Functional exploration of SNP mutations in HIF2αb gene correlated with hypoxia tolerance in blunt snout bream (Megalobrama amblycephala)

Blunt snout bream (Megalobrama amblycephala) is sensitive to hypoxia environment. Hypoxia-inducible factor (HIF) is the most critical factor in the HIF pathway, which strictly regulates the hypoxia stress process of fish. In this study, we found six hifα genes in blunt snout bream that demonstrated different expressions under hypoxia conditions. In HEK293T cells, all six hifαs were detected to activate the HRE region by luciferase reporter assay. More importantly, we identified two linkage-disequilibrium SNP sites at exon 203 and 752 of the hif2αb gene in blunt snout bream. Haplotype II (A203A752) and its homozygous diplotype II (A203A203A752A752) appeared frequently in a selected strain of blunt snout bream with hypoxia tolerance. Diplotype II has a lower oxygen tension threshold for loss of equilibrium (LOEcrit) over a similar range of temperatures. Moreover, its erythrocyte number increased significantly (p < 0.05) than those in diplotype I and diplotype III strains at 48 h of hypoxia. The enzymes related with hypoxia tolerant traits, i.e., reduced glutathione, superoxide dismutase, and catalase, were also significantly (p < 0.05) induced in diplotype II than in diplotype I or III. In addition, the expression of epo in the liver of diplotype II was significantly (p < 0.01) higher than that in the diplotype I or III strains at 48 h of hypoxia. Taken together, our results found that the hypoxia-tolerant-related diplotype II of hif2αb has the potential to be used as a molecular marker in future genetic breeding of hypoxia-tolerant strain.


Introduction
Blunt snout bream (Megalobrama amblycephala) is a native species in the affiliated lakes of Yangtze River (Ke 1965;Li et al. 1993). It is an herbivorous freshwater fish species with a high economic value and high disease resistance in China Zhao et al. 2020). The production of blunt snout bream was more than 7.6 × 10 5 tons per year in 2019 (FBMA 2020). However, blunt snout bream is a hypoxia-sensitive species. A seasonal change in temperature or water quality that decreases the dissolved oxygen (DO) concentration in ponds would cause adverse effects in aquaculture (Guan et al. 2017). Therefore, it needs to breed new fish varieties with relatively higher hypoxia tolerance. In 2020, a hypoxia-tolerant strain "Pujiang No. 2" has been developed in our lab. Compared with "Pujiang No. 1," "Pujiang No. 2" has a 27% increase in hypoxia tolerance, and the key dissolved oxygen value of body imbalance (LOE crit , 25 °C) in the fingerling stage has dropped below 0.90 mg·L −1 . At present, the hypoxia tolerance of blunt snout bream is still insufficient, and further research on hypoxia molecular breeding is still needed.
Oxygen is indispensable for survival, growth, development, and reproduction of organisms (Semenza 2014). Therefore, metazoans have evolved complex cellular metabolism and physiological systems to maintain oxygen homeostasis (Dunwoodie 2009). In order to adapt to the hypoxia environment, fish bodies have a series of mediation mechanisms, including changing the respiratory surface area, stimulating angiogenesis, increasing the number of red blood cells or the oxygen-carrying capacity of hemoglobin, activating the antioxidant defense system, and changing the expression of related genes (Huang et al. 2015;Nilsson and Renshaw 2004;Stecyk et al. 2004;Turko et al. 2012). Hypoxia-inducible factor (HIF) is the most critical factor identified in the HIF pathway, which strictly regulates this physiological process in fish (Kaelin and Ratcliffe 2008;Semenza 2010;Weidemann and Johnson 2008). It consists of an erratic alpha subunit and a steady beta subunit (Guan et al. 2014;Rytkonen et al. 2011). In invertebrates, there is only one HIFα, but in vertebrates, there are at least three functional HIFα subtypes, i.e., HIF1α, which regulates the acute hypoxia response, HIF2α regulates chronic hypoxia response, and HIF3α inhibits the activity of two other subtypes (Prabhakar and Semenza 2012;Holmquist-Mengelbier et al. 2006;Rahman and Thomas 2007;Zhang et al. 2012).
Whole-genome replication has played a key role in the adaptive evolution of vertebrates (Holland et al. 1994;Sidow 1996). The third round of wholegenome replication in teleost is thought to drive the diversification of teleost (Kollitz et al. 2014;Opazo et al. 2013). Early in the evolution of teleost, the third round of whole-genome replication specific to teleost yielded six hifα genes: hif1αa/b, hif2αa/b, and hif3αa/b, with deletions of one member of each hifαa/b paralog pair in most teleost, but not in the cyprinids (Rytkonen et al. 2013). Hifα genes have been studied in teleost, including zebrafish (Kopp et al. 2011;Rytkonen et al. 2013) and other cyprinids (Guan et al. 2014;Law et al. 2006;Rytkonen et al. 2013). However, not all the six hifα genes have been found in blunt snout bream. In this study, we isolated six hifα genes in blunt snout bream for the first time, analyzed their amino acid structures, detected gene expression after hypoxia, and verified for HREdependent transcriptional activity. In addition, we found that two SNP (single nucleotide polymorphism) sites on hif2αb gene were identified with a selected strain of blunt snout bream with hypoxia tolerance. Our results have implications for studying the HIF system and how teleost HIF plays a role in hypoxia stress in teleost fish. And the hypoxia-tolerant-related diplotype II of hif2αb has the potential to be used as a molecular marker in future genetic breeding of hypoxic-tolerant strain.

Experimental fish
Blunt snout bream specimens were obtained from the Bream Genetics and Breeding Center (BGBC) of Shanghai Ocean University, Shanghai, China. Specimens that belonged to "Pujiang No. 2" breed and wild blunt snout bream were used. The new variety of "Pujiang No. 2" was developed by taking the wild blunt snout bream collected in Poyang Lake as the basic population and adopting the population breeding technology supplemented by the hypoxia stress technology through four successive generations. Juvenile fish (30 ± 4.6 g) were acclimatized in indoor tanks for 2 weeks (25 ± 0.3 °C). The DO concentrations in the waters were 8.5 ± 0.5 mg·L −1 .

Hypoxia treatment
Thirty-six blunt snout bream were kept in two 25 L glass tanks filled with dechlorinated tap water for 2 days. The bubbling rate of N 2 and air was controlled to achieve a DO value of 2.0 mg·L −1 in the water for hypoxia treatment. Oxygen levels in water were monitored using oxygen electrodes (YSI, ProODO, Germany). Three individuals were sampled at 0, 6, 12, 24, and 48 h of hypoxia, respectively. Fish were euthanized by 100 mg·L −1 MS-222 (tricaine methane sulfonate, Sigma, USA), and then, the liver tissues were rapidly excised and stored at − 80 °C.

Quantitative real-time PCR (qRT-PCR)
The total RNA was extracted from the livers by RNAiso Plus (Takara, Japan), and the cDNA was synthesized by the PrimeScript RT reagent kit (Takara, Japan) after excluding genomic DNA from the RNA. Instrument, reagents, and method are as previously described . After comparison of internal reference genes, it was found that 18S was most stably expressed in the liver tissues of blunt snout bream. Therefore, 18S was used as the control. Primer pairs used are shown in Additional file 1. All experiments were performed for more than three replicates. The expression level of different genes was analyzed using the 2 −△△CT method after the PCR program.

Plasmids, cell culture, and transfection
For functional analysis, open reading frame (ORF) DNA of hifαs was cloned into a pCS2 + vector to obtain pCS2-hifαs constructs. Primers for this part are shown in Additional file 1. HEK293T cells were grown in medium M199 (Gibco, USA) supplemented with 10% FBS (Gibco, Australia) and 0.5% penicillin and streptomycin (Sangon, China), in a constant temperature cell incubator at 37 °C under normal oxygen (about 5% CO 2 ). Cells were plated onto 24-well plates (1-2 × 10 5 cells/well) 24 h to 55% confluency prior to transfection. The plasmids were transiently transfected into HEK293T cells using Lipofectamine™ 2000 Transfection Reagent (Invitrogen, USA). The cells were then incubated for 48 h to allow DNA uptake and gene expression. Each experiment was repeated three times.

Screening of SNP sites in hifα genes
After examination of the transcriptome data from "Pujiang No. 2" strain (the NCBI Sequence Read Archive website under accession number PRJNA723430), the coding sequences of hifα genes were aligned with the whole genome sequence of blunt snout bream (Liu et al. 2017). To further PCR amplify different SNP sites at 203 and 752 of hif2αb gene, we designed primer sequences covering all the coding regions of the gene (Additional file 1). PCR was performed on samples collected from 100 fish of "Pujiang No. 2," 100 fish of wild blunt snout bream to amplify partial cDNA fragments. Qualified products were sequenced by the Sangon Biotech Company (Shanghai, China).

Oxygen tension threshold for loss of equilibrium
Feeding was stopped before the start of the experiment. Then, determination of LOE crit was initiated. Approximately 90 individuals of the DI, DII, and DIII strains at 2 different water temperatures (15 °C and 25 °C) were used to determine LOE crit . The glass tanks and the methods for treatments were performed as previously described by Wu et al. (2020), and the DO value was decreased in a stepwise manner (decreased over 30 min and then held at the new level for 30 min), in increments of 0.5 mg·L −1 , 242 Fish Physiol Biochem (2023) 49:239-251 to a final level of 0 mg·L −1 . Oxygen levels were monitored using oxygen electrodes. When a fish showed loss of equilibrium, the DO and time were recorded. LOE crit was calculated for individual fish using Brett's equation (Brett 1964). This experiment for LOE crit was repeated 5 times.

Blood analysis
Firstly, 18 individuals of the DI, DII, and DIII strains were subjected to hypoxia with a DO value of 2.0 mg·L −1 . Then, three individuals from each strain were sampled at 0 and 48 h of hypoxia, respectively. Blood samples were immediately drawn from the caudal vein using 1 mL plastic syringes flushed with heparin. The fresh blood was diluted 100 times, and then, a drop of the dilution was placed on a Neubauer board, observed, and counted by a light microscope (Eclipse 80i; Nikon, Japan).

Enzyme activity assays
Then, the liver tissues of the blunt snout bream were weighed and homogenized (dilution 1:10) in ice-cold saline and centrifuged at 2500 rpm for 10 min at 4 °C. The supernatant was obtained and stored at − 20 °C for further determination. The protein concentration was determined by Coomassie Brilliant Blue method (Bradford 1976) using total protein quantitative assay kit (Jiancheng Bioengineering Institute, China). The activity of enzymes was detected using commercial kits including catalase (CAT) assay kit, superoxide dismutase (SOD) assay kit, and reduced glutathione (GSH) assay kit produced by Jiancheng Bioengineering Institute (Nanjing, China).

Statistical analysis
Data from this study were expressed as means ± SE ( n = 3 ). Differences among groups were analyzed by one-way analysis of variance (one-way ANOVA) followed by Fisher's post hoc tests or unpaired t-tests using SPSS Statistics 17.0 software. Significance was accepted at p < 0.05 or p < 0.01.

Characterizations of blunt snout bream hifαs
The NCBI and BioEdit software were used to analyze and classify the six hifα genes to determine the sequence divergence of hifαs (Table 1). PROSITE analysis illustrated that HIFαs of blunt snout bream contained bHLH, PAS-A, PAS-B, and PAC domains (Additional file 2). The phylogenetic tree analysis of vertebrate HIFαs showed that blunt snout bream HIFαs clustered well with teleost orthologs (Additional file 3). The secondary protein structures of HIFαs in blunt snout bream were predicted by Phyre 2 program and PyMOL software. It was found that the secondary structures of HIFαs in blunt snout bream mainly included α-helix and random coil, and they are slightly different (Fig. 1).

Transcription of Hifαs after hypoxia and activation of the HRE region
The relative expression of six hifα mRNAs in the liver of blunt snout bream after hypoxia showed different results (Fig. 2). The expression of hif1αa mRNA decreased significantly (p < 0.05) at 24 h and 48 h (0.34-fold and 0.37-fold, respectively) The expression of hif2αa mRNA was stable after hypoxia, while the expression of hif2αb mRNA in the liver increased significantly (p < 0.05) at 12 h to 48 h (3.26-fold, 7.31-fold, and 7.20-fold, respectively) than that at 0 h of hypoxia. The expression of hif3αs Fig. 1 Predicted secondary structure of HIFαs in blunt snout bream. Amino acid sequences of HIFαs were analyzed by Phyre 2 program and PyMOL software mRNAs in the liver did not change at 6 h compared with that at 0 h of hypoxia and then gradually increased after 6 h, reaching a peak at 48 h (13.00fold and 14.61-fold, respectively). Coexpression of pCS2-hifαs with pGL3-HRE resulted in a highly significant (p < 0.05) increase in reporter activity in HEK293T cells (Fig. 3). Compared with the control group, blunt snout bream hifαs caused 3.13-fold, 2.84-fold, 2.22-fold, 4.88-fold, 4.70-fold, and 3.49-fold increase, respectively (Fig. 3).

Analysis of SNP sites in hifα genes of blunt snout bream
Two linkage disequilibrium SNP sites at 203 and 752 within exons of hif2αb gene with high genetic diversity (PIC > 0.5) were identified 18S mRNA was used as internal control. The results are given as mean ± SE for separate fish ( n = 3 ). Differences among groups were analyzed by unpaired t-tests. Columns marked with different letters are significantly different (p < 0.05)

Fig. 3
Dual-luciferase report analyzed hifαs gene expression in HEK293T cells. Firefly luciferase activity was normalized to Renilla luciferase activity. Data are from three transfection experiments with assays replicated at least three times. The results are given as mean ± SE for separate fish ( n = 3 ). Columns marked with different letters are significantly different (p < 0.05) after examination of transcriptome data from blunt snout bream of "Pujiang No. 2" (Table 2). Two SNP sites are nonsynonymous mutations, and the codon change type of T 203 A 203 site is that AAA-ATA encodes lysine and isoleucine located within the PAS-A domain, respectively. The codon change type of C 752 A 752 site is TAT-TCT, encoding tyrosine and serine located within the PAS-B domain, respectively. There are two haplotypes, haplotype I (HI, T 203 C 752 ) and haplotype II (HII, A 203 A 752 ) (  (Table 4). The oxygen tension threshold for loss of equilibrium (LOE crit ) in blunt snout bream of DII was significantly (p < 0.05) lower than those in blunt snout bream with other genetic combinations, both at 15 °C and 25 °C (Table 4). At 15 °C, the LOE crit in blunt snout bream of DII was 0.64 mg·L −1 , while that of DI and DIII was 0.73 mg·L −1 and 0.71 mg·L −1 , respectively. Similarly, the LOE crit of DII was 0.71 mg·L −1 at 25 °C, while LOE crit of DI and DIII was 0.88 mg·L −1 and 0.89 mg·L −1 , respectively.

Enzyme activity, erythrocyte counts, and epo expression induced by hypoxia
To further examine if blunt snout bream of DI, DII, and DIII strains has different hypoxia tolerances, we measured the activity of antioxidant enzymes in the livers at 48 h of hypoxia. As shown in Fig. 4A,  Also, we measured the erythrocyte counts of DI, DII, and DIII strains after hypoxia (Fig. 4B). At 48 h of hypoxia, the erythrocyte counts of blunt snout bream were significantly (p < 0.05) higher than those at 0 h of hypoxia. And blunt snout bream of DII strain showed high erythrocyte count than those of DI and DIII strain types at 48 h of hypoxia and were significant (p < 0.05). The erythrocyte number was 1.97 × 10 12 L −1 for DII strain, compared with 1.76 × 10 12 L −1 for DI and 1.78 × 10 12 L −1 for DIII.
To explore the molecular mechanism of hypoxia tolerance in DII strain, the relative expression of epo (erythropoietin) mRNAs in the liver of blunt snout bream after hypoxia was analyzed by qRT-PCR (Fig. 4C). At 48 h of hypoxia, the expression of Fig. 4 Enzyme activity, erythrocyte counts, and epo expression induced by hypoxia. A Reduced glutathione (GSH), superoxide dismutase (SOD), and catalase (CAT) levels in the liver, B erythrocyte counts, and C Epo mRNA expression in the liver of blunt snout bream (diplotype I, diplotype II, and diplotype III strains) of hypoxia treatment at 15 °C. The O 2 concentration under hypoxia was kept at 2.0 mg·L. −1 . The results are given as mean ± SE for separate fish ( n = 3 ). Differences among groups were analyzed by unpaired t-tests. Columns marked with different letters are significantly different (p < 0.05). **p < 0.01 and *p < 0.05 epo mRNAs of blunt snout bream was significantly (p < 0.05) higher than that at 0 h of hypoxia. The relative expression of epo gene in the liver of the three strains of blunt snout bream was almost the same at 0 h of hypoxia. However, at 48 h of hypoxia, the relative expression level of epo gene in the liver of DII strain (20.79-fold) was significantly (p < 0.01) higher than that of DI and DIII strains (11.01-fold and 10.85-fold, respectively).

Discussion
The hypoxia signaling pathway is very conservative in the animal kingdom, so fish living in water can adapt to broader oxygen environments (Bickler and Buck 2007). The regulation of HIF is the main way to regulate hypoxia signaling transduction. In this study, we analyzed the structures of six HIFαs in blunt snout bream, consisting of bHLH, PAS-A, PAS-B, and PAC domains. The bHLH and PAS domains are well known to be involved in DNA binding and dimer formation with HIFβ (Liu et al. 2018;Bracken et al. 2003;Bruick 2003). These conserved structural residues in HIFαs imply that the functions may be conserved in oxygen-dependent degradation (Zhang et al. 2012). Phylogenetic tree analysis indicated that blunt snout bream HIFαs clustered well with teleost orthologs, which confirmed that six kinds of hifα genes in teleost would have evolved from a duplication event. And the difference in the secondary structure of a group of repeated HIFαs often means that their functions are different. After analyzing the structures of six HIFα, we found structural and functional polymorphisms in HIFα genes related to hypoxia adaptation in fish. After analyzing the structures of six HIFα genes, we identified structural and functional polymorphisms in the HIFα gene associated with fish hypoxia adaptation to maintain survival. Rytkonen et al. (2013) showed that the transcription of hif1αs decreased after hypoxic insult in adult zebrafish, but the transcription of hif2αa remained stable and hif2αb transcription increased significantly. Our results complement the relevant expression of hif3as. In the present study, the expression of hif1αs was decreased in the liver of juvenile blunt snout bream after hypoxia, but the expression of hif2αa remained stable in the liver after hypoxia and the expressions of hif2αb and hif3αs were both increased in the liver after hypoxia. These results indicated that all six hifαs could be induced by hypoxia, but they each had different functions and tissue differences. It suggested the uniqueness and importance of each hifα for the hypoxia stress in blunt snout bream.
The bHLH and PAS domains of HIFα proteins form dimer with HIFβ to recognize hypoxia response elements (HREs) on hypoxia-responsive genes and regulate transcription of downstream oxygensensitive genes, thereby regulating cell proliferation and apoptosis and other physiological processes, and help the organism cope with hypoxic environments (Xiao 2015;Yang et al. 2013). In this study, functional analysis results show that six HIFαs have significant HRE-dependent transcriptional activity. This means that although the structure and function of HIFαs changed during whole-genome replication, they still play roles in the hypoxia response of the body. HIFα is stable under hypoxic conditions, and HIFs bind to the HRE domain in the promoter regions of downstream genes to regulate expression of the downstream genes.
SNPs (single nucleotide polymorphisms) are DNA genetic polymorphisms caused by changes in individual nucleotides in the genome, which are widely distributed in the genome and have stable heritability, and are a very important molecular marker (Kim and Misra 2007;Zhang and Zhao 2000). In this study, two linkage-disequilibrium SNP sites at 203 and 752 within exons of the hif2αb gene with high genetic diversity (PIC > 0.5) were identified. The two SNP sites within exons of the hif2αb gene cause missense mutations in codons. Changes in amino acid sequence may lead to changes in protein function. Then, we further found that there are differences in LOE crit , enzyme activity, and gene expression among three types of blunt snout bream.
LOE crit was frequently used as an indicator of the hypoxia tolerance of fish species, and time to LOE crit was a standard proxy measure of hypoxia survival in fish, where lower values indicate a higher hypoxia tolerance performance (Chapman et al. 2002;He et al. 2015;Mandic et al. 2013). Therefore, in this study, the LOE crit value was used to compare the hypoxic tolerance of the three diplotypes of blunt snout bream. The diplotype II blunt snout bream has lower LOE crit values over a similar range of temperatures. These suggest that HII and its DII form may be correlated with the 248 Fish Physiol Biochem (2023) 49:239-251 hypoxia-tolerant performance of "Pujiang No. 2." In order to verify whether DII has the characteristics of hypoxia tolerance, we further measured the activity of antioxidant enzymes after hypoxia. The hypoxia-tolerant trait-related enzyme activities of GSH, SOD, and CAT in DII of blunt snout bream were significantly higher than those in blunt snout bream with other genetic combinations. In addition, we measured the erythrocyte counts. Higher erythrocyte count may enhance the blood oxygencarrying capacity in response to hypoxia (Wang et al. 2020;Wu et al. 2020). In this study, the DII strain exhibited significantly (p < 0.05) higher erythrocyte count than the DI and DIII strains after 48 h of hypoxia. This extremely high erythrocyte number diffuses enough oxygen into the blood.
In general, fish adapt to hypoxia by increasing the exposure area of the gill lamella or increasing the oxygen carrying capacity of red blood cells (Brauner and Rombough 2012;Phuong et al. 2017). However, previous research found that "Pujiang No. 2" mainly adapted to hypoxia by increasing the ability to transport oxygen in the blood (Wang et al. 2020;Wu et al. 2020). Interestingly, HIF2α proved to be closely related to angiogenesis and erythropoiesis (Takeda et al. 2004). Li et al. (2015) found that erythropoietin (epo) mRNA was significantly upregulated under hypoxia in blunt snout bream, which may result in an increase in erythrocyte count. We detected high expression of the epo gene in DII strain after hypoxia, compared to the other two strain. The results of these studies give us a hint (Fig. 5). Under hypoxia condition, HIFα exists stably and forms heterodimer with HIFβ. Then, HIFs and p300/CBP were successfully enriched, and HRE domains in the promoter region of downstream genes were identified. Finally, the downstream genes were activated and expressed. SNP mutations in DII hif2αb may increase the expression of downstream genes (such as increasing the expression of epo gene, thus increasing the number of red blood cells), thus increasing the hypoxia tolerance of blunt snout bream DII strain. Therefore, the hypoxia-tolerantrelated diplotype II of hif2αb has the potential to be used as a molecular marker in future genetic breeding of hypoxia-tolerant strain. In conclusion, we found six hifα genes in blunt snout bream and analyzed their clustering with homologs, amino acid domain, and secondary protein structure. Different expressions of six hifα genes after hypoxia were detected in vivo. In vitro, six hifαs were detected to have significant HRE-dependent transcriptional activity. More importantly, we found two linkage-disequilibrium SNP sites at exon 203 and 752 of hif2αb gene. Diplotype II was significantly correlated with hypoxia tolerance. Therefore, the hypoxia-tolerantrelated diplotype II of hif2αb has the potential to be used as a molecular marker in future genetic breeding of hypoxia-tolerant strain.