higher vertebrates during hypoxia. Presently, the expressions of HIF-1α and HIF-2α were significantly higher in brain and liver tissues of "Huangyou-1" catfish during hypoxia as compared to the control group. Similar results were observed in D. labrax, Carassius auratus, and Micropogonias undulatus (Sollid et al. 2005;Terova et al. 2008;Thomas and Rahman 2009). It has been documented that HIF-α can be hydroxylated to limit accumulation under normal oxygen conditions and that short-term hypoxia stimulation can inhibit hydroxylation, resulting in the binding of HIF-α to HIF-β to activate the downstream cascade (Thomas and Rahman 2009;Walmsley et al. 2005). This may be why HIF-1α and HIF-2α tended to be upregulated after the establishment of hypoxia. In our study, PHD2, Vhl, and FIH-1 were highly expressed during hypoxia and returned to the initial Fig. 6 Western blot analysis of proteins related to oxygen sensors and apoptosis indices in the brain and liver of "Huangyou-1" catfish during acute hypoxia and reoxygenation. FIH-1, Bax, and β-actin proteins were evident at approximately 40, 21, and 42 kDa, respectively, following SDS-PAGE. Signif-icant differences compared with the control group are denoted by * (P < 0.05), ** (P < 0.01), and *** (P < 0.001). Densitometry analysis was performed using ImageJ software. a Brain (FIH-1). b Liver (FIH-1). c Brain (Bax). d Liver (Bax) level after reoxygenation, similar to the T. fasciatus, M. amblycephala, I. punctatus, and P. vachelli (Geng et al. 2014;Li et al. 2019;Zhang et al. 2016a, b, c;Zhang et al. 2017). We suspect that the upregulation of PHD2, Vhl, and FIH-1 might act as a feedback mechanism to terminate hypoxic responses to minimize the exposure of the brain and liver to hypoxic stress (D'Angelo et al. 2003).
The energy supply of organisms under hypoxic stress is mainly dependent on a powerful energy supply system. We observed that the expressions of HK1, PFK, PGK1, and PGAM2 were increased under hypoxic conditions, and recent studies have indicated that these genes may play vital roles in the evaluation of energy supply in aquatic animals (D'Angelo et al. 2003). For example, both HK1 and PFK are rate-limiting enzymes for glycolysis, and an increase in their expression indicates an increase in energy metabolism. The upregulation of HK1 and PFK expressions was detected in Oreochromis niloticus and Liposarcus pardalis under hypoxic conditions, similar to our results (Li et al. 2018;Treberg et al. 2007). The results suggest that the anaerobic metabolism of these organisms is promoted. During hypoxia, we detected significant upregulation of LDH in the liver and brain of "Huangyou-1" catfish, similar to O. niloticus, Leiostomus xanthurus, and Astronotus crassipinnis (Almeida-Val et al. 2011;Cooper et al. 2002;Li et al. 2018). We speculate that the production of ATP is reduced under hypoxic conditions, and that the large consumption of glycogen and glucose leads to a pronounced accumulation of lactic acid. LDH can catalyze the conversion of pyruvate to lactic acid, and its activity can reflect the degree of anaerobic respiration. The increased activity of LDH also indicates that anaerobic respiration metabolism was dominant. The results support the view that anaerobic metabolism is promoted in "Huangyou-1" catfish under acute hypoxic conditions, which supplements the energy demand of cells to some extent. No significant changes in CS and SDH were evident in the liver and brain tissue under acute hypoxic conditions, with a gradual increase in their levels after reoxygenation. Similar results have been described in Astronotus ocellatus, T. fasciatus, and P. vachelli (Baptista et al. 2016;Li et al. 2019). We speculate that this undifferentiated activity of SDH and CS in liver and brain tissues reflects the low dependence on oxidative metabolism to generate energy under hypoxia. Of note, the SDH activity of D. rerio was lower than that of the control group during 3 weeks of hypoxia (10% air/90% N 2 saturated water), and the expression and activity of SDH in the liver of Pseudosciaena crocea was also decreased within 48 h of hypoxia (Jaspers et al. 2014;Zeng et al. 2016). In addition, the enzyme activity of CS was decreased during 24 h of hypoxia in Cyprinus carpio, and the enzyme activity of CS in the liver of A. ocellatus was decreased after 48 h of hypoxia (Baptista et al. 2016;Zhou 2000). The aforementioned studies indicate that the aerobic metabolism of fish may be affected by the length of hypoxia. The enzyme activity-associated aerobic metabolism was lower or appeared not to change significantly compared with the control during short-term hypoxia stress, while it inhibited the tricarboxylic acid cycle during long-term hypoxia stress. The release of most of the energy in mitochondrial respiratory metabolism depends on the electron transport system or electron transport chain. Cyt C is an important carrier of the electronic respiratory chain. Cyt C binds to the surface of the inner mitochondrial membrane through negatively charged phospholipids. Under acute hypoxia exposure, the expression of Cyt C was decreased in liver and brain tissues, which is presumably related to the inhibition of the respiratory chain under hypoxic conditions. Ferrero believes that mitochondrial respiratory chains are inhibited to produce large amounts of endogenous ROS under hypoxic conditions, and that increased ROS can inhibit Krebs cycling (Ferrero et al. 2011). Tissue quantification results indicated that genes involved in energy metabolism and respiratory metabolism were widely expressed in heart and muscle tissues, indicating that they were the main organs of energy consumption. In general, anaerobic metabolic pathways are the main source of ATP in fish in response to acute hypoxia challenge. This was presently confirmed by the changes in the enzymes related to glycolytic and Krebs cycle of "Huangyou-1" catfish in a hypoxic environment. In other words, during hypoxia, metabolic activity changes to reduce energy consumption.
Activation of the biological antioxidant defense system prevents damage due to ROS when the production rate of ROS is faster than the scavenging rate of oxygen free radicals (Leveelahti et al. 2014). Presently, the expressions of GSH-Px, SOD1, SOD2, and CAT reflected the changes in the organism's antioxidant capacity. SOD can catalyze the conversion of superoxide radicals to H 2 O 2 , while CAT and GSH-Px catalyze the conversion of H 2 O 2 to H 2 O. MDA can reflect the degree of lipid peroxidation (Keramati et al. 2010). We observed that the expressions of SOD1, CAT, and GSH-Px were highest at H6.5 or R1.5 in the liver of "Huangyou-1" catfish, indicating that the oxidative stress defense initiated under hypoxia/reoxygenation in vivo. Similar results have been observed in Threespine stickleback, Cyphocharax abramoides, and Leiostomus xanthurus (Cooper et al. 2002;Johannsson et al. 2018;O'Connor et al. 2011). The enzyme activities of SOD, GSH-Px, and CAT were highest at R1.5 compared with the control group in the serum and liver of "Huangyou-1" catfish. The time difference in the enzyme activities and expressions of these antioxidant-related genes may be due to the existing of post-transcriptional initiation and termination of transcription, as well as posttranslational changes. The CAT and SOD activities at 10% oxygen saturation were significantly greater than those at 25% and 100% oxygen saturation in L. xanthurus (Cooper et al. 2002). Similar to our experiment, the SOD1 and SOD2 enzyme activities were increased in the liver of P. vachelli under hypoxia exposure, and were highest at 1.5 h of reoxygenation (Zhang et al. 2016a, b, c). It is worthy of mentioning that the level of MDA was not significantly different from those of the control group at H4, and it increased significantly during H6.5 to R4 in serum and liver, suggesting that lipid peroxidation occurred in the fish body under hypoxia/reoxygenation. In addition, the process of oxygen recovery from hypoxia to normoxia is relatively rapid in the experiment, so it is easy to cause oxidative stress in "Huangyou-1." The formulation of the hypothesis of "preparation for oxidative stress" means a decrease in some antioxidant defenses will be expected during hypometabolic states, while transient ROS-induced oxidative stress increases rapidly as oxygen re-enters anoxic tissue (Oliveira et al. 2005;Mohibbullah et al. 2015;Wei et al. 2015). Therefore, the enhancement of antioxidant defense in "Huangyou-1" under hypoxia may not only prevent the oxidative stress injury caused by hypoxia, but also to that caused by reoxygenation. Our experiments confirmed that oxidative damage often occurs in fish during hypoxia/reoxygenation and the existence of antioxidant system could protect cells and reduce the damage caused by oxidative stress.
This study focused on the physiological compensation of the organism under hypoxic stimulation. Changes of these genes in fish can sense oxygen levels and produce appropriate adaptive responses. When the tolerance limit is exceeded, programmed cell death results (Mazure and Pouysségur 2010;Shimizu et al. 1995). Hypoxia-induced apoptosis is mainly caused by the external death receptor pathway and the endogenous mitochondrial pathway to induce apoptosis in the liver and brain (Carmeliet et al. 1998;Yin et al. 2018). These actions induce apoptosis under specific conditions. Previous reports have suggested that ROS-induced oxidative stress is one of the main causes of mitochondrial autophagy, and that oxidative stress-mediated elimination of ROS can inhibit mitochondrial apoptosis (Tanaka et al. 2002). Under hypoxic conditions, ROS produced by mitochondria can oxidize the critical thiol groups in adenine nucleotide translocase, which causes the release of Cyt C to further aggravate mitochondrial apoptosis (Kluck 1997;Luo et al. 1998). Therefore, we focused on the study of the endogenous mitochondrial pathway. Under acute hypoxia exposure, the expressions of Apaf-1, Caspase3, Caspase 9, and the gene encoding Bcl2-associated x protein (Bax) were increased in liver and brain tissues, while the expressions of the gene encoding Bcl-2 was decreased. Similar to our results, the expression of Bcl-2 was decreased in gill of I. punetaus under hypoxia (Yuan et al. 2016). On the other hand, Bcl-2 was increased in liver of the C. carpio L. during 4 days of hypoxia (0.5 mg/L) (Poon et al. 2007). The reason may be that the Bcl-2 protein is inhibited and accelerates apoptosis under short-term hypoxic stimulation, while an increased proportion of Bcl-2/Bax inhibits apoptosis and protects cells under long-term hypoxic stimulation Riva et al. 2001). This could also be a mechanism by which fish adapt to hypoxic stress. In addition, we found that the expression of Bcl-2 was increased and the expression of Bax was decreased after reoxygenation. This may be due to the formation of Bax-Bcl-2, which was more stable than Bax-Bax in the cells, which ultimately led to the inhibition of apoptosis. Marzo et al. found that the opening of PTP is regulated by the Bcl-2 protein family, which can affect the barrier function of the membrane by forming pores on the mitochondrial membrane (Brenner et al. 2000;Marzo et al. 1998). Bcl-2 protein can inhibit the pore formation of Bax, and both Bcl-2 chemical modification and proteolysis can affect its activity (Ding et al. 2014). Therefore, the ratio of Bcl-2/Bax is crucial for the direction of apoptosis. The expression of Caspase 3 was significantly increased in the central nervous system of Acipenser schrenckii after 30 min of acute hypoxia (15% oxygen saturation), Similar results was observed in liver and brain tissues of "Huangyou-1" catfish (Lu et al. 2005). Caspase 3 is a common effector of the mitochondrial apoptotic pathway and the death receptor pathway. Under normal conditions, Caspase 3 exists in normal cells in the form of a dormant cryptogen, which can cleave many protein substrates to cause apoptosis (Hua et al. 2015;Lu et al. 2005). Moreover, we observed very similar expression profiles of Caspase 3 and Caspase 9 during hypoxia, which may be related to the cascade of upstream and downstream of Caspase 9 and Caspase 3. Under hypoxic stress, the cells release Cyt C from the mitochondria into the cytoplasm, which activates Caspase 9 and increases the expression of Caspase 3 (Hu 1999). The release of Cyt C by mitochondria is also an important signal of apoptosis (Mansfield et al. 2005). When cells were subjected to hypoxia stress, Cyt C was separated from the mitochondrial inner membrane and released into the cytoplasm, along with apoptotic protease activating factor 1 (Apaf-1) and Caspase 9. p53 is a common tumor suppressor that is related to negative growth regulation and apoptosis. p53 has been widely reported in teleost fish (Bratton 2001;Gupta and Knowlton 2002). p53 has an effect on both exogenous and endogenous apoptotic pathways, whereas p53 non-transcription-dependent proapoptotic functions act mainly through the mitochondrial pathway (Erster and Moll 2004;Moll and Zaika 2001). Sansome et al. found that when hypoxia-mediated apoptosis occurs, a portion of the induced p53 protein is specifically localized to the mitochondria and interacts with Bcl-2 family members located on the mitochondrial membrane to mediate apoptosis (Chipuk 2004;Li 1999;Suzuki et al. 2001). We detected an increase of p53 in liver and brain under hypoxia, and the corresponding degree of p53 activity in liver tissue was more positive than the corresponding relationship in brain tissue. As well, to some extent, the apoptotic response in liver tissue was more severe than the response in Fig. 7 Putative mechanism of the studies of acute hypoxia on the oxygen sensors, respiratory metabolism, oxidative stress, and apoptosis of "Huangyou-1" catfish. The arrows in the fig-ure indicate changes in gene expression during acute hypoxia. (↑ and underscore increase, ↓ decrease, -unchanged) brain tissue. Other studies reported that p53 has a regulatory effect on glycolysis and oxidative phosphorylation (Corcoran et al. 2014). The present data concerning hypoxia-induced p53 activity require further and more in-depth study. Nonetheless, the upregulation of p53 activity in "Huangyou-1" catfish indicates that it has a certain promoting effect on apoptosis signaling under hypoxia.

Conclusion
Oxygen sensing pathway molecules are the most important factors for an organism to sense the oxygen concentration. These molecules are upregulated during hypoxic stress and can be used as molecular indicators of hypoxia. In addition, the HIF-1 signaling pathway can also respond to changes in energy supply under hypoxia by regulating the expression of genes involved in respiratory metabolism. We observed that the antioxidant system in the liver and brain was activated to protect cells from oxidative stress and apoptosis. The physiological regulation mechanism of "Huangyou-1" catfish under the condition of acute hypoxia and reoxygenation was studied by measuring the physiological indicators of oxygen sensors, respiratory metabolism, oxidative stress, and apoptosis (Fig. 7). However, this was not completely effective in protecting the fish from pronounced changes in oxidative conditions. Although our research has not fully explored the molecular regulation mechanism of hybrid yellow catfish "Huangyou-1" under hypoxia stress, it provides useful evidence to further elucidate the effects of acute hypoxia and reoxygenation on oxygen sensors, respiratory metabolism, oxidative stress, and cell apoptosis. Data availability All datasets for this study are included in the manuscript/supplementary files.

Declarations
Ethical approval All experiments were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals in China. This study was also approved by the Ethics Committee of Experimental Animals at Nanjing Normal University (grant No. SYXK 2015-0028, Jiangsu).

Conflict of interest
The authors declare competing interests.